Anticoagulants in atrial fibrillation patients with chronic kidney disease

Anticoagulants in atrial fibrillation patients with chronic kidney diseaseRobert G. Hart, John W. Eikelboom, Alistair J. Ingram and Charles A. Herzog Abstract Atrial fibrillation is an important cause of preventable, disabling stroke and is particularly frequent in patients with chronic kidney disease (CKD). Stage 3 CKD is an independent risk factor for stroke in patients with atrial fibrillation. Warfarin anticoagulation is efficacious for stroke prevention in atrial fibrillation patients with stage 3 CKD, but recent observational studies have challenged its value for patients with end-stage renal disease and atrial fibrillation. Novel oral anticoagulants such as dabigatran, apixaban and rivaroxaban are at least as efficacious as warfarin with reduced risks of intracranial haemorrhage. However, all these agents undergo renal clearance to varying degrees, and hence dosing, efficacy, and safety require special consideration in patients with CKD. Overall, the novel oral anticoagulants have performed well in randomized trials of patients with stage 3 CKD, with similar efficacy and safety profiles as for patients without CKD, albeit requiring dosing modifications. The required period of discontinuation of novel oral anticoagulants before elective surgery is longer for patients with CKD owing to their reduced renal clearance. Although much remains to be learned about the optimal use of these new agents in patients with CKD, they are attractive anticoagulation options that are likely to replace warfarin in coming years.
Hart, R. G. et al. Nat. Rev. Nephrol. 8, 569–578 (2012); published online 24 July 2012; doi:10.1038/nrneph.2012.160 Introduction
Atrial fibrillation is a frequent cause of disabling ischae-
Consequently, the efficacy for stroke prevention and the mic stroke owing to embolism of stasis-precipitated safety of anticoagulants have been poorly defined for thrombi forming in the left atrial appendage in patients CKD patients with atrial fibrillation.
with this common cardiac dysrhythmia. Atrial fibril- Because of the practical challenges associated with lation and chronic kidney disease (CKD) frequently warfarin use that include ongoing adjustment of war- coexist: about one-third of outpatients with atrial fibril- farin dose to maintain anticoagulation intensity in a lation have CKD,1 and 15% of patients with CKD have relatively narrow therapeutic range, novel oral anti- atrial fibrillation based on ascertainment by electro- coagulants have recently been introduced that are more cardiography or patient self-report.2,3 Atrial fibrillation selective in their anticoagulant mechanisms, and are is nearly three times as frequent in patients with stage 3 easier to dose without the need for regular laboratory CKD as in age-matched and sex-matched patients monitoring of anticoagulant effect.4 The relative roles of warfarin versus the novel oral anticoagulants are areas Multiple randomized trials have established warfa- of ongoing research and controversy,5,6 as much remains Division of Neurology rin anticoagulation to be highly efficacious for stroke to be learned about the optimal use of the novel oral (R. G. Hart), Division of prevention with acceptably low bleeding rates for most anticoagulants outside of clinical trials and in specific Hematology and Thromboembolism patients with atrial fibrillation. Warfarin is currently patient subgroups, particularly in patients with CKD. (J. W. Eikelboom), recommended by most guidelines for patients with Nevertheless, it is likely that the current generation of Division of Nephrology (A. J. Ingram), atrial fibrillation who have a substantial absolute risk of novel oral anticoagulants will eventual y replace warfarin stroke. Patients with advanced renal disease have been as the preferred anticoagulant for many, and probably Medicine, McMaster excluded from participation in recent randomized trials most, patients with atrial fibrillation.
University, 1280 Main Street West, Hamilton, of antithrombotic therapies in patients with atrial fibril- Here we review available data relevant to anticoagula- ON L8S 4L8, Canada. lation because of their increased risk of bleeding and, for tion of CKD patients with atrial fibrillation, considering Division of Cardiology, Department of some agents, renal clearance of the drugs being tested. patients with stages 3 and 4 CKD and end-stage renal Medicine, Hennepin disease (ESRD) separately. Although dual antiplatelet County Medical Center, Competing interests therapy with clopidogrel and aspirin also reduces the 701 Park Avenue, Minneapolis, R. G. Hart and J. W. Eikelboom declare associations with the incidence of stroke in patients with atrial fibrillation, it following companies: Bayer Pharmaceuticals, Boehringer is less efficacious than warfarin and insufficient data are Ingelheim, Bristol–Myers Squibb. C. A. Herzog declares an available concerning the safety of long-term dual anti- association with the following company: Johnson and Johnson. Correspondence to: See the article online for full details of the relationships. platelet therapy in patients with CKD; therefore, dual A. J. Ingram declares no competing interests.
antiplatelet therapy is not considered here.7,8 NATURE REVIEWS NEPHROLOGY VOLUME 8 OCTOBER 2012 569
2012 Macmillan Publishers Limited. All rights reserved Key points
renal clearance, with half-lives prolonged in patients with CKD (Table 1).
■ Atrial fibrillation is particularly frequent in patients with chronic kidney disease Four recent large randomized trials have assessed the ■ Stage 3 CKD is an independent risk factor for stroke in patients with atrial efficacy and safety of the three novel oral anti coagulants (dabigatran, apixaban, and rivaroxaban) that are cur- ■ Recent observational studies have challenged the value of warfarin rently approved for clinical use (Table 2).9–12 These three anticoagulation for patients with end-stage renal disease and atrial fibrillation agents were shown to be noninferior or superior to ■ Novel oral anticoagulants such as dabigatran, apixaban and rivaroxaban adjusted-dose warfarin for stroke prevention. Serious are noninferior or superior to warfarin, with reduced risks of intracranial bleeding with the new agents in clinical trials has been comparable to, or less than that with high-quality war- ■ In randomized trials to date, the novel oral anticoagulants have performed well farin anticoagulation. The reduced risk of intracranial in patients with stage 3 CKD, with similar efficacy and safety profiles as for patients without CKD, although dosing modifications are required haemorrhage with the new oral anticoagulants com- ■ The required period of discontinuation of novel oral anticoagulants before pared with warfarin was unexpected, but consistent elective surgery is longer for patients with CKD than for patients without renal and important as intra cranial haemorrhage is the most devastating complication of warfarin anticoagulation.17 Compared with warfarin, dabigatran was associated with significantly reduced cardiovascular mortality, apixaban Novel oral anticoagulants
was associated with significantly reduced total mortal- In the past decade, four novel oral anticoagulants (two ity, and rivaroxaban was associated with a trend toward direct thrombin inhibitors ximelagatran and dabigatran, reduced mortality (P = 0.15). Of note, concern has been and two factor Xa inhibitors apixaban and rivaroxaban) raised about bleeding risks in frail, elderly patients with have been tested in large phase III randomized trials renal impairment who are given dabigatran, the novel for prevention of stroke in patients with atrial fibrilla- oral anticoagulant most widely used to date outside of tion.9–14 Phase III trial results regarding a fifth agent (the clinical trials.18,19 factor Xa inhibitor edoxaban) are anticipated in late 2012 (Table 1).15 All trials have excluded participants with Warfarin in patients with CKD
severe renal impairment. Ximelagatran was withdrawn The efficacy of warfarin for stroke prevention in from the market in 2006 due to rare, but serious, hepato- patients with atrial fibrillation is proportional to the toxicity. Additional novel oral anti coagulants, some with quality of anticoagulation (that is, by sustaining the reduced renal clearance, are under development, but are optimal intensity of warfarin anticoagulation over not anticipated to be available for clinical use in the near time). Observational studies suggest that increased future. The novel anticoagulants mentioned above share time-in-therapeutic range (TTR) with an international the advantage of not requiring regular anticoagulation normalized ratio (INR) of 2–3 predicts improved clini- monitoring and frequent dose adjustments. On the cal outcomes, including reduction in stroke and bleed- downside, no antidote to rapidly reverse anticoagulant ing.20–22 However, a higher achieved average TTR in these effect has been established to date for any of the novel studies might reflect healthier patients (for example, anticoagulants in the event of acute serious bleeding fewer concomitant medications, less liver disease and (other than acute haemo dialysis for those agents that are less frequent heart failure with hepatic congestion), and not highly protein-bound such as dabigatran).16 All cur- better outcomes could be a result of better concomitant rently available novel oral anticoagulants have substantial care for which INR control is a marker. Hence, the link Table 1 Key pharmacological characteristics of novel oral anticoagulants
Coagulation target Bioavailability (%) Protein binding (%) Dosing frequency* Renal clearance (%) Routine monitoring Drug interactions CYP3A4 and P-glycoprotein CYP3A4 and P-glycoprotein CYP3A4 and P-glycoprotein Approved for ESRD *For patients with atrial fibrillation. Abbreviation: ESRD, end-stage renal disease. Adapted with permission from Wolters Kluwer Health Eikelboom, J. W. & Weitz, J. I. New anticoagulants. Circulation 121 (13), 1523–1532 (2010).
