CAST: Time-Varying Treatment Effects with Application to Chemotherapy and Radiotherapy on Head and Neck Squamous Cell Carcinoma

2025-05-09

AI Summary

This paper addresses the challenge of modeling dynamic causal effects in medical survival data. We propose CAST, the first causal framework capable of learning a continuous-time treatment effect function—overcoming the limitation of existing methods (e.g., causal survival forests) that estimate effects only at discrete time points. CAST integrates parametric basis functions with nonparametric survival forests, specifically designed for right-censored survival data, and supports both individual- and population-level time-varying effect estimation. Evaluated on the RADCURE cohort (2,651 patients with head-and-neck squamous cell carcinoma), CAST is the first method to systematically characterize the onset, peak, and decay phases of chemotherapy and radiotherapy effects. It significantly improves medium- to long-term survival prediction (C-index +0.07) and provides interpretable, quantitative temporal guidance for personalized timing of clinical interventions.

Technology Category

• Machine Learning: Causal Learning
• Reasoning under Uncertainty: Causality
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• User Modeling, Personalization and Recommendation: Studies of user behavior, including longitudinal effects of personalized systems
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Causal machine learning (CML) enables individualized estimation of treatment effects, offering critical advantages over traditional correlation-based methods. However, existing approaches for medical survival data with censoring such as causal survival forests estimate effects at fixed time points, limiting their ability to capture dynamic changes over time. We introduce Causal Analysis for Survival Trajectories (CAST), a novel framework that models treatment effects as continuous functions of time following treatment. By combining parametric and non-parametric methods, CAST overcomes the limitations of discrete time-point analysis to estimate continuous effect trajectories. Using the RADCURE dataset [1] of 2,651 patients with head and neck squamous cell carcinoma (HNSCC) as a clinically relevant example, CAST models how chemotherapy and radiotherapy effects evolve over time at the population and individual levels. By capturing the temporal dynamics of treatment response, CAST reveals how treatment effects rise, peak, and decline over the follow-up period, helping clinicians determine when and for whom treatment benefits are maximized. This framework advances the application of CML to personalized care in HNSCC and other life-threatening medical conditions. Source code/data available at: https://github.com/CAST-FW/HNSCC

Problem

• Estimates dynamic treatment effects over time
• Overcomes limitations of fixed-time effect analysis
• Models chemotherapy and radiotherapy evolution in HNSCC

Innovation

• Models treatment effects as continuous time functions
• Combines parametric and non-parametric survival analysis
• Estimates dynamic treatment effects for personalized care