Abstract A013: Role of EGFR in KRAS-mediated resistances in colorectal cancer. </title>

Abstract

Abstract The standard of care therapies for metastatic colorectal cancer (mCRC) patients with wild-type (WT) RAS or BRAF consists of chemotherapy combined with targeted anti-EGFR antibodies. Still, only half of the RAS WT mCRC patients are sensitive to anti-EGFR therapy1, and eventually some of them stop responding due to intrinsic and acquired resistance to anti-EGFR treatment2. Moreover, recent clinical studies have also shown that left-sided mCRC respond better to anti-EGFR treatment for reasons that are still unknown. Studies from our lab and others using genetically engineered mouse models (GEMMs) showed that EGFR expressing tumor-associated myeloid cells release inflammatory cytokines that have a pro-tumorigenic role in CRC3,4. Deletion of EGFR in tumor cells of these mice had no effect in tumor size, whereas deletion of EGFR in myeloid cells impaired CRC development3. Moreover, human CRC biopsies confirmed that presence of EGFR+ myeloid cells is associated with a worse overall survival of mCRC patients. From our results in GEMMs, it seems that the therapeutic action of EGFR inhibitors is mediated via EGFR inhibition in myeloid cells. Assuming that tumor cells are not the primary target of anti-EGFR therapies, it is puzzling that KRAS mutations in tumor cells confer resistance to anti-EGFR treatment in mCRC patients. We hypothesize that epithelial KRAS mutations alter the tumor microenvironment (TME) towards an immunosuppressive milieu. We are testing this hypothesis by comparing expression signatures derived from tumor and stroma of human mCRC biopsies that are WT or mutant for KRAS to elucidate the molecular signature driven by KRAS that explains anti-EGFR treatment resistance. In parallel, we employ CRC mouse models generated by orthotopic implantation of WT and RAS mutant mouse CRC organoids into GEMMs lacking the EGFR in different cell types to mimic right and left-sided metastatic disease. Targets identified in human mCRC patients are being mechanistically validated in organoids and derived tumors by CRISPR/Cas9-mediated genetic engineering. The results of these experiments will be presented. The expected results will provide a better understanding of how anti-EGFR antibodies work, how resistance mechanisms are developed, and possible new treatment combinations for KRAS mutated CRC.