A real-world study of US patients with metastatic ovarian, fallopian tube, and peritoneal cancer (mOFPC) using integrated electronic health records (EHR) and claims datasets.

e17548 Background: Ovarian cancer is the most lethal gynecologic cancer. Detailed epidemiologic descriptions from a real-world database of a large population of ovarian and related cancer patients can provide insights into the characteristics, treatment patterns, and outcomes of these patients, and could serve as a useful benchmark while developing new therapies. Methods: This retrospective analysis uses the ConcertAI RWD360 dataset which is a real-world de-identified dataset of more than 6 million patients from across the United States derived from various EHR systems and integrated with a large administrative open claims dataset (>90% overlap). 60429 patients in this dataset had OFPC out which 12406 patients had metastatic cancer. We present here the key characteristics of this mOFPC cohort including histology, biomarkers and top treatment patterns. We also performed overall survival (OS), time to treatment discontinuation (TTD) and time to next treatment (TTNT) analyses starting from initiation of 1st line of therapy (LoT) for all patients who received systemic treatment after metastatic diagnosis (8308 patients). Results: The top 5 histologies and biomarkers tested in this mOFPC cohort are presented in Table. The median OS from start of 1st LoT for this cohort was 2.45 years. The median TTD and TTNT across all 1st line therapies were 2.5 months and 5.5 months respectively. Top 2 treatments given as 1st LoT after metastatic diagnosis are also presented in Table. Chemotherapy (cis/carboplatin + pacli/docetaxel) and chemotherapy + bevacizumab were the most common 1st LoTs and there was no difference in the OS of patients treated with these two LoTs. Conclusions: This study describes the characteristics of a large US real-world cohort of mOFPC patients. Here we have presented data on the overall cohort, but further analysis of outcomes stratified by treatment arms, biomarkers and histology can be performed to better inform best treatments for specific sub-cohorts. [Table: see text]