Hereditary Diffuse Gastric Cancer (CDH1 and CTNNA1)

Autosomal dominant syndrome of diffuse gastric cancer and lobular breast cancer (diffuse gastric and lobular breast cancer syndrome), caused by germline loss-of-function variants in CDH1 (E-cadherin) and, in a minority of families, by truncating variants in CTNNA1 (α-E-catenin).

Literature and regulatory information updated: 6 October 2026.

0. Initial Reference Table

1. Introduction

Since the work of Pekka Laurén published in 1965, gastric cancer has been classified into two main histological types: the gland-forming intestinal type, whose principal known cause is Helicobacter pylori infection, and the diffuse type, consisting of poorly cohesive cells that infiltrate the gastric wall without forming a mass and, at an advanced stage, produce the appearance of linitis plastica [25, 26]. Approximately 10% of gastric cancers show familial aggregation, but only 1–3% have a demonstrated hereditary cause, and most families with several cases have neither Lynch syndrome nor an E-cadherin variant [23, 26]. In the United Kingdom, lifetime gastric cancer risk in the general population is 2.3% in men and 1.2% in women, with a mean age at diagnosis of 72 and 75 years and 5-year survival of approximately 10%, whereas in Japan systematic endoscopic screening has raised early cancer survival to more than 90% [26]. Within this epidemiological context, hereditary diffuse gastric cancer is distinguished by its young age at onset and diffuse histology.

The molecular basis was described in 1998 by Parry Guilford and colleagues, who studied a large Māori family from New Zealand (Aotearoa) with early-onset, poorly differentiated, high-grade diffuse gastric cancer. Linkage analysis showed a significant association with markers flanking the gene encoding E-cadherin, a calcium-dependent cell adhesion molecule, and sequencing identified a G>T substitution at the exon 7 splice donor site, with a truncated product; the gene’s role was confirmed by an exon 15 frameshift variant and an exon 13 premature termination codon in two other families [27]. In 1999, a group of geneticists, gastroenterologists, surgeons, oncologists, pathologists and molecular biologists, who would form the International Gastric Cancer Linkage Consortium (IGCLC), reviewed the available pathology, confirmed that all gastric cancers in families with a CDH1 variant were diffuse-type cancers, and published the first criteria and management guidelines [28].

In 2001, Huntsman and colleagues described five asymptomatic carriers, aged between 22 and 40 years, from two families, who underwent prophylactic gastrectomy: all had superficial signet ring cell infiltrates, multifocal in three, and in one individual 65 of 140 blocks were involved, although these represented less than 2% of the gastric mucosa [29]. In the same year, Pharoah and colleagues estimated, in 11 families with at least three cases of diffuse gastric cancer, cumulative risks by age 80 years of 67% in men and 83% in women, and 39% for breast cancer in women [30]. The IGCLC guidelines were revised in 2010 (broader testing criteria and recommendation for breast surveillance from age 35 years), in 2015 (with the Cambridge endoscopic protocol) and in 2020, at the Wānaka meeting, which relaxed the age limits in the criteria, introduced the category of hereditary lobular breast cancer and recognised variation in gastric risk between families [2, 31, 32].

In 2013, through exome sequencing in a large family with hereditary diffuse gastric cancer without a CDH1 variant, Majewski and colleagues identified a truncating germline CTNNA1 allele, present in two relatives with invasive cancer and four with intramucosal signet ring cells detected during surveillance; the second allele was silenced in the tumours and signet ring cells analysed [33]. Portuguese research played a relevant role in this history. Researchers from Porto whose affiliations with the Instituto de Investigação e Inovação em Saúde (i3S) and the Instituto de Patologia e Imunologia Molecular da Universidade do Porto (IPATIMUP) appear in the 2020 guideline, such as Fátima Carneiro, Raquel Seruca and Carla Oliveira, authored central studies: the model of early diffuse gastric cancer development in carriers, the description of germline CDH1 deletions, the demonstration that c.1901C>T is a founder splice variant in northern Portugal, the ERN GENTURIS genotype-first study and the 2024 GeneReviews revision [1, 2, 16, 24, 34, 35].

The past decade has changed three clinical assumptions. The first was penetrance: multigene panels identified many carriers without a family history of gastric cancer, and estimates of lifetime gastric cancer risk fell from 40–70% in men and 56–83% in women to 37–42% and 24–33%, respectively, and, for advanced gastric cancer, to 7–10%, with marked dependence on family history [10, 11, 17, 36]. The second was the significance of intramucosal lesions: stage T1a signet ring cell carcinoma foci are present in almost all carriers’ stomachs, but only a fraction progress, leading NCCN panel members and American and European centres to advocate active endoscopic surveillance as an alternative to gastrectomy [17, 37, 38]. The third was the genetic spectrum: the study of 351 families carrying CTNNA1 variants, published in 2026, characterised this gene as having moderate penetrance and proposed specific testing criteria (“Porto criteria”) [13]. This article integrates these data with the 2020 IGCLC guideline, which remains the most comprehensive international reference.

2. Molecular Genetics and Pathophysiology

2.1. E-cadherin, α-1 catenin and the adherens junction

CDH1 (16q22.1) encodes E-cadherin (epithelial cadherin, cadherin-1), a transmembrane glycoprotein expressed mainly on the basolateral membrane of epithelial cells, where it mediates calcium-dependent cell–cell adhesion and suppresses invasion; it is essential to establish and maintain polarised, differentiated epithelia during development and participates in signal transduction, differentiation, gene expression, cell motility and inflammation [1]. The five repeated extracellular cadherin domains establish homophilic bonds, and the cytoplasmic tail binds to the catenin complex [39]. Adhesive activity depends on this connection to the actin cytoskeleton through α, β and γ catenins; α-1 catenin (α-E-catenin), encoded by CTNNA1 (5q31.2), is the main bridge between adherens junction cadherins and actin, interacts with proteins from at least four cancer-associated signalling pathways (Wnt/β-catenin, NF-κB, Hippo-YAP and hedgehog) and has reported functions in inhibiting proliferation, promoting apoptosis and repressing invasion and metastasis [1]. ClinGen assigns sufficient evidence for haploinsufficiency (score 3) and no evidence for triplosensitivity to both genes [3, 5].

2.2. Two-hit model and initiation of diffuse gastric cancer

CDH1 behaves as a tumour suppressor gene: the inactivated germline allele requires a second somatic hit, which may be promoter hypermethylation, mutation or loss of heterozygosity; hypermethylation is more frequent in primary tumours and somatic deletion in lymph node metastases [1, 32]. Cells lacking E-cadherin lose the ability to adhere to one another and become invasive [1]. E-cadherin also acts as a spatial reference for mitotic spindle orientation: in its absence, divisions no longer occur in the epithelial plane and daughter cells may be displaced into the glandular lumen or lamina propria, a proposed mechanism for the formation of intramucosal foci [1, 7].

The histological sequence was reconstructed from prophylactic gastrectomies. In nine carriers’ stomachs, Carneiro and colleagues found between 1 and 161 foci of early diffuse gastric cancer, occupying 0.005% to 2.96% of the mucosa, in situ signet ring cell carcinoma in seven, and pagetoid spread of signet ring cells beneath the epithelium of the gastric pits and glands; there was no H. pylori or antral clustering of lesions, which were distributed throughout the mucosa, and the authors concluded that random or geographically targeted biopsies would be unlikely to provide sufficient sensitivity [34]. Humar and colleagues located initiation in the proliferative zone (upper glandular neck), where slow proliferation of neoplastic cells forms intramucosal signet ring cell carcinoma with differentiated features and relatively indolent behaviour; invasion beyond the mucosa is associated with dedifferentiation, increased proliferation, c-Src activation and epithelial–mesenchymal transition [40]. A spatial multiomic analysis of 20 gastrectomy samples (2025) showed that signet ring cell lesions lack alterations in known gastric cancer drivers (TP53, ARID1A, KRAS), whereas advanced diffuse gastric cancer re-expresses E-cadherin, harbours somatic TP53 and ERBB3 mutations and overexpresses CTNNA1, MYC and MET; the authors propose considering signet ring cell lesions as precancerous lesions [41].

A mouse model published in 2026 (preclinical data) showed that Cdh1 inactivation in Cd44-positive gastric cells was sufficient to produce in situ signet ring cells (pTis) within one week, intramucosal carcinoma (pT1a) within 2 months and advanced carcinoma (pT3) in 50% of animals at 9 months, without additional detected genetic events; concomitant Trp53 inactivation accelerated progression to pT3 within 3 months [42]. These animal data do not estimate human risk, but reinforce the interpretation that loss of E-cadherin initiates disease and other alterations modulate progression.

The precursor lesions recognised exclusively in CDH1 carriers are in situ signet ring cell carcinoma (cells with hyperchromatic, depolarised nuclei within the basement membrane of a gland) and pagetoid spread of a row of signet ring cells beneath the preserved epithelium; both in situ and T1a foci overexpress cytokeratin 7 [1, 2]. Carriers of truncating CTNNA1 variants also have multifocal intramucosal signet ring cell carcinoma, but in situ foci have not been observed to date [1].

2.3. CDH1 variant spectrum

Pathogenic CDH1 variants are predominantly truncating (nonsense and frameshift), splice variants or missense variants with a demonstrated effect on splicing [1, 43]. In the ClinGen Variant Curation Expert Panel (VCEP) review of 273 variants, those classified as pathogenic were nonsense, frameshift, splice or initiation codon variants, and the only pathogenic or likely pathogenic missense variants had a demonstrated effect on splicing [43]. Intragenic or whole-exon deletions account for less than 5% of pathogenic variants according to the IGCLC; in a series of 160 families meeting criteria, they were found in 4% of HDGC families (six of 93 probands without a point mutation, all from low-incidence populations in the United Kingdom and North America), raising overall detection of CDH1 alterations to approximately 46% (73/160); in that series, two families shared a 193,593 bp deletion encompassing the whole of CDH3 and CDH1 exons 1 and 2, and the predominant mechanism was non-allelic homologous recombination between Alu sequences [2, 35]. In the European genotype-first study, the most frequent single-nucleotide variant was c.1901C>T (13 families) and the most frequent copy number variation was deletion of exons 1–2 (10 families) [24].

The same study, involving 854 carriers of 398 rare variants from 29 institutions in ten ERN GENTURIS countries, showed that pathogenic and likely pathogenic variants were associated with lobular breast cancer (odds ratio 12.39; 95% CI 2.66–57.74), diffuse gastric cancer (8.00; 2.18–29.39) and gastric cancer (7.81; 2.03–29.96), whereas missense variants of uncertain significance showed no association with these phenotypes; no variant specifically predisposed to lobular cancer [24]. The practical conclusion is that detection of a CDH1 missense variant on a panel does not, in itself, justify high-risk management.

2.4. Founder variants and standardised numbering

The following table lists the founder variants mentioned in GeneReviews, with nomenclature confirmed in VariantValidator for the MANE Select transcript NM_004360.5 and in ClinVar [1, 6, 44–46].

The c.1901C>T variant illustrates a limitation of predicted protein nomenclature: VariantValidator describes p.(Ala634Val), because it derives the consequence from the coding sequence, but RNA analysis in carriers showed that the cryptic site was preferred over the canonical site in 100% of sequenced clones, producing only transcripts with a premature termination codon; the old designation “p.A634V” led to interpretation as a missense variant for years [6, 16]. The Newfoundland variant was characterised in 2007 in four families with a shared haplotype, two of which had initially undergone BRCA investigation because of a concentration of lobular cancer; in that study, c.1901C>T and c.1137G>A occurred on shared haplotypes in two families and different haplotypes in a third, demonstrating that recurrence results from both common ancestry and independent mutations [47].

2.5. Genotype–phenotype correlations

Variants associated with cleft lip and/or palate cluster in the linker regions between extracellular domains, and the calcium-chelating region between amino acids 253 and 260 is particularly intolerant of missense substitutions [1, 48]. In a cohort of 153 families, 24% of cleft-associated variants were deletions of at least one exon and the authors found no genotype–phenotype pattern, qualifying the structural correlation [14]. In CTNNA1 carriers, truncating variants are associated with diffuse gastric cancer and non-truncating variants with pattern macular dystrophy 2, particularly when located in the M fragment of α-E-catenin [13].

2.6. CTNNA1: mechanism and variant classes

In the study of 1308 individuals from 351 families carrying CTNNA1 variants and 37,428 non-carriers, transcripts with truncating variants were degraded by nonsense-mediated decay (NMD), carriers’ diffuse gastric cancers lost α-E-catenin and truncating transcripts were non-functional in a humanised Drosophila model, unlike non-truncating transcripts; diffuse gastric cancer risk was eight times greater in carriers of truncating variants than in carriers of non-truncating variants [13]. The 2021 systematic review had already shown that truncating variants in the final exon of CTNNA1, which escape NMD, have limited deleteriousness, and that only pathogenic (rather than likely pathogenic) variants proved actionable in HDGC families [8]. Examples of recurrent truncating variants, with numbering confirmed in ClinVar for NM_001903.5, are c.1351C>T p.(Arg451Ter), classified as pathogenic, and c.2023C>T p.(Gln675Ter), with discordant classifications; in the 2026 study, 33.3% (3/9) of families with c.1351C>T and 7.7% (2/26) of families with c.2023C>T met the Porto criteria [13, 49].

2.7. Allelic disorders, somatic alterations and other genes

Blepharocheilodontic syndrome 1 (OMIM 119580), with lower eyelid ectropion, distichiasis, cleft lip and palate and ectodermal abnormalities, is caused by CDH1 missense variants (five variants in 11 patients from 8 families in the original study, with accelerated E-cadherin degradation) or by truncating variants in CTNND1, which encodes p120-catenin [1, 50]. Pattern macular dystrophy 2 (OMIM 608970), including butterfly-shaped pigment dystrophy, was associated in 2016 with CTNNA1 missense variants in three families and a mouse model, with loss of retinal pigment epithelium integrity; in September 2026, ClinGen classified the relationship between CTNNA1 and hereditary macular dystrophy as definitive, separate from its oncological curation because of phenotypic variability and differences in variant type [5, 51].

In the synthesis of the 2021 French PNDS (French National Diagnosis and Care Protocol), which brings together 19 families with CDH1-associated blepharocheilodontic syndrome, most variants cluster in two hotspots. In seven families, the substitution causes skipping of exon 9, which removes residues Tyr380 to Lys440 in frame; one example is NM_004360.5:c.1320G>T, at the last nucleotide of the exon, classified in ClinVar as likely pathogenic and associated with the syndrome. In eight families, the substitution affects the conserved sequence Asp254-Gln255-Asn256-Asp257, which participates in calcium binding, such as NM_004360.5:c.760G>A p.(Asp254Asn), with conflicting classifications in ClinVar, c.760G>T p.(Asp254Tyr) and c.768T>A p.(Asn256Lys), both likely pathogenic; the remaining families have isolated missense substitutions or in-frame deletions of one amino acid, of less clear biological effect [15, 52]. Variants associated with non-syndromic clefting, with or without cancer in the family, do not differ in type or position from the usual oncological variants, which are mainly truncating and distributed along the gene [15]. Functional assays illustrate the limits of the correlation: c.760G>A p.(Asp254Asn) and c.2351G>A p.(Arg784His), the latter of uncertain significance in ClinVar, relocate E-cadherin to the cytoplasm without conferring invasive capacity on the cells, which could explain the absence of cancer in the respective families; conversely, a variant from a patient with clefting and no cancer increased cell invasion, and five variants found in families with clefting without cancer were also observed in unrelated families meeting criteria for HDGC or hereditary lobular cancer. The PNDS concludes that these assays cannot be used to estimate individual cancer risk [15, 52].

