PDX validation of a 3D microtumor platform.

person Ellen Sampson SageMedic Corp, Redwood City, CA info_outline Ellen Sampson, Katya Nikolov, Paul T. Henderson, Christian Apfel, Chong-xian Pan, Maike Zimmermann, Ai-Hong Ma

Authors person Ellen Sampson SageMedic Corp, Redwood City, CA info_outline Ellen Sampson, Katya Nikolov, Paul T. Henderson, Christian Apfel, Chong-xian Pan, Maike Zimmermann, Ai-Hong Ma Organizations SageMedic Corp, Redwood City, CA, University of California Davis Comprehensive Cancer Center, Sacramento, CA, University of California San Francisco, San Francisco, CA, Division of Hematology/Oncology, University of California Davis Cancer Center, Sacramento, CA Abstract Disclosures Research Funding Pharmaceutical/Biotech Company Title: Patient-derived xenograft validation of a 3D microtumor platform Background: Patient-derived xenograft (PDX) mouse models are thought to most closely reflect the biology of a patient’s cancer. Unfortunately, growing sufficient tumor in a PDX model takes several months and more often than not, the tumor fails to grows at all. The SAGE Direct Platform, an in-vitro model, can create hundreds of live microtumors from virtually every patient’s viable biopsy and test a panel of clinically relevant drugs within no more than 1 week. Thus, concordance of results from a PDX model with results of the SAGE Direct Platform would support a rational for the platform to be potentially useful to predict tumor response in cancer patients. Methods: A bladder cancer from a 77 year old female was used to establish a PDX model. Mice were divided into three groups receiving either saline (control), cisplatin, or gemcitabine intraperitoneal on the days 1, 8, and 13, and tumor growth was observed. One tumor sample was used to create 3D microtumors and those were tested using the same drugs. Results: Tumor growth (exceeding 1,000 mm 3 ) was similar after cisplatin compared to control (4.8 vs. 3.7 weeks). After gemcitabine tumors initially shrank and only started growing a couple of weeks after the end of treatment so that 1,000 mm 3 was only reached after 10.2 weeks (p<0.001 compared to cisplatin and control). In the SAGE Direct Platform the EC 50 of cisplatin was 97.3 µM and thus two orders of magnitudes higher than the EC50 of gemcitabine, which was 0.7 µM. Conclusions: Both the PDX model and the SAGE Direct Platform have shown this bladder cancer to be virtually resistant to cisplatin while very sensitive to gemcitabine. The next steps of these preliminary data could be to repeat this experimental design with other tumors and/or to start an observational cohort study in patients correlating the SAGE Direct Platform results to patient outcomes.