Stereotactic Body Radiation Therapy Vs Hypofractionated Intensity Modulated Radiation Therapy for Localized Prostate Cancer (SHRAPNEL): An International Meta-Analysis of 24 Prospective Trials

Abstract

We aimed to characterize the efficacy and safety of stereotactic body radiation therapy (SBRT) vs. hypofractionated intensity modulated radiation therapy (HIMRT) for the definitive treatment of localized prostate cancer. The primacy outcome measure was efficacy, defined by 5-year disease free survival (DFS), based on the “prostate specific antigen nadir + 2 ng/mL” definition. The secondary outcome measure was safety, defined as incidence of grade ≥2 late genitourinary (GU) or gastrointestinal (GI) toxicities per the Common Terminology Criteria for Adverse Events or Radiation Therapy Oncology Group. PICOS/PRISMA/MOOSE methods were used to identify eligible studies. Inclusion criteria were: (1) patients diagnosed with localized prostate cancer; (2) HIMRT, >2 (<5) Gy/fraction; or SBRT, ≥ 5 Gy/fraction in ≤ 5 fractions; (3) either no control group, or another definitive RT control arm; (4) at least one of the primary or secondary outcome measures were reported; (5) single or multi arm phase III (HIMRT) and/or phase I-II (SBRT) prospective study. Androgen deprivation therapy (ADT) was prescribed per protocol specifications. Weighted random effects meta-analyses using the DerSimonian and Laird method were conducted to characterize summary effect sizes for each outcome. Meta-regression and Wald-type tests were used to compare each outcome by treatment, where the null hypothesis was rejected for p<0.05. After initial screening of 130 studies, 24 prospective trials containing 6,952 patients (3,027 SBRT [57% low-risk, 37% intermediate, 6% high] and 3,925 HIMRT [25% low, 60% intermediate, 15% high]) from 3 countries were included. ADT use was assessed (12% SBRT and 61% HIMRT). The median follow-up was 5.8 years for SBRT vs. 5.2 years for HIMRT. Prescription doses for SBRT ranged from 19-38 Gy in 2-4 fractions to 50 Gy in 5 fractions, and those of HIMRT ranged from 51.6 Gy in 12 fractions to 70.2 Gy in 26 fractions. The 5-year DFS rates for SBRT ranged from 81%-100% vs. 79%-91% for HIMRT. The random effects estimate for 5-year DFS was 95.3% (95% confidence Interval [CI]: 92.1%-97.8%) for SBRT vs. 85.7% (95% CI: 82.7%-88.5%) for HIMRT (p=0.0004). Rates of late grade 2+ GU toxicity for SBRT ranged from 0%-32.1% vs. 6.1%-49.3% for HIMRT. The random effects estimate for late grade 2+ GU toxicity was 14.4% (95% CI: 10.9%-18.3%) for SBRT vs. 20.9% (95% CI: 9.8%-34.7%) for HIMRT (p=0.27). Rates of late grade 2+ GI toxicity ranged from 0%-29% for SBRT vs. 7.4%-25.6% for HIMRT. The random effects estimate for late grade 2+ GI toxicity was 2.7% (95% CI: 1%-4.9%) for SBRT vs. 14.7% (95% CI: 9.4%-20.9%) for HIMRT (p<0.0001). In this select patient population with prospectively collected outcomes, SBRT had superior biochemical control with fewer long-term GU and GI toxicity rates over HIMRT in the treatment of localized prostate cancer. Prospective clinical trials directly comparing these therapies are pending, but these results support standard use of SBRT in the interim.