Somatic loss of E-cadherin is frequent in sporadic gastric, breast, ovarian, pancreatic and non-melanoma skin carcinomas, through mutation, loss of heterozygosity or promoter hypermethylation, and somatic CTNNA1 variants have been described in various tumours and myeloid leukaemia; in these circumstances predisposition is not hereditary [1]. In breast cancer, CDH1 is among the genes with recurrent somatic deletions and mutations and those enriched in metastases [53]. In gastric adenocarcinoma, The Cancer Genome Atlas (TCGA) “genomically stable” subtype is enriched for diffuse histology [54]. ClinGen excluded somatic ovarian, prostate and endometrial phenotypes, susceptibility to early-onset breast cancer and blepharocheilodontic syndrome from its oncological curation of CDH1 [3].

The IGCLC states that, beyond CTNNA1, no genes specifically predisposing to diffuse rather than intestinal gastric cancer have been identified, despite panel and exome studies; in unexplained families, PALB2 and other genes associated with gastric risk (ATM, BRCA2, Lynch genes, APC, TP53) may be investigated [2]. The current Orphanet definition includes MAP3K6 among the genes for the entity, which is not supported by the IGCLC or GeneReviews; this article follows these two sources [1, 2, 4].

3. Clinical Spectrum and Manifestations by System

3.1. Diffuse gastric cancer

Diffuse gastric cancer is a poorly differentiated adenocarcinoma, also termed signet ring cell carcinoma or isolated cell carcinoma, which infiltrates and thickens the gastric wall without forming a distinct mass [1]. The signet ring appearance results from intracellular mucin accumulation displacing the nucleus to the periphery, and no clearly defined preneoplastic lesion is identified; malignant cells infiltrate beneath histologically normal mucosa, producing rigidity and diffuse thickening (linitis plastica) [1]. In the 2019 World Health Organization (WHO) classification it corresponds to poorly cohesive carcinoma; in Japan and South Korea, use of the Japanese Gastric Cancer Association classification is recommended for signet ring cell carcinoma, and index cases from new families with advanced disease may present as poorly differentiated adenocarcinoma of non-solid type [2].

In the largest CDH1 carrier cohorts, mean age at onset ranges from 42 to 49 years, with cases from age 14 years; age and behaviour vary between and within families [1]. In northern Portuguese families with c.1901C>T, mean age at clinical diffuse gastric cancer was 33 ± 12 years (27 ± 7 in men and 37 ± 13 in women), and the Oxford Desk Reference reported a mean of 38 years in HDGC families, compared with 60–70 years in sporadic gastric cancer [16, 26]. In the CTNNA1-associated form, mean age at onset is approximately 40 years [1, 8].

Initial symptoms are non-specific and tend to be dismissed by patients and doctors; when specific symptoms appear, disease is usually advanced [1]. Late manifestations include abdominal pain, nausea, vomiting, dysphagia, postprandial fullness, anorexia and weight loss, sometimes a palpable mass; spread is predominantly to the gastrointestinal tract, peritoneum, omentum, mesocolon and female genital tract, and is rare in other organs [1]. Clinically evident disease coexists, in asymptomatic carriers, with occult intramucosal foci: signet ring cell carcinomas have been found in prophylactic gastrectomies from women with lobular cancer without a personal or family history of gastric cancer, and in a prospective cohort 93.8% (15/16) of patients from HLBC families undergoing risk-reducing gastrectomy had occult carcinoma, a prevalence similar to that in HDGC families (94.7%) [1, 55]. At an American centre, 12 of 20 carriers had signet ring cell carcinoma on endoscopy and/or gastrectomy, almost always minute and confined to the lamina propria, and 7 of these 12 had no diffuse gastric cancer in first-degree relatives [56].

3.2. Gastric cancer risk: the problem of ascertainment

Penetrance estimates depend on how families were selected, whether T1a lesions detected in prophylactic gastrectomies were included, and the period. The table compares the principal estimates with denominators and outcomes.

Sources: Pharoah 2001; Kaurah 2007; Hansford 2015; Xicola 2019 and Roberts 2019 with sex-specific and subgroup values from the IGCLC supplementary material; Ryan 2024; Barbosa-Matos 2021 [2, 9–11, 16, 30, 36, 47].

The 2024 North American study warrants careful reading. Observed gastric cancer prevalence among carriers was 13.9% (123/883), and breast cancer prevalence among female carriers was 26.3% (144/547); the hazard ratio for advanced gastric cancer was 33.5 at age 30 and 3.5 at age 70 years, and risk for a carrier with three affected first-degree relatives was estimated at approximately 38% (95% CI 25–64%) [11]. GeneReviews adds that including all gastric cancers, including occult stage IA cancers found at gastrectomy, approximately doubles the estimate (from 7–10% to 13–19%) [1]. The IGCLC emphasises that variation in risk between families is probably influenced by genetic background and lifestyle factors, so a family’s historical risk should not automatically be extrapolated to younger generations [2]. The figures in the table are not interchangeable: “gastric cancer risk” in clinical cohorts may include occult T1a lesions found at gastrectomy, of uncertain prognostic significance; “advanced gastric cancer risk” more closely approximates the clinically relevant outcome; and the proportion of patients in a Portuguese series measures disease diagnosed at a particular time, rather than lifetime penetrance. Even in a region with a high incidence of gastric cancer, fewer than 20% of c.1901C>T carriers had developed diffuse gastric or lobular cancer, according to GeneReviews [1, 16].

Older textbooks and guidelines reproduce values from selected families: gastric cancer penetrance of 70–80% in the 2017 Oxford Desk Reference, 50–80% diffuse gastric cancer and 55% lobular cancer in the same book’s cancer patterns table, 35–45% diffuse gastric cancer and 40% breast cancer in the 2023 ESMO guideline table, and a relative risk of 6.6 with an absolute risk of 53% by age 80 years for breast cancer (lobular) in the Brazilian genetic syndromes textbook table [18, 26, 57, 58]. The ESMO genomics handbook reports lobular cancer risk of 39–52% [53].

3.3. Lobular breast cancer

Lobular cancer is the second most frequent cancer in the syndrome and may present as invasive lobular carcinoma or lobular carcinoma in situ, with a mean age at onset between 45 and 54 years in female CDH1 carriers [1]. Tumour cells are small, poorly cohesive, dispersed in the stroma or in single files, and lack E-cadherin; no histological or immunohistochemical features distinguish the hereditary from the sporadic form, and cytoplasmic p120-catenin staining with negative β-catenin helps confirm the lobular phenotype [1, 2]. Bilateral, diffuse foci of atypical lobular hyperplasia, lobular carcinoma in situ and small invasive foci have been found in risk-reducing mastectomies from carriers [2]. Invasive lobular carcinoma accounts for 5–15% of invasive breast cancers, and a pathogenic CDH1 variant is found in 0.1–0.5% of women with lobular cancer; in a cohort of invasive and in situ lobular cancers, 8% of patients with bilateral cancer were carriers [1]. In a case–control study of 1434 invasive lobular cancers in women up to age 60 years, 11% of patients up to age 40 years had variants in known susceptibility genes, including CDH1 [59]. Metastatic lobular disease spreads to the gastrointestinal and female genital tracts and the peritoneum, and gastric metastases from lobular cancer are among the histological mimics of signet ring cell carcinoma [1, 2].

Female CDH1 carriers may have invasive carcinoma of no special type (formerly “ductal”); the IGCLC considers these cases probably sporadic and states that CDH1 testing should only be considered in women with confirmed lobular cancer [2]. In CTNNA1 carriers, GeneReviews describes lobular cancer as very rare, whereas the study of 351 families describes it as recurrent, including families not meeting HDGC criteria; in that study’s laboratory cohort, however, the lobular cancer odds ratio was not significant for truncating CTNNA1 (1.20; 95% CI 0.59–2.14), unlike CDH1 (9.62; 7.36–12.46) [1, 13].

3.4. Cleft lip and palate and other non-oncological findings

Cleft lip with or without cleft palate was reported in at least one member of 19% of CDH1 families (29/153) and in 2.7% (8/299) of confirmed carriers; families with and without clefts had similar gastric and breast cancer frequencies [14]. Information on type and severity is limited, presence varies within the same family, and it has not been associated with truncating CTNNA1 variants [1]. A Chinese study of 121 gastric cancer patients identified three carriers of germline CDH1 variants (2.5%), and described congenital tooth agenesis in two patients, a finding the authors associate with CDH1 variants and which, in their view, requires further studies [60]. Blepharocheilodontic syndrome and pattern macular dystrophy, which are allelic disorders, are described in section 2.

The French PNDS (French National Diagnosis and Care Protocol) describes the developmental spectrum as a continuum, from isolated cleft lip or cleft lip and palate, paramedian and almost always bilateral when linked to CDH1, to blepharocheilodontic syndrome [15]. In the latter, all patients have anomalies of the eyelids, hair and teeth: euryblepharon (bilateral horizontal widening of the palpebral fissure), lagophthalmos, lower eyelid ectropion and lacrimal duct anomalies, sometimes distichiasis and ankyloblepharon; sparse hair with a high hairline and a prominent forehead; delayed eruption, conical teeth with abnormal enamel and dental agenesis that may amount to severe oligodontia, in a picture of ectodermal dysplasia of variable expression. Cleft lip and palate is present in the great majority of patients and is usually accompanied by hypertelorism. Less frequently, there is congenital hypothyroidism due to thyroid hypoplasia or aplasia, anorectal malformation, neural tube closure defect (spinal meningocele, anencephaly) and cutaneous syndactyly [15]. Inter- and intrafamilial variability is wide: some patients have severe forms and others only a slight widening of the palpebral fissures and hypodontia, so that the diagnosis is sometimes made retrospectively in a parent after molecular confirmation in the child [15].

The relationship with cancer varies according to the mode of ascertainment. Of the 26 families with CDH1-associated non-syndromic clefting gathered in the PNDS, six had cancers of the syndrome spectrum, with 28 patients with cancer among 32 carriers (ages not available), and the other 20 families, with 79 carriers, had no history of cancer; three of these were large families, one with 25 carriers of p.(Asp254Asn) over four generations, another with 7 carriers of p.(Arg784His) over four generations and another with 11 carriers of c.468G>C p.(Trp156Cys) (likely pathogenic in ClinVar) over five generations, two of them aged 56 and 60 years [15, 52]. Blepharocheilodontic syndrome was not recognised as a condition of cancer risk before the identification of CDH1; variants of residue Asp254 have been found in three families with the syndrome and have never been associated with cancer [15]. In two families, however, the variant segregated with cancer: in the first, with p.(Asn256Lys), the proband had diffuse gastric cancer at age 37 years and the mother mixed ductal-lobular breast cancer at age 48 years; in the second, with exon 9 skipping, the patient and her daughter had the syndrome, the patient's mother, without facial dysmorphism, had died of gastric cancer at age 35 years, and it was screening gastroscopy that revealed diffuse gastric cancer in the patient [15]. Outside these situations, neither the clinical picture nor the characteristics of the variant explain the occurrence of cancer, and the PNDS accepts, except for the Asp254 variants, a probable excess risk that cannot be quantified [15]. In blepharocheilodontic syndrome caused by CTNND1, the cancer risk has not been established [15].

3.5. Other cancers

The IGCLC concludes that there is no strong evidence of a significant increase in other cancer types and no basis for additional colorectal screening beyond population programmes [2]. Laboratory data show enrichment of variants in specific settings: among 212,944 patients tested at an American laboratory, 141 (0.06%) had a pathogenic CDH1 variant, with the highest frequency in diffuse gastric cancer (27/408; 6.6%), followed by colorectal signet ring cell carcinoma (3/79; 3.8%), gastric cancer (56/2756; 2%) and lobular cancer (22/6809; 0.3%) [61]. In an American study, 12% of families with a pathogenic variant had neither gastric nor breast cancer and colorectal cancer was the third most frequent, whereas in 176 European families colorectal cancer accounted for less than 2% [1]. The 2015 guideline proposed colonoscopy from age 40 years, or 10 years before the youngest case, at 3–5-year intervals, only in families with colorectal cancer in carriers; the 2020 guideline did not retain an additional recommendation [2, 32]. Recent isolated reports describe poorly differentiated gallbladder carcinoma with reduced E-cadherin expression in a 67-year-old carrier and signet ring cell carcinoma arising in ectopic gastric mucosa in the duodenal bulb; these are single-case observations without risk estimates [62, 63].

3.6. CTNNA1-associated spectrum

In a systematic review of 105 individuals from 41 families with germline CTNNA1 variants, only 24% of families met HDGC criteria and affected individuals predominantly had diffuse gastric cancer at around age 40 years; in the remaining 76%, breast cancer of unspecified histology or absence of cancer predominated, and only 2 of the 105 individuals had lobular cancer [1, 8]. In the 2026 study, the odds ratio for gastric cancer in truncating CTNNA1 variant carriers, compared with 37,428 individuals without a variant in either CTNNA1 or CDH1, was 6.97 (95% CI 2.91–14.13), compared with 37.55 (26.69–52.28) in CDH1 carriers; the authors classify CTNNA1 as a moderate-penetrance HDGC gene [13].

4. Clinical Recognition and Warning Signs

Recognition depends on a pedigree covering at least three generations, including all cancers, ages at diagnosis, institution and, whenever possible, histology reports, and on review of the histological classification of gastric (diffuse or non-diffuse) and breast cancer (lobular or non-lobular) [1, 2, 26]. Absence of a family history does not exclude the syndrome, because penetrance is reduced and cancer in an earlier generation may not have been classified [1]. Warning signs that should trigger genetics referral are:

Two pitfalls are frequent. Confirmed intestinal-type gastric cancer and non-lobular breast cancers do not count towards IGCLC criteria, and the 2015 guideline advised against CDH1 investigation in families whose confirmed gastric cancers were intestinal-type [2, 32]. Conversely, the requirement for histological confirmation means that many carriers do not meet criteria: in a series of 112 carriers identified mainly through panels, only 19% met the 2015 or 2020 criteria without assuming diffuse histology in cases without a report [65]. Therefore, the IGCLC guidelines consider confirmation of reports and, preferably, review of gastric specimens by a dedicated gastric cancer pathologist essential [2, 32].

In developmental anomalies, suspicion follows a different pathway [15]. Paramedian cleft lip and palate occurs in about 1 in every 1,000 births and is usually multifactorial; there is no pathognomonic sign of CDH1, and without a family history of clefting or of cancer of the spectrum the PNDS (French National Diagnosis and Care Protocol) does not recommend looking for a monogenic cause, not even in the prenatal period. With a suggestive family history, targeted fetal ultrasound may raise the suspicion. In blepharocheilodontic syndrome, prenatal findings (clefting, particularly bilateral, choanal atresia, neural tube defect or anencephaly, cutaneous syndactyly, digestive signs of anorectal malformation) only point to the diagnosis if there is a family context; after birth, it is the association of bilateral clefting with eyelid malformations, congenital hypothyroidism due to thyroid hypoplasia or aplasia, anorectal malformation or neural tube defect that suggests it, and the ectodermal signs, with dental agenesis in the foreground, become evident during childhood [15].

