Nneka I. Comfere, MD, Department of Dermatology, Mayo Clinic, 200 First Street SW, Rochester, MN 55905, USA email: email@example.com
Background: Wegener’s granulomatosis (WG), a systemic vasculitis, can be associated with cutaneous signs and symptoms before, during or after the diagnosis of systemic disease.
Methods: We reviewed clinical and histologic features of cutaneous lesions from 17 patients with WG. The temporal relationship between development of cutaneous symptoms and onset of systemic disease was determined, and antineutrophil cytoplasmic antibody (ANCA) status of the patients was also established.
Results: In six patients, systemic and cutaneous disease developed concurrently. In eight patients, cutaneous disease developed after patients received the diagnosis of systemic disease. In three patients, cutaneous disease preceded systemic disease. Cytoplasmic ANCA or proteinase-3-ANCA [c-ANCA/proteinase 3 (PR3)-ANCA] serologic test results were negative for one patient when cutaneous disease developed, and one patient had c-ANCA/PR3-ANCA seroconversion a year before systemic disease developed. Histopathologic features of cutaneous WG were not limited to leukocytoclastic vasculitis; they also included acneiform perifollicular and dermal granulomatous inflammation and palisaded neutrophilic and granulomatous inflammation.
Conclusions: Patients with WG can present initially with cutaneous symptoms. Histopathologic patterns vary, but leukocytoclastic vasculitis is most commonly noted. Patients with WG and skin lesions are likely to have positive c-ANCA/PR3-ANCA serologic test results.
Patients with Wegener’s granulomatosis (WG), an uncommon systemic vasculitis, can present with skin lesions that often appear concurrently with visceral involvement or after systemic signs and symptoms develop. About 15% of patients with WG will develop specific skin lesions at some time during the course of the disease. The most common skin lesion specific to WG is palpable purpura with the histopathologic correlate of leukocytoclastic vasculitis. However, a wide array of clinical and histologic features may be seen.1–6
No reliable test can identify which patients with typical WG skin lesions will eventually show systemic signs and symptoms of WG. Serologic tests, including measurement of antineutrophil cytoplasmic antibody (ANCA) titers, have played important roles in identifying patients with WG.7,8 ANCA testing uses a well-described indirect immunofluorescence technique, and positive cytoplasmic ANCA (c-ANCA) patterns correlate well with a diagnosis of WG.9–16 Proteinase 3 (PR3) is the predominant antigen of c-ANCA.17–19 Target antigen-specific solid-phase assays show that most patients with c-ANCAs have PR3-specific antibodies (PR3-ANCA).20,21 The combination of positive test results from indirect immunofluorescence assays identifying c-ANCA and target antigen-specific solid-phase assays identifying PR3-ANCA is the most sensitive and specific criterion for the diagnosis of WG.22
Knowledge of c-ANCA/PR3-ANCA levels helps dermatopathologists specify the diagnosis of cutaneous WG when interpreting a skin biopsy with compatible characteristics. The histologic identification of leukocytoclastic vasculitis, clinical findings of cutaneous WG without systemic involvement and positive c-ANCA/PR3-ANCA serologic test results should raise clinical suspicion for the subsequent development of systemic disease. Such patients should have frequent follow-up evaluations to screen for internal organ involvement, and appropriate management should be initiated early in the disease course. For patients with WG, serial tests for ANCAs are recommended because ANCA levels correlate with disease activity. Furthermore, disease relapse in the presence of negative ANCA test findings is unusual.
The primary objective of this retrospective study was to determine how often patients with clinicopathologic confirmation of cutaneous WG have negative c-ANCA/PR3-ANCA serologic test results when skin lesions occur. Secondary objectives of the study included expanding our clinical and histopathologic experience with cutaneous WG and identifying atypical or unusual cutaneous patterns or serologic test results.
Patients and methods
Seven hundred sixty-six patients with WG underwent evaluation and treatment at Mayo Clinic from 1996 through 2004, and about 10% of patients had cutaneous involvement during the disease course. Our study examined 17 patients with cutaneous WG, for whom histologic skin specimens and ANCA serologic test results were available for review.
The study was approved by the Mayo Foundation Institutional Review Board. Criteria for inclusion were clinical, histologic and serologic (c-ANCA/PR3-ANCA) findings that were characteristic of cutaneous WG. Patient cases without complete documentation of characteristic findings were excluded.
