763 POU2F3 Diagnostic Validation in 1365 Epithelial and Neuroendocrine Neoplasms Highlights Low Sensitivity but High Specificity for the Diagnosis of Neuroendocrine Carcinoma

Cameron Williams; Matthew Gosse; Natasha Rekhtman; Andrew Bellizzi

doi:10.1016/j.labinv.2025.105050

Background POU2F3 has emerged as a diagnostic marker for general neuroendocrine marker weak-to-negative small cell lung cancer (SCLC). POU2F3-positive neuroendocrine carcinomas (NEC) have been subsequently described in other organs. ∼10% of SCLCs are POU2F3-positive, with one group also finding positivity in ∼20% of basaloid squamous cell carcinomas (BSCC) of lung. More recently, POU2F3-positivity was reported in 12.4% of 6064 epithelial neoplasms, with positivity typically seen in <10% of tumor cells. We performed an extensive diagnostic validation of POU2F3 before placing the test into clinical service. Design POU2F3 immunohistochemistry (clone E5N2D; 1:200; Cell Signaling) was performed on a Dako Autostainer Link 48 after high pH retrieval. Tissue microarrays (TMA; triplicate 1 mm cores) of the following 229 tumors were stained locally and at a peer lab using a clone 6D1-based assay: SCLC, 26; extrapulmonary visceral (EPV) NEC, 18; Merkel, 37; non-NEC lung metastases (mets), 67; non-NEC brain mets, 81. TMAs of an additional 1136 tumors were stained locally: paraganglioma, 128; pheochromocytoma, 109; pancreatic neuroendocrine tumor (NET), 113; duodenal NET, 51; ileal NET, 17; colon adenocarcinoma (AdCA), 108; lung AdCA, 80; clear cell Mullerian AdCA, 56; urothelial carcinoma, 87; BSCC, 98; lung SCC, 23; other SCC, 266. Intensity (0-3+) and extent (0-100%) of staining were assessed with an H-score calculated. Results are binned as follows: negative (H-score=0), rare (1-49), focal (50-99), multifocal (100-199), diffuse (>/=200). The prevalence of NEC was estimated at 4%. Results Any POU2F3-positivity was seen in 20% of tumors but it was focal, multifocal, or diffuse in only 1.3%, 1.3%, and 0.6%, respectively. Considering H-scores >/=50 as positive and <50 as negative, E5N2D and 6D1-based assays were concordant-positive in 6 (all NEC), concordant-negative in 213, and discordant (positive E5N2D/negative 6D1) in 10 (4 renal cell carcinoma, 4 endometrial AdCA, 1 breast AdCA, 1 Merkel). At least multifocal POU2F3-positivity had the following test characteristics: 6% sensitivity, 98% specificity, 14% positive predictive value; 96% negative predictive value. Tumor types with any multifocal staining are listed in the Table. Conclusions At least multifocal POU2F3-positivity is insensitive though fairly specific for NEC. Rare staining is frequent and should be ignored. We discovered clone-specific differences in staining. Expression in endometrial and renal cell carcinoma warrants further attention.

763 POU2F3 Diagnostic Validation in 1365 Epithelial and Neuroendocrine Neoplasms Highlights Low Sensitivity but High Specificity for the Diagnosis of Neuroendocrine Carcinoma

View Online

Abstract

Details

Metrics