Genomic Medicine in the Developing World: Cancer Spectrum, Cumulative Risk and Survival Outcomes for Lynch Syndrome Variant Heterozygotes with Germline Pathogenic Variants in the MLH1 and MSH2 Genes

Abstract

Background: Although genetic testing has improved our ability to diagnose Lynch syndrome (LS), there is still limited information on the extent of variations in the clinical and genetic landscape among LS variant heterozygotes (LSVH) in Africa. We sought to investigate the cancer spectrum, cumulative risk, and survival outcomes of LSVH with pathogenic/likely pathogenic variants (P/LPVs) in the MLH1 and MSH2 genes using a LS registry in South Africa over the last 30 years. Methods: A retrospective study was conducted to retrieve demographic, clinical, and genetic data of all LSVH with P/LPVs in the MLH1 and MSH2 genes from our LS registry. Genetic data were analyzed according to cancer spectrum, cumulative risk, and crude survival. We used the Chi-squared and t-test to assess differences between groups, and Kaplan-Meier survival analyses were used to analyze the cumulative risk and crude survival outcomes. A p-value < 0.05 at a 95% confidence interval was considered statistically significant. Results: We analyzed a total of 577 LSVH from 109 families. About 450 (78%) and 127 (22%) LSVH harbored a disease-causing mutation in MLH1 and MSH2, respectively. A South African founder PV (MLH1:c.1528C>T) accounted for 74% (n = 426) of all LSVH. CRC was the most common diagnosed cancer in both MLH1 and MSH2 LSVH. MLH1 LSVH had a younger age at cancer diagnosis than MSH2 LSVH (43 vs. 47 years, respectively, p = 0.015). Extracolonic cancers were predominantly higher in female LSVH (n = 33, 35%) than in male LSVH (n = 8, 7%) with the MLH1:c.1528C>T founder PV. The cumulative risk of any cancer and CRC at any age was higher in MLH1 LSVH than in MSH2 LSVH (p = 0.020 and p = 0.036, respectively). LSVH with the MLH1:c.1528C>T PV had a better 10-year overall survival after the first cancer diagnosis, particularly for CRC. Conclusions: LSVH with P/LPVs in the MLH1 and MSH2 genes exhibited significant gene- and sex-specific differences in cancer spectrum, cumulative risk and survival outcomes. Cancer risk and survival estimates described in this study can be used to guide surveillance and genetic counselling for LSVH in our population.

Keywords: Lynch syndrome registry; South African founder variant; colorectal cancer; crude survival; cumulative risk; extracolonic cancer; genetic variants; pathogenic variants.

Figure 1

A comprehensive description of our LS registry and patient recruitment. (A) The flowchart outlines a step-by-step structured process for managing and categorizing individuals suspected or confirmed to have LS. It begins with the “Initial Screening” based on clinical criteria or family history (Bethesda and Amsterdam criteria I/II for LS), followed by Risk Assessment & Genetic Counselling. Individuals meeting eligibility criteria are referred for Genetic Testing, leading to three distinct outcomes: Positive for LS, Variants of Uncertain Significance (VUS), or Negative for LS. Each outcome has specific downstream pathways for inclusion in the LS registry, follow-up testing, or standard clinical surveillance, respectively. The flowchart also details strategies for Family Cascade Testing, comprehensive management, and research integration. (B) A flow diagram showing the number of families and individuals who were genetically diagnosed with LS in our registry, categorized by the specific MMR gene with germline P/LPVs. Note: While families and individuals with MSH6 and PMS2 P/LPVs are shown here for completeness, they were excluded from further analyses due to small number of cases.

Figure 2

Distribution and clinical characteristics of P/LPVs in the MLH1 gene. (A) The exon-intron structure of the MLH1 gene, highlighting functional domains: ATPase domain (exons 1–5, red), MutS interaction domain (exons 6–11, yellow), and PMS2/MLH3/PMS1 interaction domain (exons 13–19, gray). Symbols represent different types of variants: intronic changes (●), missense mutations (■), stop codon/frameshift mutations (▲), and deletions (—). (B) A description table provides a detailed summary of nucleotide and protein changes for each P/LPV, their corresponding exons, and the number of individuals carrying these PLPVs. Exon 13 shows the highest variant frequency, with 426 LSVH with the PV c.1528C>T (p.Gln510Ter). The associated cancer phenotypes indicate a predominance of CRC among LSVH, while extracolonic cancer incidence remains low.

Figure 3

Distribution and clinical characteristics of P/LPVs in the MSH2 gene. (A) A panel displays the exon–intron structure of the MSH2 gene, illustrating its functional domains: DNA binding domain (exons 1–6, red), MSH3/MSH6 interaction domain (exons 7–12, yellow), and MutL interaction domain (exons 13–16, gray). Different variant types are represented by specific symbols: missense mutations (■), stop codon/frameshift mutations (▲), and deletions (—). (B). A table details the nucleotide and protein changes for each PV, the specific exons affected, and the number of LSVH. Exon 7 shows the highest variant frequency, with 17 LSVH with the PV c.1219delC (p.Leu407SerfsTer5). The associated cancer phenotypes indicate a predominance of CRC among LSVH, while extracolonic cancer incidence remains low.

Figure 4

Pie charts showing the distribution of first cancers in LSVH according to sex. (A) The spectrum of first cancers in female LSVH (n = 93), where CRC was the predominant first cancer type, accounting for 65% of all first cancers. (B) The distribution in male LSVH (n = 111) with a dominance of CRC (93%) as the first cancer diagnosed, with limited occurrence of other cancers.

Figure 5

Box plots and whiskers showing age at CRC diagnosis in prevalent and incident cases stratified by mutated gene status. (A) In Group 1A, the prevalent cases were diagnosed with cancer at a significantly younger age compared to the incident cases. (B) In Group 2, there was no significant difference in the age at cancer diagnosis between the prevalent and incident cases. (C) In Group 1B, the prevalent cases were diagnosed with cancer significantly at a younger age compared to incident cases.

Figure 6

Cumulative risk of LS-associated cancers by gene and sex. (A–C) The cumulative risk curves comparing MLH1 (black) and MSH2 (red) LSVH. (D–F) Cumulative risk of LS-associated cancers by sex (male = red, female = black). (G–I) Results in Group 1B, showing similar sex-specific trends.