5. Diagnosis: Criteria, Tests and Algorithms

5.1. Operational definitions

The IGCLC abandoned the clinical definition in 2020: HDGC became defined by the presence of a germline pathogenic or likely pathogenic CDH1 or CTNNA1 variant in an isolated individual with diffuse gastric cancer, or in a family with one or more cases of diffuse gastric cancer in first- or second-degree relatives [2]. Hereditary lobular breast cancer (HLBC) is defined by the presence of a pathogenic CDH1 variant in an isolated individual with lobular cancer or a family with one or more lobular cancers, without known diffuse gastric cancer; the family is recategorised as HDGC if diffuse gastric cancer or a precursor lesion develops [2]. “HDGC-like” families meet family criteria 1 or 2 in the following table without an identified pathogenic variant, therefore having at least one confirmed diffuse gastric cancer and another gastric or lobular cancer in first- or second-degree relatives [2]. GeneReviews establishes a DGLBCS diagnosis in a proband with suggestive findings and a heterozygous germline pathogenic (or likely pathogenic) CDH1 variant, or a heterozygous germline truncating CTNNA1 variant [1]. The terms “pathogenic” and “likely pathogenic” are treated as clinically equivalent, although the IGCLC notes that “likely pathogenic” corresponds to approximately 90% probability of pathogenicity and may subsequently be reclassified [1, 2].

5.2. IGCLC 2020 genetic testing criteria (full text of Panel 1)

CDH1 testing is recommended when one of the following criteria is met, after confirmation of cancer diagnoses. When a criterion involves two or more cancers, at least one should have confirmed histology, and the remaining relevant cancers should be confirmed whenever possible. Histologically confirmed intestinal-type gastric cancer and non-lobular breast cancers should not be used to fulfil criteria. Those meeting criteria without a CDH1 variant should undergo CTNNA1 investigation [2].

*Family members must be first- or second-degree biological relatives of one another. Whenever possible, test an affected individual; if no affected individuals are alive, consider tissue testing (tumour or normal) from a deceased relative; if this is not possible, consider indirect testing in unaffected relatives [2]. In the latter scenario, the 2015 guideline recommended preferentially testing at least three first-degree relatives simultaneously and explaining that each has only a 50% probability of having inherited the variant [32].

The IGCLC also recommends CDH1 testing in patients with multiple signet ring cell carcinoma lesions (endoscopy or gastrectomy) and all Māori patients with confirmed diffuse gastric cancer [2]. In 2015, the detection rate using the 2010 criteria had fallen to 10–18% in low-incidence countries, being higher in Māori and Newfoundland, and between 8% and 15% in small series from South Korea, Japan and Portugal [32]. The Oxford Desk Reference reproduces the 2015 algorithm, with testing from the age of informed consent including MLPA (multiplex ligation-dependent probe amplification) [26].

5.3. Alternative criteria and subsequent proposals

Sources: Lerner 2025 (Table 1) and Lerner 2023; Garcia-Pelaez 2023 (Figure 3B, read as an image); Lobo 2026 (Figure 4A, read as an image); Corso 2018 [13, 17, 24, 65, 66].

Divergence between sets reflects different choices between sensitivity and specificity. The NCCN and Yale criteria do not require histological confirmation and capture carriers with gastric cancer of unknown histology and families with breast cancer alone, at the cost of testing many women with non-lobular breast cancer; European criteria maintain emphasis on lobular or diffuse histology, which in the ERN GENTURIS study was the phenotype most strongly associated with pathogenic variants [17, 24]. In the European study, 18 of the 29 carrier families not meeting the 2020 criteria had confirmed lobular cancer and were covered by the expanded criteria, whereas the authors considered that the Yale criteria would result in testing an excess of breast cancer patients with a very low yield [24].

5.4. Laboratory strategy

Germline testing should be performed in a certified laboratory (CLIA, ISO 15189 or equivalent) and include sequencing of the entire coding region with intron–exon boundaries and copy number analysis of individual exons [2]. GeneReviews describes two approaches: targeted testing, as sequential single-gene testing (CDH1 sequencing and deletion/duplication analysis and, if negative, CTNNA1) or a multigene panel including CDH1, CTNNA1 and differential diagnosis genes, and comprehensive genomic testing by exome or genome sequencing [1]. In individuals without an identified variant, CDH1 and CTNNA1 RNA analysis may reveal splice defects [1]. A 2025 RNA sequencing study in lymphoblastoid lines described eleven alternative CDH1 splicing events, including an isoform lacking exon 11 with high expression in controls, and allowed reclassification of c.1008+1G>A, c.1936+5G>A and a duplication of exons 4 to 11 as pathogenic [67]. This finding highlights that detection of aberrant transcripts must be compared with the gene’s physiological splicing profile.

Source: GeneReviews, Table 1 [1].

The percentages in the second and third columns express the proportion of known pathogenic variants detected by each method, rather than the clinical sensitivity of testing in a patient meeting criteria, which is limited by the fraction of families without an identified cause. A negative result does not exclude hereditary predisposition when the phenotype is suggestive, and these families continue surveillance according to empirical risk [1, 2]. Technical limitations include deep intronic or regulatory variants, whole-gene duplications, epimutations, regulatory region defects and somatic mosaicism, topics the IGCLC lists as research priorities for variants of uncertain significance and families without a clinically relevant variant [2]. Parental mosaicism may explain absence of the variant in a proband’s parents (section 8) [1].

The report should state the gene, transcript with version (NM_004360.5 for CDH1; NM_001903.5 for CTNNA1), HGVS (Human Genome Variation Society) nomenclature at DNA and protein levels, zygosity, the five-tier ACMG/AMP classification with the specification used, and method limitations [1, 6]. For CDH1, classification should follow ClinGen VCEP specifications; no specific specifications exist for CTNNA1 and general ACMG/AMP rules are used, although the 2021 review advocated using the CDH1 rules [1, 8]. Variants of uncertain significance do not justify predictive testing in relatives, and published in vitro or in silico functional assays for missense variants should not be used for classification until clinically validated, remaining research tools [2]. A positive direct-to-consumer test result must be confirmed in a certified laboratory [2].

5.5. CDH1-specific ClinGen rules

The ClinGen Gastric Cancer VCEP published CDH1 specifications in 2018, following evaluation of approximately 827,000 sequenced alleles, and updated them in 2023 with eleven major modifications; using the refined rules, 97% (36/37) of variants with conflicting ClinVar interpretations and 35% (15/43) of variants of uncertain significance were resolved, and 88% of the 273 curated variants fell outside the uncertain category [43, 68]. Version 3.1 is published in the ClinGen specification registry and the panel has completed 335 CDH1 variant classifications [3, 21].

Source: ClinGen CDH1 v3.1 specification (registry GN007) and variant c.2647T>C confirmed in VariantValidator [6, 21].

5.6. Endoscopy and histopathology as diagnostic tools

Endoscopy does not diagnose predisposition, but identifies disease and informs decisions; biopsies from individuals with a family history should be examined by a pathologist experienced in HDGC, with haematoxylin and eosin at three levels and periodic acid–Schiff with diastase (PAS-D) [32]. Typical signet ring cell carcinoma foci and any features of an aggressive phenotype (pleomorphic or bizarre cells, mucin loss, spindle-shaped or small cells, mitoses) should be described, as they suggest progression and should trigger staging [2]. Mimics include glassy cell change, globoid change of the foveolar epithelium, procedure-related traumatic signet ring pseudocells, epithelial vacuolation, Russell body gastritis, pseudocells in chronic gastritis, lobular cancer metastasis (with positive oestrogen receptor), xanthomatous cells and neuroendocrine tumour [2].

Prophylactic gastrectomy specimens should be assessed using the 2019 WHO and Laurén classifications, absence of residual gastric mucosa at the margins should be confirmed (proximal oesophageal squamous and distal duodenal mucosa), and a checklist including the number, location, size and aggressive features of pT1a foci, stromal reaction, margins and lymph nodes should be followed [2]. Mapping with complete mucosal embedding requires approximately 120 to 270 blocks and ten times the resources of a conventional gastrectomy, so the IGCLC proposes three levels: level 1 (minimum: photograph, margins, lymph nodes and sampling of all areas, approximately 20–50 blocks), level 2 (embedding the entire mucosa with sectioning of a subset) and level 3 (sectioning and examination of all blocks) [2]. If no foci are found, the report should not state “negative”, but “no carcinoma found in xx% of the mucosa examined” [2]. The a priori probability of at least one signet ring cell carcinoma focus in a CDH1 carrier’s gastrectomy is 95%, and yield is higher with complete embedding (122/128, 95.3%) than without it (25/40, 62.5%) [2, 17].

5.7. Algorithm

The 2020 IGCLC flowchart, consulted in the original figure (Figure 1 of Blair et al., available at PMC7116190, Figure 1), organises decision-making into four branches [2]: (1) criteria met or alternative identification pathways (cleft lip/palate, multigene panel, validated direct-to-consumer testing) lead to CDH1 and CTNNA1 testing; (2) CDH1 carriers from families with diffuse gastric cancer (HDGC) receive a PTG recommendation, annual surveillance if they decline or defer, less emphasis on PTG if the family history is weak, gastrectomy if biopsy-positive, annual breast surveillance and the option of bilateral mastectomy; carriers from breast-cancer-only families (HLBC) and carriers without gastric or breast cancer in the family receive annual gastric surveillance, gastrectomy if biopsy-positive and consideration of PTG, with reassessment if the family history changes; (3) CDH1 variants of uncertain significance and “HDGC-like” families receive annual gastric surveillance for at least 2 years, with intervals extendable thereafter, without PTG and with individualised breast management; (4) a negative test without family criteria 1 or 2 suggests sporadic diffuse gastric or lobular cancer. The CTNNA1 pathway is described in the guideline text (section 7).

5.8. CDH1-associated developmental anomalies: confirmation and initial assessment

The French PNDS (French National Diagnosis and Care Protocol) proposes molecular testing of CDH1 in three situations: non-syndromic clefting with a personal or family history of diffuse gastric cancer or lobular breast cancer, suspected blepharocheilodontic syndrome and, in clefts without a history of cancer, as part of a syndromic cleft gene panel or exome or genome sequencing [15]. Confirmation requires a class 4 (likely pathogenic) or class 5 (pathogenic) variant. As most variants found in these forms are rare missense variants, classification should be carried out with a laboratory or team experienced with the gene and may require functional assays [15]. In families in which the variant is already known, the PNDS proposes targeted testing of relatives from the age of majority, because confirmation does not change management in childhood, which is more restrictive than the IGCLC's openness to testing at age 16 years when onset in the family is early [2, 15]. Identification of the variant in a child with clefting or blepharocheilodontic syndrome creates the dilemma of diagnosing a cancer predisposition with no direct benefit at that age, to be weighed against the benefit for asymptomatic carrier relatives; without a family history or with a de novo variant, the cancer risk cannot be estimated reliably [15].

Source: HAS PNDS 2021, sections 2.4 and 2.6 [15].

6. Differential Diagnosis

An estimated 5–10% of gastric cancers show familial aggregation, but only a small fraction has an identified genetic cause [1]. HDGC identification is complicated by three factors: several syndromes share gastric and breast cancer, hereditary risk may not be apparent in a young patient with an unknown family history, and detailed histological classification (diffuse vs non-diffuse; lobular vs non-lobular) is indispensable [1]. The table integrates the GeneReviews differential diagnosis with textbook and ESMO guideline data.

The combination of early diffuse gastric cancer and cleft lip or palate points towards CDH1, whereas gastric polyps, mucocutaneous pigmentation, sarcomas or endometrial tumours suggest other syndromes. In a family with gastric and breast cancer without histology, the differential diagnosis includes CDH1, BRCA2, PALB2 and TP53: families clinically classified as HDGC may have BRCA2 variants with different clinical implications [2, 9]. In HDGC families without a CDH1 variant, a study of 144 probands found candidate variants in 11%, in CTNNA1, BRCA2, STK11, SDHB, PRSS1, ATM, MSR1 and PALB2 [9].

7. Treatments and Risk Management Strategies

7.1. General principles

No drug prevents diffuse gastric cancer and no medicine is authorised for the syndrome; risk management is surgical and endoscopic, and treatment of established cancer follows standards for sporadic gastric cancer and lobular cancer [1, 39]. IGCLC recommendations are based on expert opinion and observational studies, with evidence quality graded as low to moderate using GRADE, because there are no specific randomised trials [2]. Decisions are shared, with a multidisciplinary team including clinical and molecular genetics, gastroenterology, oesophagogastric and breast surgery, histopathology, nutrition, psychology, oncology, pharmacy and, for women, obstetrics and maternal medicine [1, 2, 26]. After molecular diagnosis, referral to a reference centre is recommended for endoscopy with multiple biopsies, H. pylori testing, breast assessment in women, assessment by a cleft team where applicable and genetic counselling [1].

7.2. Gastric management by risk category

Sources: Blair 2020 (text, Panel 3 and Figure 1); GeneReviews (Targeted treatment, Tables 7a and 7b); Lerner 2025 [1, 2, 17].

According to GeneReviews, surveillance in carriers begins in early adulthood or 5–10 years before the earliest gastric cancer in the family, with a minimum age of 18 years; Table 7a of the same chapter states “from age 40 years (or 5–10 years before the earliest diagnosis)”, which does not agree with the summary and initial assessment table, which refer to early adulthood [1]. Read together with the 2020 guideline, which recommends PTG between ages 20 and 30 years and annual surveillance for those deferring it, the applicable indication for carriers is surveillance from early adulthood [2].

Divergence between the IGCLC and the NCCN panel members’ position reflects evolving data. The 2020 guideline retained PTG as the cornerstone, but expressed growing confidence in surveillance at specialist centres for those wishing to defer surgery or with poorly defined risk [2]. NCCN panel members argue that T1a foci are almost universal (87.5–97% of PTG specimens), that >T1a lesions are found in only 2–3% of specimens, that advanced cancer risk is 7–10%, and that the reviewed surveillance series reported no stage III/IV cancers or gastric cancer deaths (there were T3N0, stage II cases), acknowledging short follow-up, patient selection and conduct in specialist centres [17]. A meta-analysis of 21 studies with 967 carriers, published in September 2026, showed that 62.9% underwent PTG, specimens contained >pT1a cancer in 1.1% and ≤pT1a in 79%, and the pooled proportion undergoing PTG fell from 83.2% (2001–2010) to 71.0% (2011–2020) and 67.5% (2021–2025), without a statistically significant trend (p=0.12) [70].

7.3. Prophylactic total gastrectomy

Selection and preparation. The decision should be considered and involve the patient, family and care coordinators, with discussion of risks, long-term consequences and, ideally, the surgeon’s or institution’s outcomes; preoperative psychological counselling and contact with previously operated patients are recommended [2]. Untreated addictions (food, drugs, alcohol, tobacco) should be addressed beforehand, and PTG should be avoided, if possible, in patients with severe eating disorders or refractory psychiatric illnesses [2]. Baseline endoscopy is mandatory before surgery, to exclude visible cancer requiring staging and identify conditions such as Barrett’s oesophagus that alter the proximal resection limit; H. pylori should be sought and eradicated [1, 2]. Signet ring cell carcinoma is detected in 40–61% of carriers, often at the initial endoscopy [1]. PTG is not recommended before full physical maturity and should be avoided in adolescents; it is not usually recommended after age 70 years [1]. A Markov model cited by the NCCN panel estimated that quality-adjusted life-years would be maximised with PTG at age 39 years in men and 30 in women, but relied on probably overestimated mortality rates [17].

Technique. PTG should only be performed by surgeons at institutions with transparent outcome data and a demonstrated ability to prevent and treat complications; access (open, laparoscopic or robotic) matters less than experience [2]. Gastrectomy must be total, with intraoperative confirmation of oesophageal squamous mucosa at the proximal margin and duodenal mucosa at the distal margin; as perigastric lymph node metastases are exceptional in true PTGs, extended D2 lymphadenectomy is unnecessary and discouraged, while perigastric D1 dissection is reasonable to avoid understaging a rare T2 tumour [2]. Reconstruction is usually Roux-en-Y, with biliopancreatic diversion 40–50 cm downstream of the oesophageal anastomosis to prevent bile reflux, closure of mesenteric defects to reduce internal hernias, and the surgeon’s choice regarding a jejunal pouch or anastomotic technique; prophylactic cholecystectomy is not routinely recommended (absolute post-gastrectomy gallstone risk 3.5–12%, with an estimated more than 32 cholecystectomies required to prevent one symptomatic cholecystitis) [2].

