Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Pharmaceutical Sciences


Biomedical and Pharmaceutical Sciences

First Advisor

Xuerong Wen


The introduction of direct oral anticoagulants (DOACs) in the last decade have greatly expanded available treatment options for the management of thromboembolic diseases. Based on their non-inferiority in efficacy and safety, and more convenient dose monitoring and adjustment compared to warfarin, DOACs are now considered the first-line treatment option in the management and prevention of venous thromboembolism (VTE) and nonvalvular atrial fibrillation (NVAF). While the use of DOACs continues to increase in the general population, limited evidence exists to support its use in chronic liver disease (CLD). Of note, individuals with elevated liver enzymes, liver cirrhosis or CLD were excluded from the pivotal randomized clinical trials (RCTs) for DOACs. There is also a paucity of large, population-based comparative effectiveness and safety studies of DOACs and warfarin in individuals with indication for oral anticoagulation (NVAF or VTE) and CLD. To this end, we conducted two studies to assess the comparative effectiveness and safety of DOACs (individually and as a class) and warfarin in patients with CLD and NVAF (Aim 1) or VTE (Aim 2).

Oral anticoagulants interfere with the normal hemostatic process, inadvertently increasing the risk of bleeding. Consequently, among individuals with NVAF on oral anticoagulants, bleeding is the most prevalent major treatment-related adverse events, with gastrointestinal bleeding (GIB) the most common. Individuals who survive a major bleeding event while on oral anticoagulant treatment face the competing risks of stroke (if oral anticoagulation is discontinued), recurrent bleeding (if oral anticoagulation is restarted), and mortality. The general consensus in review articles and contemporary guidelines is to restart oral anticoagulation (often within a month) after hospitalization for major GIB. However, there is limited comparative effectiveness and safety data to guide the selection of DOACs versus warfarin after major GIB. To address this evidence gap, in Aim 3, we evaluated the effectiveness and safety for DOACs versus warfarin in individuals with NVAF who restarted oral anticoagulant treatment after surviving a major GIB event.

These three aims are presented in a manuscript format, with three distinct manuscripts each containing an abstract, introduction, methods, results, discussion and conclusion section. All three aims were conducted using 2010 – _2017 data submitted to the Optum Clinformatics database - a deidentified claims data of beneficiaries of commercial and Medicare Advantage health insurance plans throughout the US.

Manuscript 1. In this retrospective cohort study, 10,209 subjects with NVAF and CLD initiated oral anticoagulation with DOACs (apixaban, dabigatran, edoxaban or rivaroxaban) or warfarin during the study period. The primary endpoint was net adverse clinical outcome (NACE); a composite of ischemic stroke, systemic embolism (SE), major bleeding and all-cause mortality. Compared to warfarin, DOACs users had 24% reduced risk of NACE (HR [95% CI]: 0.76 [0.66, 0.82]), suggesting that larger net clinical benefit associated with DOACs compared to warfarin among patients with NVAF and CLD. In exposure to specific DOAC versus warfarin comparisons, and pairwise comparisons of DOACs, we observed comparable effectiveness, but differential safety across DOACs. Overall, apixaban was associated with consistently larger clinical benefit compared with warfarin, and in head-to-head DOAC comparisons with rivaroxaban or dabigatran. These findings on varying magnitude of clinical benefit between oral anticoagulants may help inform treatment choice, particularly among patients with NVAF and CLD who are at higher risk of bleeding.

Manuscript 2. We identified 8477 patients with VTE and CLD that initiated oral anticoagulation with DOACs or warfarin during the study period. Compared to warfarin, DOACs users had significantly reduced risk of the composite of recurrent VTE, all-cause mortality, and clinically relevant bleeding (HR: 0.81; 95% CI: 0.74, 0.88). The findings of larger clinical benefit with DOACs versus warfarin remained largely unchanged by dose (reduced or standard doses of DOACs), and among individuals with cirrhosis or decompensated cirrhosis. The magnitude of risk reduction for the composite and individual clinical outcomes however varied between DOACs, with consistently lowest risks observed with apixaban when compared to warfarin or rivaroxaban.

Manuscript 3. We identified 4,389 patients with NVAF hospitalized for major GIB while receiving oral anticoagulants. During the follow-up period, 1116 subjects had NACE, with 288 (35.2 per 100 person-years) and 828 (45.8 per 100 person-years) in the DOAC and warfarin treatment groups respectively. The weighted hazards ratio (HR) among individuals that resumed DOACs compared to warfarin after major GIB was 0.89 (95% CI: 0.76, 1.05) for NACE; 0.84 (95% CI: 0.51, 1.40) for ischemic stroke/SE; 0.89 (95% CI: 0.74, 1.06) for all-cause mortality; 0.75 (95% CI: 0.60, 0.94) for major bleeding, and 0.76 (95% CI: 0.60, 0.98) for recurrent major GIB. Therefore, among individuals that survived and resumed oral anticoagulation within the first 90-days after hospitalization for a major GIB event, restarting oral anticoagulation treatment with DOACs was associated with lower risk of the composite outcome of ischemic stroke, systemic embolism, all-cause mortality, recurrent or incident major bleeding. Future research is needed to confirm these findings.

Available for download on Friday, May 17, 2024