# Development and validation of a risk prediction model for overall survival in cervical cancer patients under 50: a prospective cohort study in southwest China

## TL;DR

This study developed a reliable survival prediction tool for cervical cancer patients under 50 in southwest China, using clinical and lab data to guide personalized treatment.

## Contribution

A novel nomogram for predicting overall survival in cervical cancer patients under 50, validated in a Chinese cohort and outperforming FIGO staging.

## Key findings

- The nomogram included pathology, FIGO staging, treatment, β2-microglobulin, NLR, and albumin as independent risk factors.
- The model achieved concordance indices of 0.818 (training) and 0.747 (validation), with strong calibration and AUC-ROC values.
- The nomogram outperformed the FIGO staging system in predictive accuracy and net benefit.

## Abstract

Accurately predicting the prognosis of cervical cancer in younger patients is increasingly important due to the rising incidence of the disease in China and the growing number of cases among individuals under 50. This study aimed to develop a nomogram to predict overall survival (OS) in cervical cancer patients under 50 in southwest China.

Clinicopathological and follow-up data for cervical cancer patients under 50 were prospectively collected as part of an ongoing longitudinal cohort study at Chongqing University Cancer Hospital between January 1, 2015, and May 31, 2019. A training cohort (n = 703) and a validation cohort (n = 301) were randomly selected. Variables associated with OS were assessed using a Cox regression model. Multivariate analysis was used to construct the nomogram and identify independent prognostic factors. The model’s performance was evaluated using decision curve analysis (DCA), calibration curves, area under the receiver operating characteristic curve (AUC-ROC), and the concordance index.

The final model identified pathology, International Federation of Gynecology and Obstetrics (FIGO) staging, treatment, β2-microglobulin, neutrophil-lymphocyte ratio (NLR), and albumin as independent risk factors for OS in patients under 50. The concordance index for OS was 0.818 in the training cohort and 0.747 in the validation cohort. Calibration curves in both cohorts showed strong agreement between predicted and observed survival probabilities. In the training cohort, AUCs for 1-, 3-, and 5-year OS were 0.851, 0.847, and 0.816, respectively; in the validation cohort, they were 0.810, 0.733, and 0.730. Compared to the FIGO staging system, the nomogram demonstrated superior predictive accuracy and net benefit, as shown by the net reclassification index (NRI) and DCA.

The nomogram provides a reliable tool for predicting overall survival in cervical cancer patients under 50, supporting more personalized treatment planning.