Mendelian randomization analysis reveals causal effects of inflammatory bowel disease and autoimmune hyperthyroidism on diffuse large B-cell lymphoma risk

Abstract

The clinical phenomenon whereby diffuse large B-cell lymphoma (DLBCL) occurs in patients with a history of autoimmune disease (AD) has been noted, but it remains controversial. This study aimed to evaluate the causal associations between nine ADs and DLBCL via a Mendelian randomization (MR) study. Single-nucleotide polymorphism (SNP) obtained from published genome-wide association studies (GWAS) was chosen as instrumental variable (IV). A total of nine ADs of European ancestry including asthma (56,167 cases and 352,255 controls), psoriasis (4,510 cases and 212,242 controls), autoimmune hyperthyroidism (962 cases and 172,976 controls), inflammatory bowel disease (31,665 cases and 33,977 controls), type 1 diabetes (6,683 cases and 12,173 controls), multiple sclerosis (14,498 cases and 24,091 controls), sarcoidosis (2,046 cases and 215,712 controls), ankylosing spondylitis (9,069 cases and 1,550 controls), and celiac disease (12,041 cases and 12,228 controls), were set as the exposure and DLBCL (209 cases and 218,583 controls) of European ancestry as the outcome. Inverse-variance weighted (IVW) was used as the primary analysis method, and the weighted median and MR-Egger method were used as supplementary methods. The sensitivity analyses employed in this study include the MR-Egger intercept, MR-PRESSO global test, Cochran’s Q test, leave-one-out analysis, and funnel plot. IVW showed that inflammatory bowel disease (OR = 1.241, 95% CI 1.009–1.526, P = 0.040) and autoimmune hyperthyroidism (OR = 1.464, 95% CI 1.103–1.942, P = 0.008) increased the risk of DLBCL without significant heterogeneity or horizontal pleiotropy, and the results remained stable according to the leave-one-out analysis. The IVW results revealed no associations between the other seven ADs and DLBCL: asthma (OR = 0.782, 95% CI 0.395–1.546, P = 0.159), psoriasis (OR = 0.842, 95% CI 0.669–1.060, P = 0.143), type 1 diabetes (OR = 1.071, 95% CI 0.860–1.334, P = 0.537), multiple sclerosis (OR = 1.331, 95% CI 0.941–1.883, P = 0.105), sarcoidosis (OR = 1.324, 95% CI 0.861–2.038, P = 0.200), ankylosing spondylitis (OR = 1.884, 95% CI 0.776–4.573, P = 0.161), and celiac disease (OR = 1.003, 95% CI 0.854–1.178, P = 0.969). Although no significant heterogeneity or horizontal pleiotropy was detected in these seven ADs and DLBCL, these results did not pass the leave-one-out analysis; therefore, the results need to be interpreted with caution. Inflammatory bowel disease and autoimmune hyperthyroidism may increase the onset of DLBCL. The risk of DLBCL should be considered in specific types of ADs.