Circulating miR-21 and miR-181a as Biomarkers for Predicting Postoperative Complications Following Colorectal Cancer Resection: A Longitudinal Observational Study

<b>Background/Objectives:</b> Early and accurate detection of postoperative complications (POCs) remains a major challenge in colorectal cancer (CRC) surgery, underscoring the need for reliable molecular biomarkers. This study evaluated whether plasma miR-21 and miR-181a can predict POCs following left-sided CRC resection. <b>Methods:</b> This longitudinal observational sub-study was conducted within a randomized controlled trial. Adult patients undergoing elective left-sided CRC resection were included. Plasma miR-21 and miR-181a levels were measured preoperatively and on postoperative day (POD) 6 using RT-qPCR. POCs were assessed according to Clavien-Dindo classification. Of 40 enrolled patients, 38 were included in the final analysis (15 with and 23 without postoperative complications). Discriminative performance was assessed using receiver operating characteristic analysis and correlations with inflammatory markers were evaluated. <b>Results:</b> No significant differences in plasma miR-21 or miR-181a levels were observed between groups at baseline or POD6 (all <i>p</i> > 0.05). Both biomarkers showed limited discriminative ability (AUC = 0.61 and 0.54, respectively), while a combined model of miR-181a + TNF-α improved performance (AUC = 0.76, 95% CI: 0.57, 0.94, <i>p</i> = 0.01). At baseline, miR-21 correlated strongly with miR-181a (ρ = 0.81, <i>p</i> < 0.001) and moderately inversely with TNF-α (ρ = -0.35, <i>p</i> = 0.043). <b>Conclusions:</b> MiR-21 or miR-181a measured at baseline or POD6 show limited predictive value for POCs after CRC surgery. Further studies would benefit from larger sample sizes and optimized sampling strategies that reflect possible early dynamic changes in these biomarkers.