Adjunctive diagnostic utility of POU2F3 in predominantly pulmonary high-grade neuroendocrine carcinomas with limited or discordant conventional marker expression: a case-control study.

BACKGROUND: The diagnosis of high-grade neuroendocrine carcinoma may be challenging when conventional neuroendocrine markers show limited or discordant expression. POU2F3 is associated with a neuroendocrine-low subset of small cell carcinoma, but its incremental contribution in routine diagnostic practice remains incompletely defined. OBJECTIVE: To evaluate the adjunctive diagnostic utility of POU2F3 in predominantly pulmonary high-grade neuroendocrine carcinomas, particularly in cases with limited or discordant conventional marker expression. METHODS: This retrospective case-control study included 140 neuroendocrine carcinomas and 100 poorly differentiated non-neuroendocrine controls. POU2F3 expression was assessed in biopsy and resection specimens and analyzed according to histological subtype, chromogranin A status, marker-discordant status, and diagnostic-panel interpretation. Diagnostic performance was reported with 95% confidence intervals, and paired changes after incorporation of POU2F3 were evaluated using the exact McNemar test. RESULTS: POU2F3 was positive in 28 of 140 neuroendocrine carcinomas and 2 of 100 controls. Positivity was more frequent in chromogranin A-negative/low than in chromogranin A-positive neuroendocrine carcinomas (34.5% vs. 10.6%; odds ratio, 4.46; 95% confidence interval, 1.84-10.82; P = 0.001). Positivity was identical in biopsy and resection specimens (20.0% vs. 20.0%; P = 1.000). POU2F3 alone showed 20.0% sensitivity and 98.0% specificity. Incorporation of POU2F3 increased panel sensitivity from 80.7% to 91.4% and accuracy from 87.1% to 93.3%, while specificity remained 96.0% (exact McNemar P < 0.001). CONCLUSION: POU2F3 has limited stand-alone sensitivity but provides incremental diagnostic support in selected neuroendocrine carcinomas with limited or discordant conventional marker expression. It should be used selectively and interpreted together with morphology and established neuroendocrine markers.