Whole-Exome Sequencing Among Chinese Patients With Hereditary Diffuse Gastric Cancer

Abstract

Importance: The E-cadherin gene, CDH1, and the α-E-catenin gene, CTNNA1, were previously identified as hereditary diffuse gastric cancer (HDGC) susceptibility genes, explaining 25% to 50% of HDGC cases. The genetic basis underlying disease susceptibility in the remaining 50% to 75% of patients with HDGC is still unknown.

Objective: To assess the incidence rate of CDH1 germline alterations in HDGC, identify new susceptibility genes that can be used for screening of HDGC, and provide a genetic landscape for HDGC.

Design, setting, and participants: This cohort study conducted retrospective whole-exome and targeted sequencing of 284 leukocyte samples and 186 paired tumor samples from Chinese patients with HDGC over a long follow-up period (median, 21.7 [range, 0.6-185.9] months). Among 10 431 patients diagnosed with gastric cancer between January 1, 2002, and August 31, 2018, 284 patients who met the criteria for HDGC were included. Data were analyzed from August 1 to 30, 2020.

Main outcomes and measures: Incidence rate of CDH1 germline alterations, identification of new HDGC susceptibility genes, and genetic landscape of HDGC.

Results: Among 284 Chinese patients, 161 (56.7%) were female, and the median age was 35 (range, 20-75) years. The frequency of CDH1 germline alterations was 2.8%, whereas the frequency of CDH1 somatic alterations was 25.3%. The genes with the highest incidence (>10%) of private germline alterations (including insertions and deletions) in the HDGC cohort were MUC4, ABCA13, ZNF469, FCGBP, IGFN1, RNF213, and SSPO, whereas previously reported germline alterations of CTNNA1, BRCA2, STK11, PRSS1, ATM, MSR1, PALB2, BRCA1, and RAD51C were observed at low frequencies (median, 4 [range, 1-12] cases). Furthermore, enrichment of the somatic variant signature of exposure to aflatoxin suggested potential interaction between genetics and environment in HDGC. Double-hit events in genes such as CACNA1D were observed, which suggested that these events might serve as important mechanisms for HDGC tumorigenesis. In addition, germline variants of FSIP2, HSPG2, and NCKAP5 and somatic alterations of FGFR3, ASPSCR1, CIC, DGCR8, and LZTR1 were associated with poor overall survival among patients with HDGC.

Conclusions and relevance: This study provided a genetic landscape for HDGC. The study's findings challenged the previously reported high germline alteration rate of CDH1 in HDGC and identified new potential susceptibility genes. Analyses of variant signatures and double-hit events revealed potentially important mechanisms for HDGC tumorigenesis. Findings from the present study may provide helpful information for further investigations of HDGC.

Figure 1.. Germline Alteration Landscape of Hereditary Diffuse Gastric Cancer

Oncoplot showing the germline alteration landscape among 284 Chinese patients with hereditary diffuse gastric cancer. The top panel shows the total of all nonsilent germline alterations for each sample. The middle panel shows the alteration details; only genes with alteration frequencies of 8% or greater are shown. The bottom panel shows clinical information. Cancer staging was based on the AJCC Cancer Staging Manual, 7th edition. EBER indicates Epstein-Barr virus–encoded small RNAs.

Figure 2.. Germline Alterations of CDH1 and MUC4

Figure 3.. Somatic Alteration Landscape of Hereditary Diffuse Gastric Cancer

Oncoplot showing the somatic alteration landscape among 186 Chinese patients with hereditary diffuse gastric cancer. The top panel shows the total of all nonsilent somatic alterations for each sample. The middle panel shows the alteration details; only the 24 most altered genes with alteration frequencies of 7% or greater are shown, with red font indicating genes with significant alterations. The bottom panel shows clinical information. Cancer staging was based on the AJCC Cancer Staging Manual, 7th edition. EBER indicates Epstein-Barr virus–encoded small RNAs.

Figure 4.. Significantly Altered Genes and Somatic Copy Number Variations

A, Significantly altered genes (Q < .05) were detected using the MutSigCV algorithm. The low Q values were limited to a fixed value because of the processing precision limit. B, The G scores of genomic regions were plotted along the genome, with orange spikes representing significant copy number gains and dark gray spikes representing significant copy number losses. Light gray spikes indicate insignificant gain and loss. Key genes associated with gastric cancer are shown in red type, and cancer census genes (from the Catalogue of Somatic Mutations in Cancer signature database) associated with other tumor types are shown in green type.