Acute and Chronic Management of Atrial Fibrillation in Patients With Late-Stage CKD

Prevalence and Incidence of AF 

AF is the most common sustained cardiac arrhythmia. The prevalence and incidence of AF are high, particularly with advancing age.1, 2, 3, 4, 5 In the Framingham Heart Study, the prevalence of chronic AF was 2% in persons aged 60 to 69 years, 5% in persons aged 70 to 79 years, and 9% in persons aged 80 to 89 years.1 Similarly, in 5,201 persons aged 65 years and older in the Cardiovascular Health Study (CHS), the prevalence of AF was 5% in women and 6% in men,4 while in another study of 1,563 community-dwelling elderly individuals with mean age of 80 years, the prevalence of chronic AF was 9%.5 In the CHS, the incidence of AF was 19.2 per 1,000 person-years.6

AF may be paroxysmal, persistent, or permanent. Episodes of paroxysmal AF may last from a few seconds to several weeks. Notably, 68% of persons presenting with AF of less than 72 hours' duration spontaneously convert to sinus rhythm.7 Episodes of persistent AF last longer than 7 days but less than 1 year. AF in which cardioversion has failed or lasts longer than 1 year is usually termed permanent.

CKD is associated with an increased prevalence and incidence of AF. Across multiple observational studies, the prevalence of AF in individuals with CKD varies from 7% to 27% (Table 1).8, 9, 10, 11, 12, 13 The incidence of AF is also quite high; in the largest study of dialysis patients to date, Abbott et al report an incidence rate of AF of 13 per 1,000 person-years among 3,374 patients receiving either hemodialysis or peritoneal dialysis.14 In all of these studies, patients with AF were older than patients with sinus rhythm.8, 9, 10, 11, 12, 13, 14

Table 1. Prevalence of Atrial Fibrillation in Patients With Late-stage CKD

	Study	N	Mean Age	Kidney Disease Status	Prevalence of Atrial Fibrillation
	Abe et al8	72	55	On hemodialysis	7%
	Fabbian et al9	316	63	On hemodialysis	23%
	Vazquez et al10	190	60	On hemodialysis	14%
	Genovesi et al11	488	67	On hemodialysis	27%
	Das et al12	53	70	With stage 3 to 4 CKD	19%
	Atar et al13	275	49	On hemodialysis	11%

Abbreviation: CKD, chronic kidney disease.

Risks Associated With AF 

In the general population, AF is clearly associated with worse outcomes. For example, in the Framingham Study population, a longitudinal cohort of community-dwelling adults from Framingham, Massachusetts, the incidence of death from cardiovascular causes was 2.7 times higher in women and 2.0 times higher in men with chronic AF than in women and men with sinus rhythm.15 The Framingham Study also demonstrated that, after adjustment for preexisting cardiovascular conditions, the odds ratio for mortality in persons with AF was 1.9 (95% CI, 1.5-2.2) in women and 1.5 (95% CI, 1.2-1.8) in men.16 Similarly, in a study of 1,359 elderly long-term care patients with known heart disease and mean age of 81 years, those individuals with chronic AF were at 2.2 times (95% CI, 1.8-2.8) increased risk of having new coronary events compared with patients with sinus rhythm after controlling for other prognostic variables.17 In the Copenhagen City Heart Study (CCHS), the effect of AF on the risk of cardiovascular death was significantly increased 4.4 times (95% CI, 2.9-6.5) in women and 2.2 times (95% CI, 1.6-3.1) in men.18

Critically, AF is an independent risk factor for stroke, especially in older persons.1, 2 Research from the Framingham Study showed that the relative risk (RR) of stroke in patients with nonvalvular AF compared with patients with sinus rhythm was increased 2.6 times in patients aged 60 to 69 years, increased 3.3 times in patients aged 70 to 79 years, and increased 4.5 times in patients aged 80 to 89 years.1 Similarly, in a study of 2,101 elders with mean age of 81 years residing in a long-term health care facility, chronic AF was an independent risk factor for thromboembolic stroke, with a RR of 3.3 (95% CI, 2.4-4.5).2 The incidence of thromboembolic stroke in this study was 38% and 72% in older persons with chronic AF versus 11% and 24% in older persons with sinus rhythm at 3 and 5 years, respectively.2 In a second study of 1,476 patients who had 24-hour ambulatory electrocardiograms, the incidence of thromboembolic stroke after 37 months of follow up was 43% among 201 patients with AF (RR, 3.3 [95% CI, 2.4-4.5] compared with sinus rhythm), 17% among 493 patients with paroxysmal supraventricular tachycardia, and 18% among 782 patients with sinus rhythm.20 This risk may be greater in women than in men; for example, in the Euro Heart Survey on Atrial Fibrillation, women with AF had a 1.83 times (95% CI, 1.10-3.03) greater risk of stroke than men with AF.19

