Abstract 5439: The chronological analysis of previously-reported serum microRNA expression in colorectal cancer

Background: Colorectal cancer (CRC) is a major cause of cancer-associated death in the world. Despite surgical resection and pre- and postoperative chemotherapy, recurrence is common in patients with CRC. Many studies have examined potential biomarker of recurrence in CRC in attempt to improve patients’ prognosis. MicroRNAs (miRNAs), the small noncoding RNAs that are associated with the development of cancer, in serum have been shown to be potential biomarkers in various type of cancers and several miRNAs have been reported in colorectal cancer. However, no miRNAs has been applied in a clinical setting because of small amount of each samples. In the present study, we evaluated the clinical effectiveness of previously-reported serum miRNAs expression in CRC using periodically gathered serum. Materials and Methods: We obtained 328 serum samples, which gathered periodically (preoperative, postoperative one month, three months, six months, one year, and two years), from 71 patients with stageII/III CRC (stageII: 31 patients, stageIII: 40 patients). Thirteen patients had a recurrence. The miRNAs were extracted from serums, and miRNA microarray analysis was performed. Previously-reported sixteen miRNAs (let-7a, miR-15b, miR-18a, miR-19a, miR-21, miR-23a, miR-29a, miR-31, miR-92a, miR-150, miR-181b, miR-203, miR-223, miR-335, miR-1229 and miR-1246) were quantified by microRNA microarray analysis. Results: In no recurrence cases, preoperative miR-1246 expression is higher than postoperative miR-1246 expression; preoperative; 5.6 ± 1.8, postoperative one month: 4.1 ± 1.6 (p < 0.05), postoperative three months: 3.4 ± 1.9 (p < 0.05) , postoperative six months: 4.8 ± 1.5 (p < 0.05) , postoperative one year: 4.0 ± 1.6 (p < 0.05) , postoperative two years: 4.8 ± 1.5 (p < 0.05). We observed trend of miR-1246 re-elevation before recurrence. The sensitivity and specificity of miR-1246 for detection of CRC are 52.8% and 84.9% (AUC=0.746). This result is almost as same as that of serum CEA which is major tumor marker of CRC (sensitivity: 22-48%, specificity: 63-100%). In our data, the sensitivity of CEA is 38% and the specificity is 94% (cutoff: > 5 ng/ml). The sensitivity and specificity of miR-1246 is higher than of CEA. Conclusion: Serum miR-1246 are best promising biomarker for CRC because of high sensitivity and specificity. On the other hand, previous reported miRNAs excluding miR-1246 are inadequate as biomarker of CRC in our data. Next we are going to identify novel