Abstract 988: Short-term culture of organotypic tumor spheroids derived from patient xenografts in a novel 3D microfluidic chip predicts in vivo response of targeted therapies

Introduction: Patient-derived xenografts (PDX) are increasingly adopted to evaluate the efficacy of new treatment approaches, including the development of combination therapies and testing of predictive biomarkers hypotheses. However, PDXs can take months to generate and are impractical for screening large sets of drug combinations. Here we report on a novel 3D microfluidic platform that allows for evaluating ex vivo responses to targeted therapies from genotyped defined patient-derived tumor spheroids (XDOTS) (1). Methods: Non-small cell lung cancer (NSCLC) PDXs with known genotype, including EGFR ex19 del, HER-2 ins, and BRAF V600E, were generated from either core needle biopsies or pleural effusions under an IRB-approved protocol. XDOTS ( 40μm) were grown from PDX tumors using DAX-1 3D cell culture chips (AIM, Singapore) and treated with genotype matched inhibitors that are either approved (e.g., erlotinib for EGFR mut) or are being tested in the clinic (neratinib for HER-2 ins) at known peak plasma concentrations for 3 days. Live/death quantification was performed by dual labeling fluorescence microscopy using acridine orange for live and propidium iodide for dead cells. Cell type characterization was performed by immunofluorescence in the devices. Results: We first characterized our XDOT system on models for which we had orthogonal validation. We show by dual labeling fluorescence microscopy that PC9GR4 (EGFR Del19/T790M) cells are sensitive to osimertinib (IC50 = 12nM), but not gefitinib (IC50 >10uM). We further show that HCC827GR6 (Del 19/MET Amp) and DFCI32 (EML4-ALK/EGFR/HER2) cells are sensitive to crizotinib/gefitinib combo, but not to each of the single agents. Results were validated using the IncuCyte Live Cell analysis. We then went on to show that genotype-specific responses can be recapitulated in tumor spheroids generated from PDXs. In DFCI 359, a PDX derived from a HER2 mutant (755_757 LRdelinsPR), we show the antitumor efficacy of afatinib and confirm that effect is specific to tumor cells and not stromal. We observe significant drug efficacy only on cells growing in our microfluidic devices and not on spheroids growing 3D in 384 well plates. Similar results were obtained for additional PDXs with known NSCLC driver mutations. Finally, we evaluate our system on a small cohort of direct patient specimens. Conclusions: Taken together, our data show that short culture experiments using XDOTS are feasible and recapitulate key futures of in vivo response and resistance ex vivo. Reference: 1. Jenkins RW, et al. Cancer Discov DOI:10.1158/2159-8290. Citation Format: Amir Aref, Elena Ivanova, Emily Chambers, Andrew Portell, Russell Jenkins, Mika Lin, Man Xu, Jacob Haworth, Magda Bahcall, Cloud Paweletz, Pasi Janne. Short-term culture of organotypic tumor spheroids derived from patient xenografts in a novel 3D microfluidic chip predicts in vivo response of targeted therapies [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 988.