IgG4-Rich Lesions Associated With Intranasal Drug Use Can Mimic IgG4-Related Disease: A Distinct Clinicopathologic Scenario Represents a Potential Diagnostic Pitfall Open Access

As per the Columbia University (New York, New York) Institutional Review Board/Privacy Board policy, this retrospective case series does not meet the definition of “human subjects research” and is accordingly considered exempt from institutional review board oversight. Assembly of this series entailed routine clinical review of 2 analogous cases (cases 1 and 2) by a pathologist, and subsequent consultation with an otolaryngologist regarding the latter’s recollection of additional similar cases (case 3). Clinical features of each case, comprising patients’ subjective histories as well as findings from physical examination, radiologic studies, and surgical management, were compiled from the electronic medical record. Histologic morphology was evaluated by review of hematoxylin-eosin–stained slides prepared from surgical specimens. Fungal stains were performed to rule out infection. CD31, desmin, and elastic stains facilitated exclusion of vasculitis. Stains for CD3, CD20, CD138, and IgG were used to highlight T cells, B cells, and plasma cells within the inflammatory infiltrates. Polytypic composition of plasma cell populations was verified with in situ hybridization for κ and λ light chains. IgG4 abundance was assessed according to previously published protocols.^37,43  Immunohistochemical staining for IgG4 was compared to CD138 and IgG in order to visualize the subset of IgG4-positive cells among all plasma cells. For each case, 3 “hot spot” areas of highest-density IgG4 expression were identified and quantified at ×400 magnification.

Clinical features of each case are summarized in the Table, with representative imaging studies shown in Figure 1, A and B. Each patient underwent endonasal endoscopic debridement after an initial evaluation. Tissue sections sampled from the surgical specimens displayed acutely and chronically inflamed respiratory-type mucosa with squamous metaplasia, overlying seromucous glandular tissue with stromal edema. The mixed inflammatory infiltrates comprised brisk neutrophilia, eosinophilia, and plasmacytosis, with abundant IgG4-positive plasma cells (Figure 2, A through F; Table). Crystalline forms reminiscent of Charcot-Leyden crystals were present in case 2, as is frequently the case with eosinophil-rich inflammation. Mucosal ulceration was evident in cases 1 and 2, with liquefactive necrosis and devitalized debris (Figure 2, A). Case 2 also showed variably dense fibrosis, with some areas exhibiting a vaguely bandlike or fascicular pattern, apparently tracking along small-caliber vasculature (Figure 2, D). This finding, together with the prominent eosinophilia and plasmacytic infiltration, warranted consideration of EAF. The distribution of fibrosis in case 2 differed from the typical pattern of EAF, however, and this somewhat tenuous connection was accordingly insufficient for diagnosis of that entity.

According to Deshpande et al,¹  3 major histopathologic features characterize IgG4-RD: (1) dense lymphoplasmacytic inflammation; (2) fibrosis, arranged at least focally in a storiform pattern; and (3) obliterative phlebitis.¹  Additional findings associated with IgG4-RD, albeit with less sensitivity and specificity, include increased numbers of eosinophils, and phlebitis without obliteration of vascular lumina. Conversely, granulomata comprising epithelioid histiocytes and prominent neutrophilic infiltrates render a diagnosis of IgG4-RD less likely.

Comparison between these canonical findings and features of cases in the current series portrays these cases as distinct from IgG4-RD. Although these cases share some aspects in common with IgG4-RD, other definitive characteristics are absent, and notable differences exist. Cases 1 to 3 all showed marked lymphoplasmacytic infiltration, in which eosinophils but also neutrophils figured prominently. As above, the latter are generally inconsistent with IgG4-RD. Whereas case 2 demonstrated fibrosis, its configuration did not resemble the storiform pattern that is stereotypical of IgG4-RD. None of these cases exhibited phlebitis (obliterative or otherwise), which is expected in IgG4-RD, although like IgG4-RD, they also lacked epithelioid granulomata. Furthermore, patients in this series did not present with tumefactive lesions. Ultimately, despite some similarities, the pathologic processes and IgG4-rich infiltrates in these 3 cases seem to represent a different phenomenon that is not (or phenomena that are not) IgG4-RD.

