Assessment of specificity of dermatopathologic criteria for IgG4‐related skin disease

IgG4‐related disease (IgG4‐RD) represents a recently characterized multisystemic fibroinflammatory condition that can manifest a spectrum of skin findings (IgG4‐related skin disease; IgG4‐RSD). Histopathologic and immunohistochemical criteria have been proposed; however, the specificity of these criteria merits scrutiny given the potential histopathologic overlap of IgG4‐RSD and both neoplastic and inflammatory skin conditions featuring lymphoplasmacytic infiltrates (IgG4‐RSD mimics). This study sought to assess the specificity of the criteria by quantifying the frequency by which an expanded spectrum of IgG4‐RSD mimics meet proposed thresholds.

expanded spectrum of potential IgG4-RSD mimics (to include cutaneous marginal zone lymphoma, syphilis, necrobiosis lipoidica, lichen sclerosus, ALHE, psoriasis, lymphoplasmacytic plaque, EED, and erosive pustular dermatosis), highlights the relative nonspecificity of lymphoplasmacytic infiltrates contrasted with the stringency of storiform fibrosis and obliterative fibrosis.Furthermore, an IgG4+ cell count of >10 per hpf and an IgG4/IgG ratio of >0.4 are not specific to IgG4-RSD alone.In the appropriate clinical context for IgG4-RSD, histopathologic features still represent the entry threshold for diagnosis consideration, which then allows for further screening by immunohistochemical criteria.

K E Y W O R D S
IgG4-related disease, IgG4-related skin disease, plasmacytic dermatoses

| INTRODUCTION
IgG4-related disease (IgG4-RD) is a rare, recently characterized chronic fibroinflammatory condition that can affect multiple organ systems.2][3][4][5][6] Affected organs often develop mass lesions, and serum studies may demonstrate elevated levels of IgG, IgG4, and IgE. 7e histopathologic findings of IgG4-RD are reportedly similar across all organs and are characterized by storiform fibrosis, obliterative phlebitis, and a dense lymphoplasmacytic infiltrate enriched with IgG4+ plasma cells demonstrating an IgG4/IgG ratio of >0.4. 8,9G4-RD has been reported to also manifest a spectrum of skin findings, collectively termed IgG4-related skin disease (IgG4-RSD).10 IgG4-RSD appears to be rare but tends to occur in middle-aged or older patients with a male predominance.Clinically, erythematous papules, plaques, and nodules are the most common morphologies and may be associated with pruritus.11 Lesions tend to be located on the head and neck, particularly involving the cheek, periauricular, and mandibular areas.IgG4-RSD typically co-occurs with extracutaneous disease of the head and neck, most commonly involving lymph nodes, lacrimal glands, parotid glands, and submandibular glands.[11][12][13] However, skin-limited manifestations have been documented as the first presentation of IgG4-RD and may precede the development of systemic manifestations by multiple years.[14][15][16] To define and distinguish IgG4-RD from other systemic fibroinflammatory disorders, diagnostic criteria have been proposed and refined over the past decade.8,9,17 Current criteria are based upon reported clinicopathologic features in classically involved organ systems. Hoer, their sensitivity and specificity remain unknown, particularly when applied to less frequently affected organs.The consensus statement on the pathology of IgG4-RD 8 proposed a set of pathologic criteria in which the characteristic morphologic findings form the fundamental basis for diagnosis.However, the diagnosis cannot be confirmed in the absence of IgG4+ plasma cells, making immunohistochemistry an essential part of diagnostic evaluation.In an effort to account for varying levels of fibrosis and cellularity among different involved organs, organ-specific IgG4+ cell counts diagnostic thresholds were proposed.A threshold of >200 IgG4+ cells per highpowered field (hpf) was suggested for the skin.Notably, in this proposed schema, the skin is unable to obtain the designation of "histologically highly suggestive" due to the paucity of IgG4-RSD data. Elvated IgG4+ cell counts and IgG4/IgG ratios have been shown to be nonspecific in extracutaneous tissues 18 ; however, the specificity of these findings in the skin is less well characterized.Lehman et al. 19 demonstrated that certain plasma cell-rich dermatoses and reactive inflammatory infiltrates can meet the consensus criteria 8,9 for IgG4+ cell count threshold and IgG4/IgG ratio.Whether additional plasma cell-rich dermatoses or cutaneous neoplasms can also meet the consensus criteria is as of yet unstudied.
In this study, we sought to expand upon the number of previously studied inflammatory skin conditions and investigate as of yet unstudied neoplastic skin conditions featuring lymphoplasmacytic infiltrates, which might enter into the pathologic differential diagnosis of IgG4-RSD (IgG4-RSD mimics), assessing them for potentially distinctive histopathologic characteristics, mean IgG4+ cell counts, and mean IgG/IgG4 ratios.Moreover, we further evaluated the specificity of proposed histopathologic and immunophenotypic criteria for the diagnosis of IgG4-RSD by determining the frequency at which these IgG4-RSD mimics meet proposed thresholds.This work will serve to further evaluate the utility of current nonorgan-specific diagnostic criteria in the diagnosis of IgG4-RD in the skin.

