Initial Experience Using Daily Artificial Intelligence-Assisted Adaptive Radiotherapy on Cone-Beam CT for Bladder Cancer

Abstract

<h3>Purpose/Objective(s)</h3> Adaptive radiation therapy (ART) provides a method to observe daily changes in the patient's internal anatomy and revise the radiation plan. In patients undergoing radiation for bladder cancer, changes in the bladder contour due to bladder filling result in significant changes in the radiation dose distribution potentially reducing the target coverage. Large margins are required in order to account for these anatomic changes, increasing the dose to the nearby organs at risk (OARs). Daily ART is an attractive method for bladder cancer in an attempt to reduce treatment margins. Herein we report our initial experience using commercially available, online adaptive platform with Artificial Intelligence (AI)-assisted workflows on daily cone-beam computed tomography (CBCT) for bladder cancer. <h3>Materials/Methods</h3> 9 consecutive patients eligible for bladder conserving radiotherapy underwent CT-simulation (CT-sim) and planned for ART. Clinical Target Volume included the bladder and proximal prostatic urethra +/- pelvic lymph nodes with an additional 0.5-1cm margin for the planning target volume (PTV). Fractionation schemas utilized included 64Gy/32 fractions, 55Gy/20 fractions, and 36Gy/6 fractions. A reference IMRT plan was created and approved using institutional target goals and OAR constraints. For treatment, the patient was aligned and CBCT acquired. The CT-sim was registered to the CBCT using deformable registration and a synthetic CT scan was generated. AI-based auto-contour and structure deformation of OARs and Targets on CBCT were reviewed and edited by the treating physician. 2 plans were generated including, a CT sim-based plan with deformed structures (scheduled) and a re-optimized plan (adaptive) of which both plans evaluated and the best one approved and delivered. Statistical analyses were performed comparing PTV coverage and OAR constraints between adaptive versus scheduled plans and a dosimetric comparison of ART compared to non-ART large-margin (1.5cm) treatment. <h3>Results</h3> 174 comparative adaptive and scheduled plans were generated of which the adaptive plan was chosen in 91.0% of sessions. V95 PTV mean coverage was improved with adaptation 97.0% versus 86.9% (p<.001). Average PTV volume change during ART was 16.4% (Range:0.4%-98.0%). All completed ART courses met institutional OAR constraints. With adaptation, there was a mean decrease in rectal V45%, V75%, V100% by 6% (95% CI 4.2%-7.6%, p<.003), 2.3% (95% CI 1.6%-3%, p<.003), 0.7% (95% CI 0.5%-0.8%, p<.001) respectively, and an increase in bowel V85%, V90% by 0.5% (95% CI 0.3%-0.8%, p<.04), 0.6% (95% CI 0.4%-0.8%, p<.01), respectively. Compared to a plan with large margins, there was a significant reduction in all bowel and rectal parameters with daily ART and smaller PTV margins. <h3>Conclusion</h3> Daily AI-assisted ART using CBCT for bladder cancer appears feasible and may potentially improve the therapeutic index by reducing the need for large PTV margins while ensuring acceptable target coverage.