3279 - Tumor Control and Normal Tissue Complication Probability Modeling for High Dose Rate Brachytherapy Followed by 5-Fraction SBRT Treatment of Unfavorable Intermediate or High-risk Prostate Cancer Patients with Biodegradable Hydrogel Rectal Spacer Placement

Purpose/Objective(s): We present the efficacy of a plan evaluation tool for a 2-course of radiation consisting of a highly heterogeneous high dose rate (HDR) brachytherapy plan followed by curative 5-fraction SBRT for unfavorable intermediate or high-risk prostate cancer patients with biodegradable hydrogel spacer for rectal sparing. For a more accurate plan assessment, this novel method calculates the 3D voxelized Equivalent Dose in 2 Gy (EQD2) from each course and combines them into a 3D EQD2 distribution.

Materials/Methods: Under the IRB approved protocol, 12 previously treated HDR prostate (15 Gy × 1) patients followed by 5-Fraction SBRT (5 Gy × 5) were included in the study. HDR plans used NRG-GU009 protocol with V100 > 95% for prostate, D10 < 118% for urethra, and V75 < 1 cc for rectum and bladder. In SBRT plans, CTV was contoured by attending physician and was expanded by 5 mm in all directions except for superior and posterior (3 mm) for PTV. To spare the rectum during HDR and SBRT, ultrasound-transparent hydrogel was injected between rectum and prostate before and after HDR delivery. Goal was to achieve 100% dose coverage for CTV while V100 < 15% for bladder, rectum and bowel bag for SBRT. To account for radiobiological effects of HDR, ultrasound and CT images were registered by experienced attending physician. The 3D EQD2 for both plans, and dose sum were calculated utilizing standard Linear Quadratic model with contours for CTV and OARs from CT. The a/ß ratio of 1.5 Gy for prostate and 3 Gy for bladder and rectum was used. Predicted TCP value for CTV was evaluated using the voxelized TCP model based on Poisson statistics. The D₅₀ and ?₅₀ parameters used in the TCP model for stage III prostate cancer were 46.25 Gy and 0.95, adapted from literature. NTCP was evaluated via Lyman-Kutcher-Burman model with n, m, and TD50 values of 0.12, 0.15, and 80 Gy for rectum, and 0.5, 0.11, and 80 Gy for bladder.

Results: For these patients, in combined courses, mean EQD2 for CTV was 186.9 ± 15.1 (168–221.7) Gy. The predicted 5-year TCP for prostate was 92.7 ± 3 (88.1–97.9)%. The D0.03cc, D1cc, and Deff (for NTCP model) of EQD2 for rectum were 113.9 ± 57 (62.5–273.3) Gy, 64.2 ± 14.4 (42.2–86) Gy, and 62.8 ± 47.9 Gy. For bladder, D0.03cc and D1cc were 135.5 ± 58.9 (55.7–239.4) Gy and 71.7 ± 11.9 (50.5–90.4) Gy. Predicted NTCP values for rectum and bladder were 3.0 ± 5.7 (0–18.6)% and 0.

Conclusion: This composite 3D EQD2 distribution provided a means of better evaluation of dose delivered to prostate and OARs via HDR plus SBRT plans. Utilizing the knowledge from traditional studies of 2 Gy per fraction, these predictive models may allow for more personalized treatment for HDR and followed by SBRT or EBRT; potentially enhancing therapeutic ratios in managing unfavorable intermediate or high-risk prostate cancer patients. Benchmarking these predictive models by comparing them with clinical outcome results is underway.