Patient-derived tumor organoids (PDTOs) have been shown to be predictors of clinical response. While evidence suggests PDTOs are a more physiologically relevant cell model to use for drug discovery, challenges still exist to fully characterize their response to chemotherapies. Assays which attempt to quantify drug response in preclinical models are often end-point and cannot separate responses of heterogeneous cultures or capture dynamic information. Image based technologies attempt to counteract these shortcomings but often require fluorescent protein expression or cell staining to observe changes long-term. We utilized a label-free approach, fluorescence lifetime imaging microscopy (FLIM) and phasor analysis, to determine the metabolic state of individual PDTOs and correlate our findings to traditional cell viability assays (e.g., ATP measurements, phototoxic dyes) in response to drug perturbations.

By measuring the fluorescence lifetime of NADH, an autofluorescent metabolite, we were able to quantify the metabolic signature of KRAS mutant (KRASMUT) colorectal cancer (CRC) PDTOs. The emission decay at each pixel was Fourier transformed and represented on a phasor plot. Pixels that had a shorter lifetime of NADH were indicative of glycolysis (GLY) while longer lifetimes indicated regions of oxidative phosphorylation (OXPHOS). PDTOs were treated with clinically-relevant drugs, SN-38 (an active metabolite of irinotecan), 5-FU, and cetuximab (CTX). FLIM images were taken after 6 and 72 hours of treatment. KRASMUT tumors do not respond to CTX, thus this was used as a negative control. Staurosporine (ST) was used as a positive control for cell death. A shift in metabolic signature towards OXPHOS in a dose-dependent manner was seen in PDTOs treated with ST or SN-38. These shifts correlated with cell death measured using the CellTiter Glo cell viability assay and confocal fluorescence images of PDTOs stained with a vital dye, DRAQ7. Changes in metabolism were seen within hours while DRAQ7 signal was detected after a few days. No changes in metabolic signature, cell viability, or DRAQ7 levels were seen in CTX treated PDTOs as expected. 5-FU was found to be an ineffective drug for the specific PDTO tested in our assay as shown by the lack of metabolic shifts compared to untreated PDTOs and no decrease in cell viability. Metabolic heterogeneity was also seen with FLIM, including distinguishing different cell types in co-cultures of PDTOs and cancer-associated fibroblasts (CAFs). We were able to separate the metabolic signatures of each cell type and determine that CAFs have a more GLY signature. Taken together, FLIM supported label-free imaging of multicellular 3D organoid models for the purposes of determining dynamic tumor drug response. We are adapting this workflow to a high-throughput multimodal imaging platform for drug screening to make more informed decisions on patient care.

Citation Format: Emma Jane Fong, Mathias Bigger, Seungil Kim, Michael Doche, Scott Valena, Pratiksha Kshetri, Shannon Mumenthaler. Fluorescence lifetime imaging microscopy to assess drug response in patient-derived colorectal cancer organoids [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2383.