Cutis

The Diagnosis: Nonuremic Calciphylaxis

Histopathologic findings revealed ischemic necrosis and a subepidermal blister (Figure 1) with arteriosclerotic changes and fat necrosis. Foci of calcification were noted within the fat lobules. Arterioles within the deeper dermis and subcutis showed thickened hyalinized walls, narrowed lumina, and medial calcification (Figure 2). Multiple sections did not reveal any granulomatous inflammation. Periodic acid-Schiff and Gram stains were negative for fungal and bacterial elements, respectively. No dense neutrophilic infiltrate was seen. Multifocal calcific deposits within fat lobules and vessel walls (endothelium highlighted by the CD31 stain) suggested calciphylaxis.

Laboratory test results revealed a normal white blood cell count, international normalized ratio level of 4 (on warfarin), and an elevated sedimentation rate at 72 mm/h (reference range, 0-20 mm/h). Serum creatinine was 1.1 mg/dL (reference range, 0.6-1.2 mg/dL) and the calcium-phosphorous product was 40.8 mg²/dL (reference range, <55 mg²/dL). Hemoglobin A_1C (glycated hemoglobin) was 8.2% (reference range, 4%-7%). Wound cultures grew Proteus mirabilis sensitive to cefazolin. Acid-fast bacilli and fungal cultures were negative. Computed tomography of the left lower leg without contrast showed no evidence of osteomyelitis. Of note, the popliteal arteries and distal vessels showed moderate vascular calcification.

Histopathology findings as well as a clinical picture of painful ulceration on the distal extremities and uncontrolled diabetes with normal renal function favored a diagnosis of nonuremic calciphylaxis (NUC). The patient was treated with intravenous infusions of sodium thiosulfate 25 mg 3 times weekly and oral cefazolin for superadded bacterial infection. Local wound care included collagenase dressings with light compression. Warfarin was discontinued, as it can worsen calciphylaxis. Complete reepithelialization of the ulcer along with substantial reduction in pain was noted within 4 weeks.

Ulceration of the lower legs is a relatively common condition in the Western world, the prevalence of which increases up to 5% in patients older than 65 years.¹ Of the myriad of causes that lead to ulceration of the distal aspect of the leg, NUC is a rare but known phenomenon. The pathogenesis of NUC is complicated based on theories of derangement of receptor activator of nuclear factor κβ, receptor activator of nuclear factor κβ ligand, and osteoprotegerin, leading to calcium deposits in the media of the arteries.² This deposition precipitates vascular occlusion coupled with ischemic necrosis of the subcutaneous tissue and skin.³ Some of the more common causes of NUC are primary hyperparathyroidism, malignancy, and rheumatoid arthritis. Type 2 diabetes mellitus is a less common cause but often is found in association with NUC, as noted by Nigwekar et al.² According to their study, the laboratory parameters commonly found in NUC included a calcium-phosphorous product greater than 50 mg²/dL and serum creatinine of 1.2 mg/dL or less.²

Our patient displayed these laboratory findings. However, distinguishing NUC from other atypical lower extremity ulcers such as Martorell hypertensive ischemic ulcer, pyoderma gangrenosum, and warfarin necrosis can pose a challenge to the dermatologist. Martorell hypertensive ischemic ulcer is excruciatingly painful and occurs more frequently near the Achilles tendon, responding well to surgical debridement. Histopathologically, medial calcinosis and arteriosclerosis are seen.⁴

Pyoderma gangrenosum is a neutrophilic dermatosis wherein the classical ulcerative variant is painful. It occurs mostly on the pretibial area and worsens after debridement.⁵ Clinically and histopathologically, it is a diagnosis of exclusion in which a dense neutrophilic to mixed lymphocytic infiltrate is seen with necrosis of dermal vessels.⁶ 

Warfarin necrosis is extremely rare, affecting 0.01% to 0.1% of patients on warfarin-derived anticoagulant therapy.⁷ Necrosis occurs mostly on fat-bearing areas such as the breasts, abdomen, and thighs 3 to 5 days after initiating treatment. Histologically, fibrin deposits occlude dermal vessels without perivascular inflammation.⁸

Necrobiosis lipoidica is a rare cutaneous entity seen in 0.3% of diabetic patients.⁹ The exact pathogenesis is unknown; however, microangiopathy in collaboration with cross-linking of abnormal collagen fibers play a role. These lesions appear as erythematous plaques with a slightly depressed to atrophic center, ultimately taking on a waxy porcelain appearance. Although most of these lesions either resolve or become chronically persistent, approximately 15% undergo ulceration, which can be painful. Histologically, with hematoxylin and eosin staining, areas of necrobiosis are seen surrounded by an inflammatory infiltrate comprised mainly of histiocytes along with lymphocytes and plasma cells.⁹

Nonuremic calciphylaxis can mimic the aforementioned conditions to a greater extent in female patients with obesity, diabetes mellitus, and hypertension. However, microscopic calcium deposition in the media of dermal arterioles, extravascular calcification within fat lobules, and cutaneous necrosis, along with remarkable response to intravenous sodium thiosulfate, confirmed a diagnosis of NUC in our patient. Sodium thiosulfate scavenges reactive oxygen species and promotes nitric oxygen generation, thereby reducing endothelial damage.¹⁰ Although there are no randomized controlled trials to support its use, sodium thiosulfate has been successfully used to treat established cases of NUC.¹¹

