instruction
stringlengths
1.07k
7.56k
output
stringlengths
1.53k
7.69k
Please analyze the following clinical case and the related question: <clinical_case> An otherwise healthy Hispanic patient in their 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on their extremities and trunk. The Hispanic patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as their medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on their right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. </clinical_case> <question> An otherwise healthy man in his 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on his extremities and trunk. The patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as his medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on his right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. What Is Your Diagnosis? </question> <options> A. Epidermolysis bullosa pruriginosa B. Prurigo nodularis C. Pemphigoid nodularis D. Acquired perforating dermatosis </options>
C. Pemphigoid nodularis Histopathological examination revealed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (Figure, C). Results of DIF showed linear deposition of immunoglobulin G (IgG) (Figure, D) and C3 along the epidermal basement membrane zone (BMZ). Indirect immunofluorescence for anti–basement membrane antibody (IgG) was positive, with a titer of 1:256. Levels of autoantibodies for BP180 and BP230 detected via ELISA were elevated (116 and 101 [normal <9] U/mL, respectively). Based on these findings, we diagnosed pemphigoid nodularis (PN). The patient was treated with oral methylprednisolone, 28 mg/d (0.4 mg/kg per day), and topical halometasone cream twice a day. Eight weeks later, most of the papulonodular lesions had subsided, and no new lesions were observed. We gradually reduced oral methylprednisolone until a dosage of 7 mg/d (0.1 mg/kg per day) was reached. No recurrence was detected during a 1-year follow-up.Pemphigoid nodularis, first described by Yung et al1 in 1981, is a rare variant of bullous pemphigoid that combines the nodular prurigo–like skin lesions, pemphigoid blisters, and immunological characteristics of bullous pemphigoid. Pemphigoid nodularis predominantly affects older adult women, more than half of whom are of a non-White race or ethnicity.2 Pemphigoid nodularis initially manifests as multiple pruritic papulonodular lesions with or without blisters, symmetrically distributed in the extensor limbs and trunk. The face, groin, and popliteal fossae are usually not involved. Mucosal involvement has only been occasionally reported,2,3 a feature that should aid in excluding mucous membrane pemphigoid.3 Intensely itchy skin nodules cause the itch-scratch cycle, which might be responsible for disseminated skin lesions. The nodular prurigo–like lesions may precede the appearance of blisters by weeks or even months.2There are different viewpoints on the pathogenesis of PN. Some researchers have suggested that severe scratching and localized inflammation may expose cryptic epitopes within the BMZ, leading to epitope spreading and production of BP230 and/or BP180 autoantibodies.3,4 Previous research has shown that elevation of only BP180 autoantibodies is more common, and both elevations may occur.2 Others have found that drugs, including nifedipine, etanercept, psoralens, and psoralen–UV-A, may also trigger PN.3Differential diagnosis includes acquired perforating dermatosis, epidermolysis bullosa pruriginosa, and prurigo nodularis. Acquired perforating dermatosis is a group of papulonodular skin disorders featured with transepidermal excretion of dermal connective tissues. Pathologically, acquired perforating dermatosis typically presents as cup-shaped epidermal depressions with keratotic plugs and locally visible dermal connective tissues penetrating vertically through the epidermis.5 Epidermolysis bullosa pruriginosa is a rare variant of dystrophic epidermolysis bullosa presenting with trauma-induced blisters, linear plaques formed by fusion of prurigo-like nodules on the shins, nail dystrophy, and onset early in life.6 All the aforementioned features are obviously different from PN. Prurigo nodularis is a chronic inflammatory skin disease characterized by pruritus and nodular lesions.7 Considerable clinical and histological overlaps between prurigo nodularis and PN, as well as the absence of blisters throughout the course of the latter in some cases, make the diagnosis challenging. Clinically, PN can be easily misdiagnosed as prurigo nodularis. For patients initially diagnosed with prurigo nodularis, PN should be considered with the following situations: (1) severely itchy, generalized, and papulonodular lesions, and no characteristic “butterfly” sign on the back; (2) resistance to the conventional therapy regimen for prurigo nodularis, and the skin lesions continued to progress; and (3) no blisters and eosinophil infiltration on pathology, but positive findings of DIF (previous research on pathologic findings revealed that 35.1% had spongiosis/cleft and 62.2% had dermal eosinophils out of 37 cases2). Linear deposition of IgG and C3 along the BMZ under DIF, positive indirect immunofluorescence finding of anti–basement membrane antibody (IgG), and elevated BP180 and/or BP230 autoantibodies levels can provide definite clues for distinguishing these 2 entities.Treatment of PN is similar to that for bullous pemphigoid and remains challenging. Older age and neurological comorbidities as risk factors may increase the mortality of this disease.8 First-line therapy for bullous pemphigoid includes oral and topical corticosteroids. Based on the equivalent effect to oral corticosteroids, topical treatment with clobetasol propionate is considered as a valid choice for older adult patients with numerous internal disorders.9 Immunosuppressive agents are important choices as second-line treatments.9 A recent (2020) study showed that dupilumab may provide an additional supplement to the treatment of bullous pemphigoid and its variants.10
Please analyze the following clinical case and the related question: <clinical_case> An otherwise healthy Arab man in his 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on his extremities and trunk. The Arab patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as his medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on his right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. </clinical_case> <question> An otherwise healthy man in his 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on his extremities and trunk. The patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as his medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on his right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. What Is Your Diagnosis? </question> <options> A. Epidermolysis bullosa pruriginosa B. Prurigo nodularis C. Pemphigoid nodularis D. Acquired perforating dermatosis </options>
C. Pemphigoid nodularis Histopathological examination revealed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (Figure, C). Results of DIF showed linear deposition of immunoglobulin G (IgG) (Figure, D) and C3 along the epidermal basement membrane zone (BMZ). Indirect immunofluorescence for anti–basement membrane antibody (IgG) was positive, with a titer of 1:256. Levels of autoantibodies for BP180 and BP230 detected via ELISA were elevated (116 and 101 [normal <9] U/mL, respectively). Based on these findings, we diagnosed pemphigoid nodularis (PN). The patient was treated with oral methylprednisolone, 28 mg/d (0.4 mg/kg per day), and topical halometasone cream twice a day. Eight weeks later, most of the papulonodular lesions had subsided, and no new lesions were observed. We gradually reduced oral methylprednisolone until a dosage of 7 mg/d (0.1 mg/kg per day) was reached. No recurrence was detected during a 1-year follow-up.Pemphigoid nodularis, first described by Yung et al1 in 1981, is a rare variant of bullous pemphigoid that combines the nodular prurigo–like skin lesions, pemphigoid blisters, and immunological characteristics of bullous pemphigoid. Pemphigoid nodularis predominantly affects older adult women, more than half of whom are of a non-White race or ethnicity.2 Pemphigoid nodularis initially manifests as multiple pruritic papulonodular lesions with or without blisters, symmetrically distributed in the extensor limbs and trunk. The face, groin, and popliteal fossae are usually not involved. Mucosal involvement has only been occasionally reported,2,3 a feature that should aid in excluding mucous membrane pemphigoid.3 Intensely itchy skin nodules cause the itch-scratch cycle, which might be responsible for disseminated skin lesions. The nodular prurigo–like lesions may precede the appearance of blisters by weeks or even months.2There are different viewpoints on the pathogenesis of PN. Some researchers have suggested that severe scratching and localized inflammation may expose cryptic epitopes within the BMZ, leading to epitope spreading and production of BP230 and/or BP180 autoantibodies.3,4 Previous research has shown that elevation of only BP180 autoantibodies is more common, and both elevations may occur.2 Others have found that drugs, including nifedipine, etanercept, psoralens, and psoralen–UV-A, may also trigger PN.3Differential diagnosis includes acquired perforating dermatosis, epidermolysis bullosa pruriginosa, and prurigo nodularis. Acquired perforating dermatosis is a group of papulonodular skin disorders featured with transepidermal excretion of dermal connective tissues. Pathologically, acquired perforating dermatosis typically presents as cup-shaped epidermal depressions with keratotic plugs and locally visible dermal connective tissues penetrating vertically through the epidermis.5 Epidermolysis bullosa pruriginosa is a rare variant of dystrophic epidermolysis bullosa presenting with trauma-induced blisters, linear plaques formed by fusion of prurigo-like nodules on the