A very low incidence of BRAF mutations in Middle Eastern colorectal carcinoma

A complex network of genes are required for maintaining the cellular homeostasis in colorectal tissue. In the original multistep progression model of colorectal cancer (CRC) first proposed by Vogelstein et al, normal colonic epithelium gets transformed into benign (adenoma) neoplastic epithelium followed by full blown invasive cancer and eventually metastasis[1]. Over the years, significant advances in molecular genetics and caner biology have led to the understanding that there exist 3 distinct molecular phenotypes of CRC: (1) chromosomal instability(CIN)[2]; (2) microsatellite instability(MSI)[3]; and (3) the recently discovered sessile and traditional serrated adenomas, a subset of hyperplasic polyps that progress to serrated carcinomas[4].

Some of the key signaling pathways implicated in colorectal carcinogenesis are Epidermal Growth Factor Receptor Signaling (EGFR) pathway, WNT signaling[1] and transforming growth factor-β (TGF β) signaling[5]. Translational research has resulted in the bench-to-bedside application of biomarkers like KRAS, BRAF and PI3K for personalized medicine[6]. One of the key determinants of response to panitumumab and cetuximab therapy in colorectal cancer is the presence of mutations in the KRAS gene[7, 8]. Testing for certain “activating” mutations in the KRAS gene is currently one of the most widely employed methods to predict responsiveness to EGFR inhibitors. Patients with presence of KRAS mutations in their tumors do not respond to EGFR inhibitors. On the other hand, a significant proportion of CRC patients with the wild-type (normal) KRAS gene fail to respond to EGFR inhibitors and mutations in other genes such as PIK3CA/BRAF/NRAS/PTEN/TP53 have been implicated for resistance in this subgroup of patients[7‐9]. BRAF mutations are known to be mutually exclusive with KRAS mutations and Vaughn et al have reported that almost 8% of the CRC subgroup with wild type KRAS gene had BRAF mutations; these patients would receive EGFR inhibitors but would be unresponsive to therapy[8].

Constitutive activation of the mitogen-activated protein kinase (MAPK) pathway regulates cellular proliferation, differentiation, and death. Oncogenic mutations in the v-Raf murine sarcoma viral oncogene homolog B1 (BRAF) in CRC were first described by Davies et. al and the commonest mutation was a single phosphomimetic substitution in the kinase activation domain (V600E) that led to activation of the MAPK pathway[10] Although BRAF mutations have been identified in a variety of cancers, their highest incidence to the tune of almost 60% is seen in melanomas[10]. The CRC subset that originates from advanced serrated polyps have a distinct molecular phenotype characterized by widespread hypermethylation of CpG islands in the promoter regions of genes, referred to as the CpG island methylator phenotype (CIMP)[11], microsatellite instability (MSI) and oncogenic mutations in BRAF gene[4, 12, 13]. Testing for BRAF mutation is beneficial in judicious selection of patients for targeted therapy[8, 9] and also a cost effective approach in the HNPCC (Hereditary Non-polyposis Colorectal Cancer) workup; presence of BRAF mutation with absence of MLH1 protein is indicative of sporadic CRC[14, 15].

Accumulating evidence implicates BRAF mutations to have diagnostic, therapeutic and prognostic significance in CRC[10, 19‐22]. In our study, the BRAF-V600E mutation was identified in only 2.5% of all CRC, which is significantly lower than earlier reports of BRAF mutation in CRC worldwide (5-15%;[22‐26]). The overall low frequency of the BRAF-V600E mutation and serrated adenomas in this unique ethnic group of Saudi colorectal cancer cases suggests a very small role of the BRAF gene in the development of CRC and raises some interesting future questions on the burden of familial colorectal cancers like Hereditary non polyposis colorectal cancer cancer(HNPCC) syndrome in Saudi Arabia.

The reported frequency of BRAF mutations in different populations varies widely from a low incidence of 1% in Taiwan to 19.8% and 21.8% in Netherlands and USA respectively[22‐25, 27‐40]. A lower frequency has been observed in most of the Asian population as is evident from the lowest BRAF mutation incidence of 1% in Taiwan where BRAF mutation was seen in 2/182 CRC. A similar low incidence was seen in China (3.8% and 7%) and Japan (4.7% and 6.7%). BRAF mutation was also observed at a lower rate of 5% in Israel[39]. Interestingly the frequency of the BRAF V600E mutation varies widely between population groups, both within single studies (5.8% in the Ashkenazi versus 3.2% in the non-Ashkenazi) and also between reports from different groups but within the same region/country[36, 41]. The lower incidence of BRAF mutations in Saudi CRC patients could be due to a different ethnic populations with varied underlying genetic predisposition to BRAF-mutated tumors, role of environmental influences like diet, smoking and other unknown factors. Different methods used to detect gene mutation such as 454 next generation sequencing, Sanger sequencing, pyrosequencing and melting curve analysis also influence the detection rate of a gene mutation[42]. An earlier study has highlighted the differing rates of BRAF mutation in distinct ancestral populations with a lower frequency observed in Asian population as compared to White and Black patients in their study[43].

