To Space or Not to Space: The EPIC Question for Prostate Stereotactic Radiotherapy (SBRT) with or without Hydrogel Rectal Spacer (RS)

Abstract

Introduction: Following prostate radiotherapy (RT), bowel, urinary, and sexual side effects and quality-of-life (QOL) declines are common. Phase 3 trials of rectal spacers (RS) using ≥20 fractions found clinical/dose benefits and reduced QOL declines. However, the role of RS in stereotactic body radiotherapy (SBRT) is undefined.

Methods: A prospective single-institution registry of prostate SBRT from 2012-2023 was analyzed by RS use (n=290) vs no RS (n=1815). QOL was collected via EPIC-26 at baseline and up to 5 years post-RT. Treatment used CT/MRI fusion and 3-6 fiducials for real-time tracking with Robotic SBRT (CyberKnife®, Accuray Inc.). CTV included prostate plus proximal seminal vesicles. PTV margins were 5 mm except 3 mm posteriorly. 35-36.25 Gy was delivered in 5 fractions. The primary endpoint was QOL trend over time by RS vs no RS as evaluated by linear mixed-effects models which accounted for within-subject variability by controlling for key clinical/demographic characteristics. Clinically important change analyses were conducted using established minimally important difference (MID) thresholds to compare proportion of patients in each group with meaningful QOL declines at each timepoint.

Results: There were no differences in age, PSA, or prostate volume between groups. RS was associated with more recent treatment (p<0.001), intermediate/high risk disease (96% vs 85% p<0.001), ADT use (52% vs 39% p<0.001), and Caucasian patients (63% vs 55% p<0.001). Baseline EPIC scores were similar. Declines in EPIC scores post-SBRT were small, approaching baseline after 6 months and remaining stable to 5 years. There were no clinically significant differences in QOL trend over time by RS vs no RS. For the 2-month post-RT timepoint alone, the RS group had more favorable QOL with 1 × MID and/or 2 × MID thresholds met for urinary irritation, bowel, and vitality domains. No durable clinically significant QOL differences occurred between RS groups even in the baseline sexual domain EPIC≥60/no ADT subgroup.

Conclusions: SBRT produced only modest, largely transient QOL declines that resolved by ∼6 months. RS did not confer a durable clinically meaningful QOL improvement; an isolated 2 × MID signal at 2 months favored RS in select domains, but this was transient, and non-durable.