The Dosimetric Effects of Long On-Couch Time of Adaptive Ultra-Hypofractionated Radiotherapy for Prostate Cancer Patients on a 1.5 MR Linac

Abstract

<h3>Purpose/Objective(s)</h3> To quantitatively characterize the dosimetric effects of long on-couch time of adaptive radiotherapy for prostate cancer patients on 1.5-T MR-Linac. <h3>Materials/Methods</h3> Seventeen patients consecutively treated with ultra-hypofractionated radiotherapy (UHF-RT) on 1.5-T MR-Linac were recruited from an on-going prospective phase II trial (NCT05183074). UHF-RT with dose of 36.25Gy in 5 fractions to whole prostate w/o seminal vesicles (SVs) (N=12) was delivered every other day, with a boost dose of 40Gy to whole prostate (CTV4000, N=16). Except the Pre-MR for registration and re-contouring and Position Verification-MR (PV-MR) for dose/position check which were conventionally used in Adapt-To-Shape (ATS) workflow, we also collected 3D-MR during beam-on phase (Bn-MR) and after RT (Post-MR). For each fraction, the contours of targets and organs-at-risk (OARs) on PV-MR, Bn-MR and Post-MR were projected from the Pre-MR by deformable registration, and manually adapted by the physician, followed by dose-recalculation of the ATS plan. The dose parameters of target volumes and OARs on each scan were generated in a commercially available software for final analysis. <h3>Results</h3> In total, 290 MR-scans were collected, including 85 Pre-MR, 85 PV-MR, 49 Bn-MR and 71 Post-MR scans, respectively. For each fraction, the volume differences of prostate and CTV on different scans were less than 3.0cc, indicating good alignment of target contouring. With median 49 (24-78) minutes of on-couch time, the mean PTV-V95% (34.4Gy) of plans on PV-MR, Bn-MR and Post-MR was 97.69±2.03%, 97.04±2.29%, and 96.82±2.59%, respectively, compared with 99.27±0.75% of ATS plan. Furthermore, the mean CTV-V100% (36.25Gy) of plans on PV-MR, Bn-MR and Post-MR was 99.32±1.20%, 98.59±1.84% and 98.65±1.87, respectively, compared with 99.93±0.30% of ATS plan. With excellent dose coverage of prostate-V100% (36.25Gy) (99.73±0.79% on PV-MR, 99.41±1.35% on Bn-MR and 99.38±1.35% on post-MR), the main reason of lower CTV-V100% was due to mild underdose of SVs (96.97±6.26% on PV-MR, 93.62±12.94% on Bn-MR and 94.69±9.22% on post-MR). With comparison to values of ATS plan, the actual dose of rectum wall during the whole workflow also varied, with mean △V38_Gy of 0.23±0.28cc on PV-MR, 0.41±0.51cc on Bn-MR and 0.39±0.52cc on post-MR, with mean △V36_Gy of 0.27±0.57cc on PV-MR, 0.39±0.71cc on Bn-MR and 0.30±0.68cc on post-MR, respectively. However, the V29_Gy of rectum wall increased by >15% was only observed in one scan. On the contrary, no obvious increase of bladder wall dose was seen due to gradual growth of bladder over the workflow. <h3>Conclusion</h3> This dosimetry analysis demonstrated reliable dose coverage of target volumes during beam-on period with adaptive ATS workflow on 1.5-T MR-Linac. The 3 mm CTV-PTV margin applied in our study have proven to be large enough for prostate irradiation, but might be insufficient for SVs incidentally. More attention should be paid to restrict the rectum wall high dose when optimizing ATS plan.