Development and validation of a novel nomogram to predict overall survival in patients with 2018 FIGO IB cervical cancer

Abstract

Abstract Objective: To develop a nomogram for Overall Survival (OS) prediction in 2018 FIGO staging IB cervical cancer patients. Methods and Materials: A total of 1185 2018 FIGO staging IB cervical cancer patients were analyzed. They were randomly divided into the training set (n=889,75%) and the validation set (n=296,25%). Data collected included demographics, clinical characteristics, and follow-up information. Missing data were estimated by multiple imputations. LASSO and Cox analysis was used to evaluate risk factors of OS rate. Nomograms were developed and internally validated. Three risk groups were identified based on the nomo-score of the patient. Results: The 5-year OS of IB1 patients was 96.0%, for IB2 patients was 93.1%, and for IB3 patients was 90.3%. We established a multivariate Cox regression model, including 2018 FIGO staging, age, Lvsi and deep stromal invasion. The c-index and AUC for 2-, 3-, and 5-year OS revealed that the nomogram model significantly outperformed the 2018 FIGO staging. Furthermore, nomograms performed well in clinical applicability calibration and decision curve analysis. The OS of patients differed significantly (P<0.001) across the three risk groups stratified by the nomo-score. Conclusions: This study developed a novel nomogram to predict overall survival in 2018 FIGO IB cervical cancer patients. The nomogram had significant discrimination ability and prediction accuracy, making it an excellent supplement to FIGO staging.