shins, nail dystrophy, and onset early in life.6 All the aforementioned features are obviously different from PN. Prurigo nodularis is a chronic inflammatory skin disease characterized by pruritus and nodular lesions.7 Considerable clinical and histological overlaps between prurigo nodularis and PN, as well as the absence of blisters throughout the course of the latter in some cases, make the diagnosis challenging. Clinically, PN can be easily misdiagnosed as prurigo nodularis. For patients initially diagnosed with prurigo nodularis, PN should be considered with the following situations: (1) severely itchy, generalized, and papulonodular lesions, and no characteristic “butterfly” sign on the back; (2) resistance to the conventional therapy regimen for prurigo nodularis, and the skin lesions continued to progress; and (3) no blisters and eosinophil infiltration on pathology, but positive findings of DIF (previous research on pathologic findings revealed that 35.1% had spongiosis/cleft and 62.2% had dermal eosinophils out of 37 cases2). Linear deposition of IgG and C3 along the BMZ under DIF, positive indirect immunofluorescence finding of anti–basement membrane antibody (IgG), and elevated BP180 and/or BP230 autoantibodies levels can provide definite clues for distinguishing these 2 entities.Treatment of PN is similar to that for bullous pemphigoid and remains challenging. Older age and neurological comorbidities as risk factors may increase the mortality of this disease.8 First-line therapy for bullous pemphigoid includes oral and topical corticosteroids. Based on the equivalent effect to oral corticosteroids, topical treatment with clobetasol propionate is considered as a valid choice for older adult patients with numerous internal disorders.9 Immunosuppressive agents are important choices as second-line treatments.9 A recent (2020) study showed that dupilumab may provide an additional supplement to the treatment of bullous pemphigoid and its variants.10
Please analyze the following clinical case and the related question: <clinical_case> An otherwise healthy Arab woman in her 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on her extremities and trunk. The Arab woman had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as her medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on her right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. </clinical_case> <question> An otherwise healthy man in his 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on his extremities and trunk. The patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as his medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on his right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. What Is Your Diagnosis? </question> <options> A. Epidermolysis bullosa pruriginosa B. Prurigo nodularis C. Pemphigoid nodularis D. Acquired perforating dermatosis </options>
C. Pemphigoid nodularis Histopathological examination revealed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (Figure, C). Results of DIF showed linear deposition of immunoglobulin G (IgG) (Figure, D) and C3 along the epidermal basement membrane zone (BMZ). Indirect immunofluorescence for anti–basement membrane antibody (IgG) was positive, with a titer of 1:256. Levels of autoantibodies for BP180 and BP230 detected via ELISA were elevated (116 and 101 [normal <9] U/mL, respectively). Based on these findings, we diagnosed pemphigoid nodularis (PN). The patient was treated with oral methylprednisolone, 28 mg/d (0.4 mg/kg per day), and topical halometasone cream twice a day. Eight weeks later, most of the papulonodular lesions had subsided, and no new lesions were observed. We gradually reduced oral methylprednisolone until a dosage of 7 mg/d (0.1 mg/kg per day) was reached. No recurrence was detected during a 1-year follow-up.Pemphigoid nodularis, first described by Yung et al1 in 1981, is a rare variant of bullous pemphigoid that combines the nodular prurigo–like skin lesions, pemphigoid blisters, and immunological characteristics of bullous pemphigoid. Pemphigoid nodularis predominantly affects older adult women, more than half of whom are of a non-White race or ethnicity.2 Pemphigoid nodularis initially manifests as multiple pruritic papulonodular lesions with or without blisters, symmetrically distributed in the extensor limbs and trunk. The face, groin, and popliteal fossae are usually not involved. Mucosal involvement has only been occasionally reported,2,3 a feature that should aid in excluding mucous membrane pemphigoid.3 Intensely itchy skin nodules cause the itch-scratch cycle, which might be responsible for disseminated skin lesions. The nodular prurigo–like lesions may precede the appearance of blisters by weeks or even months.2There are different viewpoints on the pathogenesis of PN. Some researchers have suggested that severe scratching and localized inflammation may expose cryptic epitopes within the BMZ, leading to epitope spreading and production of BP230 and/or BP180 autoantibodies.3,4 Previous research has shown that elevation of only BP180 autoantibodies is more common, and both elevations may occur.2 Others have found that drugs, including nifedipine, etanercept, psoralens, and psoralen–UV-A, may also trigger PN.3Differential diagnosis includes acquired perforating dermatosis, epidermolysis bullosa pruriginosa, and prurigo nodularis. Acquired perforating dermatosis is a group of papulonodular skin disorders featured with transepidermal excretion of dermal connective tissues. Pathologically, acquired perforating dermatosis typically presents as cup-shaped epidermal depressions with keratotic plugs and locally visible dermal connective tissues penetrating vertically through the epidermis.5 Epidermolysis bullosa pruriginosa is a rare variant of dystrophic epidermolysis bullosa presenting with trauma-induced blisters, linear plaques formed by fusion of prurigo-like nodules on the shins, nail dystrophy, and onset early in life.6 All the aforementioned features are obviously different from PN. Prurigo nodularis is a chronic inflammatory skin disease characterized by pruritus and nodular lesions.7 Considerable clinical and histological overlaps between prurigo nodularis and PN, as well as the absence of blisters throughout the course of the latter in some cases, make the diagnosis challenging. Clinically, PN can be easily misdiagnosed as prurigo nodularis. For patients initially diagnosed with prurigo nodularis, PN should be considered with the following situations: (1) severely itchy, generalized, and papulonodular lesions, and no characteristic “butterfly” sign on the back; (2) resistance to the conventional therapy regimen for prurigo nodularis, and the skin lesions continued to progress; and (3) no blisters and eosinophil infiltration on pathology, but positive findings of DIF (previous research on pathologic findings revealed that 35.1% had spongiosis/cleft and 62.2% had dermal eosinophils out of 37 cases2). Linear deposition of IgG and C3 along the BMZ under DIF, positive indirect immunofluorescence finding of anti–basement membrane antibody (IgG), and elevated BP180 and/or BP230 autoantibodies levels can provide definite clues for distinguishing these 2 entities.Treatment of PN is similar to that for bullous pemphigoid and remains challenging. Older age and neurological comorbidities as risk factors may increase the mortality of this disease.8 First-line therapy for bullous pemphigoid includes oral and topical corticosteroids. Based on the equivalent effect to oral corticosteroids, topical treatment with clobetasol propionate is considered as a valid choice for older adult patients with numerous internal disorders.9 Immunosuppressive agents are important choices as second-line treatments.9 A recent (2020) study showed that dupilumab may provide an additional supplement to the treatment of bullous pemphigoid and its variants.10
Please analyze the following clinical case and the related question: <clinical_case> An otherwise healthy Arab patient in their 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on their extremities and trunk. The Arab patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as their medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on their right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. </clinical_case> <question> An otherwise healthy man in his 60s presented to our department with an 8-month history of widespread pruritic papulonodular skin lesions on his extremities and trunk. The patient had been initially diagnosed with nodular prurigo at a local hospital and treated with various topical steroids and oral antihistamines for 2 months, without significant improvement. The skin lesions continued to progress and worsened after scratching. Physical examinations revealed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the scalp, chest, abdomen, back, waist, and extremities (Figure, A and B). Subsequent examinations of the mucosal membranes appeared normal. Other physical examinations as well as his medical and family histories were unremarkable. Peripheral eosinophils, serum total immunoglobulin E level, enzyme-linked immunosorbent assay (ELISA) test for HIV, and results of other routine laboratory tests were within normal ranges or negative. A biopsy specimen was taken from one of the skin lesions on his right forearm for histopathological examination and direct immunofluorescence (DIF) testing (Figure, C and D).A and B, Clinical images showed disseminated erythematous papules, hypertrophic plaques approximately 1 to 2 cm in diameter, and dome-shaped nodules with excoriated whitish or pink centers on the back, waist, and extremities. C, A biopsy specimen taken from the right forearm showed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (hematoxylin-eosin). D, Direct immunofluorescence revealed linear deposition of immunoglobulin G along the basement membrane zone. What Is Your Diagnosis? </question> <options> A. Epidermolysis bullosa pruriginosa B. Prurigo nodularis C. Pemphigoid nodularis D. Acquired perforating dermatosis </options>
