Optimization of routine KRAS mutation PCR-based testing procedure for rational individualized first-line-targeted therapy selection in metastatic colorectal cancer

Abstract

KRAS mutation detection represents a crucial issue in metastatic colorectal cancer (mCRC). The optimization of KRAS mutation detection delay enabling rational prescription of first-line treatment in mCRC including anti-EGFR-targeted therapy requires robust and rapid molecular biology techniques. Routine analysis of mutations in codons 12 and 13 on 674 paraffin-embedded tissue specimens of mCRC has been performed for KRAS mutations detection using three molecular biology techniques, that is, high-resolution melting (HRM), polymerase chain reaction restriction fragment length polymorphism (PCR-RFLP), and allelic discrimination PCR (TaqMan PCR). Discordant cases were assessed with COBAS 4800 KRAS CE-IVD assay. Among the 674 tumor specimens, 1.5% (10/674) had excessive DNA degradation and could not be analyzed. KRAS mutations were detected in 38.0% (256/674) of the analysable specimens (82.4% in codon 12 and 17.6% in codon 13). Among 613 specimens in whom all three techniques were used, 12 (2.0%) cases of discordance between the three techniques were observed. 83.3% (10/12) of the discordances were due to PCR-RFLP as confirmed by COBAS 4800 retrospective analysis. The three techniques were statistically comparable (κ > 0.9; P < 0.001). From these results, optimization of the routine procedure consisted of proceeding to systematic KRAS detection using HRM and TaqMan and PCR-RFLP in case of discordance and allowed significant decrease in delays. The results showed an excellent correlation between the three techniques. Using HRM and TaqMan warrants high-quality and rapid-routine KRAS mutation detection in paraffin-embedded tumor specimens. The new procedure allowed a significant decrease in delays for reporting results, enabling rational prescription of first-line-targeted therapy in mCRC.

Keywords: Colorectal cancer; HRM; KRAS; PCR-RFLP; TaqMan PCR.

Figure 1

Macrodissection step to ensure a minimum of 20% tumor tissue content. The hematoxylin–eosin slide with selected area contains more than 20% tumor cells.

Figure 2

(A) KRAS mutation analysis using PCR-RFLP. DNA extracts from tumor samples were submitted to double PCR amplification after BstXI and XcmI enzymatic digestion allowing discrimination of codons 12 and 13 mutations. Codons 12 and 13 mutated DNA were used as positive control. Wild-type KRAS DNA and water were used as negative controls. (B) Example of codon 12 KRAS mutation detection using TaqMan PCR. Left panel represents amplification control (VIC). Right panel represents G12S mutation detection (FAM). (C) Example of codon 12 KRAS mutation detection using HRM. Depending on the presence or the absence of mutation, the melting temperature is different (left panel). The plot of the relative signal difference against the temperature allows to evidence the presence or the absence of KRAS mutation (right panel).

Figure 3

Comparison of mutation frequency as determined by the different detection methods. Overall as well as for each technique, all mutations frequencies were compared with average data from the Sanger Cosmic data base and were found to be fully consistent with the theoretical frequencies (chi-square test nonsignificant for all the data).