This series provides the first demonstration of increased IgG4 expression in nasal lesions associated with intranasal drug use. Although these 3 cases exhibit numerous IgG4-positive plasma cells, other clinical and pathologic findings distinguish them from IgG4-RD. Since 2008, various investigators have reported 37 cases of IgG4-RD involving the nasal cavity and paranasal sinuses (counting articles published in English), with the largest cohort comprising 7 individuals. In 31 of these 37 instances, the disease manifested as a “mass,” “tumor,” “tumor-like” condition, or “polypoid lesion.”^* Authors describe the other 6 cases as thickened, edematous, and/or swollen tissue,^18,20,22  or in terms of radiologic features such as a “soft tissue shadow,”⁸  “‘lozenge’-shaped radiopacity,”⁹  or “enhancement with contrast uptake.”²⁵  Thirty-four of these reports mention fibrosis, elaborated as storiform fibrosis in 20 patients.^† The inflammatory infiltrates, characterized as dense,^‡ heavy,⁹  intense,¹²  severe,¹⁸  abundant,^22,31  florid,²⁴  and/or extensive,³¹  showed submucosal/subepithelial,^8,19,31  interlobular/periductal⁷  distribution. Plasma cells represented the dominant cell type, with^§ or without^{10,20,21,24,27,31}  lymphocytes, sometimes forming “accumulations”/aggregates^15,17,19,27  or follicles with germinal centers.^11,28  In some cases, eosinophils^{6,7,15,18,22,28,29}  and (less frequently) neutrophils^7,10,28  were also present. Histiocytes were sometimes evident,^15,28  but granulomatous inflammation was absent.^20,28  Thirteen of these lesions displayed phlebitis, usually obliterative phlebitis,^{7,15,18,19,23,28}  occasionally “nonobliterative.”^22,28 

The 3 lesions described in our cohort differ from these previously reported cases of IgG4-RD in several respects. First, these 3 patients presented with destructive/degenerative or obstructive nasal changes, rather than tumors or tumor-like conditions. Second, neutrophils figured prominently among the inflammatory infiltrates in all 3 cases, whereas these cells are generally inconspicuous in IgG4-RD. Third, although phlebitis represents a recurrent finding in IgG4-RD, this feature was absent from these 3 drug-related processes. In terms of similarities (other than abundant IgG4-positive plasma cells), case 2 showed fibrosis, albeit with inconsistent density throughout the specimen, and lacking the canonical storiform arrangement that is characteristic of IgG4-RD. Notwithstanding the similarity of infiltration comprising numerous IgG4-positive plasma cells and the vague analogy pertaining to fibrosis in 1 case, the disease (or diseases) affecting these 3 patients diverges substantially from IgG4-RD.

Some features of these cases overlap to some extent with those of EAF, which perhaps represents a manifestation of IgG4-RD. These features include inflammatory infiltrates with prominent eosinophilia, and (in case 2) fibrosis distributed in vague apposition with small-caliber vasculature. As with IgG4-RD (or IgG4-RD in general), however, the phenomenon (or phenomena) described herein seems different from EAF for several reasons. First, EAF usually presents as an obstructive mass lesion, which did not occur in our cohort. Second, EAF—as its name conveys—entails inflammatory infiltration comprising primarily eosinophils. Although eosinophils were amply present in these drug-associated lesions, plasma cells represented the dominant cell type. Third, whereas fibrosis in EAF shows a consistent and characteristic angiocentric pattern, our case 2 displayed fibrosis with variable configuration, generally more fascicular than angiocentric. Its association with vasculature was irregular and relatively loose.

Our observations correspond to other reports that demonstrate a relationship between IgG4 activity and contact with certain exogenous substances. AERD exemplifies this relationship.⁴¹  Stimulation of IgG4 production by chronic high-dose antigen exposure also underlies the effect of allergen-specific immunotherapy, a therapeutic modality for allergic asthma. This process involves repetitive subcutaneous, sublingual, or oral administration of high-dose allergen, leading to induction of, and marked increases in, allergen-specific IgG4.^44–48  Collectively, entities such as AERD and allergen-specific immunotherapy, as well as the IgG4-rich lesions described in this report, implicate IgG4 in a physiologic response to antigenic materials.

Asthma, AERD, and—as we report here—intranasal drug use entail upregulation of IgG4, but are distinct from IgG4-RD, which does not result from antigenic stimulus. Accordingly, pathologists and our patient-facing colleagues should not misinterpret drug-related lesions showing extensive infiltration by IgG4-positive cells as IgG4-RD. The distinction between this clinicopathologic scenario and IgG4-RD recalls the distinction between IgG4-RD and CRS. According to Deshpande et al, “rhinosinusitis… fall[s] outside the bounds of IgG4-related disease, despite the presence in some cases of increased numbers of IgG4+ plasma cells.”¹  Similarly, the lesions described in this report, associated with intranasal drug use, contain abundant IgG4-positive cells, but should not be mistaken for or conflated with IgG4-RD.

Several limitations apply to the current study. It comprises a small, nonconsecutive cohort of patients treated at a single institution, and does not incorporate data pertaining to serum levels of IgG4 (these values were not measured, and are consequently unavailable). Despite these shortcomings, this series conveys a noteworthy reminder regarding unusual or perplexing lesions with numerous plasma cells affecting the nasal cavity and paranasal sinuses. In this anatomic location, high density of IgG4-positive plasma cells is not always tantamount to a diagnosis of IgG4-RD, and immunohistochemical quantification of IgG4 may present a proverbial “red herring.” Additional characterization of the histologic changes that result from intranasal drug use may thus represent a worthwhile endeavor, to facilitate better understanding of this clinicopathologic condition and further refinement of its associated differential diagnosis.