Outcomes. PTG mortality in a healthy young adult, in experienced hands, is below 1%, and morbidity, in terms of sequelae, is 100%: rapid intestinal transit, dumping syndrome, diarrhoea, altered eating habits and weight loss, with increased risk of malabsorption, malnutrition, nephrolithiasis, cholelithiasis and osteopenia/osteoporosis [1]. Expected weight loss is approximately 15–20%, stabilising within 3–6 months according to the IGCLC; the 2015 guideline reported a median loss of 10 kg after 1 year [2, 32].

Sources: Lerner 2025; Gallanis 2024; Gallanis 2025; Salehi 2026 [17, 71–73].

Quality-of-life surveys show mixed findings: one prospective study found 8% regret at 24 months despite frequent symptoms, another found 43% (30/70) expressed some degree of regret, associated with complications and absence of carcinoma in the specimen, and in a multicentre cohort overall quality of life (70.6) exceeded population reference values, with symptoms such as anxiety, pain, altered taste, dyspnoea and diarrhoea predicting poorer quality of life [17, 74]. The IGCLC reports that most patients operated on in specialist centres express little or no regret [2].

7.4. Endoscopic surveillance: Cambridge and Bethesda protocols

Objective. Given that the probability of at least one signet ring cell carcinoma focus in a carrier’s gastrectomy is 95% and superficial foci may remain indolent for years, surveillance aims not to detect all T1a foci, but to exclude deeply infiltrative lesions, detect large or numerous T1a lesions and identify histological or endoscopic changes suggesting more aggressive behaviour [2]. Patients should be informed that surveillance may delay diagnosis and it is unknown whether it improves life expectancy, and all programmes should be audited and, ideally, incorporated into prospective trials [2].

Cambridge protocol (IGCLC). In the 2015 version, endoscopy used a high-definition white-light endoscope in a dedicated session lasting at least 30 minutes, with repeated insufflation and deflation, washing with a mucolytic (N-acetylcysteine) and antifoaming agent (simeticone), photographic or video documentation, biopsy of all visible lesions (including pale areas) and five random biopsies from each of six zones (prepyloric, antrum, transitional zone, body, fundus and cardia), a minimum of 30, preferably with spiked forceps; chromoendoscopy with Congo red and methylene blue was no longer recommended because of toxicity concerns [32]. The 2020 revision retained the ≥30-minute session, preferred propofol sedation, recommended assessing distensibility first (a rigid, non-distensible stomach requires deep biopsies and computed tomography with endoscopic ultrasonography), repeating endoscopy in 4–6 weeks if an infiltrative lesion is suspected with negative histology, using contrast-enhancement techniques (narrow-band imaging) and inspecting and biopsying islands of heterotopic gastric mucosa in the oesophagus (inlet patches) [2]. As body foci were the most frequently missed, the number of random biopsies changed to 28–30: 3–5 from the cardia, 5 from the fundus, 10 from the body, 5 from the transitional zone and 5 from the antrum [2]. Centres with demonstrated expertise may limit random biopsies during follow-up when baseline biopsies are negative, because increasing biopsy numbers produces scars that impair lesion recognition; Fujita and colleagues estimated that 1768 random biopsies per patient would be required for a 90% probability of sampling a focus [2].

Bethesda protocol. The NIH protocol consists of assessment of 22 photographed anatomical sites, with four non-targeted biopsies from each site and biopsy of any abnormality [1]. In a retrospective study of 120 patients, endoscopic detection of signet ring cell carcinoma was 15% (4/26) with Cambridge and 36% (40/109) with Bethesda, and endoscopic false negatives, compared with the gastrectomy specimen, were 80% (12/15) and 37.7% (17/45), respectively [75].

Findings that change management. Erosive lesions, altered vascular or crypt patterns and histological signs of invasion of the muscularis mucosae or beyond require staging [2]. If a signet ring cell carcinoma focus is found without these indicators, age and comorbidities may justify deferring PTG following multidisciplinary discussion, but with six-monthly endoscopy [2]. In the GASTRIC consortium (390 carriers from 235 families at 12 academic centres, published in 2026), thickened folds, nodularity, masses and intestinal metaplasia were associated with signet ring cell carcinoma on biopsies; of 196 gastrectomies, 11 (5.6%) had advanced cancer and 10 of these had endoscopic abnormalities; carcinoma detection on baseline endoscopy had sensitivity of 0.81 and specificity of 0.74 for advanced disease, with a positive predictive value of 0.21, while masses and thickened folds had specificities of 0.99 and 0.96; negative predictive values were 0.94–1.0 [76]. The Cambridge endoscopic criteria had sensitivity of 82.8% and specificity of 78.2% in 537 endoscopies, and a clinical decision tree, prospectively validated in 66 carriers, had sensitivity of 77.8% and specificity of 90.7%, improving expert and non-expert performance [77].

Sources: Lim 2014; Mi 2018; Jacobs 2019; van Dieren 2020; Lee 2023; Asif 2023; Benusiglio 2026; Wu 2026; Li 2026; Schoeneich 2026 [37, 38, 56, 78–84].

These studies are not comparable with one another or with PTG: they differ in selection (families meeting criteria or panel-identified carriers), protocol (number of biopsies, Cambridge or Bethesda), definition of positivity (T1a foci on biopsy, which are almost universal) and follow-up duration, almost always less than 5 years against a risk spread over decades. The proportion of patients with advanced cancer under surveillance is low, but not zero (two pT3N0 at the NIH, one T3N0 after 9.4 years in another series, two >T1a in Massachusetts, one T2 elsewhere), and absence of deaths in small, recent series does not equate to zero lifetime risk [17, 37, 83, 84]. In 90 first-degree relatives from 40 “HDGC-like” families screened at two Dutch centres, pT1a foci were found in 4 people from one family, subsequently explained by a CTNNA1 variant, and an advanced poorly cohesive carcinoma in another family (approximately 6% overall), as well as intestinal metaplasia in 42% and mucosal scarring in 44%, which may impair detection of typical pale lesions [85].

Probe-based confocal laser endomicroscopy (pCLE) remains investigational: criteria defined off-line had sensitivity of 80.9% and specificity of 70.6% for early signet ring cell carcinoma, and an NIH phase 2 confocal endomicroscopy trial (37 enrolled, 36 analysed) recorded detection of 16.7% versus 11.1% with white light, with results published only in the registry [86, 87]. The IGCLC recommends use only in research protocols [2].

7.5. Breast risk management

Sources: Blair 2020 (Panel 2 and text); GeneReviews (Table 7a, Table 6); ESMO 2023; Oxford Desk Reference; Corso 2018 [1, 2, 18, 66, 88].

The IGCLC recommends annual breast surveillance even in female CDH1 carriers without a family history of gastric or breast cancer, whereas the GeneReviews surveillance table restricts it to carriers with a personal or family history of breast cancer of any type [1, 2]. Mammography has low sensitivity for lobular cancer, whose subtle pattern includes spiculated masses, architectural distortion and ill-defined asymmetric densities; MRI better assesses margins, size and multifocality [1]. The IGCLC asks that discussion of surveillance versus mastectomy address limited knowledge about breast cancer in the syndrome, the lack of prospective screening imaging data, breast density and background enhancement on MRI, the woman’s tolerance of MRI, what happens if cancer is detected, chemoprevention, information on gadolinium and screening harms such as additional recalls [2]. Modifiable risk factors (alcohol, exercise, weight) should be discussed [2].

A 2026 scoping review identified 11 studies with 234 female CDH1 carriers, of whom 75 (32.1%) underwent prophylactic bilateral or contralateral risk-reducing mastectomy; 81.3% of these operations were reported in North America and 18.7% in Europe, 81.8% of women had a family history of breast cancer and almost all specimens had at least one lobular carcinoma in situ focus [89]. The 2023 ESMO guideline recommends intensive surveillance from age 30 years or 5 years before the youngest case and discussion of mastectomy following individualised assessment; validated risk models such as CanRisk may support the decision [18].

When lobular cancer develops, treatment follows standard practice (surgery, endocrine therapy, perioperative and/or adjuvant chemotherapy according to stage and biomarkers) [1, 2]. There is a divergence here: GeneReviews identifies mastectomy as the preferred treatment, whereas the IGCLC accepts breast-conserving surgery if the woman chooses it, recommending that bilateral risk-reducing mastectomy also be considered, and allows skin- and nipple-sparing mastectomy with immediate reconstruction if margins are adequate [1, 2]. Lobular carcinoma in situ, usually an incidental finding, does not require risk-reducing mastectomy, but this should be discussed alongside surveillance and chemoprevention [2]. In a carrier with invasive carcinoma of no special type, without a history of lobular or diffuse gastric cancer, once misclassification has been excluded, the tumour is probably sporadic and breast-conserving surgery is acceptable with subsequent surveillance [2].

7.6. Helicobacter pylori

The association between H. pylori and HDGC has not been demonstrated, but the IGCLC recommends testing for and eradicating infection, especially in carriers choosing surveillance, because it is a WHO class 1 carcinogen; rapid urease testing is preferred at baseline endoscopy, with antral and body biopsies because of patchy colonisation [2, 32]. Indirect evidence comes from a Korean randomised trial in 1,838 first-degree relatives of gastric cancer patients (not selected by CDH1), in which the eradication treatment used in the trial (lansoprazole 30 mg, amoxicillin 1000 mg and clarithromycin 500 mg, twice daily for 7 days) reduced gastric cancer incidence from 2.7% to 1.2% over a median of 9.2 years (hazard ratio 0.45; 95% CI 0.21–0.94), with more frequent mild adverse events (53.0% versus 19.1%) [90]. This is the trial regimen, not a recommendation for CDH1 carriers, and regimen selection should follow local antibiotic resistance guidance. In the labels, the Portuguese SmPC of lansoprazole (Ogasto®) indicates 30 mg twice daily for 7 days (usually 7, sometimes up to 14 days, according to local guidance) with clarithromycin 250–500 mg and amoxicillin 1 g, both twice daily, or with clarithromycin 250 mg and metronidazole 400–500 mg twice daily, which covers the trial regimen [91]; the US label (Prevacid®) approves the same triple therapy with lansoprazole 30 mg, amoxicillin 1 g and clarithromycin 500 mg twice daily, but for 10 or 14 days, so that the 7-day duration falls outside the US label conditions [92]. In the Portuguese family with c.1901C>T described by Gullo and colleagues, all relatives had H. pylori gastritis, with CagA-positive strains in all specimens except that of the proband’s sister [93].

7.7. Care after total gastrectomy

The LAP-TG Delphi consensus (2022), developed with clinicians, healthcare professionals and patient representatives, acknowledges very low to low evidence for almost all statements [22].

Source: Roberts 2022, Table 3 and statements 1–7 [22]. The recommendations apply to adults after total gastrectomy. The article reports that the oral B12 dose in use is 1 mg/day (with no maximum dose defined in the source), with efficacy demonstrated in a Portuguese prospective study, and that high-dose vitamin D supplementation normalised vitamin D and PTH in a study of gastric cancer survivors [22]. The oral dose coincides with the Portuguese SmPC of oral cyanocobalamin (Permadoze®, 1 mg tablets), which is indicated for vitamin B12 malabsorption after gastrectomy: 2 tablets twice daily until remission and 1 tablet daily for maintenance or prophylaxis, with meals; the same SmPC considers the parenteral route preferable in overt deficiency, because of faster remission [94]. For the parenteral route, the US label of injectable cyanocobalamin indicates, in pernicious anaemia (an analogous malabsorption condition), 100 µg daily intramuscularly or by deep subcutaneous injection for 6–7 days, then on alternate days (7 doses) and every 3–4 days for 2–3 weeks, followed by 100 µg monthly for life, and advises against the intravenous route [95]. The multivitamin, calcium and vitamin D are mentioned as supplements; the doses in the table are those of the guideline [22].

IGCLC Table 1 adds guidance for complications: early (15–30 minutes) and late (1.5–3 hours) dumping with small, well-chewed meals without liquids, 6–8 meals daily, low in sugar and high in protein; lactose intolerance; fat malabsorption (low-fat diet, pancreatic enzymes, fat-soluble vitamins); bacterial overgrowth (antibiotics, surgery); anastomotic stricture (endoscopic dilatation); osteopenia/osteoporosis (baseline densitometry and every 2–5 years, divided calcium citrate, vitamin D, weight-bearing exercise, intravenous bisphosphonate if osteoporosis); cholelithiasis; bile reflux (adequate Roux-en-Y limb length, head of bed elevated >30°, fasting 2–3 hours before bedtime, bile acid sequestrants and sucralfate); persistent nausea (assess and replace thiamine, orodispersible ondansetron if necessary); and early satiety [2]. Patients should know the symptoms of late internal hernia, an urgent complication that can occur at any time [2].

Medicines after gastrectomy. The IGCLC recommends informing all patients about altered absorption; replacing solid oral formulations with chewable, dispersible or liquid formulations (beware of sugar in liquids, which may precipitate dumping); considering sublingual, topical, vaginal, rectal or parenteral routes; crushing tablets or opening capsules when there is no alternative and it is safe; avoiding prolonged- or delayed-release formulations; avoiding oral contraceptives (use an implant or intrauterine device); avoiding mucosal irritants (non-steroidal anti-inflammatory drugs, corticosteroids, oral bisphosphonates, acetylsalicylic acid, doxycycline and some antibiotics, potassium chloride) and drugs favouring gallstones (gemfibrozil); seeking alternatives to drugs requiring an acidic environment (azole antifungals, carbamazepine, phenytoin, selegiline); separating iron and calcium by 4–5 hours; exercising caution with alcohol and narrow-therapeutic-index drugs [2].

7.8. Treatment of established cancer

In early diffuse gastric cancer, total gastrectomy includes D2 dissection, Roux-en-Y oesophagojejunostomy and proximal margins ensuring removal of all gastric mucosa; in advanced disease, perioperative and/or adjuvant therapy (chemotherapy, radiotherapy) is added, as decided by an experienced team [1]. In children and adolescents with pT1a lesions detected during surveillance, chemotherapy and radiotherapy are not indicated: surveillance is intensified (interval from 12 to 6 months) until gastrectomy is possible at physical maturity; systemic therapy is reserved for advanced disease [1]. Survival and disease-free survival in advanced CDH1- or CTNNA1-associated diffuse gastric cancer are considered similar to sporadic disease, and targeted therapies and response predictors specific to diffuse gastric or lobular cancer are scarce [1].

Biomarkers for advanced gastro-oesophageal adenocarcinoma are HER2, mismatch repair deficiency/microsatellite instability, PD-L1 and claudin 18.2 (CLDN18.2), with parallel immunohistochemical tests to avoid delays; the genomically stable subtype, enriched for diffuse histology, is associated with poorer chemotherapy response [54]. GeneReviews notes that anti-HER2, anti-VEGFR2 and anti-PD-1 therapies are approved for gastric adenocarcinoma, but with weak data for the diffuse type, and describes zolbetuximab, a monoclonal antibody against CLDN18.2: in the phase 2 MONO trial (NCT01197885), 22 of 54 patients had advanced diffuse gastric cancer with CLDN18.2 expression in ≥50% of cells, the clinical benefit rate was 23%, and whether benefiting patients had diffuse histology was not reported [1].

