MicroRNA Expression Profiling Predicts Nodal Status and Disease Recurrence in Patients Treated with Curative Intent for Colorectal Cancer
Matthew G. Davey, Gerard Feeney, Heidi Annuk, Maxwell Paganga, Emma Holian, Aoïfe Lowery, Michael J. Kerin, Nicola Miller
Abstract
Background: Approximately one-third of colorectal cancer (CRC) patients will suffer recurrence. MiRNAs are small non-coding RNAs that play important roles in gene expression. We aimed to correlate miRNA expression with aggressive clinicopathological characteristics and survival outcomes in CRC. Methods: Tumour samples were extracted from 74 CRC patients. MiRNAs were quantified using real-time reverse transcriptase polymerase chain reaction. Descriptive statistics and Cox regression analyses were performed to correlate miRNA targets with clinicopathological and outcome data. Results: Aberrant miR-21 and miR-135b expression correlate with increased nodal stage (p = 0.039, p = 0.022). Using univariable Cox regression analyses, reduced miR-135b (β-coefficient −1.126, hazard ratio 0.324, standard error (SE) 0.4698, p = 0.017) and increased miR-195 (β-coefficient 1.442, hazard ratio 4.229, SE 0.446, p = 0.001) predicted time to disease recurrence. Survival regression trees analysis illustrated a relative cut-off of ≤0.488 for miR-195 and a relative cut-off of >−0.218 for miR-135b; both were associated with improved disease recurrence (p < 0.001, p = 0.015). Using multivariable analysis with all targets as predictors, miR-195 (β-coefficient 3.187, SE 1.419, p = 0.025) was the sole significant independent predictor of recurrence. Conclusion: MiR-195 has strong value in predicting time to recurrence in CRC patients. Additionally, miR-21 and miR-135b predict the degree nodal burden. Future studies may include these findings to personalize therapeutic and surgical decision making.