C. Pemphigoid nodularis Histopathological examination revealed marked hyperkeratosis, acanthosis with irregular elongation of the rete ridges, fibrotic dermis, and perivascular lymphocytic infiltrate (Figure, C). Results of DIF showed linear deposition of immunoglobulin G (IgG) (Figure, D) and C3 along the epidermal basement membrane zone (BMZ). Indirect immunofluorescence for anti–basement membrane antibody (IgG) was positive, with a titer of 1:256. Levels of autoantibodies for BP180 and BP230 detected via ELISA were elevated (116 and 101 [normal <9] U/mL, respectively). Based on these findings, we diagnosed pemphigoid nodularis (PN). The patient was treated with oral methylprednisolone, 28 mg/d (0.4 mg/kg per day), and topical halometasone cream twice a day. Eight weeks later, most of the papulonodular lesions had subsided, and no new lesions were observed. We gradually reduced oral methylprednisolone until a dosage of 7 mg/d (0.1 mg/kg per day) was reached. No recurrence was detected during a 1-year follow-up.Pemphigoid nodularis, first described by Yung et al1 in 1981, is a rare variant of bullous pemphigoid that combines the nodular prurigo–like skin lesions, pemphigoid blisters, and immunological characteristics of bullous pemphigoid. Pemphigoid nodularis predominantly affects older adult women, more than half of whom are of a non-White race or ethnicity.2 Pemphigoid nodularis initially manifests as multiple pruritic papulonodular lesions with or without blisters, symmetrically distributed in the extensor limbs and trunk. The face, groin, and popliteal fossae are usually not involved. Mucosal involvement has only been occasionally reported,2,3 a feature that should aid in excluding mucous membrane pemphigoid.3 Intensely itchy skin nodules cause the itch-scratch cycle, which might be responsible for disseminated skin lesions. The nodular prurigo–like lesions may precede the appearance of blisters by weeks or even months.2There are different viewpoints on the pathogenesis of PN. Some researchers have suggested that severe scratching and localized inflammation may expose cryptic epitopes within the BMZ, leading to epitope spreading and production of BP230 and/or BP180 autoantibodies.3,4 Previous research has shown that elevation of only BP180 autoantibodies is more common, and both elevations may occur.2 Others have found that drugs, including nifedipine, etanercept, psoralens, and psoralen–UV-A, may also trigger PN.3Differential diagnosis includes acquired perforating dermatosis, epidermolysis bullosa pruriginosa, and prurigo nodularis. Acquired perforating dermatosis is a group of papulonodular skin disorders featured with transepidermal excretion of dermal connective tissues. Pathologically, acquired perforating dermatosis typically presents as cup-shaped epidermal depressions with keratotic plugs and locally visible dermal connective tissues penetrating vertically through the epidermis.5 Epidermolysis bullosa pruriginosa is a rare variant of dystrophic epidermolysis bullosa presenting with trauma-induced blisters, linear plaques formed by fusion of prurigo-like nodules on the shins, nail dystrophy, and onset early in life.6 All the aforementioned features are obviously different from PN. Prurigo nodularis is a chronic inflammatory skin disease characterized by pruritus and nodular lesions.7 Considerable clinical and histological overlaps between prurigo nodularis and PN, as well as the absence of blisters throughout the course of the latter in some cases, make the diagnosis challenging. Clinically, PN can be easily misdiagnosed as prurigo nodularis. For patients initially diagnosed with prurigo nodularis, PN should be considered with the following situations: (1) severely itchy, generalized, and papulonodular lesions, and no characteristic “butterfly” sign on the back; (2) resistance to the conventional therapy regimen for prurigo nodularis, and the skin lesions continued to progress; and (3) no blisters and eosinophil infiltration on pathology, but positive findings of DIF (previous research on pathologic findings revealed that 35.1% had spongiosis/cleft and 62.2% had dermal eosinophils out of 37 cases2). Linear deposition of IgG and C3 along the BMZ under DIF, positive indirect immunofluorescence finding of anti–basement membrane antibody (IgG), and elevated BP180 and/or BP230 autoantibodies levels can provide definite clues for distinguishing these 2 entities.Treatment of PN is similar to that for bullous pemphigoid and remains challenging. Older age and neurological comorbidities as risk factors may increase the mortality of this disease.8 First-line therapy for bullous pemphigoid includes oral and topical corticosteroids. Based on the equivalent effect to oral corticosteroids, topical treatment with clobetasol propionate is considered as a valid choice for older adult patients with numerous internal disorders.9 Immunosuppressive agents are important choices as second-line treatments.9 A recent (2020) study showed that dupilumab may provide an additional supplement to the treatment of bullous pemphigoid and its variants.10
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old White girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old White boy presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old White child presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Black girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Black boy presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Black child presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Asian girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Asian boy presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Asian child presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Hispanic girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Hispanic boy presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Hispanic child presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Arab girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Arab boy presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 15-year-old Arab child presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. </clinical_case> <question> A 15-year-old girl presented to the otorhinolaryngologic department for a right-side postauricular mass. The mass had been present since birth but had slowly grown larger in the previous 2 years. On physical examination, a 2.5 × 1.5-cm, nontender, blue, soft, cystic mass was observed in the right postauricular area. There were no fistular openings around the auricle or in the external auditory canal (EAC), and there were no additional cysts, sinuses, or fistulas at the right neck from the hyoid bone and chin to the EAC. On temporal bone computed tomography, a 2.5 × 1.5-cm, hypodense, oval-shaped cystic mass was revealed at the subcutaneous fat layer superoposterior to the right auricle. On temporal magnetic resonance imaging (MRI), an oval cystic mass was revealed at the same location, with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (Figure, A-D). In surgery, the well-demarcated cystic mass was completely excised. Yellowish fluid leaked out of the cyst during excision. On histopathologic examination, the cyst was lined by keratinizing stratified squamous epithelium with hair follicles and sebaceous glands. The cyst contained keratinous debris (Figure, E). There was no recurrence 6 months after surgery.A-D, Magnetic resonance imaging showed a 2.5 × 1.5-cm, oval cystic mass at the subcutaneous fat layer superoposterior to the right auricle with high signal intensity on both T1-weighted and T2-weighted images and low signal intensity on enhanced T1-weighted fat suppression and T2-weighted short T1 inversion recovery images (in order from the left). E, Hematoxylin-eosin staining showed a cyst composed of keratinizing stratified squamous epithelium, hair follicles, and sebaceous glands. What Is Your Diagnosis? </question> <options> A. Hemangioma B. Type 1 first branchial cleft cyst C. Pilomatricoma D. Epidermoid cyst </options>
B. Type 1 first branchial cleft cyst Branchial cleft anomaly (BCA) is caused by abnormal embryological development of 1 of the 4 branchial arches.1 First BCA is rare and comprises 1% to 8% of all BCAs.2 Although typical imaging features of BCA have been reported, imaging findings can be atypical under specific pathologic conditions, including secondary infection, hemorrhage, or malignant transformation.1Duplication or failure of obliteration of the embryologic tract is thought to be the cause of BCA. For first BCA, sinuses and fistulae develop in infants and children, whereas cysts are more common in older ages.2 Branchial cleft anomaly commonly affects female patients and the left side of the head and neck.2,3 First BCA can present as either a cyst, sinus, or fistula in the Poncet triangle, which has its apex at the EAC and base between the chin and midpoint of the hyoid bone, with chronic or recurrent upper neck infections.3First BCA was divided into types 1 and 2 by the Work classification system in 1972.2-5 Type 1 anomaly is usually discovered in very young children as a thin-walled, soft mass located medially, inferiorly, or posteriorly to the pinna and superficial to the facial nerve.2,4 Type 2 anomaly is usually discovered at a later age as a fistulous opening within the Poncet triangle.4For BCA, MRI is a useful diagnostic modality to identify the extent of the lesion, the relationship with the facial nerve, and the risk of facial nerve damage for operative planning.3 Branchial cleft cyst reveals high signal on T2-weighted imaging.5,6 However, T1-weighted imaging can show diversity due to variation in proteinaceous contents,6 with high signal intensity with high protein content of mucus within the BCA.7 Atypical imaging features