(Sunday, 7/14/2019)
Room: ePoster Forums
Purpose: The purpose of this study was to investigate the relationship between model predicted tumor control probability (TCP) and observed in-field local recurrences for patients treated for non-small cell lung cancer (NSCLC) with proton therapy. Differences between treatment planning system (TPS) constant relative biological effectiveness (RBE)-weighted dose and Monte Carlo (MC) constant and variable RBE-weighted dose (cRWD and vRWD) were investigated.
Methods: A cohort of 44 patients treated for NSCLC on clinical trial with passively-scattered proton therapy was selected based on data integrity and availability. Of the selected patients, 7 patients eventually presented with in-field local tumor recurrence. Patient dose distributions were recalculated with an in-house MC system including linear energy transfer and RWD calculations. Equivalent uniform dose (EUD) and TCP were subsequently calculated for all dose distributions based on published models. Parameters were compared across different dose calculation methods as well as for in-field local recurrence status.
Results: Median predicted TCP for TPS dose, MC cRWD, and MC vRWD were 84%, 71%, and 55%, respectively, for patients without in-field local recurrence and 83%, 69%, and 50%, respectively, for patients with local recurrence. For each dose distribution type, neither TCP nor EUD was significantly associated with in-field local recurrence according to a univariate Wilcoxon rank sum test. Median model predicted TCP for TPS dose was most similar to the observed rate of in-field local control for the cohort (84%).
Conclusion: TCP for either constant or variable RBE-weighted dose calculated with MC techniques was on average lower than that for treatment planning system dose across the cohort of NSCLC patients. EUD and TCP were not significantly different for patients with or without local recurrence across all dose types, which could be attributable to the small sample size of the current study.
Keywords
Taxonomy
TH- Radiobiology(RBio)/Biology(Bio): RBio- LQ/TCP/NTCP/outcome modeling
Contact Email
