Design: Retrospective immunohistochemistry-based molecular subtyping with survival analysis
Population: 42 biliary tract neuroendocrine carcinomas
Primary outcome: Transcription-factor subtype and disease-specific survival
Effect: Null type predicted shorter survival (hazard ratio 3.48, P=0.034); POU2F3 enriched in gallbladder

Neuroendocrine carcinoma of the biliary tract is rare and poorly characterised. This retrospective series of 42 cases applied the transcription-factor subtyping used in small cell lung cancer, classifying tumours by ASCL1, NEUROD1, POU2F3 or a null pattern, and linked subtype to outcome.

Subtypes were 21% ASCL1, 40% NEUROD1, 26% POU2F3 and 12% null, with the POU2F3 type enriched in gallbladder tumours. NEUROD1 tumours were more often resectable, POU2F3 more often unresectable, and disease-specific survival differed significantly by subtype, with the null type independently predicting shorter survival (hazard ratio 3.48, P=0.034). The diagnostically important point: several POU2F3 tumours lacked the conventional neuroendocrine markers chromogranin A, synaptophysin and INSM1.

For sign-out, the message is twofold. Subtyping may become a practical prognostic adjunct, and a high-grade biliary carcinoma that is negative for standard neuroendocrine markers can still be a neuroendocrine carcinoma, so consider POU2F3 before excluding the diagnosis.

42 biliary tract neuroendocrine carcinomas subtyped as ASCL1 (21%), NEUROD1 (40%), POU2F3 (26%) or null (12%).
The null subtype independently predicted shorter disease-specific survival (hazard ratio 3.48, P=0.034).
POU2F3 tumours were enriched in the gallbladder and more often unresectable.
Some POU2F3 tumours lacked chromogranin A, synaptophysin and INSM1.
Consider POU2F3 staining before excluding neuroendocrine carcinoma in a marker-negative high-grade biliary tumour.

Why it matters

It warns that relying on chromogranin, synaptophysin and INSM1 alone can miss a POU2F3 neuroendocrine carcinoma, changing how a marker-negative high-grade tumour is worked up.

The statistics, in plain English

The survival difference and hazard ratio come from only 42 rare, retrospectively collected tumours, so the prognostic signal is promising but imprecise and not yet a standard; the marker-negativity point is a practical caution regardless of sample size.