Investigators noted that prior randomized trials published in 2017 supported the use of postoperative SRS in resected brain metastases, but the 1-year surgical bed recurrence rates with single-fraction SRS in those studies were 40% and 28%, respectively. Data from one of those trials also showed that larger preoperative tumor size was linked with greater recurrence risk; tumors larger than 2.5 to 3.5 cm (HR, 8.3; 95% CI, 2.5-27.6; P = .0005) and larger than 3.5 cm (HR, 7.1; 95% CI, 2.1-24.1; P = .0018) carried higher risk vs those 2.5 cm or smaller. Retrospective series of fSRS in this setting reported 1-year surgical bed control rates ranging from 84% to 93%.
Patients eligible for A071801 were aged 18 years or older with 1 to 4 brain metastases, including 1 completely resected metastasis measuring at least 2 cm within 30 days of registration, with a resection cavity smaller than 5.0 cm. Patients could have up to 3 unresected brain metastases smaller than 4 cm and were required to have a Karnofsky performance status (KPS) of at least 60. Those with small cell carcinoma, germ cell tumors, lymphoma, leptomeningeal disease (LMD), peri-optic or peri-brainstem lesions within 5 mm, prior whole-brain radiotherapy (WBRT), inability to tolerate contrast-enhanced brain MRI, or pregnancy were excluded.
After pre-registration and completion of electronic patient-reported outcome measures, patients were randomly assigned to single-fraction SRS or fSRS delivered in 3 to 5 fractions. Randomization was stratified by age, use of targeted therapy, use of immunotherapy, and cavity diameter. In the SRS arm, the resection cavity received 12 Gy to 20 Gy based on cavity volume. In the fSRS arm, cavities smaller than 30 cc received 27 Gy in 3 fractions (91% of patients) and those 30 cc or larger received 30 Gy in 5 fractions (9%). Central radiotherapy quality assurance and site-based SRS credentialing were employed.
The primary end point was surgical bed control; WBRT or death without prior surgical bed recurrence was not considered an event. Secondary end points included OS, patterns of recurrence, time to salvage WBRT, quality of life (QoL), emotional well-being, functional independence, cognitive function, adverse effects and radiation necrosis. The trial assumed 1-year surgical bed control rates of 65% with SRS and 82.4% with fSRS, with 90% power and a 2-sided alpha of 0.05, for a target enrollment of 238 patients. The primary analysis was planned at 67 surgical bed recurrences; a backstop analysis was ultimately used with 46 events after 3 years of follow-up in all patients.
Between October 2019 and October 2022, 254 patients were pre-registered and 242 were randomly assigned to SRS (n = 122) or fSRS (n = 120). A total of 113 and 110 patients, respectively, were evaluable. The median age was 65.0 years in the SRS arm and 63.0 years in the fSRS arm, and 59.8% and 48.3% of patients, respectively, were female. Lung was the most common primary tumor site (44.3% vs 40.0%), and the median maximum diameter of the resected metastasis was 3.3 cm vs 3.4 cm. Investigators reported that the study arms were well balanced.
The 1-year rate of freedom from salvage WBRT was 94% with fSRS vs 84% with SRS (stratified log-rank P = .06). At 1 year, rates of local failure in unresected metastases were 3% vs 7% (P = .06), new brain metastases were 34% vs 38% (P = .19), nodular meningeal disease were 7% vs 5% (P = .24), classical LMD were 3% vs 4% (P = .60), and any intracranial recurrence were 43% vs 47% (P = .14) in the fSRS and SRS arms, respectively. A stratified multivariable Cox model adjusting for KPS, primary tumor site, gender, race, and number of unresected brain metastases showed a similar association between study arm and OS (HR, 0.66; 95% CI, 0.46-0.94; P = .02).
Rates of radiation necrosis were 13% with fSRS and 11% with SRS, and rates of cerebral edema were 8% and 9%, respectively. Patient-reported outcomes did not differ significantly between arms, including FACT-BR total score (150.9 vs 147.8; P = .95), FACT-BR emotional well-being (17.9 vs 18.4; P = .65), and LASA QoL (74.8 vs 72.2; P = .54). Median duration of functional independence by KPS was 11.9 months with fSRS vs 9.0 months with SRS (P = .09) and 25.4 months vs 23.0 months by Barthel index (P = .29). Median time to cognitive progression was 11.4 months vs 13.2 months (P = .94).
Investigators concluded that the improvement in surgical bed control with fSRS was achieved without an apparent increase in toxicity, supporting fSRS as a preferred postoperative approach for patients with a resected brain metastasis of 2 cm or larger.
Reference
Aizer AA, Brown PD, Ballman KV, et al. Postoperative single fraction SRS vs. fractionated SRS (fSRS) for resected brain metastases: results from the phase III Alliance A071801 randomized trial. Presented at: 2026 ASTRO Annual Meeting; September 26-30, 2026.
