Background: Standard two-dimensional cell cultures do not retain the key characteristics of the human cancers from which they are derived and treatment effects are not always able to be replicated in vivo, making the development of alternative culturing systems paramount. Specifically, commercially available cell lines do not fully represent the mutation profiles seen in human cancers. Here we investigate the feasibility of three-dimensional PDOCS to more accurately represent the cancers from which they are derived and to predict treatment sensitivity in a clinically meaningful time frame.

Methods: Surgical resection, core needle biopsies, paracentesis or thoracentesis samples from patients with various types of cancer were obtained under an approved IRB protocol, digested and spheroid cultures grown suspended in Matrigel. PDOCS were grown for up to two weeks and passaged at least once prior to treatment. PDOCS were imaged using brightfield imaging (4X) prior to treatment with vehicle or 5-fluorouracil (5-FU; 1, 10, or 100 µM) and/or radiation (2 or 5 Gy). After 2 days of treatment, the 5-FU was removed and the cultures were allowed to grow for an additional 2 days. PDOCS were re-imaged and the relative change in diameter was calculated using ImageJ software and compared to untreated controls. Optical metabolic imaging (OMI) was performed with a multiphoton microscope to probe the fluorescence lifetime and optical redox ratio of metabolic co-enzymes NAD(P)H and FAD. Single-cell analysis of each image was completed using Cell Profiler software to parse resistant cell populations in each PDOCS sample tested. DNA for mutation profile analysis was isolated and sequenced using a QIAseq targeted panel.

Results: PDOCS were successfully isolated from a variety of cancers including colorectal (CRC), pancreas, lung, neuroendocrine, liver, breast, and ovarian. Key phenotypic characteristics of the tumors were retained in PDOCS cultured including crypt-like structures, mucin production and Ki67 proliferation rates. Cancer hot spot sequencing was performed comparing PDOCS and the tumors from which they were derived. Over 90% of the nonsynonymous mutations were identical, except in the setting of microsatellite instability. All driver mutations were identical (i.e., APC, KRAS, PIK3CA, TP53). Differential sensitivity to chemoradiation was observed among 4 different colorectal PDOCS treated with 5-FU and radiation (Median % PDOCS diameter change vs control: Patient A 17.1, p=0.64; Patient B -3.05, p=0.02; Patient C -19.4, p=2x10-5; Patient D -31.3, p=0.002). Similar response data were determined using OMI; however, single-cell analyses identified potentially resistant cell populations.

Conclusions: PDOCS retain key characteristics of the cancers from which they are derived and can be utilized for treatment sensitivity testing in a clinically meaningful time frame.

Citation Format: Cheri A. Pasch, Peter F. Favreau, Alex E. Yueh, Kwang P. Nickel, Christopher P. Babiarz, Philip B. Emmerich, Rosabella T. Pitera, Susan N. Payne, Demetra P. Korkos, Joseph T. Sharick, Carley M. Sprackling, Linda Clipson, Kristina A. Matkowskyj, Michael A. Newton, Melissa C. Skala, Michael F. Bassetti, Randall J. Kimple, Dustin A. Deming. Patient-derived organotypic cancer spheroids (PDOCS) as predictive models for the treatment of cancer in a clinically meaningful time frame [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 5011.