Clinical risk factors and prediction model for gonadal impairment in pediatric patients following hematopoietic stem cell transplantation

Gonadal impairment constitutes a prevalent endocrine complication in pediatric patients following hematopoietic stem cell transplantation (HSCT), yet no prediction model is currently available to assess its risk. This study aimed to investigate clinical risk factors and develop a prediction model for post-HSCT gonadal impairment. This retrospective cohort included 196 children undergoing HSCT (2016–2023), randomly allocated to derivation and validation groups at a 7:3 ratio. Gonadal impairment served as the primary endpoint. Key predictors were identified through the least absolute shrinkage and selection operator regression and multivariable Cox analysis. A nomogram was developed and internally validated using bootstrap resampling. Performance metrics included time-dependent receiver operating characteristic curves, calibration curves, and decision curve analysis. The cumulative incidence of gonadal impairment was 56.6% (median onset: 37.4 months, 95% CI 31.7–43.0). Independent risk factors included older age at HSCT (HR = 1.239, 95% CI 1.148–1.337), female sex (HR = 4.987, 95% CI 3.040–8.182), and cyclophosphamide equivalent dose (CED) of conditioning regimen ≥ 6300 mg/m2 (HR = 1.938, 95% CI 1.217–3.087). The nomogram exhibited strong discrimination (area under the curve > 0.75 for 1-, 3-, and 5-year predictions) and calibration, effectively stratifying high-risk patients at a cutoff score of 83.734. This study identified older age at HSCT, female sex, and high-dose CED as critical predictors of gonadal impairment. The validated nomogram enables risk stratification, guiding intensified surveillance for at-risk children. Early recognition using this tool may facilitate timely interventions to preserve endocrine health in HSCT survivors.