Development and validation of a prediction model for recurrence based on ProMisE molecular classifier and...

Objective: This study aims to develop a robust prediction model using the ProMisE molecular classification and the prognostic immune-inflammatory-nutritional score to predict recurrence in stage I-III endometrial cancer, thereby enabling risk stratification of high-risk patients. Methods: The clinical data of 582 patients (365 in the training cohort and 217 in the validation cohort) were collected from multiple large cancer centers from patients with stage I-III endometrial cancer who underwent surgical resection between August 2019 and February 2022. Cox proportional hazards regression analysis was used to identify the risk factors for recurrence-free survival (RFS). The concordance index (C-index), area under the receiver operating characteristic (ROC) curves, calibration plots, and decision curve analyses (DCA) were used to assess discrimination and clinical utility of the model. Results: Patients with a hemoglobin, albumin, lymphocyte, and platelet (HALP) score ≤ 31.70 tended to have lower BMI (P = 0.017), advanced FIGO stage (P = 0.016), deep myometrial invasion (P < 0.001), and higher serum Ca125 levels (P < 0.001). Multivariate Cox regression analysis showed that age, FIGO stage, grade, LVSI, Ca125, ProMisE molecular subgroup, HALP score, and adjuvant therapy were independent prognostic factors for RFS in patients with endometrial cancer. A nomogram for predicting RFS was established, and patients were stratified into high- and low-risk groups based on the RFS model. Conclusions: The preoperative HALP score serves as a reliable predictor of RFS in endometrial cancer. A nomogram combining the HALP score, ProMisE molecular subtyping, and clinical parameters can assist clinicians in identifying high-risk patients for recurrence. These patients may benefit from early triage and more intensive monitoring.