Novel multiple, monoallelic KRAS mutations at codon 12 and 13

Abstract

The KRAS proto-oncogene codes for a small guanosine triphosphate (GTP) binding protein that is involved in transmitting epidermal growth factor receptor (EGFR) induced signals to the cytoplasm and nucleus. Oncogenic mutations of the KRAS gene are found in 30–40% of colorectal carcinomas. Up to 90% of the mutations occur at codon 12 or 13, which affect the G-protein binding domain of the protein leading to a constitutive activation of its intrinsic GTPase activity. The identification of KRAS mutations has recently gained high attraction due to the demonstration that in several phase II and III clinical trials, only patients with metastatic colorectal cancer with wild-type KRAS in their tumors had a benefit from an antibody-based therapy against the EGFR receptor, such as cetuximab or panitumumab. According to these results, the European Medicines Agency (www.emea.europa.eu/) had restricted the use of cetuximab and panitumumab to patients with metastatic colorectal cancer with the wild-type KRAS gene in their tumors. Furthermore, the American Society of Clinical Oncology provisional clinical opinion recently stated, that KRAS mutation analysis of tumors from patients who are candidates for anti-EGFR antibody treatment should be performed and only patients with wild type KRAS should receive anti-EGFR antibody therapy. Different KRAS mutation detection methods are currently used, and until now there is no standardized assay. Direct sequencing allows the detection of all mutations and was considered as the gold standard for mutation detection. A number of additional techniques are available, which have been developed based on the fact that the number of KRAS mutations is discrete. In a series of 111 consecutive specimens, consisting of 75 primary colorectal carcinomas and 36 metastases analyzed in a diagnostic setting for KRAS mutation in the year 2008 we identified mutations in 38 (34%) tumors by direct sequencing of exon 2, encompassing the most frequently altered codons 12 and 13 of the KRAS gene. Twenty-three (30%) mutations were found in primary tumors and 15 (42%) in metastases. In one primary tumor and in 3 metastases (1 in the lung, 1 in the liver and 1 in a local recurrence), we identified double mutations. Two double mutations were found at codon 12 and 2 double mutations at codon 13, each present in the heterozygote state. To determine whether these double mutations were located on 1 allele leading to unusual amino acid exchanges or in both different alleles, we subcloned the PCR products of the 4 samples. Sequencing analysis of 5–10 subclones revealed the wild type sequence in each sample and demonstrated that all 4 double mutations affect 1 codon on 1 allele. In addition in 1 subclone deriving from the lung metastasis with a double mutation at codon 13, a monoallelic triple mutation including mutation of the last wild type nucleotide at this codon was identified. Mutations are summarized in Table I. In both metastases where the corresponding primaries were available, we detected the identical double mutations in the primary tumor tissues. The occurrence of more than 1 mutation at codon 12 or 13 has occasionally been reported in the literature, but the monoor biallelic nature of these mutations has rarely been specified. Mutation data of the KRAS gene obtained from the Sanger Institute Catalogue of Somatic Mutation in Cancer (http://www.sanger.ac.uk/cosmic) include the G13E substitution, which we also identified in 1 patient, though the zygosity is indicated as unknown. The other 4 amino acid substitutions at codon 12 and 13, G12I, G12W, G13K and G13F, have not been reported in this database, which currently comprises more than 12,200 KRAS mutations from various tumors and thus have to be considered as novel oncogenic mutations. We believe that our results are important under aspects related to tumor biology as well as to assay considerations for KRAS testing in the diagnostic setting.