Although BRAF mutations occur early in colorectal carcinogenesis and have been observed in colorectal adenomas, tumor progression needs additional acquired DNA microsatellite instability caused by hypermethylation of MLH1 in the setting of the CpG island methylator phenotype (CIMP)[12, 44]. These precursor lesions that harbor BRAF mutation, microsatellite instability and CIMP phenotype have a distinct morphology and are termed serrated adenomas. A retrospective review of all the available biopsies of hyperplastic polyps and adenomas biopsies confirmed the presence of only one case that had a morphology consistent with serrated adenomas. Additional efforts to identify serrated adenomas from the FFPE archives of Department of Pathology confirmed the fact that serrated adenomas were exceedingly rare. We have analyzed CIMP in 500 CRC patients and have observed a very low frequency of CIMP of 4.8% (data not shown). In concordance with an earlier study we hypothesize that lack of a propensity for hypermethylation and other factors in Saudi CRC prevents the driver BRAF mutation in premalignant lesions to progress to full blown cancers[13]. Although smoking is one of the factors implicated in CIMP phenotype and the incidence of smoking in Saudi Arabia is quite high[45, 46], the prevalence of BRAF mutations, CIMP and serrated adenomas in our study population was on the lower side. Moreover, the significantly lower incidence of BRAF mutations in our study suggests the existence of a possibly higher number of familial syndromes like HNPCC that are characterized by a virtual lack of BRAF abnormalities in colorectal cancers. A study is being conducted to determine the true incidence of HNPCC in Saudi Arabia. Presence of BRAF V600E mutation in CRC tumors that lacks the expression of MLH1 by IHC does not warrant further genetic testing and excludes the possibility of the presence of a germline mutation in mismatch repair (MMR) genes. This exquisite sensitivity of BRAF V600E somatic mutation for sporadic MSI has led to the development of an algorithm for screening of LS as a multistep diagnostic approach in several guidelines and recommendations[15, 47]. Since a significant proportion of sporadic CRC get excluded, the use of BRAF V600E as a screening tool to identify sporadic MSI CRC tumors is highly cost effective[15, 47, 48]. Newer developments in the detection of BRAF V600E mutation in CRC by IHC have shown to have a comparable sensitivity and specificity as PCR testing[49]. After stringent validation of BRAF IHC in each lab, BRAF testing by IHC will prove to be a simple, economical, labor and time saving test that can be performed even in small biopsies that yield a small amount of DNA. Finally Weisenberger et al investigated CIMP in CRC and demonstrated that MSI-H CRC are either HNPCC or MSI-H and CIMP + with or without MLH1 methylation[50]. The authors concluded that CIMP + encompass almost all sporadic MSI-H CRC while MLH1 methylation constitute a part of, but not all CIMP + CRC tumors. Since BRAF mutations distinctly correlate with CIMP + CRC, BRAF testing could outperform MLH1 methylation in identifying sporadic MSI-H CRC tumors in the diagnostic approach of LS[50].

Emerging studies highlight the critical significance of the impact of BRAF mutations on colorectal research. Although there has been remarkable success in using BRAF inhibitors in melanomas with a response rate of over 80%, a lower response rate of about 10% is seen in CRCs[20]. A key mechanism causing resistance to BRAF inhibitors in CRC is upregulation of the EGFR pathway and a combination therapy with EGFR inhibitors and BRAF inhibitors might prove effective[20]. Dual synergism could guide strategies that aim to improve outcomes especially in patients who are refractory to initial lines of therapy.

In conclusion, we observed a very low incidence of CRC with BRAF mutations that showed a significant association with right sided tumors, CIMP high phenotype and microsatellite instability. Future work would be directed to establish an effective screening program to study the prevalence of HNPCC cases in Saudi Arabia and also try to understand the complex interaction between genetics and environmental factors that contribute to this low incidence of BRAF mutations in this region.