Comparison of online adaptation strategies for magnetic resonance guided prostate radiation therapy

MR-guided adaptive radiation therapy allows for daily plan adaptation to account for anatomical changes. Two common strategies are adapt-to-position (ATP), involving re-optimization based on isocenter shifts, and adapt-to-shape (ATS), involving full recontouring and reoptimization. This study provides a dosimetric comparison of ATP and ATS using accumulated dose.

Dose accumulation was performed for 35 patients with prostate cancer treated on a 1.5 T MR-Linac. All patients received ATS-based treatment with either 30.0 Gy in 5 fractions (30.0/5) or 42.7 Gy in 7 fractions (42.7/7), using a 5 mm isotropic PTV margin. ATP plans were retrospectively simulated. For each fraction, dose was mapped to a reference image using deformable image registration and summed across fractions. Fractional and accumulated dose-volume histogram (DVH) metrics were compared between ATS and ATP and correlated with daily anatomical variation.

ATP and ATS achieved equivalent accumulated CTV D95 and D98 for both regimens. In the 30.0/5 cohort, small but statistically significant differences in OAR dose were observed: accumulated bladder D40 was 4 % lower for ATP (1.27 Gy; p = 0.0004), and rectum D50 was 1 % lower for ATP (0.40 Gy; p = 0.0008). Differences in rectum D1cc and bladder D5cc were not significant. In the 42.7/7 cohort, femur D5 was 3 % higher for ATP (0.83 Gy; p = 0.02); other differences were insignificant. Dosimetric differences across strategies correlated with interfraction motion.

ATP and ATS provided equivalent target coverage. OAR differences were statistically significant in some cases but remained within clinical tolerances, suggesting minimal clinical impact.