In International journal of radiation oncology, biology, physics
PURPOSE : To develop an externally validated model for predicting liver toxicity after radiotherapy in hepatocellular carcinoma (HCC) patients that can integrate both photon and proton dose distributions with patient-specific characteristics.
METHODS : Training data consisted all HCC patients treated between 2008 and 2019 at our institution (n=117, 60/40% photon/proton). We developed a shallow convolutional neural network (CNN) to predict post-treatment liver dysfunction from the differential dose-volume histogram (DVH) and baseline liver metrics. To reduce bias and improve robustness, we used ensemble learning (CNNE). Following a pre-registered study analysis plan, we evaluated stability using internal bootstrap resampling and generalizability using a dataset from a different institution (n=88). Finally, we implemented a class activation-map method to characterize the critical DVH sub-regions and benchmarked the model against logistic regression (LR) and XGBoost. The models were evaluated using the area under the receiver operating characteristics curve (AUC) and precision-recall curve (AUPRC).
RESULTS : The CNNE model showed similar internal performance and robustness compared to the benchmarks. CNNE exceeded the benchmark models in external validation, with AUC = 0.78 versus 0.55-0.70, and AUPRC=0.6 versus 0.43-0.52. The model showed improved predictive power in the photon group, excellent specificity in both modalities, and high sensitivity in the photon high-risk group. Models built solely on DVHs confirm outperformance of the CNNE and indicate that the proposed structure efficiently abstracts features from both proton and photon dose distributions. The activation map method demonstrates the importance of the low-dose bath and its interaction with low liver function at baseline.
CONCLUSION : We developed and externally validated a patient-specific prediction model for hepatic toxicity based on the entire DVH and clinical factors that can integrate both photon and proton therapy cohorts. This model complements the new ASTRO clinical practice guidelines and could support value-driven integration of proton therapy into the management of HCC.
Chamseddine Ibrahim, Kim Yejin, De Brian, Naqa Issam El, Duda Dan G, Wolfgang John A, Pursley Jennifer, Wo Jennifer Y, Hong Theodore S, Paganetti Harald, Koay Eugene J, Grassberger Clemens
2023-Feb-03
