Trans-anastomotic stent does not impact the predictive value of drain fluid amylase after pancreatoduodenectomy

BACKGROUND: Drain fluid amylase (DFA) guides risk stratification and drain management after pancreatoduodenectomy (PD), but its interpretation in patients with trans-anastomotic stents (TS) remains poorly defined. METHODS: We retrospectively analyzed multicenter PD data, focusing on TS-managed patients. DFA was measured on postoperative days (POD) 1, 3, and 5. The primary outcome was clinically relevant postoperative pancreatic fistula (CR-POPF), with major morbidity (Clavien-Dindo ≥ IIIa) as a secondary outcome. Discrimination was evaluated by adding log-transformed DFA to a baseline model including high-risk PD and center. RESULTS: Among 1156 patients, 464 (40.1%) received TS. The baseline model yielded an AUC of 0.645, which increased to 0.748 with POD1 log(DFA) and to 0.786 with POD3 log(DFA); POD3 was not significantly different from POD1 (P = 0.056). In complete-case analyses including POD5, POD5 further improved discrimination compared with POD1 (P = 0.008), but was not statistically different from POD3 (P = 0.126). Although median DFA values were higher in TS patients (reflecting a greater prevalence of high-risk PD), DFA dynamics were similar to those of non-TS patients. DISCUSSION: Early DFA remains informative with TS. Integrating DFA with pre- and intraoperative risk features supports serial risk assessment in the early postoperative period.