Copyright 2006 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.2006
Neutrophilic dermatoses are a collection of diseases with varying presentation unified by clinical and histologic features. Neutrophilic dermatosis of the dorsal hands is a recently described clinical entity and an evolving disease concept. Its relationship to acute febrile neutrophilic dermatosis (Sweet syndrome), pyoderma gangrenosum, and a primary vasculitis has been debated.
We present 9 cases (8 women and 1 man) of neutrophilic dermatosis of the dorsal hands, all with consistent histologic features. Two cases had histologic evidence of vasculitis, and 3 had clinical extension of lesions onto the forearms. Most showed fever, leukocytosis, and/or elevated erythrocyte sedimentation rate. Individual cases were associated with leukemia, lung carcinoma, and inflammatory bowel disease. All 9 patients responded to systemic corticosteroid therapy, with additional response to dapsone, methotrexate, and potassium iodide therapies in several cases. Of the 9 patients, 5 showed complete resolution of their skin disease, whereas 4 required ongoing therapy. We assessed the 43 cases previously reported in the literature.
The clinical presentation, laboratory data, histologic features, and response to corticosteroid therapy offer strong evidence that neutrophilic dermatosis of the dorsal hands is a localized variant of Sweet syndrome and is also identical to atypical pyoderma gangrenosum when that condition presents on the hands.
Neutrophilic dermatosis of the dorsal hands (NDDH) is a disease concept that continues to evolve. The relationship of NDDH to pustular vasculitis, acute febrile neutrophilic dermatosis (Sweet syndrome), and pyoderma gangrenosum (PG) remains undecided. The presentation of violaceous papulonodules on the radial aspects of the dorsal hands was originally noted by Strutton et al,1 and Galaria et al2 proposed the term neutrophilic dermatosis of the dorsal hands for this disease. Others have since reported cases and offered insights into the categorization of this entity.3- 5 We report 9 new cases of NDDH, all with features of Sweet syndrome, including histologic features (brisk dermal neutrophilic infiltrate), laboratory findings (elevated inflammatory markers and neutrophilia), constitutional symptoms (fever, malaise), associated illnesses (hematologic malignancy, inflammatory bowel disease), and response to treatment with corticosteroids. We discuss these 9 cases in the context of the 43 cases previously reported. When features of NDDH are compared with what is accepted to constitute Sweet syndrome, encompassing clinical, laboratory, and histologic characteristics, we conclude that NDDH is best recognized as a distributional variant of Sweet syndrome. Moreover, cases of atypical PG, when presenting in its usual location of the upper extremity, are clinically and histologically indistinguishable from NDDH and display a similar pattern of underlying systemic disease. Others have noted the similarity between atypical PG and bullous Sweet syndrome.6,7 Thus, atypical PG, bullous Sweet syndrome, and pustular vasculitis of the hands are best understood as variations of a single disease entity, which we propose is most appropriately designated NDDH.
A 61-year-old woman was seen for a painful ulcer on her right dorsal hand, which began as a pink papule 2 weeks prior. She had been treated with antibiotics (cephalexin hydrochloride and ampicillin sodium–sulbactam sodium) for the past 10 days. She denied having fever or prior trauma, though reported that similar lesions had developed intermittently over the joints of her dorsal hands for the past 10 years, typically lasting 1 to 2 weeks and treated with wound care and occasional oral antibiotics. Physical examination revealed a 6 × 4-cm violaceous plaque with central nonundermining superficial ulceration at the right dorsal hand over the second and third metacarpophalangeal joints extending onto the index finger (Figure 1A). Lymphadenopathy was absent. Findings from laboratory studies showed moderate leukocytosis, neutrophilia, anemia, and elevated inflammatory indices (Table 1). Results of serum electrolyte assessment, renal and liver function panel, serum protein electrophoresis, and urinalysis, as well as hepatitis serologies and autoantibody profile (antinuclear antibody, SS-A, SS-B, double-stranded DNA, antineutrophil cytoplasmic antibody, and thyroid peroxidase), were normal. A cutaneous biopsy was performed (Table 1). Bacterial, fungal, and mycobacterial cultures were negative. She was initially treated with dapsone, 100 mg/d. Prednisone, 60 mg/d, was added 5 days later, with improvement over the next week (Figure 1B). The lesions resolved, and both medications were slowly tapered.
A, Violaceous plaque on right dorsal hand; B, clinical improvement 2 weeks later under therapy with prednisone and dapsone.
