Endocrine Abstracts

JOINT1335

Introduction: Bilateral macronodular adrenocortical disease (BMAD), formerly known as PBMAH for primary bilateral adrenocortical hyperplasia is an uncommon disease responsible for about 2% of overt Cushing’s syndrome. The disease is clinically, biologically and radiologically heterogeneous. Our team has described four microscopic subtypes based on macronodule architecture and cell type proportion (clear, eosinophilic and oncocytic). Subtypes 1 and 2 correlate with the presence of pathogenic variants in ARMC5 and KDM1A genes, respectively. ARMC5 inactivation leads to accumulation of the A subunit of RNA polymerase II (POLR2A). We have also shown that bi-allelic inactivation of KDM1A occurs via loss of heterozygosity with deletion of the short arm of chromosome 1. The aim of this study is to describe the immunohistochemical and in situhybridization characteristics of BMAD subtypes.

Patients and methods: Immunohistochemistry and in situhybridization were performed on 4 patients of each subtype from the cohort of 35 patients previously described at our center. We performed immunohistochemistry (IHC) using antibodies targeting alpha inhibin, DAB2, HSD3B1, HSD3B2, CYP11B1, CYP11B2, CYP17A1, KDM1A and POLR2A proteins. Fluorescent in situhybridization (FISH) was performed with 1p36/1q25 commercial probe.

Results: In each subtype, HSD3B2 preferentially stains clear cells whereas CYP17A1 stains eosinophilic cells contrary to cortisol producing adenomas (CPA) that co-express CYP17A1 and HSD3B2 (Kubota Hum Pathol 2016). HSD3B2 uniformly stains clear cells only in patients with a pathogenic variant of ARMC5, in BMAD. HSD3B1 and CYP11B1 stain all cell types in each subtype. In subtype 1 (ARMC5), there is a population of columnar eosinophilic cells expressing DAB2 without expressing CYP11B2. In patients with a pathogenic variant of ARMC5, contrary to other subtypes, all nodular cells strongly express POLR2A in contrast with non-nodular adrenal gland. In subtype 2 (KDM1A), alpha inhibin is highly expressed in eosinophilic cells, and KDM1A immunoexpression is lower than in the adjacent adrenal gland. Similarly, this subtype is the only one in which a deletion of the short arm of chromosome 1 is demonstrated by the 1p36/1q25 FISH probe.

Conclusion: The absence of HSD3B2 and CYP17A1 co-expression appears to distinguish BMAD from CPA, and these IHCs could provide a diagnostic marker for BMAD. HSD3B2 and POLR2A are strongly correlated with ARMC5 pathogenic variants and, KDM1A IHC and 1p36/1q25 FISH probe distinguish KDM1A altered BMAD. These markers could be used to guide genetic investigations or to confirm the pathogenic nature of germline variants of undetermined significance.