Case Letter

To the Editor:

Cutaneous manifestations of leukemia can be defined as specific or nonspecific skin lesions. Specific skin lesions include papules, nodules, tumors, or ulcerating plaques caused by the infiltration of leukemic cells (eg, leukemia cutis), while nonspecific skin lesions include petechiae, purpura, ecchymosis, or infection. Leukemia cutis is a rare entity that denotes a poor prognosis. We present a case of leukemia cutis with scrotal ulcerations, which heralded the onset of acute myelogenous leukemia (AML). Other causes of painful cutaneous ulcers to consider in a patient with a hematologic malignancy include a variety of infectious and inflammatory etiologies.

A 63-year-old man with long-standing essential thrombocytosis was admitted to the hospital with increasing fatigue and fever. A complete blood cell count revealed anemia, thrombocytopenia, neutropenia, and leukocytosis with 49% blasts. He was immediately placed on broad-spectrum antifungal, antiviral, and antibiotic agents after blood and urine cultures were obtained and while subsequent diagnostic tests were being performed. The patient developed antibiotic-associated Clostridium difficile diarrhea after several days in the hospital. The nursing staff noted “genital irritation from feces” while cleaning the patient and reported the findings to the primary team. Dermatology was consulted for treatment recommendations and wound care.

A full-body skin examination revealed discrete and coalescing painful scrotal and perineal ulcers with fecal contamination (Figure 1), along with numerous petechial macules on the anteromedial thighs. The patient was unsure of the onset of the ulcers, but he claimed to have scrotal and perineal pain prior to admission, which suggested their long-standing presence. No crepitus or rapid contiguous spread was noted on serial physical examination over the next 12 to 24 hours. Punch biopsies for tissue culture and histopathologic examination were obtained from the scrotum.

Figure 1. Discrete and coalescing painful scrotal and perineal ulcers with fecal contamination.

Subsequent diagnostic studies, including a bone marrow biopsy touch preparation that showed 44.7% blasts, along with cytogenetics analysis yielded the diagnosis of AML with an unfavorable karyotype: 5q deletion, 17p deletion, and monosomy 16. Induction chemotherapy with multiple cycles of a combination of cytarabine and daunorubicin was initiated, albeit with an unfavorable response. Blood, urine, bone marrow, and stool cultures remained negative. Serologic testing for herpes simplex virus (HSV), cytomegalovirus, and Epstein-Barr virus was negative.

Histopathology showed an ulcerated epidermis along with a dense dermal perivascular and interstitial infiltrate of myeloid blasts (Figure 2). Lesional cells were diffusely CD68⁺ (Figure 3), myeloperoxidase positive (Figure 4), and CD34^-, correlating with findings on the bone marrow biopsy. Gomori methenamine-silver, periodic acid–Schiff, and Gram stains highlighted several bacterial cocci, yeast forms, and pseudohyphae on the ulcer surface with no organisms seen in the dermis. Tissue culture did not yield any organisms. There was no overt evidence of HSV infection. The histopathologic features were consistent with a diagnosis of leukemia cutis in the setting of AML. After failed chemotherapy and a complicated hospital course with a poor long-term prognosis, the patient and his family opted for home hospice care.

Figure 2. Dense dermal infiltrate of myeloid blasts (H&E, original magnification ×10).

Figure 3. CD68 stain highlighted lesional cells (original magnification ×20).

Cases of leukemia cutis manifesting as scrotal ulcerations are rare.¹ Typically, most dermatologic lesions seen in leukemia patients are of a nonleukemic origin.² Leukemia cutis is a rare entity that is associated with a poor prognosis.³ It was accordingly an ominous finding in our patient, as his AML had an unfavorable karyotype and was unresponsive to chemotherapy.

Figure 4. Myeloperoxidase stain highlighted lesional cells (original magnification ×10).

In hematologic malignancies, the morphology and distribution of specific cutaneous lesions tends to vary by the type of leukemia. For example, chronic lymphocytic leukemia presents with papules, nodules, and ulcerating plaques more often than acute lymphocytic leukemia.⁴ Acute myelocytic papules, nodules, or tumors tend to appear on the face, while chronic lesions commonly are seen on the face and extremities.⁵ Generally, for all forms of leukemia, lesions on the genitalia are uncommon.

This case of ulcerative scrotal AML lesions is unique, as myelogenous leukemia cutis typically appears as a tumor that infrequently ulcerates and rarely is located on the genitalia.⁶ Tissue culture, blood culture, viral serology, and biopsy failed to reveal an infectious etiology of the genital ulcerations in our patient. Therefore, the leukemic infiltrate likely preceded the ulceration and an antecedent infection did not elicit an inflammatory response, resulting in the attraction of a leukemic infiltrate to the area. This latter phenomenon also has been demonstrated in the setting of recent surgical scars, trauma, burns, herpes zoster, HSV, and intramuscular injection sites.^1,7

The differential diagnosis of painful cutaneous ulcers in the setting of hematologic malignancy includes a variety of infectious and inflammatory etiologies such as ulcerative HSV, cytomegalovirus, Epstein-Barr virus, amoebic skin ulcerations, opportunistic fungal infections, bacterial infections, pyoderma gangrenosum, erythema nodosum, acute febrile neutrophilic dermatosis (Sweet syndrome), histoplasmosis,¹ Fournier gangrene, and vasculitis, all of which were ruled out in our case.

We report a case of a patient with a history of essential thrombocytosis who developed leukemia cutis presenting as scrotal ulcerations, which preceded the diagnosis of AML. Ulcerative leukemic lesions on the genitalia are atypical and extremely rare. However, the lesions are notable because they may be suggestive of a poor prognostic outcome, as in our patient whose induction therapy with cytarabine and daunorubicin failed. We propose that leukemia cutis be included in the differential diagnosis of chronic skin ulcerations of the genitalia.

To the Editor:

Cutaneous manifestations of leukemia can be defined as specific or nonspecific skin lesions. Specific skin lesions include papules, nodules, tumors, or ulcerating plaques caused by the infiltration of leukemic cells (eg, leukemia cutis), while nonspecific skin lesions include petechiae, purpura, ecchymosis, or infection. Leukemia cutis is a rare entity that denotes a poor prognosis. We present a case of leukemia cutis with scrotal ulcerations, which heralded the onset of acute myelogenous leukemia (AML). Other causes of painful cutaneous ulcers to consider in a patient with a hematologic malignancy include a variety of infectious and inflammatory etiologies.

A 63-year-old man with long-standing essential thrombocytosis was admitted to the hospital with increasing fatigue and fever. A complete blood cell count revealed anemia, thrombocytopenia, neutropenia, and leukocytosis with 49% blasts. He was immediately placed on broad-spectrum antifungal, antiviral, and antibiotic agents after blood and urine cultures were obtained and while subsequent diagnostic tests were being performed. The patient developed antibiotic-associated Clostridium difficile diarrhea after several days in the hospital. The nursing staff noted “genital irritation from feces” while cleaning the patient and reported the findings to the primary team. Dermatology was consulted for treatment recommendations and wound care.

A full-body skin examination revealed discrete and coalescing painful scrotal and perineal ulcers with fecal contamination (Figure 1), along with numerous petechial macules on the anteromedial thighs. The patient was unsure of the onset of the ulcers, but he claimed to have scrotal and perineal pain prior to admission, which suggested their long-standing presence. No crepitus or rapid contiguous spread was noted on serial physical examination over the next 12 to 24 hours. Punch biopsies for tissue culture and histopathologic examination were obtained from the scrotum.

Figure 1. Discrete and coalescing painful scrotal and perineal ulcers with fecal contamination.

Subsequent diagnostic studies, including a bone marrow biopsy touch preparation that showed 44.7% blasts, along with cytogenetics analysis yielded the diagnosis of AML with an unfavorable karyotype: 5q deletion, 17p deletion, and monosomy 16. Induction chemotherapy with multiple cycles of a combination of cytarabine and daunorubicin was initiated, albeit with an unfavorable response. Blood, urine, bone marrow, and stool cultures remained negative. Serologic testing for herpes simplex virus (HSV), cytomegalovirus, and Epstein-Barr virus was negative.

Histopathology showed an ulcerated epidermis along with a dense dermal perivascular and interstitial infiltrate of myeloid blasts (Figure 2). Lesional cells were diffusely CD68⁺ (Figure 3), myeloperoxidase positive (Figure 4), and CD34^-, correlating with findings on the bone marrow biopsy. Gomori methenamine-silver, periodic acid–Schiff, and Gram stains highlighted several bacterial cocci, yeast forms, and pseudohyphae on the ulcer surface with no organisms seen in the dermis. Tissue culture did not yield any organisms. There was no overt evidence of HSV infection. The histopathologic features were consistent with a diagnosis of leukemia cutis in the setting of AML. After failed chemotherapy and a complicated hospital course with a poor long-term prognosis, the patient and his family opted for home hospice care.

Figure 2. Dense dermal infiltrate of myeloid blasts (H&E, original magnification ×10).

Figure 3. CD68 stain highlighted lesional cells (original magnification ×20).

Cases of leukemia cutis manifesting as scrotal ulcerations are rare.¹ Typically, most dermatologic lesions seen in leukemia patients are of a nonleukemic origin.² Leukemia cutis is a rare entity that is associated with a poor prognosis.³ It was accordingly an ominous finding in our patient, as his AML had an unfavorable karyotype and was unresponsive to chemotherapy.

Figure 4. Myeloperoxidase stain highlighted lesional cells (original magnification ×10).

In hematologic malignancies, the morphology and distribution of specific cutaneous lesions tends to vary by the type of leukemia. For example, chronic lymphocytic leukemia presents with papules, nodules, and ulcerating plaques more often than acute lymphocytic leukemia.⁴ Acute myelocytic papules, nodules, or tumors tend to appear on the face, while chronic lesions commonly are seen on the face and extremities.⁵ Generally, for all forms of leukemia, lesions on the genitalia are uncommon.

This case of ulcerative scrotal AML lesions is unique, as myelogenous leukemia cutis typically appears as a tumor that infrequently ulcerates and rarely is located on the genitalia.⁶ Tissue culture, blood culture, viral serology, and biopsy failed to reveal an infectious etiology of the genital ulcerations in our patient. Therefore, the leukemic infiltrate likely preceded the ulceration and an antecedent infection did not elicit an inflammatory response, resulting in the attraction of a leukemic infiltrate to the area. This latter phenomenon also has been demonstrated in the setting of recent surgical scars, trauma, burns, herpes zoster, HSV, and intramuscular injection sites.^1,7

The differential diagnosis of painful cutaneous ulcers in the setting of hematologic malignancy includes a variety of infectious and inflammatory etiologies such as ulcerative HSV, cytomegalovirus, Epstein-Barr virus, amoebic skin ulcerations, opportunistic fungal infections, bacterial infections, pyoderma gangrenosum, erythema nodosum, acute febrile neutrophilic dermatosis (Sweet syndrome), histoplasmosis,¹ Fournier gangrene, and vasculitis, all of which were ruled out in our case.

