Mixed- and multi-relative biological effectiveness model simultaneous optimization in carbon ion radiotherapy: A proof-of-concept

Background and purpose: In carbon ion radiotherapy (CIRT), different relative biological effectiveness (RBE) models have been used for calculating RBE-weighted dose (DRBE). Conversion between current RBE predictions and introduction of novel approaches remains a challenging task. Our aim is to introduce a framework considering multiple RBE models simultaneously during CIRT plan optimization, easing the translation between D RBE prescriptions. Materials and methods: An in-house developed Monte Carlo treatment planning system was extended to incorporate the local effect model version I (LEM-I), the modified microdosimetric kinetic model (mMKM) and the MKM-derived Japanese biological model (NIRS-MKM). Four clinical cases (two head-and-neck and two prostate patients), initially optimized with LEM-I for both targets and organs at risk (OARs), underwent two further optimizations: to fulfill mMKM/NIRS-MKM-based target prescriptions (mixed-RBE approach) or to simultaneously consider two biological models within the target regions (multi-RBE approach). Both approaches retained LEM-I-derived dose constraints for OARs. Results: The developed optimization strategies have been successfully applied, fulfilling all the clinical criteria for the applied RBE models. One of the RBE models showed unfavorable dose distribution when not explicitly considered in the optimization, while multi-RBE model optimization allowed meeting dose objectives for the selected OARs for both models simultaneously. Conclusions: The introduced optimization approaches allow for mixed- or multi-RBE optimization in CIRT through the selection of RBE models independently for each region of interest. This capability addresses challenges of adhering to multiple RBE frameworks and proposes an advanced solution for tailored patient treatment plans.