Abstract P6-16-05: New graded prognostic assessment (GPA) in breast cancer brain metastases in a contemporary cohort

Abstract

Abstract Background: Breast cancer is the second most common cause of brain metastases (BCBM). We evaluated prognostic factors for overall survival (OS) in contemporary cohort of BCBM patients treated at a tertiary care institution. Methods: With IRB approval, Cleveland Clinic’s database was used to identify BCBM patients treated between 2000 and 2013. OS from the diagnosis of BCBM was the primary end point. Cox proportional hazards models with stepwise variable selection were used for data analysis. Results: 562 female patients were included for this analysis. Karnofsky performance scale (KPS) was 90-100 in 204 patients (41%), 70-80 in 223 (44%) and <70 in 76 (15%) patients. Two hundred, eighty nine (52%) patients were treated with whole brain radiation therapy (WBRT), 89 (16%) with stereotactic radiosurgery (SRS), 66 (12%) with WBRT and SRS, 44 (8%) underwent surgery and WBRT, 16 (3%) underwent surgery and SRS. Median OS for the entire cohort was 11.2 months (95% C.I., 10.0-13.1). Median OS in Luminal B (triple positive), Her 2 positive, Luminal A (ER/PR positive, Her 2 negative) and basal (triple negative) patients was 23.3, 14.5, 10.0 and 8.3 months respectively. Disease specific graded prognostic assessment (GPA) for BCBM is based on KPS, age at diagnosis, and subtype. Overall it was prognostic for OS (p <.0001), however separation between groups was variable. Starting de-novo, KPS, age at diagnosis of BCBM, and subtype were again identified as prognostic, although the scoring was different. Additional independent predictors were the number of sites of extra- cranial metastases, leptomeningeal disease, control of primary cancer and location of BCBM were found to be independently prognostic for survival. A new GPA was formulated by assigning "points" (weights) to the levels within each of these seven factors that are proportional to the regression coefficient estimates in the final model and then adding the total number of points present. Prognostic factors in BCBM on multivariate analysisFactorPointsHazard RatiopKPS 0.41 (0.32-0.52)<0.00180-10013 <700 Breast Cancer Subtype ( basal as reference) Luminal B100.49 (0.32-0.74)<0.001HER280.58 (0.43-0.77)0.007Luminal A60.63 (0.48-0.77)0.002Basal0 Number of sites of Extra-Cranial Metastases 0.65 (0.56-0.77)<0.0010 or 112 2-34 >30 Leptomeningeal Disease 0.51 (0.38-0.67)<0.001No10 Yes0 Primary Controlled 0.68 (0.53-0.88)0.003Yes5 No0 Age at Diagnosis of BCBM 0.75 (0.59-0.96)0.02<504 >/=500 Location of Brain Metastasis 0.76 (0.61-0.96)0.02Infratentorial or Supratentorial4 Both0 The new GPA consists of four groups: unfavorable, intermediate 1, intermediate 2 and favorable with OS of 2.7, 9.1, 18.5 and 29.5 months respectively. New GPA and Conventional GPANew GPAGroupNumber of pointsNumberMedian OS (months)Unfavorable</=2184 (19%)2.7Intermediate-122-36153 (34%)9.1Intermediate-237-45133 (30%)18.5Favorable>4578 (17%)29.5Conventional GPAUnfavorable0-155 (12%)4.8Intermediate-11.5-2130 (29%)9.1Intermediate-22.5-3148 (33%)12.9Favorable3.5-4116 (26%)20.4 Conclusions: A new GPA for BCBM is proposed. Citation Format: Manmeet S Ahluwalia, Ming Chi, Vyshak Alva Venur, Thomas Budd, Lilyana Angelov, Samuel Chao, Paul Elson, Gene H Barnett, Jame Abraham. New graded prognostic assessment (GPA) in breast cancer brain metastases in a contemporary cohort [abstract]. In: Proceedings of the Thirty-Seventh Annual CTRC-AACR San Antonio Breast Cancer Symposium: 2014 Dec 9-13; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2015;75(9 Suppl):Abstract nr P6-16-05.