Alliance A071801 NCT04114981
For1-4 brain metastases, one lesion >2cm resected, postop cavity RT
87% vs 81%
1 year surgical bed control, fSRS vs SRS, p=0.046
TL;DR1 year surgical bed control 87% vs 81% (p=0.046) favoring fSRS over single-fraction SRS after brain met resection; mOS 29 vs 20 mo.
Intact metastases were also treated per arm, so the OS signal (median 29 vs 20 months, HR=0.69, p=0.035) reflects a fractionation strategy across all treated lesions, not only the cavity. fSRS did not lower grade 1+ radiation necrosis (14% vs 10%), so the case for fractionating rests on control, not safety.
In pts with 1 to 4 brain metastases after resection of a lesion >2cm, this supports fractionated over single-fraction cavity SRS; it does not address unresected-only disease, neoadjuvant SRS, or WBRT.
Intact metastases were also treated per arm, so the OS signal (median 29 vs 20 months, HR=0.69, p=0.035) reflects a fractionation strategy across all treated lesions, not only the cavity. fSRS did not lower grade 1+ radiation necrosis (14% vs 10%), so the case for fractionating rests on control, not safety.
10 details 3 trials watching
Phase III multi-institutional cooperative group randomized trial, 242 randomized between October 2019 and October 2022. Stratified by age (60 years cut), targeted therapy or immunotherapy use within 4 weeks after RT, and maximal resection cavity diameter (3cm cut). Median follow-up 48 months.
1 to 4 brain metastases with resection of one lesion >2cm. Baseline characteristics reported as well-balanced; median surgical cavity target volume 16.5 for fSRS and 14.8 for SRS (p=0.41, units printed as cm in source).
SRS: 12-20Gy by volume in 1 fraction. fSRS: 27Gy/3 for <30cc or 30Gy/5 for >30cc. Intact metastases received SRS or fSRS per assigned arm, so fractionation differed across all treated lesions, not only the cavity.
Primary: time to surgical bed failure. Secondary: local control of intact brain metastases, overall survival, radiation necrosis.
Primary endpoint met in favor of fSRS, with OS and intact-lesion control moving the same direction (see table). No HR or CI for the primary endpoint in source.
| Endpoint | fSRS | SRS | Effect |
|---|---|---|---|
| Surgical bed control, 1 year | 87% | 81% | p=0.046 |
| Failure of treated intact mets | 4% | 11% | p=0.063 |
| Median OS | 29 months | 20 months | HR=0.69, p=0.035 |
| Grade 1+ radiation necrosis | 14% | 10% | no significant difference |
| Grade 1+ cerebral edema | 8% | 9% | no significant difference |
No significant differences in adverse event rates between arms. Grade 1+ radiation necrosis 14% fSRS vs 10% SRS; grade 1+ cerebral edema 8% vs 9%. Symptomatic or higher-grade necrosis not broken out in source.
N107C set postoperative single-fraction SRS as a standard against WBRT, and the MD Anderson randomised trial of cavity SRS vs observation tied local failure to larger cavities. This trial tests fractionation as the answer to that size problem, a practice the source describes as often favored but lacking high-level comparison.
Abstract only: the time-to-event primary is reported as a 1 year rate with a p-value near threshold, with no HR, CI, or per-arm N. Cavity-size subgroup results, the split between 27Gy/3 and 30Gy/5, leptomeningeal failure, and cause of death are not reported.
The local-control gain is modest and the survival difference is larger than that gain alone would predict, which points at intact-lesion control, neurologic death, or chance in a secondary endpoint. Authors conclude fSRS should be considered a standard of care after resection of larger brain metastasis; the necrosis data do not add a safety argument.
Randomised cooperative-group phase III, primary endpoint met with 48-month median f/u. fSRS already often favored for large cavities per source; this supplies the missing high-level data.
- Does the fSRS benefit hold in cavities 3cm or smaller?
- Postoperative fSRS vs neoadjuvant SRS for resectable brain metastasis active Neoadjuvant vs. Intraoperative vs. Adjuvant Resection Cavity Radiotherapy of Brain Metastases Phase NAn=90 · primary completion 2027-05 · preop vs intraop vs postop RT for resected brain metsrecruiting Trial of Preoperative Radiosurgery Versus Postoperative Stereotactic Radiotherapy for Resectable Brain Metastases Phase NAn=200 · primary completion 2028-12 · 1-fx preop SRS vs 5-fx postop SRT; 12 mo LMD incidencenot yet Study Comparing Neoadjuvant Versus Adjuvant Stereotactic Radiotherapy of Brain Metastases Phase NAn=68 · primary completion 2030-06 · neoadjuvant vs adjuvant SRT; local control at 6/12/24 mo
- What drives the OS difference: neurologic death or intact-lesion control?
📚 Sources · 📄 1 paper
Abstract
The longer read
The source frames this trial as the randomised comparison that fractionated cavity radiosurgery never had. Postoperative single-fraction SRS became a standard through N107C and the MD Anderson cavity trial, and the weakness that work exposed was local control as the target grew. Fractionation spread into practice on retrospective series and radiobiologic reasoning. Here the prespecified primary endpoint is positive: 1 year surgical bed control of 87% with fSRS versus 81% with SRS, p=0.046.
How much weight that carries depends on details the abstract withholds. The endpoint is time to surgical bed failure, yet the result is given as a landmark rate and a p-value sitting close to the threshold, with no hazard ratio, no confidence interval, and no per-arm denominators. A single-fraction arm controlling 81% of cavities at 1 year is not a failing comparator, so the question for the full report is whether the curves continue to diverge across the 48 months of median follow-up or whether the difference is established early and then holds flat. A reader should treat the direction as credible and the magnitude as provisional.
The survival result is the harder one to interpret. Median OS of 29 versus 20 months, HR=0.69, p=0.035, and a multivariable model that barely moves it (HR=0.68, p=0.032), is a larger effect than a modest difference in cavity control would be expected to produce. Comparisons of local brain-directed strategies have generally not shifted survival. Two features of the design offer a partial explanation. Intact metastases were treated according to the assigned arm, so the experiment is fractionation across every treated lesion, and intact-lesion failure was 11% with SRS versus 4% with fSRS (p=0.063). The other explanations are less comfortable: chance in one of several secondary endpoints, or an imbalance in systemic therapy after radiotherapy that stratification on planned use within 4 weeks would not capture. The source gives no cause of death, and without a neurologic versus systemic split the survival finding is better held as a hypothesis than as a second reason to change practice.
The toxicity data cut against a common rationale. Fractionation is often justified as kinder to surrounding brain, but grade 1+ radiation necrosis was 14% with fSRS and 10% with SRS, without a significant difference. Grade 1+ includes imaging-only change, and symptomatic rates are not reported, so this neither condemns nor supports fSRS on safety. It does mean the argument rests on control.
Transfer to practice has two gaps. Median cavity target volumes of 16.5 and 14.8 sit below the 30cc threshold, so the 27Gy/3 schedule probably carried most of the fSRS arm, and the evidence for 30Gy/5 in the largest cavities may be thin; the split is not reported. Eligibility began at a resected lesion >2cm, which reaches cavities many clinicians would still treat in one fraction, and although cavity diameter at 3cm was a stratification factor, no subgroup result is given. Leptomeningeal failure and any comparison with neoadjuvant radiosurgery sit outside what this abstract can answer.