570 OCTOBER 2012 VOLUME 8
2012 Macmillan Publishers Limited. All rights reserved Table 2 Overview of phase III randomized trials of new oral anticoagulants*
Mean time in Main results‡ Stroke, non-CNS embolism and cardiovascular mortality reduced by dabigatran 150 mg vs warfarin; major haemorrhage reduced by dabigatran 110 mg vs warfarin; intracranial bleeding reduced by both doses of dabigatran vs warfarin; no significant difference in total mortality Double-blind; Serum Stroke and non-CNS embolism reduced by apixaban vs aspirin; major haemorrhage and >221 μmol/l creatinine intracranial bleeding comparable with both ≥133 μmol/l plus agents; no significant difference in age ≥80 years or cardiovascular or total mortality ROCKET AF11 Rivaroxaban Double-blind; eCrCl 15 mg per day if Rivaroxaban noninferior to warfarin for stroke CrCl <50 ml/min and non-CNS embolism; major haemorrhage comparable with both agents; intracranial bleeding reduced by rivaroxaban vs warfarin; no significant difference in cardiovascular or total mortality ARISTOTLE12 Apixaban Stroke, non-CNS embolism, major haemorrhage, intracranial bleeding and total >221 μmol/l creatinine mortality reduced by apixaban vs warfarin; ≥133 μmol/l plus no significant difference in cardiovascular age ≥80 years or *Publication of the phase III ENGAGE AF-TIMI 48 trial testing the factor Xa inhibitor edoxaban is anticipated in late 2012.15 ‡Among all participants; for results in subgroups of patients with stage 3 CKD, see Table 3. Abbreviations: CKD, chronic kidney disease; CNS, central nervous system; eCrCl, estimated creatinine clearance; INR, international normalized ratio; NA, not available.
between higher TTR and improved clinical outcomes Stage 3 CKD patients with atrial fibrillation
may be causal or reflect an association, or likely both. Stroke risk and warfarin efficacy For patients with atrial fibrillation receiving warfarin in Atrial fibrillation is not uncommon (18% in one clinical practice, TTRs average 55%,23,24 and in recent large study2) among patients with predialysis CKD. randomized trials, the mean TTRs ranged from 55% to Conversely, one-third of outpatients with atrial fibrilla- 64% (Table 2).9,11,12 Among patients with atrial fibrilla- tion have CKD.1 Among atrial fibrillation participants tion assigned to warfarin in the ROCKET AF trial, 1,476 in recent randomized trials of anticoagulant therapy participants with estimated creatinine clearance (eCrCl) with exclusion criteria based on renal function and of 30–49 ml/min had a median TTR of 58%, identical to whose patients were younger than population-based that of other participants.25 atrial fibrillation cohorts, 15–21% had an eCrCl of A general impression exists that anticoagulation control is particularly difficult in patients with ESRD and Stage 3 CKD is an independent predictor of stroke in atrial fibrillation who are given warfarin, but data are patients with atrial fibrillation (hazard ratio [HR] 1.5), limited and inconsistent. Studies have typically reported after adjustment for other risk factors.1,29–31 The mecha- the average of all recorded INRs or the fraction of all nisms underlying the increased risk of stroke conferred INRs that fall within the therapeutic range, but these by stage 3 CKD status in patients with atrial fibrillation metrics generally underestimate the TTR. In a small ret- are unclear and probably multiple.1 Stage 3 CKD may rospective study of 11 patients with ESRD given warfarin be a marker for end-organ damage from hypertension (five for atrial fibrillation and six for venous thrombo- and diabetes mellitus, adding predictive information embolism), the TTR for the conventional target INR that is not captured by considering just the prevalence range of 2–3 was 50%.26 In a prospective study from an (but not the severity, duration, or treatment) of vascu- anticoagulation clinic, patients with severe renal disease lar factors. The magnitude conferred by stage 3 CKD (47 of 53 undergoing dialysis) fol owed for approximately status on stroke risk is similar to other predictors in 1 year, 40% of all INRs were in the therapeutic range (the the widely used CHADS scheme.32,33 Among 89 par- TTR was not provided), and the average maintenance ticipants in the Stroke Prevention in Atrial Fibrillation warfarin dosage was significantly lower than for patients (SPAF) III trials with stage 3 CKD and a CHADS score without renal failure (3.9 mg per day versus 4.8 mg per of 0, the observed stroke risk without anticoagulation day, respectively).27 A trial of low-intensity warfarin was 2.3% per year (that is, moderate risk), albeit based (target INR 1.4–1.9) in 56 patients on haemodialysis on only three events.30 In our view, all atrial fibrilla- reported 47% of INRs to be in the target range.28 tion patients with stage 3 CKD should be considered to NATURE REVIEWS NEPHROLOGY VOLUME 8 OCTOBER 2012 571
2012 Macmillan Publishers Limited. All rights reserved Table 3 Phase III trials of antithrombotic therapies in atrial fibrillation patients with moderate CKD
Outcomes in CKD participants eGFR 30–59 ml/min* Warfarin INR 2–3 vs Ischaemic stroke/systemic embolism low, ineffective-dose reduced by 76% (95% CI 42–90), by warfarin plus aspirin warfarin (P <0.001) 30–49 ml/min‡ Warfarin INR 2–3 vs Stroke/non-CNS embolism rate 2.8% per dabigatran 150 mg or year with warfarin, 2.2% per year with 110 mg twice daily dabigatran 110 mg twice daily (NS), and 1.5% per year with dabigatran 150 mg twice daily (P <0.01); similar rates of major haemorrhage in all three treatment arms eGFR 30–59 ml/min* Apixaban 5 mg twice Stroke/non-CNS embolism rate 5.6% per year with aspirin vs 1.8% per year with apixaban (P <0.001); major bleeding 2.2% per year on aspirin vs 2.5% per year with apixaban (NS) 25–50 ml/min‡ Apixaban 5 mg twice Stroke/non-CNS embolism rate 2.7% per daily vs warfarin year with warfarin, 2.1% per year with apixaban (NS); major bleeding reduced by half with apixaban vs warfarin (P <0.01) 30–49 ml/min‡ Warfarin INR 2–3 vs Stroke/non-CNS embolism rate 3.4% per rivaroxaban 15 mg year with warfarin, 3.0% per year with rivaroxaban (NS); major bleeding rates nearly equal *Based on the CKD–EPI equation.71 ‡eCrCl using the Cockcroft–Gault formula.39 §Restricted to patients with atrial fibrillation deemed unsuitable for adjusted-dose warfarin.10 Dosage reduced to 2.5 mg twice daily for participants with two of the following criteria: age ≥80 years, body weight ≤60 kg, or serum creatinine concentration ≥133 μmol/l. Abbreviations: CKD, chronic kidney disease; CNS, central nervous system; eCrCl, estimated creatinine clearance; eGFR, estimated glomerular filtration rate; INR, international normalized ratio; NS, not significant.
