Key Points
Question
In individuals with CDH1 gene variants, what is the lifetime risk of gastric and breast cancer?
Findings
In this multicenter, retrospective cohort and modeling study that included 213 families with more than 1 family member with a CDH1 gene variant, the cumulative risk of diffuse gastric cancer by age 80 years was 7% to 10%, and the cumulative risk of breast cancer among female carriers was 37%.
Meaning
Among individuals with a CDH1 gene variant, the risk of gastric cancer was lower than previously described, while risk of breast cancer among female carriers was similar to previous estimates.
Abstract
Importance
Approximately 1% to 3% of gastric cancers and 5% of lobular breast cancers are hereditary. Loss of function CDH1 gene variants are the most common gene variants associated with hereditary diffuse gastric cancer and lobular breast cancer. Previously, the lifetime risk of gastric cancer was estimated to be approximately 25% to 83% and for breast cancer it was estimated to be approximately 39% to 55% in individuals with loss of function CDH1 gene variants.
Objective
To describe gastric and breast cancer risk estimates for individuals with CDH1 variants.
Design, Setting, and Participants
Multicenter, retrospective cohort and modeling study of 213 families from North America with a CDH1 pathogenic or likely pathogenic (P/LP) variant in 1 or more family members conducted between January 2021 and August 2022.
Main Outcomes and Measures
Hazard ratios (HRs), defined as risk in variant carriers relative to noncarriers, were estimated for each cancer type and used to calculate cumulative risks and risks per decade of life up to age 80 years.
Results
A total of 7323 individuals from 213 families were studied, including 883 with a CDH1 P/LP variant (median proband age, 53 years [IQR, 42-62]; 4% Asian; 4% Hispanic; 85% non-Hispanic White; 50% female). In individuals with a CDH1 P/LP variant, the prevalence of gastric cancer was 13.9% (123/883) and the prevalence of breast cancer among female carriers was 26.3% (144/547). The estimated HR for advanced gastric cancer was 33.5 (95% CI, 9.8-112) at age 30 years and 3.5 (95% CI, 0.4-30.3) at age 70 years. The lifetime cumulative risk of advanced gastric cancer in male and female carriers was 10.3% (95% CI, 6%-23.6%) and 6.5% (95% CI, 3.8%-15.1%), respectively. Gastric cancer risk estimates based on family history indicated that a carrier with 3 affected first-degree relatives had a penetrance of approximately 38% (95% CI, 25%-64%). The HR for breast cancer among female carriers was 5.7 (95% CI, 2.5-13.2) at age 30 years and 3.9 (95% CI, 1.1-13.7) at age 70 years. The lifetime cumulative risk of breast cancer among female carriers was 36.8% (95% CI, 25.7%-62.9%).
Conclusions and Relevance
Among families from North America with germline CDH1 P/LP variants, the cumulative risk of gastric cancer was 7% to 10%, which was lower than previously described, and the cumulative risk of breast cancer among female carriers was 37%, which was similar to prior estimates. These findings inform current management of individuals with germline CDH1 variants.
This cross-sectional study describes the risk estimates for gastric and breast cancer for individuals with CDH1 variants.
Introduction
Germline loss of function variants in the CDH1 gene (OMIM:192090) have a variant allele frequency of approximately 5/100 000 to 10/100 000 and are causally linked to the diffuse gastric and lobular breast cancer syndrome, with or without cleft lip/palate.1,2,3,4 The risk of diffuse gastric cancer has been estimated as up to 70% for male and 83% for female individuals. The estimated risk of breast cancer in females with pathogenic CDH1 variants is 39% to 52%.1,5,6,7 Risk-reducing total gastrectomy is currently recommended for CDH1 pathogenic and likely pathogenic (P/LP) variant carriers as early as age 20 years.8 Surgery for cancer prevention may be considered for individuals with high lifetime cancer risk; however, some families with CDH1 P/LP variants may have lower risk of gastric cancer.9,10 More accurate estimates of cancer risk in people with CDH1 P/LP variants are needed.
Original estimates of CDH1-associated cancer risk were established from fewer than a dozen high-risk kindreds that included the indigenous Māori people of Aotearoa New Zealand.1,6 Subsequent analyses of data from individuals undergoing genetic testing in commercial laboratories for hereditary cancer indications estimated that lifetime cancer risk ranged from 25% to 42% for gastric cancer and 42% to 55% for breast cancer.11,12 However, these studies were limited by small size and potential sampling bias. Reliable cancer risk estimates are important because they may influence clinical care including decisions to undergo risk-reducing surgery and cancer surveillance. This retrospective cohort and modeling study reports estimates of gastric and breast cancer risk due to CDH1 P/LP variants in a North American cohort.