The Diagnosis: Nonuremic Calciphylaxis

Histopathologic findings revealed ischemic necrosis and a subepidermal blister (Figure 1) with arteriosclerotic changes and fat necrosis. Foci of calcification were noted within the fat lobules. Arterioles within the deeper dermis and subcutis showed thickened hyalinized walls, narrowed lumina, and medial calcification (Figure 2). Multiple sections did not reveal any granulomatous inflammation. Periodic acid-Schiff and Gram stains were negative for fungal and bacterial elements, respectively. No dense neutrophilic infiltrate was seen. Multifocal calcific deposits within fat lobules and vessel walls (endothelium highlighted by the CD31 stain) suggested calciphylaxis.

Laboratory test results revealed a normal white blood cell count, international normalized ratio level of 4 (on warfarin), and an elevated sedimentation rate at 72 mm/h (reference range, 0-20 mm/h). Serum creatinine was 1.1 mg/dL (reference range, 0.6-1.2 mg/dL) and the calcium-phosphorous product was 40.8 mg²/dL (reference range, <55 mg²/dL). Hemoglobin A_1C (glycated hemoglobin) was 8.2% (reference range, 4%-7%). Wound cultures grew Proteus mirabilis sensitive to cefazolin. Acid-fast bacilli and fungal cultures were negative. Computed tomography of the left lower leg without contrast showed no evidence of osteomyelitis. Of note, the popliteal arteries and distal vessels showed moderate vascular calcification.

Histopathology findings as well as a clinical picture of painful ulceration on the distal extremities and uncontrolled diabetes with normal renal function favored a diagnosis of nonuremic calciphylaxis (NUC). The patient was treated with intravenous infusions of sodium thiosulfate 25 mg 3 times weekly and oral cefazolin for superadded bacterial infection. Local wound care included collagenase dressings with light compression. Warfarin was discontinued, as it can worsen calciphylaxis. Complete reepithelialization of the ulcer along with substantial reduction in pain was noted within 4 weeks.

Ulceration of the lower legs is a relatively common condition in the Western world, the prevalence of which increases up to 5% in patients older than 65 years.¹ Of the myriad of causes that lead to ulceration of the distal aspect of the leg, NUC is a rare but known phenomenon. The pathogenesis of NUC is complicated based on theories of derangement of receptor activator of nuclear factor κβ, receptor activator of nuclear factor κβ ligand, and osteoprotegerin, leading to calcium deposits in the media of the arteries.² This deposition precipitates vascular occlusion coupled with ischemic necrosis of the subcutaneous tissue and skin.³ Some of the more common causes of NUC are primary hyperparathyroidism, malignancy, and rheumatoid arthritis. Type 2 diabetes mellitus is a less common cause but often is found in association with NUC, as noted by Nigwekar et al.² According to their study, the laboratory parameters commonly found in NUC included a calcium-phosphorous product greater than 50 mg²/dL and serum creatinine of 1.2 mg/dL or less.²

Our patient displayed these laboratory findings. However, distinguishing NUC from other atypical lower extremity ulcers such as Martorell hypertensive ischemic ulcer, pyoderma gangrenosum, and warfarin necrosis can pose a challenge to the dermatologist. Martorell hypertensive ischemic ulcer is excruciatingly painful and occurs more frequently near the Achilles tendon, responding well to surgical debridement. Histopathologically, medial calcinosis and arteriosclerosis are seen.⁴

Pyoderma gangrenosum is a neutrophilic dermatosis wherein the classical ulcerative variant is painful. It occurs mostly on the pretibial area and worsens after debridement.⁵ Clinically and histopathologically, it is a diagnosis of exclusion in which a dense neutrophilic to mixed lymphocytic infiltrate is seen with necrosis of dermal vessels.⁶ 

Warfarin necrosis is extremely rare, affecting 0.01% to 0.1% of patients on warfarin-derived anticoagulant therapy.⁷ Necrosis occurs mostly on fat-bearing areas such as the breasts, abdomen, and thighs 3 to 5 days after initiating treatment. Histologically, fibrin deposits occlude dermal vessels without perivascular inflammation.⁸

Necrobiosis lipoidica is a rare cutaneous entity seen in 0.3% of diabetic patients.⁹ The exact pathogenesis is unknown; however, microangiopathy in collaboration with cross-linking of abnormal collagen fibers play a role. These lesions appear as erythematous plaques with a slightly depressed to atrophic center, ultimately taking on a waxy porcelain appearance. Although most of these lesions either resolve or become chronically persistent, approximately 15% undergo ulceration, which can be painful. Histologically, with hematoxylin and eosin staining, areas of necrobiosis are seen surrounded by an inflammatory infiltrate comprised mainly of histiocytes along with lymphocytes and plasma cells.⁹

