Identifying miRNA Biomarkers and Predicted Targets Associated with Venous Thromboembolism in Colorectal Cancer Patients

Abstract

Introduction Venous thromboembolism (VTE) is the second leading cause of mortality among cancer patients and is potentially preventable through the use of anticoagulation therapy. Risk models exist to guide use of prophylactic therapy but have low positive predictive value. New biomarkers are needed, particularly in intermediate-risk cancer types. Extracellular small RNAs, such as miRNAs, are promising biomarkers that have been implicated in tumor-dependent modification of platelets, a key component of thrombus formation. In this pilot study, we explore miRNAs as potential biomarkers for VTE risk in patients with colorectal cancer. Methods We conducted a case-control study utilizing specimens from a population enrolled in a prospective colorectal cancer biorepository at the Cleveland Clinic. Cases were defined as patients who developed VTE, including deep vein thrombosis and pulmonary embolism, within 6 months after cancer diagnosis and had their blood drawn prior to VTE. Cases were matched to controls (who did not have VTE and had a minimum of 6 months survival after cancer diagnosis) on a 1:2 ratio based on age, sex, cancer stage at diagnosis, and cancer treatment received prior to blood collection for a total of 21 patients. Total RNA from plasma specimens were sequenced on a Ion Proton platform (Thermo Fisher). Sequencing data were analyzed using the limma-voom R package. As this study was meant to be exploratory, miRNA were determined to be differentially expressed at a corrected Benjamini-Hochberg false detection rate (FDR) < 0.2. Target genes of differentially expressed miRNAs were predicted using mirDB and target gene pathways constructed with PANTHER. Results The study population had a median age of 65 (IQR 51-72). Of these patients, 85.7% were male; 42.9% had stage I/II cancer, 42.9% stage III cancer, 14.3% stage IV cancer; 71.4% had received no treatment prior to blood collection, 14.3% received chemo/chemoradiation therapy, and 14.3% received surgery. A total of 2426 unique miRNAs (median 1524, IQR 1371-1665) were expressed in the study group. Of these, 9 miRNAs were significantly differentially expressed (FDR < 0.2) and downregulated in cases compared to controls: hsa-miR-4451, 942-3p, 8063, 3132, 3118, 105-5p, 891a-5p, 200a-5p, and 6832-3p. From these miRNAs, 609 target genes were predicted and classified into 75 pathways, including angiogenesis, G-protein coupled receptors (GPCRs), inflammation mediated by chemo/cytokines, and integrin signaling. Target genes within these notable pathways included EPHA3, PDGFA, PTK2/FAK1, and IL15. Conclusions We identified 9 significantly downregulated miRNAs in the blood of colorectal cancer patients who developed VTE compared to controls in this pilot study. These data suggest that colorectal cancer patients may express unique miRNA profiles prior to VTE development which may be useful as biomarkers in future predictive models. In addition, this study identified potential new mechanistic targets for understanding cancer-associated thrombosis. While the role of GPCRs, integrins, and inflammation in platelet activation and function is widely known, this study also identified factors within the angiogenesis pathway that have been linked to increased platelet aggregation and tissue factor activation. Thus, downregulation of inhibitory miRNAs may cause disinhibition of pathways important for platelet and vascular function and other prothrombotic factors. Disclosures Khorana: Janssen: Consultancy; Bayer: Consultancy; Pfizer: Consultancy; Sanofi: Consultancy. McCrae:Pfizer Pharmaceutical: Membership on an entity's Board of Directors or advisory committees; Dova Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees; Rigel Pharmaceutical: Membership on an entity's Board of Directors or advisory committees; Sanofi Pharmaceuticals: Membership on an entity's Board of Directors or advisory committees.