have at least moderate stroke risk, independent of other participants with an eCrCl of 30–49 ml/min in the RE-LY predictive factors. Among those with stage 3 CKD, the study had a substantial rate of major haemorrhage (5.4% CHADS score seems to additionally stratify stroke risk per year), which was higher than that of other partici- based on two studies.29,30 Of note, these correlations pants (3.2% per year).9 However, these higher rates are are based on eCrCl or estimated glomerular filtration not adjusted for differences in age. A longitudinal cohort rate (eGFR) calculated from a single measurement of analysis found stage 3 CKD not to be an independent serum creatinine and sensitive to variations and error in predictor of major or minor haemorrhage during war- farin anticoagulation.27 In short, it is clear that atrial Warfarin anticoagulation markedly reduces the fibrillation patients with stage 3 CKD have about twice incidence of stroke in stage 3 CKD patients with atrial the rate (averaging about 5% per year in recent clinical fibrillation. In a subgroup analysis of 516 atrial fibrilla- trials) of major bleeding during warfarin anticoagulation tion participants with stage 3 CKD in the randomized compared with those with better renal function, but it SPAF III trial, ischaemic stroke or systemic embo- is uncertain whether this outcome is accounted for by lism was reduced by 76% (95% CI 42–90, P <0.001) by differences in age and other associated comorbidities.
adjusted-dose warfarin compared with aspirin plus low, Four schemes to stratify risk of bleeding during war- ineffective doses of warfarin (Table 3).30 farin anticoagulation have been published to date,34–37 and three have included abnormal renal function as a Bleeding during warfarin anticoagulation risk factor.34,36,37 In the derivation dataset of the ATRIA In recent randomized trials, participants with stage 3 scheme, an eCrCl of <30 ml/min was an independ- CKD had consistently higher rates of major haemor- ent predictor of major haemorrhage.36 None of these rhage during warfarin anticoagulation than did other schemes has yet been sufficiently validated for general participants, but whether this outcome would persist if adjusted for age and underlying cause of CKD is unclear. In the ROCKET AF trial, major haemorrhage with war- Novel oral anticoagulants farin occurred at a rate of 3.2% per year in those with All recent phase III randomized trials evaluating the an eCrCl of >50 ml/min (mean age 71 years) versus novel oral anticoagulants in patients with atrial fibril- 4.7% per year in those with an eCrCl of 30–49 ml/min lation have included participants with stage 3 CKD (mean age 79 years).25 In the ARISTOTLE trial, major (Table 2). In most trials, results were reported for sub- haemorrhage was more than twice as frequent among groups with an eCrCl (assessed using the Cockcroft– participants with an eCrCl of 25–50 ml/min given war- Gault formula)39 between 30 ml/min and 49 ml/min farin compared with other participants (6.4% per year (that is, the lower two-thirds of the conventional stage 3 versus 2.5% per year, respectively).12 Warfarin-assigned 572 OCTOBER 2012 VOLUME 8
2012 Macmillan Publishers Limited. All rights reserved In the RE-LY randomized trial, which compared Stroke or systemic embolism two doses of dabigatran with warfarin, patients were Major haemorrhage excluded if their eCrCl was <30 ml/min.9 Dabigatran trough levels correlate with efficacy for stroke pre- vention and are altered in patients with CKD because about 80% of dabigatran is excreted unchanged by the kidney.40 The efficacy of dabigatran compared with war- farin in a subgroup of 3,505 participants with an eCrCl of 30–49 ml/min has been reported.9 The rate of stroke Relative risk reduction or non-central-nervous-system embolism was 2.8% per year among those assigned to warfarin, 1.5% per year (P <0.01) in those given dabigatran 150 mg twice daily, and 2.2% per year in those given dabigatran 110 mg twice daily (Table 3, Figure 1).9 Although significantly fewer major haemorrhages occurred in those assigned Figure 1 Relative risk reductions in stroke or systemic
to dabigatran 110 mg twice daily versus warfarin among embolism and major haemorrhage by novel oral all RE-LY participants, major haemorrhage rates were anticoagulants versus warfarin in patients with moderate about equal with warfarin and dabigatran among those CKD.9,12,25 Patients with CKD had estimated creatinine with an eCrCl of 30–49 ml/min.
clearances of 30–49 ml/min, except for those treated with 41 However, there was apixaban (25–50 ml/min). Risk reductions were no statistical heterogeneity of effect according to renal statistically significant for dabigatran 150 mg on stroke function, so cautious interpretation requires that the and for apixaban on major haemorrhage. Abbreviation: reduced rate of major haemorrhages observed with the CKD, chronic kidney disease.
dabigatran 110 mg dose in the total cohort be applied to participants with stage 3 CKD. In summary, those with an eCrCl 30–49 ml/min in the RE-LY trial had a signifi- was not statistically different from that of other partici- cantly reduced rate of stroke with the 150 mg twice daily pants.12 There was a significant interaction between the dose compared with warfarin, and with a similar rate of effect of apixaban (given in the reduced dose to many major haemorrhage.9,41 For the dabigatran 110 mg twice patients with reduced eCrCl per protocol) versus war- daily dose compared with warfarin, there was no signifi- farin on major haemorrhage according to renal impair- cant differences in stroke or major haemorrhage among ment (P = 0.03); those with an eCrCl of 25–50 ml/min participants with an eCrCl of 30–49 ml/min.
had half the rate of major haemorrhage with apixaban The AVERROES trial compared apixaban with aspirin (3.3%) versus warfarin (6.7%). Based on the available in patients with atrial fibrillation deemed unsuitable for results of the ARISTOTLE trial, apixaban administered warfarin, mostly owing to a perceived risk of bleeding according to the renal dose-adjusted scheme resulted or patient preference (40% had previously received a in a trend toward superior efficacy and significantly vitamin K antagonist).10 Participants were assigned to less bleeding than warfarin for those with reduced apixaban 5 mg twice daily, reduced to 2.5 mg twice daily eCrCl (Figure 1).
