PURPOSE: Ovarian cancer is a heterogeneous disease with multiple histological subtypes and a wide range of genetic aberrations. While high-grade serous carcinoma (HGSC) is the most common type, clear cell carcinoma (CCC) of the ovary is notoriously the most challenging to treat and exhibits very low response rates to standard therapies. However, efforts to understand CCC and develop new therapeutic approaches to this subtype have been limited because it represents a minority of ovarian cancer cases in the United States and Europe. In contrast, CCC accounts for more than 30% of all ovarian cancer in Japan. To improve the survival of patients with ovarian CCC, a deeper understanding of the molecular features of available model systems is needed. Our goal is to characterize a panel of CCC cell lines genomically and functionally and identify those that can serve as tractable model systems for future in vivo drug discovery studies. METHOD: First, we characterize 9 CCC cell lines (ES-2, TOV21G, OVTOKO, OVMANA, OCI-C5x, JHOC-5, JHOC-7, JHOC-9, and OVISE) with whole exome sequence and proteomics approaches (reverse phase protein array; RPPA). Secondly, we performed in vitro assays, including soft-agar colony formation assays and MTT assays using standard chemotherapies. Thirdly, to test in vivo tumorigenic potential, by injecting 5 million cells of luciferized CCC lines in NSG female mice using both the subcutaneous route and the intraperitoneal route. Imaging was performed weekly using the In Vivo Imaging System. RESULTS: Among the 9 CCC lines, seven (TOV21G, OVTOKO, OVMANA, OCI-C5x, JHOC-5, JHOC-9, and OVISE) harbor ARID1A mutation, which is the most prevalent mutation in CCC (50%). The second common mutation in CCC (40%), PIK3CA mutation, and was detected in 4 cells (TOV21G, OVMANA, OCI-C5x, and OVISE). The ES-2 line has TP53 and BRAF mutation and its genomic profile is inconsistent with CCC. A BRCA2 mutation was found in OVTOKO and JHOC-5, finding not previously reported. Principal component analysis of RPPA data showed distinct clusters between the 9 CCC lines and the 6 HGSC lines. Interestingly, we also observed two distinct clusters within the CCC lines. Consistent with our genomic analysis, ES-2 correlated with HGSC lines based on RPPA data. In our in vitro drug studies, OVTOKO, OCI-C5x, and OVISE exhibited a Carboplatin/Paclitaxel resistance phenotype. In xenograft study, 4 cell lines (ES-2, TOV21G, OVTOKO, and OCI-C5x) formed tumor within a month, suggesting they are useful tools for in vivo studies. In contrast, OVMANA, JHOC-7, JHOC-9, and OVISE take more than 100 days to form the tumors. Soft-agar colony formation results correlated with how long it took to develop xenograft tumor in each line. CONCLUSION: Our genomic studies identified aberrations in CCC lines not previously described. The proteomics data revealed two clusters within the CCC lines that might represent functionally distinct groups. Finally, we identified four cell lines (ES-2, TOV21G, OVTOKO, and OCI-C5x) that readily form tumors in mice and could be used for future in vivo drug studies. However, ES-2 appears to cluster more closely with HGSC and may not represent the CCC histotype. Citation Format: Yasuto Kinose, Dorothy Hallberg, Gordon Mills, Tan Ince, Victor Velculescu, Fiona Simpkins, Ronny Drapkin. COMPREHENSIVE GENOMIC, PROTEOMIC, AND EXPERIMENTAL CHARACTERIZATION OF OVARIAN CLEAR CELL CARCINOMA CELL LINES FOR IMPROVED DRUG DEVELOPMENT [abstract]. In: Proceedings of the 12th Biennial Ovarian Cancer Research Symposium; Sep 13-15, 2018; Seattle, WA. Philadelphia (PA): AACR; Clin Cancer Res 2019;25(22 Suppl):Abstract nr GMM-036.