We report a case of a patient with a history of essential thrombocytosis who developed leukemia cutis presenting as scrotal ulcerations, which preceded the diagnosis of AML. Ulcerative leukemic lesions on the genitalia are atypical and extremely rare. However, the lesions are notable because they may be suggestive of a poor prognostic outcome, as in our patient whose induction therapy with cytarabine and daunorubicin failed. We propose that leukemia cutis be included in the differential diagnosis of chronic skin ulcerations of the genitalia.

To the Editor:

Cutaneous manifestations of leukemia can be defined as specific or nonspecific skin lesions. Specific skin lesions include papules, nodules, tumors, or ulcerating plaques caused by the infiltration of leukemic cells (eg, leukemia cutis), while nonspecific skin lesions include petechiae, purpura, ecchymosis, or infection. Leukemia cutis is a rare entity that denotes a poor prognosis. We present a case of leukemia cutis with scrotal ulcerations, which heralded the onset of acute myelogenous leukemia (AML). Other causes of painful cutaneous ulcers to consider in a patient with a hematologic malignancy include a variety of infectious and inflammatory etiologies.

A 63-year-old man with long-standing essential thrombocytosis was admitted to the hospital with increasing fatigue and fever. A complete blood cell count revealed anemia, thrombocytopenia, neutropenia, and leukocytosis with 49% blasts. He was immediately placed on broad-spectrum antifungal, antiviral, and antibiotic agents after blood and urine cultures were obtained and while subsequent diagnostic tests were being performed. The patient developed antibiotic-associated Clostridium difficile diarrhea after several days in the hospital. The nursing staff noted “genital irritation from feces” while cleaning the patient and reported the findings to the primary team. Dermatology was consulted for treatment recommendations and wound care.

A full-body skin examination revealed discrete and coalescing painful scrotal and perineal ulcers with fecal contamination (Figure 1), along with numerous petechial macules on the anteromedial thighs. The patient was unsure of the onset of the ulcers, but he claimed to have scrotal and perineal pain prior to admission, which suggested their long-standing presence. No crepitus or rapid contiguous spread was noted on serial physical examination over the next 12 to 24 hours. Punch biopsies for tissue culture and histopathologic examination were obtained from the scrotum.

Figure 1. Discrete and coalescing painful scrotal and perineal ulcers with fecal contamination.

Subsequent diagnostic studies, including a bone marrow biopsy touch preparation that showed 44.7% blasts, along with cytogenetics analysis yielded the diagnosis of AML with an unfavorable karyotype: 5q deletion, 17p deletion, and monosomy 16. Induction chemotherapy with multiple cycles of a combination of cytarabine and daunorubicin was initiated, albeit with an unfavorable response. Blood, urine, bone marrow, and stool cultures remained negative. Serologic testing for herpes simplex virus (HSV), cytomegalovirus, and Epstein-Barr virus was negative.

Histopathology showed an ulcerated epidermis along with a dense dermal perivascular and interstitial infiltrate of myeloid blasts (Figure 2). Lesional cells were diffusely CD68⁺ (Figure 3), myeloperoxidase positive (Figure 4), and CD34^-, correlating with findings on the bone marrow biopsy. Gomori methenamine-silver, periodic acid–Schiff, and Gram stains highlighted several bacterial cocci, yeast forms, and pseudohyphae on the ulcer surface with no organisms seen in the dermis. Tissue culture did not yield any organisms. There was no overt evidence of HSV infection. The histopathologic features were consistent with a diagnosis of leukemia cutis in the setting of AML. After failed chemotherapy and a complicated hospital course with a poor long-term prognosis, the patient and his family opted for home hospice care.

Figure 2. Dense dermal infiltrate of myeloid blasts (H&E, original magnification ×10).

Figure 3. CD68 stain highlighted lesional cells (original magnification ×20).

Cases of leukemia cutis manifesting as scrotal ulcerations are rare.¹ Typically, most dermatologic lesions seen in leukemia patients are of a nonleukemic origin.² Leukemia cutis is a rare entity that is associated with a poor prognosis.³ It was accordingly an ominous finding in our patient, as his AML had an unfavorable karyotype and was unresponsive to chemotherapy.

Figure 4. Myeloperoxidase stain highlighted lesional cells (original magnification ×10).

In hematologic malignancies, the morphology and distribution of specific cutaneous lesions tends to vary by the type of leukemia. For example, chronic lymphocytic leukemia presents with papules, nodules, and ulcerating plaques more often than acute lymphocytic leukemia.⁴ Acute myelocytic papules, nodules, or tumors tend to appear on the face, while chronic lesions commonly are seen on the face and extremities.⁵ Generally, for all forms of leukemia, lesions on the genitalia are uncommon.

This case of ulcerative scrotal AML lesions is unique, as myelogenous leukemia cutis typically appears as a tumor that infrequently ulcerates and rarely is located on the genitalia.⁶ Tissue culture, blood culture, viral serology, and biopsy failed to reveal an infectious etiology of the genital ulcerations in our patient. Therefore, the leukemic infiltrate likely preceded the ulceration and an antecedent infection did not elicit an inflammatory response, resulting in the attraction of a leukemic infiltrate to the area. This latter phenomenon also has been demonstrated in the setting of recent surgical scars, trauma, burns, herpes zoster, HSV, and intramuscular injection sites.^1,7

The differential diagnosis of painful cutaneous ulcers in the setting of hematologic malignancy includes a variety of infectious and inflammatory etiologies such as ulcerative HSV, cytomegalovirus, Epstein-Barr virus, amoebic skin ulcerations, opportunistic fungal infections, bacterial infections, pyoderma gangrenosum, erythema nodosum, acute febrile neutrophilic dermatosis (Sweet syndrome), histoplasmosis,¹ Fournier gangrene, and vasculitis, all of which were ruled out in our case.

We report a case of a patient with a history of essential thrombocytosis who developed leukemia cutis presenting as scrotal ulcerations, which preceded the diagnosis of AML. Ulcerative leukemic lesions on the genitalia are atypical and extremely rare. However, the lesions are notable because they may be suggestive of a poor prognostic outcome, as in our patient whose induction therapy with cytarabine and daunorubicin failed. We propose that leukemia cutis be included in the differential diagnosis of chronic skin ulcerations of the genitalia.

To the Editor:

Remission of psoriasis after bone marrow transplant has been reported¹; however, follow-up typically has been limited. We describe a case of a remote recurrence of psoriasis 13 years after autologous stem cell transplant.

In September 1997, a 48-year-old woman with a 20-year history of moderate to severe plaque psoriasis, previously managed with oral methotrexate (up to 30–40 mg weekly), presented to her primary care physician with a persistent cough and newly pervasive arthralgia and myalgia. A complete blood cell count showed a white blood cell count of 6.9 mg/dL (reference range, 3.8–10.8 mg/dL), hemoglobin level of 11.6 mg/dL (reference range, 11.7–15.5 mg/dL), and hematocrit level of 34% (reference range, 35.0%–45.0%). Bone survey demonstrated a 4×2-cm lytic expansile lesion in the occipital bone, and a bone marrow aspirate and biopsy revealed 40% plasma cells. She was subsequently diagnosed with stage IIIA IgA λ and free λ light chain multiple myeloma with an IgA level of 62.1 g/L (reference range, 1.24–2.17 g/L).

Therapy was initiated with dexamethasone in combination with pamidronate and she was transitioned to cyclophosphamide chemotherapy. A peripheral autologous blood stem cell transplant followed. Throughout her initial treatment, the psoriasis continued to flare and she remained on methotrexate to help maintain control of the skin disease.

In November 1998, the patient underwent an autologous CD34⁺ stem cell transplant. There were no complications following the transplant and the multiple myeloma has remained in complete remission to date. Following the stem cell transplant, the patient remained free of psoriatic disease for just over 13 years without any intervention. Recurrent erythematous plaques with thick scale on the bilateral lower legs prompted her return to dermatology in August 2011. To date, she continues to note intermittent though mild flares of her psoriasis. She has responded well to methotrexate 7.5 mg weekly and clobetasol propionate ointment 0.05% as needed with satisfactory control of her disease.

The development of autoimmune disease following an allogenic bone marrow transplant such as psoriasis, thyroiditis, myasthenia gravis, insulin-dependent diabetes mellitus, and vitiligo has been well described.^2-4 The causal relationship between allogenic hematopoetic stem cell transplant (HSCT) and subsequent autoimmune disease has been attributed to T cells and passive transfer of autoimmune disease from the donor to the recipient.¹

In the case of autologous HSCT, the remission of autoimmune disease thereafter has been attributed to myeloablative conditioning, whereby high-dose chemotherapy eliminates self-reactive lymphocytes, resulting in ablation of peripheral reactive and alloreactive T cells with depletion of thymus and reactive B cells.¹

In 1997, Cooley et al⁵ described 4 patients with autoimmune disease who experienced complete but temporary (up to 26 months) remission following autologous bone marrow or stem cell transplant. A comprehensive review of psoriasis cases that resolved after HSCT published by Kaffenberger et al⁶ in 2013 cited 6 cases of autologous and 13 cases of allogenic transplant that resulted in resolution of psoriatic disease. Recurrence of disease was witnessed in 3 of 13 allogenic HSCT patients and 5 of 6 autologous HSCT patients. Of note, all 5 of these patients developed recurrent disease within 24 months following autologous HSCT, albeit with more benign disease.⁶

Similar to those cases presented by Kaffenberger et al,⁶ our patient’s psoriatic disease was notably mild compared to her disease prior to HSCT. Although an unintended but welcome outcome of multiple myeloma treatment, the nearly quiescent nature of her skin disease following autologous HSCT has led to a dramatic improvement in her quality of life.

We highlight a unique case in which psoriatic skin disease recurred more than 1 decade after autologous HSCT.

To the Editor:

Remission of psoriasis after bone marrow transplant has been reported¹; however, follow-up typically has been limited. We describe a case of a remote recurrence of psoriasis 13 years after autologous stem cell transplant.

In September 1997, a 48-year-old woman with a 20-year history of moderate to severe plaque psoriasis, previously managed with oral methotrexate (up to 30–40 mg weekly), presented to her primary care physician with a persistent cough and newly pervasive arthralgia and myalgia. A complete blood cell count showed a white blood cell count of 6.9 mg/dL (reference range, 3.8–10.8 mg/dL), hemoglobin level of 11.6 mg/dL (reference range, 11.7–15.5 mg/dL), and hematocrit level of 34% (reference range, 35.0%–45.0%). Bone survey demonstrated a 4×2-cm lytic expansile lesion in the occipital bone, and a bone marrow aspirate and biopsy revealed 40% plasma cells. She was subsequently diagnosed with stage IIIA IgA λ and free λ light chain multiple myeloma with an IgA level of 62.1 g/L (reference range, 1.24–2.17 g/L).

