<i>Helicobacter pylori</i> infection, treatment and colorectal cancer risk by genetic predisposition: evidence from two randomised trials

Background Previous studies have suggested a possible association between Helicobacter pylori ( H. pylori ) infection and an increased risk of colorectal cancer (CRC). Objective We examined the associations of H. pylori infection and H. pylori treatment with incident CRC risk and evaluated whether these associations were modified by genetic susceptibility. Design This study was based on two randomised trial cohorts: the Shandong Intervention Trial (SIT; n=3365; 1995–2024) and the Mass Intervention Trial in Linqu, Shandong Province (MITS; n=180 284; 2011–2024). Within MITS, we further conducted a case-cohort study to assess CRC risk according to seropositivity for H. pylori virulence factors and host genetic predisposition. Results Compared with H. pylori -negative individuals, H. pylori -positive individuals who did not receive antibiotic treatment had a significantly higher risk of CRC (SIT: HR=2.96, 95% CI 1.30 to 6.71; MITS: HR=1.27, 95% CI 1.04 to 1.55). The increased risk was particularly evident among individuals seropositive for four H. pylori -specific antigens (CagA, HpaA, Omp and HP0305) and those at high genetic risk (top decile of the Polygenic Risk Score). In SIT, H. pylori treatment was associated with a significantly reduced CRC risk over 29.4 years of follow-up (HR=0.47, 95% CI 0.22 to 0.99), with a greater reduction observed among participants with successful eradication (HR=0.38, 95% CI 0.15 to 0.94). In MITS, no overall benefit was observed 13.8 years after treatment (HR=1.17, 95% CI 0.95 to 1.43). However, a protective effect was apparent among individuals at high genetic risk and among those seropositive for key H. pylori -specific antigens. Conclusion In post hoc observational analyses of two established trial cohorts, H. pylori infection was associated with an increased risk of CRC. H. pylori treatment reduced CRC risk in SIT, whereas in MITS, the benefit appeared to be limited to individuals at high genetic risk or those infected with specific virulence factor subtypes.