Warfarin and Stroke Outcomes in Hemodialysis Patients with Atrial Fibrillation

The prevalence of atrial fibrillation in ESRD is extremely high, reaching 27%,1 and fibrillation, as in the general population, also is associated with increased mortality in hemodialysis patients.2 A large number of trials show the usefulness of oral anticoagulation therapy1 for primary and secondary prevention of stroke in patient populations with atrial fibrillation absent ESRD.3 Recently, a large study demonstrated the superiority of oral anticoagulation therapy compared with the combination of clopidogrel plus aspirin with regard to stroke prevention, with no added risk of bleeding.4 Even trials performed in patients with high hemorrhagic risk who took warfarin, particularly the elderly, show that benefits of treatment exceed the risks when the international normalized ratio (INR) is monitored correctly. The decision to use oral anticoagulation therapy, particularly warfarin, in patients with atrial fibrillation involves weighing the risk of a thromboembolic event without therapy, or with inadequate anticoagulation, against the risk of a hemorrhagic event on therapy, particularly over-anticoagulation. Efficacy and safety of anticoagulation in atrial fibrillation depend on maintaining the INR between 2 and 3,5 as recommended by most practice guidelines.6 Recently an INR of 3.0 to 3.4 has been proposed to achieve optimal anticoagulation intensity in patients with atrial fibrillation.7 However, dialysis populations are different from the general population. The association between renal dysfunction and bleeding has long been recognized, even as long as 200 yr ago,8 and morbidity and mortality from bleeding remain a significant clinical problem in ESRD. Impaired platelet function is one of the main determinants of uremic bleeding. This impairment is multifactorial and includes defects that are intrinsic to platelets and abnormal platelet–endothelial interactions. Uremic toxins and anemia also play a role.9 Moreover, hemodialysis patients, unlike other patient settings, are exposed to continuous anticoagulation with heparin during the course of hemodialysis. Therefore, the problem of how to anticoagulate dialysis patients with atrial fibrillation is a truly huge concern. In this issue of JASN, Chan et al.10 report an increased risk of stroke (presumably hemorrhagic) associated with warfarin use in dialysis patients with atrial fibrillation. The authors stress the risk of unmonitored anticoagulation in this subpopulation and caution the careful use of warfarin in these patients. Although the study was performed in a large number of patients, it contains various limitations, as do all association studies—association studies are hypothesis generating: There is no formal documentation of arrhythmias (and thus a correct indication for anticoagulation therapy); in many cases, the exact nature of the strokes (ischemic or hemorrhagic) is not identified; the number of patients on warfarin treatment without INR monitoring is high; and the INR value at the moment of the stroke is often unknown. All of these factors limit the strength of the results and the conclusions of the work. However, the study confirms the importance of identifying patients at special risk for warfarin-associated stroke to optimize the risk/benefit of anticoagulation therapy. Risk of ischemic stroke in the general population is best estimated with the CHADS2 score (congestive heart failure, hypertension, age ≥75 yr, diabetes, 1 point each; prior stroke or transient ischemic attack, 2 points). For patients with atrial fibrillation and a CHADS2 score ≥2, anticoagulation with warfarin is recommended, although the sensitivity of this scoring system for dialysis populations is unclear.11 Access to high-quality monitoring of anticoagulation is crucial in the decision to use warfarin. The use of warfarin without adequate monitoring is more dangerous than the choice not to use it, even in the presence of increased thromboembolic risk, and this seems to apply particularly to patients with ESRD. It should be emphasized that dialysis patients already demonstrate an ischemic or hemorrhagic stroke prevalence that is much higher than that of the general population, and the magnitude of this excess stroke risk is greater for ischemic stroke than that for hemorrhagic stroke.12,13 Thus, it is also difficult to understand which percentage of strokes reported by Chan et al.10 really had a thromboembolic cause, perhaps preventable by proper warfarin therapy, and which percentage of the hemorrhagic strokes would have occurred regardless of such treatment. Nevertheless, the study by Chan et al.10 poses another important question that nephrologists have to deal with frequently in light of the growing importance of cardiovascular comorbidities in hemodialysis patients: Are the anticoagulation guidelines for the general population at risk for stroke applicable to patients who concomitantly suffer from cardiovascular disease and ESRD? The Chan et al.10 study raises concern about offering standard medical treatments established for heart disease patients without renal failure and then translating them to dialysis patients with cardiac diseases. This issue urges the development of a prospective clinical trial where dialysis patients with documented atrial fibrillation are randomized to treatment with oral anticoagulation therapy or placebo to evaluate the efficacy of anticoagulant therapy in this population. An evidence-based answer to the issue is urgently needed. In conclusion, in our opinion and for the moment, dialysis patients with atrial fibrillation should be treated with oral anticoagulation only if they have a CHADS score ≥2 and after carefully weighing the risk/benefit ratio. In subjects on renal replacement therapy, proper anticoagulation control along with aggressive BP management is also important to minimize the risk of stroke in these high-risk patients. Careful INR monitoring is already feasible, because dialysis treatment requires most patients to be in contact with medical staff more than once a week. However, even these cautious measures of monitoring may not protect our patients from the risk of stroke, and on this issue we are in agreement with Chan et al.10 The future also could be brighter for dialysis patients. We need antithrombotic agents that are more effective than aspirin and safer to use than adjusted-dose warfarin. Additional options for stroke prevention may be on the horizon. Several large, randomized trials are currently testing novel oral anticoagulants, such as dabigatran and combinations of antiplatelet agents,14,15 but for the time being, given the ongoing epidemic of atrial fibrillation in ESRD, all nephrologists should be aware of this common and possibly iatrogenic cause of preventable stroke. Disclosures None.