Therapy was initiated with dexamethasone in combination with pamidronate and she was transitioned to cyclophosphamide chemotherapy. A peripheral autologous blood stem cell transplant followed. Throughout her initial treatment, the psoriasis continued to flare and she remained on methotrexate to help maintain control of the skin disease.

In November 1998, the patient underwent an autologous CD34⁺ stem cell transplant. There were no complications following the transplant and the multiple myeloma has remained in complete remission to date. Following the stem cell transplant, the patient remained free of psoriatic disease for just over 13 years without any intervention. Recurrent erythematous plaques with thick scale on the bilateral lower legs prompted her return to dermatology in August 2011. To date, she continues to note intermittent though mild flares of her psoriasis. She has responded well to methotrexate 7.5 mg weekly and clobetasol propionate ointment 0.05% as needed with satisfactory control of her disease.

The development of autoimmune disease following an allogenic bone marrow transplant such as psoriasis, thyroiditis, myasthenia gravis, insulin-dependent diabetes mellitus, and vitiligo has been well described.^2-4 The causal relationship between allogenic hematopoetic stem cell transplant (HSCT) and subsequent autoimmune disease has been attributed to T cells and passive transfer of autoimmune disease from the donor to the recipient.¹

In the case of autologous HSCT, the remission of autoimmune disease thereafter has been attributed to myeloablative conditioning, whereby high-dose chemotherapy eliminates self-reactive lymphocytes, resulting in ablation of peripheral reactive and alloreactive T cells with depletion of thymus and reactive B cells.¹

In 1997, Cooley et al⁵ described 4 patients with autoimmune disease who experienced complete but temporary (up to 26 months) remission following autologous bone marrow or stem cell transplant. A comprehensive review of psoriasis cases that resolved after HSCT published by Kaffenberger et al⁶ in 2013 cited 6 cases of autologous and 13 cases of allogenic transplant that resulted in resolution of psoriatic disease. Recurrence of disease was witnessed in 3 of 13 allogenic HSCT patients and 5 of 6 autologous HSCT patients. Of note, all 5 of these patients developed recurrent disease within 24 months following autologous HSCT, albeit with more benign disease.⁶

Similar to those cases presented by Kaffenberger et al,⁶ our patient’s psoriatic disease was notably mild compared to her disease prior to HSCT. Although an unintended but welcome outcome of multiple myeloma treatment, the nearly quiescent nature of her skin disease following autologous HSCT has led to a dramatic improvement in her quality of life.

We highlight a unique case in which psoriatic skin disease recurred more than 1 decade after autologous HSCT.

To the Editor:

Remission of psoriasis after bone marrow transplant has been reported¹; however, follow-up typically has been limited. We describe a case of a remote recurrence of psoriasis 13 years after autologous stem cell transplant.

In September 1997, a 48-year-old woman with a 20-year history of moderate to severe plaque psoriasis, previously managed with oral methotrexate (up to 30–40 mg weekly), presented to her primary care physician with a persistent cough and newly pervasive arthralgia and myalgia. A complete blood cell count showed a white blood cell count of 6.9 mg/dL (reference range, 3.8–10.8 mg/dL), hemoglobin level of 11.6 mg/dL (reference range, 11.7–15.5 mg/dL), and hematocrit level of 34% (reference range, 35.0%–45.0%). Bone survey demonstrated a 4×2-cm lytic expansile lesion in the occipital bone, and a bone marrow aspirate and biopsy revealed 40% plasma cells. She was subsequently diagnosed with stage IIIA IgA λ and free λ light chain multiple myeloma with an IgA level of 62.1 g/L (reference range, 1.24–2.17 g/L).

Therapy was initiated with dexamethasone in combination with pamidronate and she was transitioned to cyclophosphamide chemotherapy. A peripheral autologous blood stem cell transplant followed. Throughout her initial treatment, the psoriasis continued to flare and she remained on methotrexate to help maintain control of the skin disease.

In November 1998, the patient underwent an autologous CD34⁺ stem cell transplant. There were no complications following the transplant and the multiple myeloma has remained in complete remission to date. Following the stem cell transplant, the patient remained free of psoriatic disease for just over 13 years without any intervention. Recurrent erythematous plaques with thick scale on the bilateral lower legs prompted her return to dermatology in August 2011. To date, she continues to note intermittent though mild flares of her psoriasis. She has responded well to methotrexate 7.5 mg weekly and clobetasol propionate ointment 0.05% as needed with satisfactory control of her disease.

The development of autoimmune disease following an allogenic bone marrow transplant such as psoriasis, thyroiditis, myasthenia gravis, insulin-dependent diabetes mellitus, and vitiligo has been well described.^2-4 The causal relationship between allogenic hematopoetic stem cell transplant (HSCT) and subsequent autoimmune disease has been attributed to T cells and passive transfer of autoimmune disease from the donor to the recipient.¹

In the case of autologous HSCT, the remission of autoimmune disease thereafter has been attributed to myeloablative conditioning, whereby high-dose chemotherapy eliminates self-reactive lymphocytes, resulting in ablation of peripheral reactive and alloreactive T cells with depletion of thymus and reactive B cells.¹

In 1997, Cooley et al⁵ described 4 patients with autoimmune disease who experienced complete but temporary (up to 26 months) remission following autologous bone marrow or stem cell transplant. A comprehensive review of psoriasis cases that resolved after HSCT published by Kaffenberger et al⁶ in 2013 cited 6 cases of autologous and 13 cases of allogenic transplant that resulted in resolution of psoriatic disease. Recurrence of disease was witnessed in 3 of 13 allogenic HSCT patients and 5 of 6 autologous HSCT patients. Of note, all 5 of these patients developed recurrent disease within 24 months following autologous HSCT, albeit with more benign disease.⁶

Similar to those cases presented by Kaffenberger et al,⁶ our patient’s psoriatic disease was notably mild compared to her disease prior to HSCT. Although an unintended but welcome outcome of multiple myeloma treatment, the nearly quiescent nature of her skin disease following autologous HSCT has led to a dramatic improvement in her quality of life.

We highlight a unique case in which psoriatic skin disease recurred more than 1 decade after autologous HSCT.

To the Editor:

Benign cephalic histiocytosis (BCH) falls into the group of non–Langerhans cell histiocytosis (non-LCH), which is characterized by a benign course and tendency toward spontaneous remission. Apart from BCH, the main types of non-LCH include juvenile xanthogranuloma, generalized eruptive histiocytoma, and xanthoma disseminatum.¹

Benign cephalic histiocytosis is a rare form of cutaneous histiocytosis in young children. It presents as a papular eruption on the head and has not been associated with internal organ involvement.^2-4 It was described in 1971 by Gianotti et al⁵ and was named infantile histiocytosis with intracytoplasmic worm-like bodies because electron microscopy revealed histiocytes with large cytoplasmatic inclusions composed of wormlike membranous profiles and absence of Birbeck granules. In BCH, skin lesions are located on the head including the face and sometimes on the neck. Lesions occasionally may appear on the trunk, buttocks, and thighs. Mucous membranes are not involved. The onset of disease is typical in the first year of life; however, the disease may begin within the first 3 years of life. An eruption is characterized by small, 2- to 8-mm, discrete, asymptomatic, tan to red-brown macules and papules. The lesions may persist for several months or years and subsequently flatten, becoming hyperpigmented briefly. They often completely resolve with time. Most children are otherwise healthy and developmentally normal^6-9; however, diabetes insipidus has been reported in some children with BCH.¹⁰

Histologic examination of skin samples reveals the infiltrate of histiocytes, which closely approaches the epidermis, accompanied by scattered lymphocytes and a few eosinophils.^1,2,11 The histiocytes express a typical macrophage marker CD68, whereas immunostaining for Langerhans cell markers such as CD1a and S-100 is negative.^3,9,12,13

A 2.5-year-old boy was admitted to our dermatology department with suspected cutaneous mastocytosis (CM). Since the age of 13 months, he had developed small, 4- to 8-mm, dark pinkish macules and papules localized on the cheeks (Figure 1). Physical examination performed in our center revealed yellowish macules and flat papules limited to the cheeks. Darier sign was negative. The boy was otherwise healthy and developmentally normal. All laboratory tests were within reference range and his family history was uneventful.

Figure 1. Maculopapular eruption of benign cephalic histiocytosis on the cheeks (A and B).

Histopathologic examination of the skin sample revealed a normotypic epidermis and scattered subepidermal infiltrates in the upper dermis. The infiltrates were composed of predominating histiocytes and a few mast cells and eosinophils (Figure 2). The histiocytes were slightly pleomorphic and had abundant clear cytoplasm, vesicular nuclei, and prominent nucleoli. Mitoses were absent in these cells. The majority of cells within the infiltrate expressed CD68 and were CD1a^- and S-100^-. However, occasional CD1a⁺ cells were seen. Immunostaining for mast cell marker CD117 was negative. Cutaneous mastocytosis was excluded and non-LCH was recognized. Based on the typical location, morphology, and immunophenotype of skin lesions, BCH was diagnosed. At 24-month follow-up, spontaneous regression of skin lesions was observed.

Gianotti et al⁷ described BCH as a separate entity of the non-LCH group of disorders and established its diagnostic criteria: (1) onset of disease within the first 3 years of life; (2) location of skin lesions on the scalp and lack of lesions on the hands, feet, mucous membranes, and internal organ involvement; (3) spontaneous complete remission of symptoms; and (4) monomorphic infiltration of histiocytes that do not express S-100 and CD1a.

The macular and flat, papular, pink-yellow or orange lesions visible on the face of our patient are characteristic of BCH. Moreover, the cheeks are the most typical location of a BCH eruption, as noted in our patient.^6,7,12 The presence of histiocytic infiltrates composed of CD68⁺ cells strongly support the diagnosis.^3,4,9-13 In contrast to other reports, occasional CD1a⁺ cells of Langerhans phenotype were found in our case.^3,9,11,12 The proliferation of Langerhans cells in the skin and internal organs and presence of langerin are characteristic of Langerhans cell histiocytosis (LCH).^1,4,14 The presence of a few CD1a cells in cases with clinical features compatible with non-LCH may suggest that some of these cases may represent a papular self-healing variant of LCH or may indicate that there is an overlap among the histiocytic syndromes (eg, non-LCH and LCH). Furthermore, many of benign histiocytic lesions may evolve over the course of time into the others.^12,13 Differential diagnosis of BCH should include other benign forms of cutaneous histiocytosis, particularly the small nodular variant of juvenile xanthogranuloma and generalized eruptive histiocytoma (GEH). Juvenile xanthogranuloma pre-sents as disseminated, yellowish, nodular lesions and may be associated with ocular involvement, whereas GEH is characterized by rapid onset of the disease and disseminated nodular eruption.^1,4

Figure 2. Skin section showing the upper and mid dermis infiltrated with slightly pleomorphic epithelioid histiocytic cells with clear cytoplasm and vesicular nuclei. Few accompanying lymphocytes and eosinophils were visible (H&E, original magnifications ×200 and ×400).