could be observed in the presence of infection, hemorrhage, or carcinomatous transformation, resulting in difficult diagnosis. Intracystic hemorrhage is usually caused by secondary infection or biopsy attempts.1 Cholesterol content of a branchial cleft cyst could be related to high signal on T1-weighted imaging.8 The cyst contents, such as cholesterol, could be related to its fatlike appearance on MRI.The treatment of first BCA is complete surgical excision to prevent recurrence.2-5 Definitive diagnosis is possible on histopathologic examination after surgical excision.1 Histopathologically, branchial cleft cysts are usually lined with squamous epithelium and may show concentrations of lymphoid tissue.8 Recurrence is dependent on preoperative procedures and infections and complete excision.3Pilomatricoma is an uncommon, benign congenital tumor originating from the hair follicles. Pilomatricoma occurs commonly in the head and neck regions, especially in the neck and preauricular area.5,9 It appears as a well-defined subcutaneous nodule with low-intermediate signal intensity on T1-weighted imaging and variable signal intensity on T2-weighted imaging.5 Computed tomography can reveal a sharply demarcated, calcified, subcutaneous nodule with mild to moderate enhancement.9Epidermoid cyst appears as a soft, nontender, cystic mass.9 It is revealed as a well-defined mass with high signal background and internal low signal on T2-weighted imaging and high signal foci on T1-weighted imaging with enhancement of a thin wall.10 Epidermoid cyst shows restricted diffusion on diffusion-weighted imaging.Hemangioma is a benign vascular soft tissue tumor and occurs most commonly in the head and neck of children.9 Hemangioma shows intermediate signal intensity on T1-weighted imaging and high signal intensity on T2-weighted imaging, with prominent internal flow voids and enhancement.5 In addition, half of hemangiomas regress spontaneously by 5 years of age.9First branchial cleft cyst could occur as an atypical appearance on physical examination and imaging findings mimicking those of other benign tumors. However, it should be considered in the differential diagnosis for postauricular cyst in a young pediatric patient. The present case mimicked lipoma on all MRI sequences. The only way to reach a definitive diagnosis was histopathologic confirmation. There was no recurrence 6 months after surgery.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old White woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The White patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the White patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old White man with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, he was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The White man underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. He was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the White man reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old White patient with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, they was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The White patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. They was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the White patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Black woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Black patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Black patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Black man with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, he was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Black man underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. He was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Black man reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Black patient with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, they was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Black patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. They was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Black patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Asian woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Asian patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Asian patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Asian man with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, he was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Asian man underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. He was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Asian man reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Asian patient with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, they was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Asian patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. They was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Asian patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Hispanic woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Hispanic patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Hispanic patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Hispanic man with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, he was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Hispanic man underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. He was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Hispanic man reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Hispanic patient with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, they was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Hispanic patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. They was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Hispanic patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Arab woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Arab patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Arab patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Arab man with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, he was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Arab man underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. He was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Arab man reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 66-year-old Arab patient with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, they was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The Arab patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. They was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the Arab patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). </clinical_case> <question> A 66-year-old woman with chronic hepatitis B infection and hepatocellular carcinoma (HCC) presented with subacute epigastric pain. Three years prior, she was diagnosed with a solitary HCC and underwent surgical resection with curative intent. Pathology revealed poorly differentiated HCC, 16 cm in largest dimension, with large vein involvement (T3bNXM0, R0 resection). The patient underwent expectant observation. One and a half years prior to the current presentation, laboratory results revealed an elevated α-fetoprotein (AFP), and cross-sectional imaging revealed multiple lung nodules consistent with metastases. She was treated with nivolumab every other week and attained a biochemical (baseline, 212.2 ng/dL; best response, 3.3 ng/dL) and radiographic complete response.In the current presentation, the patient reported nonradiating epigastric pain of moderate intensity for 1 month. Associated symptoms included nausea, anorexia, and water brash without emesis. Review of systems was positive for unchanged chronic abdominal bloating and negative for other constitutional or gastrointestinal symptoms. Laboratory results, including liver function tests, amylase, lipase, and AFP, were within normal limits, and hepatitis B virus polymerase chain reaction was undetectable. Computed tomography with multiphasic liver protocol showed new enlarged mesenteric nodes and diffuse thickening of the gastric body (Figure, A). Esophagogastroduodenoscopy (EGD) identified diffuse inflammation and ulcerations in the gastric body and antrum (Figure, B). Biopsy specimens showed chronic and active mixed inflammatory infiltrate, including intraepithelial lymphocytosis (Figure, C). No overt viral cytopathic effects were observed, and immunostaining results for Helicobacter pylori, cytomegalovirus, and adenovirus were negative.A, Coronal view of restaging computed tomography of chest, abdomen, and pelvis with contrast. The yellow circle notes borderline enlarged mesenteric nodes inferior to the stomach, and the yellow arrowheads show thickened gastric body wall. B, Gross appearance of the gastric mucosa visualized via upper endoscopy, which showed diffuse inflammation and ulceration in the gastric body and antrum. C, Biopsy of the gastric mucosa showed marked chronic and active intraepithelial infiltration of lymphocytes (blue arrowheads) and neutrophils (yellow arrowhead) and associated glandular epithelial injury with withered glands and gland dropout (red circle) (original magnification ×100). What Is Your Diagnosis? </question> <options> A. Immunotherapy-related gastritis B. Liver cancer progression C. Infectious gastritis D. Portal hypertensive gastropathy </options>
