Abstract
OBJECTIVE: Current clinical staging of pancreatic ductal adenocarcinoma (PDAC) relies predominantly on anatomical resectability, thus limiting its prognostic utility. We developed and validated a pretreatment prognostic grading system incorporating multidimensional parameters. METHODS: Patients with histologically confirmed PDAC undergoing curative-intent pancreatectomy were retrospectively enrolled. Independent prognostic determinants of overall survival (OS) and disease-free survival (DFS), identified through multivariable Cox proportional hazards regression, provided the basis for deriving the Tianjin Prognostic Score and its corresponding risk stratification scheme. RESULTS: Resectability status, lymph node metastasis indicated by imaging, pretreatment serum CA19-9 levels, and the prognostic nutritional score (PNS) independently predicted both OS and DFS. These parameters were integrated into the Tianjin Prognostic Score for PDAC prognosis stratification. The Tianjin-Grade system, subsequently established according to this score, segregated patients into 4 discrete prognostic cohorts with significantly divergent survival outcomes. This system exhibited significantly greater discriminatory ability for prognosis than conventional serum CA19-9 and resectability criteria. Notably, patients classified as having high risk or extremely high risk derived substantial survival benefits from neoadjuvant chemotherapy (NAC), whereas those with low or intermediate risk demonstrated comparable survival outcomes regardless of NAC administration. CONCLUSIONS: The Tianjin-Grade system provides accurate pretreatment prognosis prediction in patients with PDAC through integration of anatomical and biological parameters, thus serving as a reliable tool for prognostic assessment. This system facilitates the development of personalized preoperative therapeutic strategies.