Clinical and demographic information included patient sex, age at diagnosis, location and type of cutaneous lesions, systemic organ involvement and ANCA status when cutaneous and systemic disease developed. The temporal relationship between the development of c-ANCA/PR3-ANCA and the cutaneous and systemic WG was determined. Histopathologic features of each case were also recorded.
Summary of clinical and demographic characteristics
Eleven women and six men had confirmed WG with evidence of organ involvement, positive histopathologic findings and positive ANCA serologic test results. The mean age at onset of systemic disease was 47.4 years (range 10–65 years). Clinical characteristics of the patients are summarized in Table 1; correlation of disease state with ANCA status is shown in Table 2.
Table 1. Clinical characteristics of patients with Wegener’s granulomatosis
All patients had positive serologic test results at some point during the disease course.
20 months later
20 months earlier
8 years later
8 years earlier
7 years earlier
4 years later
2 months later
2 months earlier
6 years later
4 years earlier
1 years earlier
1 years later
4 months later
4 months earlier
Several week later
Several weeks earlier
2 years later
2 years earlier
1 years earlier
Eleven patients had involvement of the lungs or upper respiratory tract (or both). Systemic involvement also included kidneys (five patients), joints (four patients), auditory canals (two patients), gastrointestinal tract (two patients), episclera (one patient), retinal artery (one patient), heart (one patient) and central nervous system (one patient). Six patients had simultaneous systemic and cutaneous disease with positive c-ANCA/PR3-ANCA serologic test findings. For eight, cutaneous disease developed several weeks to 8 years after diagnosis of systemic WG. Of these eight patients, seven had positive c-ANCA findings when systemic disease developed. One had systemic symptoms for 4 years before cutaneous lesions developed but did not have c-ANCA tests until skin lesions appeared. In retrospect, the patient’s systemic symptoms were attributable to WG. Three patients had cutaneous disease that preceded the onset of systemic symptoms by 1–7 years. One patient had negative c-ANCA/PR3-ANCA serologic test results when palpable purpura developed. One had palpable purpura and leukocytoclastic vasculitis for 7 years before systemic symptoms developed and c-ANCAs were identified. However, c-ANCA status at the time of leukocytoclastic vasculitis development was unknown (assays were not performed). One patient had skin disease, and c-ANCA/PR3-ANCA seroconversion occurred 1 year before systemic disease developed.
Target antigen-specific solid-phase assays for PR3 and myeloperoxidase were conducted in 10 patients. Nine patients had positive findings for PR3, and all correlated with positive c-ANCA status and active WG. Neither perinuclear ANCAs (p-ANCAs) nor myeloperoxidase ANCAs were detected in any of the 10 patients.
Summary of histopathologic characteristics
Erythema nodosum (EN-like) histomorphology was observed in patient 1; it was characterized by a predominantly septal and focal lobular panniculitis with a mixed inflammatory infiltrate that consisted of histiocytes, lymphocytes, occasional plasma cells and singly disposed foreign body giant cells with neutrophils. Variable septal fibrosis was evident. Focal lipomembranous and liposclerotic changes were also observed (Fig. 1). The inflammatory infiltrate also extended into the adjacent reticular dermis. No histologic evidence of vasculitis was observed within the skin biopsy specimens from the EN-like subcutaneous nodules. Direct immunofluorescence tests showed a nondiagnostic inflammatory reaction pattern. However, this patient also had concurrent skin biopsies performed on other sites; those specimens were consistent with a granulomatous and neutrophilic vasculitis. Direct immunofluorescence studies were supportive of vasculitis. The deep dermal vessels stained strongly for immunoreactants including immunoglobulin (Ig) M, IgA and C3.
Patient 2 showed dense, perifollicular, suppurative and granulomatous inflammation (lymphocytes, histiocytes, plasma cells, multinucleated giant cells, cellular debris and leukocytoclasis) and follicular destruction and disruption. Gomori methenamine silver stain and Gram stain assays were negative for microorganisms. Swab cultures were also negative for bacteria, fungi and mycobacteria.
The histologic findings of three patients (patients 7, 14 and 17) included palisaded extravascular necrotizing granuloma (Churg-Strauss-like granuloma). The patients typically had focal palisaded neutrophilic and granulomatous inflammation with central basophilic collagen degeneration within the superficial and deep dermis. Also, prominent leukocytoclasis usually was present.