in participants with a serum creatinine concentration The ROCKET AF randomized trial excluded partici- ≥1.5 mg/dl (≥133 μmol/l) and either age ≥80 years or pants with an eCrCl <30 ml/min and reduced the dose of body weight ≤60 kg. In the subgroup of patients with rivaroxaban to 15 mg per day for those with an eCrCl of stage 3 CKD (n = 1,697, 30% of the cohort, mean eGFR 30–49 ml/min.11,25 Among these 2,950 participants (21% 49 ml/min), apixaban significantly reduced the rate of of the total trial cohort), the median TTR was 58% and stroke compared with aspirin (1.8% per year versus was not different from the TTR in those with a higher 5.6% per year, respectively; HR 0.32, 95% CI 0.18–0.55, eCrCl.25 By intention-to-treat analysis, the rate of stroke P <0.001).29 No significant difference was found in major or non-central-nervous-system embolism was 3.4% per haemorrhage in patients with stage 3 CKD by treatment: year with warfarin and 3.0% per year with rivaroxa- 2.2% per year with aspirin versus 2.5% per year with ban (HR 0.86, 95% CI 0.63–1.2) with no heterogeneity apixaban (HR 1.2, 95% CI 0.65–2.1).
of treatment effect compared with other participants. The ARISTOTLE randomized trial assessed apixaban In participants with CKD, the rates of the composite 5 mg twice daily in 18,201 patients with atrial fibrillation of major haemorrhage or clinically relevant nonmajor and reported superiority to warfarin in preventing stroke bleeding were similar among those given rivaroxaban or systemic embolism (HR 0.79, 95% CI 0.66–0.95), with and warfarin (HR 0.98, 95% CI 0.84–1.1).25 A reduction less bleeding and lower mortality.12 The dose of apixa- in intracranial haemorrhage with rivaroxaban compared ban was reduced to 2.5 mg twice daily for participants with warfarin was evident in those with reduced eCrCl who had a serum creatinine concentration ≥1.5 mg/dl given rivaroxaban 15 mg daily, albeit not statistically sig- (≥133 μmol/l) and either age ≥80 years or a body weight nificant (HR 0.81, 95% CI 0.41–1.6), but commensurate ≤60 kg. Among 3,017 participants with an eCrCl of with the effect in the entire trial. In short, there was no 25–50 ml/min (89% between 31–50 ml/min), stroke rates evidence of heterogeneity for any outcome comparing the were higher than in other participants, but the efficacy of treatment effects of rivaroxaban versus warfarin between apixaban relative to warfarin in this subgroup (HR 0.78) those with versus those without reduced eCrCl.25 NATURE REVIEWS NEPHROLOGY VOLUME 8 OCTOBER 2012 573
2012 Macmillan Publishers Limited. All rights reserved per year and a 0.3% per year increase in major haem- Warfarin INR 2.0–3.0 0.24 (0.10–0.38) orrhage with apixaban versus aspirin.29 Based on these Apixaban 2.5/5.0 mg twice daily 0.32 (0.18–0.55) limited data from randomized trials, the benefit seems to substantially exceed the risk, and most patients with stage 3 CKD and atrial fibrillation are likely to impor- tantly benefit from anticoagulation. That participants in clinical trials are selected as healthier and less prone to bleeding complications than the average patient seen in Figure 2 Hazard ratios for subgroups of patients with stage 3 CKD from two
clinical practice is an important caveat, however.18 randomized trials comparing anticoagulation with aspirin.29,30 For warfarin Randomized comparisons of three novel oral anti- comparison from the SPAF III study, the outcome was ischaemic stroke and coagulants with warfarin involving 9,472 participants systemic embolism and the aspirin group additionally received low, ineffective with CKD indicate superiority or noninferiority to war- doses of warfarin.30 For apixaban the outcome was stroke and systemic farin for stroke prevention (Figure 3) with comparable embolism.29 Abbreviations: CKD, chronic kidney disease; HR, hazard ratio; INR, international normalized ratio.
or reduced bleeding risks (Figure 1). Consequently, the novel oral anticoagulants seem to be reasonable alterna- tives to warfarin for stroke prevention in stage 3 CKD Dabigatran 110 mg patients with atrial fibrillation. The FDA,42,43 Health 0.77 (0.51–1.18) Canada,44,45 and European Medicines Agency46,47 have Dabigatran 150 mg approved dabigatran and rivaroxaban for use in atrial 0.55 (0.40–0.81) Rivaroxaban 15 mg fibrillation patients with stage 3 CKD; apixaban is cur- 0.86 (0.63–1.17) rently undergoing evaluation by these regulatory agen- Apixaban 2.5/5.0 mg cies (Table 4). In our view, it cannot be determined with 0.79 (0.57–1.20) confidence which of the novel oral anticoagulants is preferred in stage 3 CKD patients with atrial fibrillation. Single trials, wide confidence intervals surrounding the estimated effects in subgroups of patients with CKD, absence of significant interactions of CKD subgroups anticoagulant better with overall effects, differences in trial populations, Figure 3 Hazard ratios for subgroups of patients with stage 3 CKD (estimated
differing dosing regimens, and lack of head-to-head creatinine clearances 30–49 ml/min or 25–50 ml/min for apixaban) from comparisons combine to prevent reliable comparisons.
randomized trials comparing novel oral anticoagulants with warfarin for the primary outcome of stroke and systemic embolism.9,12,25 The width of the 95% CI are Stage 4 CKD patients with atrial fibrillation
estimated from published figures for dabigatran and apixaban.9,12 Abbreviations: CKD, chronic kidney disease; HR, hazard ratio.
Small numbers of participants with stage 4 CKD (eGFR 15–29 ml/min) were included in the ARISTOTLE randomized trial comparing apixaban with warfarin Based on this absence of hetero geneity, the overall (n = 270)12 and in the AVERROES randomized trial trial results are assumed to apply to best characterize comparing apixaban with aspirin (n = 70),29 but no the effect in the subgroup with eCrCl of 30–49 ml/min results for these subgroups have been published. The treated with rivaroxaban 15 mg per day: rivaroxaban is FDA has approved a reduced dose of dabigatran (75 mg noninferior to warfarin for prevention of stroke with twice daily) for patients with stage 4 CKD,42 based on similar risks of major bleeding and reduced risks of pharmaco kinetic and pharmacodynamic studies.48 intracranial and fatal bleeding.11 Other regulatory agencies have not approved the use of dabigatran in patients with stage 4 CKD (Table 4). Randomized trials and regulatory agency approvals Rivaroxaban 15 mg per day has been approved by some Five randomized trials that included 11,685 patients with major regulatory agencies for stage 4 CKD,43,47 although stage 3 CKD (with some with stage 4 CKD) and atrial such patients were excluded from participation in the fibril ation have tested four anticoagulants (Table 3). The ROCKET AF trial (Table 4). We are unaware of clini- two trials (SPAF III30 and AVERROES29) that compared cal end point data supporting efficacy or safety of either oral anticoagulation with aspirin in 2,213 patients with dabigatran or rivaroxaban in patients with stage 4 CKD.
