Dissecting the causal role of immunophenotypes in primary sclerosing cholangitis risk: A Mendelian randomization study

Abstract

Primary sclerosing cholangitis (PSC), a chronic cholestatic liver condition, is frequently associated with inflammatory bowel disease. Specific immune cells have been implicated in PSC pathogenesis with the emergence of the “microbiota” and “gut lymphocyte homing” hypotheses, albeit their identities remain controversial. The first genome-wide association analysis leveraged nonoverlapping data from 3757 Europeans to evaluate 731 immunophenotypes. A genome-wide association analysis comprising 2871 cases and 12,019 controls yielded summary statistics for PSC. An inverse-variance weighted (IVW) analysis was performed to identify immunophenotypes causally related to PSC, and the results were validated using weighted mode, MR-Egger, and weighted median methods. Comprehensive sensitivity analyses were performed to verify the robustness, heterogeneity, and horizontal pleiotropy of the results. IVW analysis revealed 26 immune traits exhibiting causal associations with PSC. CD3 on HLA-DR+ CD4+ (IVW odds ratio [OR]: 0.904; 95% confidence interval [CI]: 0.828–0.986, P = .023) and CD3 on secreting Treg (IVW OR: 0.893; 95% CI: 0.823–0.969, P = .007) were negatively associated with PSC susceptibility and demonstrated high consistency across the 3 validation methods. Moreover, 7 other immune traits, including CD39+ resting Treg absolute cell (IVW OR = 1.083, 95% CI: 1.013–1.157, P = .019), CD39+ secreting Treg absolute cell (IVW OR = 1.063, 95% CI: 1.012–1.118, P = .015), CD3 on naive CD8br (IVW OR = 0.907, 95% CI: 0.835–0.986, P = .022), CD3 on CD39+ activated Treg (IVW OR = 0.927, 95% CI: 0.864–0.994, P = .034), CD28 on resting Treg (IVW OR = 0.724, 95% CI: 0.630–0.833, P = 5.95E-06), and CD39 on CD39+ CD4+ (IVW OR = 1.055, 95% CI: 1.001–1.112, P = .044) exhibited consistent results in the Weighted Median and Weighted Mode validation methods. Moreover, no significant heterogeneity or horizontal pleiotropy was observed across the single nucleotide polymorphisms. The leave-one-out results revealed that sequentially eliminating each single nucleotide polymorphism had no significant influence on model effect estimates or qualitative inference. This study evaluated potential causal links between 731 immune traits and PSC susceptibility. Twenty-six immune traits were identified using the IVW method. Verification across multiple methods revealed 9 immune traits with a plausible causal connection to PSC. These findings may uncover mechanistic pathways and novel therapeutic approaches.