A. Immunotherapy-related gastritis This case highlights an increasingly common circumstance encountered in the management of patients with advanced liver cancer as well as other solid tumors—diagnostic considerations for an acute medical issue in a patient with a long-term response to ongoing immunotherapy.1,2 Here, the patient had been receiving treatment for 18 months without evidence of active liver cancer and presented with epigastric pain, a common clinical complaint with a broad differential diagnosis.3 The evaluation must consider the presence of chronic liver disease and metastatic HCC, as well as the potential complications of immunotherapy.As the median progression-free survival with metastatic HCC being treated with frontline immunotherapy is only 3.7 to 6.8 months, worsening abdominal pain warrants immediate assessment for progressive cancer.1,2 Normal AFP level and atypical imaging findings of gastric body thickening are inconsistent with HCC disease recurrence. Secondary malignant neoplasm (eg, gastric cancer) occurs in about 3% of patients with HCC and is a consideration given radiographic findings, though the biopsy specimen showed no malignant neoplasm.4Portal hypertensive gastropathy occurs due to cirrhosis or in the noncirrhotic liver affected with extensive cancer or portal vein tumor thrombus.5 Portal vein involvement by HCC occurs in up to 40% of patients with advanced disease and may lead to abnormal EGD findings.1 Portal hypertensive gastropathy rarely manifests as epigastric discomfort and has a characteristic fine reticular appearance of the gastric mucosa (snakeskin appearance) on EGD with pathologic findings of capillary ectasia and dilated mucosal and submucosal veins.5Here, pathology review showed lymphocytic gastritis, which can be associated with celiac disease, Crohn disease, infection, and drug effect. Although infectious gastritis (ie, H pylori and cytomegalovirus) is plausible, no pathogens were detected on special stains, and the radiographic findings were atypical for this entity.The diagnostic workup was most consistent with immunotherapy-related gastritis as a late complication of long-term anti–programmed cell death 1 (PD-1) therapy. Immune-related adverse events (irAEs) are now well documented and are an on-target toxic effect of PD-1/PD-L1 ligand 1 (PD-L1) and cytotoxic T-lymphocyte–associated antigen 4 (CTLA-4) pathway interference—reactivating T-cell–mediated antitumor activity with a cost of breaking immunological self-tolerance. Immune-related adverse events may affect multiple organs with varying severity and time to onset. In a meta-analysis of 7936 patients with advanced solid tumors, gastrointestinal toxic effects were the second most common irAEs observed in 10.7% (95% CI, 8.9%-13.0%) of patients treated with nivolumab, and in 33.6% (95% CI, 27.2%-40.6%) of patients receiving nivolumab plus ipilimumab.6 Only about 1% of patients treated with immune checkpoint inhibitor therapy have been reported to have immune-mediated gastritis, which presents with epigastric pain, anorexia, and emesis, with varied time to onset from weeks to up to 3 years.7-9 As in this case, EGD often shows fragile antral mucosa with erosions or ulcerations.8,9 Biopsy is characterized by severe inflammatory cell infiltration in the epithelium and lamina propria.Treatment of immune-mediated gastritis includes proton pump inhibitors and glucocorticoids, with options for refractory cases including use of infliximab, vedolizumab, or mycophenolate mofetil.7,9 As with other irAEs, retreatment with immune checkpoint inhibitor therapy may be possible; however, up to 26% may experience recurrence of the same irAE, while another 26% may develop a new irAE.10 As the treatment of lymphocytic gastritis varies considerably based on cause, it is critical for pathologic review and special stains to rule out alternative causes.The patient’s symptoms improved with cessation of anti–PD-1 therapy, a course of corticosteroids, sucralfate, and indefinite proton pump inhibitor therapy. Twenty months later, she developed a new solitary lung metastasis. She received 5 additional doses of nivolumab without recurrence of gastritis and underwent successful oligo-metastasectomy. At last follow-up it was 7 years since her initial HCC diagnosis, and she was alive without evidence of disease.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old White patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The White patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old White woman was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. She had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, she did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The White woman used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old White patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. They had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, they did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The White patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Black patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Black patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Black woman was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. She had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, she did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Black woman used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Black patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. They had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, they did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Black patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Asian patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Asian patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Asian woman was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. She had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, she did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Asian woman used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Asian patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. They had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, they did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Asian patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Hispanic patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Hispanic patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Hispanic woman was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. She had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, she did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Hispanic woman used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Hispanic patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. They had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, they did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Hispanic patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Arab patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Arab patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Arab woman was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. She had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, she did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Arab woman used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 65-year-old Arab patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. They had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, they did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The Arab patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. </clinical_case> <question> A 65-year-old patient was referred to the ear, nose, and throat department with left-sided hearing loss during the past 6 months without evident cause. He had a medical history of myocardial infarction, sleep apnea, Ménière disease, and congenital cataract. Furthermore, he did not report otalgia, otorrhea, a recent episode of Ménière disease, or other ear, nose, and throat or systemic symptoms. The patient used atorvastatin and acetylsalicylic acid, did not smoke, and reported drinking 2 glasses of wine daily. Otoscopy of the left ear revealed a yellow-whitish calcareous lobular mass located in the mesotympanum anterior to the malleus with an intact tympanic membrane. The lesion was hard and painful during palpation (Figure 1A). Otoscopy results for the right side were normal.Conductive hearing loss of the left ear because of a middle ear mass, seen on otoscopy (A) and on computed tomography (CT) (B).Pure-tone audiometry results showed a left-sided mixed hearing loss with a pure tone average of 110 dB HL and bone conduction levels at approximately 60 dB HL. The tympanogram showed a normal A-type curve with a normal to low middle ear compliance of 0.5 mm on the left side. Computed tomography (CT) imaging of the middle ear showed a heterogeneous hyperdense round mass of 5.5 mm in the left middle ear with a close relation to the malleus, eardrum, and tegmen (Figure 1B). There were no other abnormalities in the mastoid, ossicular chain, facial nerve canal, or surrounding blood vessels. What Is Your Diagnosis? </question> <options> A. Tympanosclerosis B. Tophaceous gout C. Cholesteatoma D. Osteoma </options>
