|
内容記述 |
Purpose: The LET trilemma—an inherent conflict among target dose homogeneity, range robustness, and high dose-averaged linear energy transfer (LETd)—poses a major challenge in treatment optimization. To ensure accurate delivery of multi-ion therapy, this study evaluated the effects of range and setup uncertainties on LETd optimized treatment plans and explored strategies to overcome this trilemma, framed within the phase I LETd escalation trial for head and neck cancers.Methods: Six head and neck cancer patients representing diverse tumor locations and sizes were selected. Based on a previous study, the ion combinations (carbon, oxygen, and neon) were determined according to the gross tumor volume (GTV) size and the target LETd level. Multi-ion therapy plans were then simultaneously optimized using two fields to achieve a target LETd of 90 keV/μm (the final LETd level of the phase I trial). These plans were recalculated to incorporate uncertainties in stopping power ratio (SPR) conversion and random daily setup variations across the 16 treatment fractions. The combined effects of these uncertainties on the dose and LETd distributions were subsequently evaluated at the limits of the 87% confidence interval (approximately 1.5σ limits for a normal distribution). Additionally, to explore strategies to increase plan robustness, five modified plans—target expansion, orthogonal arrangement, opposing arrangement, four-field arrangement, and heavier ion selection—were evaluated for one patient identified as particularly susceptible to uncertainties.Results: SPR conversion error was the dominant contributor to degraded plan quality across the 16 treatment fractions in multi-ion therapy, substantially outweighing daily patient setup error. Small, centrally located tumors were most susceptible, exhibiting dose inhomogeneity of approximately 11%. This was primarily due to the steep dose gradients formed at the field junctions in the LETd optimized plan, which created dose hot and cold spots within the GTV under range overshoot and undershoot scenarios, respectively. In contrast, LETd distributions within the GTV remained stable, with absolute deviations of less than 3.5 keV/μm. Importantly, all treatment plans satisfied dose constraints for normal tissues even under the upper boundary scenarios. The most effective strategy for improving robustness was the heavier ion selection, which involved replacing the original carbon−oxygen combination with oxygen ions for both beam ports. This strategy reduced dose inhomogeneity by more than 7%, resulting in an approximately 4% inhomogeneity while maintaining normal tissue sparing adjacent to the target.Conclusion: Optimization toward higher LETd makes treatment plans susceptible to range uncertainties, primarily by introducing steep dose gradients that lead to dose inhomogeneity within small, deep-seated tumors. Employing heavier ions is a practical and effective strategy to overcome this challenge, enabling robust target coverage by leveraging their inherently higher LETd, while preserving normal tissue sparing, planning simplicity, and beam delivery throughput. These findings provide a key rationale for ion selection in the design of robust multi-ion therapy. |
