SRS Improves Survival Vs Hippocampal Avoidant-WBRT in SCLC Brain Metastases

Stereotactic radiosurgery (SRS) did not significantly improve time to neurocognitive failure compared with hippocampal-avoidant whole brain radiotherapy (HA-WBRT) plus memantine (Namenda) but was associated with significantly longer overall survival (OS) among patients with brain metastases from small cell lung cancer (SCLC), according to findings from the phase 3 NRG-CC009 trial (NCT04804644) presented at the 2026 American Society for Radiation Oncology (ASTRO) Annual Meeting.

The median OS was 17.4 months (95% CI, 10.9-19.6) with SRS vs 8.6 months (95% CI, 5.4-13.4) with HA-WBRT plus memantine (HR, 0.63; 95% CI, 0.41-0.96; log-rank P = .03). The 1-year OS rates were 58.9% vs 40.8%, respectively, and the 2-year OS rates were 31.1% vs 18.5%.

For the primary end point, the 6-month neurocognitive failure rate was 61.7% in the SRS arm vs 59.2% in the HA-WBRT plus memantine arm (HR, 0.77; 95% CI, 0.50-1.16; Gray P = .66).

“These findings support SRS as a treatment option for patients with [brain metastases] from [small cell lung cancer],” presenting author Chad G. Rusthoven, MD, associate professor of radiation oncology at the University of Colorado Anschutz, said during the presentation. He added that the superior OS observed with SRS addressed a fundamental historical objection to upfront SRS alone in this population, namely the concern that omitting WBRT could worsen survival.

Although SRS is the preferred treatment for limited brain metastases across most tumor histologies because of better cognitive preservation and quality of life vs WBRT, patients with SCLC were excluded from the randomized trials that established SRS. Investigators had been concerned that omitting WBRT in SCLC could lead to rapid central nervous system progression, increased neurologic mortality, and worse OS. As a result, WBRT has remained the standard of care for brain metastases from SCLC, with HA-WBRT plus memantine representing the preferred whole brain approach for cognitive preservation per NCCN guidelines. According to Rusthoven, NRG-CC009 is the first randomized phase 3 trial to compare SRS with HA-WBRT plus memantine in this setting.

Several shifts in SCLC management have made the comparison more pressing, the presentation noted. Reduced use of prophylactic cranial irradiation, paired with increased brain MRI surveillance, has led to more frequent identification of brain metastases, including limited disease that may be amenable to SRS. At the same time, immunotherapy and other systemic advances have improved SCLC prognoses overall, magnifying the importance of long-term neurocognitive preservation.

On multivariable analysis, SRS remained associated with superior OS (HR, 0.52; 95% CI, 0.34-0.81; P = .0034), and a Karnofsky performance status (KPS) of 70 to 80 vs 90 to 100 was associated with worse OS (HR, 2.60; 95% CI, 1.67-4.05; P <.0001). On cause-specific multivariable analysis of neurocognitive failure, there was no significant difference by treatment arm (HR, 0.90; 95% CI, 0.58-1.40; P = .6339), while older age was associated with a higher risk of neurocognitive failure (HR, 1.03; 95% CI, 1.00-1.06; P = .0262).

Regarding secondary end points, there were no significant differences in neurologic mortality between arms, with 1-year rates of 4.6% with SRS vs 9.4% with HA-WBRT plus memantine and 2-year rates of 11.5% vs 20.6% (Gray P = .36). The cumulative incidence of intracranial failure at 12 months was higher with SRS at 27.9% vs 12.6% (Gray P = .0034); however, this difference was not statistically significant on cause-specific multivariable analysis (HR, 1.16; 95% CI, 0.64-2.09; P = .6253). The 1-year local brain failure rates were 15.3% vs 12.6% (Gray P = .13), and the 1-year rates of leptomeningeal disease were 1.5% vs 0% (Gray P = .19).

Treatment-related adverse effects (TRAEs) were more frequent with HA-WBRT plus memantine overall, with no TRAEs reported in 60.9% of patients in the SRS arm vs 38.8% in the HA-WBRT plus memantine arm. However, rates of grade 3 or 4 TRAEs did not differ significantly (8.7% vs 14.9%; P = .26). The most common TRAEs with HA-WBRT plus memantine were fatigue (35.8%), dizziness (14.9%), memory impairment (13.4%), decreased appetite (11.9%), and alopecia (11.9%). With SRS, the most common TRAEs were fatigue (18.8%), headache (11.6%), nausea (5.8%), and alopecia (5.8%).

In the trial, patients with brain metastases from SCLC underwent neurocognitive function testing and were randomly assigned to receive SRS or HA-WBRT plus memantine. Randomization was stratified by Diagnosis-Specific Graded Prognostic Assessment (DS-GPA) score and number of brain metastases. HA-WBRT was delivered at 30 Gy in 10 fractions, memantine was recommended for a 6-month duration, and single-fraction or multifraction SRS was allowed at physician discretion with volume-based dose guidelines.

Eligible patients had a KPS of at least 70, no prior brain radiation, and brain metastases measuring at least 5 mm outside the hippocampi and no more than 4 cm in largest diameter, with a total brain metastasis volume of no more than 30 cm³. There was no upper limit on the number of brain metastases after an amendment to the initial limit of 10. Neurocognitive failure was defined as a reliable change index–defined decline on at least 1 test in a battery that included the Hopkins Verbal Learning Test–Revised, Controlled Oral Word Association, and Trail Making Test Parts A and B, with testing and brain MRI performed at 2, 4, 6, 9, and 12 months.

The trial enrolled patients from February 2021 to June 2026 at 42 institutions in the US and Canada. Of 158 patients enrolled, 151 were randomly assigned to SRS (n = 75) or HA-WBRT plus memantine (n = 76); the trial closed to accrual once 91 events were reported, before reaching its initial goal of 200 patients. Overall, 15 patients (9.9%) did not receive any treatment, and patients were analyzed on an intention-to-treat basis. At the time of analysis, median follow-up was 9.5 months for living patients.

Baseline characteristics were well balanced between arms. The median age was 67 years in the SRS arm and 68.5 years in the HA-WBRT plus memantine arm, the median number of brain metastases was 2 in both arms, and extracranial metastases were present in 52% and 53% of patients, respectively.

Patient-reported quality of life and salvage therapy analyses will be reported in the future.

Reference

Rusthoven CG, Paulus R, Gondi V, et al. NRG-CC009: a phase III trial of stereotactic radiosurgery (SRS) vs hippocampal-avoidant whole brain radiotherapy (HA-WBRT) for brain metastases from small cell lung cancer (SCLC). Presented at: 2026 American Society for Radiation Oncology Annual Meeting; Boston, MA. LBA 2.