Nonuremic calciphylaxis can mimic the aforementioned conditions to a greater extent in female patients with obesity, diabetes mellitus, and hypertension. However, microscopic calcium deposition in the media of dermal arterioles, extravascular calcification within fat lobules, and cutaneous necrosis, along with remarkable response to intravenous sodium thiosulfate, confirmed a diagnosis of NUC in our patient. Sodium thiosulfate scavenges reactive oxygen species and promotes nitric oxygen generation, thereby reducing endothelial damage.¹⁰ Although there are no randomized controlled trials to support its use, sodium thiosulfate has been successfully used to treat established cases of NUC.¹¹

The Diagnosis: Nonuremic Calciphylaxis

Histopathologic findings revealed ischemic necrosis and a subepidermal blister (Figure 1) with arteriosclerotic changes and fat necrosis. Foci of calcification were noted within the fat lobules. Arterioles within the deeper dermis and subcutis showed thickened hyalinized walls, narrowed lumina, and medial calcification (Figure 2). Multiple sections did not reveal any granulomatous inflammation. Periodic acid-Schiff and Gram stains were negative for fungal and bacterial elements, respectively. No dense neutrophilic infiltrate was seen. Multifocal calcific deposits within fat lobules and vessel walls (endothelium highlighted by the CD31 stain) suggested calciphylaxis.

Laboratory test results revealed a normal white blood cell count, international normalized ratio level of 4 (on warfarin), and an elevated sedimentation rate at 72 mm/h (reference range, 0-20 mm/h). Serum creatinine was 1.1 mg/dL (reference range, 0.6-1.2 mg/dL) and the calcium-phosphorous product was 40.8 mg²/dL (reference range, <55 mg²/dL). Hemoglobin A_1C (glycated hemoglobin) was 8.2% (reference range, 4%-7%). Wound cultures grew Proteus mirabilis sensitive to cefazolin. Acid-fast bacilli and fungal cultures were negative. Computed tomography of the left lower leg without contrast showed no evidence of osteomyelitis. Of note, the popliteal arteries and distal vessels showed moderate vascular calcification.

Histopathology findings as well as a clinical picture of painful ulceration on the distal extremities and uncontrolled diabetes with normal renal function favored a diagnosis of nonuremic calciphylaxis (NUC). The patient was treated with intravenous infusions of sodium thiosulfate 25 mg 3 times weekly and oral cefazolin for superadded bacterial infection. Local wound care included collagenase dressings with light compression. Warfarin was discontinued, as it can worsen calciphylaxis. Complete reepithelialization of the ulcer along with substantial reduction in pain was noted within 4 weeks.

Ulceration of the lower legs is a relatively common condition in the Western world, the prevalence of which increases up to 5% in patients older than 65 years.¹ Of the myriad of causes that lead to ulceration of the distal aspect of the leg, NUC is a rare but known phenomenon. The pathogenesis of NUC is complicated based on theories of derangement of receptor activator of nuclear factor κβ, receptor activator of nuclear factor κβ ligand, and osteoprotegerin, leading to calcium deposits in the media of the arteries.² This deposition precipitates vascular occlusion coupled with ischemic necrosis of the subcutaneous tissue and skin.³ Some of the more common causes of NUC are primary hyperparathyroidism, malignancy, and rheumatoid arthritis. Type 2 diabetes mellitus is a less common cause but often is found in association with NUC, as noted by Nigwekar et al.² According to their study, the laboratory parameters commonly found in NUC included a calcium-phosphorous product greater than 50 mg²/dL and serum creatinine of 1.2 mg/dL or less.²

Our patient displayed these laboratory findings. However, distinguishing NUC from other atypical lower extremity ulcers such as Martorell hypertensive ischemic ulcer, pyoderma gangrenosum, and warfarin necrosis can pose a challenge to the dermatologist. Martorell hypertensive ischemic ulcer is excruciatingly painful and occurs more frequently near the Achilles tendon, responding well to surgical debridement. Histopathologically, medial calcinosis and arteriosclerosis are seen.⁴

Pyoderma gangrenosum is a neutrophilic dermatosis wherein the classical ulcerative variant is painful. It occurs mostly on the pretibial area and worsens after debridement.⁵ Clinically and histopathologically, it is a diagnosis of exclusion in which a dense neutrophilic to mixed lymphocytic infiltrate is seen with necrosis of dermal vessels.⁶ 

Warfarin necrosis is extremely rare, affecting 0.01% to 0.1% of patients on warfarin-derived anticoagulant therapy.⁷ Necrosis occurs mostly on fat-bearing areas such as the breasts, abdomen, and thighs 3 to 5 days after initiating treatment. Histologically, fibrin deposits occlude dermal vessels without perivascular inflammation.⁸

Necrobiosis lipoidica is a rare cutaneous entity seen in 0.3% of diabetic patients.⁹ The exact pathogenesis is unknown; however, microangiopathy in collaboration with cross-linking of abnormal collagen fibers play a role. These lesions appear as erythematous plaques with a slightly depressed to atrophic center, ultimately taking on a waxy porcelain appearance. Although most of these lesions either resolve or become chronically persistent, approximately 15% undergo ulceration, which can be painful. Histologically, with hematoxylin and eosin staining, areas of necrobiosis are seen surrounded by an inflammatory infiltrate comprised mainly of histiocytes along with lymphocytes and plasma cells.⁹

