Clinical, Molecular, and Functional Characterization of a Diverse Collection of Patient-derived Colorectal Cancer Organoids from the Human Cancer Models Initiative

Background

In the United States, colorectal cancer (CRC) is the third most common cause of cancer-related mortality and is second in mortality rate. There is a need for clinically representative, in vitro CRC models that reflect the genomic and phenotypic diversity of disease seen in patient populations. The Human Cancer Models Initiative (HCMI) is a collaboration between the National Cancer Institute, Cancer Research UK, Wellcome Sanger Institute, Hubrecht Organoid Technology, and ATCC that has generated hundreds of novel patient-derived cancer models that are supported with clinical and molecular annotation. These models were manufactured, characterized, and validated at ATCC and are available for academic and commercial use to support cancer research and drug discovery.

Methods

CRC organoids available in the ATCC collection of HCMI organoid models (n=70 unique models, n=67 unique donors) were subjected to whole-exome sequencing. Models were derived from primary tissues including colon, rectosigmoid junction, and rectum across 10 different anatomical sites. Data were compared to publicly available clinical and genotyping results from the tumor of origin and large cancer datasets such as The Cancer Genome Atlas (TCGA). A subset of models was tested for sensitivity to a bespoke panel of compounds (n=24) that included drugs in clinical use, chemotherapeutics, and experimental compounds with a variety of mechanisms of action and molecular targets. Models were tested with a 12-point, half log dilution dose curve and drug sensitivity was calculated as the half-maximal inhibitory concentration (IC₅₀). Post-treatment cultures were stained with live- and dead-cell fluorescent dyes and a luminescent ATP viability assay to assess response. 

Results 

CRC organoids from the HCMI were predominantly derived from primary tumors (83%) with an adenocarcinoma histological type (94%). Models had mutations in disease-relevant driver genes including APC, TP53, KRAS, PIK3CA, BRAF, ARID1A, GNAS, and SMAD4 at frequencies similar to the TCGA COAD cohort. Drug sensitivity testing revealed donor to donor variability in response to targeting compounds, which supplemented with WES data may suggest variant specific sensitivity to compounds of interest.

Conclusion

Patient-derived CRC organoids from the HCMI recapitulate genotypes seen in patient populations and respond to therapeutic treatment in vitro.