E. Strong (ed.). Gastric cancer. Principles and practice. Springer (2015)
HDGC is caused by germline mutations of the E-cadherin gene or (CDH1). CDH1 mutations were first identified in HDGC kindreds by linkage analysis of three New Zealand Maori families with multigeneration, early onset DGC in 1998 . The family in which the CDH1 gene mutation was identified was originally published in 1965 , and remains the largest published HDGC kindred. Family members from this kindred had a median age of death from GC of 33 years, with the youngest documented death occurring at the age of 14 [12, 13]. Since this discovery, mutations in CDH1 have been identified in HDGC families from diverse ethnic backgrounds [14–24]. The pattern of inheritance is autosomal dominant with incomplete penetrance.
CDH1 (MIM# 192090) encodes epithelialcadherin (E-cadherin), a transmembrane calciumdependent cell adhesion glycoprotein that plays an essential role in morphogenesis, the maintenance of normal polarised epithelium, and interacts through catenins with the actin cytoskeleton [25–27]. CDH1 is located on 16q21.1, and consists of 16 exons with a CpG island upstream of the coding region . The translated E-cadherin protein consists of three major domains, a large extracellular domain (encoded by exons 4–13) and smaller transmembrane (exons 13–14) and cytoplasmic domains (exons 14–16) [25, 26].
Somatic mutations in the CDH1 gene were described in DGC before the recognition of HDGC [29–33]. Abnormal E-cadherin expression or mutations have also been found in sporadic lobular breast cancer (LBC) , prostate cancer  and carcinomas of the endometrium and ovary .
In HDGC, mutations are seen over the length of the CDH1 coding sequence, and include point mutations, and small insertions and deletions. More recently, larger deletions have also been identified [37–39]. Table 6.1 describes the published CDH1 mutations found in HDGC families. The frequency of CDH1 mutations in families that meet the International Gastric Cancer Linkage Consortium (IGCLC) criteria for HDGC (see below) appears to be inversely proportional to the incidence of GC in the general population from which the family is drawn, with countries with high incidence of GC having lower incidence of germline CDH1 mutations identified in patients meeting the IGCLC testing criteria . As yet, no genotype–phenotype correlations are apparent.
Table 6.1. Published CDH1 mutations in HDGC families
Mutations with unknown pathogenic relevance, or identified in cancers other than DGC excluded
a Referenced in this paper
b Family does not meet HDGC criteria
As per Knudson’s two-hit hypothesis of tumour suppressor gene inactivation, a second event is required to account for loss or inactivation of the wild-type CDH1 allele . Promoter hypermethylation of the second CDH1 allele has been demonstrated by several groups as the most common mechanism inactivating the wild-type CDH1 allele in HDGC [42, 74]. CDH1 promoter hypermethylation has been found in prostate, breast and sporadic GC [75, 76], and was demonstrated by Grady et al. in 2000 to be the “2nd hit” in some HDGC patients . Grady et al. also demonstrated in vitro that the demethylating agent, 5-azacytidine restored E-cadherin expression in a GC cell line that tested positive for CDH1 promoter methylation, revealing that methylation was the mechanism of silencing. Other mechanisms include somatic mutation, one case of an intragenic deletion has been identified, and it is thought that histone modifications may also be important [42, 68, 74, 78]. In GC tumours from HDGC patients’, loss of heterozygosity is another mechanism for loss of the wild-type CDH1 allele [12, 15, 19, 74].
It is not uncommon for HDGC patients to have multiple foci of tumour in their gastrectomy specimens. Genetic analysis of multiple tumours in the same individual reveal that different mechanisms of silencing of the second allele occur independently at multiple sites in metastatic deposits  and within lesions of the stomach  .
