Development and validation of an ovarian cancer risk assessment tool for first-degree relatives of patients in the Chinese population

Abstract

Objective

To develop and validate an ovarian cancer risk assessment tool for first-degree relatives of patients in the Chinese population.

Methods

A bidirectional multicenter cohort was established, including 529 probands and 3141 first-degree relatives. Cancer incidence was analyzed using the standardized incidence ratio (SIR). Significant variables were identified through Cox regression analyses and visualized via a nomogram. Model performance was evaluated using the C-index, with first-degree relatives stratified into high- and low-risk groups based on a 10 % cancer risk threshold.

Results

Among 1596 first-degree female relatives, 57 ovarian cancer cases were identified, demonstrating a significant increase in SIR (SIR = 9.19; 95 % CI, 7.03–11.83; p < 0.001). In 980 relatives with germline mutations, elevated SIRs were observed for ovarian cancer (SIR = 23.33; 95 % CI, 16.51–32.09; p < 0.001) and breast cancer (SIR = 3.56; 95 % CI, 2.46–5.00; p < 0.001). Cox regression analyses identified key risk factors, including the proband's age of onset, tumor histology, gene mutation status, family history of breast cancer, and relationship to the proband. The nomogram demonstrated good predictive accuracy, with C-indices of 0.75 (training set), 0.75 (internal validation), and 0.71 (external validation). Calibration plots and Kaplan-Meier curves confirmed strong agreement and significant differences between high- and low-risk groups (cut-off value = 2.1).

Conclusions

This study develops and preliminarily validates a risk assessment tool for first-degree relatives of ovarian cancer patients in China, utilizing accessible clinical and familial data to enable early identification of high-risk individuals.