A 44-year-old woman with ulcerative colitis was seen for recurrent ulcerations on the dorsal hands. Lesions, associated with fever and chills, began as violaceous edematous plaques and would ulcerate within 2 days. A biopsy was performed (Table 1). Cultures for bacteria, fungi, and mycobacteria were negative. She responded rapidly to prednisone therapy, 1 mg/kg, tapered over 6 weeks as dapsone therapy was initiated.
A 48-year-old woman was seen for a painful pustular eruption on her dorsal fingers. The lesions started as small, edematous, pink papules evolving into pustules with subsequent ulceration. She did not improve with ceftriaxone sodium and azithromycin therapies. Complete blood cell count revealed leukocytosis with neutrophilia. A biopsy was performed (Table 1). Cultures for bacteria, fungi, and mycobacteria were negative. Her skin lesions, fever, and laboratory abnormalities resolved with prednisone tapered over 8 weeks, and she continues dapsone therapy without recurrence.
A 71-year-old man was referred for ulcerations on his hands, which developed from spontaneous papules. Treatment with cephalexin hydrochloride and then amoxicillin sodium–clavulanic acid had been ineffective, as had acetic acid soaks and mupirocin cream. He reported chills but denied having arthralgias. He had a history of lung adenocarcinoma treated with lobectomy. Metastases to the mediastinum and left adrenal gland were discovered, and he had completed chemotherapy 8 months prior to presentation. Physical examination revealed superficial ulcerations with raised violaceous nonundermined borders on the dorsal aspect of both index fingers and the left fourth finger (Figure 2A). Laboratory studies revealed leukocytosis, neutrophilia, and elevated erythrocyte sedimentation rate and C-reactive protein level. Radiography of the hands was unremarkable, and bacterial cultures were negative. A biopsy was obtained (Figure 2B; Table 1). He was treated with prednisone, 60 mg/d (0.8 mg/kg) for 4 days and then tapered over 3 weeks, and continued local wound care, with rapid improvement. He died of metastatic lung cancer 4 months later.
A, Superficial ulcerations with raised violaceous borders on the dorsal index fingers and left fourth finger; B, cutaneous biopsy specimen showing spongiosis, compact parakeratosis, and a dense upper dermal neutrophilic infiltrate (hematoxylin-eosin, original magnification ×40).
A 30-year-old woman was seen for an enlarging ulcer on her left dorsal hand for 4 days. She denied having fever, prior trauma, or a history of similar lesions. Physical examination revealed a 4 × 4-cm violaceous, firm, centrally necrotic plaque on the left dorsal hand with 2 erythematous 4-mm papules on the forearm. Laboratory studies revealed elevated inflammatory indexes and leukocytosis. A cutaneous biopsy was performed (Table 1). She was treated with prednisone, 60 mg/d (tapered over 2 weeks), and topically with a combination of bacitracin zinc and polymyxin B sulfate (Polysporin; Glaxo Wellcome, Research Triangle Park, NC) and adaptic gauze, with resolution.
A 69-year-old woman was seen for a tender ulcer on her right thumb for 2 weeks, which was treated with oral antibiotic without improvement. She then sustained cat bites on her right index and middle fingers with resulting pathergic ulcerations, which was unresponsive to treatments with intravenous (IV) levofloxacin and metronidazole hydrochloride and then ampicillin-sulbactam. A biopsy was performed (Table 1). On transfer to our institution, physical examination revealed a 2-cm ulceration with a violaceous border on her right dorsal thumb with smaller ulcers on bilateral second and third dorsal fingers. Laboratory studies revealed elevated inflammatory indexes and a monoclonal gamma globulinemia. Antibiotic treatments were discontinued, and prednisone therapy was started at 60 mg/d. The lesions slowly resolved, and prednisone therapy was tapered over 4 months. Subsequent oncologic evaluation disclosed a low-grade B-cell lymphoma, which was monitored clinically.
A 71-year-old woman was seen for a pruritic eruption initially appearing on her right wrist and hand and spreading to the forearms and trunk. The eruption had begun 2 weeks after right knee arthroplasty. She denied having fever, local trauma, or a history of similar lesions. Physical examination revealed violaceous 3- to 4-cm plaques on the right dorsal wrist and hand, with smaller (1- to 1.5-cm) erythematous, succulent plaques on the thighs. A cutaneous biopsy was performed (Table 1). Treatment with prednisone, 30 mg/d, and 0.1% triamcinolone acetonide ointment resulted in improvement after 1 week. After 3 weeks of tapering prednisone therapy, her lesions flared, prompting a 4-week tapering of prednisone therapy, while dapsone therapy was started at 50 mg/d. She continues to receive treatment with dapsone and 0.1% tacrolimus ointment without recurrence.