Methods
A multi-institutional cohort study was conducted through collaborative data transfer agreements between partnering institutions. The organizing site (National Cancer Institute) analyzed data as part of its prospective cohort study of hereditary gastric cancers, which was approved by the National Institutes of Health institutional review board. Each participating institution received written informed consent from patients as part of their institution’s research. Between January 2021 and August 2022, each institution provided data from pedigrees that were derived by certified genetic counselors from 1 or more family members who presented with a personal and/or family history suggestive of hereditary cancer susceptibility. Twelve families were associated with more than 1 of the study institutions, which was accounted for in the analysis. Genotyping was conducted at clinical diagnostic laboratories that adhere to the American College of Medical Genetics and Genomics/Association for Molecular Pathology variant interpretation recommendations and the ClinGen tailored curation guidelines for CDH1.13,14,15 All sites followed standard CDH1 genetic testing criteria, such that genotyping was performed with single-site analysis if a CDH1 variant existed in the family, single CDH1 gene analysis was performed if the personal and family history suggested hereditary diffuse gastric cancer, or multigene panel analysis if the personal and/or family history overlapped with more than 1 syndrome.8,16 The study followed reporting guidelines for cohort studies (Strengthening the Reporting of Observational Studies in Epidemiology) and genetic risk prediction studies (GRIPS).17
Inclusion and Exclusion Criteria
Families in which at least 1 individual had a germline CDH1 P/LP variant were included in analyses. The first individual to test positive for a CDH1 variant in a family was considered the proband. Classification of CDH1 variants was confirmed manually using publicly available data curation in ClinVar and/or in consultation with the ClinGen CDH1 variant curation expert panel.18 Families with CDH1 variants of uncertain significance and hereditary diffuse gastric cancer–like syndrome without a CDH1 P/LP variant were excluded from the study; however, clinical management was provided based on International Gastric Cancer Linkage Consortium guidelines (Figure 1).8 Variants that had been downgraded from likely pathogenic to variants of uncertain significance at the time of data analysis were excluded. Missense variants were included if they impacted splicing and were classified as P/LP. eTable 1 in Supplement 1 lists the P/LP variants present in families in the current study cohort. A comprehensive list of CDH1 variants can be found in ClinVar (https://www.ncbi.nlm.nih.gov/clinvar). Deidentified pedigrees that included cancer histories, genotypes, and demographics were included in the final dataset.
Figure 1. North American Cohort Assembly.
Kindred indicates an extended family; P/LP, pathogenic/likely pathogenic.
aFamilies that fulfill hereditary diffuse gastric cancer genetic testing criteria but have no identifiable CDH1 or CTNNA1 variant.
Self-reported race and ethnicity were collected in an open-ended manner and to document potential differences in distribution and type of CDH1 variants by race and ethnicity. Pedigrees and CDH1 genotypes were reviewed by coauthors (G.A.F. and A.D.) for accuracy. If a family was present at more than 1 institution, that pedigree was represented only once in the final analysis.
Outcome Measures
Self-reported personal and family histories of diffuse gastric cancer and breast cancers (ductal, lobular, and unspecified) were included in the penetrance analysis; however, lobular carcinoma in situ was not. Pathology reports were not requested or required to confirm cancer diagnoses but were reviewed when provided. A diagnosis of occult, intramucosal signet ring cell carcinoma (pT1aN0, American Joint Committee on Cancer [AJCC] stage IA) was considered separate and distinct from the clinical diagnosis of advanced (AJCC stage II-IV) gastric cancer.19 Specifically, the diagnosis of clinically occult gastric cancer detected by random gastric biopsy during endoscopic surveillance in an asymptomatic patient was considered clinically not important (ie, not life-threatening); such patients were included in cumulative risk analyses because they retained their stomachs and were assumed to have similar advanced gastric cancer risk to those in whom surveillance did not detect occult cancer cells. Similarly, in the primary analysis of cumulative cancer risk, the diagnosis of an early gastric cancer (pT1aN0, AJCC stage IA) at the time of risk-reducing total gastrectomy was considered to be not clinically important. This is because stage IA gastric cancer is a frequent (88%-100%) diagnosis among asymptomatic CDH1 P/LP variant carriers who undergo risk-reducing total gastrectomy and, to our knowledge, no cancer recurrences have been reported in such cases.20,21,22 A secondary analysis was performed that included all diagnoses of gastric cancer irrespective of pathologic stage or clinical presentation.
