A model-based risk-minimizing proton treatment planning concept for brain injury prevention in low-grade glioma patients

H. Sallem,S. Harrabi,E. Traneus,K. Herfarth,J. Debus,J. Bauer
DOI: https://doi.org/10.1016/j.radonc.2024.110579
IF: 6.901
2024-10-12
Radiotherapy and Oncology
Abstract:Purpose Late-occurring contrast-enhancing brain lesions (CEBLs) have been observed on MRI follow-up in low-grade glioma (LGG) patients post-proton therapy. Predictive risk-models for this endpoint identified a dose-averaged linear energy transfer (LET d )-dependent proton relative biological effectiveness (RBE) effect on CEBL occurrence and increased radiosensitivity of the cerebral periventricular region (VP 4mm ). This work aimed to design a stable risk-minimizing treatment planning (TP) concept addressing these intertwined risk factors through a classically formulated optimization problem. Material and Methods The concept was developed in RayStation-research 11B IonPG featuring a variable-RBE-based optimizer involving 20 LGG patients with varying target volume localizations and risk-factor contributions. Classical cost functions penalizing dose, dose-volume-histogram points, and equivalent uniform dose were used to formulate the optimization problem, and a new set of structures was introduced to actively spare the VP 4mm , control high LET d regions, and de-escalate the dose outside the gross tumor volume. Target volume coverage and organ-at-risk sparing were robustly evaluated, and Normal Tissue Complication Probabilities (NTCP) for CEBL occurrence were quantified. Results The concept yielded stable optimization outcomes for all considered subjects. Risk hot spots were successfully mitigated, and an NTCP reduction of up to 79 % was observed compared to conventional TP while maintaining target coverage, demonstrating the feasibility of the chosen model-based approach. Conclusion With the proposed TP protocol, we close the gap between predictive risk-modeling and practical risk-mitigation in the clinic and provide a concept for CEBL avoidance with the potential to advance treatment precision for LGG patients.
oncology,radiology, nuclear medicine & medical imaging
What problem does this paper attempt to address?