Individuals with Lynch syndrome (LS) are at increased risk for gastric cancer (GC) and duodenal cancer (DC), particularly those with specific mismatch repair (MMR) gene mutations or a family history of GC. However, the benefit of upper gastrointestinal surveillance with esophagogastroduodenoscopy (EGD) remains unclear, with heterogeneous recommendations among international guidelines. This study aimed to estimate the incidence of GC and DC in LS patients, assess the risk in high-risk subgroups, and evaluate the effect of EGD surveillance.

We conducted a systematic review and meta-analysis of cohort studies and clinical trials identified in PubMed and EMBASE (inception to April 2025). Cancer incidence rates per 1,000 person-years (p-y) were calculated overall and stratified by gene mutation and family history. When applicable, relative risks (RRs) and 95% confidence intervals (CIs) were computed to compare patients undergoing EGD surveillance with those without surveillance. Random-effects meta-analyses were performed, and heterogeneity was assessed using the I² statistic.

Thirty-four studies were included, comprising 36,402 LS patients. Among studies with calculable person-years, the pooled incidence rate of GC was 1.13 per 1,000 p-y (95% CI, 0.62–1.63) and 1.92 per 1,000 p-y (0.78–3.05) for DC. By MMR gene, GC incidence rates were 0.62 (MLH1), 1.25 (MSH2), and 0.12 (MSH6) per 1,000 p-y. In supplementary proportion-based analyses, overall GC and DC incidences were 2.71% (95% CI, 2.21–3.21) and 1.71% (95% CI, 1.30–2.12), respectively; GC incidences were 2.66% for MLH1, 4.05% for MSH2, 0.40% for MSH6, and 10.38% (95% CI, 3.25–17.51) in patients with a family history of GC. DC incidences by gene were 1.65% for MLH1, 1.75% for MSH2, and 0.64% for MSH6. In analyses by mode of diagnosis, incidence rates were higher in screen-detected than in symptom-detected cohorts (GC 1.35 vs 0.74 per 1,000 p-y; DC 2.70 vs 0.67 per 1,000 p-y). In sensitivity analyses, the pooled GC incidence was 0.76 per 1,000 p-y when restricting to studies with ≥250 participants and 0.73 per 1,000 p-y in prospective-only cohorts. In two comparative studies of surveillance, EGD was associated with a lower risk of both GC (RR 0.31; 95% CI, 0.14–0.69) and DC (RR 0.29; 95% CI, 0.16–0.52).

Patients with LS, particularly MSH2 carriers and those with a family history of GC, have an increased incidence of upper gastrointestinal cancers. EGD surveillance was associated with a lower incidence of both GC and DC in selected patients, although evidence is based on a limited number of comparative cohorts. Further prospective studies are needed to define optimal surveillance strategies.