Statistical Analyses
Penetrance analyses assumed a CDH1 P/LP variant allele frequency (VAF) of 10/100 000. We queried the UK Biobank (http://www.ukbiobank.ac.uk, under project ID 54389, April 2021), which demonstrated CDH1 P/LP VAF of 5/100 000 to 15/100 000. We queried gnomAD v4 and found CDH1 P/LP VAF of 5/100 000.23 According to ancestral descent, CDH1 P/LP VAF is reported as 1/12 000 among the European ancestral group, 1/10 000 among the African ancestral group, and 1/13 000 among the Asian ancestral group. In gnomAD v4, no CDH1 P/LP variants were reported in Ashkenazi Jewish, Finnish, and Middle Eastern ancestral groups. The exact VAF is expected to have little impact on results because a sensitivity analysis by Pharoah et al6 showed that risk estimates tend to be almost identical for frequencies less than 1 in 100. The gastric and female breast cancer risks for carriers of CDH1 P/LP variants were estimated using segregation analysis, with models fitted in R version 4.2.0 (R Core Team) by the method of maximum likelihood.24,25 Pedigree likelihoods were calculated by the R package clipp version 1.1.1 and the likelihood optimization was performed by the R function mle. All estimates were appropriately adjusted for the clinic-based ascertainment of families using a retrospective likelihood approach, in which each pedigree’s data were conditioned on the proband’s genotype and the ages and affected statuses of all family members.26 The expected number of CDH1 P/LP variant carriers in a given group was obtained by summing the carrier probabilities over all people in the group, where each person’s carrier probability was calculated by the R package clipp version 1.1.1 based on family structure and known CDH1 genotypes only.
Age-specific hazard ratios (HRs), which are the age-specific cancer incidences for people who carry the gene variant divided by the incidences in people who do not carry the gene variant, were estimated by the method of maximum likelihood. Incidences for noncarriers (with mean polygenic risk, see below) were set equal to the age- and sex-specific annual population data from the Surveillance, Epidemiology, and End Results Program for the period 2015 to 2019. Cancers at different anatomical sites (stomach or breast) within the same person were assumed to be conditionally independent, given the person’s genotype (ie, given the person’s major gene and polygenic genotypes, see below). Breast cancer HRs among male carriers were held fixed at the neutral value of 1. Other HRs were modeled as step functions that were constant in the 5 age groups (0-39, 40-49, 50-59, 60-69, and ≥70 years old).
A genetic mixed model that incorporated unmeasured polygenic factors in addition to the major gene CDH1 was used in the segregation analyses to account for unexplained familial aggregation of gastric cancer risk.24,25,27 The polygenic factor’s standard deviation was chosen to correspond to a familial relative risk of 2.35 and its mean was chosen so that the mean HR due to the polygenic factors was 1.28
For each combination of sex and anatomical site, the age-specific cumulative risk CR(t) of cancer by age t years (penetrance to age t years) was calculated from the estimated HRs as the mean over the polygenic factor of
 ,
|
where λ(s) is the product of the polygenic HR, the HR at age s for the CDH1 genotype, and the relevant population incidence at age s. For unaffected carriers aged t years, risks of cancer within the next decade were calculated as
 .
|
Confidence intervals for the cumulative risks and decadal risks were calculated using a parametric bootstrap. For various levels of family history (ie, number of relatives diagnosed with cancer), the lifetime gastric cancer risk for a CDH1 P/LP variant carrier was estimated as the mean over the carrier’s polygene of the corresponding lifetime risks, as in the calculation of CR(t) above, though now with the distribution of the polygene being conditional on the given family history. See the eMethods in Supplement 1 for more details.
All statistical analyses were performed in R version 4.2.0, using the R package clipp version 1.1.1 to calculate pedigree likelihoods and genotype probabilities and using the R function mle to optimize the likelihood.
Results
A total of 213 families including 7323 individuals were identified (Figure 1). Race and ethnicity recorded for kindreds included Asian (9 [4%]), Black (4 [2%]), Hispanic (9 [4%]), and White (181 [85%]) (Table 1). The personal cancer history among probands included 21.6% (46/213) with advanced (stage ≥II) gastric cancer and 32.9% (70/213) with breast cancer. CDH1 genotyping was reported in 1348 individuals, of whom 883 (65.5%) were diagnosed with a CDH1 P/LP variant. Among CDH1 P/LP variant carriers, the prevalence of gastric cancer was 13.9% (123/883), the prevalence of breast cancer among female carriers was 26.3% (144/547), and the prevalence of breast cancer among male carriers was 0.6% (2/336).
