Abstract
Neoadjuvant therapy (NAT) followed by resection is recommended for borderline resectable or locally advanced pancreatic ductal adenocarcinoma (PDAC), yet a substantial proportion of patients derive no survival benefit. This multicenter retrospective study (2020-2024) included 529 patients undergoing NAT and pancreatectomy, divided into derivation (n = 375) and external validation (n = 154) cohorts. Futile resection, defined as death or recurrence within six months, occurred in 24.4% of patients. Multivariable logistic regression identified elevated post-NAT direct bilirubin, larger post-NAT tumor size, ASA class III, lower post-NAT white blood cell count, decreased albumin, and percent change in CA19-9 as independent predictors. The model demonstrated good discrimination (AUC = 0.815 in derivation; 0.763 in validation) and calibration. An online calculator was developed to facilitate individualized risk assessment. This model may support preoperative decision-making and reduce non-beneficial resections in NAT-treated PDAC patients.