P03.03.A STEREOBRAIN: Stereotactic radiosurgery versus whole-brain radiotherapy in patients with 4-10 brain metastases - a non-randomized controlled trial

Abstract Background Since the JLGK0901 study described non-inferiority of stereotactic radiosurgery (SRS) of 5-10 compared to 2-4 brain metastases in 2014, it has been widely discussed if SRS should replace whole brain radiotherapy (WBRT) in 5 - 10 brain metastases though lacking randomized evidence. At our department, WBRT has been the treatment of choice in 4-10 brain metastases until 2017. This paradigm was entirely changed to include patients within the STEREOBRAIN trial (DRKS00014694). We designed this prospective controlled single arm trial to systematically introduce SRS to 4 to 10 brain metastases comparing this new treatment regimen with the former approach. We aimed at showing a survival benefit of this novel treatment paradigm. Material and Methods Inclusion criteria were 4-10 brain metastases, largest diameter 2.5 cm, all histologies except SCLC, germ cell tumors, lymphoma, and ECOG ≤ 2. The retrospective WBRT cohort was identically selected from consecutive patients from 2012 to 2017 in a 1:2 fashion. At a significance level of 5%, a total number of 99 events (deaths) across the two groups allows to detect an increase in median OS from 6 months (retrospective WBRT cohort) to 11 months (SRS) considered clinically relevant (and corresponding to a hazard ratio of 0.55 assuming exponential distributions of survival times within groups) with a power of 80%. Propensity-score matching was performed to take confounders into account given the inherent bias caused by different treatment periods. Multiple brain mets SRS was performed using a single-isocenter technique. Results Patients were recruited from 2017 - 2020 and end of F/U was July 1st 2021. Forty patients were recruited to the SRS cohort and 70 patients were eligible within the WBRT cohort (81 events altogether). Median follow-up, OS and intracranial progression free survival were 21.6 months (95%-CI 19.8-NA), 10.4 months (95%-CI 9.3-NA) and 7.1 months (95%-CI 3.9-14.2) for the SRS-cohort, and 61.4 months (95% 54.6-NA), 6.5 months (95%-CI 4.9-10.4) and 5.9 months (95%-CI 4.1-8.8) for the WBRT-cohort, respectively. The Cox-model yielded a trend for improved survival within the SRS-cohort, HR 0.65 (95%-CI 0.4-1.1); p=0.076. After optimal propensity score matching, OS was significantly superior for the SRS-cohort (HR 0.53 (95%-CI 0.32-0.86), p=0.01). No grade III toxicities were observed in the SRS-cohort. Conclusion Despite all limitations of a historical control and missing the target number of