Mutations in EGFR have attracted attention because of their common occurrence in lung cancers of non-smokers (in particular in adenocarcinoma, women, and patients of Asian origin) and because of their significance as predictors of response to therapeutic tyrosine kinase inhibitors
[19, 34, 35]. In lung cancer, EGFR mutations cluster in exons 18 to 21, encoding the domain of the tyrosine kinase that contains the ATP binding pocket. The most common mutations are short, in frame deletions in exon 19 (34%) and missense mutations at codon 858 (p.L858R) in exon 21 (36%) (http://www.sanger.ac.uk/genetics/CGP/cosmic/). These mutations modify the geometry of the ATP binding cleft in the tyrosine kinase, resulting in a hyperactive form of the receptor. These are not restricted to lung adenocarcinoma and we have reported two mutations among four lifetime never-smokers with squamous cell carcinoma of the lung
. Mutations are infrequent in other cancer pathologies analyzed to date
Based on the hypothesis that these mutations may preferentially occur in a context of non-tobacco dependent carcinogenesis, we investigated whether EGFR mutations could be detected in three series of ESCC from central Asia (two high incidence areas in Northern Iran (Golestan) and Northern India (Kashmir) and one low incidence area in Iran (Tehran). The etiology of ESCC in these areas has been addressed in a number of studies
[5, 7, 8, 28, 37, 38]. Overall, epidemiological studies have consistently reported that tobacco usage is not a significant risk factor, in contrast with ESCC detected in most Western countries and in Japan. Sequencing of exons 18 to 21 of EGFR in a total of 152 ESCC cases detected 14 variations (9.2%). Comparison with COSMIC database indicates that four of these variations are known activating mutations in EGFR TK domain (2.6%), including a common deletion (p.Q746_A750del; 560 occurrences in the COSMIC database) and three less common missense mutations. The other variations identified in this study are known polymorphisms except for two unknown never reported variations which is not clear whether they may activate the tyrosine kinase in the same way as well-characterized EGFR activating mutations.
Previous studies have shown that EGFR activating mutations were rare in ESCC. In a study of 57 cases, Guo et al. reported three EGFR mutations, including a truncating mutation at E872 and two silent mutations at G873 and P753
. In another series of 40 cases (15 of which with amplification of Egfr), Hanawa et al. did not detect any mutation in exon 19 or 21
. To our knowledge, the only study in which a significant proportion of cases contained EGFR mutation (14%) was focused on a rare variant form, basaloid squamous cell carcinoma, suggesting that aberrations in EGFR may be involved in this particularly aggressive form of the disease
. Notably, the series analyzed here does not include basaloid forms of ESCC but comprises cases from Golestan, an area in which we have previously reported an extremely high rate of TP53 mutations (90%), suggestive of the role of diverse mutagens in esophageal carcinogenesis
. Thus, contrary to our hypothesis, and regardless of involvement of environmental mutagens in ESCC from Golestan, EGFR mutations in non-smoking ESCC patients appears to be a rare event which may not play a significant role in the pathogenesis of ESCC. Further studies are needed to determine such EGFR mutations tend to occur at higher frequency in specific groups of ESCC patients.
Despite infrequent mutations in EGFR, over expression of the protein, with or without gene amplification, is a relatively frequent event in ESCC
[16–18, 39–41]. In 2006, Hanawa et al. reported that, among 53 cases of ESCC with high protein expression levels, FISH analysis of amplification revealed clear amplification in 15 cases, evidence for modest changes in copy numbers in 32 cases and no evidence for amplification in six cases, clearly showing that amplification is only partially correlated with expression
. In the present series, we detected Egfr over expression in 65% of the cases, irrespective of tumor grade and stage. However, we have not analyzed the amplification status of the EGFR locus. The only case with known EGFR activating mutation expressed Egfr at moderate levels. The frequent over-expression of Egfr has led to the concept that Egfr-targeting therapies may have some benefit in patients with advanced esophageal cancer. A phase II study of Gefinitib, an Egfr tyrosine kinase inhibitor, in second-line treatment of advanced esophageal cancer, reported that a significantly higher disease control rate (response plus stable disease) was observed in patients with ESCC histology or with high Egfr expression
. These results concur with ours to suggest that further efforts should be made to select esophageal cancer patients who may benefit from Egfr-targeted therapies.