A 31-year-old man presented in November 1998 with spontaneously arising bullae on his dorsal hands and fingers, which expanded into tender ulcers. A biopsy revealed a dense neutrophilic infiltrate with papillary dermal edema without vasculitis. Laboratory studies revealed leukocytosis, neutrophilia, and elevated liver function test results. Findings from serum protein electrophoresis, urinalysis, and urine porphyrins were normal. Bullous Sweet syndrome was diagnosed. He began treatment with prednisone, 60 mg/d, and topical Polysporin, with initial improvement. His disease flared over the next 6 months during any attempt to taper prednisone therapy, and he did not tolerate treatments with colchicine or dapsone. Flares responded briefly to treatment with IV methylprednisolone sodium phosphate. Minocycline hydrochloride therapy was added in August 1999 without significant benefit. A reevaluation in October 1999 revealed violaceous ulcers on the right first and left fourth fingers (2 cm) and the left dorsal hand (5 cm). A second biopsy (Table 1) revealed vasculitis, and his condition was now interpreted as atypical PG. Findings from colonoscopy and bone marrow biopsies were normal. Treatment with topical tacrolimus was not helpful, and 3-month trials of cyclosporine and thalidomide failed. By May 2000, lesions had developed on his ears and shoulders, which were poorly controlled with prednisone therapy, 40 mg/d, with methylprednisolone pulses every 3 to 4 weeks. Cyclophosphamide, 500 mg, was administered IV 4 times over 2 months with improvement, allowing for tapering of prednisone therapy to 25 mg/d. Therapy with saturated solution of potassium iodide was added in October 2000. He developed indurated papules on his forearms, and a biopsy confirmed lymphocytic infiltrate of Jessner. Therapy with hydroxychloroquine sulfate, 200 mg/d, was added. His disease flared despite increased use of corticosteroid, and in December 2001, cyclophosphamide was resumed monthly for 5 months, allowing for tapering of prednisone therapy to 20 mg/d. Cyclophosphamide therapy was subsequently discontinued because of concern for toxic effects, and he began methotrexate therapy in July 2002, at a weekly dose of 17.5 mg, allowing for tapering of prednisone therapy to 10 mg/d. His disease was well controlled in May 2003, at the time of his accidental death.
A 51-year-old woman developed expanding ulcers on her hands and arms for 2 weeks. The onset was associated with scratches from her domestic cat. A local physician began therapy with antibiotics (amoxicillin–clavulanic acid and then IV ampicillin-sulbactam) without improvement. Bacterial and fungal cultures were negative, and she denied having chills or sweats. Her medical history included depression, migraines, and syncopal episodes during the past year, which were diagnosed as transient ischemic attacks. On physical examination, superficially ulcerated plaques with violaceous undermined borders were present on the dorsal hands, left fifth finger, and left dorsal forearm, without purulent discharge (Figure 3A). Laboratory studies revealed an elevated C-reactive protein level, leukocytosis, neutrophilia, and anemia. Therapy with oral antibiotics was discontinued, and her wounds were treated with bacitracin zinc and nonadherent gauze. Prednisone therapy was begun at 60 mg/d (1.25 mg/kg), and she was hospitalized. A skin biopsy was performed (Figure 3B; Table 1). The ulcerations were improving 3 days later, and she was discharged on therapy with tapering doses of prednisone and fluocinonide ointment. New plaques on the hands developed as the prednisone dose was decreased to 15 mg/d, and dapsone therapy was added (50 mg/d), with doses changed to 10 mg/d and 100 mg/d, respectively, after 4 weeks. Two months later, dapsone therapy was discontinued owing to hemolytic anemia, and the prednisone dose was increased to 20 mg/d. Shortly thereafter she was hospitalized for confusion and new-onset seizures; encephalopathic Sweet syndrome was considered. Magnetic resonance imaging disclosed chronic ischemia of the right frontal lobe; findings from electroencephalography and lumbar puncture were normal. A bone marrow biopsy revealed mildly hypercellular marrow. She began gabapentin therapy for her seizure disorder. During this time, she developed new ulcers on her hands and arms at venipuncture sites, prompting an increase in the prednisone dose to 30 mg/d. Colchicine therapy was not tolerated. Therapy with saturated solution of potassium iodide was begun (15 drops 3 times daily). Her cutaneous lesions gradually regressed (Figure 3C). Prednisone and topical corticosteroid treatments were withdrawn over 2 months, and therapy with saturated solution of potassium iodide was tapered over the following 2 months.