CKD confirm that the large reduction in stroke con- Similarly, no data exist to support the efficacy and ferred by warfarin and apixaban in patients with atrial safety of warfarin for stage 4 CKD patients with atrial fibrillation overall extends to those with stage 3 CKD fibrillation, although advocated by some guidelines.49 (Figure 2). Given the substantial rates of major haemor- Stage 4 CKD emerged as an independent predictor of rhage during warfarin anticoagulation in patients with major haemorrhage during warfarin anticoagulation stage 3 CKD described above, does the absolute reduc- from multivariate analysis of one large outpatient cohort tion in stroke outweigh the absolute increase in major study of patients with atrial fibrillation.36 haemorrhage? Too few major bleeding events were observed in the SPAF III trial for meaningful analysis.30 Interrupting anticoagulation for surgery
Among participants with stage 3 CKD in the AVERROES Because the half-life of the novel oral anticoagulants is trial, there was an absolute reduction in stroke of 3.8% prolonged in patients with CKD, longer interruption 574 OCTOBER 2012 VOLUME 8
2012 Macmillan Publishers Limited. All rights reserved Table 4 Major regulatory agency recommendations for novel oral anticoagulants in patients with CKD*
Stage 3 CKD: 150 mg twice daily 15 mg daily for CrCl 15–49 ml/min Stage 4 CKD: 75 mg twice daily‡ European Medicines Agency46,47 Stage 3 CKD: 110 mg twice daily if aged 15 mg daily for CrCl 15–49 ml/min >80 years or at high risk of bleedingStage 4 CKD: not approved Health Canada44,45 CrCl 30–50 ml/min: either 110 mg or 150 mg 15 mg daily for CrCl 30–49  ml/min twice daily except 110 mg twice daily for Stage 4 CKD: not approved those aged >75 years and CrCl <50 ml/minStage 4 CKD: not approved *Edoxaban has not been considered by these agencies. ‡Reduce the dose to 75 mg twice daily in stage 3 CKD when given with systemic ketoconazole or dronedarone; avoid use of dabigatran in stage 4 CKD with P-glycoprotein inhibitors. Abbreviations: CKD, chronic kidney disease; CrCl, creatinine clearance; NR, no recommendations to date.
of treatment for these patients is required before elec- accurately estimate the rate of stroke in ESRD patients tive surgery.50 For patients with an eCrCl of 30–49 ml/ with atrial fibrillation who are not receiving anti- min, the half-life of dabigatran is estimated to be 18 h. thrombotic therapy. For example, in one large dialysis Withdrawing dabigatran for 2–4 days (and at least 5 days clinic-based study, strokes were identified only at the for patients with stage 4 CKD) and a normal activated time of hospitalization (that is, probable underdetection), partial thromboplastin time before surgery is recom- patients with transient ischaemic attack were included, mended.51 For rivaroxaban, with less renal clearance than and nearly half had received warfarin.52 A stroke rate of dabigatran, withholding treatment for 2 days in patients about 7% per year for all strokes (including intracerebral with stage 3 CKD and 3 days in patients with stage 4 haemorrhage) is a reasonable estimate based on available CKD is recommended.
data.52,56–58 These imperfect data are, however, consistent In the event of massive haemorrhage, haemodialysis with the notion that although the rates of stroke in ESRD can be used in patients receiving dabigatran, but not for patients with atrial fibrillation are higher than for ESRD the more highly protein-bound rivaroxaban and apixa- patients without atrial fibrillation, the relative increase ban (Table 1). Although factor VIIa and four-factor does not seem to be as large as that for patients without prothrombin complex concentrates have been used in ESRD, in part because of the higher background stroke these situations, their value in reversing the clinical anti- rate in patients with ESRD.
coagulant effects and controlling clinical haemorrhage Limited data exist regarding stratification of stroke risk is uncertain.16 Humanized Fab fragment for dabigatran, in ESRD patients with atrial fibrillation. A large retro- and recombinant, active-site-blocked G1a-domainless spective study of patients on haemodialysis with atrial factor Xa to neutralize rivaroxaban and apixaban, are fibrillation reported that increasing age, heart failure, under development.
and systolic blood pressure correlated with stroke risk among haemodialysis patients with atrial fibrillation, ESRD patients with atrial fibrillation
but whether these predictors resulted from multi variate Among patients on haemodialysis, the overal prevalence analy sis was not clear.52 Multivariate analysis of another of atrial fibrillation is variously estimated at 7–20%, at study reported prior stroke, diabetes mellitus, and least double that of age-matched patients without advancing age to be independently predictive of hospi- ESRD.52–54 The frequency is most strongly related to age; talization for stroke, but hypertension and heart failure in the study by Genovesi et al., the cross-sectional preva- were not.56 Both of these studies plus a third study59 lence of atrial fibrillation was 17% in patients on haemo- reported that the CHADS scheme successfully strati- dialysis aged 51–60 years, increasing to 37% for those fied stroke risk in ESRD patients with atrial fibrillation, aged 71–80 years.53 In an analysis of USRDS/Medicaid but the contribution of individual components of the data, there was a graded increase from approximately CHADS score seemed to differ. It is therefore uncer- 2% to 17% over the age range <55 years to >85 years.54 tain how reliably the CHADS scheme and other stroke Atrial fibrillation was an independent risk factor for risk stratification schemes apply to ESRD patients with ischaemic stroke in patients with ESRD in the study by Vázquez et al. (odds ratio 2.3, 95% CI 1.2–11)55 and in the ESRD seems to be an independent risk factor for major DOPPS I and II analysis (HR 1.3, 95% CI 1.0–1.6).56 By haemorrhage during warfarin therapy.27,60 In studies of contrast, the prospective study by Genovesi et al. found US dialysis patients with atrial fibrillation between 1996 no increase in stroke in patients on haemodialysis with and 2004, 26–44% were treated with warfarin.52,56 Despite atrial fibrillation.57 the frequent use of warfarin in patients on haemo dialysis, From available studies that are limited by small patient data on bleeding rates are meagre. A systematic review numbers, concurrent antithrombotic treatment, different of bleeding rates during warfarin anticoagulation in methods of stroke detection and uncertain reliability of patients on haemodialysis published in 2007 identified identification of atrial fibrillation, it is not possible to only three studies and no randomized trials evaluating NATURE REVIEWS NEPHROLOGY VOLUME 8 OCTOBER 2012 575
2012 Macmillan Publishers Limited. All rights reserved of responders recently changed opinion regarding the Table 5 Anticoagulation options for CKD patients with atrial fibrillation*
risk/benefit of warfarin for atrial fibrillation patients on Anticoagulant options dialysis and 71% believed that guidelines for warfarin Stage 3 (eGFR 30–59 ml/min) Warfarin (target INR 2–3) anti coagulation established for the general population Dabigatran 110 mg or 150 mg twice daily should not be extrapolated to patients on dialysis.70 In Apixaban 5 mg twice dailyRivaroxaban 15 mg daily for eGFR 30–49 ml/ our view and concordant with newly revised guide- min; 20 mg daily for eGFR 50–59 ml/min lines,69 there are at present insufficient data to recom- Stage 4 (eGFR 15–29 ml/min) Warfarin (target INR 2–3) mend routine anticoagulation with warfarin for ESRD Dabigatran 75 mg twice daily patients with atrial fibrillation for the primary preven- Rivaroxaban 15 mg daily tion of stroke. Previous nonlacunar cardioembolic stroke ESRD (eGFR <15 ml/min or dialysis) Primary prevention: no anticoagulation‡ or transient ischemic attack, however, are such potent Secondary prevention: warfarin (target INR 2–3) risk factors for subsequent disabling stroke that warfarin *Recommendations based on results of large randomized trials for stage 3 CKD but not for stage 4 CKD or anticoagulation seems reasonable (but not mandatory) ESRD. Regulatory approvals are inconsistent (see Table 4). ‡Relatively recent recommendation. Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate; ESRD, end-stage for secondary prevention of stroke in ESRD patients renal disease; INR, international normalized ratio.
