Predictive microRNAs for lymph node metastasis in endoscopically resectable submucosal colorectal cancer

This is the authors' abstract. We don't add key points for this paper.

// Chan Kwon Jung 1, * , Seung-Hyun Jung 2, 3, 6, * , Seon-Hee Yim 3 , Ji-Han Jung 1 , Hyun Joo Choi 1 , Won-Kyung Kang 4 , Sung-Won Park 3, 6 , Seong-Taek Oh 4 , Jun-Gi Kim 4 , Sug Hyung Lee 5, 6 , Yeun-Jun Chung 2, 3 1 Department of Hospital Pathology, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea 2 Department of Microbiology, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea 3 Integrated Research Center for Genome Polymorphism, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea 4 Department of Surgery, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea 5 Department of Pathology, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea 6 Cancer Evolution Research Center, College of Medicine, The Catholic University of Korea, Seoul 06591, Republic of Korea * These authors have contributed equally to this work Correspondence to: Yeun-Jun Chung, email: yejun@catholic.ac.kr Keywords: endoscopically resectable colorectal cancer, microRNA, lymph node metastasis Received: September 03, 2015      Accepted: March 28, 2016      Published: April 16, 2016 ABSTRACT Accurate prediction of regional lymph node metastasis (LNM) in endoscopically resected T1-stage colorectal cancers (CRCs) can reduce unnecessary surgeries. To identify miRNA markers that can predict LNM in T1-stage CRCs, the study was conducted in two phases; (I) miRNA classifier construction by miRNA-array and quantitative reverse transcription PCR (qRT-PCR) using 36 T1-stage CRC samples; (II) miRNA classifier validation in an independent set of 20 T1-stage CRC samples. The expression of potential downstream target genes of miRNAs was assessed by immunohistochemistry. In the discovery analysis by miRNA microarray, expression of 66 miRNAs were significantly different between LNM-positive and negative CRCs. After qRT-PCR validation, 11 miRNAs were consistently significant in the combined classifier construction set. Among them, miR-342-3p was the most significant one ( P =4.3×10 -4 ). Through logistic regression analysis, we developed a three-miRNA classifier (miR-342-3p, miR-361-3p, and miR-3621) for predicting LNM in T1-stage CRCs, yielding the area under the curve of 0.947 (94% sensitivity, 85% specificity and 89% accuracy). The discriminative ability of this system was consistently reliable in the independent validation set (83% sensitivity, 64% specificity and 70% of accuracy). Of the potential downstream targets of the three-miRNAs, expressions of E2F1, RAP2B, and AKT1 were significantly associated with LNM. In conclusion, this classifier can predict LNM more accurately than conventional pathologic criteria and our study results may be helpful to avoid unnecessary bowel surgery after endoscopic resection in early CRC.