Predictive value of a CLIA-approved organoid based drug sensitivity test.

Abstract

3630 Background: Precision medicine integrates genetic, molecular, and clinical information to optimize therapy selection for cancer patients. Ex vivo drug testing has the potential to match the right drug to the right patient. We developed a CLIA-certified functional drug assay for all solid tumors which provides an actionable report of organoid sensitivity to targeted, endocrine and chemotherapy agents as a tool for therapeutic decisions. Objectives: To establish the predictive power of the test in relation to well-known genomic biomarkers as well as prior treatments to identify drug sensitivity. To demonstrate that functional drug testing increases the actionability of genomic reports. Methods: From 2016 to 2019, 240 organoids from cancer patients were subjected to functional testing at SEngine Precision Medicine. Patients with advanced primary or metastatic cancer (solid tumors) who were treatment naïve or had previous therapies fail. Fresh samples of tumor cells from core biopsies, surgical excisions, or fluids arrived <48 hrs following collection and were cultured as 3D organoids. They were evaluated using a multi-dose response format with a library of up to 130 compounds. Drug sensitivity was quantified using a score that combines sensitivity and personalization of each patient’s response relative to a reference population. Known genomic actionability from levels of evidence 1-2 from MSKCC OncoKB were queried against results for correlation. Results: Organoids were derived from breast (18.7%), ovarian (18.3%), colorectal (17.9%), pancreatic (6.7%), and others solid tumors (38,3%). Median age of patients was 53 (r5-83). 68 drugs on average were tested per patient with a mean turnaround time of 18 days (r9 -37). A mean of 7 drugs per patient were identified as top scoring drugs. In 75 patients with genomic data, we found high concordance of drug sensitivity with known genomic anchors (e.g., inhibitors of BRCA1/ PARP, ERBB2/HER2, FGFR1-2/FGFR, KRAS, PIK3CA/PI3K), measured as sensitivity to drugs among this targeted groups. However, several patient samples demonstrated sensitivity to targeted agents in the absence of known genomic biomarkers. Most important, analysis of previous treatments indicated >90% of retrospective concordance. Conclusions: Organoid based drug testing exhibits strong concordance with genomic and retrospective clinical evidence. In addition, functional testing identifies candidate therapies in patients with no known biomarkers and can identify the significance of variants currently not validated.