Dosage (identical in the FDA label and European SmPC): 800 mg/m² intravenously for the first dose, followed by 600 mg/m² every 3 weeks or 400 mg/m² every 2 weeks, in combination with chemotherapy, until progression or unacceptable toxicity, in adults; the dose is calculated by body surface area, with no absolute maximum dose defined in the label; dose reduction is not recommended; antiemetic premedication before each infusion; administration by intravenous infusion only [19, 20]. Adverse effects: hypersensitivity reactions in 18% (severe in 2%, including anaphylaxis), infusion reactions, nausea (82% with mFOLFOX6 and 69% with CAPOX) and vomiting (67% and 66%), fatigue, anorexia, diarrhoea, peripheral sensory neuropathy, abdominal pain, weight loss and fever; monitoring during and for up to 2 hours after infusion; do not breastfeed during treatment or for 8 months afterwards; paediatric safety not established [19]. Grade 2 reaction management involves interrupting infusion until grade ≤1 and resuming at a reduced rate; grade 3–4 reactions or anaphylaxis require permanent discontinuation [19].

Source: FDA Vyloy® label [19]. The trials did not stratify by diffuse histology or germline CDH1 status; the absolute median overall survival benefit was 2.2–2.7 months, without an increase in objective response rate, and any extrapolation to patients with HDGC is indirect. The 2022 ESMO guideline contains no specific recommendations for gastric cancer in germline variant carriers beyond referral to genetics [23].

7.9. Ongoing studies, recent developments and hierarchy of evidence

Sources: ClinicalTrials.gov records accessed on 7 October 2026 [87, 96–101].

Results from NCT04535414 and NCT03648879 are registry data not published in a peer-reviewed journal and do not permit conclusions: the randomised trial’s primary endpoint depended on comparison with the gastrectomy specimen and could only be assessed in two participants, and the difference in crude detection, without specimen confirmation, demonstrates neither greater sensitivity nor clinical benefit [97]. There are no randomised trials of PTG versus surveillance.

Over the past 12 months, evidence has accumulated at four distinct levels. In published peer-reviewed human studies: the GASTRIC consortium defined endoscopic predictors of advanced disease; the Cambridge cohort showed stability of the positive-biopsy burden over time; series from Massachusetts, Paris and Australia reported surveillance without gastric cancer deaths, but with occasional >T1a cancers; a meta-analysis documented a non-significant fall in the proportion undergoing PTG; and the study of 351 families redefined CTNNA1 as a moderate-penetrance gene [13, 38, 70, 73, 76, 82, 83]. In preclinical research: the Cd44-Cdh1 mouse model and the Drosophila model for CTNNA1 variants [13, 42]. In curation: ClinGen classified the association of CTNNA1 with hereditary macular dystrophy as definitive in September 2026 [5]. No syndrome-specific authorisation applications, opinions or regulatory approvals were identified in the sources consulted; the only recent regulatory change identified with relevance to advanced diffuse gastric cancer was the zolbetuximab EPAR update in September 2026 [20]. The IGCLC guideline lists chemoprevention, synthetic lethality and gene replacement among research priorities, all still without clinical application [2].

7.10. CDH1-associated developmental anomalies: treatment and cancer risk management (HAS PNDS 2021)

There is no aetiological treatment for CDH1-related developmental anomalies [15]. Management is multidisciplinary, coordinated by a clinical geneticist, preferably trained in oncogenetics, with the paediatrician or attending physician maintaining routine follow-up, and includes primary treatment of the cleft and its aesthetic and functional consequences by a team of the MAFACE network (TeteCou network), otorhinolaryngological follow-up of hearing, speech and language therapy and dental and maxillofacial follow-up of dental agenesis [15]. In blepharocheilodontic syndrome, management also includes treatment of congenital hypothyroidism by paediatric endocrinology, ophthalmological treatment of eyelid malformations, which may be surgical, with prevention of exposure keratitis, correction of anorectal malformation by paediatric digestive surgery and neurosurgical and paediatric neurological treatment of neural tube defects; in the form caused by CTNND1, the recommended management is the same, without the oncological component [15]. The PNDS (French National Diagnosis and Care Protocol) does not specify drugs with their own doses; a cancer of the spectrum is treated by a specialist team according to the standards of the HAS and the Institut National du Cancer (INCa) [15].

Cancer risk management depends on the family history [15]. With a personal or family history of cancer, the oncogeneticist first confirms that the cancers in the family belong to the CDH1 spectrum; if they do, the patient is managed as an at-risk carrier, with testing offered to adult relatives. The PNDS cites cumulative incidences of gastric cancer at 80 years of 42% (95% CI 30–56%) in men and 33% (21–43%) in women, rising to 64% and 47% in families with three or more gastric cancers (the PNDS states "more than three"), and agrees with the IGCLC on prophylactic total gastrectomy between ages 20 and 30 years and on annual endoscopy according to the Cambridge protocol, in a specialist centre, from age 18 years, while surgery is deferred [2, 10, 15]. Gastric and breast decisions are taken at a multidisciplinary meeting with a geneticist, oncogeneticist, molecular biologist experienced with the gene, gastroenterologist, digestive surgeon, gynaecologist with oncological training, breast surgeon and oncologist [15]. The inclusion of colorectal cancer in the spectrum remains controversial, because cohorts show no excess risk relative to the general population, which is consistent with the absence of additional colorectal screening in the IGCLC [2, 15].

Without a history of cancer, or with a de novo variant, the risk cannot be estimated precisely; in practice the question arises mainly in blepharocheilodontic syndrome, because CDH1-associated non-syndromic clefts are generally identified from a family history of cancer [15]. The PNDS sets out the dilemma between proposing a gastrectomy, with a mortality of about 1%, short- and long-term adverse effects and an impact on quality of life, to a family without cancer, and failing to prevent a cancer with high mortality; it considers endoscopy according to the Cambridge protocol, in a reference centre and with data recording, appropriate in this situation, and recommends discussing and offering breast surveillance, which is less invasive [15]. The mortality figure is higher than the figure below 1% in experienced hands given by GeneReviews (section 7.3) [1, 15].

Sources: HAS PNDS 2021 (sections 2.4, 2.7, 3.4); Blair 2020; GeneReviews; Lerner 2025 [1, 2, 15, 17].

The main divergences concern the breast and the age of testing. The PNDS follows the French high-risk recommendations, with mammography combined with MRI from age 30 years and transition to mammography alone at 65 years, whereas the IGCLC defers mammography to age 40 years (35–40 case by case), does not recommend it before 35 years without suspicious findings and does not define an end age for MRI, in a context in which mammography has low sensitivity for lobular cancer (section 7.5) [1, 2, 15]. In the same PNDS, the initial assessment refers to breast management "from age 25–30 years" and the surveillance section sets the start at 30 years, except in situations discussed at a meeting; this article adopts 30 years, in line with the IGCLC, ESMO and GeneReviews [1, 2, 15, 18]. As regards testing of minors, the PNDS is more restrictive than the IGCLC, and the choice of age should take into account the age of onset in the family, because cancers before age 18 years have been described (section 8.2) [2, 15].

8. Reproductive Options and Family Counselling

8.1. Inheritance and risks to relatives

The syndrome is autosomal dominant. Most patients with the CDH1-associated form inherited the variant from a parent who, because of reduced penetrance, may not have had cancer; some patients with the CTNNA1 form have a first- or second-degree relative with gastric or breast cancer; de novo CDH1 variants have been reported [1]. If the proband appears to be the only affected individual, testing the parents for the identified variant is recommended, both to refer them and to estimate sibling risk; if neither parent has the variant and parentage is confirmed, a de novo variant or parental gonadal (or somatic and gonadal) mosaicism should be considered, which leucocyte testing may not detect [1]. Siblings have a 50% risk if one parent is a carrier; if the variant is not detected in the parents, risk is slightly above population risk; absence of disease in a parent of unknown genotype does not predict sibling risk [1]. Each child of a carrier has a 50% probability of inheriting the variant, and age at onset and manifestations may differ between siblings with the same variant [1]. In developmental anomalies, the French PNDS (French National Diagnosis and Care Protocol) estimates the recurrence risk at about 1% for a couple whose child has a de novo variant, corresponding to the theoretical risk of germline mosaicism, and emphasises that intrafamilial variability and possible non-penetrance make it impossible to predict the severity of malformations in a carrier child [15].

In families without an identified variant, predictive testing is not possible; first-degree relatives receive surveillance tailored to personal and family history, the search for the genetic cause should continue in the affected proband and DNA banking should be considered [1]. For familial aggregation of gastric cancer without a defined syndrome, the Oxford Desk Reference reports a 2- to 3-fold greater risk in patients’ relatives, 1 in 8 to 1 in 10 with two affected relatives and 1 in 3 with three affected relatives, empirical figures that do not apply to known variant carriers [26].

8.2. Predictive testing and family communication

Predictive testing should take place within formal genetic counselling, discussing socioeconomic consequences, the need for long-term follow-up, and testing capabilities and limitations [1]. It should be offered from the legal age of consent (usually 16–18 years), with testing of younger relatives considered according to family history; the IGCLC accepted that testing at age 16 years may be considered when onset in the family is early, and syndrome-related cancers have been reported before age 18 years [1, 2]. The 2015 guideline noted that the risk of clinically significant disease before age 20 years is low, although cases occur before 18 [32]. Parental requests to test asymptomatic minors should be received sensitively, with comprehensive counselling for parents and child [1]. In the Portuguese family with c.1901C>T, the proband had metastatic diffuse gastric cancer at age 18 years and his 14-year-old sister underwent PTG, which showed intramucosal foci [93].

Counselling should include a three-generation pedigree, assessment for cleft lip or palate and histological confirmation of cancers and precursor lesions, attention to psychosocial needs and support in communicating the diagnosis to at-risk relatives, with a communication plan and, when useful, an extended family meeting [2]. The 2015 guideline identified six areas of psychosocial difficulty: coping with risk, practical problems (insurance, loans, employment), family problems (communication, sense of responsibility), concerns about children, living with cancer and emotions (anxiety, anger, loss, but also relief) [32]. In carriers identified through panels without a history of diffuse gastric cancer, counselling must emphasise risk uncertainty [2].

When the variant is identified in a child with clefting or blepharocheilodontic syndrome, the French PNDS asks that the result be discussed beforehand at a multidisciplinary meeting with the prescribing physician and with those locally responsible for gastric and breast risk management, and that it be communicated at a dedicated consultation by the geneticist who raised the suspicion or by the prescriber, preferably with a psychologist [15]. With a family history of cancer of the spectrum, the risk is explained and preventive measures are offered; in sporadic cases or without cancer in the family, it is explained that the risk has not been established, particularly for the Asp254 variants. A second consultation, which in the case of a child may take place with the parents alone, discusses the oncological aspects and the balance between benefits and risks of the measures, followed by consultations with gastroenterology and gynaecology and by psychological support; the written report is sent to the patient and to all physicians involved [15].

8.3. Reproductive options

Prenatal diagnosis and preimplantation genetic testing should be discussed and made available to CDH1 and CTNNA1 carriers, and adults of reproductive age should receive reproductive genetic counselling, ideally before pregnancy [1, 2]. Perspectives on these techniques vary among professionals and families, and the decision is personal [1].

For developmental anomalies, the French PNDS describes a dedicated pathway [15]. In each pregnancy of a carrier, a detailed anomaly scan targeting clefting, neural tube defects and digestive tract anomalies should be performed. Molecular prenatal diagnosis may be offered by chorionic villus sampling at about 12 weeks of gestation or by amniocentesis from 16 weeks, and requests for prenatal and preimplantation diagnosis are discussed at the multidisciplinary prenatal diagnosis centre (CPDPN); if the fetus is a carrier, the possibility of a mild form should be discussed. With clefting or blepharocheilodontic syndrome and a family history of cancer, predisposition to diffuse gastric cancer may be considered an admissible ground for a request for termination of pregnancy for medical reasons made by the pregnant woman, because of the severity of the cancer and the burden of preventive gastrectomy; in familial clefts without cancer, the CPDPN weighs the uncertainty of the risk, and in blepharocheilodontic syndrome without cancer it also weighs the spectrum of expression, which ranges from anencephaly to a slight widening of the palpebral fissures with oligodontia, the ultrasound findings and the family's experience of the disease [15]. This framework corresponds to French legislation and cannot be transposed to other countries without adaptation.

8.4. Pregnancy and contraception after gastrectomy

The nutritional consequences of gastrectomy should be discussed before and during pregnancy, which should be deferred for at least 6–12 months after surgery to allow weight stabilisation and nutritional recovery [1, 2]. Healthy pregnancies after PTG have been described (Kaurah 2010 reported pregnancies in three female CDH1 carriers and a non-carrier woman who underwent gastrectomy for early cancer), but, by analogy with bariatric surgery, there is a higher risk of preterm birth, growth restriction and neonatal intensive care admission, so pregnant women should be closely followed by an obstetrician and dietitian [1, 2, 102]. Vitamin A supplementation in pregnancy requires caution [22]. An intrauterine device or another method independent of gastrointestinal absorption is recommended, because oral contraceptive absorption is compromised; the LAP-TG consensus mentions reports of unplanned pregnancies on oral contraception after PTG [1, 2, 22]. Gastrectomy and bilateral mastectomy affect sexuality and body image, and including gynaecology/obstetrics and maternal medicine in the team is useful [2].

9. Special Considerations: Population Diversity and Epidemiology

9.1. Frequency

The IGCLC and the 2022 ESMO guideline estimate the population incidence of HDGC at approximately 5–10 per 100,000 births [2, 23]. Orphanet assigns a European annual incidence class of 1–9/100,000 marked as unvalidated, and the GeneGuidance database advises against using it, because this annual incidence is implausible for the entity; Orphanet point prevalence is unknown [39]. Clinically useful figures are yields in affected groups: a pathogenic CDH1 variant in approximately 7% (6.6%; 27/408) of diffuse gastric cancer patients tested at an American laboratory, up to 3% of gastric cancer patients according to the ESMO handbook, 1–3% hereditary gastric cancers across syndromes, and 0.1–0.5% of women with lobular cancer [1, 23, 54, 61]. In a laboratory population of 212,944 cancer patients, 0.06% were carriers [61]. The prevalence of CDH1-related developmental anomalies is unknown; the French PNDS (French National Diagnosis and Care Protocol) gathers 26 families with non-syndromic clefting and 19 families (about 29 patients) with blepharocheilodontic syndrome, against a frequency of about 1 in 1,000 births for cleft lip and palate as a whole, mostly multifactorial [15].

9.2. Founder populations and groups with specific data

Māori (Aotearoa New Zealand). The family in which CDH1 was discovered was Māori, one of the few populations with a higher gastric cancer incidence and predominantly diffuse type [27, 64]. In an unselected series of 94 Māori gastric cancer patients and 200 controls, 18% of all cases, 34% of diffuse cases and 67% of those diagnosed before age 45 years had a pathogenic variant; after adjustment for screening of known families, an estimated 6% of all advanced gastric cancers and 13% of advanced diffuse cancers were attributable to germline variants, with four equally frequent variants [1, 64]. Hence the testing criterion for all Māori patients with diffuse gastric cancer [2].

Newfoundland (Canada). Four families from the south-east coast share the c.2398del variant on a common haplotype, and two were initially investigated as hereditary breast cancer because of a concentration of lobular cancer; in these families, risk by age 75 years was 40% (men) and 63% (women) for clinical gastric cancer and 52% for breast cancer [47].