Table 1. Characteristics of Kindreds and Individuals With Germline CDH1 P/LP Variants.
| Characteristic | No. (%) |
|---|---|
| Kindredsa | 213 |
| Individuals per kindred, mean (IQR) | 30 (22-45) |
| Individuals genotyped per kindred, mean (IQR)b | 6 (2-8) |
| Total individuals | 7323 |
| Female | 3555 (49) |
| Male | 3623 (49) |
| Not specified | 145 (2) |
| Race and ethnicity (by kindred) | |
| Asian | 9 (4) |
| Black or African American | 4 (2) |
| Hispanic or Latino | 9 (4) |
| Middle Eastern or North African | 4 (2) |
| Multiracial | 5 (2) |
| White | 181 (85) |
| Not reported | 1 (0.5) |
| Cancer history (by kindred) | |
| Both gastric and breast cancer | 109 (51) |
| Gastric cancer only | 52 (24) |
| Breast cancer only | 43 (20) |
| Neither | 9 (4) |
| All genotyped individuals | 1348 |
| Female | 833 (62) |
| Male | 515 (38) |
| Individuals with CDH1 P/LP germline variantc | |
| Total | 883 |
| Female | 547 (62) |
| Male | 336 (38) |
| All CDH1 P/LP variant carriers | 883 |
| Breast cancer | 146 (17) |
| Female | 144 (26) |
| Male | 2 (0.6) |
| Gastric cancer | 123 (14) |
| Female | 61 (11) |
| Male | 62 (18) |
| Age at cancer diagnosis, mean (IQR), y | |
| Breast cancer | 51 (44-58) |
| Gastric cancer | 49 (39-59) |
Abbreviations: P/LP, pathogenic or likely pathogenic; IQR, interquartile range.
Kindred, an extended family.
Genotype refers to germline genetic testing for specific gene variants or mutations.
Germline variant, a gene change in a reproductive cell (egg or sperm) that becomes incorporated into the DNA of every cell in the body of the offspring.
A total of 103 unique gene variants were identified among 883 individuals with CDH1 P/LP variants. The most frequent variants were frameshift (38/103 [36.9%]), followed by splice site variants (24/103 [23.3%]), nonsense variants (23/103 [22.3%]), large deletions of 1 or more exons (10/103 [9.7%]), and missense variants that affected splicing (6/103 [5.8%]) (eTable 1 in Supplement 1). Variants were distributed throughout the CDH1 gene and observed in all 16 exons. The most common CDH1 variant, c.2064_2065del, was observed in 77 individuals (77/883 [8.7%]) and 15 kindreds (15/213 [7%]). Among people with this variant, the prevalence of gastric cancer was 13% and the prevalence of breast cancer was 25%. When variants identified in single individuals were excluded, the CDH1 variant c.48+1G>A, a canonical splice site variant, was associated with the highest frequency of gastric cancer.
Observed Cancer Rates in the CDH1 Cohort
Among individuals with CDH1 P/LP variants (age range, 17-84 years), the prevalence of gastric cancer was 13.9% and the prevalence of breast cancer was 26.3%. Advanced gastric cancer (AJCC stage ≥II) was more common in male carriers (62/336 [18.5%]) than female carriers (61/547 [11.2%]). The median age at gastric cancer diagnosis was similar among CDH1 P/LP variant carriers and the overall cohort (genotyped and not genotyped) (49 vs 50 years, P = .39). Breast cancer among female carriers was more frequent among individuals with CDH1 P/LP variants (146/883 [16.5%]) compared with genotyped individuals who tested negative for a CDH1 P/LP variant (18/465 [3.9%]). Among people with a CDH1 P/LP variant, breast cancer among female carriers was diagnosed at a median age of 51 years compared with 55 years in the entire cohort. Overall, 5 men had breast cancer, including 2 with the CDH1 P/LP variant, 1 without a CDH1 P/LP variant, and 2 who did not have genotyping.
