Introduction

Detailed family history contributed greatly to the understanding of hereditary cancer syndromes as well as to the evolution of molecular genetic of these familial diseases. Cancer family history investigation and genetic counselling have an unquestionable role in selecting families with suspicious hereditary cancer syndromes [1].

In particular, the recent molecular discoveries in gastric cancer (GC), where better defined when this disease occurs as sporadic, familial or hereditary. The knowledge of genetic basis is necessary because it opens different approaches for genetic testing and for clinical management of an asymptomatic population.

Among the hereditary forms, the main genetic mechanism is related with the E-cadherin (E-cad) germline mutation, causing the hereditary diffuse GC (HDGC) syndrome [2, 3]. Moreover, GC can participate in other inherited predispositions as a component of the hereditary nonpolyposis colorectal cancer (HNPCC), Li-Fraumeni (LF) disease and Peutz-Jeghers syndrome (PJS). In this brief review, we described these principal hereditary syndromes offering a simple management to physicians where are these diseases diagnosed.

Familial gastric cancer and environmental factor exposure

About 10% of GC shows a familial clustering and of these only 1–3% constitutes a clear hereditary form [4]; conversely, the majority of GC patients (90%) presents a sporadic tumor. To assess the diagnosis of GC with an hereditary setting, pedigree should present multiple cases of diagnosed GC and members should carry at last a documented germline mutation with high penetrance genotype, otherwise, we could define the form as familial cluster or aggregation. It seems that in a restricted geographic area, environmental factors without a clear genetic cause, such as diet and particular traditional habits contribute to the familial GC development. Specific environmental factors may play a role in generating mismatch repair (MMR) system deficiency leading to microsatellite instability (MSI) phenotype in some GC with familial cluster. Indeed, Palli et al. showed in a study performed in Tuscany, Italy, a geographical region with high rate for GC, an association between MSI, positive family history of GC and high consumption of red meat and nitrates, suggesting that environmental factors, such as nutritional habits, may play a key role in inducing genomic instability and an increased risk for GC with familial clustering [5, 6]. Aggressive and continuous general environmental exposures, such as these particular diets, are considered as potential generators of increased microsatellite alterations in tumors with an increased tolerance to DNA damage associated with reduced MMR activity [7]. Tobacco consumption is also considered another risk factor for gastric carcinoma development [8], leading to a functional deficiency of the MMR system, as for example the XRCC1 gene; we demonstrated recently that the −77T > C XRCC1 promoter polymorphism is related with tobacco consumption and GC with familial aggregation for lung cancer [9].

E-cadherin and hereditary diffuse gastric cancer

HDGC is an autosomal dominant cancer disease caused by germline mutation of E-cad gene (CDH1) [10] that maps to chromosome 16q22.1 and encodes for the protein E-cad, a member of the transmembrane glycoprotein family expressed on epithelial tissue, responsible for calcium-dependent, cell-to-cell adhesion [11, 12]. In 1998 Guilford et al. firstly identified CDH1 germline mutations in Maori kindreds from New Zealand with a strong familial cluster for DGC [2]; subsequently the International GC Linkage Consortium (IGCLC) in 1999 defined the clinical criteria to select pedigree for candidate to the E-cad genetic screening [3]. The clinical criteria were subsequently revised by Brooks-Wilson et al. in 2004 in which patients with sporadic early onset DGC with age at diagnosis ≤35 years were also considered for the genetic screening. In few cases, germline mutations of E-cad can be identified also in pedigree showing DGC and lobular breast cancer and/or mucinous colon cancer [13]. Families fulfilling the clinical criteria for HDGC established by the IGCLC showed a high penetrance of CDH1 gene mutations, with an estimated risk of >80% and an individual CDH1 germline mutation carrier presents more than 80% lifetime risk of developing GC. About 25–30% of families fulfilling the criteria for HDGC have germline mutations of the E-cad gene. Clinical criteria for HDGC definition revised by the IGCLC consensus are the following: (a) two GC cases in a family, one confirmed <50; (b) three confirmed DGC cases in 1st or 2nd degree relatives independent of age; (c) age <40; (d) personal or family history of DGC and lobular breast cancer, 1 < 50 [14].

To date, about 100 CDH1 germline mutations have been reported; from these the most frequent are insertion/deletion, nonsense, splice site (72%), instead missense variations occurred in about 28% of DGC patients [15]. To assess the pathogenic role of CDH1 missense mutations, in silico and in vitro analysis is required [16].

