Preoperative Risk Assessment of Pancreatic Fistula Impact on Adjuvant Chemotherapy Delivery After Pancreatoduodenectomy

Background: Up to 40% of patients undergoing pancreatoduodenectomy (PD) for resectable pancreatic ductal adenocarcinoma do not receive adjuvant chemotherapy (aCT). This study aimed to evaluate the impact of postoperative pancreatic fistula (POPF) on aCT delivery and timing and to explore how preoperative variables influence these outcomes according to the occurrence of a POPF. Methods: This multicenter retrospective study included patients from 25 pancreatic centers. Propensity score matching was performed based on anatomical, biological, and conditional variables. Multivariable regression analyses were used to identify independent predictors of aCT omission and delay. Results: Among 1590 patients, 267 (16.8%) developed a POPF. Overall, aCT was administrated in 1,146 patients (72.1%) with a median time to first dose delivery of 56 days (26). After matching, POPF was associated with a significantly lower likelihood of aCT delivery (p < 0.001) and a significant delay in its initiation (p < 0.001). Independent predictors of aCT omission were age ≥ 70 (odds ratio [OR] 2.480, 95% confidence interval [CI] 1.439-4.274; p < 0.001), chronic renal failure (OR 4.554, 95% CI 1.320-15.708; p = 0.016), and chronic obstructive pulmonary disease (OR 2.775, 95% CI 1.021-7.546; p = 0.045) when POPF occurred. In the absence of POPF, apart from age ≥ 70, venous contact (OR 1.574, 95% CI 1.114-2.224; p = 0.010) and tumor size > 20 mm (OR 0.713, 95% CI 0.523-0.972; p = 0.032) were predictors of aCT delivery. Conclusions: Postoperative pancreatic fistula is a key driver of aCT delivery after pancreatoduodenectomy. Its interaction with patient frailty highlights the need for preoperative risk assessment to better select candidates for upfront surgery in resectable pancreatic ductal adenocarcinoma.