Genome-wide association assessment between immune cells and osteoarthritis: a bidirectional Mendelian randomization study.

BACKGROUND

The synovial immune microenvironment plays a critical role in the onset and advancement of osteoarthritis (OA), but previous findings on some immune cells were inconsistent. This study seeks to comprehensively investigate the causal association between a multitude of immune cell traits and OA.

METHODS

We performed this bidirectional Mendelian randomization (MR) analysis between a genome-wide association studies (GWAS) summary statistics containing 407,746 European ancestry and the largest GWAS data on 731 immune phenotypes. A replication analysis was conducted on a dataset containing 63,556 participants for validating the positive results. The causal effects were primarily estimated through inverse variance weighted (IVW) method, with four other methods (MR Egger, weighted median, simple mode, weighted mode) to reinforce the strength of causal evidence. Multiple sensitivity analyses (MR Egger, IVW method, leave-one-out analysis) were applied to mitigate the impact of heterogeneity and horizontal pleiotropy. Additionally, we employed a bioinformatics analysis by xCell algorithm to examine the expression of these immune cell phenotypes in OA and normal synovial tissues.

RESULTS

After false discovery rate (FDR) correction test, thirteen immune cell traits exhibited significant causal relationships with OA. These immune cell phenotypes came from seven groups, including B cell (n=3), conventional dendritic cell (cDC) (n=3), monocyte (n=3), myeloid cell (n=1), T cell, B cell, natural killer (NK) cell (TBNK) (n=2), regulatory T cell (Treg) (n=1). The strongest effects on OA were found in "CD64 on CD14 CD16 monocyte" [odds ratio (OR): 1.044; 95% confidence interval (CI): 1.012-1.076; P=0.03] and "CD16 monocyte %monocyte" (OR: 0.948; 95% CI: 0.916-0.980; P=0.009). Sensitivity analyses did not detect any evidence of heterogeneity and horizontal pleiotropy. We also identify five immune traits influenced by OA. Additionally, replication analysis reconfirmed the causal effect of "CD64 on CD14 CD16 monocyte" (OR: 1.102; 95% CI: 1.046-1.161; P<0.001) and "HLA DR NK %NK" (OR: 0.945; 95% CI: 0.908-0.983; P=0.03) on OA.

CONCLUSIONS

Our findings reveal the causal relationships between specific immune cells and OA, offering genetic insights into the role of immune cells in OA pathogenesis and guiding the exploration of novel immunological treatments for OA.