Nonuremic calciphylaxis can mimic the aforementioned conditions to a greater extent in female patients with obesity, diabetes mellitus, and hypertension. However, microscopic calcium deposition in the media of dermal arterioles, extravascular calcification within fat lobules, and cutaneous necrosis, along with remarkable response to intravenous sodium thiosulfate, confirmed a diagnosis of NUC in our patient. Sodium thiosulfate scavenges reactive oxygen species and promotes nitric oxygen generation, thereby reducing endothelial damage.¹⁰ Although there are no randomized controlled trials to support its use, sodium thiosulfate has been successfully used to treat established cases of NUC.¹¹

Imipramine is a tricyclic medication uncommonly used to treat depression, anxiety, and other psychiatric illnesses. Although relatively rare, it has been associated with hyperpigmentation of the skin including slate gray discoloration of sun-exposed areas.

We present the case of a 63-year-old woman who had been taking imipramine for more than 20 years when she developed bluish gray discoloration on the face and neck. Histopathology of biopsy specimens showed numerous perivascular and interstitial brown globules in the dermis that were composed of melanin only, as evidenced by positive Fontana-Masson staining and negative Perls Prussian blue staining. A diagnosis of imipramine-induced hyperpigmentation was made based on histopathology and clinical history.

In addition to the case presentation, we provide a review of drugs that commonly cause hyperpigmentation as well as their associated histopathologic staining characteristics.

Case Report

A 63-year-old woman presented with blue-gray discoloration on the face and neck. She first noted the discoloration on the left side of the forehead 3 years prior; it then spread to the right side of the forehead, cheeks, and neck. She denied pruritus, pain, redness, and scaling of the involved areas; any recent changes in medications; or the use of any topical products on the affected areas. Her medical history was remarkable for hypertension, which was inconsistently controlled with lisinopril and hydrochlorothiazide, and depression, which had been managed with oral imipramine.

Physical examination disclosed blue-gray hyperpigmented patches with irregular borders on the bilateral forehead, temples, and periorbital skin (Figure 1). Reticulated brown patches were noted on the bilateral cheeks, and the neck displayed diffuse muddy brown patches with sparing of the submental areas.

Punch biopsies obtained from the lateral forehead showed an unremarkable epidermis with deposition of numerous golden brown granules in the upper and mid dermis and in perivascular macrophages (Figure 2). The pigmented granules showed positive staining with Fontana-Masson (Figure 3), and a Perls Prussian blue stain for hemosiderin was negative. Based on the clinical history, a diagnosis of imipramine-induced hyperpigmentation was made.

The patient revealed that she had taken imipramine for more than 20 years for depression as prescribed by her mental health professional. She had tried several other antidepressants but none were as effective as imipramine. Therefore, she was not willing to discontinue it despite the likelihood that the hyperpigmentation would persist and could worsen with continued use of the medication. Diligent photoprotection was advised. Additionally, she started taking lisinopril some time after the appearance of the hyperpigmentation presented and had not taken hydrochlorothiazide consistently for several years. Although these drugs are known to cause various cutaneous reactions, it was not considered likely in this case.

Comment

Drug-induced hyperpigmentation accounts for 10% to 20% of all cases of acquired hyperpigmentation.¹ Common causative drugs include amiodarone, antimalarials, minocycline, and rarely psychotropics including phenothiazines and tricyclic antidepressants such as imipramine.^1-4 Although amiodarone-induced hyperpigmentation is associated with lipofuscin in addition to melanin, most other medications, including imipramine, induce cutaneous effects through deposition of melanin and/or hemosiderin. A review of the histopathologic staining characteristics in pigment anomalies caused by these drugs is summarized in the Table.

Imipramine-induced hyperpigmentation presents as slate gray discrete macules and patches on sun-exposed skin that may appear anywhere from 2 to 22 years after initiating the medication.^1-4 Affected areas include the malar cheeks, temples, periorbital areas, hands, forearms, and seldom the iris and sclera.^2-4 Although the blue to slate gray coloring is classic, other colors have been described including brown, golden brown, and purple.²

Histopathology of imipramine-induced hyperpigmentation shows golden brown, round to oval granules in the superficial dermis and within dermal macrophages.^1,3 Generally, Fontana-Masson staining is positive for melanin and Perls Prussian blue staining is negative for iron.^1,2,4

Imipramine-induced hyperpigmentation likely results from photoexcitation of imipramine or one of its metabolites. These compounds activate tyrosinase, increasing melanogenesis and leading to formation of melanin-imipramine or melanin-metabolite complexes.^1-3 Complexes are deposited in the dermis and basal layer or are engulfed by dermal macrophages and darkened on sun exposure due to their high melanin content.¹ Other possible mechanisms of hyperpigmentation include nonspecific inflammation caused by the drug in the skin, hemosiderin deposition from vessel damage and subsequent erythrocyte extravasation, or deposition of newly formed pigments related to the drug.¹

Most patients report satisfactory resolution of imipramine-induced discoloration within 1 year of stopping imipramine or switching to a different antidepressant.^1,4 Patients who are unwilling to discontinue imipramine may achieve resolution with alexandrite or Q-switched ruby laser therapy.^1,4 Strict sun protective measures are necessary, both to prevent new deposition of melanin and to prevent darkening of existing pigment.