A close histologic relationship and presence of overlapping symptoms observed among BCH, GEH, and juvenile xanthogranuloma indicate that these entities fall into a spectrum of the same disorder. However, the presence of a uniform infiltrate of large foamy histiocytes readily distinguishes xanthomas from BCH.⁴ In some unusual clinical presentations of CM or in cases of the nodular form of the condition, there is a need to distinguish between non-LCH and CM, as in our patient. Darier sign, consisting of urtication and erythema appearing after mechanical irritation of the skin lesion, is pathognomonic for CM. Nevertheless, Darier sign is not sufficient to confirm CM when it is not pronounced. Therefore, histologic examination with the use of immunostaining plays a key role in the differential diagnosis of these disorders in children.¹⁵ Treatment of BCH is not recommended because of spontaneous remission of the disease.^1-5

Benign cephalic histiocytosis is a rare clinical form of non-LCH. No systemic or mucosal involvement has been described. Lesions often are confused with plane warts, but a biopsy is definitive. Therapy is not effective but fortunately none is necessary.

To the Editor:

Benign cephalic histiocytosis (BCH) falls into the group of non–Langerhans cell histiocytosis (non-LCH), which is characterized by a benign course and tendency toward spontaneous remission. Apart from BCH, the main types of non-LCH include juvenile xanthogranuloma, generalized eruptive histiocytoma, and xanthoma disseminatum.¹

Benign cephalic histiocytosis is a rare form of cutaneous histiocytosis in young children. It presents as a papular eruption on the head and has not been associated with internal organ involvement.^2-4 It was described in 1971 by Gianotti et al⁵ and was named infantile histiocytosis with intracytoplasmic worm-like bodies because electron microscopy revealed histiocytes with large cytoplasmatic inclusions composed of wormlike membranous profiles and absence of Birbeck granules. In BCH, skin lesions are located on the head including the face and sometimes on the neck. Lesions occasionally may appear on the trunk, buttocks, and thighs. Mucous membranes are not involved. The onset of disease is typical in the first year of life; however, the disease may begin within the first 3 years of life. An eruption is characterized by small, 2- to 8-mm, discrete, asymptomatic, tan to red-brown macules and papules. The lesions may persist for several months or years and subsequently flatten, becoming hyperpigmented briefly. They often completely resolve with time. Most children are otherwise healthy and developmentally normal^6-9; however, diabetes insipidus has been reported in some children with BCH.¹⁰

Histologic examination of skin samples reveals the infiltrate of histiocytes, which closely approaches the epidermis, accompanied by scattered lymphocytes and a few eosinophils.^1,2,11 The histiocytes express a typical macrophage marker CD68, whereas immunostaining for Langerhans cell markers such as CD1a and S-100 is negative.^3,9,12,13

A 2.5-year-old boy was admitted to our dermatology department with suspected cutaneous mastocytosis (CM). Since the age of 13 months, he had developed small, 4- to 8-mm, dark pinkish macules and papules localized on the cheeks (Figure 1). Physical examination performed in our center revealed yellowish macules and flat papules limited to the cheeks. Darier sign was negative. The boy was otherwise healthy and developmentally normal. All laboratory tests were within reference range and his family history was uneventful.

Figure 1. Maculopapular eruption of benign cephalic histiocytosis on the cheeks (A and B).

Histopathologic examination of the skin sample revealed a normotypic epidermis and scattered subepidermal infiltrates in the upper dermis. The infiltrates were composed of predominating histiocytes and a few mast cells and eosinophils (Figure 2). The histiocytes were slightly pleomorphic and had abundant clear cytoplasm, vesicular nuclei, and prominent nucleoli. Mitoses were absent in these cells. The majority of cells within the infiltrate expressed CD68 and were CD1a^- and S-100^-. However, occasional CD1a⁺ cells were seen. Immunostaining for mast cell marker CD117 was negative. Cutaneous mastocytosis was excluded and non-LCH was recognized. Based on the typical location, morphology, and immunophenotype of skin lesions, BCH was diagnosed. At 24-month follow-up, spontaneous regression of skin lesions was observed.

Gianotti et al⁷ described BCH as a separate entity of the non-LCH group of disorders and established its diagnostic criteria: (1) onset of disease within the first 3 years of life; (2) location of skin lesions on the scalp and lack of lesions on the hands, feet, mucous membranes, and internal organ involvement; (3) spontaneous complete remission of symptoms; and (4) monomorphic infiltration of histiocytes that do not express S-100 and CD1a.

The macular and flat, papular, pink-yellow or orange lesions visible on the face of our patient are characteristic of BCH. Moreover, the cheeks are the most typical location of a BCH eruption, as noted in our patient.^6,7,12 The presence of histiocytic infiltrates composed of CD68⁺ cells strongly support the diagnosis.^3,4,9-13 In contrast to other reports, occasional CD1a⁺ cells of Langerhans phenotype were found in our case.^3,9,11,12 The proliferation of Langerhans cells in the skin and internal organs and presence of langerin are characteristic of Langerhans cell histiocytosis (LCH).^1,4,14 The presence of a few CD1a cells in cases with clinical features compatible with non-LCH may suggest that some of these cases may represent a papular self-healing variant of LCH or may indicate that there is an overlap among the histiocytic syndromes (eg, non-LCH and LCH). Furthermore, many of benign histiocytic lesions may evolve over the course of time into the others.^12,13 Differential diagnosis of BCH should include other benign forms of cutaneous histiocytosis, particularly the small nodular variant of juvenile xanthogranuloma and generalized eruptive histiocytoma (GEH). Juvenile xanthogranuloma pre-sents as disseminated, yellowish, nodular lesions and may be associated with ocular involvement, whereas GEH is characterized by rapid onset of the disease and disseminated nodular eruption.^1,4

Figure 2. Skin section showing the upper and mid dermis infiltrated with slightly pleomorphic epithelioid histiocytic cells with clear cytoplasm and vesicular nuclei. Few accompanying lymphocytes and eosinophils were visible (H&E, original magnifications ×200 and ×400).

A close histologic relationship and presence of overlapping symptoms observed among BCH, GEH, and juvenile xanthogranuloma indicate that these entities fall into a spectrum of the same disorder. However, the presence of a uniform infiltrate of large foamy histiocytes readily distinguishes xanthomas from BCH.⁴ In some unusual clinical presentations of CM or in cases of the nodular form of the condition, there is a need to distinguish between non-LCH and CM, as in our patient. Darier sign, consisting of urtication and erythema appearing after mechanical irritation of the skin lesion, is pathognomonic for CM. Nevertheless, Darier sign is not sufficient to confirm CM when it is not pronounced. Therefore, histologic examination with the use of immunostaining plays a key role in the differential diagnosis of these disorders in children.¹⁵ Treatment of BCH is not recommended because of spontaneous remission of the disease.^1-5

Benign cephalic histiocytosis is a rare clinical form of non-LCH. No systemic or mucosal involvement has been described. Lesions often are confused with plane warts, but a biopsy is definitive. Therapy is not effective but fortunately none is necessary.

To the Editor:

Benign cephalic histiocytosis (BCH) falls into the group of non–Langerhans cell histiocytosis (non-LCH), which is characterized by a benign course and tendency toward spontaneous remission. Apart from BCH, the main types of non-LCH include juvenile xanthogranuloma, generalized eruptive histiocytoma, and xanthoma disseminatum.¹

Benign cephalic histiocytosis is a rare form of cutaneous histiocytosis in young children. It presents as a papular eruption on the head and has not been associated with internal organ involvement.^2-4 It was described in 1971 by Gianotti et al⁵ and was named infantile histiocytosis with intracytoplasmic worm-like bodies because electron microscopy revealed histiocytes with large cytoplasmatic inclusions composed of wormlike membranous profiles and absence of Birbeck granules. In BCH, skin lesions are located on the head including the face and sometimes on the neck. Lesions occasionally may appear on the trunk, buttocks, and thighs. Mucous membranes are not involved. The onset of disease is typical in the first year of life; however, the disease may begin within the first 3 years of life. An eruption is characterized by small, 2- to 8-mm, discrete, asymptomatic, tan to red-brown macules and papules. The lesions may persist for several months or years and subsequently flatten, becoming hyperpigmented briefly. They often completely resolve with time. Most children are otherwise healthy and developmentally normal^6-9; however, diabetes insipidus has been reported in some children with BCH.¹⁰

Histologic examination of skin samples reveals the infiltrate of histiocytes, which closely approaches the epidermis, accompanied by scattered lymphocytes and a few eosinophils.^1,2,11 The histiocytes express a typical macrophage marker CD68, whereas immunostaining for Langerhans cell markers such as CD1a and S-100 is negative.^3,9,12,13

A 2.5-year-old boy was admitted to our dermatology department with suspected cutaneous mastocytosis (CM). Since the age of 13 months, he had developed small, 4- to 8-mm, dark pinkish macules and papules localized on the cheeks (Figure 1). Physical examination performed in our center revealed yellowish macules and flat papules limited to the cheeks. Darier sign was negative. The boy was otherwise healthy and developmentally normal. All laboratory tests were within reference range and his family history was uneventful.

Figure 1. Maculopapular eruption of benign cephalic histiocytosis on the cheeks (A and B).

Histopathologic examination of the skin sample revealed a normotypic epidermis and scattered subepidermal infiltrates in the upper dermis. The infiltrates were composed of predominating histiocytes and a few mast cells and eosinophils (Figure 2). The histiocytes were slightly pleomorphic and had abundant clear cytoplasm, vesicular nuclei, and prominent nucleoli. Mitoses were absent in these cells. The majority of cells within the infiltrate expressed CD68 and were CD1a^- and S-100^-. However, occasional CD1a⁺ cells were seen. Immunostaining for mast cell marker CD117 was negative. Cutaneous mastocytosis was excluded and non-LCH was recognized. Based on the typical location, morphology, and immunophenotype of skin lesions, BCH was diagnosed. At 24-month follow-up, spontaneous regression of skin lesions was observed.

Gianotti et al⁷ described BCH as a separate entity of the non-LCH group of disorders and established its diagnostic criteria: (1) onset of disease within the first 3 years of life; (2) location of skin lesions on the scalp and lack of lesions on the hands, feet, mucous membranes, and internal organ involvement; (3) spontaneous complete remission of symptoms; and (4) monomorphic infiltration of histiocytes that do not express S-100 and CD1a.