A, Superficially ulcerated plaques with violaceous undermined borders on the left dorsal hand and fifth finger; B, cutaneous biopsy specimen showing dense dermal neutrophilic infiltrate with small subcorneal pustules, subepidermal vesiculation, papillary dermal edema, and hemorrhage (hematoxylin-eosin, original magnification ×40); and C, postinflammatory hyperpigmentation of dorsal hands after 9 months of therapy.
Since the original description of acute febrile neutrophilic dermatosis by Sweet8 in 1964, diagnostic criteria have been variably proposed and generally include the abrupt onset of typical skin lesions with characteristic histopathologic features as major criteria and elevation of serum inflammatory markers, presence of constitutional symptoms or associated infection, inflammatory, or malignant disease, and responsiveness to corticosteroid therapy as minor criteria.9 In 1995, Strutton et al1 reported 6 patients with violaceous plaques on the radial aspect of the dorsal hands, with histologic features of papillary dermal edema, brisk neutrophilic infiltrate, and leukocytoclastic vasculitis. The clinicopathologic features and responsiveness to corticosteroid therapy were suggestive of acute febrile neutrophilic dermatosis (Sweet syndrome), though the authors termed the presentation pustular vasculitis of the hands to reflect the vascular damage. Pustular vasculitis had been initially applied to purpuric lesions of Behçet syndrome and bowel-associated dermatosis-arthrosis syndrome.10 In 2000, Galaria et al2 presented 3 similar cases that lacked leukocytoclastic vasculitis and proposed the term neutrophilic dermatosis of the dorsal hands as a localized variant of Sweet syndrome. In 2002, DiCaudo and Connolly3 reported 7 additional cases, all with vasculitis and unassociated with systemic illness. In 2004, Weenig et al5 presented 4 new cases and reviewed the other 32 cases then in the literature, concluding that the presentation warranted evaluation for occult malignancy or other systemic disease. Three additional cases have since been reported.11- 13 Many of these authors have commented on the morphologic similarity of NDDH to Sweet syndrome.1- 4 Clinical features of our 9 patients are summarized in Table 1.
We agree with Galaria et al2 that NDDH is a subset of Sweet syndrome.2 Assimilation of reported cases supports this idea. Of 52 patients, 36 (69%) were female, in accordance with the female preponderance for Sweet syndrome. Of 52 cases of NDDH, 11 (21%) (1 in our series and 10 in the review by Weenig et al5) had associated leukemia, myelodysplasia, or other hematologic disease. This is in agreement with the incidence of hematopoietic disease (16%-54%) in 3 large reviews of Sweet syndrome.14- 16 Of 52 cases of NDDH, 8 (15%) (1 in our series and 7 in the review by Weenig et al5) had associated inflammatory bowel disease, similar to the association (up to 19%16) with Sweet syndrome. Most of our 9 patients showed fever, peripheral neutrophilia, leukocytosis, and/or elevated erythrocyte sedimentation rate or C-reactive protein level. Of the 52 assimilated cases, 17 (33%) mentioned fever and 17 mentioned leukocytosis or peripheral neutrophilia, and 37 (71%) responded to systemic corticosteroid therapy.
While an ulcerative pattern is uncommon in Sweet syndrome (and when present may suggest hematologic malignancy17), ulceration is not unusual in cases of NDDH (27 of 52 cases) and is the clinical sine qua non of PG. While NDDH is clearly distinct from classic PG, it bears resemblance to vesiculobullous PG, also called atypical PG, which is also similar and may be identical to bullous or atypical Sweet syndrome.7 Atypical PG characteristically presents as hemorrhagic bullous lesions, which superficially ulcerate and is most commonly found on the dorsal hands.18 Characteristically, PG involves the lower extremities as deeper ulcers with overhanging borders and often excruciating pain. Atypical PG is distinguished from typical PG histologically by the frequent presence of a diffuse neutrophilic infiltrate with associated vascular reaction and therapeutically by a much prompter remission. Each of these features of atypical PG is similar to NDDH.
It is also likely that many of the cases termed atypical Sweet syndrome or PG-Sweet overlap17,19,20 belong to the spectrum of neutrophilic dermatoses now increasingly recognized as NDDH when present in this distribution.5 Atypical Sweet syndrome has been reported to occur frequently in patients with hematologic malignancy,14 and many of the published photographs of cancer-associated Sweet syndrome21 closely resemble lesions of NDDH seen in the presence and absence of malignancy. Callen6,7 has also noted the similarity between bullous Sweet syndrome and atypical PG and opined that pustular vasculitis of the hands is a variant of atypical PG. At present, as more cases of NDDH have been recognized and reported, the association has crystallized and can be stated more definitively: these 3 entities in fact represent the same disease. When features of NDDH, atypical PG, and Sweet syndrome are numerically compared (Table 2), the similarity is striking.
