Abstract
Introduction/BackgroundAlthough most patients with ovarian cancer (OC) respond to first line (1L) treatment, 70% of women experience disease progression (PD) within 3 years. Identifying prognostic factors that impact survival is crucial to identify patients who may benefit from new treatment regimens such as maintenance therapies. The objective of this study was to assess the association between visible residual disease (VRD) following interval (IDS) or primary debulking surgery (PDS) and other clinical factors, and the risk of PD or death in patients with advanced OC in a real-world setting.MethodologyThis retrospective cohort study included patients diagnosed with invasive ovarian cancer between January 1, 2011 and February 29, 2020, from the Flatiron Health electronic health record-derived de-identified US database (most OC patients (87%) originate from community oncology practices). Inclusion/exclusion criteria are shown in table 1. The index date (ID) was defined as the last date of 1L treatment. Multivariate Cox regression models were used to identify demographic and clinical factors associated with time to next treatment (TTNT,a proxy for PD), defined as time from ID to start date of second line treatment, death, or last confirmed structured activity. Overall survival (OS) was defined as time from ID to death or last confirmed structured activity. Kaplan Meier analysis was used to assess median TTNT and OS.ResultsA total of 1,920 advanced OC patients with a median (25th, 75th percentile) age of 67.0 (58.0, 75.0) years were included. Most patients were white (74%) and originated from a community oncology practice (88%). While 67% of patients had evidence of a BRCA biomarker test, only 5% had evidence of a homologous recombination deficient (HRD) test.Results are shown in table 2. Statistically significant predictors of OS were VRD, age, BRCA status, Eastern Cooperative Oncology Group (ECOG) score, extent of debulking, histology, stage of disease, and therapy modality. Statistically significant predictors of TTNT included: VRD, BRCA status, extent of debulking, histology, practice type, race, stage of disease, and therapy modality.ConclusionVRD was a key predictor of TTNT and mortality in patients with advanced OC. Among biomarkers, BRCA status was a key predictor of OC outcomes; HRD could not be assessed due to lack of data. A key strength of this study is that it presents outcomes from community practices primarily, whereas most retrospective studies contain a higher proportion of academic practices.Abstract 374 Table 1Inclusion/Exclusion criteriaAbstract 374 Table 2TTNT and OS results (unadjusted)DisclosuresDr. Chase reports speakers’ bureau fees from GSK.Dr. González-Martín reports personal fees and non-financial support from AstraZeneca; Grant and personal fees from GSK, Clovis Oncology, Roche Holding AG, Merck & Co., Inc., Genmab, INMUNOGEN, Pharma Mar, S.A., and Oncoinvent AS.Drs. Kalilani, Perhanidis, Sansbury, Woodward and Gupta are employees of GlaxoSmithKline.