Radiotherapy for localized prostate cancer has increasingly moved toward shorter treatment schedules, making the balance between treatment convenience, quality of life, toxicity, and disease control increasingly relevant. The phase 3 NRG-GU005 trial compared stereotactic body radiotherapy (SBRT) delivered in 5 fractions with moderately hypofractionated intensity-modulated radiation therapy (MH-IMRT) in patients with localized intermediate-risk prostate cancer.
The results were published in JAMA in the original investigation, “Stereotactic Body Radiotherapy vs Moderately Hypofractionated IMRT for Localized Intermediate-Risk Prostate Cancer: A Randomized Clinical Trial.” The study was published online on August 13, 2026.
Authors: Rodney J. Ellis, Stephanie L. Pugh, James B. Yu, Felix Y. Feng, Andre A. Konski, Robert L. Grubb III, Robert E. Wallace, David J. Gladstone, Cynthia Ménard, Joseph A. Miccio, Arthur J. Frazier, J. Daniel Pennington, Jeff M. Michalski, Daniel E. Spratt, Alvaro Martinez, Scott C. Morgan, Alina Mihai, Abhishek Ashok Solanki, Asim Amjad, Michael W. Straza, Guila Delouya, Thomas M. Schroeder, David T. Marshall, Nirav Kapadia, Akshar N. Patel, Terrence P. Cescon, Ali El-Gayed, Harold A. Yoon, Rebecca Paulus, Howard M. Sandler, and the NRG-GU005 Collaborative Authors.
Comparing 5-Fraction SBRT With MH-IMRT
NRG-GU005 (NCT03367702) was an international, open-label, randomized phase 3 trial conducted across 136 centers in Asia, Canada, Europe, and the United States. The study was activated in November 2017 and closed to accrual in June 2022.
Eligible patients had previously untreated, localized intermediate-risk prostate cancer. This included clinical stage T1-T2b disease with either Gleason 3+4 and a PSA level below 20 ng/mL, or Gleason 3+3 and a PSA level between 10 and 20 ng/mL. Patients with a prostate volume of 70 mL or greater or T3 disease on diagnostic MRI were excluded.
A total of 698 patients were randomized 1:1 to SBRT or MH-IMRT. The SBRT group received 36.25 Gy in 5 fractions, while the MH-IMRT group received either 70 Gy in 28 fractions or 60 Gy in 20 fractions. SBRT was delivered 2 to 3 fractions per week, at least every other day, and within fewer than 17 business days.
Patients were stratified by risk group and use of rectal manipulation, including a rectal balloon and/or rectal spacing gel. Treatment was analyzed according to the randomized assignment using an intention-to-treat approach.
Median age was 68 years, and median follow-up was 3.2 years. Most patients had Gleason 3+4 disease, with 73.5% having Gleason 3+4 and PSA below 10 ng/mL. Most also had T1c disease, and the majority had a Zubrod performance status of 0. More than half of patients used rectal spacing alone.
Treatment adherence was high. Almost all patients assigned to SBRT received the planned 36.25 Gy in 5 fractions, while in the MH-IMRT group, 70.6% received 28 fractions and 26.6% received 20 fractions. Treatment was delivered per protocol or with acceptable variation in more than 97% of patients in both groups.
Coprimary Endpoints: Quality of Life and DFS
The trial had two coprimary objectives. The first was to determine whether SBRT was superior to MH-IMRT in terms of patient-reported urinary irritative/obstructive and bowel quality of life at 2 years, measured using the Expanded Prostate Cancer Index Composite-26 (EPIC-26).
The second coprimary objective was to determine whether SBRT was superior to MH-IMRT in disease-free survival (DFS) at 3 years. At 2 years, there was no significant difference in the proportion of patients with a minimal clinically important decline in urinary irritative/obstructive quality of life. A minimal clinically important decline occurred in 35.4% of patients treated with SBRT and 33.7% receiving MH-IMRT, with a mean difference of −1.7% and a P value of .68.
Bowel-related quality of life, however, favored SBRT. At 2 years, a clinically important decline in the bowel domain was observed in 34.9% of patients receiving SBRT compared with 43.8% receiving MH-IMRT, corresponding to a mean difference of 8.9% and a P value of .03.
Secondary Patient-Reported Outcomes
Several secondary EPIC-26 outcomes also favored SBRT. At 1 year, a clinically important decline in the bowel domain occurred in 33.1% of patients in the SBRT group compared with 45.9% in the MH-IMRT group (P=.002). A decline in the sexual domain at 1 year occurred in 34.3% versus 43.9%, respectively (P=.03).
