Evaluation of Cumulative Daily Dose Mismatches In Prostate Proton Therapy: A Quantitative Incentive for Online Adaptive Replanning
Abstract
Purpose
To quantify the discrepancies that occur when a proton treatment plan is delivered on daily anatomy, following the clinical standard of marker-based registration using daily CT imaging. This study aims to demonstrate the necessity for real-time adaptive replanning by evaluating the mismatch between planned and delivered dose distributions due to anatomical changes that occur from day-to-day.
Methods
Prostate patients receiving proton therapy underwent a daily CT prior to each fraction. The original treatment plan was recalculated onto the daily CTs following a rigid marker-based registration with the treatment planning CT (TPCT) on Raystation. Using MIM software, the dose distributions from the daily CTs were mapped back onto the TPCT coordinate system. 3D Volumetric Gamma Index analysis (3%/3mm, 5% lower limit threshold) using PTW-VeriSoft was performed along with dose-volume-histogram (DVH) evaluations for single-fraction and cumulative dose comparisons against the original treatment plan.
Results
Initial comparisons between single fractions revealed significant spatial and dosimetric shifts between the delivered dose on a daily CT and the planned dose on the TPCT. The 3D gamma analysis between single fractions revealed an 81.1% agreement, while the cumulative comparison dropped to 75.0%. The DVHs showed a decrease in tumor dose coverage by an average of 0.38 Gy per fraction while 0.96 Gy for the cumulative plan, while an increase in dose to the two organs-at-risk (OAR), i.e. rectum and bladder, by an average of 0.12 Gy and 0.52 Gy, respectively.
Conclusion
Marker-based registration is insufficient to maintain the high-dose conformality required for proton prostate therapy due to daily changes in patient's internal anatomy. These observed dose mismatches serve as a quantitative incentive for the implementation of real-time adaptive proton therapy to ensure accurate target coverage and OAR sparing.