Extranodal Extension Adds Prognostic Information Beyond Nodal Count in Node‐Positive Periampullary and Pancreatic Head Carcinoma: A Hypothesis‐Generating Analysis of 304 Consecutive Pancreatoduodenectomies

BACKGROUND: Count-based nodal staging classifies periampullary and pancreatic head carcinoma by the number of involved nodes. Because extranodal extension (ENE) occurs only in node-positive disease, its prognostic value must be judged against node-positive patients without ENE. We examined whether ENE adds prognostic information beyond nodal count among node-positive patients after pancreatoduodenectomy. METHODS: We analysed 304 consecutive pancreatoduodenectomies (October 2019-May 2025; median lymph node yield 24). The primary analysis was restricted to node-positive patients (n = 125), with node-positive ENE-negative patients as reference, using Cox regression adjusted for N2 versus N1, perineural invasion, age and pancreatic head primary. Sensitivity analyses added pathological covariates and adjuvant therapy, applied a 90-day landmark, and excluded non-adenocarcinoma histology. RESULTS: ENE was identified in 26 patients (8.6%), constituting 20.8% of node-positive cases. Among node-positive patients, 2-year overall survival was 55.2% with ENE versus 80.1% without (log-rank p < 0.001). ENE was independently associated with mortality (HR 3.23, 95% CI 1.56-6.67; p = 0.002), whereas N2 versus N1 (HR 1.15; p = 0.69) and perineural invasion (HR 1.00; p = 0.99) were not. The association persisted in all sensitivity analyses (HR range 2.49-3.14), including the 90-day landmark (HR 2.97) and adenocarcinoma-only (HR 2.49) cohorts. CONCLUSIONS: Among node-positive patients, ENE was associated with survival beyond nodal count. These findings are hypothesis-generating; they support routine synoptic reporting of ENE and warrant multi-centre validation before any staging implication is considered.