The macular and flat, papular, pink-yellow or orange lesions visible on the face of our patient are characteristic of BCH. Moreover, the cheeks are the most typical location of a BCH eruption, as noted in our patient.^6,7,12 The presence of histiocytic infiltrates composed of CD68⁺ cells strongly support the diagnosis.^3,4,9-13 In contrast to other reports, occasional CD1a⁺ cells of Langerhans phenotype were found in our case.^3,9,11,12 The proliferation of Langerhans cells in the skin and internal organs and presence of langerin are characteristic of Langerhans cell histiocytosis (LCH).^1,4,14 The presence of a few CD1a cells in cases with clinical features compatible with non-LCH may suggest that some of these cases may represent a papular self-healing variant of LCH or may indicate that there is an overlap among the histiocytic syndromes (eg, non-LCH and LCH). Furthermore, many of benign histiocytic lesions may evolve over the course of time into the others.^12,13 Differential diagnosis of BCH should include other benign forms of cutaneous histiocytosis, particularly the small nodular variant of juvenile xanthogranuloma and generalized eruptive histiocytoma (GEH). Juvenile xanthogranuloma pre-sents as disseminated, yellowish, nodular lesions and may be associated with ocular involvement, whereas GEH is characterized by rapid onset of the disease and disseminated nodular eruption.^1,4

Figure 2. Skin section showing the upper and mid dermis infiltrated with slightly pleomorphic epithelioid histiocytic cells with clear cytoplasm and vesicular nuclei. Few accompanying lymphocytes and eosinophils were visible (H&E, original magnifications ×200 and ×400).

A close histologic relationship and presence of overlapping symptoms observed among BCH, GEH, and juvenile xanthogranuloma indicate that these entities fall into a spectrum of the same disorder. However, the presence of a uniform infiltrate of large foamy histiocytes readily distinguishes xanthomas from BCH.⁴ In some unusual clinical presentations of CM or in cases of the nodular form of the condition, there is a need to distinguish between non-LCH and CM, as in our patient. Darier sign, consisting of urtication and erythema appearing after mechanical irritation of the skin lesion, is pathognomonic for CM. Nevertheless, Darier sign is not sufficient to confirm CM when it is not pronounced. Therefore, histologic examination with the use of immunostaining plays a key role in the differential diagnosis of these disorders in children.¹⁵ Treatment of BCH is not recommended because of spontaneous remission of the disease.^1-5

Benign cephalic histiocytosis is a rare clinical form of non-LCH. No systemic or mucosal involvement has been described. Lesions often are confused with plane warts, but a biopsy is definitive. Therapy is not effective but fortunately none is necessary.

To the Editor:

A 26-year-old man was referred to our clinic for evaluation of a persistent red rash to rule out cutaneous lupus erythematosus (LE). The patient was diagnosed at 12 years of age with autosomal-recessive chronic granulomatous disease (CGD)(nitroblue tetrazolium test, 5.0; low normal, 20.6), type p47^phox mutation. At that time the patient had recurrent fevers, sinusitis, anemia, and noncaseating granulomatous liver lesions, but he lacked any cutaneous manifestations. The patient was then treated for approximately 2 years with interferon therapy but discontinued therapy given the absence of any signs or symptoms. He remained asymptomatic until approximately 16 years of age when he experienced the onset of an intermittently painful and pruritic rash on the face that slowly spread over the ensuing years to involve the trunk, arms, forearms, and hands. Although he reported that sunlight exacerbated the rash, the rash also persisted through the winter months when the majority of the sun-exposed areas of the trunk and arms were covered. He denied exposure to topical products and denied the use of any oral medications (prescription or over-the-counter).

Review of systems was negative for fever, fatigue, malaise, headaches, joint pain, arthritis, oral ulcers, dyspnea, or dysuria. Physical examination revealed a well-defined exanthem comprised of erythematous, mildly indurated papules coalescing into larger plaques with white scale that were exclusively limited to the photodistributed areas of the face (Figure 1), neck, arms, forearms, hands, chest, and back. Laboratory test results included the following: minimally elevated erythrocyte sedimentation rate of 31 mm/h (reference range, 0–15 mm/h) and rheumatoid factor of 45 IU/mL (reference range, <20 IU/mL; negative antinuclear antibody screen, Sjögren syndrome antigens A and B, double-stranded DNA, anti–extractable nuclear antigen antibody test, and anti-Jo-1 antibody; complete blood cell count revealed no abnormalities; basic metabolic panel, C3 and C4, CH50, glucose-6-phosphate dehydrogenase activity, total plasma porphyrins, and testing for hepatitis B and C virus and human immunodeficiency virus serologies were negative. Skin biopsy from a lesion on the lateral arm showed features consistent with interface dermatitis (Figure 2). Additional skin biopsies for direct immunofluorescence showed linear deposition of IgG at the dermoepidermal junction, both from involved and uninvolved neck skin (more focally from the involved site). Extensive photopatch testing did not show any clinically relevant positive reactions.

Figure 1. Erythematous patches and thin plaques in photodistributed areas on the face, with sparing of the nasolabial folds. Figure 2. A biopsy from the right lateral arm showed basal layer vacuolization; scattered individually necrotic keratinocytes (black arrows; H&E, original magnification ×400 [inset]); increased dermal mucin; and a mild superficial, predominantly perivascular, lymphocytic infiltrate with mild leukocytoclasis and focal extravasated erythrocytes (H&E, original magnification ×100).

Given the patient’s history of CGD and the extensive negative workup for rheumatologic, photoallergic, and phototoxic causes, the patient was diagnosed with a lupus-like rash of CGD. The rash failed to respond to rigorous sun avoidance and a 3-week on/1-week off regimen of high-potency class 1 topical steroids to the trunk, arms, forearms, and legs, and lower-potency class 4 topical steroid to the face, with disease flaring almost immediately on cessation of treatment during the rest weeks. Given the marked photodistribution resembling subacute cutaneous LE, oral hydroxychloroquine 200 mg (5.7 mg/kg) twice daily was initiated in addition to continued topical steroid therapy.

Four months after the addition of hydroxychloroquine, the patient showed considerable improvement of the rash. Seven months after initiation of hydroxychloroquine, the photodistributed rash was completely resolved and topical steroids were stopped. The rash remained in remission for an additional 24 months with hydroxychloroquine alone, at which time hydroxychloroquine was stopped; however, the rash flared 2 months later and hydroxychloroquine was restarted at 200 mg twice daily, resulting in clearance within 3 months. The patient was maintained on this dose of hydroxychloroquine.

During treatment, the patient had an episode of extensive furunculosis caused by Staphylococcus aureus that was successfully treated with a 14-day course of oral doxycycline 100 mg twice daily. He has since been maintained on prophylactic intranasal mupirocin ointment 2% for the first several days of each month and daily benzoyl peroxide wash 10% without further episodes. He also developed a single lesion of alopecia areata that was successfully treated with intralesional steroid injections.

Chronic granulomatous disease can either be X-linked or, less commonly, autosomal recessive, resulting from a defect in components of the nicotinamide adenine dinucleotide phosphate oxidase complex that is necessary to generate reactive oxygen intermediates for killing phagocytosed microbes. Cutaneous manifestations are relatively common in CGD (60%–70% of cases)¹ and include infectious lesions (eg, recurrent mucous membrane infections, impetigo, carbuncles, otitis externa, suppurative lymphadenopathy) as well as the less common chronic inflammatory conditions such as lupus-like eruption, aphthous stomatitis, Raynaud phenomenon, arcuate dermal erythema, and Jessner lymphocytic infiltrate.² The pathognomonic clinical feature of CGD is the presence of characteristic multinucleated giant cell granulomas distributed in multiple organ systems such as the gastrointestinal system, causing pyloric and/or small bowel obstruction, and the genitourinary system, causing ureter and/or bladder outlet obstruction.³

Importantly, CGD patients also demonstrate immune-related inflammatory disorders, most commonly inflammatory bowel disease, IgA nephropathy, sarcoidosis, and juvenile idiopathic arthritis.³ In addition, both CGD patients and female carriers of X-linked CGD have been reported to demonstrate lupus-like rashes that share overlapping clinical and histologic features with the rashes seen in true discoid LE and tumid LE patients without CGD.^4-6 This lupus-like rash is more commonly observed in adulthood and in carriers, possibly secondary to the high childhood mortality rate of CGD patients.^4,6

De Ravin et al³ proposed that autoimmune conditions arising in CGD patients who have met established criteria for a particular autoimmune disease should be treated for that condition rather than consider it as a part of the CGD spectrum. This theory has important therapeutic implications, including initiating paradoxical corticosteroid and/or steroid-sparing immunosuppressive agents in this otherwise immunocompromised patient population. They reported a 21-year-old man with cutaneous LE lesions and negative lupus serologies whose lesions were refractory to topical steroids but responded to systemic prednisone, requiring a low-dose alternate-day maintenance regimen.³ Beyond the development of a true autoimmune disease associated with CGD, systemic medications, specifically voriconazole, have been implicated as an alternative etiology for this rash in CGD.⁷ While important to consider, our patient’s rash presented in the absence of any systemic medications, supporting the former etiology over the latter.

Our case demonstrates the utility of hydroxychloroquine to treat the lupus-like rash of CGD. Similarly, the lupus-like symptoms of female carriers of X-linked CGD, predominantly with negative lupus serologies, also have been reported to respond to hydroxychloroquine and mepacrine.^4,5,8-10 Interestingly, the utility of monotherapy with hydroxychloroquine may extend beyond treating cutaneous lupus-like lesions, as this regimen also was reported to successfully treat gastric granulomatous involvement in a CGD patient.¹¹

Chronic granulomatous disease often is fatal in early childhood or adolescence due to sequelae from infections or chronic granulomatous infiltration of internal organs. Residual reactive oxygen intermediate production was shown to be a predictor of overall survival, and CGD patients with 1% of normal reactive oxygen intermediate production by neutrophils had a greater likelihood of survival.¹² In this regard, the otherwise good health of our patient at the time of presentation was consistent with his initial nitroblue tetrazolium test showing some residual oxidative activity, emphasizing the phenotypic variability of this rare genetic disorder and the importance of considering CGD in the diagnosis of seronegative cutaneous lupus-like reactions.

To the Editor:

A 26-year-old man was referred to our clinic for evaluation of a persistent red rash to rule out cutaneous lupus erythematosus (LE). The patient was diagnosed at 12 years of age with autosomal-recessive chronic granulomatous disease (CGD)(nitroblue tetrazolium test, 5.0; low normal, 20.6), type p47^phox mutation. At that time the patient had recurrent fevers, sinusitis, anemia, and noncaseating granulomatous liver lesions, but he lacked any cutaneous manifestations. The patient was then treated for approximately 2 years with interferon therapy but discontinued therapy given the absence of any signs or symptoms. He remained asymptomatic until approximately 16 years of age when he experienced the onset of an intermittently painful and pruritic rash on the face that slowly spread over the ensuing years to involve the trunk, arms, forearms, and hands. Although he reported that sunlight exacerbated the rash, the rash also persisted through the winter months when the majority of the sun-exposed areas of the trunk and arms were covered. He denied exposure to topical products and denied the use of any oral medications (prescription or over-the-counter).

