3326 - Dosimetric Correlates of Long-Term Quality of Life In a Randomized Trial of Prostate Radiation Therapy with and without a Hydrogel Spacer
Presenter(s)
O. Awad1,2, Z. A. Siddiqui1, C. G. Candano1, Z. A. Seymour3, T. J. Quinn3, C. Knill4, H. Mekdash1, D. A. Hamstra1, and A. S. Mohamed1,5; 1Department of Radiation Oncology, Dan L. Duncan Comprehensive Cancer Center, Baylor College of Medicine, Houston, TX, 2Clinical Oncology Department, Faculty of Medicine, Alexandria University, Alexandria, Egypt, 3Department of Radiation Oncology, Corewell Health William Beaumont University Hospital, Royal Oak, MI, 4Univ of Michigan, Ann Arbor, MI, 5Department of Radiation Oncology, The University of Texas M.D. Anderson Cancer Center, Houston, TX
Purpose/Objective(s): We aim to evaluate the dosimetric predictors of long-term quality of life (QOL) in patients undergoing prostate radiation, as assessed using the Expanded Prostate Cancer Index Composite (EPIC). Specifically, we examine the dosimetric impact of a prostate rectal spacer (SpaceOAR) on patient-reported outcomes. We hypothesized that segmental rectal DVH analysis would identify the dose threshold most predictive of bowel QOL deterioration.
Materials/Methods: In a single-blind phase III trial, 222 men receiving 79.2 Gy prostate IG-IMRT were randomized 2:1 to SpaceOAR hydrogel or control. The rectum was segmented into inferior, middle, and superior thirds alongside whole rectum. DVH parameters were extracted for bladder, bladder trigone, urethra, and bilateral internal pudendal artery (IPA). Bowel and urinary QOL were assessed using EPIC; minimally important difference (MID) was defined as deterioration = 5 for bowel and = 6 points for urinary function. DVH parameters were compared between MID groups using Mann-Whitney U test. Recursive partitioning analysis (RPA) identified optimal thresholds. Paired Wilcoxon test compared SpaceOAR pre vs post-hydrogel dosimetry. Akaike information criterion (AIC) evaluated whether additional parameters improved prediction beyond the primary threshold. Conditional RPA identified the next actionable constraint among patients already meeting the primary threshold.
Results: At 36 months, 132 patients with completed questionnaires and available DVH were analyzed. Bowel MID = 5 was in 38% of control vs 14% of SpaceOAR patients (p=0.005), while urinary MID = 6 showed no significant difference (29% vs 17%, p=0.16). Whole rectum DVH parameters in the range V38-V83 were all significantly higher in patients with bowel MID=5, followed by inferior rectum V7-V47 and middle rectum V55-V77 (all p<0.05). RPA identified whole rectum V75=9% as the most significant predictor of worse bowel QOL (p=0.00008), with 67% of patients exceeding this threshold experiencing bowel MID=5 compared to 16% of those below. AIC confirmed no additional whole rectum parameter improved prediction beyond V75. Rectal V75= 9% was observed in 34% of control compared to 2% in the SpaceOAR patients (p<0.0001). In SpaceOAR pre-hydrogel plans, 47% exceeded this threshold, reduced to 2% after hydrogel placement. Conditional RPA among SpaceOAR patients meeting V75<9% (n=62) identified inferior rectum V38=17% as the next predictor: 32% bowel MID=5 above vs 0% below (p=0.008). IPA DVH showed no association with bowel or urinary MID.
Conclusion: We identified rectal V75 = 9% as a dosimetric predictor of long-term worsened bowel QOL. SpaceOAR reduces the proportion exceeding this threshold from 47% to 2%, explaining superior bowel QOL in the SpaceOAR arm. After this threshold is met, inferior rectum V38< 17% eliminates residual bowel MID risk in SpaceOAR patients, establishing a hierarchical two-constraint framework and supporting the role of prostate apex spacing in optimizing bowel QOL