Non-CDH1 hereditary diffuse gastric cancer
About 25 to 30% of patients meeting current clinical criteria for HDGC are found to have germline mutations in CDH1, meaning that up to 70% of families have no identifiable mutation . Using the nomenclature reported by Blair et al., families that fulfil the IGCLC criteria for HDGC (see below) but have no identified CDH1 mutation are designated familial diffuse gastric cancer (FDGC), HDGC refers only to families with a pathogenic CDH1 mutation . CDH1 genetic testing involves sequencing and multiplex ligation-dependent probe amplification (MLPA) to detect large deletions. While this testing is currently the gold standard, there may still be mutations that are missed due to technological limitations. Families without identified CDH1 mutations have been investigated for other potential candidate genes involved in FDGC. In a Dutch kindred with FDGC, a mutation has recently been identified in CTNNA1, which encodes α-E-catenin, making this a potential causative mutation in FDGC . Alpha-E-catenin, in a complex with β-catenin, binds the cytoplasmic domain of E-cadherin to the cytoskeleton [32, 81, 82]. Loss of CTNNA1 in animal models induces altered cell polarity, hyperproliferation, and increase in Rasand mitogen-activated kinase (MAPK) activity—features which are consistent with the potential to induce a malignant phenotype . While the CTNNA1 mutation in this family is suspicious of pathogenicity for HDGC, the phenotype suggests older onset of DGC and mutations in CTNNA1 have not been found in other families to validate the result. It is likely that some families with FDGC may harbour mutations in other genes that have yet to be identified.
Other malignancies associated with HDGC
Other malignancies have shown higher prevalence in families with HDGC, the most prominent of which is Lobular Breast Cancer (LBC) [10, 16, 18, 48]. The risk for developing LBC for females with CDH1 mutations is approximately 60% by age 80 . Therefore, LBC is considered a cancer in the HDGC syndrome that warrants specific management. Colorectal cancer has been identified in some HDGC kindreds [19, 20, 41], although the numbers are small and given this is such a common cancer in the community, direct pathogenesis from a CDH1 mutation has not been established .
Other hereditary syndromes associated with gastric cancer
Other hereditary syndromes are associated with increased risk for GC, including Lynch syndrome or hereditary non-polyposis colorectal cancer (HNPCC), hereditary breast and ovarian cancer, familial adenomatous polyposis (FAP), Cowdens syndrome and Peutz–Jeghers syndrome. Table 6.2 describes the associated risk of GC, which can be intestinal or diffuse type, with selected syndromes where it is known. Patients with hereditary breast and ovarian cancer syndrome have increased risks of malignancies in addition to breast and ovarian cancer, including gastric cancer although the risks are not well quantified [84, 85]. Patients with Lynch syndrome, in addition to colorectal cancer, are at risk of GC, other gastrointestinal malignancies, endometrial carcinoma and carcinomas of the renal tract [6, 86, 87].
Polyposis of the gastrointestinal (GI) tract is associated with a number of familial cancer syndromes. Upper GI polyposis is frequently found associated with FAP. In the stomach this is manifested as fundic gland polyposis [89, 90]. Although sporadic fundic gland polyps are considered to be non-neoplastic, in FAP patients some have been found to harbour dysplasia which may evolve into invasive GC [89, 91, 92]. Peutz– Jeghers and Cowden syndrome patients present with polyposis of multiple organs. Only rare cases of GC have been reported in Cowden syndrome, suggesting this is not a common manifestation of this disease [93, 94]. The risk of cancer in Peutz–Jeghers syndrome is higher, with a lifetime risk of GC of 29% . Juvenile polyposis syndrome is another syndrome associated with hamartomatous polyp formation of the GI tract, with associated increased risk of GI malignancy mainly related to colorectal cancer, but GC has also been documented [95, 96]. GC has also been seen in families with Li–Fraumeni syndrome [97, 98], and some FDGC families have been identified with TP53 mutations suggesting GC is a component of Li–Fraumeni syndrome spectrum . Although there appears to be no significant increase in the incidence of GC in patients with MUTYH-associated polyposis, a autosomal recessive disorder caused by germline mutations in the base excision repair gene MUTYH , monoallelic MUTYH mutation carriers have been reported to have a higher incidence of GC than the general population . There are reports of an increased incidence of GC in relatives of patients with Fanconi’s anaemia . In 2012 a new autosomal dominant condition associated with gastric polyposis and intestinal GC was described. This featured proximal polyposis of the stomach and GC and was named gastric adenocarcinoma and proximal polyposis of the stomach (GAPPS) [102, 103]. The causative gene for this syndrome is unknown.
Table 6.2. Inherited cancer syndromes with associated GC risk
|Cancer syndrome||Gene||Gastric cancer risk lifetime risk (%)|
|Hereditary breast/ovarian cancer||BRCA1||5.5 (3.4–7.5)|
|Lynch syndrome||MLH1, MSH2, MSH6, PMS2||4.4–19.3|
[contact-form-7 id=»5168″ title=»Контактная форма 1″]