Clinically Targetable Mutation Status Associated with Local Control Following Radiation Treatment for Brain Metastases

Abstract

Brain metastases can harbor genetic mutations distinct from the primary tumor. It is unclear whether brain metastasis that contain clinically targetable mutations confer a local control benefit following radiation therapy (RT). Thus, we examined targetable mutation status within patients with whole-exome sequenced brain metastases that received radiation and assessed association with RT response. We retrospectively reviewed 58 patients with brain metastases that received RT from 2003-2014 and underwent whole-exome sequencing of matched brain metastases, primary tumor, and normal tissues. We have previously reported that whole exome analysis evaluated 95,431 gene alterations across our entire dataset and found 330 that met TARGET criteria as being clinically informative. Association between clinical factors, genetic alternation status, and local control was evaluated by Cox regression analysis. We performed Kaplan-Meier to assess local failure and overall survival based on presence of clinically targetable mutation status. 58 patients within our dataset received intracranial radiation therapy for brain metastases. The median follow-up was 16.5 months. Thirteen patients received SRS alone, 25 patients received WBRT alone, and 20 patients received SRS and WBRT. Forty-eight lesions were treated with SRS, 31 lesions (65%) were treated with single fraction (median dose 18 Gy, range 16-22 Gy) and 17 lesions were treated with a hypofractionated regimen (25 Gy in 5 fractions). The median dose for WBRT was 35 Gy in 2.5 Gy fractions (range 30-39 Gy). The most common tumor histologies were lung adenocarcinoma (N=22, 38%) and breast carcinoma (N=14, 24%). Patients were identified with brain metastases that contained a potentially favorable mutation status for treatment sensitivity to inhibitors for the cyclin-dependent kinase (CDK) pathway (n= 27), the PI3K/AKT/mTOR pathway (n= 26), the Her2/EGFR pathway (n= 18), and the MAPK pathway (n= 18). Patients that received targeted therapy (n=18) and immunotherapy (n=3) comprised 36% of the dataset. 19 radiographic local failure events were observed with a median time to failure of 9.2 months (range, 1-40 months). A unique clinically actionable mutation was observed in 16 of 58 (29%) of brain metastases not detected in the clinically sampled primary tumor. Clinically actionable mutation status is associated with local control following radiation therapy for brain metastases. Prospective trials involving patients with genetically characterized brain metastases treated with targeted therapeutic strategies in combination with radiation are needed.