| METHODS
Following institutional IRB approval (approval number: 201710826), we searched the University of Iowa pathology database (from 2009 to 2022) for the following skin conditions associated with elevated numbers of plasma cells on skin biopsy: erythema elevatum diutinum (EED), lymphoplasmacytic plaque, cutaneous marginal zone lymphoma (CMZL), syphilis, autoimmune blistering disease (pemphigus vulgaris and bullous pemphigoid), psoriasis, angiolymphoid hyperplasia with eosinophilia, lichen sclerosus, plasma cell dyscrasia, Zoon balanitis, necrobiosis lipoidica, necrobiotic xanthogranuloma (NXG), morphea, and erosive pustular dermatosis.Skin biopsy slide diagnoses, as originally interpreted by a board-certified staff dermatopathologist who signed out the case, were all re-reviewed by one of the boardcertified dermatopathologist study authors (V.L.) and one pathology resident physician (D.P.) and included if diagnostically fully concordant.Cases met inclusion criteria if they were classified under one of the above-identified diagnoses, demonstrated at least 10 plasma cells per hpf (Â40 objective lens), and had features of at least one of the major histopathologic criteria proposed by the consensus statement. 8e major histopathologic criteria include (i) dense lymphoplasmacytic infiltrate; (ii) fibrosis, at least focally storiform; (iii) obliterative phlebitis.Cases without significant plasma cells, major histopathologic features, or sufficient tissue for IHC analysis were excluded from the study.A chart review of patients' age and gender was completed.
The hematoxylin and eosin slides were assessed for the following histopathologic characteristics: dense lymphoplasmacytic infiltrate (categorized as absent, mild, moderate, or severe), presence of inflammatory infiltrate (single, multifocal, or diffuse), location of inflammatory infiltrate (in papillary dermis only, reticular dermis only, or both), presence of fibrosis (absent, present with no storiform features, or storiform fibrosis), obliterative phlebitis (present or absent), nonobliterative phlebitis (present or absent), and increased tissue eosinophils (absent, mild, moderate, or abundant).
IgG and IgG4 IHC stains were performed for each specimen using the formalin-fixed, paraffin-embedded tissue and a standard IHC technique with appropriate positive and negative controls.For each case, reviewers blinded to diagnoses identified and photographed 3 hpf (approximately 1 mm 2 ) with the highest subjective densities of IgG4+ plasma cells using a Â40 objective lens, as were the corresponding areas on the IgG-stained slides.Cell counting was conducted through visual inspection at a high-power field of Â400.Plasma cells with cytoplasmic immunostaining were counted if greater than 50% of the individual cell was present in the field of view.For each hpf, IgG4+ plasma cell and IgG + plasma cell counts were obtained.Mean IgG4+ plasma cell counts per hpf were calculated and recorded.The mean IgG4/IgG ratio for each specimen was calculated and recorded.These histopathologic and IHC results were analyzed to determine if three specific thresholds were met:  2. Moderate to severe dense lymphoplasmacytic infiltrate was observed in 41 (59%) of 69 specimens.Fibrosis was seen in 36 (52%) of 69 specimens; however, a predominantly storiform pattern of fibrosis was observed in only 1 (1%) (EED case) of T A B L E 1 Analysis of IgG4-related quantitative thresholds for IgG4-RSD histopathologic mimics.

