Non-bleeding Adverse Events with the Use of Direct Oral Anticoagulants: A Sequence Symmetry Analysis

Introduction Postmarketing pharmacovigilance reports have raised concerns about non-bleeding adverse events associated with direct oral anticoagulants (DOACs), but only limited results are available from large claims databases. Objective The aim of this study was to assess the potential association between DOAC initiation and the onset of four types of non-bleeding adverse events by sequence symmetry analysis (SSA). Methods SSA was performed using nationwide data from the French National Healthcare databases (Régime Général, 50 million beneficiaries) to assess a cohort of 386,081 DOAC new users for the first occurrence of four types of non-bleeding outcomes: renal, hepatic, skin outcomes identified by using hospitalization discharge diagnoses, and gastrointestinal outcomes by using medication reimbursement. Asymmetry in the distribution of each investigated outcome occurring before and after initiation of DOAC therapy was used to test the association between DOAC therapy and these outcomes. SSA inherently controls for time-constant confounders, and adjusted sequence ratios were computed after correcting for temporal trends. Negative (glaucoma) and positive (bleeding, depressive disorders) control outcomes were used and analyses were replicated on a cohort of 310,195 patients initiating a vitamin K antagonist (VKA). Results This study demonstrated the expected positive association between either DOAC or VKA therapy and hospitalised bleeding and initiation of antidepressant therapy, while no association was observed between either DOAC or VKA therapy and initiation of antiglaucoma medications. For DOAC therapy, signals were the associations with hepatic outcomes, including acute liver injury [for the 3-month time window, aSR3 = 2.71, 95% confidence interval (CI) 1.79–4.52]; gastrointestinal outcomes, including initiation of drugs for constipation and antiemetic drugs (aSR3 = 1.31, 95% CI 1.27–1.36; and 1.17, 95% CI 1.12–1.22, respectively); and kidney diseases (aSR3 = 1.33, 95% CI 1.29–1.37). Conclusion Results of this nationwide study suggest that DOACs are associated with rare but severe liver injury and more frequent gastrointestinal disorders. A low risk of kidney injury with DOAC therapy can also not be excluded.