Anticoagulation Management in Patients on Hemodialysis with Atrial Fibrillation More Questions than Answers

Is atrial fibrillation (AF) in patients on dialysis an actual effector of cardioembolic events, or is it a surrogate marker for cardiovascular disease?

Overall, direct oral anticoagulants (DOACs) have a superior benefit to a risk profile compared with vitamin K antagonists (VKAs), such as warfarin. When it comes to patients on hemodialysis (HD), however, the confusion lies in which, if any, anticoagulants are appropriate. Patients receiving maintenance HD have a high incidence of stroke, which typically warrants the use of anticoagulation. However, patients on HD also have an increased risk of bleeding because they are routinely heparinized three times each week and have platelet dysfunction (1). A meta-analysis of 13 studies among patients on HD reported a stroke rate of 5.2 per 100 patient-years with AF versus 1.9 per 100 patient-years without AF (2). This suggests that AF is a risk factor in patients on HD. Interestingly, in a cohort of 1382 patients on HD, AF was not significantly associated with new stroke (3). Potential reasons for this are the high competing risk of mortality in patients on HD, a possible protective effect of heparin administration during the dialysis procedure, and the high prevalence of subclinical AF in the “no AF” cohort in observational studies of patients on HD (4). As a result, the use of anticoagulation in patients on HD with AF remains controversial.

Although VKAs are the mainstay therapy for thromboembolic issues, there is a paucity of evidence to support a reduction of risk of stroke in patients on HD. Considering this and the increased risk of bleeding and calciphylaxis, the use of VKA in patients on HD should be questioned (4, 5). On the other hand, the dependence of DOACs on kidney clearance, bioavailability, and bleeding risk is a factor to consider when treating patients on HD with AF (6).

One major advantage of using direct oral agents…is that there is no need for measurement of INR levels or specialdietary restrictions.

Should patients with kidney failure on HD with AF receive anticoagulants?

In a meta-analysis of 12 cohort studies comprising over 17,000 patients on HD with AF, VKAs had a non-significant (26%) reduction of ischemic stroke rate, no effect on total mortality, a 21% increase in total bleeding risk, and a doubling of the incidence of hemorrhagic stroke (7). On the other hand, in a meta-analysis of 15 studies that had more than 47,000 patients on HD with AF, the use of VKAs did not reduce ischemic stroke or all-cause mortality and increased risk of hemorrhagic stroke but did not affect overall risk of bleeding (8).

Overall, when compared with no anticoagulation, neither VKAs nor the DOAC apixaban (standard and reduced dose) were associated with a reduction in stroke or systemic embolism, although heterogeneity was high (9). In summary, the data are mixed, and we are still left with more questions than answers for stroke reduction using anticoagulation in patients on HD with AF.

Several randomized controlled trials assessing stroke and bleeding risk of oral anticoagulants versus VKAs or no anticoagulation in patients on HD with AF are currently ongoing and can provide better answers to this question (NCT02933697: AXADIA, NCT03987711: SAFE-D, and NCT03969953: TRACK).

Patients with AF on apixaban, a factor Xa inhibitor, qualify for either standard or reduced dosing. The standard dose for stroke risk reduction in patients with non-ventricular AF (NVAF) is typically 5 mg twice a day. A reduced dose of apixaban—2.5 mg twice daily—is warranted for patients meeting two of the three following criteria: ≥80 years old, serum creatinine ≥1.5 mg/dL, or weight ≤60 kg. The rationale for this dose adjustment is the greater risk of bleeding and mortality in patients with NVAF and in patients with at least two of the mentioned dose-adjustment criteria compared with patients with one or fewer of the criteria (10).

When it comes to bleeding events, the risk of fatal or intracranial bleeding increased in patients on apixaban (4.9 events/100 patient-years) versus those who received no anticoagulation (1.6 events/100 patient-years); this was true for apixaban 5 mg but not 2.5 mg twice a day. Apixaban at 2.5 mg twice a day dosing had a higher rate of myocardial infarction or ischemic stroke versus no anticoagulation. Apixaban resulted in lower all-cause mortality compared with those receiving no anticoagulation (11). Mortality risk is lowest with apixaban 5 mg compared with VKA, apixaban 2.5 mg, and no anticoagulation (11).

