JMIR Publications

ABSTRACT

Background:

Progression-free survival (PFS) is a critical endpoint in oncology, yet real-world applications of individualised, explainable machine-learning (ML) predictions remain limited.

Objective:

This study aims to develop and validate explainable ML models to predict PFS using retrospective data from a national prostate cancer cohort in Brunei Darussalam.

Methods:

We analysed a retrospective cohort of 212 patients (478 longitudinal observations) treated at the Brunei Cancer Centre (January 2018 to December 2024). Clinical, laboratory, and treatment data were harmonised, with missing values imputed via Extremely Randomised Trees. Longitudinal patterns were captured using a recurrent autoencoder to generate latent representations. We compared four modelling approaches: Cox Proportional Hazards (CPH), Random Survival Forests (RSF), Gradient Boosting Survival (GBS), and Deep Neural Network Survival models. Performance was evaluated using time-dependent AUC, Harrell’s C-index, and Integrated Brier Score (IBS), with SHAP (Shapley Additive exPlanations) used for interpretability.

Results:

RSF demonstrated improved discriminative performance and balanced calibration, achieving a C-index of 0.906 and AUCs of 0.941 at both 4 and 5 years (IBS = 0.0698). In contrast, the traditional CPH model performed poorly (C-index 0.531; AUC 0.706 at 4 years). Deep survival (AUCs of 0.941 at 4 years and 0.941 at 5 years, C-index 0.719, IBS=0.0590) and GBS (AUCs of 0.765 at 4 years and 0.833 at 5 years, C-index 0.844, IBS=0.0887) models showed moderate performance. SHAP analysis identified sodium (Na), alanine aminotransferase (ALT), MCH, platelet count, and specific treatment categories as key drivers of increased progression risk.

Conclusions:

Tree-based ensemble approaches, particularly RSF integrated with SHAP, offer high accuracy for personalised risk stratification in prostate cancer. These findings highlight the potential of explainable ML to enhance clinical decision-making. However, external validation in larger multi-institutional, multi-omics dataset is required before routine clinical implementation.