Focal Dose Intensification to the Dominant Intraprostatic Lesion (DIL) Using an MR-Linac Adaptive-Planning Approach for Prostate Cancer: Dosimetric Outcomes and Early Toxicity

Abstract

<h3>Purpose/Objective(s)</h3> Focal boosting of the dominant intraprostatic lesion (DIL) improves biochemical control. The MR-linear accelerator (MRL) may improve the precision of DIL boosting with ultrahypofractionation by accounting for daily changes in the target and organs at risk (OAR). We report our early experience using this approach. <h3>Materials/Methods</h3> Fifteen patients with intermediate risk prostate cancer and a dominant lesion were treated on the MRL with stereotactic body radiation (SBRT) to 40Gy in 5 fractions, with the DIL boosted to 45Gy. A daily adaptive plan was created on initial daily imaging starting with bladder empty using adapt-to-shape (ATS) method. A verification image was acquired and further adjustments made using ATS or adapt-to-position (ATP) to account for target motion during initial planning. Contours were retrospectively adjusted on verification and post treatment scans to account for anatomic changes. The daily adapted plan was recalculated on these scans to estimate true dose delivered to the prostate and DIL. Prostate and DIL coverage on the pre-treatment plan (plan 1), was compared with recalculated plans on the verification (plan 2) and post treatment (plan 3) scans, based on values averaged over patients' 5 fractions (n=15). Differences across all fractions (n=75) were evaluated using a linear mixed model. Dosimetric impact of covariables including use of secondary ATS/ATP, and bladder and rectal volume changes were evaluated. Bladder filling was evaluated from fraction 1 to 5 to evaluate for any trends as treatment progresses. <h3>Results</h3> The median prostate D95 for plan 1, 2 and 3 was 40.3Gy, 40.5Gy and 40.3Gy respectively and DIL D95 was 45.7Gy, 45.2Gy and 44.6Gy. On a per fraction level, prostate coverage did not significantly change during planning time (plan 1 vs plan 2) or treatment delivery time (plan 2 vs plan 3), however there was a trend towards reduced DIL coverage between plan 1 vs plan 2, and plan 2 vs plan 3 (p=0.08, 0.06 respectively). Increased bladder volume reduced DIL coverage (p=0.03, plan 1 vs 2), and prostate coverage (p= 0.03, plan 2 vs 3). Bladder filling was associated with fraction number, fraction 1 being higher than the other 4 (all p-values < 0.02). Treatment time on table was associated with change in bladder filling (p=0.01) and after accounting for this, the associations between change in bladder filling and fraction number were weaker. At a median follow-up of 9 mo, the maximum reported toxicity was G2 GU (n=2) and G2 GI (n=1). The median change in IPSS was 3.5. <h3>Conclusion</h3> SBRT with focal boosting of the DIL on the MRL with daily adaptive planning is feasible and results in excellent planned and delivered dosimetry and acceptable toxicity. Intra-fractional bladder filling may impact target coverage. When considering focal boosting with SBRT, adaptive planning with MRL may provide the greatest confidence for delivering the prescribed dose with least normal tissue exposure.