Search

Colorectal cancer (CRC) is the second leading cause of cancer death in the United States. The RAS pathway is activated in more than 55% of CRC and has been targeted for therapeutic intervention with MEK inhibitors. Unfortunately, many patients have de novo resistance, or can develop resistance to this new class of drugs. We have hypothesized that much of this resistance may pass through SRC as a common signal transduction node, and that inhibition of SRC may suppress MEK inhibition resistance mechanisms. CRC tumors of the Consensus Molecular Subtype (CMS) 4, enriched in stem cells, are difficult to successfully treat and have been suggested to evade traditional chemotherapy agents through resistance mechanisms. Here, we evaluate targeting two pathways simultaneously to produce an effective treatment by overcoming resistance. We show that combining Trametinib (MEKi) with Dasatinib (SRCi) provides enhanced cell death in 8 of the 16 tested CRC cell lines compared to treatment with either agent alone. To be able to select sensitive cells, we simultaneously evaluated a validated 18-gene RAS pathway activation signature score along with a 13-gene MEKi resistance signature score, which we hypothesize predict tumor sensitivity to this dual targeted therapy. We found the cell lines that were sensitive to the dual treatment were predominantly CMS4 and had both a high 18-gene and a high 13-gene score, suggesting these cell lines had potential for de novo MEKi sensitivity but were subject to the rapid development of MEKi resistance. The 13-gene score is highly correlated to a score for SRC activation, suggesting resistance is dependent on SRC. Our data show that gene expression signature scores for RAS pathway activation and for MEKi resistance may be useful in determining which CRC tumors will respond to the novel drug combination of MEKi and SRCi.

Simple Summary Colorectal cancer (CRC) is a major cause of cancer deaths. Cetuximab is an FDA-approved, underutilized therapeutic targeting the epidermal growth factor receptor (EGFR) in metastatic CRC. To date, despite selection of patients with wild-type RAS, it is still difficult to identify patients who may benefit from EGFR inhibitor (e.g., cetuximab) therapy. Our aim is to molecularly classify CRC patients to better identify subpopulations sensitive to EGFR targeted therapy. APC and TP53 are two major tumor suppressor genes in CRC whose mutations contribute to tumor initiation and progression and may identify cetuximab-sensitive tumors. Recently, it has been suggested that the consensus molecular subtype (CMS) classification may be used to help identify cetuximab-sensitive patients. Here, we report an analysis of multiple CRC tumor/PDX/cell line datasets using combined APC and TP53 mutations to refine the CMS classification to better predict responses to cetuximab to improve patient outcomes. Abstract Recently, it was suggested that consensus molecular subtyping (CMS) may aide in predicting response to EGFR inhibitor (cetuximab) therapies. We recently identified that APC and TP53 as two tumor suppressor genes, when mutated, may enhance cetuximab sensitivity and may represent easily measured biomarkers in tumors or blood. Our study aimed to use APC and TP53 mutations (AP) to refine the CMS classification to better predict responses to cetuximab. In total, 433 CRC tumors were classified into CMS1-4 subtypes. The cetuximab sensitivity (CTX-S) signature scores of AP vs. non-AP tumors were determined across each of the CMS classes. Tumors harboring combined AP mutations were predominantly enriched in the CMS2 class, and to a lesser degree, in the CMS4 class. On the other hand, AP mutated CRCs had significantly higher CTX-S scores compared to non-AP CRCs across all CMS classes. Similar results were also obtained in independent TCGA tumor collections (n = 531) and in PDMR PDX/PDO/PDC models (n = 477). In addition, the in vitro cetuximab growth inhibition was preferentially associated with the CMS2 cell lines harboring A/P genotypes. In conclusion, the AP mutation signature represents a convenient biomarker that refines the CMS classification to identify CRC subpopulations predicted to be sensitive to EGFR targeted therapies.