The role of multi-organ cancer predisposition genes in the risk of inherited and histologically diverse gastric cancer

Abstract

Background: Approximately 10% of cases with gastric cancer (GC) exhibit familial clustering, however, only 1-3% of cases can be explained by two known hereditary syndromes: Hereditary Diffuse Gastric Cancer (HDGC) caused by CDH1 and CTNNA1 pathogenic germline variants; and Gastric Adenocarcinoma and Proximal Polyposis of the Stomach (GAPPS), caused by germline variants in APC 1B promoter. Familial intestinal gastric cancer (FIGC) has been defined clinically, but it remains mostly genetically unexplained. Likewise, the heritability of mixed histology GC remains to be known. We aimed to estimate the frequency of known cancer predisposition gene variants in GC cases with and without a cancer family history, diverse histological subtypes, and varied age of onset.

Methods: We evaluated the contribution of pathogenic or likely pathogenic (P/LP) variants in well-established moderate-to-high penetrance multi-organ cancer predisposition genes for GC risk in a large international multi-centre retrospective cohort study involving 750 patients with GC of early-onset or family history of cancer, either by panel sequencing or whole exome sequencing (WES). Panel sequencing was conducted on 328 cases, while WES was performed on the remaining 422. Tumour sequence analyses were performed on samples from 15 patients with P/LP variants. Mutations identified in five index cases were also tested in their relatives.

Findings: We identified 45 patients (6%) with P/LP variants in: ATM (17 cases), BRCA2 (10 cases), MLH1 (five cases), TP53 (three cases), BRCA1, PALB2, RAD51D, and CHEK2 (two patients each), and RAD51C and PMS2 (one case each), all of which were mutually exclusive. The P/LP variant prevalence was higher in intestinal (9.8%) than in diffuse (4.3%) or mixed GC (4.5%) (p-value = 0.023), without difference per mutated gene by histological subtypes. Only 16 of the 45 patients who carried P/LP variants fulfilled the National Comprehensive Cancer Network genetic testing criteria of at least one cancer predisposition syndrome.

Interpretation: Our findings indicate that a broader panel of cancer predisposition genes, beyond CDH1 and CTNNA1, should be included in gene panels to investigate germline variants in patients with GC. This would be especially beneficial when there is a family history of cancer, irrespective of histology subtype, as it would increase the chance of identifying patients who could benefit from risk reduction, targeted treatment, and surveillance of other cancer types.

Funding: National Cancer Institute of the National Institutes of Health, USA; Universidad del Tolima, Colombia; MINCIENCIAS, Colombia; L'OREAL-UNESCO-ICETEX-COLCIENCIAS, Colombia; Instituto Nacional de Cancerología, Colombia; American Association for Cancer Research, USA; ANID Ministerio de Ciencia, Chile; Fondecyt, Chile; CONICYT/ANID FONDAP, Chile; Instituto Mexicano del Seguro Social and Consejo Nacional de Ciencia y Tecnología, México; IPO Porto, Portugal; Liga Portuguesa Contra o Cancro, Portugal; Fundacao para a Ciencia e Tecnologia, Portugal; The Auburn Community Cancer Endowed Chair in Basic Research, USA; The Heart, BrEast, and BrAin HeaLth Equity Research (HEAL HER) program, a program made possible by residual class settlement funds in the matter of April Krueger v. Wyeth, Inc., Case No. 03-cv-2496 (US District Court, SD of Calif.), USA.

Keywords: Familial aggregation; Gastric cancer; Genetic predisposition; Germline variants; Homologous recombination; Stomach cancer.

Fig. 1

Landscape of pathogenic and likely pathogenic germline variants identified in index patients with gastric cancer. Characteristics of 45 patients carrying germline pathogenic or likely pathogenic variants. In the first panel, each column represents a patient, and each row a gene, colour indicates variant classification type as indicated in legend. The second panel shows the clinical characteristics of each patient, colour coded as indicated in the panel legend.

Fig. 2

Somatic analysis of patients with gastric cancer, carriers of pathogenic germline variants. Oncoplot depicting somatic mutations from patients with P/LP germline variant in which somatic data available by whole exome (n = 7), or targeted panel (n = 8). Each column represents a patient and each row a gene, colour indicates variant classification type as indicated in legend. The table below indicates detailed patient and mutation information. Genes in red represent mutations in a gene within the same pathway as the initial germline variation. Homologous recombination deficiency (HRD) score calculated based on loss of heterozygosity (LOH), telomeric allelic imbalance, and large-scale state transition for whole exome data only. Microsatellite instability (MSI) score was calculated using MSISensor-Pro for whole exome data only and confirmed by analysing the fragment length variations of five markers within the “Bethesda panel”. Tumour Mutation Burden (TMB) was calculated for exome data only as number of somatic mutations per megabase.