Despite the advent of new psychotropic medications, imipramine remains the antidepressant of choice for many patients. Although rare, it is important to be able to recognize imipramine-induced hyperpigmentation and to encourage patient-psychiatrist communication to determine an antidepressant regimen that avoids unnecessary cutaneous side effects.

Imipramine is a tricyclic medication uncommonly used to treat depression, anxiety, and other psychiatric illnesses. Although relatively rare, it has been associated with hyperpigmentation of the skin including slate gray discoloration of sun-exposed areas.

We present the case of a 63-year-old woman who had been taking imipramine for more than 20 years when she developed bluish gray discoloration on the face and neck. Histopathology of biopsy specimens showed numerous perivascular and interstitial brown globules in the dermis that were composed of melanin only, as evidenced by positive Fontana-Masson staining and negative Perls Prussian blue staining. A diagnosis of imipramine-induced hyperpigmentation was made based on histopathology and clinical history.

In addition to the case presentation, we provide a review of drugs that commonly cause hyperpigmentation as well as their associated histopathologic staining characteristics.

Case Report

A 63-year-old woman presented with blue-gray discoloration on the face and neck. She first noted the discoloration on the left side of the forehead 3 years prior; it then spread to the right side of the forehead, cheeks, and neck. She denied pruritus, pain, redness, and scaling of the involved areas; any recent changes in medications; or the use of any topical products on the affected areas. Her medical history was remarkable for hypertension, which was inconsistently controlled with lisinopril and hydrochlorothiazide, and depression, which had been managed with oral imipramine.

Physical examination disclosed blue-gray hyperpigmented patches with irregular borders on the bilateral forehead, temples, and periorbital skin (Figure 1). Reticulated brown patches were noted on the bilateral cheeks, and the neck displayed diffuse muddy brown patches with sparing of the submental areas.

Punch biopsies obtained from the lateral forehead showed an unremarkable epidermis with deposition of numerous golden brown granules in the upper and mid dermis and in perivascular macrophages (Figure 2). The pigmented granules showed positive staining with Fontana-Masson (Figure 3), and a Perls Prussian blue stain for hemosiderin was negative. Based on the clinical history, a diagnosis of imipramine-induced hyperpigmentation was made.

The patient revealed that she had taken imipramine for more than 20 years for depression as prescribed by her mental health professional. She had tried several other antidepressants but none were as effective as imipramine. Therefore, she was not willing to discontinue it despite the likelihood that the hyperpigmentation would persist and could worsen with continued use of the medication. Diligent photoprotection was advised. Additionally, she started taking lisinopril some time after the appearance of the hyperpigmentation presented and had not taken hydrochlorothiazide consistently for several years. Although these drugs are known to cause various cutaneous reactions, it was not considered likely in this case.

Comment

Drug-induced hyperpigmentation accounts for 10% to 20% of all cases of acquired hyperpigmentation.¹ Common causative drugs include amiodarone, antimalarials, minocycline, and rarely psychotropics including phenothiazines and tricyclic antidepressants such as imipramine.^1-4 Although amiodarone-induced hyperpigmentation is associated with lipofuscin in addition to melanin, most other medications, including imipramine, induce cutaneous effects through deposition of melanin and/or hemosiderin. A review of the histopathologic staining characteristics in pigment anomalies caused by these drugs is summarized in the Table.

Imipramine-induced hyperpigmentation presents as slate gray discrete macules and patches on sun-exposed skin that may appear anywhere from 2 to 22 years after initiating the medication.^1-4 Affected areas include the malar cheeks, temples, periorbital areas, hands, forearms, and seldom the iris and sclera.^2-4 Although the blue to slate gray coloring is classic, other colors have been described including brown, golden brown, and purple.²

Histopathology of imipramine-induced hyperpigmentation shows golden brown, round to oval granules in the superficial dermis and within dermal macrophages.^1,3 Generally, Fontana-Masson staining is positive for melanin and Perls Prussian blue staining is negative for iron.^1,2,4

Imipramine-induced hyperpigmentation likely results from photoexcitation of imipramine or one of its metabolites. These compounds activate tyrosinase, increasing melanogenesis and leading to formation of melanin-imipramine or melanin-metabolite complexes.^1-3 Complexes are deposited in the dermis and basal layer or are engulfed by dermal macrophages and darkened on sun exposure due to their high melanin content.¹ Other possible mechanisms of hyperpigmentation include nonspecific inflammation caused by the drug in the skin, hemosiderin deposition from vessel damage and subsequent erythrocyte extravasation, or deposition of newly formed pigments related to the drug.¹

Most patients report satisfactory resolution of imipramine-induced discoloration within 1 year of stopping imipramine or switching to a different antidepressant.^1,4 Patients who are unwilling to discontinue imipramine may achieve resolution with alexandrite or Q-switched ruby laser therapy.^1,4 Strict sun protective measures are necessary, both to prevent new deposition of melanin and to prevent darkening of existing pigment.

