Radiosurgery Compared To External Beam Radiotherapy for Localized Spine Metastasis: Phase III Results of NRG Oncology/RTOG 0631

Purpose/Objective(s): This study aimed to compare pain relief between radiosurgery (SRS)/stereotactic body radiation therapy (SBRT) and conventional external beam radiation therapy (cEBRT) for patients with 1 to 3 sites of spine metastases. Materials/Methods: In this open-label randomized phase 3 study, we enrolled patients with limited spine metastases involving 1-3 separate sites, each site involving up to 2 contiguous spine segments. Minimal epidural lesions at least a 3 mm gap from the spinal cord were included. The study randomly assigned patients to receive either SRS/SBRT 16 or 18 Gy in one fraction to the involved spine segment(s) or cEBRT 8 Gy in one fraction to the involved spine including one additional spine segment above and below the index level. The treatments were performed under robust protocol guidelines with central rapid review. The primary endpoint was pain control, defined as a 33 point improvement on the Numerical Rating Pain Scale (NRPS), at the treated spine segment at 3 months post-treatment. The study had 80% power with 352 patients to show a 40% improvement in pain response with 2:1 randomization in favor of SRS/SBRT. Results: Of the 353 patients, 339 were eligible with 209 randomized to SRS/SBRT and 130 to cEBRT. The baseline mean pain score at the index spine site was 6.06 (standard deviation [SD]=2.61) in the SRS/SBRT arm and 5.88 (SD=2.41) in the cEBRT arm. The 3 month change in pain score at the index site was -3.00 (SD=3.34) in the SRS/SBRT arm compared to -3.83 (SD=2.97) in the cEBRT arm. There was no difference in pain response between SRS/SBRT and cEBRT at 3 months (40.3% vs. 57.9%, respectively, one-sided p=0.99). All the evaluable cases were planned per protocol or with an acceptable variation for delineation of target volume and organs at risk, as well as for dosimetric compliance on both arms. There was no significant difference between the treatment arms in rates of adverse events or in FACT-G, BPI and EQ-5D scores. Conclusion: SRS/SBRT for spine metastases treating only the involved spine was safely performed without causing any increase in adverse effects, as compared to conventional EBRT treating to one additional vertebral body above and below the metastatic lesion(s). However, pain control at 3 months from the patient’s perspective was not improved using SRS/SBRT, due to the lower pain control rate than expected in the SRS arm. Acknowledgements: This project was supported by grants UG1CA189867 (NCORP) and the Division of Cancer Prevention from the National Cancer Institute (NCI)