Association between Genetic Polymorphisms and Bleeding in Patients on Direct Oral Anticoagulants

Abstract

Objectives: The purpose of our study is to investigate the effects of apolipoprotein B (APOB) and APOE gene polymorphisms on bleeding complications in patients receiving direct oral anticoagulants (DOACs). Methods: A total of 16 single nucleotide polymorphisms (SNPs) in 468 patients were genotyped. Six SNPs of ABCB1 (rs3842, rs1045642, rs2032582, rs1128503, rs3213619, and rs3747802), one SNP of CYP3A5 (rs776746), seven SNPs of APOB (rs1042034, rs2163204, rs693, rs679899, rs13306194, rs13306198, and rs1367117), and two SNPs of APOE (rs429358 and rs7412) were analyzed by a TaqMan genotyping assay. Multivariable logistic regression analysis with selected variables was performed for the construction of a risk scoring system. Two risk scoring systems were compared (demographic factors only vs. demographic factors and genetic factors). Results: In the multivariable analyses, two models were constructed; only demographic factors were included in Model I and both demographic factors and genetic factors in Model II. Rivaroxaban and anemia showed significant association with bleeding in both models. Additionally, ABCB1 rs3842 variant homozygote carriers (CC) and APOB rs13306198 variant allele carriers (AG, AA) had a higher risk of bleeding risk compared with that of wild-type allele carriers (TT, TC) and wild-type homozygote carriers (GG), respectively. Whereas the area under the receiver operating characteristic curve (AUROC) value using demographic factors only was 0.65 (95% confidence interval (CI): 0.56-0.74), the AUROC increased to 0.72 by adding genetic factors (95% CI: 0.65-0.80). The predicted bleeding risks of bleeding in patients with 0, 1, 2, 3, 4, 5, 6, 7 and 8 points from the logistic regression curve were 0.8%, 2.0%, 5.4%, 5.2%, 12.5%, 26.9%, 47.0%, 64.3% and 82.3%, respectively. Conclusions: The study results can be used for enhancing individualized treatment strategies in patients taking DOACs, helping clinicians predict the bleeding risk.