B. Tophaceous gout Gout is an inflammatory joint disease characterized by depositions of monosodium urate crystals that are typically found in subcutaneous tissue or peripheral joints. Macroscopic depositions (ie, tophi) in the head and neck region are reported in the auricular helix, nasal bridge, larynx, and cricoarytenoid and the temporomandibular and sternoclavicular joints.1 Risk factors for these tophi include obesity, hypertension, alcohol use, a fructose-enriched diet, meat and fish consumption, and medication, including acetylsalicylates and diuretics.1-3Although conductive hearing loss due to a middle ear mass is common, a tophus in the middle ear is an uncommon cause of a middle ear mass.3-8 The low incidence combined with the absence of clinical manifestations of gout or hyperuricemia in all known cases, including in this patient, explains why tophaceous gout in the middle ear is often misidentified as osteoma, cholesteatoma, or tympanosclerosis.3,6,7 This emphasizes the importance of considering the diagnosis of a tophus in patients with conductive hearing loss and a mass in the middle ear.Diagnostic features first include a lobular and white aspect of the mass during otoscopy, which is caused by the accumulation of crystals in the tophus, whereas osteoma and tympanosclerosis are denser and more rounded.7 Second, a tophus often appears heterogeneously hyperdense (ie, semolina-like structure) on CT imaging, whereas an osteoma appears as a compact mass, and tympanosclerosis or cholesteatoma as a more homogenous mass.7,8 Finally, magnetic resonance imaging can be used as an additional instrument to differentiate between tophi or cholesteatoma, because the latter is T2 hyperintense.9 For this patient, diagnosis was made using otoscopy and confirmative CT.Surgical removal is performed in most cases, in which the ossicular chain remained intact in 5 of 6 cases.4-8 Furthermore, guidelines from the American College of Rheumatology strongly recommend for all patients with tophaceous gout the use of urate-lowering therapy (eg, allopurinol or febuxostat), with a serum urate target of less than 0.36 mmol/L or less than 6 mg/dL, and a 6-month use of an anti-inflammatory prophylaxis therapy (eg, colchicine).10In this case, the patient first started to make lifestyle adjustments but came back to undergo surgery because of progressive hearing loss. Using a postauricular, transmeatal approach, a brittle mass was removed from the tensor tympani, malleus, and incudostapedial joint (Figure 2). The chorda tympani remained intact, and the incus and stapes were reconnected using bone cement. Postoperatively, there was a complete closure of the air-bone gap. Histology showed crystals matching with the diagnosis of gout.Intraoperative lateral view of the left middle ear cavity via a postauricular, transmeatal approach containing a brittle 5-mm mass connected to the tensor tympani, malleus, and incudostapedial joint (probably originating from the joint).This case demonstrated the characteristic otoscopy and CT image of an uncommon cause for a common middle ear mass. Tophaceous gout can be diagnosed on otoscopy and should be considered in the differential diagnosis of middle ear masses.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old White woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old White man with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. He had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. He was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old White patient with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. They had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. They was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Black woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Black man with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. He had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. He was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Black patient with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. They had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. They was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Asian woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Asian man with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. He had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. He was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Asian patient with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. They had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. They was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Hispanic woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Hispanic man with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. He had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. He was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Hispanic patient with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. They had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. They was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Arab woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Arab man with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. He had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. He was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A 59-year-old Arab patient with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. They had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. They was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan </clinical_case> <question> A 59-year-old woman with a 5-year history of intermittent bilateral uveitis and scleritis presented with painful left uveitis and an episcleral nodule (Figure 1). Ultrasound biomicroscopy revealed an elevated and tender uveal-episcleral lesion. Visual acuity was 20/30 in the left eye and 20/25 in the right eye. She had been taking 60 mg of oral prednisone; topical prednisolone acetate, 1%; and oral ibuprofen for 10 days without relief. She was suspected of being a steroid responder due to elevated intraocular pressure bilaterally. Three years prior, sinus pressure symptoms led to magnetic resonance imaging that revealed a nasal sinus mass. Biopsy demonstrated a histiocytic infiltrate with emperipolesis, consistent with Rosai-Dorfman-Destombes (RDD) disease. Timolol-dorzolamide topical drops were initiated. Staging evaluations showed no other sites of RDD disease.Pretreatment. External photograph (A) and corresponding ultrasound biomicroscopy (B) demonstrating an elevated sclerouveal nodule.Initiate an alternative systemic therapy (ie, an immunomodulatory agent)Consider treating the left eye with intra-arterial melphalan What Would You Do Next? </question> <options> A. Initiate an alternative systemic therapy (ie, an immunomodulatory agent) B. Consider treating the left eye with intra-arterial melphalan C. Administer subtenon triamcinolone to the left eye D. Switch oral ibuprofen to oral flurbiprofen </options>
B. Consider treating the left eye with intra-arterial melphalan Uveitis and episcleral/uveal tract nodules can be manifestations of RDD disease, but RDD disease can be refractory to steroids, making an alternative treatment necessary. RDD disease is a subtype of histiocytosis, in which clonal expansion of activated histiocytes accumulate in tissues including the lymph nodes and possibly the eye. Given the myelogenous derivation of histiocytic disease, targeted delivery of melphalan (choice C) is emerging as a treatment option for isolated, symptomatic disease, particularly of ocular structures. It is standard care for retinoblastoma and has few adverse events.1 Following 3 monthly infusions of intra-arterial melphalan (9 mg) administered to the ophthalmic artery, the episcleral/uveal nodule regressed, the pain abated, oral steroids and all topical drops were discontinued, and the corrected visual acuity improved to 20/20 (Figure 2).Posttreatment. External photograph and corresponding ultrasound biomicroscopy (inset) demonstrating regression of the sclerouveal nodule. Arrowheads demonstrate the site of regressed disease.Histiocytoses are rare hematologic disorders characterized by pathological accumulation and infiltration of activated histiocytes in affected tissues, including ophthalmic structures. The non–Langerhans cell histiocytoses include RDD disease and are clinically heterogenous, including vision-threatening structures (such as the orbit, uveal tract, optic nerves, or the hypothalamic-pituitary-adrenal axis).2,3 This case demonstrated an episcleral nodule with sclerouveal infiltration seen with ultrasound biomicroscopy. Despite oral and topical steroids, she had no improvement. Switching to a different nonsteroidal anti-inflammatory drug (choice A), administering local steroids (choice D), or initiating an alternative immunomodulatory agent (choice B) would likely have provided no relief or only temporary therapeutic relief for this histiocytic-derived disease.4 Given the threat to her vision, an escalated treatment strategy was needed.Because histiocytic neoplasms are hematologic in their pathogenesis, we considered intra-arterial melphalan as a potential therapeutic option. Melphalan is an alkylating agent of mustard gas derivation used for blood-based neoplasms, notably multiple myeloma.5 Small case reports/series suggest histiocytic disease is sensitive to melphalan and similarly derived drugs.6 We have reported a small series of patients with neurologic or orbital histiocytosis treated with intra-arterial melphalan.7The advantage of intra-arterial chemotherapy is delivering a therapeutic dose to the target lesion while minimizing systemic administration. We are familiar with the safety and toxicity of intra-arterial melphalan administration to the eye because of our extensive experience in the treatment of retinoblastoma.1,8 Melphalan may be appealing for local administration because it is rapidly eliminated and has no neurotoxicity and limited ophthalmic toxicity with intra-arterial delivery.1 Melphalan is myelosuppressive in a dose-dependent manner, and particularly at systemic doses.This patient had no myelosuppression or other systemic toxic reactions following her 3 courses of 9-mg intra-arterial melphalan. Instead, she had a sustained ocular response, which allowed her to discontinue systemic steroids, oral nonsteroidal anti-inflammatory drugs, and all eye drops. She continued to have a sustained response at 10 months’ follow-up (Figure 2). Focal radiotherapy could have been a consideration; however, outcomes in RDD disease affecting neurologic and ophthalmic structures have been variable.4 Systemic therapies such as thalidomide, cladribine, and other chemotherapeutic/immunomodulatory agents would have been reasonable2; however, we attempted local therapy owing to her site-limited disease.In summary, we highlight that ocular histiocytosis may masquerade as refractory nodular scleritis. We describe implementation of intra-arterial chemotherapy for this vision-threatening intraocular histiocytosis with sclerouveal infiltration. This treatment approach might benefit from broader evaluation of its feasibility and safety across a variety of tumor sites and arterial targets, as well as its efficacy.
