Abstract
BACKGROUND AND PURPOSE: Nuclear interaction correction (NIC) and trichrome fragment spectra modelling improve relative biological effectiveness-weighted dose (D(RBE)) and dose-averaged linear energy transfer (LET(d)) calculation for carbon ions. The effect of those novel approaches on the clinical dose and LET distributions was investigated. MATERIALS AND METHODS: The effect of the NIC and trichrome algorithm was assessed, creating single beam plans for a virtual water phantom with standard settings and NIC + trichrome corrections. Reference D(RBE) and LET(d) distributions were simulated using FLUKA version 2021.2.9. Thirty clinically applied scanned carbon ion treatment plans were recalculated applying NIC, trichrome and NIC + trichrome corrections, using the LEM low dose approximation and compared to clinical plans (base RS). Four treatment sites were analysed: six prostate adenocarcinoma, ten head and neck, nine locally advanced pancreatic adenocarcinoma and five sacral chordoma. The FLUKA and clinical plans were compared in terms of D(RBE) deviations for D(98%), D(50%), D(2%) for the clinical target volume (CTV) and D(50%) in ring-like dose regions retrieved from isodose curves in base RS plans. Additionally, region-based median LET(d) deviations and global gamma parameters were evaluated. RESULTS: Dose deviations comparing base RS and evaluation plans were within ± 1% supported by γ-pass rates over 97% for all cases. No significant LET(d) deviations were reported in the CTV, but significant median LET(d) deviations were up to 80% for very low dose regions. CONCLUSION: Our results showed improved accuracy of the predicted D(RBE) and LET(d). Considering clinically relevant constraints, no significant modifications of clinical protocols are expected with the introduction of NIC + trichrome.