Purpose: To use dose accumulation to directly compare two strategies for online adaptation on an MR-Linac (MRL): the adapt-to-position (ATP) isocentric shift approach versus the full re-optimization adapt-to-shape (ATS) approach for prostate SBRT treatment.
Methods: This study analyzed 12 prostate SBRT patients treated at our centre with 30 Gy in 5 fractions on a 1.5 T MRL using 9–field IMRT with MR-based reference planning. The ATP workflow adapts the beams following translation-only rigid registration of the prostate. The ATS workflow generates a new treatment plan to match the daily anatomy using deformable image registration (DIR) and manual online contour editing by a Radiation Oncologist. Treatment plans on the localization MR images were created for all fractions using both ATP (simulated) and ATS (clinical) workflows. All plans were imported into a commercial treatment planning system to perform DIRs and accumulate dose on the reference MR images. Clinical goal DVH metrics for accumulated ATP and ATS dose distributions were compared with Student’s t-tests.
Results: Mean (± STD) CTV D98 was not significantly different between ATP (3132 ± 129 cGy) and ATS (3180 ± 90 cGy), and satisfied the > 2850 cGy clinical goal for 11 of 12 patients for ATP and all patients for ATS. DVH goals for femurs (D5 < 1200 cGy), rectum (D1cm3 < 3000 cGy, D20 < 2000 cGy, D50 < 1000 cGy) and bladder (D5cm3 < 3000 cGy) were not significantly different between ATP and ATS. Organ-at-risk clinical goals were not met in 6 patients (12/72 goals) for ATP and 2 patients (3/72 goals) for ATS.
Conclusion: Dose accumulation demonstrates that CTV coverage and OAR sparing are statistically equivalent for ATP and ATS workflows. The ATP workflow may be preferred for MR-guided adaptive treatments to eliminate need for online recontouring and thus reduce the overall treatment time.