Proton Tumor Control Relative Biological Effectiveness (RBE) Estimated for Various Fractionation Schedules and Hypoxia Levels

Abstract

<h3>Purpose/Objective(s)</h3> We previously modeled clinical proton local control data of early-stage lung cancer using a simulation model based on an energy budget. The fitted modeling results suggest that proton therapy is nearly independent of hypoxia reoxygenation which is a key process to model photon therapy results. In this work, we investigated the effect of variable hypoxia levels on the local control efficacy of proton therapy compared to photon, in terms of RBE, for various fractionation schedules, searching for the optimal treatment option. <h3>Materials/Methods</h3> Model simulations were performed with a previously developed tumor simulation model with clinically derived parameter values for photon (CCR; 23; 5469-79; 2017) and proton (presented at RRS 2021). For five different levels of hypoxia, represented by growth fractions (GF), where lower GF represents more hypoxic condition, simulations were performed for 14 clinically applied fractionation schedules (see Table). For each proton schedule, the cell survival fraction was estimated and a partner simulation was performed with photon parameters, to find the dose level with the same fractionation scheme that would lead to an equivalent cell survival level. Finally, the proton RBE was calculated from the ratio of determined photon physical dose to physical dose of proton. <h3>Results</h3> The estimated proton RBE depends on both the hypoxia level as well as the fractionation schedule (see Table). The proton RBE increases with increasing hypoxic level (decreasing GF) for all fractionation schedules. For shorter fractionation schedules (3-5 fxs), proton therapy is always more effective than photon, with RBEs of 1.06-1.18. For longer fractionation schedules (>10), however, the RBE decreases as the GF increases (less hypoxia) and even drops below 1.0 (down to 0.81) for some schedules, due to the ability of photon therapy to take advantage of long term reoxygenation. For each hypoxia level, an optimal schedule exists with the highest RBE value: longer schedules (10-12 fx) are better for hypoxic tumors; 5 fx schedules are better for less hypoxic tumors. <h3>Conclusion</h3> The results predict that proton vs. photon local control will depend critically on the hypoxic level, with proton therapy more effective as the level of hypoxia increases. For longer treatment schedules, proton therapy loses this advantage with decreasing hypoxia, suggesting that longer proton schedules may be less effective for well-oxygenation tumors. These predicted effects should be further probed through both preclinical experiments and clinical data analyses.