Comparison of COBAS 4800 KRAS, PCR TaqMan, and PCR high-resolution melting assays for the detection of KRAS somatic mutations in formalin-fixed paraffin-embedded colorectal tumors.

Abstract

103 Background: Colorectal cancer (CRC) is the 2nd most common cancer with more than one million new cases diagnosed every year. From 2006 to 2008, several studies showed the importance of the KRAS oncogene in the treatment of metastatic CRC (mCRC) and response to anti-EGFR therapies like cetuximab or panitumumab. KRAS is a small G protein which can bear activating mutations in 40% cases of mCRC. Numerous methods have been described for the detection of KRAS mutations in FFPE tissues. Direct sequencing is the gold standard but the most sensitive methods are real time PCR assays based. The new COBAS4800 KRAS developed by Roche Diagnostics is a CE IVD validated assay using real time PCR and TaqMelt technology. This assay is validated for the detection of 19 KRAS somatic mutations on exon 2 (codons 12/13) and exon 3 (codon 61) in specimen containing at least 5% of mutated tumoral cells. Methods: We compared COBAS with previously validated PCR TaqMan and PCR High Resolution Melting (HRM) assays. The PCR TaqMan and HRM are a real time PCR based assays. TaqMan can detect accurately 0.5 to 1% of mutated tumoral DNA on codon 12 (G12A, G12C, G12D, G12R, G12S and G12V) and on codon 13 (G13D). HRM can detect 5 to 10% of tumoral DNA on codon 12 and 13. 80 FFPE samples of colorectal tumors resections or biopsies have been collected and processed with the 3 assays. Results have been compared using kappa statistics. Results: On 80 samples, 69 were interpretable with COBAS and TaqMan and 62 with COBAS and HRM. No statistically significant difference between COBAS and TaqMan was found (k=0.942 ; p<0.001). Only 2 discordant cases were found mainly regarding codon 61 mutations not detected with TaqMan. No statistically significant difference between COBAS and HRM was found (k=0.903 ; p<0.001). Three discordant cases were found two regarding codon 61, not detected with HRM and one G12S mutation (detected by COBAS and TaqMan). Conclusions: The 3 assays can detect accurately KRAS mutations. Less ininterpretable results were found with COBAS and TaqMan assays. A pre screening with COBAS and mutation characterization with TaqMan seem to be a good scheme for routine diagnostic of KRAS mutations.