Please analyze the following clinical case and the related question: <clinical_case> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A White woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The White patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A White man in him late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The White man lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A White patient in them late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The White patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Black woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Black patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Black man in him late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Black man lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Black patient in them late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Black patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Asian woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Asian patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Asian man in him late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Asian man lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Asian patient in them late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Asian patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Hispanic patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic man in him late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Hispanic man lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic patient in them late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Hispanic patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Arab woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Arab patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Arab man in him late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Arab man lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A Arab patient in them late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The Arab patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. </clinical_case> <question> A woman in her late 70s with a history of immunoglobulin A monoclonal gammopathy of unknown significance presented with a tender, draining lesion of the central face despite several courses of treatment with antibiotics (Figure, A). The patient lived in Southern California but had visited Kenya for an extended trip a few years prior. Review of systems was negative for fevers, night sweats, weight loss, headache, vision changes, respiratory distress, or arthralgias. On the left nasal dorsum and medial cheek, there was a 3 × 4 cm violaceous plaque with focal purulent drainage. Within several weeks, the lesion rapidly expanded into a vegetative, freely draining plaque covering the cheeks and glabella and approaching the medial canthi (Figure, B). Computed tomography revealed findings consistent with abscess that did not involve the orbits, sinuses, or bone. Chest radiography results were unremarkable. Punch biopsy specimens were obtained for histopathological analysis and tissue cultures (Figure, C). Tissue culture results at 2 weeks remained negative, and a QuantiFERON-TB Gold test result was negative.A, Violaceous plaque on the nose and left cheek. B, Rapid progression of the lesion involved the cheeks and glabella, with substantially worsening drainage at 6-week follow-up. C, Punch biopsy of lesion showing suppurative and granulomatous dermatitis (hematoxylin-eosin). D, Posttreatment, there was improvement of the lesion from initial presentation, but substantial residual facial scarring remained. What Is Your Diagnosis? </question> <options> A. Pyoderma gangrenosum B. Lupus vulgaris C. Granulomatosis with polyangiitis D. Coccidioidomycosis </options>
A. Pyoderma gangrenosum Biopsies demonstrated suppurative and granulomatous dermatitis with an unremarkable epidermis; infectious stain results were negative. An extensive infectious workup, including final tissue cultures/polymerase chain reactions and serology, yielded negative results. Autoimmune markers, including antineutrophil cytoplasmic antibodies, were negative. Ophthalmologic examination was normal. A diagnosis of pyoderma gangrenosum (PG) was made. While the results of an infectious workup were pending, the patient did not respond to high-potency topical or intralesional corticosteroids. She received prednisone, 0.5 mg/kg per day, but developed new crusted scalp plaques and a dusky bulla on the left shin that progressed to an ulcer with undermined borders. A biopsy from the shin showed neutrophil-rich dermatitis with secondary vasculitic changes, further supporting the PG diagnosis. Increasing the dose of prednisone to 1 mg/kg per day arrested further growth and resolved the drainage, followed by a slow taper accomplished using steroid-sparing agents, including infliximab, colchicine, and tacrolimus ointment. At 8 months’ follow-up there was ongoing improvement but substantial residual cribriform scarring on the face (Figure, D).Head and neck PG is thought to be rare but may be underrecognized.1 Features of this case suggested superficial granulomatous PG (SGPG), an uncommon PG subtype that presents with vegetative or ulcerative lesions and may involve the face, particularly the cheeks.2 However, in contrast to the expected clinical course of SGPG, which displays indolent growth and sometimes spontaneous resolution,3 the present case exhibited rapid progression with impending ocular compromise and necessitated systemic immunosuppressive therapies to achieve remission. In addition, it lacked the classic histology of SGPG, the 3-layer granuloma (a central layer of neutrophils and necrosis surrounded by histiocytes and then an outer layer of plasma cells and eosinophils), although suppurative and granulomatous dermatitis is also characteristic of this entity.3,4 The initial vegetative/suppurative facial lesion also preceded the development of more typical bullous/ulcerative PG elsewhere. It is debated whether SGPG is a PG variant or a separate condition (superficial granulomatous pyoderma); this case seems to support the first position. A review of 170 cases of head/neck PG (only 8 of which were SGPGs) reported complications, including scarring/disfigurement (48%), local structure invasion (29%), and eye involvement (8%).1Pyoderma gangrenosum was considered a diagnosis of exclusion before validated diagnostic criteria5; however, these apply to ulcerative PG and are less useful for other variants. One helpful finding in this case was the presence of immunoglobulin A monoclonal gammopathy of unknown significance, a known PG comorbidity. Regardless, suppurative granulomas on histology require a search for infection before diagnosing PG and initiating systemic immunosuppression. This should include infectious stains and tissue cultures, because they may have differing sensitivities.6 This patient’s history of visiting a tuberculosis-endemic area raised concern for lupus vulgaris, a form of cutaneous tuberculosis affecting patients with intact cell-mediated immunity, that is characterized by red-brown papules that favor the face. Histology displays variable tuberculoid or sarcoidal granulomas, usually lacking suppuration or caseation necrosis.7 While it is uncommon to visualize acid-fast bacilli on histology, tissue cultures and/or polymerase chain reaction test results may be positive. Purified protein derivative skin test and interferon γ release assay are usually positive in lupus vulgaris.8Suppurative granulomas in a patient from California also raise concern for disseminated cutaneous coccidioidomycosis. Coccidioides immitis, a spore-forming fungus endemic to the Southwestern US, can produce suppurative plaques and nodules that favor the face. On histology, granulomatous inflammation is consistently seen; the associated inflammatory pattern varies from (most commonly) suppurative/granulomatous to lymphoplasmacytic, sarcoidal, neutrophilic, necrotizing, or eosinophilic.9 Organisms are visualized as endospore-filled spherules ranging 5 to 75 μm in diameter. The diagnosis is excluded with tissue, serologic, and imaging studies.Pyoderma gangrenosum and other neutrophilic dermatoses may demonstrate secondary vasculitic changes on histology. The coexistence of granulomas and vasculitis may indicate granulomatosis with polyangiitis, a small-/medium-vessel vasculitis associated with upper/lower respiratory and kidney complications. Atypical presentations of granulomatosis with polyangiitis featuring PG-/SGPG-like lesions have been reported, which are sometimes termed malignant pyoderma.10 In this case, the absence of systemic involvement and negative antineutrophil cytoplasmic antibodies ruled out the diagnosis. Superficial granulomatous PG is a challenging diagnosis requiring a comprehensive workup to rule out mimickers, particularly infections; however, this must be balanced with the need for prompt initiation of immunosuppression in more aggressive cases involving the face that can eventuate in permanent disfigurement.