with atrial fibrillation. These recommendations are based on such low-quality evidence that it is reasonable conventional intensity warfarin.61 In these three studies, not to discontinue warfarin in ESRD patients with atrial the annualized bleeding rates for full- intensity war- fibrillation who have had good INR control and without farin anticoagulation were very high (10–54%).61 bleeding. At present, the novel oral anticoagulants dis- Subsequently, 255 patients on haemodialysis were ret- cussed above have not been approved for use in patients rospectively analyzed according to time-dependent use with ESRD. There is an urgent need for randomized clin- of antithrombotic therapy and it was found that patients ical trials of anticoagulant (warfarin and/or other novel with ESRD on warfarin had a fourfold increase in major anti coagulants with minimal renal clearance) for stroke haemorrhage, with an absolute rate of 3.1% per year.62 prevention in ESRD patients with atrial fibrillation.
Patients on dialysis seem to spend less time within the therapeutic INR range, with a tendency to suprathera- Using eGFR and eCrCl to tailor therapy
peutic INR values.27 In addition to warfarin-associated Anticoagulation options for CKD patients with atrial haemorrhage, there is concern about accentuation of vas- fibrillation depend on the degree of renal impairment cular calcification and calciphylaxis by chronic warfarin (Table 5). The clinical studies of patients with CKD and use in patients with ESRD.63 atrial fibrillation cited above used eCrCl or eGFR using Extrapolating the striking efficacy of warfarin anti- equations based on measurements of serum creati- coagulation shown in randomized clinical trials64 to nine levels. Extrapolating the results of these studies to atrial fibrillation patients on haemodialysis, the 2005 measure ment of creatinine clearance (CrCl) using timed K/DOQI guidelines state: "Anticoagulation in non- urine col ection would be tenuous, and we do not favour valvular atrial fibrillation: Dialysis patients are at direct measurement of CrCl in routine clinical practice. In increased risk for bleeding and careful monitoring addition, for the same value of the serum creatinine con- should accompany intervention."65 Recent retrospective centration, the eCrCl will most often exceed the eGFR, analyses of large dialysis databases have raised concern with the latter used to define the stage of CKD due to the regarding the efficacy of warfarin anticoagulation in difference in commonly used estimating equations.39,71–73 haemodialysis patients with atrial fibrillation.52,56,58,66 Many clinical laboratories automatical y provide an eGFR Studies from a national dialysis network of incident that accompanies measurement of serum creatinine dialysis patients identified the use of warfarin with an levels.73 Consequently, it is challenging to apply these increased risk of stroke and overall mortality.52,66 The data to everyday clinical management, particularly the increased risk of stroke with warfarin demonstrated a results of routine laboratory values for the eGFR when dose effect with higher INRs associated with increased, clinical studies were based on eCrCl.39,73 Given the uncer- not decreased, risk of stroke.67 A study from the DOPPS tainties and pending additional information, we favour database reported increased hazard ratio for stroke for management based on eGFR calculated by the CKD–EPI those receiving warfarin.56 This study also found that equation71 or the MDRD equation.74 The eGFR should warfarin use in patients aged >75 years was associated be obtained before initiating anti coagulation and at least with an increased risk of stroke (perhaps as a result of annual y thereafter (and at least every 6 months in those haemorrhagic stroke, although this effect could not be with an eGFR <45 ml/min). Before changing the dose of determined). An unacceptably high rate of haemorrhagic a novel oral anti coagulant or discontinuing anticoagula- stroke (2.6% per year) has been reported in patients tion for declining renal function, the serum creatinine on haemodialysis given warfarin for atrial fibrillation measurement should be repeated after 1 month, and with no apparent reduction in ischaemic stroke in a eGFR reassessed due to instability related to laboratory retrospective cohort study.58 These observational data have prompted doubts about the value of warfarin anticoagulation in ESRD patients with atrial fibrillation.67–69 Of note, an inter- CKD and atrial fibrillation frequently coexist. The national survey of dialysis providers reported that 54% presence of CKD is an important factor to consider 576 OCTOBER 2012 VOLUME 8
2012 Macmillan Publishers Limited. All rights reserved when anticoagulating patients with atrial fibrillation to embolic brain ischaemia. Oral anticoagulants suitable for prevent stroke, particularly when using the novel oral use in ESRD patients with atrial fibrillation that are safer anti coagulants discussed in this Review. We speculate and easier to administer than adjusted-dose warfarin are that future studies will result in further refinement of urgently needed.
the optimal dosing of the novel oral anticoagulants in patients with CKD, but even at the current state of knowledge, they are attractive options that are likely to eventually replace warfarin for most stage 3 and 4 CKD This Review was based on a combination of the working patients with atrial fibrillation. Recent studies have chal- knowledge and expert opinion of the authors. The lenged the value of warfarin anticoagulation for ESRD authors' views were supported by citing the key relevant patients with atrial fibril ation except for those with prior publications in the field.
1. Go, A. S. et al. for the ATRIA Study Investigators. warfarin for stroke prevention in patients with fibrillation and moderate renal impairment. Eur. Impact of proteinuria and glomerular filtration nonvalvular atrial fibrillation: a randomized trial. Heart J. 32, 2387–2394 (2011).
rate on risk of thromboembolism in atrial JAMA 293, 690–698 (2005).
26. To, A. C., Yehia, M. & Collins, J. F. Atrial fibrillation. The Anticoagulation and Risk Factors 15. Ruff, C. T. et al. Evaluation of the novel factor fibrillation in haemodialysis patients: in Atrial Fibrillation (ATRIA) Study. Circulation Xa inhibitor edoxaban compared with warfarin do guidelines for anticoagulation apply? 119, 1363–1369 (2009).
in patients with atrial fibrillation: design and Nephrology (Carlton) 12, 441–447 (2007).
2. Soliman, E. Z. et al. Chronic kidney disease and rationale for the Effective aNticoaGulation with 27. Limdi, N. A. et al. Kidney function influences prevalent atrial fibrillation: the Chronic Renal factor xA next GEneration in Atrial Fibrillation- warfarin responsiveness and hemorrhagic Insufficiency Cohort (CRIC). Am. Heart J. 159, Thrombolysis In Myocardial Infarction study 48 complications. J. Am. Soc. Nephrol. 20, 912–921 1102–1107 (2010).
