What drives patient outcomes in brain metastases: Number, volume, or biology?
Central Nervous System Tumors
2017 ASCO Annual Meeting
Poster Board Number:
Poster Session (Board #313)
J Clin Oncol 35, 2017 (suppl; abstr 2071)
Author(s): Rupesh Kotecha, Jacob A Miller, Samuel T. Chao, Alireza Mohammad Mohammadi, Erin Sennett Murphy, John H. Suh, Gene H. Barnett, Michael A. Vogelbaum, Lilyana Angelov, Manmeet Singh Ahluwalia; Department of Radiation Oncology, Cleveland Clinic Foundation, Cleveland, OH; Cleveland Clinic, Cleveland, OH; The Cleveland Clinic, University Heights, OH
Background: To delineate the prognostic importance of number of brain metastases (BM), lesion volume, and biology on overall survival. Methods: Patients treated for BM with whole-brain radiation therapy (WBRT), surgery, and/or stereotactic radiosurgery (SRS) at a single tertiary care institution from 1997-2015 were reviewed. The primary outcome was overall survival. Multivariable proportional hazards regression was used to adjust for confounding factors. Results: 3,955 patients with 16,189 BM were included in the analysis. There was a reduction in median survival with increasing number of lesions [1 lesion, 10.2 months; 2-4 lesions, 7.2 months, Hazard Ratio (HR) 1.36; 5-10 lesions, 5.6 months, HR: 1.69; > 10 lesions, 5.5 months, HR: 1.57, p< 0.001]. Among 1,651 patients (35%) who underwent SRS, there was a similar reduction in median survival with increasing lesion number [1 lesion, 12.2 months; 2-4 lesions, 10.1 months, HR 1.20; 5-10 lesions, 8.4 months, HR 1.41; > 10 lesions, 6.9 months, HR 1.19], p< 0.001). Among patients who underwent upfront SRS, increasing number of BM did not adversely affect survival in those with the smallest intracranial disease volume (≤0.4 cc, 10th percentile, p= 0.39), but was associated with inferior survival in patients with larger disease volumes (≤1.1 cc, 25th percentile, p= 0.05; ≤2.6 cc, 50th percentile, p= 0.005; ≤6.3 cc, 75thpercentile, p= 0.006, and ≤12.2 cc, and 90th percentile, p =0.004). After partitioning the cohort into molecular subsets, patients with ALK+ disease had no difference in survival based on either lesion number or volume. Patients with EGFR+, HER2+, and luminal B disease had no difference in survival based on number of metastases, while patients with BRAF V600+and luminal A disease had no difference in survival based on intracranial disease volume. Conclusions: Number of BM closely correlates with survival in the majority of patients and intracranial disease volume impacts survival independent of number of metastases. For patients with certain tumor subtypes (ALK+), intracranial disease burden appears to have no correlation with survival. Molecular profile characterization is important to identify patients with favorable subtypes given available treatment options.
Categories: Brain Tumor