Abstract / Summary
Early recurrence is a major challenge for patients with pancreatic ductal adenocarcinoma (PDAC) after radical resection. Early identification of high-risk patients is crucial for formulating personalized follow-up strategies and improving prognosis. Based on a meta-analysis of 33 cohort studies involving 13,009 patients with PDAC who underwent radical resection, an early recurrence risk scoring system was established. Risk factors and their odds ratios (OR) were extracted, and only factors with statistical significance in the pooled analysis were included in the model. The validation cohort comprised patients who underwent radical resection for PDAC from a single center in China, and the optimal cutoff value was determined using the ROC curve. The Newcastle-Ottawa Scale was used to assess the risk of bias. Significant risk factors included: preoperative CA19-9 > 37 U/mL, postoperative CA19-9 > 37 U/mL, tumor size > 3.0 cm, poor tumor differentiation (G3-4), microvascular invasion, positive resection margin (R1), lymph node metastasis, no adjuvant chemotherapy, and Charlson Age-Comorbidity Index ≥ 4. The final scoring system achieved an area under the receiver operating characteristic curve of 0.753 (95% CI: 0.684–0.822) in external validation, with an optimal cutoff value of 18.5 (sensitivity 0.865, specificity 0.615). The calibration curve demonstrated good agreement between predicted probabilities and observed actual probabilities. This scoring system demonstrated preliminary predictive value in the single-center retrospective validation and is expected to serve as a reference tool to assist postoperative decision-making. However, its clinical utility still needs to be further verified in multi-center, prospective cohorts.