Review of systems was negative for fever, fatigue, malaise, headaches, joint pain, arthritis, oral ulcers, dyspnea, or dysuria. Physical examination revealed a well-defined exanthem comprised of erythematous, mildly indurated papules coalescing into larger plaques with white scale that were exclusively limited to the photodistributed areas of the face (Figure 1), neck, arms, forearms, hands, chest, and back. Laboratory test results included the following: minimally elevated erythrocyte sedimentation rate of 31 mm/h (reference range, 0–15 mm/h) and rheumatoid factor of 45 IU/mL (reference range, <20 IU/mL; negative antinuclear antibody screen, Sjögren syndrome antigens A and B, double-stranded DNA, anti–extractable nuclear antigen antibody test, and anti-Jo-1 antibody; complete blood cell count revealed no abnormalities; basic metabolic panel, C3 and C4, CH50, glucose-6-phosphate dehydrogenase activity, total plasma porphyrins, and testing for hepatitis B and C virus and human immunodeficiency virus serologies were negative. Skin biopsy from a lesion on the lateral arm showed features consistent with interface dermatitis (Figure 2). Additional skin biopsies for direct immunofluorescence showed linear deposition of IgG at the dermoepidermal junction, both from involved and uninvolved neck skin (more focally from the involved site). Extensive photopatch testing did not show any clinically relevant positive reactions.

Figure 1. Erythematous patches and thin plaques in photodistributed areas on the face, with sparing of the nasolabial folds. Figure 2. A biopsy from the right lateral arm showed basal layer vacuolization; scattered individually necrotic keratinocytes (black arrows; H&E, original magnification ×400 [inset]); increased dermal mucin; and a mild superficial, predominantly perivascular, lymphocytic infiltrate with mild leukocytoclasis and focal extravasated erythrocytes (H&E, original magnification ×100).

Given the patient’s history of CGD and the extensive negative workup for rheumatologic, photoallergic, and phototoxic causes, the patient was diagnosed with a lupus-like rash of CGD. The rash failed to respond to rigorous sun avoidance and a 3-week on/1-week off regimen of high-potency class 1 topical steroids to the trunk, arms, forearms, and legs, and lower-potency class 4 topical steroid to the face, with disease flaring almost immediately on cessation of treatment during the rest weeks. Given the marked photodistribution resembling subacute cutaneous LE, oral hydroxychloroquine 200 mg (5.7 mg/kg) twice daily was initiated in addition to continued topical steroid therapy.

Four months after the addition of hydroxychloroquine, the patient showed considerable improvement of the rash. Seven months after initiation of hydroxychloroquine, the photodistributed rash was completely resolved and topical steroids were stopped. The rash remained in remission for an additional 24 months with hydroxychloroquine alone, at which time hydroxychloroquine was stopped; however, the rash flared 2 months later and hydroxychloroquine was restarted at 200 mg twice daily, resulting in clearance within 3 months. The patient was maintained on this dose of hydroxychloroquine.

During treatment, the patient had an episode of extensive furunculosis caused by Staphylococcus aureus that was successfully treated with a 14-day course of oral doxycycline 100 mg twice daily. He has since been maintained on prophylactic intranasal mupirocin ointment 2% for the first several days of each month and daily benzoyl peroxide wash 10% without further episodes. He also developed a single lesion of alopecia areata that was successfully treated with intralesional steroid injections.

Chronic granulomatous disease can either be X-linked or, less commonly, autosomal recessive, resulting from a defect in components of the nicotinamide adenine dinucleotide phosphate oxidase complex that is necessary to generate reactive oxygen intermediates for killing phagocytosed microbes. Cutaneous manifestations are relatively common in CGD (60%–70% of cases)¹ and include infectious lesions (eg, recurrent mucous membrane infections, impetigo, carbuncles, otitis externa, suppurative lymphadenopathy) as well as the less common chronic inflammatory conditions such as lupus-like eruption, aphthous stomatitis, Raynaud phenomenon, arcuate dermal erythema, and Jessner lymphocytic infiltrate.² The pathognomonic clinical feature of CGD is the presence of characteristic multinucleated giant cell granulomas distributed in multiple organ systems such as the gastrointestinal system, causing pyloric and/or small bowel obstruction, and the genitourinary system, causing ureter and/or bladder outlet obstruction.³

Importantly, CGD patients also demonstrate immune-related inflammatory disorders, most commonly inflammatory bowel disease, IgA nephropathy, sarcoidosis, and juvenile idiopathic arthritis.³ In addition, both CGD patients and female carriers of X-linked CGD have been reported to demonstrate lupus-like rashes that share overlapping clinical and histologic features with the rashes seen in true discoid LE and tumid LE patients without CGD.^4-6 This lupus-like rash is more commonly observed in adulthood and in carriers, possibly secondary to the high childhood mortality rate of CGD patients.^4,6

De Ravin et al³ proposed that autoimmune conditions arising in CGD patients who have met established criteria for a particular autoimmune disease should be treated for that condition rather than consider it as a part of the CGD spectrum. This theory has important therapeutic implications, including initiating paradoxical corticosteroid and/or steroid-sparing immunosuppressive agents in this otherwise immunocompromised patient population. They reported a 21-year-old man with cutaneous LE lesions and negative lupus serologies whose lesions were refractory to topical steroids but responded to systemic prednisone, requiring a low-dose alternate-day maintenance regimen.³ Beyond the development of a true autoimmune disease associated with CGD, systemic medications, specifically voriconazole, have been implicated as an alternative etiology for this rash in CGD.⁷ While important to consider, our patient’s rash presented in the absence of any systemic medications, supporting the former etiology over the latter.

Our case demonstrates the utility of hydroxychloroquine to treat the lupus-like rash of CGD. Similarly, the lupus-like symptoms of female carriers of X-linked CGD, predominantly with negative lupus serologies, also have been reported to respond to hydroxychloroquine and mepacrine.^4,5,8-10 Interestingly, the utility of monotherapy with hydroxychloroquine may extend beyond treating cutaneous lupus-like lesions, as this regimen also was reported to successfully treat gastric granulomatous involvement in a CGD patient.¹¹

Chronic granulomatous disease often is fatal in early childhood or adolescence due to sequelae from infections or chronic granulomatous infiltration of internal organs. Residual reactive oxygen intermediate production was shown to be a predictor of overall survival, and CGD patients with 1% of normal reactive oxygen intermediate production by neutrophils had a greater likelihood of survival.¹² In this regard, the otherwise good health of our patient at the time of presentation was consistent with his initial nitroblue tetrazolium test showing some residual oxidative activity, emphasizing the phenotypic variability of this rare genetic disorder and the importance of considering CGD in the diagnosis of seronegative cutaneous lupus-like reactions.

To the Editor:

A 26-year-old man was referred to our clinic for evaluation of a persistent red rash to rule out cutaneous lupus erythematosus (LE). The patient was diagnosed at 12 years of age with autosomal-recessive chronic granulomatous disease (CGD)(nitroblue tetrazolium test, 5.0; low normal, 20.6), type p47^phox mutation. At that time the patient had recurrent fevers, sinusitis, anemia, and noncaseating granulomatous liver lesions, but he lacked any cutaneous manifestations. The patient was then treated for approximately 2 years with interferon therapy but discontinued therapy given the absence of any signs or symptoms. He remained asymptomatic until approximately 16 years of age when he experienced the onset of an intermittently painful and pruritic rash on the face that slowly spread over the ensuing years to involve the trunk, arms, forearms, and hands. Although he reported that sunlight exacerbated the rash, the rash also persisted through the winter months when the majority of the sun-exposed areas of the trunk and arms were covered. He denied exposure to topical products and denied the use of any oral medications (prescription or over-the-counter).

Review of systems was negative for fever, fatigue, malaise, headaches, joint pain, arthritis, oral ulcers, dyspnea, or dysuria. Physical examination revealed a well-defined exanthem comprised of erythematous, mildly indurated papules coalescing into larger plaques with white scale that were exclusively limited to the photodistributed areas of the face (Figure 1), neck, arms, forearms, hands, chest, and back. Laboratory test results included the following: minimally elevated erythrocyte sedimentation rate of 31 mm/h (reference range, 0–15 mm/h) and rheumatoid factor of 45 IU/mL (reference range, <20 IU/mL; negative antinuclear antibody screen, Sjögren syndrome antigens A and B, double-stranded DNA, anti–extractable nuclear antigen antibody test, and anti-Jo-1 antibody; complete blood cell count revealed no abnormalities; basic metabolic panel, C3 and C4, CH50, glucose-6-phosphate dehydrogenase activity, total plasma porphyrins, and testing for hepatitis B and C virus and human immunodeficiency virus serologies were negative. Skin biopsy from a lesion on the lateral arm showed features consistent with interface dermatitis (Figure 2). Additional skin biopsies for direct immunofluorescence showed linear deposition of IgG at the dermoepidermal junction, both from involved and uninvolved neck skin (more focally from the involved site). Extensive photopatch testing did not show any clinically relevant positive reactions.

Figure 1. Erythematous patches and thin plaques in photodistributed areas on the face, with sparing of the nasolabial folds. Figure 2. A biopsy from the right lateral arm showed basal layer vacuolization; scattered individually necrotic keratinocytes (black arrows; H&E, original magnification ×400 [inset]); increased dermal mucin; and a mild superficial, predominantly perivascular, lymphocytic infiltrate with mild leukocytoclasis and focal extravasated erythrocytes (H&E, original magnification ×100).

Given the patient’s history of CGD and the extensive negative workup for rheumatologic, photoallergic, and phototoxic causes, the patient was diagnosed with a lupus-like rash of CGD. The rash failed to respond to rigorous sun avoidance and a 3-week on/1-week off regimen of high-potency class 1 topical steroids to the trunk, arms, forearms, and legs, and lower-potency class 4 topical steroid to the face, with disease flaring almost immediately on cessation of treatment during the rest weeks. Given the marked photodistribution resembling subacute cutaneous LE, oral hydroxychloroquine 200 mg (5.7 mg/kg) twice daily was initiated in addition to continued topical steroid therapy.

Four months after the addition of hydroxychloroquine, the patient showed considerable improvement of the rash. Seven months after initiation of hydroxychloroquine, the photodistributed rash was completely resolved and topical steroids were stopped. The rash remained in remission for an additional 24 months with hydroxychloroquine alone, at which time hydroxychloroquine was stopped; however, the rash flared 2 months later and hydroxychloroquine was restarted at 200 mg twice daily, resulting in clearance within 3 months. The patient was maintained on this dose of hydroxychloroquine.