Cancer Risk Estimates
The estimated HRs for advanced gastric cancer ranged from 33.5 (95% CI, 9.8-112) at age 30 years to 3.5 (95% CI, 0.4-30.3) at age 70 years (Table 2). For example, this means that the incidence of advanced gastric cancer was estimated to be 33.5 times higher for individuals with CDH1 P/LP variants than for individuals without CDH1 variants at age 30 years. The cumulative risk of advanced gastric cancer by age 80 years was 10.3% (95% CI, 6%-23.6%) for male carriers who carried a CDH1 P/LP variant and 6.5% (95% CI, 3.8%-15.1%) for female carriers (Figure 2). These penetrance estimates were calculated for individuals irrespective of family history, but a CDH1 P/LP variant carrier with 3 affected first-degree relatives had an estimated penetrance of 38% (95% CI, 25%-64%) because of other shared familial risk factors for gastric cancer (Figure 3). Inclusion of all gastric cancers, specifically occult stage IA (pT1aN0) cancers diagnosed at the time of risk-reducing gastrectomy, resulted in higher risk estimates by approximately 2-fold (eFigure in Supplement 1). For example, cumulative advanced gastric cancer risk in female and male carriers by age 80 years was 12.6% and 19.1%, respectively (eTable 2 in Supplement 1). The inclusion of cases in which gastric cancer was detected at the time of a surveillance endoscopy did not meaningfully alter risk estimates (eTable 3 in Supplement 1).
Table 2. Hazard Ratio and Cumulative Risk Estimates of Advanced Gastric and Breast Cancer for CDH1 P/LP Variant Carriers.
| Age, y | Advanced gastric cancera | Breast cancer in female carriersb | ||||||
|---|---|---|---|---|---|---|---|---|
| Hazard ratio (95% CI) | Male carriers | Female carriers | ||||||
| No.c | Cumulative risk, % (95% CI) | No.c | Cumulative risk, % (95% CI) | Hazard ratio (95% CI) | No.c | Cumulative risk, % (95% CI) | ||
| 30 | 33.5 (9.8-112) | 783 | 0.2 (0.1-0.6) | 801 | 0.2 (0.1-0.6) | 5.7 (2.5-13.2) | 800 | 0.5 (0.2-1.2) |
| 40 | 13.9 (4.4-42.6) | 700 | 0.7 (0.3-2.3) | 690 | 0.8 (0.3-2.5) | 4.2 (2.2-7.9) | 682 | 3.6 (1.9-7.4) |
| 50 | 24.1 (8.4-67.9) | 435 | 1.6 (0.8-3.7) | 446 | 1.5 (0.8-3.6) | 4.3 (2.3-8.1) | 423 | 10.5 (6.9-16.9) |
| 60 | 19 (5.7-59) | 308 | 4.7 (2.5-9.9) | 322 | 3.4 (1.9-7.2) | 1.2 (0.4-3.6) | 281 | 19.3 (13.7-28.6) |
| 70 | 3.5 (0.4-30.3) | 177 | 8.9 (4.9-18.2) | 203 | 5.7 (3.2-11.8) | 3.9 (1.1-13.7) | 169 | 24.2 (18.3-35.5) |
| 80 | 3.5 (0.4-26.2) | 83 | 10.3 (6-23.6) | 103 | 6.5 (3.8-15.1) | 3.9 (1.1-13.7) | 84 | 36.8 (25.7-62.9) |
Abbreviation: P/LP, pathogenic or likely pathogenic.
Advanced gastric cancer was defined according to American Joint Committee on Cancer as T1b or greater tumor stage, N1 or greater nodal stage, and any M stage.
Breast cancer was defined as invasive carcinoma of lobular or ductal subtype; carcinoma in situ was excluded.
The expected number of nonproband carriers who are at risk of the given cancer at the given age.
Figure 2. Age-Specific Cumulative Risks and 10-Year Risks of Cancer.
Age-specific cumulative risks are shown for CDH1 pathogenic/likely pathogenic (P/LP) variant carriers, with shaded regions indicating the 95% CIs (A-C). The probabilities that a CDH1 P/LP variant carrier who is unaffected at a given age will be diagnosed in the next 10 years are shown, with shaded regions indicating the 95% CIs (D-F). For example, panel A indicates that an unaffected male carrier aged 60 years has a 4.4% chance of developing gastric cancer by the age of 70 years. The expected number of nonproband carriers who are at risk for each cancer type at the ages 20, 30, 40, and on are shown.
Figure 3. Effect of Family History on Estimated Lifetime Risk of Gastric Cancer.
Lifetime gastric cancer risks for carriers under various scenarios of family history. The lifetime risks presented are, from left to right, irrespective of family history (single open box) or with a family history of 0, 1, 2, or 3 affected first-degree relatives. The expected number of nonproband carriers in the study with the given family history are indicated (n) below each pedigree. The whiskers indicate 95% CIs. In the pedigree drawings, the CDH1 pathogenic/likely pathogenic variant carrier is indicated with an arrow, squares are male and circles are female individuals, open circles are unaffected family members, black filled circles indicate gastric cancer cases, strikethroughs indicate deceased family members, and the numbers give the ages of the carrier’s relatives in years.
