Distinct Molecular Profiles of Sporadic Early-Onset Colorectal Cancer: A Population-Based Cohort and Systematic Review

Abstract

Background and aims: The observed increase in the incidence of early-onset colorectal cancer (EOCRC) is being driven by sporadic cases, but the molecular characteristics of these tumors are not fully understood. Our objective was to investigate the prevalence of microsatellite instability (MSI) and selected mutations in sporadic EOCRC, and their association with survival.

Methods: Firstly, we compared the prevalence of molecular characteristics and survival within a population-based cohort study of 652 stage II and III colon cancer patients in Northern Ireland, comparing sporadic early-onset (<50 years, n = 35) with older (60-69 years, n = 179) patients. Secondly, a systematic review for studies reporting the prevalence of MSI, mismatch repair deficiency (dMMR), or BRAF, KRAS, NRAS, PIK3CA, and TP53 mutations in sporadic EOCRC was conducted. A meta-analysis was performed to calculate pooled estimates of the prevalence of molecular features in sporadic EOCRC.

Results: Firstly, within the cohort study, EOCRC patients did not have a significantly increased risk of colorectal cancer-specific death (adjusted hazard ratio 1.20; 95% confidence interval [CI] 0.61-2.39) compared with 60- to 69-year-olds. Second, 32 studies were included in the systematic review. The pooled analysis estimated a prevalence of 10% (95% CI 7%-14%) for MSI high/dMMR in sporadic EOCRC. BRAF and KRAS mutations had a prevalence of 1% (95% CI 0%-3%) and 32% (95% CI 23%-40%), respectively.

Conclusion: The molecular characteristics of sporadic EOCRC differ from those of cancers in older adults, particularly regarding reduced prevalence of BRAF mutations. Ten percent of sporadic EOCRC display MSI high/dMMR. Further studies are needed to address survival in sporadic EOCRC cases and whether molecular profiles influence EOCRC outcomes in this patient group.

Keywords: Microsatellite Instability; Mismatch Repair; Mutations.

Figure 1

Distribution of molecular characteristics by age category in sporadic stage II and III colon cancer.

Figure 2

Risk of CRC–specific death according to age categories in stage II and III colon cancer cases, by microsatellite instability status. ∗∗Adjusted for sex, adjuvant chemotherapy receipt, stage, tumor differentiation, family history of CRC, ECOG performance status, smoking, alcohol, inflammatory bowel disease and emergency surgery. CI, confidence interval; CRC, colorectal cancer; ECOG, Eastern Cooperative Oncology Group; HR, hazard ratio; MSI, microsatellite instability.

Figure 3

Flow chart of the selection of articles included in the review.

Figure 4

(A) Forest plot illustrating meta-analysis of the prevalence of microsatellite instability-high/deficient mismatch repair tumors in sporadic early-onset colorectal cancer. (B) Forest plot illustrating meta-analysis of the prevalence of BRAF mutations in sporadic early-onset colorectal cancer. (C) Forest plot illustrating meta-analysis of the prevalence of KRAS mutations in sporadic early-onset colorectal cancer. (D) Forest plot illustrating meta-analysis of the prevalence of NRAS mutations in sporadic early-onset colorectal cancer. (E) Forest plot illustrating meta-analysis of the prevalence of PIK3CA mutations in sporadic early-onset colorectal cancer. (F) Forest plot illustrating meta-analysis of the prevalence of TP53 mutations in sporadic early-onset colorectal cancer. Studies of Ak et al through Hamilton et al have presented microsatellite instability-high data, determined by PCR; studies of Aitchison et al through Suzuki et al have used mismatch repair immunohistochemistry. Details of molecular testing for each study are found in Table A4. CI, confidence interval; dMMR, deficient mismatch repair; ES, effect size (equivalent to proportion); MSI-H, microsatellite instability-high; PCR, polymerase chain reaction.

Figure 4

(A) Forest plot illustrating meta-analysis of the prevalence of microsatellite instability-high/deficient mismatch repair tumors in sporadic early-onset colorectal cancer. (B) Forest plot illustrating meta-analysis of the prevalence of BRAF mutations in sporadic early-onset colorectal cancer. (C) Forest plot illustrating meta-analysis of the prevalence of KRAS mutations in sporadic early-onset colorectal cancer. (D) Forest plot illustrating meta-analysis of the prevalence of NRAS mutations in sporadic early-onset colorectal cancer. (E) Forest plot illustrating meta-analysis of the prevalence of PIK3CA mutations in sporadic early-onset colorectal cancer. (F) Forest plot illustrating meta-analysis of the prevalence of TP53 mutations in sporadic early-onset colorectal cancer. Studies of Ak et al through Hamilton et al have presented microsatellite instability-high data, determined by PCR; studies of Aitchison et al through Suzuki et al have used mismatch repair immunohistochemistry. Details of molecular testing for each study are found in Table A4. CI, confidence interval; dMMR, deficient mismatch repair; ES, effect size (equivalent to proportion); MSI-H, microsatellite instability-high; PCR, polymerase chain reaction.

Figure 4

(A) Forest plot illustrating meta-analysis of the prevalence of microsatellite instability-high/deficient mismatch repair tumors in sporadic early-onset colorectal cancer. (B) Forest plot illustrating meta-analysis of the prevalence of BRAF mutations in sporadic early-onset colorectal cancer. (C) Forest plot illustrating meta-analysis of the prevalence of KRAS mutations in sporadic early-onset colorectal cancer. (D) Forest plot illustrating meta-analysis of the prevalence of NRAS mutations in sporadic early-onset colorectal cancer. (E) Forest plot illustrating meta-analysis of the prevalence of PIK3CA mutations in sporadic early-onset colorectal cancer. (F) Forest plot illustrating meta-analysis of the prevalence of TP53 mutations in sporadic early-onset colorectal cancer. Studies of Ak et al through Hamilton et al have presented microsatellite instability-high data, determined by PCR; studies of Aitchison et al through Suzuki et al have used mismatch repair immunohistochemistry. Details of molecular testing for each study are found in Table A4. CI, confidence interval; dMMR, deficient mismatch repair; ES, effect size (equivalent to proportion); MSI-H, microsatellite instability-high; PCR, polymerase chain reaction.