The fact that the eruptions of NDDH share features of both Sweet syndrome and atypical PG underscores the utility of the designation. Weenig et al5 proposed shortening the designation of NDDH to neutrophilic dermatosis of the hands, though we prefer to retain the term dorsal to reflect the primary distribution in most cases.
Many of the reported cases of NDDH discuss suspected infection and failure of antibiotic therapy prior to dermatologic consultation. Of our 9 patients, 6 had been treated as such prior NDDH diagnosis. While most cases respond to treatment with prednisone or steroid-sparing immunosuppressants, at least 10 of 52 reported cases responded to dapsone and at least 4 responded to minocycline. Interestingly, 3 of our cases demonstrated pathergy, as noted in other series.3
Much of the nosologic debate has been related to the presence of vasculitis in NDDH biopsy specimens. Cases originally described as Sweet syndrome lacked vasculitis, though vascular injury may occur secondary to a brisk neutrophilic infiltrate and the accompanying enzymatic and cytokine cascade. Leukocytoclastic vasculitis was present in 2 (22%) of 9 patients in our series and in 12 (28%) of 43 cases in the literature,3,5 or 27% (14/52) in aggregate. This agrees with 2 histologic series that identified leukocytoclastic vasculitis or vasculitislike changes in 18% to 30% of cases of typical Sweet syndrome.23,24 Malone et al23 reviewed the presence of vasculitis in Sweet syndrome in relation to the timing of the biopsy and found that vasculitis correlated with lesions of longer duration. Indeed, 1 of our patients (case 8) had an initial biopsy result negative for vasculitis, then a subsequent positive biopsy result, suggesting a difference in lesion timing or sampling rather than a fundamental change in the disease process. Cohen4 and Cohen and Kurzrock22 have argued that vasculitis in Sweet syndrome and in NDDH is an epiphenomenon in which the damaged vessel is an “innocent bystander” of an inflammatory dermatosis. We agree with these conclusions and contend that the presence or absence of vasculitis is of secondary importance in the diagnosis of NDDH.
Others have opined that NDDH may be a variant of erythema elevatum diutinum.25 In our opinion, the clinical appearance, only occasional presence of leukocytoclastic vasculitis, and infrequent association with IgA gammopathy, adequately differentiate these conditions clinically. Moreover, the histologic features of chronic erythema elevatum diutinum lesions, including fibrosis, vascular proliferation, and a lymphohistiocytic infiltrate with plasma cells, have not been noted in NDDH. The presence of vasculitis and antineutrophil cytoplasmic antibody positivity have been too infrequent to consider NDDH a variant of Wegener granulomatosis.26
In conclusion, NDDH is an evolving disease concept. The nosologic classification of this disease as a limited distributional variant of Sweet syndrome is supported by the clinical presentation, laboratory and histologic findings, presence of comorbid medical conditions, and response to treatment. When present on the dorsal hands, atypical PG is clinically indistinguishable and represents the same disease. We view NDDH as a distributional variant of Sweet syndrome, which is identical to atypical PG and pustular vasculitis of the hands, and propose that the term neutrophilic dermatosis of the dorsal hands is the most appropriate designation for such eruptions.
Correspondence: Hobart W. Walling, MD, PhD, Dermatology, P.C., 6000 University Ave, Suite 450, West Des Moines, IA 50266 (firstname.lastname@example.org).
Financial Disclosure: None.
Accepted for Publication: July 16, 2004.
Author Contributions:Study concept and design: Walling and Piette. Acquisition of data: Walling, Snipes, and Gerami. Analysis and interpretation of data: Walling. Drafting of the manuscript: Walling, Snipes, and Gerami. Critical revision of the manuscript for important intellectual content: Walling and Piette. Statistical analysis: Walling. Study supervision: Walling and Piette.
Walling HW, Snipes CJ, Gerami P, Piette WW. The Relationship Between Neutrophilic Dermatosis of the Dorsal Hands and Sweet SyndromeReport of 9 Cases and Comparison to Atypical Pyoderma Gangrenosum. Arch Dermatol. 2006;142(1):57-63. doi:10.1001/archderm.142.1.57