Abstract
BACKGROUND: Epithelial ovarian cancer (EOC) is one of the deadliest gynaecological malignancies worldwide. The aim of this retrospective study was to create a predictive scoring model based on simple immunological and inflammatory parameters to predict overall survival (OS) and progression-free survival (PFS) in patients with EOC. METHODS: We obtained 576 EOC patients and randomly assigned them to the training set (n = 405) and the validation set (n = 171) in a ratio of 7:3. We retrospectively evaluated the association between PIV and OS and PFS using a novel immunoinflammatory marker, according to the optihmal treshold of PIV, we divided the patients into two different subgroups, high PIV (PIV > 254.9) and low PIV (PIV ≤ 254.9). Pan-immune Inflammatory Value (PIV) was computed as follows: neutrophil count (10(9)/L) × platelet count (10(9)/L) × monocyte count (10(9)/L)/lymphocyte count (10(9)/L). Then developed a simple score prediction model based on several independent prognostic parameters using Cox regression analysis. We used receiver operator characteristic (ROC) curves, calibration plots, and decision analysis (DCA) curves to evaluate the performance of the model. Finally, we used Kaplan-Meier curves to ensure that the model could distinguish well between low- and high-risk groups. RESULTS: There was a significant difference in survival outcomes between high PIV (PIV > 310.2) and low PIV (PIV ≤ PIV310.2) (3-year survival rates of 61.34% and 76.71%, respectively); 5-year OS, 25.21% and 51.14%, respectively; 3-year PFS, 40.90% and 65.30%; 5-year PFS, 19.33% and 39.73%, respectively). Column plots of OS and PFS were constructed using independent prognostic factors. In the training module, the 3-, 5-, and 10-year AUCs for OS and PFS column charts were 0.713, 0.796, 0.839, and 0.730, 0.799, 0.826, respectively.In the validation cohort, the 3-, 5-, and 10-year AUCs for OS and PFS column charts were 0.676, 0.803, 0.685, and 0.700, respectively, 0.754, 0.727. The calibration curves showed good agreement between predicted survival and actual observations. The decision analysis curves also showed that the current model has good accuracy and clinical applicability. 3-year OS was 61.34% and 76.71%, respectively; 5-year OS was 25.21% and 51.14%, respectively; 3-year PFS was 40.90% and 65.30%, respectively; 5-year PFS was 19.33% and 39.73%, respectively. CONCLUSIONS: We constructed and validated a PIV-based nomogram to predict OS and PFS in EOC patients, with a view to helping gynaecologists converge on oncologists in their treatment and follow-up expertise in epithelial ovarian cancer.