Despite the advent of new psychotropic medications, imipramine remains the antidepressant of choice for many patients. Although rare, it is important to be able to recognize imipramine-induced hyperpigmentation and to encourage patient-psychiatrist communication to determine an antidepressant regimen that avoids unnecessary cutaneous side effects.

Imipramine is a tricyclic medication uncommonly used to treat depression, anxiety, and other psychiatric illnesses. Although relatively rare, it has been associated with hyperpigmentation of the skin including slate gray discoloration of sun-exposed areas.

We present the case of a 63-year-old woman who had been taking imipramine for more than 20 years when she developed bluish gray discoloration on the face and neck. Histopathology of biopsy specimens showed numerous perivascular and interstitial brown globules in the dermis that were composed of melanin only, as evidenced by positive Fontana-Masson staining and negative Perls Prussian blue staining. A diagnosis of imipramine-induced hyperpigmentation was made based on histopathology and clinical history.

In addition to the case presentation, we provide a review of drugs that commonly cause hyperpigmentation as well as their associated histopathologic staining characteristics.

Case Report

A 63-year-old woman presented with blue-gray discoloration on the face and neck. She first noted the discoloration on the left side of the forehead 3 years prior; it then spread to the right side of the forehead, cheeks, and neck. She denied pruritus, pain, redness, and scaling of the involved areas; any recent changes in medications; or the use of any topical products on the affected areas. Her medical history was remarkable for hypertension, which was inconsistently controlled with lisinopril and hydrochlorothiazide, and depression, which had been managed with oral imipramine.

Physical examination disclosed blue-gray hyperpigmented patches with irregular borders on the bilateral forehead, temples, and periorbital skin (Figure 1). Reticulated brown patches were noted on the bilateral cheeks, and the neck displayed diffuse muddy brown patches with sparing of the submental areas.

Punch biopsies obtained from the lateral forehead showed an unremarkable epidermis with deposition of numerous golden brown granules in the upper and mid dermis and in perivascular macrophages (Figure 2). The pigmented granules showed positive staining with Fontana-Masson (Figure 3), and a Perls Prussian blue stain for hemosiderin was negative. Based on the clinical history, a diagnosis of imipramine-induced hyperpigmentation was made.

The patient revealed that she had taken imipramine for more than 20 years for depression as prescribed by her mental health professional. She had tried several other antidepressants but none were as effective as imipramine. Therefore, she was not willing to discontinue it despite the likelihood that the hyperpigmentation would persist and could worsen with continued use of the medication. Diligent photoprotection was advised. Additionally, she started taking lisinopril some time after the appearance of the hyperpigmentation presented and had not taken hydrochlorothiazide consistently for several years. Although these drugs are known to cause various cutaneous reactions, it was not considered likely in this case.

Comment

Drug-induced hyperpigmentation accounts for 10% to 20% of all cases of acquired hyperpigmentation.¹ Common causative drugs include amiodarone, antimalarials, minocycline, and rarely psychotropics including phenothiazines and tricyclic antidepressants such as imipramine.^1-4 Although amiodarone-induced hyperpigmentation is associated with lipofuscin in addition to melanin, most other medications, including imipramine, induce cutaneous effects through deposition of melanin and/or hemosiderin. A review of the histopathologic staining characteristics in pigment anomalies caused by these drugs is summarized in the Table.

Imipramine-induced hyperpigmentation presents as slate gray discrete macules and patches on sun-exposed skin that may appear anywhere from 2 to 22 years after initiating the medication.^1-4 Affected areas include the malar cheeks, temples, periorbital areas, hands, forearms, and seldom the iris and sclera.^2-4 Although the blue to slate gray coloring is classic, other colors have been described including brown, golden brown, and purple.²

Histopathology of imipramine-induced hyperpigmentation shows golden brown, round to oval granules in the superficial dermis and within dermal macrophages.^1,3 Generally, Fontana-Masson staining is positive for melanin and Perls Prussian blue staining is negative for iron.^1,2,4

Imipramine-induced hyperpigmentation likely results from photoexcitation of imipramine or one of its metabolites. These compounds activate tyrosinase, increasing melanogenesis and leading to formation of melanin-imipramine or melanin-metabolite complexes.^1-3 Complexes are deposited in the dermis and basal layer or are engulfed by dermal macrophages and darkened on sun exposure due to their high melanin content.¹ Other possible mechanisms of hyperpigmentation include nonspecific inflammation caused by the drug in the skin, hemosiderin deposition from vessel damage and subsequent erythrocyte extravasation, or deposition of newly formed pigments related to the drug.¹

Most patients report satisfactory resolution of imipramine-induced discoloration within 1 year of stopping imipramine or switching to a different antidepressant.^1,4 Patients who are unwilling to discontinue imipramine may achieve resolution with alexandrite or Q-switched ruby laser therapy.^1,4 Strict sun protective measures are necessary, both to prevent new deposition of melanin and to prevent darkening of existing pigment.

