Direct Oral Anticoagulants vs. Warfarin When Treating Renal Dysfunction
July 1, 2023
By Michael H. Crawford, MD
Professor of Medicine, Lucy Stern Chair in Cardiology, University of California, San Francisco
SYNOPSIS: A patient-level meta-analysis of five pivotal trials of direct oral anticoagulants (DOACs) vs. warfarin for patients with atrial fibrillation and creatinine clearance levels ranging from normal to 25 mL/min revealed standard-dose DOACs exhibit superior safety and efficacy compared to adjusted-dose warfarin and lower-dose DOACs.
SOURCE: Harrington J, Carnicelli AP, Hua K, et al. Direct oral anticoagulants versus warfarin across the spectrum of kidney function: Patient-level network meta-analyses from COMBINE AF. Circulation 2023; Apr 12. doi: 10.1161/CIRCULATIONAHA.122.062752.[Online ahead of print].
Patients with kidney dysfunction are at higher risk of developing atrial fibrillation (AF), thromboembolic events, and bleeding, which makes anticoagulation decisions difficult. Since all direct-acting oral anticoagulants (DOACs) are cleared at least partially by the kidney, they are used less frequently or are underdosed in patients with kidney dysfunction.
The Collaboration Between Multiple Institutions to Better Investigate Non-Vitamin K Antagonist Oral Anticoagulant Use in AF (COMBINE AF) is a database of individual data on 71,683 patients from five pivotal randomized trials of DOACs (apixaban, dabigatran, edoxaban, rivaroxaban) vs. warfarin in AF, which includes 24,396 patients with a creatinine clearance (CrCl) less than 60 mL/min. Harrington et al used these data to perform a network meta-analysis of the safety and efficacy outcomes of DOACs and warfarin across a spectrum of kidney function down to a CrCl of 25 mL/min.
These studies involved standard-dose DOAC, lower-dose DOAC, and adjusted warfarin doses. Efficacy outcomes for these studies were stroke or systemic emboli (SSE) and all-cause mortality. Safety outcomes were intracranial hemorrhage (ICH) and major bleeding. For the efficacy outcomes, the authors used the intention-to-treat approach. Safety outcomes used the criterion of receiving at least one dose of the study drug. The mean age in the total population was 71 years, 37% were women, and mean CrCl levels were 76 mL/min. The median patient follow-up time was 23 months.
The incidence of SSE, mortality, ICH, and major bleeding increased significantly with declining kidney function. There was no difference in the risk of major bleeding between DOAC- and warfarin-treated patients (P = 0.61). Standard-dose DOAC use exhibited a significantly lower rate of SSE compared to warfarin (5% decrease in HR per 10 mL/min decrease in CrCl; P = 0.01). Standard-dose DOAC also showed a lower risk of ICH compared to warfarin (6% decrease in HR per 10 mL/min decrease in CrCl; P = 0.08) and a lower risk of death (2% decrease in HR per 10 mL/min decrease in CrCl; P = 0.08). Using a lower-dose DOAC was associated with a higher incidence of death and SSE compared to warfarin, without any difference in major bleeding or ICH in patients with worse kidney function.
The authors concluded standard-dose DOAC is more effective and safer than adjusted-dose warfarin with CrCl levels down to 25 mL/min. Lower-dose DOACs did not lower the risk of major bleeding or ICH compared to standard-dose DOACs, but were associated with a higher incidence of SSE and mortality.
COMMENTARY
Researchers conduct meta-analyses on groups of randomized trials when individual randomized trials of a particular hypothesis are not robust, usually because of few subjects. When large, randomized trials have been subsequently performed, their agreement with meta-analyses has been mixed. Exceptions include burning clinical questions whereby a large, randomized trial in the future is highly unlikely. Anticoagulation in AF patients with kidney dysfunction is such an issue. Also, most meta-analyses are conducted using the group data from the included studies because patient-level data were unavailable. The COMBINE AF meta-analysis authors used patient-level data of high quality in the five pivotal trials of DOACs vs. warfarin in patients with AF. Therefore, calling attention to this study is worthwhile.
In COMBINE AF, one-third of patients recorded reduced CrCl levels down to 25 mL/min. These patients were older, with lower body weight. They were more likely to be women and have experienced heart failure, have been diagnosed with coronary artery disease, and have presented with a history of bleeding. Also, these patients recorded higher CHA2DS2-VASc scores, used more antiplatelets, were living with more permanent or persistent AF, and were living with less paradoxical AF. These are the patients in whom concern about the risk of bleeding events on a DOAC occasions thoughts of using adjusted-dose warfarin or lower doses of a DOAC. Thus, it is not surprising some of the five studies in COMBINE AF included lower DOAC doses.
Surprisingly, in COMBINE AF, the benefit of standard-dose DOAC was amplified as kidney function declined. This is especially noteworthy since the incidence of adverse events increased as kidney function declined. These data prompted the authors to argue it is inappropriate and perhaps even dangerous to prescribe lower doses of DOACs for AF patients unless they meet certain criteria stated by the manufacturers and approved by the Food and Drug Administration. For example, with apixaban, the patient must meet the following three criteria: Cr less than 1.5 mg/dL, body weight < 60 kg, and age older than 80 years. If Cr is greater than 1.5 mg/dL, then the patient must meet one of the age or weight criteria. However, a CrCl level lower than 15 mL/min is a data-free zone.
Despite its utility, there were limitations to COMBINE AF. For example, the authors used baseline CrCl levels for the analyses, but they did not consider any changes in CrCl over time. Also, there were relatively few patients with a CrCl level lower than 25 mL/min. However, this was a robust meta-analysis that included randomized, individual patient data, where CrCl level was a continuous variable rather than categorical. This analysis supports using standard-dose DOACs for patients with a CrCl level down to 25 mL/min rather than a reduced-dose DOAC or adjusted-dose warfarin.