During treatment, the patient had an episode of extensive furunculosis caused by Staphylococcus aureus that was successfully treated with a 14-day course of oral doxycycline 100 mg twice daily. He has since been maintained on prophylactic intranasal mupirocin ointment 2% for the first several days of each month and daily benzoyl peroxide wash 10% without further episodes. He also developed a single lesion of alopecia areata that was successfully treated with intralesional steroid injections.

Chronic granulomatous disease can either be X-linked or, less commonly, autosomal recessive, resulting from a defect in components of the nicotinamide adenine dinucleotide phosphate oxidase complex that is necessary to generate reactive oxygen intermediates for killing phagocytosed microbes. Cutaneous manifestations are relatively common in CGD (60%–70% of cases)¹ and include infectious lesions (eg, recurrent mucous membrane infections, impetigo, carbuncles, otitis externa, suppurative lymphadenopathy) as well as the less common chronic inflammatory conditions such as lupus-like eruption, aphthous stomatitis, Raynaud phenomenon, arcuate dermal erythema, and Jessner lymphocytic infiltrate.² The pathognomonic clinical feature of CGD is the presence of characteristic multinucleated giant cell granulomas distributed in multiple organ systems such as the gastrointestinal system, causing pyloric and/or small bowel obstruction, and the genitourinary system, causing ureter and/or bladder outlet obstruction.³

Importantly, CGD patients also demonstrate immune-related inflammatory disorders, most commonly inflammatory bowel disease, IgA nephropathy, sarcoidosis, and juvenile idiopathic arthritis.³ In addition, both CGD patients and female carriers of X-linked CGD have been reported to demonstrate lupus-like rashes that share overlapping clinical and histologic features with the rashes seen in true discoid LE and tumid LE patients without CGD.^4-6 This lupus-like rash is more commonly observed in adulthood and in carriers, possibly secondary to the high childhood mortality rate of CGD patients.^4,6

De Ravin et al³ proposed that autoimmune conditions arising in CGD patients who have met established criteria for a particular autoimmune disease should be treated for that condition rather than consider it as a part of the CGD spectrum. This theory has important therapeutic implications, including initiating paradoxical corticosteroid and/or steroid-sparing immunosuppressive agents in this otherwise immunocompromised patient population. They reported a 21-year-old man with cutaneous LE lesions and negative lupus serologies whose lesions were refractory to topical steroids but responded to systemic prednisone, requiring a low-dose alternate-day maintenance regimen.³ Beyond the development of a true autoimmune disease associated with CGD, systemic medications, specifically voriconazole, have been implicated as an alternative etiology for this rash in CGD.⁷ While important to consider, our patient’s rash presented in the absence of any systemic medications, supporting the former etiology over the latter.

Our case demonstrates the utility of hydroxychloroquine to treat the lupus-like rash of CGD. Similarly, the lupus-like symptoms of female carriers of X-linked CGD, predominantly with negative lupus serologies, also have been reported to respond to hydroxychloroquine and mepacrine.^4,5,8-10 Interestingly, the utility of monotherapy with hydroxychloroquine may extend beyond treating cutaneous lupus-like lesions, as this regimen also was reported to successfully treat gastric granulomatous involvement in a CGD patient.¹¹

Chronic granulomatous disease often is fatal in early childhood or adolescence due to sequelae from infections or chronic granulomatous infiltration of internal organs. Residual reactive oxygen intermediate production was shown to be a predictor of overall survival, and CGD patients with 1% of normal reactive oxygen intermediate production by neutrophils had a greater likelihood of survival.¹² In this regard, the otherwise good health of our patient at the time of presentation was consistent with his initial nitroblue tetrazolium test showing some residual oxidative activity, emphasizing the phenotypic variability of this rare genetic disorder and the importance of considering CGD in the diagnosis of seronegative cutaneous lupus-like reactions.

To the Editor:

Marjolin ulcers are malignancies arising in nonhealing cutaneous wounds. Although burn wounds are the most common type of cutaneous trauma associated with this entity, there are a multitude of possible lesions that may initiate this disease process including traumatic wounds, venous stasis ulcers, and vaccination sites.^1,2 The most common type of malignancy reported in a Marjolin ulcer is an aggressive squamous cell carcinoma (SCC).^1-3 Less commonly, basal cell carcinoma (BCC) also has been reported.^1,3,4 However, cases of BCCs developing in surgical scars are exceedingly rare. We describe a case of a morphoeic BCC in a long-standing surgical scar in a 50-year-old woman with Crohn disease.

A 50-year-old woman presented with an intermittent ulceration within a horizontal surgical scar on the right side of the upper abdomen of 2 years’ duration that had not healed over the course of the last 6 months. The scar was present from surgeries conducted while she was a teenager for complications associated with Crohn disease. She underwent her first abdominal surgery for a partial gastric resection at 16 years of age, followed by multiple laparotomies from a perforated bile duct that occurred during the first surgery. The original incision created for the partial gastric resection was used for all subsequent surgeries.

The patient’s medical history was notable for central nervous system vasculitis with vision loss, chronic pancreatitis, Crohn disease, arthritis, multiple superficial BCCs on the back that were successfully treated with imiquimod cream, a nodular BCC on the neck that was surgically removed, and facial actinic keratoses treated with liquid nitrogen. She had Fitzpatrick skin type I. She grew up in a residential area in Southern Ontario and did not have a history of heavy sun exposure. She did not receive notable radiation from treatment of Crohn disease, and she usually wore a 1-piece bathing suit when swimming outdoors. According to the patient, she had never been exposed to arsenic. The patient’s family history was negative for skin cancer and she was a nonsmoker. She was taking methotrexate, prednisone, folic acid, pentazocine, and vitamin B₁₂ injections at the time of presentation for the aforementioned conditions.

On physical examination a 1.5-cm honey-crusted ulcer with surrounding violaceous erythema in a long-standing surgical scar was observed (Figure 1). There was no palpable adenopathy in the inguinal or axillary regions. The suspected diagnosis prebiopsy was an SCC developing within the scar tissue. On histologic examination sections of small nests and strands of basal cell infiltrating thick collagen bundles were visualized. The appearance was consistent with a morphoeic BCC. The pathologist’s interpretation indicated that the lesion appeared to be a morphoeic BCC within the scar as opposed to a BCC that appeared morphoeic because of background scarring. Histologic images stained with hematoxylin and eosin showed small nests and strands of basal cells penetrating thick collagen bundles (Figure 2).

Figure 1. Complete visualization of the surgical scar with the associated ulcer in the epigastric region (A). Magnified view of a honey-crusted ulcer with surrounding violaceous erythema (B).

Figure 2. Histology showed small nests and strands of basal cells penetrating thick collagen bundles (A and B)(both H&E, original magnifications ×40 and ×100).

Marjolin ulcer was first described in 1828 by the French surgeon Jean-Nicolas Marjolin who published 4 cases of ulcers arising from scar tissue but did not appreciate their malignant capacity. However, the term Marjolin ulcer is now widely accepted as meaning any malignant tumor occurring within scar tissue or a chronic nonhealing wound.^1,2

The exact incidence of malignant transformation in cutaneous wounds remains unknown, but this phenomenon can occur in individuals of all races and across all age groups.^1,5-7 The most prevalent malignancy identified on biopsy is SCC, followed by BCC, melanoma, osteosarcoma, fibrosarcoma, and liposarcoma.^1,2,6,7 With this entity infrequently occurring in the clinical setting, it often is overlooked or misdiagnosed.² In addition, malignancies presenting as Marjolin ulcers have a greater tendency to metastasize and are reported to have a higher associated fatality rate.^2,8 Thus, early recognition is essential, as a delay in diagnosis can potentially allow the tumor to progress to a life-threatening stage. In our patient, malignancy was clinically suspected based on the presence of an ulcer that was not healing despite adequate wound care and the location in a scar that was present for more than 30 years. The surgical scar had been a place of repeated trauma given the number of surgical procedures and the perforated bile duct, which can increase the potential for malignant transformation. Furthermore, the patient also was on immunosuppressive therapy for an extended period of time, possibly contributing to the development of this cancerous lesion and prior cutaneous malignancies.

The pathogenesis of a Marjolin ulcer is unclear, though many hypotheses have been suggested.^1,2,6,9 Theories investigating decreased vascularity, lowered immune surveillance, decreased regenerative capacity, genetic mutations, and injury-related release of toxins have all been postulated as possible explanations for the increase in potential of malignant transformation.^1-3,6,9 However, despite the pathogenesis, the mainstay of treatment remains wide local excision with at least 2-cm margins.^1-3,10 Alternatively, Mohs micrographic surgery can be considered for Marjolin ulcers, but it is less frequently conducted in comparison to wide local excision. Radiation therapy often follows excision as adjuvant therapy, depending on the type of tumor.^2,10 Prophylactic lymph node dissection is not indicated in most cases, but regional node dissection is suggested when palpable lymphadenopathy is present.^1,2,10 Moreover, amputation is indicated with deep bone or joint involvement.^1-3,10 Recurrence rates are high, ranging from 20% to 50%, and metastases to the brain, liver, lungs, kidneys, and lymph nodes have been reported.^1,3 The prognosis of the cutaneous malignancy in this setting is not as favorable, and the 5-year survival rate is cited at approximately 60%.³ Overall prognosis depends on several factors including location, type of malignancy, immune status, progression of disease, and lymph node metastasis. Our patient’s presentation with a BCC instead of the more common SCC should carry a good overall prognosis, though she will need to be closely followed for recurrence after wide local excision.

This novel presentation of a morpheaform BCC in a surgical scar may serve as a reminder to consider this diagnosis and biopsy nonhealing ulcers within any type of chronic wound or scar.

To the Editor:

Marjolin ulcers are malignancies arising in nonhealing cutaneous wounds. Although burn wounds are the most common type of cutaneous trauma associated with this entity, there are a multitude of possible lesions that may initiate this disease process including traumatic wounds, venous stasis ulcers, and vaccination sites.^1,2 The most common type of malignancy reported in a Marjolin ulcer is an aggressive squamous cell carcinoma (SCC).^1-3 Less commonly, basal cell carcinoma (BCC) also has been reported.^1,3,4 However, cases of BCCs developing in surgical scars are exceedingly rare. We describe a case of a morphoeic BCC in a long-standing surgical scar in a 50-year-old woman with Crohn disease.

A 50-year-old woman presented with an intermittent ulceration within a horizontal surgical scar on the right side of the upper abdomen of 2 years’ duration that had not healed over the course of the last 6 months. The scar was present from surgeries conducted while she was a teenager for complications associated with Crohn disease. She underwent her first abdominal surgery for a partial gastric resection at 16 years of age, followed by multiple laparotomies from a perforated bile duct that occurred during the first surgery. The original incision created for the partial gastric resection was used for all subsequent surgeries.

