SU‐GG‐T‐08: Is a Volume‐Based HDR Brachytherapy Optimization Algorithm Comparable to a Classic Line‐Based One?: Toward Tumor‐Volume Adaptive Brachytherapy for Cervical Cancer By 3T MRI Guidance

Purpose: Verify a volume‐based optimization algorithm against a classic line‐based one for intracavitary high‐dose‐rate (HDR) brachytherapy (BT) treatment planning for cervical cancer. Method and Materials: This retrospective study reviewed ten randomly selected tandem and ovoids plans. To investigate the dose‐behavior induced by different algorithms, no variations in T&O applicator geometry and imaging dataset were made. Three metrics were employed: 1) the total reference air kerma (TRAK), 2) the volume enclosed by the 100% prescription dose (VOL 100%Rx ) (both based on ICRU Report 38), and 3) the dose to Point H. Conventional HDR plans (BrachyVision™, Varian, version 6.1) were regenerated and reoptimized by the optimization algorithm capable of incorporating a volume‐based optimization (version 6.5). To optimize dwell times, both are based on Nelder‐Mead Simplex method. However, a volume‐based plan optimizes dwell times to give the desired dose to volumetric structures along with reference lines by utilizing a group of points in a structure, while a conventional algorithm affords solely reference lines. Results: A volume‐based algorithm was found to be comparable to a classic one in terms of three metrics described above. The ratio of TRAK values (i.e. a volume‐based algorithm over a classic), VOL 100%Rx , and Point H doses showed on average 1.01 (std dev ± 0.01), 1.03 (std dev ± 0.07), and 0.99 (std dev ± 0.02), respectively. To obtain an identical HDR plan was limited. However, we found a volume‐based optimization algorithm generates a plan comparable to a classic line‐based one which is relatively simple and robust. The plan generated by a volume‐based optimization algorithm also affords the ABS recommendations for BT for cervical cancer. Conclusion: A high‐dose‐rate brachytherapy plan utilizing a volume‐based optimization algorithm was comparable to a classic one. Hence, a volume‐based optimization is available to be implemented for a tumor‐volume adaptive brachytherapy.