Independent risk factors for clinically relevant postoperative pancreatic fistula after robotic pancreatoduodenectomy: the protective role of the modified Blumgart pancreatojejunostomy in a single surgeon retrospective cohort of 220 cases

Background: Clinically relevant postoperative pancreatic fistula (CR-POPF) remains the primary driver of morbidity after pancreatoduodenectomy (PD). Although robot-assisted PD (RPD) is increasingly adopted, determinants of CR-POPF in this setting—particularly during program implementation—are not well defined. This study aimed to identify independent risk factors for CR-POPF in a large cohort, with a focus on anastomotic technique and the learning curve.

Methods: We retrospectively analyzed 220 consecutive RPDs performed by a single surgeon between October 2015 and December 2024 at the University Hospital of Orléans. Univariate and multivariate logistic regression analyses were used to identify predictors of CR-POPF. The learning curve was assessed using cumulative sum (CUSUM) analysis of operative time, estimated blood loss, CR-POPF rates, risk-adjusted CUSUM (RA-CUSUM) for severe morbidity (Clavien-Dindo grade ≥ III) and RA-CUSUM for textbook outcomes (TOs).

Results: CR-POPF occurred in 17% (n=36) of patients. In multivariate analysis, the updated alternative pancreatic fistula risk score (ua-FRS) and conventional pancreatojejunostomy (versus the modified Blumgart technique) were independent risk factors; conventional pancreatojejunostomy conferred the highest risk (odds ratio 3.06; 95% confidence interval: 1.12–8.40; P=0.03). The CUSUM curve for CR-POPF, RA-CUSUM for Clavien-Dindo ≥ III and RA-CUSUM for TOs revealed a consistent inflection around the 65th case, coinciding with the near-exclusive adoption of the modified Blumgart anastomosis. No significant learning effects were observed for operative time or blood loss.

Conclusions: In RPD, the modified Blumgart pancreatojejunostomy is independently protective against CR-POPF. Standardizing this technique significantly reduces fistula rates and may shorten the learning curve. These findings support adopting the modified Blumgart reconstruction as the default anastomotic strategy during RPD program implementation.