Immunohistochemical characterization of peritoneal inclusion cysts with squamous metaplasia.

The authors received no financial support for the study.

We identified 10 PICs with squamous metaplasia (Table  1 ). Four were from our institution, and 6 were from other institutions. The average age of the patients was 51 years (age range 26 to 73); there were 8 females and 2 males. The surgical indication for five was abdominal or pelvic pain. Four others had incidental PICs found during surgeries for gynaecologic lesions, including neoplastic and non‐neoplastic conditions. The size ranged from 1.3 to 10.8 cm. The median size for symptomatic PICs ( n  = 5) was 5 cm and 3.5 cm for incidental PICs ( n  = 4). Half of the cases were unilocular cysts ( n  = 5), and half were multilocular ( n  = 5). Two of the four in‐house patients had at least 1 year of follow‐up. Neither had a recurrence of cysts, and both were doing well clinically. Demographics, clinical presentation and size of PICs LLQ, Left lower quadrant; RLQ, Right lower quadrant; RUQ, Right upper quadrant. On histology, all our cohort's PICs ( n  = 10) had diffuse squamous metaplasia (Figure  1A ). The squamous lining comprised 2–3 to 8–10 layers of cytologically bland squamous cells (Figure  1B ). The serosal surfaces were lined by a single layer of cytologically bland mesothelial cells (Figure  1B ). A full IHC panel was performed in 9 of 10 cases. A limited IHC profile (including claudin 4, calretinin, WT‐1 and p40) was recorded in one consultation case for which tissue blocks were unavailable. Immunophenotype of PICs with squamous metaplasia. ( A ) Low power magnification (H&E stain) shows squamous lined cysts with variable cystic wall thickness. ( B ) Higher power demonstrates that the squamous lining shows no cytologic atypia (H&E stain). This H&E stain shows the corresponding area of cyst lining with squamous metaplasia for immunohistochemical stains (IHCs) in (C–I). The black arrows indicate the serosal side lined by a single layer of benign mesothelial cells, whereas the asterisks indicate cyst lining with diffuse squamous metaplasia. By IHC, the cyst lining is diffusely positive for claudin 4 (with skipping of the basal layer; C), cytokeratin cocktail (D), WT1 (F), CK5/6 (H) and p40 (I) with focal staining for D2‐40 (limited to the basal layer; G) and negative staining for calretinin (E). In contrast, the serosal benign mesothelial lining (indicated by arrows) is diffusely positive for cytokeratin cocktail (D), calretinin (E) and WT1 (F) with focal and weak staining for D2‐40 (G) and negative staining for claudin 4 (C), CK5/6 (H) and p40 (I). All 10 cases showed diffuse positivity for claudin 4 in squamous lining cells with skipping of the basal layer, contrasting with serosal surface cells that were negative for claudin 4 (Figure  1C ). Both the squamous cystic lining and serosal mesothelial cells were diffusely positive for cytokeratin cocktails (Figure  1D ). For 5 cases, other epithelial markers, including MOC31, BerEP4 and B72.3, were performed at the time of their diagnosis, which showed either focal and weak staining or negative staining of internal squamous and surface mesothelial cells. All 10 cases were negative for calretinin in the cystic squamous lining, whereas the serosal mesothelial cells showed strong cytoplasmic and nuclear staining as expected (Figure  1E ). WT‐1 was positive with variable staining patterns in squamous lining cells. Most cases ( n  = 9) showed full thickness variable staining intensity to WT‐1 with accentuated staining in the basal layer (Figure  1F ); only 1 case showed diffuse staining throughout the squamous lining without a definitive basal accentuation pattern. D2‐40 staining performed in 9 cases showed immunoreactivity limited to the basal layer in 5 cases (Figure  1G ), diffuse staining with variable intensity in 3 cases and equivocal staining in 1 case. Squamous markers (CK5/6, p40/p63) were strongly and diffusely positive throughout the cystic lining for all cases ( n  = 10) tested (Figure  1H,I ). While CK5/6 is also regarded as a mesothelial marker, it is used here as a squamous marker in this context of notable squamous differentiation. IHCs for PAX‐8 were negative in 7 cases, and in two, rare cells with weak, non‐specific staining were shown. The IHCs performed for biomarkers associated with mesotheliomas, including BAP1, MTAP and merlin, showed retained expression in all cases ( n  = 9) tested, further supporting a benign process.

