5 Year Survivors from Brain Metastases Treated with Stereotactic Radiosurgery: Biology, Improving Treatments or Just Plain Luck?

Abstract

<h3>Purpose/Objective(s)</h3> Improvements in therapies have led to an increasing number of long-term survivors of brain metastases. This population has not been well characterized. The present series compares a population of 5-year survivors of brain metastases to a generalized brain metastasis population to assess for factors attributable to long term survival. <h3>Materials/Methods</h3> A single institution retrospective review was performed to identify 5-year survivors of brain metastases who have received stereotactic radiosurgery (SRS). Electronic medical records were queried to characterize this population based on clinical, demographic and treatment data. A historical control population of 747 patients with brain metastases was used to assess for similarities and differences between the long-term survivor population and the general population treated with upfront SRS. Statistical analysis was performed with SAS version 9.4 using the 2-sample t test, chi squared test, and Fine Gray competing risk analysis using death as a competing risk. <h3>Results</h3> From 9/99 to 2/17 a total of 2422 patients were treated with SRS for brain metastases at a single institution, where 95 patients (3.9%) were found to have survived over 57 months. In the historical control population, 25 of 747 patients (3.3%) lived beyond 57 months. In the long-term survivor cohort, primary tumors were from lung (53%), melanoma (19%), breast (12%), renal cell (9.5%) and other histology (7.4%). Median survival of the long-term survivor cohort was 85 months, with mean number of 2.2 (SD=2.0) metastases treated and 2% receiving WBRT prior to SRS. Two-sample t-test test found no difference between long term survivors and controls with regards to age (p=0.2), primary cancer distribution (p=0.7), number of metastases (p=0.9), insurance type (p=0.1) or median income level (p=0.3). 22% of long-term survivors were receiving systemic therapy at last follow-up, and 52% showed no evidence of clinical disease at the last follow-up. Cumulative incidence of neurologic death at 6, 8 and 10 years for the long-term survivor cohort was 5%, 15%, and 15%, respectively. In the historical controls, cumulative incidence of neurologic death reached a plateau at 34% after 4.9 years. 61% of long-term survivors received further SRS (median 2, range 2-9, IQR 2.3), while 7% received post SRS WBRT. 5-year survivors received more additional SRS (P<0.0001), while historical controls received more post SRS WBRT (P=0.0008). <h3>Conclusion</h3> Five-year survivors of brain metastasis represent a diverse histologic population, suggesting that biologically, for each cancer type, a small population of oligometastatic and indolent cancers exist. While neurologic death of this long-term survivors is lower than in historical controls, these patients can still, die of neurologic death.