Abstract 4001: Patient avatar models for ovarian cancer: Predicting patient response to treatment

Abstract Background: Despite recent advances in targeted therapies, ovarian cancer (OVCA) remains the deadliest gynecologic malignancy, with approximately 80% of cases diagnosed at advanced stages and an equally high recurrence rate despite high response rates to frontline chemotherapy. Current OVCA models fail to accurately replicate the intricate biological landscape of metastatic disease. In response, we developed an innovative 3D in vitro system that preserves the unique cellular complexity of patient’s original tumor. This model features key advancements, including prolonged (>21 days) culture of patient-derived microtumors, a perfusable microgel medium (Liquid-Like Solids, LLS), specialized microtiter plates for controlled "pumpless" media perfusion (Darcy Plates), and integrated high-resolution multiphoton confocal microscopy for in situ imaging. Methods: Tumors were collected at the time of OVCA surgery and fragmented into 1 µl sized microtumors, which were then suspended into a 3D Liquid Like Solid (LLS) medium and cultured in a perfusion (Darcy) plate. Cell viability and proliferation were assessed over 21 days using LIVE/DEAD™ Cell Imaging Kit and EDU staining, respectively. Immunofluorescence microscopy was used to identify epithelial OVCA cells and immune cell populations. Drug screens were performed in a 48-well perfusion plate using standard of care chemotherapy and targeted therapies and compared to patient response. Results: Optimal conditions were established using HPLM media supplemented with 10% FBS, 1% Pen/Strep, 2.5 mg/L Amphotericin B, 10 mM HEPES, 0.5 µM A83-01, and ITS, along with a perfusion flow rate of 100 µl/day per well. PAX8 nuclear expression confirmed epithelial OVCA cells. EDU staining demonstrated sustained cell proliferation and CD45 markers showed immune cell retention. Drug testing with standard of care chemotherapies like carboplatin and targeted agents revealed differential microtumor responses and was comparable to patient response. Conclusions: OVCA's complex biology and extensive metastatic potential present significant clinical challenges. This study highlights the potential of 3D patient-derived microtumors as a robust platform for drug screening, high-throughput assays, and biomarker discovery. By retaining tumor heterogeneity and microenvironmental interactions, this model offers a promising avenue for improving therapeutic strategies and personalized treatment approaches in OVCA management. Citation Format: Erin M. George, Yuxi Zhou, Alfonso Pepe, Athena Tsingelis, Angie Rivera, Diego Pedro, Katriana Johnson, Dominick Dag, Said Cifuentes Maury, Lucia Mazzacurati, Jack Famiglietti, Ryan Smolchek, Duy Nguyen, W. Gregory Sawyer. Patient avatar models for ovarian cancer: Predicting patient response to treatment [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 4001.