Please analyze the following clinical case and the related question: <clinical_case> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A White man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the White patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A White woman in her mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the White woman showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A White patient in their mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the White patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Black man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Black patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Black woman in her mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Black woman showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Black patient in their mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Black patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Asian man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Asian patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Asian woman in her mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Asian woman showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Asian patient in their mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Asian patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Hispanic patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic woman in her mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Hispanic woman showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Hispanic patient in their mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Hispanic patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Arab man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Arab patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Arab woman in her mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Arab woman showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.
Please analyze the following clinical case and the related question: <clinical_case> A Arab patient in their mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the Arab patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone </clinical_case> <question> A man in his mid-50s with a history of mitral valve repair presented to the emergency department with complaints of recent-onset palpitation within the last hour. On physical examination, the patient showed a rhythmic tachycardia with a rate of 178 beats/min and blood pressure of 130/80 mm Hg. The 12-lead electrocardiogram (ECG) revealed a regular tachycardia with a rate of 178 beats/min; no obvious P waves were recognizable (Figure 1A). The QRS complexes displayed 2 different morphologies occurring with a definite allorhythmic distribution: pairs of beats exhibiting complete right bundle branch block (RBBB) and QRS duration of 140 milliseconds followed by a third beat with normal configuration and QRS duration of 70 milliseconds. This trigeminal variation in intraventricular conduction occurred without any change in cycle length and was maintained. Adenosine administration quickly slowed the ventricular rate allowing recognition of an atrial tachycardia with a rate of 178/min and a 2:1 atrioventricular (AV) ratio (Figure 1B).A, Regular supraventricular tachycardia with right bundle branch block aberration disappearing every third beat. B, Lead II. Numbers within the tracing indicate the R-R intervals (in milliseconds). The variability in ventricular cycle lengths results from small changes in both the P-P and P-R intervals. ECG indicates electrocardiogram.Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone What Would You Do Next? </question> <options> A. Order an urgent electrophysiologic investigation B. Perform urgent DC shock C. Place a temporary pacemaker and start treatment with low-molecular-weight heparin and amiodarone D. Start treatment with low-molecular-weight heparin and amiodarone </options>
D. Start treatment with low-molecular-weight heparin and amiodarone The atrial rate during adenosine administration was identical to the ventricular rate during tachycardia, implying that the arrhythmia in Figure 1A reflects an atrial tachycardia with 1:1 AV conduction. The medical history strongly suggests a macroreentrant circuit (macroreentrant atrial tachycardia, also called atypical atrial flutter) developing around the atriotomy scar related to mitral valve surgery, possibly a left-sided atrial flutter such as mitral annular flutter. The R-R interval prolongation due to the shift from 1:1 to 2:1 AV ratio restored the normal right bundle branch conduction resulting in narrow QRS complexes. This supports an intermittent phase 3 RBBB. The most challenging issue in the tracing on presentation is the periodic disappearance of the RBBB aberration. The key to the correct diagnosis is the unexpected normalization of the QRS morphology occurring periodically, every third beat, despite constant cycle lengths.1-4A pattern of complete bundle branch block does not necessarily imply a total interruption of conduction in the affected bundle branch but can also result from a delay exceeding the time required for activation from the contralateral bundle branch (accepted critical delay: 40-60 milliseconds).3 In a Wenckebach series, whenever the opening beat is normally conducted whereas complete bundle branch block occurs in the second impulse, progressive conduction delays in later beats and final failure do not further affect QRS morphology and width: the bundle branch Wenckebach phenomenon, therefore, is “incompletely concealed” being inapparent on ECG only from the second to the closing beat of each sequence.3,4 As illustrated schematically in Figure 2, while the conduction through the left bundle branch is normal, the right bundle branch shows increasingly impaired conduction owing to an area of refractoriness leading to progressive prolongation of the conduction time, until a block occurs in the third impulse of each Wenckebach series. Impulse 1 conducts down both bundle branches normally (conduction time: 40 milliseconds), resulting in narrow QRS. Impulse 2 penetrates the critical area within the right bundle branch with delayed conduction (right bundle conduction time: 90 milliseconds). Although a complete block in the right bundle branch does not occur, the delayed conduction results in ventricular depolarization via the left bundle branch, resulting in a complete RBBB pattern. Impulse 3 represents further impaired conduction and a true block in the right bundle branch. The proximal bidirectional block during the third impulse of each Wenckebach sequence doubles the effective cycle length of the right bundle branch, increasing its recovery time to 2 R-R intervals, thereby allowing normal conduction of impulse 4. This prolonged cycle length, furthermore, extends the right bundle refractory period following the normalized beat so that the second impulse of the next Wenckebach sequence once again conducts with RBBB aberration, and the mechanism then repeats.1,2 Accordingly, the Wenckebach phenomenon in the right bundle branch is incompletely concealed and is evident only on the first to second beats.3,4 Incompletely concealed Wenckebach-type block in a bundle branch was first postulated by Rosenbaum in 1969 in order to explain periodical improvements of aberrant conduction unrelated to changes in cycle length.3Schematic explanation of the behavior of intraventricular conduction. Numbers above the strip are the R-R intervals. An area of refractoriness within the right bundle results in progressively impaired conduction (red: normal conduction; gray: delayed conduction; black: conduction failure). Concealed 3:2 Wenckebach-type conduction occurs over the right bundle branch. CT indicates conduction time (in milliseconds); LB, left bundle, RB, right bundle.Alternate explanations are less convincing. Disappearance of the RBBB pattern every third beat theoretically could reflect fusion of supraventricular impulses with trigeminal right ventricular extrasystoles. However, a regular repetition of such ectopics in a fixed trigeminal rhythm with perfect synchronization with atrial impulses is unlikely; furthermore, no ventricular extrasystoles were evidenced throughout the tracing.1,2 The findings in this ECG are compatible with a 3:1 RBBB. However, such a high conduction ratio in a bundle branch is extremely rare especially in a repetitive fashion. Moreover, in order to invoke a 3:1 RBBB, a very long right bundle branch refractory period, extending over 2 cardiac cycles (>675 milliseconds), has to be assumed. Nevertheless, as the rate slows with adenosine and normal right bundle conduction resumes, some R-R intervals measuring 630 to 670 milliseconds occur (Figure 1B), which excludes this possibility.3,4Although the conduction disturbance may have been the result of mitral valve surgery, a Wenckebach-type bundle branch block occurring at fast rates does not necessarily indicate a damaged conduction system, but may represent merely a functional, phase 3 delay in the distal His-Purkinje system.1 Amiodarone infusion restored sinus rhythm, leading to persistent normalization of intraventricular conduction. Given the very recent onset of tachycardia and the low risk of stroke, postcardioversion anticoagulation was omitted.5,6 No recurrence of atrial flutter and/or RBBB was documented over a 1-year follow-up while receiving oral amiodarone.