A study of 124 patients on HD found less bleeding in apixaban than VKAs (12). In a larger study of more than 25,000 patients on chronic HD with AF, 2351 were taking apixaban (44% on 5 mg twice a day and 56% on 2.5 mg twice a day), and 23,172 were taking VKAs; the risk of stroke and intracranial bleeding was identical between both agents. Apixaban showed fewer major bleeding events and a non-significant trend toward reduced mortality (13). At a dose of 5 mg, apixaban resulted in less major bleeding, lower risk of stroke, and a non-significant trend toward reduced mortality compared with VKAs. At a dose of 2.5 mg twice a day, there was a lower risk of bleeding without differences in stroke and death (13). Finally, a meta-analysis of five studies comprising more than 43,000 patients (combined chronic kidney disease and end stage kidney disease [ESKD]) demonstrated that apixaban was associated with lower risk of bleeding but similar risk of thromboembolic events when compared with VKA (14). Although the role of DOACs in patients on HD with AF remains ambiguous, apixaban is approved for use in dialysis and is a feasible alternative to VKA.

Data on the use of rivaroxaban for stroke risk reduction in patients with kidney failure and AF are limited. According to a study investigating the use of DOACs in HD patients with AF, the risk of bleeding is increased in DOACs, such as dabigatran and rivaroxaban, compared with apixaban (15). In addition, risk of hemorrhagic stroke was significantly lower in patients on dabigatran or rivaroxaban compared with VKAs, despite an overall increased bleeding risk (16). Only 33% of rivaroxaban is eliminated by the kidney with minimal dialyzability due to high protein binding. Currently, rivaroxaban can be used at a reduced dose of 15 mg in patients with reduced kidney function (creatinine clearance [CrCl] ≤50 mL/min). There are limited data for use of rivaroxaban in patients with kidney failure (15). Despite this, the renally impaired population is being exposed to both rivaroxaban and dabigatran. Further research is necessary to make a recommendation.

The Valkyrie study (5) looked at 132 patients on HD with AF who were randomized to VKAs with a target international normalized ratio (INR) of 2–3, rivaroxaban 10 mg daily, or rivaroxaban plus vitamin K2 for 18 months. The incidence of fatal and non-fatal cardiovascular events and of symptomatic limb ischemia was higher in VKAs than both rivaroxaban groups. Furthermore, death from any cause, cardiac death, and risk of stroke were not different between the groups. Life-threatening or major bleeding adjusted for competing risk of death was increased in VKAs compared with rivaroxaban (5). Overall, in patients on HD with AF, a lower dose of rivaroxaban (10 mg) showed fewer outcomes of fatal and non-fatal cardiovascular events and major bleeding complications compared with VKA (5). Trials are underway to reach more definitive conclusions.

What doses are used for direct oral agents in patients on HD?

Only rivaroxaban and apixaban are suitable for patients undergoing maintenance HD, as they have the least dependence on kidney clearance and are not substantially eliminated by HD. A pharmacokinetic study found that a 10-mg dose of rivaroxaban in HD patients without residual kidney function results in drug exposure, similarly as published for 20 mg in healthy volunteers (17). Table 1 lists the various anticoagulants used for AF and their dosage adjustments in HD patients.

Table 1

Comparison of various anticoagulants available for atrial fibrillation and their use in patients on hemodialysis

View Table

Apixaban is approved for use in patients with ESKD. Dosing recommendations for apixaban are derived from limited studies. In one study, patients with ESKD received a one-time dose of apixaban 5 mg, resulting in 36% higher area under the curve (AUC) and no increase in the maximum concentration (Cmax) of the drug compared with healthy subjects (18). In addition, levels taken after the HD session show a 13% and 14% reduction of Cmax and AUC for apixaban, respectively (18). However, in another pharmacological study of apixaban in patients on HD, they received 5 mg twice daily, which resulted in supra-therapeutic levels that should be avoided. Meanwhile, the reduced dose (2.5 mg twice a day) in patients on HD was comparable with the standard dose (5 mg twice a day) in patients with normal kidney function (19). This suggests that apixaban 2.5 mg twice a day may be a viable dose alternative for patients on HD. The Valkyrie study (5) does suggest safety of using rivaroxaban in patients on HD with AF at a 10-mg dose, but more data are needed to confirm the use of this agent in patients on HD.

In summary, the data surrounding the use of oral anticoagulation in patients on HD with AF are challenging to interpret. This is because no randomized clinical trial has definitively shown that oral anticoagulants provide protection against stroke, whereas a substantial amount of evidence reveals a significantly increased bleeding risk.