Abstract
PURPOSE: This study aimed to develop and validate a 10-point, threshold-based variation score (VS) that integrates magnetic resonance imaging (MRI)-derived spacer morphology and assesses its association with rectal dose after low-dose-rate (LDR) prostate brachytherapy. MATERIALS AND METHODS: We retrospectively studied 149 men treated with iodine-125 LDR brachytherapy and polyethylene-glycol spacers (August 2022-April 2023). Using day 30 fused computed tomography-MRI, the VS assigned one point each for total volume >5 mL; hemisphere volumes (apex, base, left, and right) >5 mL; midgland thickness >5 mm; and thickness >5 mm 1-cm toward apex, base, left, and right. Univariable linear regression tested VS versus change in rectal D1cc (maximum dose to 1 mL). Bonferroni-adjusted Fisher's exact tests evaluated component differences across prespecified VS thresholds (≥6, ≥8, and 10). RESULTS: Median rectal D1cc decreased from 92.4% pre-spacer to 53.4% of prescription at day 30. Each one-point higher VS was associated with 5.4 Gy lower D1cc (95% confidence interval, 3.4 to 7.4; p < 0.001, R² = 0.166). Threshold analyses identified key contributors: midgland separation distinguished VS <6 from 6-7 (adequacy 57%→100%; adjusted p = 0.026); uniform directional thickness, especially apical, distinguished 6-7 from 8-9 (22%→91%; adjusted p < 0.001); apical volume >5 mL was required for VS = 10 (19%→100%; adjusted p < 0.001). CONCLUSION: VS is a simple, objective metric linking spacer morphology with rectal dose reduction. Threshold-oriented targets, midgland separation for VS ≥ 6, uniform thickness for VS ≥ 8, and robust apical volume for VS = 10, may guide technical optimization and quality assurance in prostate LDR brachytherapy.