Brain metastases in ovarian cancer: Association of outcomes with timing of diagnosis.

Abstract

e17565 Background: Brain metastases (BM) in ovarian cancer (OC) are rare and late events during the disease course. We report our data on patients with OC who developed BM and compared the outcomes with subjects with evidence of BM at the time of diagnosis from the Surveillance, Epidemiologic, and End Results (SEER) national database. Methods: Retrospective chart review of all patients with brain metastasis in the setting of previously diagnosed ovarian cancer was performed in single NCI designated comprehensive cancer institute. Inclusion criteria were: primary cancer documented as ovarian or tubal, evidence of BM, treatment and follow up information. For comparison and due to rare nature of ovarian cancer cases with BM we utilized the SEER database. SEER data base reports only on patients with OC who had BM at the time of diagnosis. Study groups were: 1- BM at diagnosis 2- BM during disease course. Primary outcomes were overall death, and overall survival (OS) months. Secondary outcome was survival months after diagnosis of BM. Variables are reported as median and interquartile range (IQR) or number (%). Mean survival was calculated using Kaplan Meier analysis and Mantel-Cox rank test. Cox regression analysis was performed to calculate adjusted hazard ratios (HR). Results: 106 and 24 subjects were included with evidence of BM at diagnosis (Group 1, 2010-2016, SEER database) and during the disease (Group2, 1992-2017, institutional data), respectively. Age at diagnosis of OC was significantly higher in group 1 (64 year vs. 55 year, p = 0.002). In group 2, BM were diagnosed at median interval of 33.5 months. The median OS month was 3 [1, 10] months and 58 [21, 122] in group 1 and 2, respectively (p < 0.001). Although overall mortality rate was not statistically significantly different between the study groups, survival months after diagnosis of BM in group 2 was significantly higher than group 1 (14 months [4, 32] versus 3 months [1, 10], p = 0.002). Cox-regression analysis revealed 58% decrease in the hazard of overall death when the BM is diagnosed during the disease course. (Adjusted HR = 0.42, 95% CI: 0.19-0.91, p = 0.03). Conclusions: Outcomes of OC with BM is poor with high mortality rate regardless of timing of diagnosis of BM. Our study revealed that patients with diagnosis of BM during the disease course have statistically significantly longer survival after diagnosis of BM compared to those with BM at the time of diagnosis. In order to understand the occurrence of BM in OC and these differences, it is necessary to gain knowledge in the genetic and immune mechanisms that drive and allow the development of BM in ovarian cancer. This will allow for development of effective therapeutic strategies to treat BM, which is of critical importance now as we are gaining better control of peripheral disease in women with ovarian cancer.