At 2 years, a clinically important decline in the urinary incontinence domain was observed in 25.9% of patients treated with SBRT and 34.7% treated with MH-IMRT (P=.02). Other prespecified comparisons across urinary irritative/obstructive, sexual, and hormonal domains were not statistically different.
Longitudinal analyses were also consistent with better bowel and urinary incontinence scores in the SBRT group. The investigators additionally found that absence of a rectal balloon or spacer was associated with worse bowel scores, while no significant interaction between rectal manipulation and treatment group was observed.
SBRT Did Not Demonstrate Superior Disease Control
At the interim DFS analysis, 91 DFS events had occurred. Based on the prespecified analysis, the hypothesis that SBRT was superior to MH-IMRT for DFS was rejected. The 3-year DFS rate was 88.6% with SBRT and 92.1% with MH-IMRT. The unadjusted HR was 1.38 (95% CI, 0.91-2.09). After adjustment for covariates, there remained no statistically significant difference between groups, with an HR of 1.40 (95% CI, 0.91-2.13; P=.12).
The analysis also identified a difference in PSA-defined biochemical failure. At 3 years, biochemical failure occurred in 7.8% of patients receiving SBRT compared with 4.2% receiving MH-IMRT. The adjusted HR was 1.82 (95% CI, 1.01-3.27; P=.046). Local failure at 3 years was similar between treatment groups, occurring in 1.2% with SBRT and 1.0% with MH-IMRT. Regional and distant failures were uncommon, and the small number of events prevented further analysis of these endpoints.
There was also no significant difference in overall survival between the groups. In the adjusted analysis, the HR for overall survival was 1.15 (95% CI, 0.55-2.41; P=.70).
Safety
Grade 3 or 4 genitourinary adverse events occurred less frequently with SBRT than with MH-IMRT, at 0.6% versus 2.5%, respectively (P=.04). Any-grade rectal hemorrhage was also less frequent with SBRT, occurring in 10.5% versus 17.3% of patients (P=.01). Fatigue was reported in 39.2% versus 50.8%, respectively (P=.002).
Interpreting the Findings
The investigators emphasized that the trial was designed to test whether SBRT was superior to MH-IMRT, rather than whether it was noninferior. Therefore, the failure to demonstrate superiority for DFS does not establish that the SBRT regimen used in NRG-GU005 is noninferior to MH-IMRT.
The study also used a modest SBRT dose prescription of 36.25 Gy in 5 fractions. In their discussion, the authors noted that DFS was numerically worse with SBRT and that worse disease control could emerge with further follow-up, as reflected by the higher PSA failure rate observed at 3 years.
Several factors could influence interpretation of the disease-control endpoints. The authors discussed the possibility that the lower PSA control with SBRT could be explained by a lower-than-expected biologically effective dose due to prostate cancer biology or by marginal missing of prostate cancer due to smaller radiation treatment volumes. These explanations were presented as hypotheses rather than established mechanisms.
Limitations
NRG-GU005 had several important limitations. First, because it was designed as a superiority trial, the study cannot determine whether the protocol-defined SBRT regimen is noninferior to MH-IMRT.
Second, the trial did not evaluate focal radiation dose escalation or boosting of dominant intraprostatic lesions. The investigators also noted that benign PSA bounce, which can occur after SBRT, may have affected biochemical failure and DFS endpoints. The trial was not designed to distinguish benign PSA bounce from true PSA failure.
In addition, quality-of-life and toxicity data beyond 2 years were not available, meaning that longer-term differences between the treatment strategies remain uncertain.
Takeaway
In the phase 3 NRG-GU005 trial, 5-fraction SBRT was associated with fewer clinically important declines in bowel-related quality of life and favorable findings in several secondary patient-reported outcomes compared with MH-IMRT.
However, SBRT was not superior for urinary irritative/obstructive quality of life or disease-free survival, and PSA-defined biochemical failure was higher with the SBRT regimen used in the study. Grade 3 or 4 genitourinary adverse events were less frequent with SBRT.
Importantly, because NRG-GU005 was designed as a superiority rather than a noninferiority trial, the results do not establish that the protocol-defined SBRT regimen is noninferior to MH-IMRT for disease control. The findings highlight the balance between treatment convenience, patient-reported quality of life, toxicity, and disease control when selecting a radiotherapy approach for localized intermediate-risk prostate cancer.