Pathology of HDGC typically presents a diffuse histotype with signet ring cells; tumoral foci are multiple and spread through the entire gastric mucosa with a particular concentration in the antrum and in the transition zone between antrum and body [17]; frequently conventional endoscopy fails to detect these foci at early stage, and addition of chromoendoscopy improves the detection of early HDGC. CDH1 germline mutations are documented only in diffuse GC cases, as well as in sporadic early onset cancers and in hereditary tumors; 37 cases it has been screened showing aggregation for intestinal GC and no E-cadherin constitutional alterations have been identified [18].

DNA mismatch repair genes and hereditary nonpolyposis colorectal cancer

HNPCC or Lynch syndrome is an autosomal dominant human cancer syndrome related with constitutional mutations in the DNA MMR genes, namely MLH1, MSH2, PMS1, PMS2 or MSH6 with a MSI pattern in tumoral samples [19]; among these, germline mutations of MLH1 and MSH2 MMR genes are more frequently detected in HNPCC [20]. MMR system is involved in DNA repair during genomic replication; mutations of MMR genes causes the “mutator phenotype” that leads to the MSI phenotype, due to the accumulation of errors for incorporation of number of bases during replication of long repetitive DNA sequences [21].

Aside from colon and rectum cancer, many other type of cancers can occur in HNPCC pedigree and GC represents a common extra-colonic tumor of this hereditary disease [22, 23]. In GC cases with MMR genes carriers, cumulative incidence is 13% by 70 years of age [24]. HNPCC gastric carcinomas carrying MMR genes germline mutations (as MLH1 or MSH2), show prevalently intestinal histotype (more than 90% of the cases), and only a small percentage are diffuse. This correlation is due to the strong association between MSI tumor phenotype and intestinal GC [25].

In 1991, the International Collaborative Group on HNPCC developed the original Amsterdam Criteria for recruiting HNPCC patients in which performing the genetic testing; in this initial meeting GC was excluded. Subsequently, criteria were modified in 2004 with a revision of the Bethesda guidelines [26], due to the identification of some GC cases in HNPCC family. The novel criteria recommended MSI testing and considered gastric carcinoma as a spectrum of HNPCC [27]. A complete description of clinical criteria was reported by Vasen et al. [28]. Whenever GC patient is component of HNPCC, and fulfills to the restrict described criteria [27], the probability to find a MMR gene mutation is extremely high (>80%) [25].

TP53 gene and Li-Fraumeni syndrome

LF syndrome is a rare autosomal dominant hereditary disease associated with TP53 (chromosome 17p13) germline mutation. Germline carriers of the TP53 mutation show increased risk for multiple primary tumours, such as sarcoma, breast cancer, brain tumours and adrenocortical carcinoma. Other less common tumours have also been associated with LF syndrome, including leukaemia, lung cancer, melanoma, pancreatic cancer, prostate cancer and GC [29, 30].

Clinical criteria were firstly described by Li and Fraumeni in 1988 [31] and subsequently in 1994, Birch et al. introduced a new definition of Li–Fraumeni-like syndrome (LFL). It was based on more extensive and updated information of the type of tumours and the ages of onset in families [32, 33].

The International Agency for Research on Cancer database reports that GC frequency is up to 2.8% of LFS families. TP53 germline mutations have been identified in about 70% of patients with clinical criteria of LF syndrome; constitutional TP53 alterations can appear in both GC histotype with similar frequency, as well as in intestinal and in diffuse forms. In particular, diffuse GC patients with TP53 mutation, in some cases overlap the HDGC clinical criteria [30, 34–38]. However, TP53 constitutional mutations are very rarely documented in the overall GC spectrum [39, 40].

STK11 gene and Peutz-Jeghers syndrome

The PJS is a rare autosomal dominant inherited disease associated with the germline mutations in the tumor suppressor gene namely STK11. This syndrome relates with gastrointestinal hamartomas, polyposis and mucocutaneous pigmentations; criteria were described first by Peutz in 1921 [41] and subsequently by Jeghers in 1949 [42]. Typically, initial morphology of GC with the PJS is a gastric mucosa polyposis. Among gastrointestinal cancers, GC is the third most frequent tumor in PJS, preceded by small intestine and colorectal carcinoma [43], with a cumulative cancer risk of 47% at the age of 65 years [44].