The probabilities that an individual unaffected by cancer at the start of a given decade of life will develop gastric cancer at ages 30 to 39 years, 40 to 49 years, 50 to 59 years, 60 to 69 years, and 70 to 79 years were 0.6%, 0.8%, 3.1%, 4.4%, and 1.6%, respectively, for male carriers and 0.6%, 0.7%, 1.9%, 2.4%, or 0.9%, respectively, for female carriers (Figure 2; eTable 4 in Supplement 1). Cancer risk by decade estimated with inclusion of all cancer diagnoses is shown in eTable 5 in Supplement 1. Lifetime breast cancer risk was estimated in female carriers who were CDH1 P/LP variant carriers. The HR ranged from 5.7 (95% CI, 2.5-13.2) at age 30 years to 3.9 (95% CI, 1.1-13.7) at age 70 years (Table 2). The cumulative risk for breast cancer in female carriers by age 80 was 36.8% (95% CI, 25.7%-62.9%) (Figure 2). The probabilities of breast cancer during the decades corresponding to ages 30 to 39 years, 40 to 49 years, 50 to 59 years, 60 to 69 years, and 70 to 79 years were estimated to be 3.1%, 7.1%, 9.9%, 6.1%, and 16.6%, respectively (eTable 4 in Supplement 1).
Discussion
These analyses estimated gastric and breast cancer risk from geographically and genetically distinct kindreds with CDH1 variants in North America. Lifetime risk of advanced gastric cancer was 10% in male and 7% in female carriers, and breast cancer risk was 37%. While these estimates were for carriers irrespective of family history, individualized cancer risk was also estimated. Results showed that an individual with strong family history of gastric cancer, defined as 3 first-degree relatives with gastric cancer, can have up to a 30% higher lifetime cancer risk due to the combined effect of CDH1 variants and other heritable risk factors for cancer. Current cancer risk estimates from this study (7%-10% and 37%) closely aligned with the observed rates (13.9% and 26.3%) of gastric and breast cancer, respectively, in individuals with CDH1 P/LP variants.
These results can help communicate lifetime cancer risk to patients with CDH1 variants, and help patients and clinicians understand cancer risk with a more discrete, individualized frame of reference. These results may inform clinical decision-making for people with CDH1 variants. For example, individuals with a CDH1 P/LP variant and no family history of gastric cancer may warrant consideration of gastric surveillance rather than risk-reducing total gastrectomy because of a lower cumulative risk compared with those with 1 or more family members with gastric cancer. Findings reported here require validation in independent and diverse CDH1 cohorts to inform precision clinical management in the context of other known determinants of cancer risk.
Contemporary estimates have been hindered by evaluation of small numbers of affected kindreds or restrictive, phenotype-driven genetic testing criteria.11,12 Although the current study included families identified from clinical settings, analyses accounted for this by using a retrospective likelihood approach that gives unbiased estimates when the ascertainment scheme only depends on the families’ phenotypes.26 In contrast, studies of families identified from clinical settings with less rigorous corrections for ascertainment may result in overestimates of risk.11,12 The prevalence of advanced gastric cancer (13.9%) in the current study corresponded well with the cumulative risk estimates. When clinically irrelevant asymptomatic cases of stage IA gastric cancer were included in analysis, the cumulative risk of advanced gastric cancer by age 80 years increased from 7%-10% to 13%-19%. The evidence that early gastric lesions are a frequent and expected finding in CDH1 P/LP variant carriers, combined with current gastric cancer risk estimates, should be considered in clinical decision-making regarding referral of individuals for risk-reducing surgery and the methods of endoscopic mucosal surveillance.
Prior studies described that the risk of breast cancer due to CDH1 ranged from 39% to 55%.7,11,12 In the current study, the estimated lifetime risk for breast cancer in female carriers by age 80 years was 37%. In the current study, females aged 40 to 69 years had a higher probability of developing breast cancer than gastric cancer during the same period. These results can inform communication of breast cancer risk for people with a CDH1 P/LP variant.
Limitations
This study had several limitations. First, histologic subtype of cancer cases was not uniformly available. Second, diagnoses of cancer were self-reported and pathology data were not available to confirm all diagnoses of cancer. Third, gastric cancer risk estimates in the primary analysis were derived from diagnoses of advanced gastric cancer by excluding cases of early-stage gastric signet ring cell lesions that are frequently diagnosed with surveillance endoscopy and risk-reducing total gastrectomy and may have resulted in lower risk estimates than the true risk estimates.20,22 Fourth, the risk estimates reported here do not apply to people with risk-reducing surgery. Fifth, the effect of risk-reducing surgery or enhanced cancer surveillance on cancer rates could not be determined from these analyses. Sixth, it is possible that variant carriers from clinic-based families have higher cancer risk than variant carriers in the general population. Seventh, analyses did not adjust for confounders, such as ancestry and as yet unidentified familial risk factors. Eighth, there was a small proportion of underrepresented racial and ethnic minority groups.
