Adaptive SBRT by 1.5 T MR-linac for prostate cancer: On the accuracy of dose delivery in view of the prolonged session time.

PURPOSE Adaptive Stereotactic Body Radiotherapy (SBRT) of prostate cancer (PC) by online 1.5 T MRi-guidance prolongs session-time, due to contouring and planning tasks, thus increasing the risk of prostate motion. Hence, the interest to verify the adequacy of the delivered dose. MATERIAL AND METHODS For twenty PC patients treated by 35 Gy (Dp) in five fractions, daily pre- and post- delivery MRi scans were respectively used for adapt-to-shape (ATS) optimization, and re-computation of the delivered irradiation (Drec). Two expansion recipes, from Clinical (CTV) to Planning target volume (PTV), which slightly differed in the posterior margin were used for groups I and II, of ten patients each. Plans had to assure D95% ≥ 95%Dp to PTV, and D1cc ≤ Dp to rectum, bladder, penile bulb, and urethral planning-risk-volume (urethral-PRV). The adequacy of the delivered dose was estimated by inter-fraction average (ifa) of dose-volume metrics computed from Drec. A cumulative dose (Dsum) was calculated from the five daily Drec deformed onto the simulation MRi. RESULTS For each patient, CTV coverage resulted in D95% > 95%Dp when estimated as ifa by Drec. No significant difference for D95% and D99% metrics to CTV resulted between groups I and II. D1cc was < Dp for rectum, urethral-PRV, and penile bulb, whereas < 103.5%Dp for the bladder. Significant correlations resulted between metrics computed by Dsum and as ifa by Drec, by both linear-correlation analysis, and Receiver-Operating-Characteristic curve analysis. CONCLUSIONS Our results for PC-SBRT confirm the adequacy of the delivered dose by ATS with 1.5 T MR-linac, and the consistency between dose-volume metrics computed by Drec and Dsum.