Background
One of the main goals of cancer research is to find a way of selecting the appropriate patients for specific therapies. A promising result regarding lung cancer was the discovery that somatic activating mutations in the tyrosine kinase domain of the
EGFR gene identified a subset of non-small cell lung cancer (NSCLC) patients that may respond to tyrosine kinase inhibitors [
3‐
5]. After this seminal discovery, evidence accumulated in recent years has supported the idea that patients harbouring these mutations in their tumours show response to EGFR tyrosine kinase inhibitors [
6,
7]. Based on the results of the latest phase III trials [
8,
9] gefitinib, an EGFR tyrosine kinase inhibitor, has been approved in Europe for first line treatment of adult patients with locally advanced or metastatic NSCLC carrying activating
EGFR somatic mutations. Thus, introduction of mutational screening of somatic mutations into clinical care is already a requirement for taking therapeutic decisions [
10].
As recorded in the Sanger Institute Catalogue of Somatic Mutations in Cancer data-base (COSMIC v49 Release) [
11,
12], curated from the recent literature, a total of 22489 samples from lung cancer tumours have been analyzed for
EGFR tyrosine kinase domain mutations. The number of samples with mutations is 4722, with 5055 mutations detected, being p.L858R missense mutation and in frame-deletions in exon19 the most frequently detected alterations in
EGFR. Resequencing studies provide an opportunity to find novel variants that may be involved in the activity of the protein, and may also allow us to find germ line alterations that support implication of these genes in familial carcinogenesis.
In the course of our EGFR mutational screening in lung adenocarcinoma patients from the area of Asturias, Northern Spain, we found a tumour sample with a double somatic mutation in EGFR gene; one of the changes was demonstrated to be present as a germ-line mutation. We have determined the frequency in our population of the new germ-line mutation as well as the frequency of three additional EGFR germ-line mutations described earlier, and we have investigated the impact of the novel mutation on the biochemical properties of the EGFR protein.
Discussion
In recent years, mutational screening and re-sequencing of potential oncogenes in tumours allowed the identification of hot spots for mutations and also new genes not previously involved in tumour genesis [
12,
18]. Identification of somatic mutations in exons coding for tyrosine kinase domain of
EGFR in tumours from patients with NSCLC and their association with response with targeted therapies, has led to the introduction of mutational screening of tumours in the clinic. A substantial number of tumours have already been studied and in addition to somatic mutations, few germ line mutations have been found.
In this work we have described the identification of a new germ line mutation p.R776G in
EGFR in a patient suffering from NSCLC. In order to rule out that this mutation could be a polymorphism, we analysed 954 alleles from healthy donors and did not find this variation. In addition, we have searched for the presence of this mutation in 285 genomic DNA samples from patients with lung cancer and some history of familial cancer, and 627 samples from patients with lung cancer and no familial cancer antecedents. We have not found this mutation in any other patient. Three other germ line mutations in
EGFR have been described: the first is p.T790M, that has been found in at least five families [
14,
15,
19,
20], p.V843I, described in two cases [
17,
21] and p.P848L [
22]. It is interesting to mention that p.T790M was the first mutation in NSCLC associated with resistance to therapy with tyrosine kinase inhibitors [
23‐
25]. We also studied if these mutations were present in our cohort of individuals with lung cancer and familial history of cancer but we did not find any of them. Moreover, the sequencing of the entire exon 20 of
EGFR did not reveal any other alteration in this gene. Vikis
et al [
26] also analyzed a number of families with high susceptibility to lung cancer for the presence of p.T790M mutation and did not find this alteration in any of 237 probands. Girard
et al [
20] found p.T790M as germ-line in 2 of 369 cases of never smokers with NSCLC, both patients with a family history significant for lung cancer. Concerning p.V843I and p.P848L no studies with a large group of samples was performed previously to the present work. From these results, we can conclude that p.R776G, p.T790M, p.V843I and p.P848L are rare
EGFR alleles that have been found in germ line from patients suffering lung cancer.
