The Largest Chinese Cohort Study Indicates Homologous Recombination Pathway Gene Mutations as Another Major Genetic Risk Factor for Colorectal Cancer with Heterogeneous Clinical Phenotypes

Abstract

While genetic factors were associated with over 30% of colorectal cancer (CRC) patients, mutations in CRC-susceptibility genes were identified in only 5% to 10% of these patients. Besides, previous studies on hereditary CRC were largely designed to analyze germline mutations in patients with single genetic high-risk factor, which limited understanding of the association between genotype and phenotypes. From January 2015 to December 2018, we retrospectively enrolled 2,181 patients from 8,270 consecutive CRC cases, covering 5 categories of genetic high-risk factors. Leukocyte genomic DNA was analyzed for germline mutations in cancer predisposition genes. The germline mutations under each category were detected and analyzed in association with CRC susceptibility, clinical phenotypes, and prognoses. A total of 462 pathogenic variants were detected in 19.3% of enrolled CRC patients. Mismatch repair gene mutation was identified in 9.1% of patients, most prevalent across all high-risk groups. Homologous recombination (HR) gene mutations were detected in 6.5% of cases, penetrated in early-onset and extra-colonic cancer risk groups. Mutations in HR genes, including BARD1, RAD50, and ATM, were found to increase CRC risk with odds ratios of 2.8-, 3.1-, and 3.1-fold, respectively. CRC patients with distinct germline mutations manifested heterogeneous phenotypes in clinicopathology and long-term prognoses. Thus, germline mutation screenings should be performed for CRC patients with any of those genetic risk factors. This study also reveals that HR gene mutations may be another major driver for increased CRC risk.

Fig. 1.

Schematic of the study and sample enrollment. (A) Schematic of the study. †Samples from patients with complete response after neoadjuvant therapy were not tested with IHC. Risk factors are not mutually exclusive. A patient fitting several criteria will be counted multiple times, causing the sum from all groups to surpass the total of 2,181 patients. (B) The upset plot illustrated enrolled samples' distribution and basic clinical characteristics.

Fig. 2.

Germline mutation spectrum of 2,181 CRC patients with high genetic risk. (A) Landscape of P/LP germline mutations identified in CRC patients (n = 421), the 20 genes mentioned in NCCN guidelines are shown in the top part, and the other 12 genes are shown in the bottom part, #Heterozygous mutation. (B) According to mutation classification, count the frequency of mutations. (C) Overall detection rate of germline mutation and major gene pathways in high-risk patients (n = 2,181). (D) Forest plot displays the odds ratio of CRC susceptibility genes.

Fig. 3.

Germline mutation spectrum and prevalence with different genetic risks. (A) Landscape of P/LP germline mutation identified in patients with 5 genetic risk groups. The red background indicated the highest number of patients with the mutation. (B) Sankey plot illustrated the correlation between gene mutations and genetic risk factors. (C) Overall detection rate of germline mutation and major gene pathways in high-risk patients (n = 2,181). (D) The upset plot illustrated the enrolled sample’s distribution and germline mutation rate. (E) Relationship between dMMR tumors only or with at least one additional risk factor and the detection rate of germline mutation. The bar chart shows the average detection rate of germline mutation in patients with pMMR and other risk factors.

Fig. 4.

Clinical outcomes of patients carrying respective germline mutations. (A) Correlation analysis between mutation types and clinical outcomes. (B) Comparison of the proportion of CRC primary location in respective germline mutation carriers. (C) Age of onset distribution in individual germline mutation carriers. (D) Comparison of the number of metastasized lymph nodes observed in respective germline mutation carriers. MMR, HR, APC, and Other refer to LS patients, HR gene mutation carriers, FAP patients, and carriers of mutations in other genes, respectively. Adjusted P values were applied in multiple groups comparison.

Fig. 5.

Survival analysis of respective molecular subtypes. (A) Overall survival curves for CRC patients carrying germline mutation. (B) Recurrence-free survival curves for CRC patients carrying germline mutation. (C) Overall survival curves for CRC of respective germline mutation status. (D) Recurrence-free survival curves for CRC of respective germline mutation status. (E) Risk ratio analysis of risk factors for survival. MMR, HR, APC, and Other refer to LS patients, HR genes mutation carriers, FAP patients, and carriers of mutations in other genes, respectively.

Fig. 6.

Establishment of genetic testing recommendations for patients carrying respective genetic risks. The above diagram illustrates the effect of 5 risk factors alone on predicting germline mutation carriers, and the flowchart shows genetic testing based on risk factors.