Purpose: Inverse planning is commonly used in high-dose-rate (HDR) brachytherapy, but it tends to generate plans with large variations in needle dwell times, which may compromise target coverage when needle displacement occurs. A robust planning technique involving forward graphical optimization and post-hoc uniformity correction was developed to minimize such variations, and the robustness of generated plans was compared against inverse plans in this study.
Methods: Gynecological Syed HDR cases (n=8) were investigated. Clinical robust plans were created in Oncentra® Brachy using the robust planning technique. Inverse plans were retrospectively generated using the hybrid inverse planning optimization (HIPO) module. A Matlab-based dose engine employing a dose kernel extracted from Oncentra® Brachy was constructed to simulate the effects of needle displacement. Systematic (2 or 5 mm) or random (2±2 or 5±5 mm, 10 simulations per plan) needle retractions were induced on all needles, after which doses were recalculated. Changes in dose volume histogram (DVH) metrics, including the high-risk clinical target volume D90%, V90%, D100%, V100%, V150%, as well as bladder and rectum D2cc, were compared.
Results: The Matlab-based dose engine provides comparable dose calculation compared to Oncentra® Brachy. Similar target and organ-at-risk (OAR) DVH metrics were observed in the HIPO and robust plans at baseline. Both systematic and random needle displacements resulted in decreased target coverage. Compared to the HIPO plans, the robust plans provided improved HRCTV coverage (V100% by 0.5% to 1.2%, V90% by 0.5% to 1.0%, and D90% by 1.2% to 2.5%; P<0.05 for all). No significant difference was observed in changes in OAR D2cc.
Conclusion: A robust planning technique for HDR brachytherapy is proposed, and a Matlab-based dose engine developed for simulating the effects of needle displacement. Compared to inverse planning, the robust planning offered more robust target coverage when needle displacement occurs.