This is the first study to characterize the immunophenotype of PICs with squamous metaplasia using a comprehensive IHC panel. The cyst lining cells in these cases exhibit a squamous epithelial phenotype with diffuse expression of claudin 4 and p63/p40 in the suprabasal cells that also show consistent co‐expression of mesothelial marker WT1 in all cases. In contrast, the basal layer displays a mesothelial phenotype, demonstrating immunoreactivity to WT1 and D2‐40 and lacking expression of claudin 4. Calretinin showed complete loss of expression in both the basal and parabasal cells, arguing that during the process of squamous metaplasia of the mesothelial lining of PICs, calretinin is the first mesothelial marker that loses expression. Knowing the detailed immunophenotype of PIC with squamous metaplasia is helpful for correctly recognizing this lesion, especially when the process of squamous metaplasia in PICs is diffuse and obscures the conventional mesothelial cyst lining. The distinct differential expression of epithelial and mesothelial markers in the suprabasal and basal cells further supports the conclusion that this is likely a ‘reactive’/metaplastic rather than a neoplastic process. This conclusion is further supported by the bland cytologic features of the squamous cyst lining with no evidence of stromal invasion and preserved expression of commonly used biomarkers for which absence of staining is linked to a subset of mesotheliomas (BAP1, MTAP and merlin). All these features argue against the possibility of mesothelioma with squamous differentiation, a rare phenomenon that has been documented only in case reports. 13 , 14 , 15 Interpretation of the immunostains requires careful correlations with the morphologic features, especially for PICs with thin layers of squamous metaplasia. In this scenario, a thin cystic squamous lining could be confused with the thin layer of mesothelial cells lining the serosa of PICs. Knowing which thin layer is the cystic squamous lining and which one is the layer of mesothelial cells lining the serosa is important for accurate interpretation of the IHCs. Lastly, although our numbers are small and follow‐up is incomplete, our study confirms previous observations of a female predominance and a benign clinical course in patients with PICs. 1 , 8 , 16 One of the study's limitations is that we lacked clinical follow‐up in most of these cases to determine how often PICs with squamous metaplasia recur. In conclusion, our study demonstrates that PICs with squamous metaplasia express a unique mixed epithelial and mesothelial immunophenotype. The basal layer shows an immunophenotype consistent with mesothelial cells, and the suprabasal layers show an immunophenotype consistent with epithelial/squamous cells with co‐expression of mesothelial markers. Characterizing the immunophenotype of PICs with squamous metaplasia helps identify these benign lesions accurately.

Benign mesothelial‐lined cysts are commonly found in the peritoneum. 1 They are more likely to be found in women and can present in a wide age range. 2 These cysts can be solitary or multifocal and can be unilocular or multilocular in appearance. Various names have been used, such as benign multicystic mesothelioma, cystic mesothelioma and multilocular peritoneal inclusion cyst (PIC). 3 , 4 When describing these benign cysts, having ‘mesothelioma’ as part of the designation can be confusing. For this reason, the 5th edition of the World Health Organization (WHO) Classification of Female Genital Tumours does not recommend using the word ‘mesothelioma’, and PIC is the preferred term. 5 While PIC is regarded as a benign process, recurrence has been reported, raising the possibility that PIC is a neoplastic process. 6 , 7 Most published reports have concluded that PICs are reactive, non‐neoplastic lesions, 6 , 8 , 9 further supported by a recent comprehensive molecular study of PICs demonstrating no clonal genetic alterations. 9 Peritoneal inclusion cysts are known to be lined by a single layer of benign mesothelial cells. On occasion, PICs can develop squamous metaplasia. 1 , 6 , 7 Squamous metaplasia has been reported in the peritoneal mesothelial lining as a response to chronic irritation in patients undergoing peritoneal dialysis or abdominal surgery. 10 , 11 , 12 The exact mechanisms triggering squamous metaplasia in PICs are unknown. While unaltered PIC lining has a mesothelial immunophenotype, to our knowledge, our study is the first to describe the immunophenotype of PICs with squamous metaplasia. We aim to understand whether the squamous metaplasia process would change the lining cells' immunophenotype from mesothelial to epithelial or result in a mixed immunophenotype. Characterization of the immunophenotype of PIC with squamous metaplasia is helpful in correctly recognizing this lesion.

There are no financial disclosures to report.

This study was approved by the Institutional Review Board at the University of Michigan (#HUM0184538). We collected surgical excisions of PICs with squamous metaplasia that were received from February 2021 to August 2024. We also included cases sent to our extramural consultation service during this period. Information was collected through our electronic medical system (Epic Systems, Madison, WI) or materials provided by the referring institutions. We collected the following: age, sex, clinical presentation, prior imaging studies, gross and intraoperative findings, surgical pathology diagnosis and immunohistochemical (IHC) results at the time of diagnosis. All available slides were re‐reviewed by at least two members of our team (HC‐Y, TH). The IHC panel included one epithelial marker (claudin 4), three mesothelial markers (WT‐1, D2‐40 and calretinin), two squamous markers (CK5/6 and p40 or p63), three biomarkers known to be associated with mesotheliomas (BAP1, MTAP, merlin) and PAX‐8. We also collected data on additional epithelial markers (MOC31, B72.3 or Ber‐EP4) if they were performed at the time of diagnosis. Detailed information for antibodies used is as follows: Claudin 4 (Abcam, rabbit monoclonal EPRR17575), calretinin (Ventana, rabbit monoclonal SP65), WT‐1 (Ventana, mouse monoclonal 6F‐H2), D2‐40 (BioLegend, mouse monoclonal D2‐40), CK5/6 (Ventana, mouse monoclonal D5/16B4), p40 (Ventana, mouse monoclonal BC28), BAP1 (Santa Cruz Biotechnology, mouse monoclonal C‐4), Merlin (Cell Signalling, rabbit monoclonal D3S3W) and MTAP (Santa Cruz Biotechnology, mouse monoclonal 42‐T) and PAX‐8 (Cell Marque, rabbit polyclonal). All the IHC stains performed at our institution were performed on an automated Ventana BenchMark UTRA stainer.