Organoids are 3D in vitro cell cultures that represent simplified versions of an organ that capture the key architecture, functionality, and genetic features of original tissues. Patient-derived cancer organoids are one variation that can be derived from primary patient tumors and studied in the laboratory. These patient-derived organoids (PDOs) have served as preclinical models of disease, but their ability to accurately predict patient drug responses in the clinical has been unclear.
A recent study reported the generation of an organoid biobank from patients with metastatic gastrointestinal cancer with 110 biopsies collected from 71 patients enrolled in clinical trials. PDOs were generated from 70% of biopsies. By creating organoids from patients with clinical trail history, the authors were able to compare organoid drug responses to patient clinical responses.
The organoids had similar phenotypic and genotypic profiles to those of the patient tumor. Histology confirmed similar morphologies between the PDOs and the original patient biopsies. Immunohistochemistry and in situ hybridization of the organoids also matched expression patterns of key biomarkers from the parental tumors. Next-generation sequencing revealed remarkably similar genome profiles with 96% mutational overlap between PDOs and biopsies.
Encouraged by these results, the authors explored the feasibility of using the PDOs as a drug-screening model by examining 21 comparisons of patient clinical responses to ex vivo organoid responses. Incredibly, the authors reported 100% sensitivity, 93% specificity, 88% positive predictive value, and 100% negative predictive value in forecasting response to targeted agents or chemotherapy in patients. Altogether, this paper demonstrates that PDOs can be harnessed for functional genomics to aid clinical decision-making, reinforcing their value as a platform for drug screening and development.
The article titled, “Patient-derived organoids model treatment response of metastatic gastrointestinal cancers” was published in Science