The other patients in our study had histopathologic features of leukocytoclastic vasculitis. However, the extent and degree of involvement were quite robust. Patients had a denser infiltrate of neutrophils, considerable leukocytoclasis and vascular destruction with involvement of the superficial and mid-dermal vasculature. Histopathologic features also included variable infiltration of the vessel walls by neutrophils (extending into the perivascular area and the surrounding dermis) and prominent leukocytoclasis. Thickening and fibrinoid necrosis of the vessel walls were also observed. Occasional thrombosed vessels were present in the superficial dermis. Some patients had other inflammatory cells, including eosinophils within the superficial and mid-dermis. Papillary dermal edema and ectatic lymphatic vessels were noted for some patients. Careful examination of the perivascular dermis and the dermis adjacent to vessels showed slight to moderate basophilic collagen degeneration. However, degeneration tended to occur in conjunction with interstitial neutrophilia and leukocytoclasis. Thus, the focal basophilic collagen degeneration may represent a secondary change because of the adjacent vascular inflammation.
Our previous study2 described 30 patients with cutaneous WG, and this study expands our experience to a total of 47 patients. Although about half of the patients with WG may have cutaneous signs and symptoms, specific cutaneous lesions (directly related to WG and not the result of infection, medication sensitivity or other definable causes) appear in approximately 15% of patients at some point during the course of the illness. Furthermore, for patients with multiple organ involvement (lungs, upper respiratory tract, orbit and joints), cutaneous findings are more likely to develop. Palpable purpura developed in over 70% of the current study group and was the most common clinical finding. Nodules, papules, ulcerations and deep EN-like subcutaneous nodules complete the clinical spectrum (Figs. 2–5).
Most often, leukocytoclastic vasculitis was the corresponding histopathologic pattern for WG in the majority of patients in our study, especially in those with palpable purpura (Fig. 6). Leukocytoclastic vasculitis also may be the histopathologic correlate to ulcerations and nodules. Close inspection of the histopathologic pattern of leukocytoclastic vasculitis in our patients showed a robust neutrophilic inflammatory infiltrate that was associated with vascular destruction and extravascular inflammation. The vascular involvement was not only restricted to the superficial dermal vessels but also affected the mid dermal vasculature. We cannot draw meaningful conclusions from these findings because of the limited number of patients in our study and the inherent bias in our study population, which was drawn from patients with known extracutaneous WG.
One patient (case 2, Figs. 3 and 7) had extensive ulceration on the face and ear and was initially believed to have malignant pyoderma; his illness was reported in a 1997 follow-up summary of patients with WG .23 At this institution, most patients with an initial diagnosis of malignant pyoderma eventually received a diagnosis of WG.24 Patients who received the diagnosis of malignant pyoderma presented with extensive ulcerations that did not have the necrotic centers and undermined violaceous borders that clinically are more characteristic of pyoderma gangrenosum (PG). Furthermore, the classic disease associations of PG differed from that of malignant pyoderma. Long-term follow-up of three patients with the initial diagnosis of malignant pyoderma showed that other systemic signs of WG clearly developed for two patients.25
The histopathologic pattern of pyoderma-like facial lesions of WG is characterized by foci of palisaded neutrophilic and granulomatous dermatitis, prominent granulomatous and neutrophilic necrotizing vasculitis and basophilic collagen degeneration. However, this contrasts with classic PG, which typically shows striking neutrophilic epidermal and dermal necrosis with mononuclear cell dominant vascular inflammation; it occurs away from the zones of marked extravascular inflammation and does not have luminal or mural fibrin deposition. Histopathologic differences between pyoderma-like ulcerations of WG and PG may be more subtle in the early phases of disease, when follicular and perifollicular inflammation with intradermal abscess formation are present. Patient 2 had facial pyoderma-like lesions that showed histopathologic characteristics of acneiform folliculitis. However, during close follow-up, systemic symptoms of WG eventually developed, which made the diagnosis obvious. The histopathologic pattern of sterile neutrophilic folliculitis has been described previously by Magro and Crowson25 as a reaction pattern that reflects systemic disease. Clinical features that should raise suspicion for cutaneous WG in patients with pyoderma-like lesions or ulcerations include facial involvement (especially periauricular lesions), absence of typical PG features (no undermined violaceous borders, erythema or necrotic centers), serologic ANCA positivity, other disease associations and the development of systemic signs and symptoms of WG.4,15,26–29
