MicroRNAs and response to anti-EGFR therapy in metastatic colorectal cancer.

Abstract

e21002 Background: The discovery of KRAS gene as response predictive factor to anti-EGFR antibodies supposed a revolution in the pharmacological management of metastatic colorectal cancer (mCRC) patients (pts), since only KRAS native (WT) pts are possible responders (R) and susceptible to be treated with this targeted therapy. However, the appearance of non-responders (NR) among KRAS WT pts during the administration of anti-EGFR therapy has stimulated the research in additional predictive markers. Here, we explore the putative role of microRNA (miRNA) deregulation as predictive tool of response to anti-EGFR therapy in mCRC pts. Methods: Tissue samples of mCRC were obtained as formalin-fixed paraffin-embedded (FFPE) specimens from the archives of our institution, after obtaining informed consent of pts. Genomic DNA was extracted using Qiaamp DNA mini kit (Qiagen). For the mutational testing of KRAS gene, we use the TheraScreen KRAS PCR kit (DxS, Qiagen). miRNA-containing total RNA was extracted using Tripure (Roche). miRNA expression was analyzed by SYBR-green-based RT-qPCR in a LightCycler 480 (Roche) device using the Human Cancer RT2 miRNA PCR array (SABiosciences, Qiagen). First data analysis was performed with the application provided by supplier (SABiosciences, Qiagen). Bioinformatic analyses were performed in miRBase 18, TargetScanHuman 6.0 and TarBase 6.0 databases. Additional statistical analyses were performed on R 2.14 (R Foundation), SPSS 19 (IBM) and Prism (GraphPad). Results: We analyze 9 primary colon cancer and 3 non-matched normal adjacent (NAT) tissues. Tumor samples corresponded to 6 KRAS WT and 3 KRAS mutated pts. Among KRAS WT pts, 3 had shown response to anti-EGFR antibodies and 3 did not. Samples were analyzed in duplicate for the expression of 88 cancer-related miRNAs. We found 22 miRNAs significantly (p<0.01) upregulated (>4-fold) when comparing KRAS WT NR versus R pts. From these 22 miRNAs, we detected only 6 when we search in databases for experimentally tested or predicted miRNA interactions with KRAS. Conclusions: We found a miRNA signature which could be utilized as response predictive tool for anti-EGFR therapy in mCRC pts. Further analysis will be necessary to establish the performance of this signature.