Portugal. In northern Portugal, c.1901C>T was found between 1999 and 2021 in nine apparently unrelated families from the Porto district, living in neighbouring parishes in the north-east of the district; analysis of 12 polymorphic markers in four families estimated the most recent common ancestor at approximately 19.6 generations (approximately 490 years; 95% CI 445–10,900 years), placing the founder in the sixteenth century [16]. Of 134 individuals tested, 58 (43.3%) were carriers; 11 had clinical diffuse gastric cancer (mean age 33 years) and 6 women had lobular cancer (mean age 50 years), and 21 of 26 risk-reducing gastrectomies contained lesions [16]. In the European genotype-first study, c.1901C>T was the most recurrent single-nucleotide variant (13 families) [24]. The authors interpret the clustering as the result of genetic drift in previously isolated rural populations, in a region with high gastric cancer incidence [16]. In small series from Portugal, South Korea and Japan, the detection rate of CDH1 variants in HDGC and early gastric cancer was 8–15% [32].

Europe and North America. The ERN GENTURIS European cohort was predominantly families with gastric cancer alone (52%) or gastric and breast cancer (38%), and only 9% were breast-cancer-only families, whereas American cohorts described breast-cancer-only families in more than 35% of carriers; the authors suggest ascertainment, environmental or modifier-gene explanations [24]. Large American cohorts are predominantly of European ancestry: 85% non-Hispanic White in the cohort of 213 families, 93% in the NIH surveillance study and 76% European ancestry among families with recorded ancestry in the 2019 laboratory cohort [10, 11, 37].

East Asia. The IGCLC testing criteria were designed for countries with a low incidence of sporadic gastric cancer; in Japan and South Korea, where incidence is high, the Japanese classification of signet ring cell carcinoma is recommended, and large carrier cohorts have not been systematically studied [2, 32]. At a Chinese centre, 3 of 121 gastric cancer patients tested (2.5%) were carriers [60]. In Lynch syndrome, gastric cancer incidence before age 40 years is higher in Asian families, which matters for the differential diagnosis [1].

Brazil and other regions. The main sources consulted contain no frequency estimates or Brazilian clinical HDGC series; the Brazilian genetic syndromes textbook includes CDH1 among high-penetrance hereditary breast cancer genes (relative risk 6.6 for lobular cancer), and the 2026 multicentre CTNNA1 study has co-authors from Barretos and the Instituto Nacional de Câncer, without country-specific data in the main text [13, 58]. No data from Africa, the Middle East or Hispanic Latin America were found in the sources consulted either, which does not demonstrate absence of studies.

10. Illustrative Clinical Case

10.1. Italian family with CDH1 c.1792C>T (Cavallaro et al., 2026)

In June 2015, a 41-year-old man was admitted to a hospital in Catania (Italy) with progressive abdominal pain over two weeks, nausea and vomiting; two sisters had died from gastric cancer at ages 18 and 23 years [103]. Ultrasonography showed ascites, endoscopy a 4 cm ulcerated lesion on the posterior gastric body wall, and computed tomography extensive circumferential thickening of the greater curvature and antrum, with peritoneal nodularity suggesting carcinomatosis. He underwent en bloc subtotal gastrectomy with right and transverse colectomy and hyperthermic intraperitoneal chemotherapy; histology showed poorly differentiated/signet ring cell adenocarcinoma throughout the wall thickness, with omental metastases (pT4aN1M1, AJCC 7th edition), and he died three weeks later from multiorgan failure secondary to sepsis. Subsequent immunohistochemistry on archival tissue showed loss of membranous E-cadherin in more than 90% of tumour cells (Article Figure 2) [103].

In October 2017, a 44-year-old brother presented with persistent heartburn; endoscopy showed a deep ulcer on the greater curvature and a rigid, non-distensible antrum, and biopsies confirmed poorly differentiated adenocarcinoma. After gastric perforation and emergency gastrectomy at another hospital, he received adjuvant FLOT chemotherapy; in February 2018, he underwent reoperation for intestinal obstruction with infiltrated omentum (“omental cake”) and died a few days later from complications and sepsis. Immunohistochemistry showed preserved expression of all four mismatch repair proteins [103].

These patients’ mother had undergone surgery for breast cancer and her brother had died from gastric cancer. Next-generation CDH1 and CTNNA1 sequencing in surviving siblings and the deceased individuals’ daughters identified the heterozygous c.1792C>T (p.Arg598Ter) variant in a surviving brother (case 3), the daughter of one deceased sister and the daughter of case 1; two sisters tested negative and several relatives were not tested; the panel showed no variants in MLH1, MSH2, MSH6, PMS2, EPCAM, APC, MUTYH, BRCA1 or BRCA2 (Article Figure 3, pedigree) [103]. The variant, with nomenclature confirmed against NM_004360.5, is classified as pathogenic by an expert panel in ClinVar and appears in the GeneReviews list of frequent variants in Māori families, illustrating independent recurrence of some variants [1, 6, 45, 47].

The two asymptomatic female carriers, aged 22 and 18 years, underwent open prophylactic total gastrectomy, with resection 3 cm above the cardia and including the first centimetre of the duodenum. Specimens showed chronic gastritis (active and microerosive in one), without significant atypia or invasive carcinoma foci in the mucosa or submucosa, with clear lymph nodes and margins; H. pylori was negative in the younger woman [103]. The third carrier, a 30-year-old man informed of the risk, showed little willingness to undergo surgery or surveillance; symptoms developed in 2020, endoscopy revealed an extensive ulcerated lesion on the lesser curvature and, after four cycles of neoadjuvant FOLFOX, partial gastrectomy with D2 lymphadenectomy in March 2021 showed poorly cohesive diffuse/signet ring cell adenocarcinoma with serosal perforation, transverse colon invasion, 23 metastatic lymph nodes (ypT4b N3b, AJCC 8th edition), poor chemotherapy response, HER2 0 and proficient mismatch repair; he died a few months later [103].

Critical appraisal. The family meets several IGCLC 2020 criteria (two or more gastric cancers with one diffuse; diffuse gastric cancer before age 50 years) and shows intrafamilial heterogeneity, with deaths at ages 18, 23, 30, 41 and 44 years [2, 103]. Three aspects warrant comment in relation to guidelines. Age 18 years at one gastrectomy falls below the 20–30-year interval suggested by the IGCLC, and GeneReviews advises against PTG before full physical maturity; the authors justified the approach by high lifetime risk, difficulty of early detection and several deaths in the family [1, 2, 103]. Absence of carcinoma in both specimens contrasts with the 87.5–97% prevalence of foci in series using extensive embedding protocols; the article does not describe the sampling protocol, and the IGCLC recommends that, when no foci are found, the report state the percentage of mucosa examined [2, 17, 103]. Finally, the carrier declining surgery and surveillance presented with symptomatic advanced disease, and the case cannot assess endoscopic surveillance, which was never undertaken. Management described in this report should not be generalised as a recommendation.

10.2. Portuguese family with CDH1 c.1901C>T (Gullo et al., 2018)

A young man from northern Portugal developed extensively invasive, metastatic diffuse gastric cancer at age 18 years, and testing identified c.1901C>T, then designated p.Ala634Val; of 30 relatives tested, 15 were carriers [93]. Six carriers underwent PTG with negative preoperative investigations, including the proband’s 14-year-old sister, the youngest patient then described undergoing PTG after negative biopsies; all specimens contained intramucosal foci (1 to 33 per specimen) with an indolent phenotype, whereas the proband’s tumour had pleomorphic cells, absent E-cadherin, a high Ki-67 index and overexpression of p53, pSrc and pStat3 [93]. The authors concluded that bizarre and pleomorphic cells in endoscopic biopsies suggest advanced disease and should prompt intervention. This description is based on the article abstract; its full text and figures were not consulted; the variant was later demonstrated to be a founder splice variant (section 9) [16].

11. Prognosis and Long-Term Follow-Up

11.1. Prognosis

When sporadic diffuse gastric cancer is detected before wall invasion, 5-year survival may exceed 75–90%, depending on region; when detected late, it is below 30% [1]. Survival in advanced CDH1- or CTNNA1-associated diffuse cancer is considered similar to the sporadic form, although one study described shorter survival in CDH1 carriers than in patients meeting criteria without a variant [1]. The 2015 guideline reported that only approximately 10% of patients with symptomatic diffuse cancer have potentially curable disease and that, even among these, 5-year survival does not exceed 30% [32]. In invasive lobular cancer, overall 5-year survival is approximately 93%, with no known impact of the germline variant [1]. In carriers undergoing PTG at a high-volume centre, there were no recurrences or gastric cancer deaths at a median of 36 months [72].

11.2. Follow-up plan

Transition to adulthood coincides with the age of consent for testing and surveillance initiation, and the PTG decision should consider family age at onset, nutritional status, family planning and psychological readiness [2, 17]. A long-term relationship with an experienced team is considered beneficial for supporting decisions between surveillance and surgery [2].

11.3. Networks and resources

In Europe, the relevant reference network is ERN GENTURIS (genetic tumour risk syndromes, genturis.eu), which coordinated the genotype-first study of 29 institutions in ten countries and on whose behalf, in 2020, a database on germline CDH1 variant pathogenicity coordinated by Carla Oliveira (i3S, Porto) was being developed [2, 24]. Among patient organisations, the IGCLC lists No Stomach For Cancer (nostomachforcancer.org), Hereditary Diffuse Gastric Cancer Advocacy (hereditarydiffusegastriccancer.org), DeGregorio Family Foundation (degregorio.org) and Lobular Breast Cancer Alliance (lobularbreastcancer.org); GeneReviews adds the National Cancer Institute, American Cancer Society, CancerCare and International Society for Gastrointestinal Hereditary Tumours (InSiGHT) [1, 2].

In France, CDH1-associated developmental anomalies are followed at the reference and competence centres for developmental anomalies and malformation syndromes of the AnDDI-Rares network, with coordination of the PNDS by the reference centre of the North-West region (Jeanne de Flandre Hospital, CHU de Lille), and cleft surgery in the teams of the MAFACE network of the TeteCou network (rare diseases of the head, neck and teeth); the PNDS also lists the BRAIN-TEAM, DéfiScience, FIMARAD, FIRENDO, NeuroSphinx and SENSGENE networks for the associated manifestations, and the Institut National du Cancer for the oncological component [15]. The protocol supports the application for 100% coverage as a long-term condition outside the list (ALD hors liste) and the certificate for the Maison Départementale des Personnes Handicapées (MDPH), and assigns to the attending physician referral to genetics, surveillance of complications in liaison with the reference teams, routine follow-up of predispositions to gastric and breast cancer and support for the family at the time of diagnosis [15]. There is no patient association specific to these forms; contact between families takes place through the AnDDI-Rares network, the Fondation Maladies Rares, the Fondation ARCAD (digestive cancers) and a French-speaking group dedicated to CDH1, focused on oncological aspects; the PNDS also lists Orphanet and the Human Disease Genes database (humandiseasegenes.nl/cdh1) [15].

12. Summary and Final Considerations

Hereditary diffuse gastric cancer, now also termed diffuse gastric and lobular breast cancer syndrome, is caused mainly by germline loss-of-function CDH1 variants and, in a minority of families, truncating CTNNA1 variants. Loss of the second allele initiates multifocal intramucosal signet ring cell carcinoma foci in the neck of gastric glands, almost universal in carriers, only a fraction of which progress to infiltrative cancer with a poor prognosis; in women, the same mechanism predisposes to lobular breast cancer.

Risk depends on how carriers were identified. In families selected using clinical criteria, cumulative gastric cancer risk by age 80 years reached 56–83%; in cohorts with panel-identified carriers it is 24–42%, and advanced gastric cancer risk was estimated at 7–10%, reaching approximately 38% with three affected first-degree relatives. Breast cancer risk in female carriers is more consistent (37–55%). For CTNNA1, 2026 data indicate moderate penetrance, with gastric cancer risk approximately seven times that of the tested population and no significant increase in lobular cancer in the laboratory cohort.

Diagnosis rests on the 2020 IGCLC criteria, which require confirmed histology and relaxed age limits, and on alternative criteria with greater sensitivity (NCCN, Yale) or greater focus on lobular cancer (ERN GENTURIS) and CTNNA1 (Porto). Testing should include sequencing and copy number analysis, with classification according to ClinGen-specific rules; missense variants without an effect on splicing do not justify high-risk management. Variants of uncertain significance and families without an identified cause are managed with surveillance and research, without prophylactic surgery.

Gastric management is changing: the IGCLC retains prophylactic total gastrectomy between ages 20 and 30 years as the recommendation for carriers from HDGC families, but accepts endoscopic surveillance in specialist centres for those deferring surgery or with poorly defined risk, and NCCN panel members, in a review underpinning version 1.2024 of the American guideline, advocate active endoscopic surveillance as a first-line alternative in the absence of concerning signs. Surveillance data from 2023–2026 are encouraging but limited by short follow-up, selection and conduct in reference centres, and advanced cancers have occurred under surveillance. Gastrectomy eliminates gastric risk with mortality below 1%, but with permanent sequelae requiring lifelong nutritional follow-up. Breast surveillance with annual MRI from age 30 years and risk-reducing mastectomy are the options for lobular risk. No drug is authorised for the syndrome; zolbetuximab, authorised for advanced CLDN18.2-positive gastric cancer, has no carrier-specific data. Portugal contributed the c.1901C>T founder variant from the north of the country and central research on the gene, and ERN GENTURIS is the European reference network.

The same gene causes developmental anomalies, from isolated cleft lip and palate to blepharocheilodontic syndrome, which the 2021 French PNDS (French National Diagnosis and Care Protocol) organises into a dedicated pathway: confirmation by a class 4 or 5 variant with an experienced laboratory, multidisciplinary management of the malformations, an oncogenetics consultation at age 18 years and cancer risk management according to the family history, with acknowledged uncertainty in families without cancer and for variants of residue Asp254, which have never been associated with cancer. The divergences from the IGCLC concentrate on mammography, combined with MRI from age 30 years in the French recommendations, and on the age of testing of minor relatives [2, 15].