Rheumatoid-like papules or nodules with histopathologic features of palisaded neutrophilic and granulomatous dermatitis are compatible with rheumatoid arthritis and WG (Figs. 4 and 8). However, positive serologic test results for c-ANCA favor the diagnosis of WG. Like leukocytoclastic vasculitis, this pattern of palisaded granulomatous dermatitis is not specific to WG and may appear with several other disorders, including Churg-Strauss vasculitis.30,31 Nevertheless, the presence of papules or nodules on extensor extremities and a histopathologic correlate of palisaded granulomatous dermatitis should suggest a differential diagnosis that includes cutaneous WG, and serologic testing for c-ANCAs should be performed. Magro et al.32 described the frequent occurrence of granuloma annulare-like reactions with an associated active vasculopathy and extravascular neutrophilia on the elbow in patients with underlying systemic illness such as chronic viral infection, rheumatologic disorders or ANCA-positive vasculitides. The authors concluded that a granuloma annulare-like or necrobiosis lipoidica-like tissue reaction pattern with an active vasculopathy and an extravascular neutrophilia in the appropriate clinical context should raise clinical suspicion for an underlying systemic disease. One patient had anterior leg nodules with a corresponding histopathologic pattern bearing some resemblance to EN (Figs. 1 and 5). With the possible exception of the patient with facial ulcerations (described above), none of the clinical or histopathologic patterns observed in this study were specific to or diagnostic for WG.
Therefore, on the basis of our findings, several important principles require emphasis. In most patients, signs or symptoms of WG precede or are concurrent with the onset of cutaneous symptoms. For the majority of these patients, serologic test results for c-ANCA/PR3-ANCA were positive after cutaneous lesions manifested. In general, 75–90% of patients with active, systemic WG will have positive c-ANCA/PR3-ANCA test results at some point during their disease course. However, in 10–25% of patients with active, systemic disease and 40% of those with limited disease, ANCA serologies may be negative. The sensitivity of ANCA serologic tests is related to the extent, severity and activity of the disease at the time of testing. Thus, lack of a positive c-ANCA/PR3-ANCA test result does not exclude WG.
Other conditions that are associated with c-ANCA/PR3-ANCA positivity include microscopic polyangiitis (10–50%), Churg-Strauss syndrome, drug-induced ANCA-associated vasculitis, collagen vascular disease (rare) and subacute bacterial endocarditis.22 Although some patients with WG will have perinuclear-ANCA or myeloperoxidase ANCA, none were detected in our study group. When patients present with leukocytoclastic vasculitis and positive c-ANCA/PR3-ANCA serologies, the presence or absence of clinical or laboratory findings that are characteristic for the entities in the differential diagnosis will help establish an accurate clinical diagnosis.
Various ANCA patterns and their clinical correlates have been reviewed.22,33 By combining the clinical and histopathologic pattern with ANCA test results, cutaneous findings can be diagnostic for WG in most patients. For those with cutaneous symptoms but without systemic signs or symptoms of WG, a positive c-ANCA test result suggests that these patients need close clinical follow up because WG will probably develop in another organ. A small percentage of patients with leukocytoclastic vasculitis and negative ANCA test results have a risk of developing WG. False-negative results are possible for various reasons, but they most likely occur with limited cutaneous disease or when a patient with WG is in disease remission.34,35 In our study group, the titers of c-ANCA/PR3-ANCA diminished with resolution of cutaneous lesions. Our study design precludes a confident estimate of the percentage of patients with false-negative findings, but we believe it is small relative to the population of patients with leukocytoclastic vasculitis.
In conclusion, awareness of the common presentations of WG in the skin is essential for pathologists. Although the cutaneous patterns are not unique to WG, they are present at some point during the disease course in approximately 15% of patients with WG. The main histopathologic correlate of WG is leukocytoclastic vasculitis, but granulomatous inflammation with or without vasculitis and palisaded extravascular granulomas may also occur. Serologic tests with ANCA (using indirect immunofluorescence) and PR3 (using target antigen-specific solid-phase assays) are effective tools for correlating cutaneous findings in patients who either have or are likely to develop systemic disease.
Editing, proofreading and reference verification of this article were provided by the Section of Scientific Publications, Mayo Clinic.