(ENGAGE AF-TIMI 48). Am. Heart J. 160, 3. Baber, U. et al. Association of chronic kidney 635–641 (2010).
28. Crowther, M. A. et al. Low-intensity warfarin is disease with atrial fibrillation among adults in 16. Kaatz, S. et al. Guidance on the emergent ineffective for the prevention of PTFE graft failure the United States: Reasons for Geographic and reversal of oral thrombin and factor Xa in patients on hemodialysis: a randomized Racial Differences in Stroke (REGARDS) Study. inhibitors. Am. J. Hematol. 87 (Suppl. 1), controlled trial. J. Am. Soc. Nephrol. 13, Circ. Arrhythm. Electrophysiol. 4, 26–32 (2011).
S141–S145 (2012).
2331–2337 (2002).
4. Eikelboom, J. W. & Weitz, J. I. New 17. Hart, R. G. et al. Intracranial hemorrhage in 29. Eikelboom, J. W. et al. Stroke risk and efficacy of anticoagulants. Circulation 121, 1523–1532 atrial fibrillation patients during apixaban in atrial fibrillation patients with anticoagulation with warfarin or dabigatran: moderate chronic kidney disease. J. Stroke 5. Ansell, J. New oral anticoagulants should not be the RE-LY trial. Stroke 43, 1511–1517 (2012).
Cerebrovasc. Dis. (in press).
used as first-line agents to prevent 18. Harper, P., Young, L. & Merriman, E. Bleeding 30. Hart, R. G., Pearce, L. A., Asinger, R. W. & thromboembolism in patients with atrial risk with dabigatran in the frail elderly. N. Engl. Herzog, C. A. Warfarin in atrial fibrillation fibrillation. Circulation 125, 165–170 (2012).
J. Med. 366, 864–866 (2012).
patients with moderate chronic kidney disease. 6. Granger, C. B. & Armaganijan, L. V. Newer oral 19. U. S. Food and Drug Administration. Pradaxa Clin. J. Am. Soc. Nephrol. 6, 2599–2604 (2011).
anticoagulants should be used as first-line (dabigatran etexilate mesylate): drug safety 31. Nakagawa, K. et al. Chronic kidney disease and agents to prevent thromboembolism in patients communication—safety review of post-market CHADS score independently predict with atrial fibrillation and risk factors for stroke reports of serious bleeding events [online], cardiovascular events and mortality in patients or thromboembolism. Circulation 125, 159–164 with nonvalvular atrial fibrillation. Am. J. Cardiol. 107, 912–916 (2011).
7. The ACTIVE Investigators. Effect of clopidogrel 32. Gage, B. F. et al. Validation of clinical added to aspirin in patients with atrial classification schemes for predicting stroke: fibrillation. N. Engl. J. Med. 360, 2066–2078 20. Connolly, S. J. et al. on behalf of the ACTIVE W results from the National Registry of Atrial Investigators. Benefit of oral anticoagulant Fibrillation. JAMA 285, 2864–2870 (2001).
8. Connolly, S. et al. Clopidogrel plus aspirin versus over antiplatelet therapy in atrial fibrillation 33. Stroke Risk in Atrial Fibrillation Working Group. oral anticoagulation for atrial fibrillation in the depends on the quality of international Independent predictors of stroke in patients with Atrial fibrillation Clopidogrel Trial with Irbesartan normalized ratio control achieved by centers atrial fibrillation: a systematic review. Neurology for prevention of Vascular Events (ACTIVE W): and countries as measured by time in 69, 546–554 (2007).
a randomised controlled trial. Lancet 367, therapeutic range. Circulation 118, 34. Lip, G. Y., Frison, L., Halperin, J. L. & Lane, D. A. 1903–1912 (2006).
2029–2037 (2008).
Comparative validation of a novel risk score for 9. Connolly, S. J. et al. Dabigatran versus warfarin 21. White, H. D. et al. Comparison of outcomes predicting bleeding risk in anticoagulated in patients with atrial fibrillation. N. Engl. J. Med. among patients randomized to warfarin therapy patients with atrial fibrillation: the HAS-BLED 361, 1139–1151 (2009).
according to anticoagulant control: results (Hypertension, Abnormal Renal/Liver Function, 10. Connolly, S. J. et al. for the AVERROES Steering from SPORTIF III and V. Arch. Intern. Med. 167, Stroke, Bleeding History or Predisposition, Labile Committee and Investigators. Apixaban in 239–245 (2007).
INR, Elderly, Drugs/Alcohol Concomitantly) patients with atrial fibrillation. N. Engl. J. Med. 22. Jones, M. et al. Evaluation of the pattern of score. J. Am. Coll. Cardiol. 57, 173–180 (2011).
364, 806–817 (2011).
treatment, level of anticoagulation control, and 35. Shireman, T. I. et al. Development of a 11. Patel, M. R. et al. Rivaroxaban versus warfarin in outcome of treatment with warfarin in patients contemporary bleeding risk model for elderly nonvalvular atrial fibrillation. N. Engl. J. Med. 365, with non-valvular atrial fibrillation: a record warfarin recipients. Chest 130, 1390–1396 883–891 (2011).
linkage study in a large British population. 12. Granger, C. B. et al. for the ARISTOTLE Heart 91, 472–477 (2005).
36. Fang, M. C. et al. A new risk scheme to predict Committees and Investigators. Apixaban versus 23. Baker, W. L., Cios, D. A., Sander, S. D. & warfarin-associated hemorrhage: The ATRIA warfarin in patients with atrial fibrillation. N. Engl. Coleman, C. I. Meta-analysis to assess the (Anticoagulation and Risk Factors in Atrial J. Med. 365, 981–992 (2011).
quality of warfarin control in atrial fibrillation Fibrillation) Study. J. Am. Coll. Cardiol. 58, 13. Executive Steering Committee on behalf of the patients in the United States. J. Manag. Care 395–401 (2011).
SPORTIF III Investigators. Stroke prevention with Pharm. 15, 244–252 (2009).
37. Gage, B. F. et al. Clinical classification schemes the oral direct thrombin inhibitor ximelagatran 24. Rose, A. J., Ozonoff, A., Henault, L. E. & for predicting hemorrhage: results from the compared with warfarin in patients with non- Hylek, E. M. Warfarin for atrial fibrillation in National Registry of Atrial Fibrillation (NRAF). Am. valvular atrial fibrillation (SPORTIF III): community-based practise. J. Thromb. Heart J. 151, 713–719 (2006).
randomised controlled trial. Lancet 362, Haemost. 6, 1647–1654 (2008).
38. Loewen, P. & Dahri, K. Risk of bleeding with oral 1691–1698 (2003).
25. Fox, K. A. et al. Prevention of stroke and anticoagulants: an updated systematic review 14. SPORTIF Executive Steering Committee for the systemic embolism with rivaroxaban compared and performance analysis of clinical prediction SPORTIF V Investigators. Ximelagatran vs with warfarin in patients with non-valvular atrial rules. Ann. Hematol. 90, 1191–1200 (2011).
