Development and internal validation of a predictive model for thrombocytopenia in cervical cancer patients undergoing concurrent chemoradiotherapy

Concurrent chemoradiotherapy (CCRT) enhances treatment efficacy in cervical cancer compared with radiotherapy alone but increases the risk of thrombocytopenia. This adverse effect has become a leading cause of dose reduction, treatment delays, and premature termination of CCRT, ultimately compromising patient survival. This study aimed to develop a simple and reliable model for pretherapeutic prediction of thrombocytopenia risk in patients undergoing CCRT. We retrospectively reviewed 146 cervical cancer patients treated with CCRT at the Cancer Hospital, Chinese Academy of Medical Sciences, Shenzhen Center, between April 2019 and September 2022. Patients were stratified into a thrombocytopenia group ( n = 64) and a non-thrombocytopenia group ( n = 82). Predictor variables were selected using least absolute shrinkage and selection operator (LASSO) regression, followed by multivariable logistic regression analysis. Internal validation was performed via bootstrap resampling. Chemotherapy regimen, extended-field radiotherapy, body mass index (BMI), baseline platelet count (PLT), and the percentage of bone marrow volume receiving ≥ 25 Gy (BM_V25) were identified as independent predictors of thrombocytopenia. The predictive model demonstrated good discrimination (area under the curve [AUC] = 0.840; 95% confidence interval [CI], 0.773–0.904) and calibration (Brier score = 0.162; 95% CI, 0.128–0.195). Decision curve analysis confirmed its clinical utility. After internal validation, the model maintained good performance (AUC = 0.817; Brier score = 0.176). We developed and internally validated a predictive model based on clinical and dosimetric factors that enables early effectively identification of cervical cancer patients at high risk of thrombocytopenia during CCRT, facilitating individualized treatment planning.