Despite the advent of new psychotropic medications, imipramine remains the antidepressant of choice for many patients. Although rare, it is important to be able to recognize imipramine-induced hyperpigmentation and to encourage patient-psychiatrist communication to determine an antidepressant regimen that avoids unnecessary cutaneous side effects.

To the Editor:

A 63-year-old man presented with a prior diagnosis of severe psoriasis affecting the extremities, neck, face, and scalp of 1 year’s duration. He reported pain, itching, and swelling in the affected areas. He felt the rash was worst on the hands and feet, and pain made performing activities of daily living difficult. His treatment regimen at presentation included triamcinolone cream 0.1% and azathioprine 150 mg daily as prescribed by an outside dermatologist without any response. Physical examination revealed diffuse erythema with lichenification and thick, white, flaking scale on the arms and legs (Figure 1A), face, neck, palms, and soles with islands of sparing. Multiple salmon-colored, follicular-based papules topped with central hyperkeratosis were scattered on these same areas. The palms and soles had severe confluent keratoderma (Figure 2A). Histologic examination of a follicular-based papule showed foci of parakeratosis and hypergranulosis consistent with the patient’s clinical picture of pityriasis rubra pilaris (PRP).

Baseline laboratory tests at the time of PRP diagnosis revealed 20.8% eosinophils (reference range, 0%–7%) and an absolute eosinophil count of 2.17×10⁹/L (reference range, 0–0.7×10⁹/L). Laboratory test results from an outside dermatologist conducted 10 to 12 months prior to the current presentation showed 12% eosinophils with a white blood cell count of 8.9×10⁹/L (reference range, 4.5–11.0×10⁹/L) around the time of rash onset and before treatment with azathioprine, making a drug reaction an unlikely cause of the eosinophilia.

After consulting with the hematology department, a hypereosinophilia workup including erythrocyte sedimentation rate, lactate dehydrogenase, serum protein electrophoresis, urine protein electrophoresis, tryptase, double-stranded DNA antibody, human T-lymphotrophic virus I/II, stool ova, and parasites, as well as a Strongyloides antibody titer, were performed; all were within reference range. His antinuclear antibody level was mildly elevated at 1:160, but the patient had no clinical manifestations of lupus. Given this negative workup, the most likely explanation for the hypereosinophilia was a reactive process secondary to the extreme inflammatory state.

The patient was started on isotretinoin 40 mg daily in addition to urea cream 40% mixed with clobetasol ointment at least once daily to the extremities. Hydrocortisone ointment 2.5% and petrolatum-based ointment were applied to the face, and hydroxyzine was used as needed for pruritus. One month after initiating isotretinoin, erythema had decreased and a repeat complete blood cell count with differential showed a decrease of eosinophils to 14.7% and an absolute eosinophil count of 1.56×10⁹/L. After 2 months of therapy, the patient showed remarkable improvement. After 3.5 months of therapy, the keratoderma on the palms and soles was almost completely resolved, the follicular-based papules disappeared, and the patient had no areas of lichenification (Figures 1B and 2B). After 5 months of therapy, the patient experienced resolution of the PRP, except for residual facial erythema. His eosinophil count continued to trend downward during these 5 months, reaching 7.6% with an absolute eosinophil count of 0.93×10⁹/L. Three years after the initial onset of the rash and 2 years after completing isotretinoin, his eosinophil level was normal at 5.3% with an absolute eosinophil count of 0.7×10⁹/L.

We present a case of PRP and severe eosinophilia. We initially considered a second disease process to explain the extremely elevated eosinophil count; however, a negative eosinophilia workup and simultaneous resolution of these problems suggest that the eosinophilia was related to the severity of the PRP.

To the Editor:

A 63-year-old man presented with a prior diagnosis of severe psoriasis affecting the extremities, neck, face, and scalp of 1 year’s duration. He reported pain, itching, and swelling in the affected areas. He felt the rash was worst on the hands and feet, and pain made performing activities of daily living difficult. His treatment regimen at presentation included triamcinolone cream 0.1% and azathioprine 150 mg daily as prescribed by an outside dermatologist without any response. Physical examination revealed diffuse erythema with lichenification and thick, white, flaking scale on the arms and legs (Figure 1A), face, neck, palms, and soles with islands of sparing. Multiple salmon-colored, follicular-based papules topped with central hyperkeratosis were scattered on these same areas. The palms and soles had severe confluent keratoderma (Figure 2A). Histologic examination of a follicular-based papule showed foci of parakeratosis and hypergranulosis consistent with the patient’s clinical picture of pityriasis rubra pilaris (PRP).

Baseline laboratory tests at the time of PRP diagnosis revealed 20.8% eosinophils (reference range, 0%–7%) and an absolute eosinophil count of 2.17×10⁹/L (reference range, 0–0.7×10⁹/L). Laboratory test results from an outside dermatologist conducted 10 to 12 months prior to the current presentation showed 12% eosinophils with a white blood cell count of 8.9×10⁹/L (reference range, 4.5–11.0×10⁹/L) around the time of rash onset and before treatment with azathioprine, making a drug reaction an unlikely cause of the eosinophilia.