The patient’s medical history was notable for central nervous system vasculitis with vision loss, chronic pancreatitis, Crohn disease, arthritis, multiple superficial BCCs on the back that were successfully treated with imiquimod cream, a nodular BCC on the neck that was surgically removed, and facial actinic keratoses treated with liquid nitrogen. She had Fitzpatrick skin type I. She grew up in a residential area in Southern Ontario and did not have a history of heavy sun exposure. She did not receive notable radiation from treatment of Crohn disease, and she usually wore a 1-piece bathing suit when swimming outdoors. According to the patient, she had never been exposed to arsenic. The patient’s family history was negative for skin cancer and she was a nonsmoker. She was taking methotrexate, prednisone, folic acid, pentazocine, and vitamin B₁₂ injections at the time of presentation for the aforementioned conditions.

On physical examination a 1.5-cm honey-crusted ulcer with surrounding violaceous erythema in a long-standing surgical scar was observed (Figure 1). There was no palpable adenopathy in the inguinal or axillary regions. The suspected diagnosis prebiopsy was an SCC developing within the scar tissue. On histologic examination sections of small nests and strands of basal cell infiltrating thick collagen bundles were visualized. The appearance was consistent with a morphoeic BCC. The pathologist’s interpretation indicated that the lesion appeared to be a morphoeic BCC within the scar as opposed to a BCC that appeared morphoeic because of background scarring. Histologic images stained with hematoxylin and eosin showed small nests and strands of basal cells penetrating thick collagen bundles (Figure 2).

Figure 1. Complete visualization of the surgical scar with the associated ulcer in the epigastric region (A). Magnified view of a honey-crusted ulcer with surrounding violaceous erythema (B).

Figure 2. Histology showed small nests and strands of basal cells penetrating thick collagen bundles (A and B)(both H&E, original magnifications ×40 and ×100).

Marjolin ulcer was first described in 1828 by the French surgeon Jean-Nicolas Marjolin who published 4 cases of ulcers arising from scar tissue but did not appreciate their malignant capacity. However, the term Marjolin ulcer is now widely accepted as meaning any malignant tumor occurring within scar tissue or a chronic nonhealing wound.^1,2

The exact incidence of malignant transformation in cutaneous wounds remains unknown, but this phenomenon can occur in individuals of all races and across all age groups.^1,5-7 The most prevalent malignancy identified on biopsy is SCC, followed by BCC, melanoma, osteosarcoma, fibrosarcoma, and liposarcoma.^1,2,6,7 With this entity infrequently occurring in the clinical setting, it often is overlooked or misdiagnosed.² In addition, malignancies presenting as Marjolin ulcers have a greater tendency to metastasize and are reported to have a higher associated fatality rate.^2,8 Thus, early recognition is essential, as a delay in diagnosis can potentially allow the tumor to progress to a life-threatening stage. In our patient, malignancy was clinically suspected based on the presence of an ulcer that was not healing despite adequate wound care and the location in a scar that was present for more than 30 years. The surgical scar had been a place of repeated trauma given the number of surgical procedures and the perforated bile duct, which can increase the potential for malignant transformation. Furthermore, the patient also was on immunosuppressive therapy for an extended period of time, possibly contributing to the development of this cancerous lesion and prior cutaneous malignancies.

The pathogenesis of a Marjolin ulcer is unclear, though many hypotheses have been suggested.^1,2,6,9 Theories investigating decreased vascularity, lowered immune surveillance, decreased regenerative capacity, genetic mutations, and injury-related release of toxins have all been postulated as possible explanations for the increase in potential of malignant transformation.^1-3,6,9 However, despite the pathogenesis, the mainstay of treatment remains wide local excision with at least 2-cm margins.^1-3,10 Alternatively, Mohs micrographic surgery can be considered for Marjolin ulcers, but it is less frequently conducted in comparison to wide local excision. Radiation therapy often follows excision as adjuvant therapy, depending on the type of tumor.^2,10 Prophylactic lymph node dissection is not indicated in most cases, but regional node dissection is suggested when palpable lymphadenopathy is present.^1,2,10 Moreover, amputation is indicated with deep bone or joint involvement.^1-3,10 Recurrence rates are high, ranging from 20% to 50%, and metastases to the brain, liver, lungs, kidneys, and lymph nodes have been reported.^1,3 The prognosis of the cutaneous malignancy in this setting is not as favorable, and the 5-year survival rate is cited at approximately 60%.³ Overall prognosis depends on several factors including location, type of malignancy, immune status, progression of disease, and lymph node metastasis. Our patient’s presentation with a BCC instead of the more common SCC should carry a good overall prognosis, though she will need to be closely followed for recurrence after wide local excision.

This novel presentation of a morpheaform BCC in a surgical scar may serve as a reminder to consider this diagnosis and biopsy nonhealing ulcers within any type of chronic wound or scar.

To the Editor:

Marjolin ulcers are malignancies arising in nonhealing cutaneous wounds. Although burn wounds are the most common type of cutaneous trauma associated with this entity, there are a multitude of possible lesions that may initiate this disease process including traumatic wounds, venous stasis ulcers, and vaccination sites.^1,2 The most common type of malignancy reported in a Marjolin ulcer is an aggressive squamous cell carcinoma (SCC).^1-3 Less commonly, basal cell carcinoma (BCC) also has been reported.^1,3,4 However, cases of BCCs developing in surgical scars are exceedingly rare. We describe a case of a morphoeic BCC in a long-standing surgical scar in a 50-year-old woman with Crohn disease.

A 50-year-old woman presented with an intermittent ulceration within a horizontal surgical scar on the right side of the upper abdomen of 2 years’ duration that had not healed over the course of the last 6 months. The scar was present from surgeries conducted while she was a teenager for complications associated with Crohn disease. She underwent her first abdominal surgery for a partial gastric resection at 16 years of age, followed by multiple laparotomies from a perforated bile duct that occurred during the first surgery. The original incision created for the partial gastric resection was used for all subsequent surgeries.

The patient’s medical history was notable for central nervous system vasculitis with vision loss, chronic pancreatitis, Crohn disease, arthritis, multiple superficial BCCs on the back that were successfully treated with imiquimod cream, a nodular BCC on the neck that was surgically removed, and facial actinic keratoses treated with liquid nitrogen. She had Fitzpatrick skin type I. She grew up in a residential area in Southern Ontario and did not have a history of heavy sun exposure. She did not receive notable radiation from treatment of Crohn disease, and she usually wore a 1-piece bathing suit when swimming outdoors. According to the patient, she had never been exposed to arsenic. The patient’s family history was negative for skin cancer and she was a nonsmoker. She was taking methotrexate, prednisone, folic acid, pentazocine, and vitamin B₁₂ injections at the time of presentation for the aforementioned conditions.

On physical examination a 1.5-cm honey-crusted ulcer with surrounding violaceous erythema in a long-standing surgical scar was observed (Figure 1). There was no palpable adenopathy in the inguinal or axillary regions. The suspected diagnosis prebiopsy was an SCC developing within the scar tissue. On histologic examination sections of small nests and strands of basal cell infiltrating thick collagen bundles were visualized. The appearance was consistent with a morphoeic BCC. The pathologist’s interpretation indicated that the lesion appeared to be a morphoeic BCC within the scar as opposed to a BCC that appeared morphoeic because of background scarring. Histologic images stained with hematoxylin and eosin showed small nests and strands of basal cells penetrating thick collagen bundles (Figure 2).

Figure 1. Complete visualization of the surgical scar with the associated ulcer in the epigastric region (A). Magnified view of a honey-crusted ulcer with surrounding violaceous erythema (B).

Figure 2. Histology showed small nests and strands of basal cells penetrating thick collagen bundles (A and B)(both H&E, original magnifications ×40 and ×100).

Marjolin ulcer was first described in 1828 by the French surgeon Jean-Nicolas Marjolin who published 4 cases of ulcers arising from scar tissue but did not appreciate their malignant capacity. However, the term Marjolin ulcer is now widely accepted as meaning any malignant tumor occurring within scar tissue or a chronic nonhealing wound.^1,2

The exact incidence of malignant transformation in cutaneous wounds remains unknown, but this phenomenon can occur in individuals of all races and across all age groups.^1,5-7 The most prevalent malignancy identified on biopsy is SCC, followed by BCC, melanoma, osteosarcoma, fibrosarcoma, and liposarcoma.^1,2,6,7 With this entity infrequently occurring in the clinical setting, it often is overlooked or misdiagnosed.² In addition, malignancies presenting as Marjolin ulcers have a greater tendency to metastasize and are reported to have a higher associated fatality rate.^2,8 Thus, early recognition is essential, as a delay in diagnosis can potentially allow the tumor to progress to a life-threatening stage. In our patient, malignancy was clinically suspected based on the presence of an ulcer that was not healing despite adequate wound care and the location in a scar that was present for more than 30 years. The surgical scar had been a place of repeated trauma given the number of surgical procedures and the perforated bile duct, which can increase the potential for malignant transformation. Furthermore, the patient also was on immunosuppressive therapy for an extended period of time, possibly contributing to the development of this cancerous lesion and prior cutaneous malignancies.

The pathogenesis of a Marjolin ulcer is unclear, though many hypotheses have been suggested.^1,2,6,9 Theories investigating decreased vascularity, lowered immune surveillance, decreased regenerative capacity, genetic mutations, and injury-related release of toxins have all been postulated as possible explanations for the increase in potential of malignant transformation.^1-3,6,9 However, despite the pathogenesis, the mainstay of treatment remains wide local excision with at least 2-cm margins.^1-3,10 Alternatively, Mohs micrographic surgery can be considered for Marjolin ulcers, but it is less frequently conducted in comparison to wide local excision. Radiation therapy often follows excision as adjuvant therapy, depending on the type of tumor.^2,10 Prophylactic lymph node dissection is not indicated in most cases, but regional node dissection is suggested when palpable lymphadenopathy is present.^1,2,10 Moreover, amputation is indicated with deep bone or joint involvement.^1-3,10 Recurrence rates are high, ranging from 20% to 50%, and metastases to the brain, liver, lungs, kidneys, and lymph nodes have been reported.^1,3 The prognosis of the cutaneous malignancy in this setting is not as favorable, and the 5-year survival rate is cited at approximately 60%.³ Overall prognosis depends on several factors including location, type of malignancy, immune status, progression of disease, and lymph node metastasis. Our patient’s presentation with a BCC instead of the more common SCC should carry a good overall prognosis, though she will need to be closely followed for recurrence after wide local excision.

This novel presentation of a morpheaform BCC in a surgical scar may serve as a reminder to consider this diagnosis and biopsy nonhealing ulcers within any type of chronic wound or scar.