Abstract Background: There are limited clinical tools for predicting the effectiveness of cancer therapies. We aim to prospectively predict patient treatment response using patient-derived organotypic cancer spheroids (PDOCS) as an in vitro model which recapitulates the genetic characteristics and 3D organization of the patient’s tumor. Using optical metabolic imaging (OMI) to analyze single cells, we can determine heterogeneous subpopulations in response to drug treatment. Further clinical validation of these techniques and analysis methods are needed before clinical translation. Methods: Tissue biopsies and gross tissue resections were acquired through the University of Wisconsin Precision Medicine Molecular Tumor Board (IRB#UW15068) and UWCCC TSB Biobank. Next-generation sequencing (NGS) from the biopsies was performed to determine molecular profiling. In alignment with the patient’s treatment course, PDOCS were treated with physiologic doses of chemotherapy or targeted therapy. Treatment response was evaluated by measuring change in diameter in conjunction with optical metabolic imaging (OMI) using a multiphoton microscope to measure the fluorescence and redox ratio of NAD(P)H and FAD as an indication of cellular metabolism. Diameter changes between control and treatment groups were compared using Glass’s delta; resistance to therapy was indicated by a Glass’s delta score of below 1.5. The optical redox ratios determined by OMI were compared using Glass’s delta, and resistance was indicated below 0.5. Clinical response was measured using RECIST v1.1 standard response assessment criteria. Results: PDOCS were successfully isolated from colorectal (CRC), lung, gastrointestinal stromal tumor (GIST), ovarian, and breast cancers. These biopsies were all obtained in the treatment refractory setting. PDOCS were generated for seven patients and treated with the same pharmacologic treatment as the patient from which the PDOCS were generated. Multiple treatments were able to be tested both in vitro and clinically for a subset of patients. Treatments included: 5-fluouracil, oxaliplatin, gemcitabine, paclitaxel, olaparib, panitumumab, osimertinib, fulvestrant, and palbociclib. In this cohort, two treatments resulted in stable disease and seven treatments resulted in disease progression. Change in spheroid diameter correlated with clinical treatment outcomes with an effect size (Glass’s delta) threshold of 1.5. OMI predicted response for all patients imaged with an effect size threshold of 0.5 which correlated with the size change analyses. Treatment heterogeneity of OMI was observed in many of the samples. Conclusions: In this largely prospective cohort of patients across disease types, changes in PDOCS size and OMI indices predict treatment benefit for individual patients. Studies on a larger scale are needed to further validate these findings. Citation Format: Carley M. Sprackling, Jeremy D. Kratz, Peter F. Favreau, Mohammad R. Karim, Christopher P. Babiarz, Cheri A. Pasch, Amani A. Gillette, Linda Clipson, Kristina A. Matkowskyj, Jens C. Eichoff, Kayla K. Lemmon, Hannah K. Houtler, Mark E. Burkard, Devon Miller, Melissa C. Skala, Dustin A. Deming. Predicting treatment response using patient derived organotypic cancer spheroids [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 3143.