13. Bibliographic References

1. Barbosa-Matos R, Córdova L, Schrader K, Oliveira C. Diffuse Gastric and Lobular Breast Cancer Syndrome. In: GeneReviews® [Internet]. Seattle (WA): University of Washington, Seattle; 1993–2026. First published 4 November 2002; comprehensive update 10 October 2024. NBK1139. PMID:20301318. Local full text accessed 7 October 2026.
2. Blair VR, McLeod M, Carneiro F, Coit DG, D'Addario JL, van Dieren JM, et al. Hereditary diffuse gastric cancer: updated clinical practice guidelines. Lancet Oncol. 2020;21(8):e386-e397. doi:10.1016/S1470-2045(20)30219-9. PMID:32758476. PMC7116190.
3. Clinical Genome Resource (ClinGen). CDH1 (HGNC:1748): gene–disease validity (Hereditary Cancer GCEP, definitive, 22/03/2024), dosage sensitivity (haploinsufficiency 3, 10/04/2020) and clinical actionability (moderate in adults, limited in children, 04/06/2019) [Internet]. Gene page. Accessed 7 October 2026.
4. Orphanet. Hereditary diffuse gastric cancer. ORPHA:26106 [Internet]. Definition, synonyms and ICD-10 mapping obtained through the Orphacodes API (record dated 2 July 2026). Disease record. Accessed 7 October 2026.
5. Clinical Genome Resource (ClinGen). CTNNA1 (HGNC:2509): gene–disease validity (diffuse gastric and lobular breast cancer syndrome, definitive, 05/07/2023; hereditary macular dystrophy, definitive, 03/09/2026) and dosage sensitivity (haploinsufficiency 3, 12/10/2022) [Internet]. Gene page. Accessed 7 October 2026.
6. VariantValidator. Validation of NM_004360.5:c.190C>T, c.1792C>T, c.1901C>T, c.2287G>T, c.2386dup, c.2398del and c.2647T>C (GRCh38/GRCh37) and MANE Select transcripts of CDH1 (NM_004360.5) and CTNNA1 (NM_001903.5) [Internet]. VariantValidator. Accessed 7 October 2026.
7. Vaessen C, Redpath K, Schulpen E, Guilford P. The Changing Landscape of Hereditary Diffuse Gastric Cancer. J Gastric Cancer. 2026;26(1):31-51. doi:10.5230/jgc.2026.26.e5. PMID:41517846. PMC12802062.
8. Lobo S, Benusiglio PR, Coulet F, Boussemart L, Golmard L, Spier I, et al. Cancer predisposition and germline CTNNA1 variants. Eur J Med Genet. 2021;64(10):104316. doi:10.1016/j.ejmg.2021.104316. PMID:34425242.
9. Hansford S, Kaurah P, Li-Chang H, Woo M, Senz J, Pinheiro H, et al. Hereditary Diffuse Gastric Cancer Syndrome: CDH1 Mutations and Beyond. JAMA Oncol. 2015;1(1):23-32. doi:10.1001/jamaoncol.2014.168. PMID:26182300.
10. Roberts ME, Ranola JMO, Marshall ML, Susswein LR, Graceffo S, Bohnert K, et al. Comparison of CDH1 Penetrance Estimates in Clinically Ascertained Families vs Families Ascertained for Multiple Gastric Cancers. JAMA Oncol. 2019;5(9):1325-1331. doi:10.1001/jamaoncol.2019.1208. PMID:31246251. PMC6604087.
11. Ryan CE, Fasaye GA, Gallanis AF, Gamble LA, McClelland PH, Duemler A, et al. Germline CDH1 Variants and Lifetime Cancer Risk. JAMA. 2024;332(9):722-729. doi:10.1001/jama.2024.10852. PMID:38873722. PMC11372503.
12. Coudert M, Drouet Y, Delhomelle H, Svrcek M, Benusiglio PR, Coulet F, et al. First estimates of diffuse gastric cancer risks for carriers of CTNNA1 germline pathogenic variants. J Med Genet. 2022;59(12):1189-1195. doi:10.1136/jmg-2022-108740. PMID:36038258.
13. Lobo S, Dias A, Pedro AM, Ferreira M, Pinto-Oliveira A, São José C, et al. Hereditary diffuse gastric cancer spectrum associated with germline CTNNA1 loss of function revealed by clinical and molecular data from 351 carrier families and over 37,000 non-carrier controls. Gut. 2026;75(5):872-885. doi:10.1136/gutjnl-2024-334601. PMID:40998418. PMC13151538.
14. Green BL, Fasaye GA, Samaranayake SG, Duemler A, Gamble LA, Davis JL. Frequent cleft lip and palate in families with pathogenic germline CDH1 variants. Front Genet. 2022;13:1012025. doi:10.3389/fgene.2022.1012025. PMID:36246616. PMC9554356.
15. Haute Autorité de Santé. Protocole National de Diagnostic et de Soins (PNDS): Anomalies du développement liées aux variants de CDH1. Saint-Denis La Plaine: HAS; 20 October 2021 (text of September 2021, Centre de référence des anomalies du développement et syndromes malformatifs de l'interrégion Nord-Ouest, filière AnDDI-Rares). [https://www.has-sante.fr/jcms/p_3293690](https://www.has-sante.fr/jcms/p_3293690).
16. Barbosa-Matos R, Leal Silva R, Garrido L, Aguiar AC, Garcia-Pelaez J, André A, et al. The CDH1 c.1901C>T Variant: A Founder Variant in the Portuguese Population with Severe Impact in mRNA Splicing. Cancers (Basel). 2021;13(17):4464. doi:10.3390/cancers13174464. PMID:34503274. PMC8430675.
17. Lerner BA, Gupta S, Burke CA, Kupfer S, Katona BW, Grady WM, et al. Advancing the Evaluation and Management of CDH1-Associated Gastric Cancer. J Natl Compr Canc Netw. 2025;23(4):e257006. doi:10.6004/jnccn.2025.7006. PMID:40203872.
18. Sessa C, Balmaña J, Bober SL, Cardoso MJ, Colombo N, Curigliano G, et al. Risk reduction and screening of cancer in hereditary breast-ovarian cancer syndromes: ESMO Clinical Practice Guideline. Ann Oncol. 2023;34(1):33-47. doi:10.1016/j.annonc.2022.10.004. PMID:36307055.
19. U.S. Food and Drug Administration. VYLOY (zolbetuximab-clzb) for injection, for intravenous use. Prescribing information (August 2026 label) and BLA 761365 approval history (initial approval on 18 October 2024; efficacy supplement on 6 June 2025) [Internet]. openFDA. Accessed 7 October 2026.
20. European Medicines Agency. Vyloy (zolbetuximab): EPAR and summary of product characteristics (CHMP opinion 25 July 2024; authorisation 19 September 2024; orphan medicine; latest update 11 September 2026) [Internet]. EPAR Vyloy. Accessed 7 October 2026.
21. ClinGen Gastric Cancer Variant Curation Expert Panel. ClinGen CDH1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines, version 3.1 (GN007) [Internet]. Criteria Specification Registry. CSpec GN007. Accessed 7 October 2026.
22. Roberts G, Benusiglio PR, Bisseling T, Coit D, Davis JL, Grimes S, et al. International Delphi consensus guidelines for follow-up after prophylactic total gastrectomy: the Life after Prophylactic Total Gastrectomy (LAP-TG) study. Gastric Cancer. 2022;25(6):1094-1104. doi:10.1007/s10120-022-01318-5. PMID:35831514. PMC9588655.
23. Lordick F, Carneiro F, Cascinu S, Fleitas T, Haustermans K, Piessen G, et al. Gastric cancer: ESMO Clinical Practice Guideline for diagnosis, treatment and follow-up. Ann Oncol. 2022;33(10):1005-1020. doi:10.1016/j.annonc.2022.07.004. PMID:35914639.
24. Garcia-Pelaez J, Barbosa-Matos R, Lobo S, Dias A, Garrido L, Castedo S, et al. Genotype-first approach to identify associations between CDH1 germline variants and cancer phenotypes: a multicentre study by the European Reference Network on Genetic Tumour Risk Syndromes. Lancet Oncol. 2023;24(1):91-106. doi:10.1016/S1470-2045(22)00643-X. PMID:36436516. PMC9810541.
25. Laurén P. The two histological main types of gastric carcinoma: diffuse and so-called intestinal-type carcinoma. An attempt at a histo-clinical classification. Acta Pathol Microbiol Scand. 1965;64:31-49. doi:10.1111/apm.1965.64.1.31. PMID:14320675.
26. Firth HV, Hurst JA. Gastric cancer. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 574–575.
27. Guilford P, Hopkins J, Harraway J, McLeod M, McLeod N, Harawira P, et al. E-cadherin germline mutations in familial gastric cancer. Nature. 1998;392(6674):402-5. doi:10.1038/32918. PMID:9537325.
28. Caldas C, Carneiro F, Lynch HT, Yokota J, Wiesner GL, Powell SM, et al. Familial gastric cancer: overview and guidelines for management. J Med Genet. 1999;36(12):873-80. PMID:10593993. PMC1734270.
29. Huntsman DG, Carneiro F, Lewis FR, MacLeod PM, Hayashi A, Monaghan KG, et al. Early gastric cancer in young, asymptomatic carriers of germ-line E-cadherin mutations. N Engl J Med. 2001;344(25):1904-9. doi:10.1056/NEJM200106213442504. PMID:11419427.
30. Pharoah PD, Guilford P, Caldas C, International Gastric Cancer Linkage Consortium. Incidence of gastric cancer and breast cancer in CDH1 (E-cadherin) mutation carriers from hereditary diffuse gastric cancer families. Gastroenterology. 2001;121(6):1348-53. doi:10.1053/gast.2001.29611. PMID:11729114.
31. Fitzgerald RC, Hardwick R, Huntsman D, Carneiro F, Guilford P, Blair V, et al. Hereditary diffuse gastric cancer: updated consensus guidelines for clinical management and directions for future research. J Med Genet. 2010;47(7):436-44. doi:10.1136/jmg.2009.074237. PMID:20591882. PMC2991043.
32. van der Post RS, Vogelaar IP, Carneiro F, Guilford P, Huntsman D, Hoogerbrugge N, et al. Hereditary diffuse gastric cancer: updated clinical guidelines with an emphasis on germline CDH1 mutation carriers. J Med Genet. 2015;52(6):361-74. doi:10.1136/jmedgenet-2015-103094. PMID:25979631. PMC4453626.
33. Majewski IJ, Kluijt I, Cats A, Scerri TS, de Jong D, Kluin RJ, et al. An α-E-catenin (CTNNA1) mutation in hereditary diffuse gastric cancer. J Pathol. 2013;229(4):621-9. doi:10.1002/path.4152. PMID:23208944.
34. Carneiro F, Huntsman DG, Smyrk TC, Owen DA, Seruca R, Pharoah P, et al. Model of the early development of diffuse gastric cancer in E-cadherin mutation carriers and its implications for patient screening. J Pathol. 2004;203(2):681-7. doi:10.1002/path.1564. PMID:15141383.
35. Oliveira C, Senz J, Kaurah P, Pinheiro H, Sanges R, Haegert A, et al. Germline CDH1 deletions in hereditary diffuse gastric cancer families. Hum Mol Genet. 2009;18(9):1545-55. doi:10.1093/hmg/ddp046. PMID:19168852. PMC2667284.
36. Xicola RM, Li S, Rodriguez N, Reinecke P, Karam R, Speare V, et al. Clinical features and cancer risk in families with pathogenic CDH1 variants irrespective of clinical criteria. J Med Genet. 2019;56(12):838-843. doi:10.1136/jmedgenet-2019-105991. PMID:31296550.
37. Asif B, Sarvestani AL, Gamble LA, Samaranayake SG, Famiglietti AL, Fasaye GA, et al. Cancer surveillance as an alternative to prophylactic total gastrectomy in hereditary diffuse gastric cancer: a prospective cohort study. Lancet Oncol. 2023;24(4):383-391. doi:10.1016/S1470-2045(23)00057-8. PMID:36990610. PMC10084814.
38. Benusiglio PR, Leenhardt R, Duros C, Sirmai L, Dardenne A, Metras J, et al. Expert endoscopic surveillance in CDH1 pathogenic variant carriers seems safe, even after positive (pT1a) biopsies. J Med Genet. 2026;63(2):132-135. doi:10.1136/jmg-2025-111222. PMID:41266136.
39. GeneGuidance. Records for the CDH1 and CTNNA1 genes and condition ORPHA:26106 (version dated 26 September 2026). User database, accessed 7 October 2026.
40. Humar B, Fukuzawa R, Blair V, Dunbier A, More H, Charlton A, et al. Destabilized adhesion in the gastric proliferative zone and c-Src kinase activation mark the development of early diffuse gastric cancer. Cancer Res. 2007;67(6):2480-9. doi:10.1158/0008-5472.CAN-06-3021. PMID:17363565.
41. Gallanis AF, Gamble LA, Oguz C, Samaranayake SG, Kedei N, Hernandez MO, et al. Spatial Analysis of Hereditary Diffuse Gastric Cancer Reveals Indolent Phenotype of Signet Ring Cell Precursors. Mol Cancer Res. 2025;23(8):699-709. doi:10.1158/1541-7786.MCR-24-1039. PMID:40192595. PMC12324967.
42. Decourtye-Espiard L, Schulpen E, McElroy K, Charlton A, van der Post RS, Godwin T, et al. E-cadherin loss in Cd44-positive gastric cells initiates diffuse gastric cancer in a murine model. Gut. 2026;75(8):1466-1479. doi:10.1136/gutjnl-2025-336182. PMID:41708310.
43. Luo X, Maciaszek JL, Thompson BA, Leong HS, Dixon K, Sousa S, et al. Optimising clinical care through CDH1-specific germline variant curation: improvement of clinical assertions and updated curation guidelines. J Med Genet. 2023;60(6):568-575. doi:10.1136/jmg-2022-108807. PMID:36600593. PMC10202836.
44. National Center for Biotechnology Information. ClinVar. VCV000012244: NM_004360.5(CDH1):c.1901C>T (p.Ala634Val), pathogenic/likely pathogenic, criteria provided, multiple submitters without conflict (last evaluation 14 January 2026) [Internet]. VCV000012244. Accessed 7 October 2026.
45. National Center for Biotechnology Information. ClinVar. VCV000012241: NM_004360.5(CDH1):c.1792C>T (p.Arg598Ter), pathogenic, reviewed by expert panel (29 August 2023) [Internet]. VCV000012241. Accessed 7 October 2026.
46. National Center for Biotechnology Information. ClinVar. VCV001319114: NM_004360.5(CDH1):c.190C>T (p.Gln64Ter), patogénica, um submissor (3 November 2021) [Internet]. VCV001319114. Accessed 7 October 2026.
47. Kaurah P, MacMillan A, Boyd N, Senz J, De Luca A, Chun N, et al. Founder and recurrent CDH1 mutations in families with hereditary diffuse gastric cancer. JAMA. 2007;297(21):2360-72. doi:10.1001/jama.297.21.2360. PMID:17545690.
48. Selvanathan A, Nixon CY, Zhu Y, Scietti L, Forneris F, Uribe LMM, et al. CDH1 Mutation Distribution and Type Suggests Genetic Differences between the Etiology of Orofacial Clefting and Gastric Cancer. Genes (Basel). 2020;11(4):391. doi:10.3390/genes11040391. PMID:32260281. PMC7231129.
49. National Center for Biotechnology Information. ClinVar. NM_001903.5(CTNNA1):c.1351C>T (p.Arg451Ter), pathogenic (variation 659776); NM_001903.5(CTNNA1):c.2023C>T (p.Gln675Ter), conflicting classifications (variation 640852) [Internet]. ClinVar 659776; ClinVar 640852. Accessed 7 October 2026.
50. Ghoumid J, Stichelbout M, Jourdain AS, Frenois F, Lejeune-Dumoulin S, Alex-Cordier MP, et al. Blepharocheilodontic syndrome is a CDH1 pathway-related disorder due to mutations in CDH1 and CTNND1. Genet Med. 2017;19(9):1013-1021. doi:10.1038/gim.2017.11. PMID:28301459.
51. Saksens NT, Krebs MP, Schoenmaker-Koller FE, Hicks W, Yu M, Shi L, et al. Mutations in CTNNA1 cause butterfly-shaped pigment dystrophy and perturbed retinal pigment epithelium integrity. Nat Genet. 2016;48(2):144-51. doi:10.1038/ng.3474. PMID:26691986. PMC4787620.