| DISCUSSION
With the emergence of IgG4-RD as a newly recognized multisystemic disorder, details of its skin manifestations merit characterization.
Histopathologic-and IgG4-related quantitative criteria adopted for T A B L E 2 Histopathologic profiles of IgG4-RSD mimics.proposed immunohistochemical criteria can be met by many non-IgG4-RSD cases. 9r results showing that both mean IgG4+ cell counts of >10 per hpf, and an IgG4/IgG ratio of >0.4 are each not specific to IgG4-RSD supports and expands the impression of prior work by Lehman et al. 19 Furthermore, our data suggest that while an IgG4/ IgG ratio of >0.4 appears more specific for IgG4-RSD than a mean IgG4+ count of >10 per hpf, application of both criteria simultaneously offers increased diagnostic rigor compared to either individual criterion alone.Despite this increased stringency, it is noteworthy that autoimmune blistering disease as a diagnostic category was able to fulfill both criteria.None of our cases met the cutaneous IgG4+ cell count threshold of >200 cells per hpf proposed by the consensus statement on IgG4-RD pathology. 8While our data supports this as a highly specific criterion, prior work has demonstrated that few, if any, cases of IgG4-RSD meet this threshold and the sensitivity may be too low to be useful.This threshold was primarily derived from a small number of case reports, and, as stated by the authors, was recommended as a supportive feature secondary in importance to more comprehensive clinicopathologic criteria. 20,21The progressive fibrosis and decreased cellularity characteristic of long-standing IgG4-RD and the numerous alternative diagnoses associated with increased tissue IgG4+ cell infiltration likely limit the utility of isolated mean IgG4+ cell count thresholds for diagnosis. 2,8,13,22ether modified skin-specific IHC criteria could differentiate IgG4-RSD from other plasma cell-rich dermatoses with higher specificity than current criteria is worth consideration.Alternative criteria for the skin have been proposed, including scaled scoring for IgG4+ cell counts, 13 but further studies are needed to understand and optimize the sensitivity and specificity of these criteria.
Histopathologic findings still form the primary basis for pathologic diagnosis according to the consensus statement criteria. 8Our results support increased specificity of the major histopathologic criteria over immunohistochemical thresholds for IgG4-RD as they are rarely seen in IgG4-RSD mimics.Prior reviews have suggested that storiform fibrosis and obliterative phlebitis occur less frequently in the skin than in other organs.Reported rates of storiform fibrosis in IgG4-RSD range from 30.3% to 76%, [11][12][13] though in our cohort, storiform fibrosis was seen in only one case of EED.The prevalence of obliterative phlebitis has been reported at 54%, 13 though in our cohort, no cases showing obliterative phlebitis were identified.While the presence of obliterative phlebitis and storiform fibrosis are highly suggestive of IgG4-RSD with supportive clinical and immunochemical features, data suggests their absence may be insufficient to rule out the disease, sacrificing sensitivity for specificity.

F I G U R E 1
Autoimmune blistering demonstrating lymphoplasmacytic infiltrates (A) H&E stain, Â40; (B) H&E stain, Â200; (C) IgG stain, Â200; (D) IgG4 stain, Â200.classically involved organ systems have been adapted to cutaneous biopsies, but their sensitivity and specificity have not yet been thoroughly evaluated.In this study, IgG4+ plasma cell counts, IgG4/IgG ratios, and histopathologic findings were evaluated for 14 plasma cellrich dermatoses identified as potential IgG4-RSD mimics to elucidate the frequency with which these mimics meet consensus histopathologic or immunohistochemical criteria.Our results demonstrate the

F I G U R E 2
Erosive pustular dermatosis demonstrating lymphoplasmacytic infiltrates at Â100, Â200.(A, B) H&E stain; (C, D) IgG stain; (E, F) IgG4 stain.This study further characterizes the specificity of current consensus diagnostic criteria for IgG4-RSD by expanding the number of analyzed skin conditions featuring lymphoplasmacytic infiltrates and assessing the frequency with which these mimic dermatoses meet histopathologic and immunophenotypic criteria.Our findings expand upon prior work that mean IgG4+ cell counts >10 per hpf and IgG4/ IgG ratios of >0.4 are not individually specific for the diagnosis of IgG4-RSD, though combining IgG4+ cell counts and IgG4/IgG ratios may offer increased diagnostic rigor.Interpretation of our findings should be tempered by study limitations, which include the small number of cases for certain diagnoses (some with only a single case), which limits the ability to detect true trends and risks observing rates of histopathological or immunohistochemical features that are not necessarily representative of the diagnosis as a whole.Overall, these results emphasize the critical importance of careful integration of clinical, serologic, histopathologic, and immunophenotypic features for accurate diagnosis of IgG4-RSD, echoing the consensus statement's 8 recommendation that immunohistochemical criteria are only relevant in the correct histopathologic context.Further studies directly comparing examples on the spectrum of IgG4-RSD with these potential histopathologic mimics should guide the development of more specific diagnostic criteria.