Abstract

Background

Cutaneous granulomatous vasculitis (CGV) is a category of inflammatory skin conditions characterized by evidence of vascular damage to small vessels with a predominance of histiocytes. CGV is often associated with systemic diseases such as sarcoidosis, systemic vasculitis, and lymphoma, underscoring the importance of accurate diagnosis and classification for effective management. It can also be associated with drug ingestion (Gibson, el-Azhary et al.

1994). However, in some cases, a histopathologist can only make a diagnosis of granulomatous vasculitis and list disease associations. An important consideration is that earlier reports of CGV predate contemporary recognition that immature neutrophils often mimic histiocytes under microscopic examination. This raises the possibility that some cases historically diagnosed as CGV may be instances of neutrophilic vasculitis in which immature neutrophils/bands or metamyelocytes are mistaken for histiocytes. The senior author has observed cases in which staining with myeloperoxidase (MPO) has shown that what appears to be granulomatous vasculitis is instead vasculitis containing many immature neutrophils. The primary aim of this project is to delineate these cases from ones in which true histiocytes and not histiocytoid myeloid cells are present. In this study, we reassess cases diagnosed as CGV from the archives of University of California, San Francisco (UCSF) and Mass General Brigham using immunohistochemistry to identify immature myeloid cells (MPO positive) and histiocytes

(CD163 positive). This project is novel in its re-evaluation of clinical CGV diagnoses using contemporary knowledge and techniques. Its importance lies in addressing potential misclassifications that may impact prognosis.

Methods

Over 300 cases from 2012-2024 were retrospectively identified from the dermatopathology archives of Mass General Brigham and UCSF using variations of the search terms “granulomatous,” “cutaneous vasculitis,” “histiocytic vasculitis,” and “histiocytoid vasculitis.” The

cases were manually reviewed, and granulomatous vasculitis was defined as the presence of well or poorly formed angiocentric granulomas associated with vascular injury, including necrosis and fibrin deposition. 40 unique patient cases meeting criteria for granulomatous vasculitis were identified, corresponding to 46 biopsy sites.5-μm sections from each block were stained for CD163, CD3, and MPO. Immunostaining was semi-quantitatively scored as 1+ (scattered perivascular staining), 2+ (moderate circumferential perivascular staining), or 3+ (robust circumferential perivascular staining). Scoring was assisted by quantitative assessment using QuPath cell detection software. Patient clinical histories were reviewed and correlated with immunostaining results for each case to identify associations between histopathologic features and clinical context.

Results

A total of 40 cases of cutaneous granulomatous vasculitis, corresponding to 46 biopsy sites, were analyzed. The cohort included 28 females (70%) with a mean age of 47.8 years, and 12 males with a mean age of 51.3 years. Necrotizing granulomatous vasculitis was identified in 9 cases (23%) and 10 biopsy sites (22%). All cases demonstrated histiocytic inflammation, confirmed by CD163 immunostaining. CD163 expression was consistently present, ranging from

1+ (>30% of nuclei) to 3+ (>70% of nuclei). Myeloperoxidase (MPO) staining, indicative of the presence of cells of myeloid lineage, was present in all but one case. Myeloperoxidase staining was graded as 1+ (~10% of cells), 2+ (10–30% of cells), and 3+ (>40% of cells). On H&E, MPO- positive cells corresponded to mononuclear elements with elongated, twisted nuclei and scant cytoplasm; mature neutrophils were not seen.  Biopsies with 1+ MPO staining showed the strongest correlation with rheumatologic diseases. Among these, 3 necrotizing cases (60%) were associated with polyarteritis nodosa or granulomatosis with polyangiitis. In contrast, 6 non- necrotizing biopsies (35%), representing 5 patients, were linked to lupus or antiphospholipid antibody syndrome. Cases with 3+ MPO staining lacked a clear rheumatologic association, possibly due to the paucity of cases (n=2). One case with 3+ MPO staining was observed in a patient with Crohn’s disease. This pattern of MPO reactivity may reflect nonspecific

inflammatory vigor rather than disease specificity. Four biopsy sites (40%) without necrosis and with 3+ CD163 staining were associated with rheumatologic diseases, notably lupus/antiphospholipid antibody syndrome (3 cases). These had a mean MPO score of 1.3. This grouping suggests that robust histiocytic infiltration accompanied by modest immature neutrophil presence (~10% of cells) may serve as a useful immunophenotypic clue to identifying rheumatologic etiologies in granulomatous vasculitis.

Conclusion

This study shows that immature myeloid cells are present are present in what has been thought of as an exclusively histiocytic process. It also shows that strong CD163 expression with modest MPO-positive immature neutrophil infiltration (~10%) defines a histiocyte-rich immunophenotypic subset of granulomatous vasculitis most often linked to rheumatologic disease, suggesting a reproducible histopathologic signature with potential diagnostic value. It remains possible that the percentage of immature myeloid cells in granulomatous vasculitis varies over time, but the absence of sequential biopsies from individual patients precludes a conclusion about this. The findings suggest that anti-neutrophilic therapies may be effective in some cases of granulomatous vasculitis.