52. National Center for Biotechnology Information. ClinVar. NM_004360.5(CDH1): c.760G>A p.(Asp254Asn), conflicting classifications (variation 834043); c.760G>T p.(Asp254Tyr), likely pathogenic (variation 438813); c.768T>A p.(Asn256Lys), likely pathogenic (variation 834044); c.1320G>T p.(Lys440Asn), likely pathogenic, conditions including blepharocheilodontic syndrome 1 (variation 438812); c.2351G>A p.(Arg784His), uncertain significance (variation 220271); c.468G>C p.(Trp156Cys), likely pathogenic (variation 4282383) [Internet]. [https://www.ncbi.nlm.nih.gov/clinvar/](https://www.ncbi.nlm.nih.gov/clinvar/). Accessed 2 November 2026.
53. Collet et al. Genomics in Breast Cancer. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 82–86.
54. Alsina M, Dienstmann R. Genomics in Gastrointestinal Cancers. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 45–48.
55. Gamble LA, Rossi A, Fasaye GA, Kesserwan C, Hernandez JM, Blakely AM, et al. Association Between Hereditary Lobular Breast Cancer Due to CDH1 Variants and Gastric Cancer Risk. JAMA Surg. 2022;157(1):18-22. doi:10.1001/jamasurg.2021.5118. PMID:34643667. PMC8515254.
56. Jacobs MF, Dust H, Koeppe E, Wong S, Mulholland M, Choi EY, et al. Outcomes of Endoscopic Surveillance in Individuals With Genetic Predisposition to Hereditary Diffuse Gastric Cancer. Gastroenterology. 2019;157(1):87-96. doi:10.1053/j.gastro.2019.03.047. PMID:30935944.
57. Firth HV, Hurst JA. Patterns of cancer (Table A14). In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 846–847.
58. Ribeiro EM, coordinator. Síndrome de mama e/ou ovário hereditário (Table 8.2.1). In: Síndromes Genéticas: Reconhecer para Tratar. Fortaleza: EdUnichristus; 2021. p. 170.
59. Petridis C, Arora I, Shah V, Moss CL, Mera A, Clifford A, et al. Frequency of Pathogenic Germline Variants in CDH1, BRCA2, CHEK2, PALB2, BRCA1, and TP53 in Sporadic Lobular Breast Cancer. Cancer Epidemiol Biomarkers Prev. 2019;28(7):1162-1168. doi:10.1158/1055-9965.EPI-18-1102. PMID:31263054.
60. Wang Z, Ding Q, Xu J, Xue L, Dong L. Dental agenesis as a novel phenotypical feature associated with hereditary diffuse gastric cancer in China. J Med Genet. 2026;63(2):79-85. doi:10.1136/jmg-2025-111081. PMID:41130721.
61. Adib E, El Zarif T, Nassar AH, Akl EW, Abou Alaiwi S, Mouhieddine TH, et al. CDH1 germline variants are enriched in patients with colorectal cancer, gastric cancer, and breast cancer. Br J Cancer. 2022;126(5):797-803. doi:10.1038/s41416-021-01673-7. PMID:34949788. PMC8888603.
62. Hayashi M, Kubota N, Abe M, Ishii M, Sasaki K, Ando M, et al. Hereditary Diffuse Gastric Cancer With Poorly Differentiated Gallbladder Adenocarcinoma: A Case Report Suggesting Carcinogenesis in the Stomach and Gallbladder Caused by a Pathogenic Germline Variant in CDH1. In Vivo. 2026;40(5):3181-3190. doi:10.21873/invivo.14465. PMID:42665410. PMC13531147.
63. Horiuchi K, Ishibashi Y, Tomioka A, Narimatsu K, Tsujimoto H, Hokari R, et al. Hereditary diffuse gastric cancer with carcinogenesis from ectopic gastric mucosa in the duodenum. Clin J Gastroenterol. 2026;19(2):388-394. doi:10.1007/s12328-025-02279-9. PMID:41591723. PMC13048960.
64. Hakkaart C, Ellison-Loschmann L, Day R, Sporle A, Koea J, Harawira P, et al. Germline CDH1 mutations are a significant contributor to the high frequency of early-onset diffuse gastric cancer cases in New Zealand Māori. Fam Cancer. 2019;18(1):83-90. doi:10.1007/s10689-018-0080-8. PMID:29589180. PMC6323075.
65. Lerner BA, Xicola RM, Rodriguez NJ, Karam R, Llor X. Simplified and more sensitive criteria for identifying individuals with pathogenic CDH1 variants. J Med Genet. 2023;60(1):36-40. doi:10.1136/jmedgenet-2021-108169. PMID:35078942. PMC9661780.
66. Corso G, Figueiredo J, La Vecchia C, Veronesi P, Pravettoni G, Macis D, et al. Hereditary lobular breast cancer with an emphasis on E-cadherin genetic defect. J Med Genet. 2018;55(7):431-441. doi:10.1136/jmedgenet-2018-105337. PMID:29929997.
67. Sebai M, Tang R, Adnani Y, Fievet A, Cabaret O, Robert de Rancher MA, et al. RNA-seq Splicing Profile of the CDH1 Gene and Its Impact on the Clinical Pathogenicity Classification of CDH1 Variants: A Description of Alternative and Pathogenic Splicing Patterns. Cancers (Basel). 2025;17(20):3320. doi:10.3390/cancers17203320. PMID:41154377. PMC12562425.
68. Lee K, Krempely K, Roberts ME, Anderson MJ, Carneiro F, Chao E, et al. Specifications of the ACMG/AMP variant curation guidelines for the analysis of germline CDH1 sequence variants. Hum Mutat. 2018;39(11):1553-1568. doi:10.1002/humu.23650. PMID:30311375. PMC6188664.
69. Harrold E, Stadler ZK. Cancer Predisposition Syndromes. In: ESMO Handbook — Cancer Genomics. Lugano: ESMO Press. p. 23–24.
70. Corso G, Quadrini G, Bagnardi V, La Vecchia C. Prophylactic total gastrectomy in germline CDH1 carriers: results across 25 years from a systematic review and meta-analysis. J Med Genet. 2026. Published online ahead of print. doi:10.1136/jmg-2026-111804. PMID:42767841.
71. Gallanis AF, Gamble LA, Samaranayake SG, Lopez R, Rhodes A, Rajasimhan S, et al. Costs of Cancer Prevention: Physical and Psychosocial Sequelae of Risk-Reducing Total Gastrectomy. J Clin Oncol. 2024;42(4):421-430. doi:10.1200/JCO.23.01238. PMID:37903316. PMC10824374.
72. Gallanis AF, Bowden C, Lopez R, Gamble LA, Samaranayake SG, Payne C, et al. Lessons learned from 150 total gastrectomies for prevention of cancer. J Gastrointest Surg. 2025;29(1):101889. doi:10.1016/j.gassur.2024.101889. PMID:39547590. PMC11710962.
73. Salehi O, Tjandra D, Karp J, Boussioutas A. Endoscopic detection of signet ring cell carcinoma in CDH1 carriers: a 15-year single-centre experience. Fam Cancer. 2026;25(3):73. doi:10.1007/s10689-026-00586-9. PMID:42406170. PMC13337865.
74. Kaurah P, Talhouk A, MacMillan A, Lewis I, Chelcun-Schreiber K, Yoon SS, et al. Hereditary diffuse gastric cancer: cancer risk and the personal cost of preventive surgery. Fam Cancer. 2019;18(4):429-438. doi:10.1007/s10689-019-00133-9. PMID:31273560. PMC8164729.
75. Curtin BF, Gamble LA, Schueler SA, Ruff SM, Quezado M, Miettinen M, et al. Enhanced endoscopic detection of occult gastric cancer in carriers of pathogenic CDH1 variants. J Gastroenterol. 2021;56(2):139-146. doi:10.1007/s00535-020-01749-w. PMID:33206267. PMC9169699.
76. Gilad O, Drogan CM, Keel E, Gao G, Swallow C, Govindarajan A, et al. Factors Associated With Gastric Signet Ring Cell Carcinoma in CDH1 Pathogenic Variant Carriers: Report From the Group of Investigators Striving Toward Research in CDH1 Consortium. Clin Gastroenterol Hepatol. 2026:S1542-3565(26)00593-8. doi:10.1016/j.cgh.2026.07.036. PMID:42604658.
77. Wu L, Honing J, Wu A, Kupfer SS, Bisseling TM, van Dieren JM, et al. Clinical decision tree for optimizing endoscopic assessment of signet ring cell carcinoma in hereditary diffuse gastric cancer surveillance. Endoscopy. 2025;57(10):1118-1127. doi:10.1055/a-2634-7895. PMID:40506018.
78. Lim YC, di Pietro M, O'Donovan M, Richardson S, Debiram I, Dwerryhouse S, et al. Prospective cohort study assessing outcomes of patients from families fulfilling criteria for hereditary diffuse gastric cancer undergoing endoscopic surveillance. Gastrointest Endosc. 2014;80(1):78-87. doi:10.1016/j.gie.2013.11.040. PMID:24472763.
79. Mi EZ, Mi EZ, di Pietro M, O'Donovan M, Hardwick RH, Richardson S, et al. Comparative study of endoscopic surveillance in hereditary diffuse gastric cancer according to CDH1 mutation status. Gastrointest Endosc. 2018;87(2):408-418. doi:10.1016/j.gie.2017.06.028. PMID:28688938. PMC5780354.
80. van Dieren JM, Kodach LL, den Hartog P, van der Kolk LE, Sikorska K, van Velthuysen MF, et al. Gastroscopic surveillance with targeted biopsies compared with random biopsies in CDH1 mutation carriers. Endoscopy. 2020;52(10):839-846. doi:10.1055/a-1157-8678. PMID:32408363.
81. Lee CYC, Olivier A, Honing J, Lydon AM, Richardson S, O'Donovan M, et al. Endoscopic surveillance with systematic random biopsy for the early diagnosis of hereditary diffuse gastric cancer: a prospective 16-year longitudinal cohort study. Lancet Oncol. 2023;24(1):107-116. doi:10.1016/S1470-2045(22)00700-8. PMID:36509094.
82. Wu L, Jun Lim H, Karthik N, Samra S, O'Donovan M, O'Neill JR, et al. Predictors of early cancer burden in CDH1 pathogenic variant carriers: a UK single-centre cohort study. EClinicalMedicine. 2026;97:104036. doi:10.1016/j.eclinm.2026.104036. PMID:42433273. PMC13351748.
83. Li M, Lynch D, Dewan T, Patel K, Patel D, Chung DC. Long-term Endoscopic Surveillance in Hereditary Diffuse Gastric Cancer Syndrome. Cancer Prev Res (Phila). 2026. Published online ahead of print. doi:10.1158/1940-6207.CAPR-26-0102. PMID:42784746.
84. Schoeneich R, Patel RR, Lee BR, Disano P, Gao MT, VanDerGraaf M, et al. Care Patterns and Gastric Cancer-Related Outcomes in Patients With Pathogenic/Likely Pathogenic Germline CDH1 Variants: A 14-Year Single-Center Analysis. Am J Clin Oncol. 2026. doi:10.1097/COC.0000000000001350. PMID:42397948.
85. van der Post RS, van Dieren J, Grelack A, Hoogerbrugge N, van der Kolk LE, Snaebjornsson P, et al. Outcomes of screening gastroscopy in first-degree relatives of patients fulfilling hereditary diffuse gastric cancer criteria. Gastrointest Endosc. 2018;87(2):397-404.e2. doi:10.1016/j.gie.2017.04.016. PMID:28455161.
86. Pilonis ND, O'Donovan M, Richardson S, Fitzgerald RC, di Pietro M. Confocal endomicroscopy diagnostic criteria for early signet-ring cell carcinoma in hereditary diffuse gastric cancer. BMC Gastroenterol. 2023;23(1):176. doi:10.1186/s12876-023-02822-3. PMID:37221458. PMC10207770.
87. ClinicalTrials.gov. NCT03648879: Phase II Study Evaluating Confocal Endoscopic Microscopy for Detection of Early Stage Gastric Cancer in Subjects With Hereditary Diffuse Gastric Cancer Syndrome (National Cancer Institute); results posted in the registry [Internet]. NCT03648879. Accessed 7 October 2026.
88. Firth HV, Hurst JA. Cancer surveillance methods. In: Oxford Desk Reference: Clinical Genetics and Genomics. 2nd ed. Oxford: Oxford University Press; 2017. p. 552–554.
89. Corso G, Vecchia C, Magnoni F, Polizzi A, Bottazzoli E, Scalzi A, et al. Bilateral Prophylactic and Contralateral Risk-Reducing Mastectomy in Women With Germline CDH1 Pathogenic or Likely-Pathogenic Variants. J Surg Oncol. 2026;134(4):475-480. doi:10.1002/jso.70323. PMID:42437507. PMC13575603.
90. Choi IJ, Kim CG, Lee JY, Kim YI, Kook MC, Park B, et al. Family History of Gastric Cancer and Helicobacter pylori Treatment. N Engl J Med. 2020;382(5):427-436. doi:10.1056/NEJMoa1909666. PMID:31995688.
91. INFARMED, I.P. Ogasto 30 mg gastro-resistant capsules (lansoprazole), Laboratório Medinfar: summary of product characteristics (Infomed, med_id 6365); text approved on 21 December 2023. Infomed. Accessed 26 October 2026.
92. Takeda Pharmaceuticals America, Inc. Prevacid (lansoprazole) delayed-release capsules and Prevacid SoluTab: US prescribing information. NDA 020406/021428; version effective 8 July 2025. openFDA. Accessed 26 October 2026.
93. Gullo I, Devezas V, Baptista M, Garrido L, Castedo S, Morais R, et al. Phenotypic heterogeneity of hereditary diffuse gastric cancer: report of a family with early-onset disease. Gastrointest Endosc. 2018;87(6):1566-1575. doi:10.1016/j.gie.2018.02.008. PMID:29454568.
94. INFARMED, I.P. Permadoze Oral 1 mg film-coated tablets (cyanocobalamin), Generis Farmacêutica: summary of product characteristics (Infomed, med_id 6816); text approved on 11 March 2022. Infomed. Accessed 26 October 2026.
95. Meitheal Pharmaceuticals Inc. Cyanocobalamin injection, USP: US prescribing information. ANDA 214316; version effective 12 June 2025. DailyMed. Accessed 26 October 2026.
96. ClinicalTrials.gov. NCT03030404: Hereditary Gastric Cancer Syndromes: An Integrated Genomic and Clinicopathologic Study of the Predisposition to Gastric Cancer (National Cancer Institute) [Internet]. NCT03030404. Accessed 7 October 2026.
97. ClinicalTrials.gov. NCT04535414: Phase II Randomized Trial of Bethesda Protocol Compared to Cambridge Method for Detection of Early Stage Gastric Cancer in CDH1 Mutation Carriers (National Cancer Institute); results posted in the registry [Internet]. NCT04535414. Accessed 7 October 2026.
98. ClinicalTrials.gov. NCT05126290: CTNNA1 Familial Expansion (CAFÉ) Study (Abramson Cancer Center, Penn Medicine) [Internet]. NCT05126290. Accessed 7 October 2026.
99. ClinicalTrials.gov. NCT04253106: Liquid Biopsies (blood, Gastric Fluid) for the Personalized Management of Patients with Hereditary Diffuse Gastric Cancer: a Pilot Project (Assistance Publique – Hôpitaux de Paris) [Internet]. NCT04253106. Accessed 7 October 2026.
100. ClinicalTrials.gov. NCT03950908: Single-bite Versus Double-bite Technique for Mapping Biopsies During Endoscopic Surveillance of Hereditary Diffuse Gastric Cancer (University of Cambridge) [Internet]. NCT03950908. Accessed 7 October 2026.
101. ClinicalTrials.gov. NCT04206891: Understanding How CDH1 Germline Mutations Affect Hereditary Lobular Breast Cancer (European Institute of Oncology) [Internet]. NCT04206891. Accessed 7 October 2026.
102. Kaurah P, Fitzgerald R, Dwerryhouse S, Huntsman DG. Pregnancy after prophylactic total gastrectomy. Fam Cancer. 2010;9(3):331-4. doi:10.1007/s10689-009-9316-y. PMID:20063069.
103. Cavallaro A, Zanghì A, Di Mattia P, Graziano G, Sanfilippo F, La Via L, et al. Hereditary diffuse gastric cancer: between underdiagnosis and overtreatment: a case series. Front Surg. 2026;13:1732442. doi:10.3389/fsurg.2026.1732442. PMID:41694804. PMC12901339.