Screening guidelines and surveillance

The correct management of hereditary cancer syndrome requires genetic counselling, information on three generation family pedigrees, histopathological confirmation of gastric carcinoma and informed consent.

  1. (a)

    Hereditary diffuse gastric cancer. The IGCLC guidelines were established in the Cambridge consensus [14]; confirmation of diffuse histotype is mandatory. Genetic testing should be initiated in affected proband and performed on blood, the recommended age for genetic testing is 16 versus 18 years; in CDH1 positive cases, prophylactic gastrectomy or endoscopic surveillance should be proposed. Prophylactic gastrectomy should be considered in truncating CDH1 germline mutations, conversely in subject carriers CDH1 missense mutation prophylactic surgical approach is debateable. Firstly, it is important to define the pathogenicity of the identified missense variation [45]; also in cases of deleterious missense mutation it seems that endoscopic surveillance is preferable [18]. In CDH1 positive cases with deleterious missense mutations or in cases that declined the surgery, endoscopic screening should be performed with chromoendoscopic technique, with a minimum of 30 biopsies and with an annual frequency.

Prophylactic surgical treatment should be proposed in CDH1 truncating carriers, with a total gastrectomy with Roux-en-Y reconstruction and performed in a high volume cancer centre.

  1. (b)

    Hereditary nonpolyposis colorectal cancer. Management of MMR gene mutation carriers is focused on endoscopic colon surveillance or prophylactic surgery; considering GC screening, data are inconsistent because the cancer risk is not extremely high and no studies have been conducted. However, with an increased rate of GC cluster, upper gastrointestinal tract endoscopy could be proposed periodically; in any case, the number of affected members is not a target of predisposition to the Lynch syndrome and the adequate surveillance for GC in Lynch syndrome remains undefined and controversial [46, 47].

  2. (c)

    Li-Fraumeni syndrome. Considering the low incidence rate of GC in pedigree with LF or LFL syndrome, the cumulative risk is unknown and no guidelines for management of TP53 gene mutation carriers have been proposed; it is reasonable to suggest an annual endoscopy in pedigree with multiple cases of GC. Certainly, detection of TP53 mutations contributes considerably to increase GC risk in family with LF or LFL syndrome [31].

  3. (d)

    Peutz-Jeghers syndrome. Considering the relative high cancer risk (47%), surveillance for PJS patients is recommended; in the specific for the gastrointestinal tract, the Dutch surveillance suggests a gastroduodenoscopy starting from 20 years with a frequency of 2–5 years, depending on findings and on the number of cases of GC [42].

Conclusion

Physicians should know GC with familial cluster through simple information as reported in this brief review. Clinical approach to patients with a suspicious inherited predisposition for gastric tumors requires a strict collaboration between geneticist for testing, surgeon, gastroenterologist and pathologist for a correct clinical management and treatment.

In the Fig. 1, we propose a flow-chart for management of familial and hereditary GC patients. The first approach claims a clinical method, providing the correct description of a detailed pedigree, with all familial information, as number of patients, sex, age at onset and at the end of the follow-up, cancer diagnosis, histotype confirmation and outcome. Detection of clinical criteria can help to understand when GC is sporadic or familial or hereditary; in the specific, management of GC with suspicious familial or hereditary firstly requires a pedigree with at least knowledge of three member generations [48, 49], and when analysed pedigree overlap the clinical criteria, genetic testing is mandatory. The identification of a germline alteration requires a specific treatment in asymptomatic patients, in particular in CDH1 mutation carriers in the HDGC syndrome. In a very low percentage of sporadic early onset DGC (3%), it has been documented CDH1 germline alterations, presenting as de novo mutations. E-cad genetic testing should be proposed also in cases of sporadic DGC with age at onset of ≤35 years (Fig. 1) [50].

Fig. 1
figure 1

Algorithm for familial gastric carcinoma management starting from clinical overview, pedigree data collection, clinical criteria, genetic testing, surveillance with endoscopy and gastrectomy

Full size image

Unfortunately, cases with familial aggregation and without specific gene involvement are the majority and represent an urgent clinical burden considering that familial GC, unlike its sporadic form, shows an increasing trend, in particular, in cases with diffuse gastric carcinoma.