After consulting with the hematology department, a hypereosinophilia workup including erythrocyte sedimentation rate, lactate dehydrogenase, serum protein electrophoresis, urine protein electrophoresis, tryptase, double-stranded DNA antibody, human T-lymphotrophic virus I/II, stool ova, and parasites, as well as a Strongyloides antibody titer, were performed; all were within reference range. His antinuclear antibody level was mildly elevated at 1:160, but the patient had no clinical manifestations of lupus. Given this negative workup, the most likely explanation for the hypereosinophilia was a reactive process secondary to the extreme inflammatory state.

The patient was started on isotretinoin 40 mg daily in addition to urea cream 40% mixed with clobetasol ointment at least once daily to the extremities. Hydrocortisone ointment 2.5% and petrolatum-based ointment were applied to the face, and hydroxyzine was used as needed for pruritus. One month after initiating isotretinoin, erythema had decreased and a repeat complete blood cell count with differential showed a decrease of eosinophils to 14.7% and an absolute eosinophil count of 1.56×10⁹/L. After 2 months of therapy, the patient showed remarkable improvement. After 3.5 months of therapy, the keratoderma on the palms and soles was almost completely resolved, the follicular-based papules disappeared, and the patient had no areas of lichenification (Figures 1B and 2B). After 5 months of therapy, the patient experienced resolution of the PRP, except for residual facial erythema. His eosinophil count continued to trend downward during these 5 months, reaching 7.6% with an absolute eosinophil count of 0.93×10⁹/L. Three years after the initial onset of the rash and 2 years after completing isotretinoin, his eosinophil level was normal at 5.3% with an absolute eosinophil count of 0.7×10⁹/L.

We present a case of PRP and severe eosinophilia. We initially considered a second disease process to explain the extremely elevated eosinophil count; however, a negative eosinophilia workup and simultaneous resolution of these problems suggest that the eosinophilia was related to the severity of the PRP.

To the Editor:

A 63-year-old man presented with a prior diagnosis of severe psoriasis affecting the extremities, neck, face, and scalp of 1 year’s duration. He reported pain, itching, and swelling in the affected areas. He felt the rash was worst on the hands and feet, and pain made performing activities of daily living difficult. His treatment regimen at presentation included triamcinolone cream 0.1% and azathioprine 150 mg daily as prescribed by an outside dermatologist without any response. Physical examination revealed diffuse erythema with lichenification and thick, white, flaking scale on the arms and legs (Figure 1A), face, neck, palms, and soles with islands of sparing. Multiple salmon-colored, follicular-based papules topped with central hyperkeratosis were scattered on these same areas. The palms and soles had severe confluent keratoderma (Figure 2A). Histologic examination of a follicular-based papule showed foci of parakeratosis and hypergranulosis consistent with the patient’s clinical picture of pityriasis rubra pilaris (PRP).

Baseline laboratory tests at the time of PRP diagnosis revealed 20.8% eosinophils (reference range, 0%–7%) and an absolute eosinophil count of 2.17×10⁹/L (reference range, 0–0.7×10⁹/L). Laboratory test results from an outside dermatologist conducted 10 to 12 months prior to the current presentation showed 12% eosinophils with a white blood cell count of 8.9×10⁹/L (reference range, 4.5–11.0×10⁹/L) around the time of rash onset and before treatment with azathioprine, making a drug reaction an unlikely cause of the eosinophilia.

After consulting with the hematology department, a hypereosinophilia workup including erythrocyte sedimentation rate, lactate dehydrogenase, serum protein electrophoresis, urine protein electrophoresis, tryptase, double-stranded DNA antibody, human T-lymphotrophic virus I/II, stool ova, and parasites, as well as a Strongyloides antibody titer, were performed; all were within reference range. His antinuclear antibody level was mildly elevated at 1:160, but the patient had no clinical manifestations of lupus. Given this negative workup, the most likely explanation for the hypereosinophilia was a reactive process secondary to the extreme inflammatory state.

The patient was started on isotretinoin 40 mg daily in addition to urea cream 40% mixed with clobetasol ointment at least once daily to the extremities. Hydrocortisone ointment 2.5% and petrolatum-based ointment were applied to the face, and hydroxyzine was used as needed for pruritus. One month after initiating isotretinoin, erythema had decreased and a repeat complete blood cell count with differential showed a decrease of eosinophils to 14.7% and an absolute eosinophil count of 1.56×10⁹/L. After 2 months of therapy, the patient showed remarkable improvement. After 3.5 months of therapy, the keratoderma on the palms and soles was almost completely resolved, the follicular-based papules disappeared, and the patient had no areas of lichenification (Figures 1B and 2B). After 5 months of therapy, the patient experienced resolution of the PRP, except for residual facial erythema. His eosinophil count continued to trend downward during these 5 months, reaching 7.6% with an absolute eosinophil count of 0.93×10⁹/L. Three years after the initial onset of the rash and 2 years after completing isotretinoin, his eosinophil level was normal at 5.3% with an absolute eosinophil count of 0.7×10⁹/L.

We present a case of PRP and severe eosinophilia. We initially considered a second disease process to explain the extremely elevated eosinophil count; however, a negative eosinophilia workup and simultaneous resolution of these problems suggest that the eosinophilia was related to the severity of the PRP.