Regarding the possibility that these mutations can represent cancer susceptibility alleles, several studies have been performed on p.T790M allele (reviewed by Suda
et al [
27]). Its detection as a somatic mutation in tumours supports this option. Moreover, this mutation has also been found together with the activating p.L858R mutation, both in tumours and in cell lines that never were in contact with tyrosine kinase inhibitor treatments, suggesting that p.T790M might be also growth promoting. Bell
et al [
14] observed that this mutation seems to occur in
cis with the p.L858R activating mutation. Kinase activity of EGFR p.T790M mutant has been reported as indistinguishable from wild-type EGFR [
23,
25], but Vikis
et al [
26] reported that p.T790M mutation alone causes increased phosphorylation levels. Godin-Heymann
et al have shown that, although p.T790M mutation has a moderate effect on EGFR function, when combined with p.L858R or p.746_750del show a remarkable enhancement of EGFR activity [
28]. Finally, animal models generated to inducibly express the p.T790M mutation have been also shown to develop lung tumours [
29]. The latency of developing tumours is longer in those mice with p.T790M mutation alone than those bearing both p.T790M and p.L858R.
In relation to p.R776G, the novel germ line mutation revealed in this work, some lines of evidence similar to the p.T790M mutation would also support its possible role as a cancer susceptibility allele. Mutations at this position in the
EGFR gene have also been discovered in tumours as somatic mutations [
30]. In the tumour sample analyzed in this work, the p.R776G mutation was found together with a p.L858R activating somatic mutation in
cis. A double mutation p.R776G/p.L858R was found by Wu
et al [
30] and the patient harbouring these two mutations did not respond to tyrosine-kinase inhibitor therapy. Alterations at R776 position giving rise to a different amino acid substitution (p.R776C) have also been described as somatic mutations and in conjunction with p.L858R [
5,
31]. Preliminary studies by transfection of cell lines with
EGFR mutant constructs revealed that in the absence of ligand (EGF), an enhanced tyrosine autophosphorylation can be observed in cells harbouring the plasmid containing EGFR-R776G when compared with wild-type EGFR, providing biochemical evidence of the functional relevance of this mutation.
Further analysis encompassing single and double mutant constructs will help to clarify the role of these alterations on EGFR function and in tumour genesis. It is tempting to speculate that, since p.R776G has appeared as germ-line mutation, it would have a weaker effect on EGFR functions than the double mutations detected within the tumours. Functional analyses of these mutants need to be evaluated by cellular assays, for example with tagged constructs [
32], by transformation assays and by tumorigenicity studies in animal models. It will also be interesting to test the sensibility of these mutants to TK inhibitors in order to gain information of clinical value.
Mutations in genes coding for proteins with tyrosine kinase activity have been one of the genetic alterations most frequently detected in cancer over the last years [
33]. Pathways activated by these proteins usually lead to proliferative advantages, thus they have received much consideration, given their association with malignant proliferation. Regarding the role of receptor tyrosine kinase alterations in familial carcinogenesis, germ line mutations in
KIT have been described in familial gastrointestinal stromal tumours (GIST),
RET mutations are involved familial medullary thyroid cancers and other endocrine cancer predisposition syndromes and
MET mutations have been found in papillary renal cancer syndrome among others [
34]. Somatic mutations in these genes have also been described in sporadic tumours and in the case of GIST, activating mutations in
KIT are predictors of response to targeted therapy with imatinib [
35‐
37].
Taking these examples as precedent, it is reasonable to hypothesize that germ-line mutations in the
EGFR gene could be involved in lung cancer pathogenesis, and probably in other tumours in whose development EGFR deregulation can play a role. Although there have been reported some evidences of mendelian inheritance in the pathogenesis of lung cancer [
38], the intrinsic genetic factors controlling susceptibility to this malignancy might be hidden by strong environmental factors like cigarette smoking and air pollution, implicated in the mutagenesis of many other genes controlling pathways involved in the development of these tumours [
18,
39].
More studies are needed in large cohorts of cancer patients with familiar history of cancer in order to link EGFR germ-line mutations to development of the disease, but these results and precedent observations do suggest that EGFR is a candidate to be involved in familial oncogenesis.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
IC: carried out mutational analysis, genetic studies and plasmid constructions. PB: participated in design of the study, analysis of clinical data and helped in draft the manuscript. AA, PM: performed pathological review and selections of all tumour samples and analyses of clinical data. IS, ASP, FI: participated in mutational analysis. FGO, JMF: carried out transfections and western blot analyses and participated in review of the manuscript. PGA, AT: provided samples and clinical data. MB: designed the study and drafted the manuscript. All authors approved the final manuscript.