Dosimetric comparison of inverse optimization with geometric optimization in combination with graphical optimization for HDR prostate implants

Abstract

The purpose of this study is to compare geometric optimization (GO) with anatomy based inverse optimization (ABIO). Five patients of carcinoma prostate treated with HDR interstitial brachytherapy had been studied. Post implant CT scans of 5 mm slice thickness were obtained; target volume and other critical structures rectum, bladder and urethra were drawn by the clinician. Plans were obtained with geometric optimization and anatomy based inversed optimization. Anatomy based inverse planning implemented currently in PLATO BPS version 14.2, is based on geometric and dose point optimization and designed to account for the critical structures. Graphical optimization (GrO) is used to fine-tune the distribution ie to reduce the dose to critical structures and to improve the target coverage in both geometric optimization and anatomy based inverse optimization plans. DVH of target, rectum, bladder and urethra were evaluated and compared, dose homogeneity index and conformity index were also evaluated for all the plans. The mean target coverage was 93.9±7%, 90.3±4%, 82±13%, 91.6±3 for different optimization techniques GO, GO_gr, ABIO and ABIO_gr respectively. The target coverage in ABIO is not clinically acceptable. Maximum dose, dose to 2% of the volume of urethra D2%,U was 137±12%, 123.2±2%, 111.5±9, 122.7±4 for GO, GO_gr, ABIO and ABIO_gr respectively. The mean conformity index values were 0.71, 0.76, 0.65, 0.82 for GO, GO_gr, ABIO, ABIO_gr respectively. ABIO_gr has a good conformity over all other optimization techniques. However the difference is not very significant between GO and GO_gr. The mean values of DHI are 0.81, 0.77, 0.65 and 0.75 for GO, GO_gr, ABIO and ABIO_gr respectively. Geometric optimization is highly homogenous compared to all other optimization techniques.To conclude, target coverage in ABIO is not clinically acceptable. However ABIO followed by graphical optimization is much superior in sparing of critical structures and conformity compared to geometrical optimization. Target coverage is marginally better in GO compared to ABIO_gr. Homogeneity is superior in GO compared to ABIO_gr. However ABIO_gr plans were clinically acceptable with respect to homogeneity. Further, dose escalation to the target is possible with ABIO, without exceeding the tolerance dose to urethra. Clinical correlation of genitourinary toxicity has to be studied.