Abstract
Osteoradionecrosis (ORN) of the mandible is one of the most severe adverse events (AEs) for head and neck (H&N) cancer radiotherapy. Previous retrospective investigations on real-world data relied heavily on conventional statistical models that primarily elucidate correlation rather than establishing causal relationships. Through the novel causal machine learning method, we aim to obtain empirical relative biological effectiveness (RBE) for mandible ORN in head and neck (H&N) cancer patients treated with pencil-beam-scanning proton therapy (PBSPT).
1,266 H&N cancer patients were included: 335 patients treated by PBSPT and 931 patients treated by volumetric-modulated arc therapy (VMAT). We use 1:1 case-matching based on propensity scores to minimize the imbalance in clinical factors between patients treated with PBSPT and VMAT. The bias test of standardized mean differences (SMD) was applied on the case-matched patient cohorts. The causal machine learning method, causal forest (CF), was adopted to investigate the causal effects between dosimetric factors and the incidence of ORN. The dose volume constraints (DVCs) for VMAT and PBSPT were derived when the critical volumes of the derived DVCs lead to the largest average causal effect (ATE). RBE values were further empirically derived based on tolerance curves formed from the critical volumes of the derived DVCs. This was accomplished by comparing the equivalent constraint doses against the actual physical doses of PBSPT. To rigorously account for statistical variability in the RBE estimates, a bootstrap resampling method was applied to generate confidence intervals, thereby quantifying the uncertainty in the analysis.
335 VMAT patients were case-matched to 335 PBSPT patients; however, standardized mean bias analysis revealed persistent covariate imbalances within each group, indicating residual confounding influence. Using CF modeling, we identified DVCs of mandible for ORN and found that PBSPT had lower critical volumes than those of VMAT, leading to empirical RBE exceeding 1.1 in the moderate dose range (1.61 at 40 Gy[RBE=1.1], 1.30 at 50 Gy, and 1.13 at 60 Gy).
This study presents a novel application of causal machine learning to evaluate mandible ORN in radiotherapy, identifying DVCs linked to the strongest causal effects and deriving empirical RBEs from equivalent constraint dose analysis based on the tolerance curve. The results indicate that proton RBE may significantly exceed 1.1 in the moderate dose range (40-60 Gy[RBE=1.1]), underscoring the importance of incorporating the variable RBE into PBSPT treatment planning to mitigate the risk of ORN.