Conclusions
Among families from North America with germline CDH1 P/LP variants, the cumulative risk of gastric cancer was 7% to 10%, which was lower than described previously, and the cumulative risk of breast cancer among female carriers was 37%, which was similar to prior estimates. These findings inform current management of individuals with germline CDH1 variants.
eFigure. Secondary Analysis of Age Specific Cumulative Risks
eTable 1. CDH1 Genotype and Variant Classification of Pathogenic and Likely Pathogenic Variants Present in Study Cohort
eTable 2. Secondary Analysis With Hazard Ratio (HR) and Cumulative Risk Estimates of Advanced Gastric and Breast Cancer for CDH1 P/LP Variant Carriers That Includes Family History of Advanced Gastric Cancer and Stage IA Gastric Cancers Detected at the Time of Prophylactic Total Gastrectomy
eTable 3. Secondary Analysis With Hazard Ratio (HR) and Cumulative Risk Estimates of Advanced Gastric and Breast Cancer for CDH1 P/LP Variant Carriers That Includes Family History of Advanced Gastric Cancer and Stage IA Gastric Cancers Detected at the Time of Prophylactic Total Gastrectomy, and Intramucosal Signet Ring Cells Detected at Endoscopy
eTable 4. The Probability (%) That an Unaffected Carrier of a Given Age Will Develop Cancer in the Next 10 Years
eTable 5. Secondary Analysis Including All Gastric Cancer Diagnoses Irrespective of Clinical Presentation or Method of Detection (i.e., Endoscopy or Gastrectomy), Demonstrating the Probability (%) That an Unaffected Carrier of a Given Age Will Develop Cancer in the Next 10 Years
eMethods. Supplementary Statistical Methods
eReferences
Data Sharing Statement
References
- 1.Guilford PJ, Hopkins JB, Grady WM, et al. E-cadherin germline mutations define an inherited cancer syndrome dominated by diffuse gastric cancer. Hum Mutat. 1999;14(3):249-255. doi: 10.1002/(SICI)1098-1004(1999)14:3<249::AID-HUMU8>3.0.CO;2-9 [DOI] [PubMed] [Google Scholar]
- 2.Guilford P, Hopkins J, Harraway J, et al. E-cadherin germline mutations in familial gastric cancer. Nature. 1998;392(6674):402-405. doi: 10.1038/32918 [DOI] [PubMed] [Google Scholar]
- 3.Figueiredo J, Melo S, Carneiro P, et al. Clinical spectrum and pleiotropic nature of CDH1 germline mutations. J Med Genet. 2019;56(4):199-208. doi: 10.1136/jmedgenet-2018-105807 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bar-Mashiah A, Soper ER, Cullina S, et al. CDH1 pathogenic variants and cancer risk in an unselected patient population. Fam Cancer. 2022;21(2):235-239. doi: 10.1007/s10689-021-00257-x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Kaurah P, MacMillan A, Boyd N, et al. Founder and recurrent CDH1 mutations in families with hereditary diffuse gastric cancer. JAMA. 2007;297(21):2360-2372. doi: 10.1001/jama.297.21.2360 [DOI] [PubMed] [Google Scholar]
- 6.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):1348-1353. doi: 10.1053/gast.2001.29611 [DOI] [PubMed] [Google Scholar]
- 7.Hansford S, Kaurah P, Li-Chang 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 [DOI] [PubMed] [Google Scholar]
- 8.Blair VR, McLeod M, Carneiro F, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Gamble LA, Rossi A, Fasaye GA, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Girardi A, Magnoni F, Vicini E, et al. CDH1 germline mutations in families with hereditary lobular breast cancer. Eur J Cancer Prev. 2022;31(3):274-278. doi: 10.1097/CEJ.0000000000000688 [DOI] [PubMed] [Google Scholar]
- 11.Roberts ME, Ranola JMO, Marshall ML, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Xicola RM, Li S, Rodriguez N, 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 [DOI] [PubMed] [Google Scholar]
- 13.Lee K, Krempely K, Roberts ME, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Richards S, Aziz N, Bale S, et al. ; ACMG Laboratory Quality Assurance Committee . Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med. 2015;17(5):405-424. doi: 10.1038/gim.2015.30 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Luo X, Maciaszek JL, Thompson BA, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.National Comprehensive Cancer Network . Gastric cancer (version 2.2022). Accessed December 21, 2022. https://www.nccn.org/professionals/physician_gls/pdf/gastric.pdf
