Time-Dynamic AI Models to Predict Quality of Life in Patients With Breast Cancer: Development and Validation Study...

Patients with breast cancer often experience health-related quality of life (HRQoL) impairments that remain difficult to predict on an individual level. Prediction models can aid in understanding individual survivorship trajectories. However, current prognostic models are based on fixed intervals, limiting their utility in clinical follow-up schedules.

This study aimed to develop and externally validate time-dynamic machine learning (ML) models that predict clinically relevant HRQoL impairments in nonmetastatic patients with breast cancer.

Using the pooled multicohort EORTC (European Organisation for Research and Treatment of Cancer) BALANCE (big data in patients with breast cancer) dataset (n=6316) containing repeated HRQoL measurements (EORTC QLQ [Quality of Life Core Questionnaire]-C30), we constructed over 70,000 patient assessment pairs. ML algorithms were trained using the earlier HRQoL assessment and clinical data to predict dichotomized impairments in QLQ-C30 domains at the later assessment between 2 weeks and 5 years ahead, reflecting the range of follow-up intervals available in the dataset. The best performing model was determined via the area under the receiver operating characteristic curve in the internal validation, and externally validated in an independent cohort of the BALANCE dataset, in which the calibration and predictive performance in risk groups (patients: postmenopause, with financial difficulties, with obesity, with 2 or more comorbidities, with lower educational status, and with frailty) were also evaluated.

ML models showed good discrimination (area under the receiver operating characteristic curve 0.64-0.84) across most domains, especially for persistent symptoms such as fatigue, financial difficulties, or functioning scales. Gradient boosting models performed best, but tended to be overconfident, with poor calibration for low-prevalence symptoms such as diarrhea or constipation. Model performance varied by risk group (eg, lower education and frailty), though no group consistently performed poorly. Performance remained stable across time windows, with prior HRQoL being the strongest predictor at the respective scale level, while clinical variables such as the type of treatment were less important for prediction.

Time-dynamic ML models can support personalized HRQoL prediction in breast cancer care. Future improvements should focus on calibration and fairness to enable equitable, clinically meaningful implementation.