Local failure after stereotactic radiosurgery (SRS) for intracranial metastasis: analysis from a cooperative, prospective national registry

Introduction: Stereotactic radiosurgery (SRS) has been increasingly employed to treat patients with intracranial metastasis, both as a salvage treatment after failed whole brain radiation therapy (WBRT) and as an initial treatment. “Several studies have shown that SRS may be as effective as WBRT with the added benefit of preserving neuro-cognition”. However, some patients may have local failure following SRS for intracranial metastasis, defined as increase in total lesion volume by 25% after at least 3 months of follow up. Methods: The SRS registry, established by the Neuro point alliance (NPA) under the auspices of the American Association of Neurological Surgeons (AANS), was queried for patients with intracranial metastasis receiving SRS at the participating sites. Demographic, clinical symptoms, tumor, and treatment characteristics as well as follow up status were summarized for the cohort. A multivariable explanatory cox- regression was performed to evaluate the impact of each of the factors on time to local failure.at last follow-up. Results: A total of 441 patients with 1255 intracranial metastatic lesions undergoing SRS were identified. The most common primary cancer histology was non-small cell lung cancer (43.8%, n = 193). More than half of the cohort had more than 1 metastatic lesion (2–3 lesions: 29.5%, n = 130; more than 3 lesions: 25.2% (n = 111). The average duration of follow-up for the cohort was found to be 8.4 months (SD = 7.61). The mean clinical treatment volume (CTV), after adding together the volume of each lesion for each patient was 5.39 cc (SD = 7.6) at baseline. A total of 20.2% (n = 89) had local failure (increase in volume by > 25%) with a mean time to progression of 7.719 months (SD = 6.09). The progression free survival (PFS) for the cohort at 3, 6 and 12 months were found to be 94.9%, 84.3%, and 69.4%, respectively. On multivariable cox regression analysis, factors associated with increased hazard of local failure included male gender (HR 1.65, 95% CI 1.03–2.66, p = 0.037), chemotherapy at or before SRS (HR = 2.39, 95% CI 1.41–4.05, p = 0.001), WBRT at or before SRS (HR = 2.21, 95% CI 1.16- 4.22, p = 0.017), while surgical resection (HR 0.45, 95% CI 0.21–0. 97, p = 0.04) and immunotherapy (0.34, 95% CI 0.16–0.50, p = 0.014) were associated with lower hazard of local failure. Conclusion: Factors found to be predictive of local failure included higher RPA score and those receiving chemotherapy, while patients undergoing surgical resection and those with occipital lobe lesions were less likely to experience local failure. Our analyses not only corroborate those previously reported but also demonstrate the utility of a multi-institutional registry to advance real-world SRS research for patients with intracranial metastatic lesions.