A comparison of anatomy-based inverse planning with simulated annealing and graphical optimization for high-dose-rate prostate brachytherapy

Abstract

Background and purpose Dose distribution in a high-dose-rate (HDR) brachytherapy implant is optimized by adjusting source dwell positions and dwell times along the implanted catheters. Inverse planning with fast simulated annealing (IPSA) is a recently developed algorithm for anatomy-based inverse planning, capable of generating an optimized plan in less than 1 min. The purpose of this study is to compare dose distributions achieved using IPSA to those obtained with a graphical optimization (GrO) algorithm for prostate HDR brachytherapy. Methods and materials This is a retrospective study of 63 consecutive prostate HDR brachytherapy implants planned and treated using on-screen GrO to a dose of 10 Gy per implant. All plans were then recalculated using IPSA, without changing any parameters (contours, catheters, number, or location of dwell positions). The IPSA and GrO plans were compared with respect to target coverage, conformality, dose homogeneity, and normal tissue dose. Results The mean volume of target treated to 100% of prescription dose ( V 100) was 97.1% and 96.7%, and mean Conformal Index 0.71 and 0.68 with GrO and IPSA, respectively. IPSA plans had a higher mean homogeneity index (0.69 vs. 0.63, p < 0.001) and lower volume of target receiving 150% (30.2% vs. 35.6%, p < 0.001) and 200% (10.7% vs. 12.7%, p < 0.001) of the prescription dose. Mean dose to urethra, rectum, and bladder were all significantly lower with IPSA ( p < 0.001). IPSA plans tended to be more reproducible, with smaller standard deviations for all measured parameters. Conclusions Plans generated using IPSA provide similar target coverage to those obtained using GrO but with lower dose to normal structures and greater dose homogeneity.