- 17.Janssens AC, Ioannidis JP, van Duijn CM, Little J, Khoury MJ; GRIPS Group . Strengthening the reporting of genetic risk prediction studies: the GRIPS statement. BMJ. 2011;342:d631. doi: 10.1136/bmj.d631 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Rehm HL, Berg JS, Brooks LD, et al. ; ClinGen . ClinGen: the Clinical Genome Resource. N Engl J Med. 2015;372(23):2235-2242. doi: 10.1056/NEJMsr1406261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Amin MB, Greene FL, Edge SB, et al. The Eighth Edition AJCC Cancer Staging Manual: continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin. 2017;67(2):93-99. doi: 10.3322/caac.21388 [DOI] [PubMed] [Google Scholar]
- 20.Asif B, Sarvestani AL, Gamble LA, 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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Charlton A, Blair V, Shaw D, Parry S, Guilford P, Martin IG. Hereditary diffuse gastric cancer: predominance of multiple foci of signet ring cell carcinoma in distal stomach and transitional zone. Gut. 2004;53(6):814-820. doi: 10.1136/gut.2002.010447 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Rocha JP, Gullo I, Wen X, et al. Pathological features of total gastrectomy specimens from asymptomatic hereditary diffuse gastric cancer patients and implications for clinical management. Histopathology. 2018;73(6):878-886. doi: 10.1111/his.13715 [DOI] [PubMed] [Google Scholar]
- 23.Karczewski KJ, Francioli LC, Tiao G, et al. ; Genome Aggregation Database Consortium . The mutational constraint spectrum quantified from variation in 141,456 humans. Nature. 2020;581(7809):434-443. doi: 10.1038/s41586-020-2308-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Antoniou AC, Pharoah PD, McMullan G, Day NE, Ponder BA, Easton D. Evidence for further breast cancer susceptibility genes in addition to BRCA1 and BRCA2 in a population-based study. Genet Epidemiol. 2001;21(1):1-18. doi: 10.1002/gepi.1014 [DOI] [PubMed] [Google Scholar]
- 25.Lange K. Mathematical and Statistical Methods for Genetic Analysis. 2nd ed. Springer; 2002. doi: 10.1007/978-0-387-21750-5 [DOI] [Google Scholar]
- 26.Kraft P, Thomas DC. Bias and efficiency in family-based gene-characterization studies: conditional, prospective, retrospective, and joint likelihoods. Am J Hum Genet. 2000;66(3):1119-1131. doi: 10.1086/302808 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Gong G, Hannon N, Whittemore AS. Estimating gene penetrance from family data. Genet Epidemiol. 2010;34(4):373-381. doi: 10.1002/gepi.20493 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Yaghoobi M, McNabb-Baltar J, Bijarchi R, Hunt RH. What is the quantitative risk of gastric cancer in the first-degree relatives of patients? a meta-analysis. World J Gastroenterol. 2017;23(13):2435-2442. doi: 10.3748/wjg.v23.i13.2435 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eFigure. Secondary Analysis of Age Specific Cumulative Risks
eTable 1. CDH1 Genotype and Variant Classification of Pathogenic and Likely Pathogenic Variants Present in Study Cohort
eTable 2. Secondary Analysis With Hazard Ratio (HR) and Cumulative Risk Estimates of Advanced Gastric and Breast Cancer for CDH1 P/LP Variant Carriers That Includes Family History of Advanced Gastric Cancer and Stage IA Gastric Cancers Detected at the Time of Prophylactic Total Gastrectomy
eTable 3. Secondary Analysis With Hazard Ratio (HR) and Cumulative Risk Estimates of Advanced Gastric and Breast Cancer for CDH1 P/LP Variant Carriers That Includes Family History of Advanced Gastric Cancer and Stage IA Gastric Cancers Detected at the Time of Prophylactic Total Gastrectomy, and Intramucosal Signet Ring Cells Detected at Endoscopy
eTable 4. The Probability (%) That an Unaffected Carrier of a Given Age Will Develop Cancer in the Next 10 Years
eTable 5. Secondary Analysis Including All Gastric Cancer Diagnoses Irrespective of Clinical Presentation or Method of Detection (i.e., Endoscopy or Gastrectomy), Demonstrating the Probability (%) That an Unaffected Carrier of a Given Age Will Develop Cancer in the Next 10 Years
eMethods. Supplementary Statistical Methods
eReferences
Data Sharing Statement