NATURE REVIEWS NEPHROLOGY VOLUME 8 OCTOBER 2012 577
2012 Macmillan Publishers Limited. All rights reserved 39. Cockcroft, D. W. & Gault, M. H. Prediction of 51. van Ryn, J. et al. Dabigatran etexilate—a novel, stroke in patients who have nonvalvular atrial creatinine clearance from serum creatinine. reversible, oral direct thrombin inhibitor: fibrillation. Ann. Intern. Med. 146, 857–867 Nephron 16, 31–41 (1976).
interpretation of coagulation assays and reversal 40. U. S. Food and Drug Administration. Dabigatran of anticoagulant activity. Thromb. Haemost. 103, 65. K/DOQI Workgroup. K/DOQI clinical practice Briefing Document [online], http://www.fda.gov/ 1116–1127 (2010).
guidelines for cardiovascular disease in 52. Chan, K. E., Lazarus, J. M., Thadhani, R. & dialysis patients. Am. J. Kidney Dis. 45, Hakim, R. M. Warfarin use associates with S1–S153 (2005).
increased risk for stroke in hemodialysis patients 66. Chan, K. E., Lazarus, J. M., Thadhani, R. & with atrial fibrillation. J. Am. Soc. Nephrol. 20, Hakim, R. M. Anticoagulant and antiplatelet 41. Eikelboom, J. W. et al. Risk of bleeding with 2 2223–2233 (2009).
usage associates with mortality among doses of dabigatran compared with warfarin in 53. Genovesi, S. et al. Prevalence of atrial fibrillation hemodialysis patients. J. Am. Soc. Nephrol. 20, older and younger patients with atrial fibrillation: and associated factors in a population of long- 872–881 (2009).
an analysis of the Randomized Evaluation of term hemodialysis patients. Am. J. Kidney Dis. 67. Yang, F., Chou, D., Schweitzer, P. & Hanon, S. Long-Term Anticoagulant Therapy (RE-LY) Trial. 46, 897–902 (2005).
Warfarin in haemodialysis patients with atrial Circulation 123, 2363–2372 (2011).
54. Wetmore, J. B. et al. The prevalence of and fibrillation: what benefit? Europace 12, 42. U. S. Food and Drug Administration. FDA factors associated with chronic atrial fibrillation 1666–1672 (2010).
approved drug products: label information for in Medicare/Medicaid-eligible dialysis patients. 68. Marinigh, R., Lane, D. A. & Lip, G. Y. Severe PRADAXA [online], http:// Kidney Int. 81, 469–476 (2012).
renal impairment and stroke prevention in 55. Vázquez, E. et al. Comparison of prognostic value atrial fibrillation: implications for of atrial fibrillation versus sinus rhythm in thromboprophylaxis and bleeding risk. J. Am. 43. U. S. Food and Drug Administration. FDA patients on long-term hemodialysis. Am. J. Coll. Cardiol. 57, 1339–1348 (2011).
approved drug products: Xarelto [online], Cardiol. 92, 868–871 (2003).
69. Herzog, C. A. et al. Cardiovascular disease in 56. Wizemann, V. et al. Atrial fibrillation in chronic kidney disease. A clinical update from hemodialysis patients: clinical features and Kidney Disease: Improving Global Outcomes associations with anticoagulant therapy. Kidney (KDIGO). Kidney Int. 80, 572–586 (2011).
44. Health Canada. Drugs and Health Products: Int. 77, 1098–1106 (2010).
70. Barany, P. Warfarin treatment in ESRD: Survey 57. Genovesi, S. et al. Atrial fibrillation and morbidity and mortality in a cohort of long-term 45. Health Canada. Drugs and Health Products: hemodialysis patients. Am. J. Kidney Dis. 51, 71. Levey, A. S. et al. for the CKD-EPI (Chronic 255–262 (2008).
Kidney Disease Epidemiology Collaboration). A 58. Winkelmayer, W. C., Liu, J., Setoguchi, S. & new equation to estimate glomerular filtration 46. European Medicines Agency. Product Choudry, N. K. Effectiveness and safety of rate. Ann. Intern. Med. 150, 604–612 (2009).
information: Pradaxa [online], http://www.ema.
warfarin initiation in older hemodialysis patients 72. Levey, A. S. & Coresh, J. Chronic kidney with atrial fibrillation. Clin. J. Am. Soc. Nephrol. 6, disease. Lancet 379, 165–180 (2012).
2662–2668 (2011).
73. Stevens, L. A., Coresh, J., Greene, T. & 59. Chou, C. Y. et al. Outcome of atrial fibrillation Levey, A. S. Assessing kidney function— 47. European Medicines Agency. Product among patients with end-stage renal disease. measured and estimated glomerular filtration information: Xarelto [online], http://www.ema.
Nephrol. Dial. Transplant. 25, 1225–1230 (2010).
rate. N. Engl. J. Med. 354, 2473–2483 (2006).
60. McMahon, D. A., Smith, D. M., Carey, M. A. & 74. Levey, A. S. et al. A more accurate method to Zhou, X. H. Risk of major hemorrhage for estimate glomerular filtration rate from serum outpatients treated with warfarin. J. Gen. Intern. creatinine: a new prediction equation. 48. Stangier, J., Rathgen, K., Stähle, H. & Mazur, D. Med. 13, 311–316 (1998).
Modification of Diet in Renal Disease Study Influence of renal impairment on the 61. Elliott, M. J., Zimmerman, D. & Holden, R. M. Group. Ann. Intern. Med. 130, 461–470 pharmacokinetics and pharmacodynamics of Warfarin anticoagulation in hemodialysis oral dabigatran etexilate: an open-label, parallel- patients: a systematic review of bleeding rates. 75. Coresh, J. et al. Calibration and random group, single-centre study. Clin. Pharmacokinet. Am. J. Kidney Dis. 50, 433–440 (2007).
variation of the serum creatinine assay as 49, 259–268 (2010).
62. Holden, R., Harman, G. J., Wang, M., Holland, D. critical elements of using equations to 49. Skanes, A. C. et al. Focused 2012 update of the & Day, A. G. Major bleeding in hemodialysis estimate glomerular filtration rate. Am. J. Canadian Cardiovascular Society atrial patients. Clin. J. Am. Soc. Nephrol. 3, 105–110 Kidney Dis. 39, 920–929 (2002).
fibrillation guidelines: recommendations for stroke prevention and rate/rhythm control. Can. 63. Holden, R. M. & Booth, S. L. Vascular Author contributions J. Cardiol. 28, 125–136 (2012).
calcification in chronic kidney disease: the role of All authors contributed equally to discussion of 50. Warkentin, T. E. et al. Recombinant factor VIIa vitamin K. Nat. Clin. Pract. Nephrol. 3, 522–523 content for the article, researching data to include in (rFVIIa) and hemodialysis to manage massive the manuscript, and reviewing and editing of the dabigatran-associated postcardiac surgery 64. Hart, R. G., Pearce, L. A. & Aguilar, M. I. Meta- manuscript before submission. R. G. Hart and bleeding. Blood 119, 2172–2174 (2012).
analysis: antithrombotic therapy to prevent J. W. Eikelboom wrote the manuscript.
578 OCTOBER 2012 VOLUME 8
2012 Macmillan Publishers Limited. All rights reserved

Source: http://mcmasterstrokeprogram.ca/workfiles/Anticoag%20of%20AF%20with%20CKD%20nrneph%202012.pdf