Assessing Stroke Risk 

The CHADS2 score (Table 2) is a validated predictive instrument used to assess stroke risk in patients with AF in the general population; it gives 1 point for congestive heart failure,1 point for hypertension, 1 point for age greater than 75 years, 1 point for diabetes mellitus, and 2 points for previous stroke or transient ischemic attack.21 Table 3 shows the relationship between the CHADS2 score and the annual risk of stroke.21 Current recommendations suggest that individuals in the general population with a CHADS2 score of 0 are at lower risk of stroke and may be treated with aspirin only, while those with a score of 1 are at moderate risk of stroke and should receive either aspirin or warfarin, and those with a score of 2 or more are at highest risk of stroke and should receive warfarin. Impressively, after 31 months of follow up in a study of 521 patients with AF, those individuals with a CHADS2 score of 5 or 6 had a 52-fold increased risk of stroke as compared with individuals with a CHADS2 score of 0.22 The CHADS2 score has not been studied in the dialysis population.

Table 2. CHADS2 Score

	Variable	Number of Points
	Congestive heart failure	1
	Hypertension	1
	Age older than 75 years	1
	Diabetes mellitus	1
	Prior stroke or transient ischemic attack	2

Source: Gage et al.21

Table 3. Relationship Between CHADS2 Score and Risk of Stroke Each Year

	CHADS2 Score	Risk of Stroke Per Year
	0	1.9%
	1	2.8%
	2	4.0%
	3	5.9%
	4	8.5%
	5	12.5%
	6	18.2%

Source: Gage et al.21

Risk Associated With AF in Kidney Disease 

Patients with kidney failure requiring dialysis have an increased prevalence of cardiovascular events and cardiovascular mortality, including stroke.23, 24, 25, 26, 27 AF may contribute to this increased risk. For example, in the Genovesi study of 476 hemodialysis patients, the presence of AF was associated with a 65% increased risk of mortality (hazard ratio, 1.65 [95% CI, 1.18-2.31]); interestingly, while the mortality risk was driven by cardiovascular death, AF was not associated with incident stroke.28 Similarly, in the Vazquez study10 of 190 hemodialysis patients, 26 patients had AF at baseline. Of these, 23% of patients with AF and 6% of patients with sinus rhythm on hemodialysis died during 1 year of follow up, while 35% of patients with AF and 4% of patients with sinus rhythm developed thromboembolic events. Unlike the findings in Genovesi et al, multivariable analysis showed that AF was the only independent predictor of thromboembolic events at 1 year (odds ratio, 8.03 [95% CI, 2.35-27.4]).10 Over a 50-month follow-up period, Vazquez et al later reported an 81% mortality rate among patients with AF at baseline versus 28.7% among those without AF (odds ratio, 2.1 [CI, 1.2-3]) in analyses adjusted for age and serum albumin.29 Of the 26 patients with AF, 11 (42.3%) had 14 thromboembolic episodes while thromboembolic events occurred in 9.7% of the patients with sinus rhythm at baseline (RR, 4.6 [95% CI, 2.4-8.6]) in unadjusted analyses.29 Vazquez et al subsequently reported on the 164 hemodialysis patients without AF at baseline, and described that, over a mean follow-up period of 47 months, 20 patients developed AF; those individuals with incident AF were at significantly increased risk of subsequent thromboembolic events (RR, 5.2 [95% CI, 2.1-12.4]).30

Use of Warfarin in AF 

In the general population, numerous prospective, randomized, double-blind, placebo-controlled studies31, 32, 33, 34, 35, 36, 37, 38, 39, 40 and prospective, nonrandomized observational data41, 42 have demonstrated that high-risk patients with nonvalvular AF treated with warfarin to maintain an international normalized ratio (INR) between 2.0 and 3.0 have a significant reduction in thromboembolic stroke with an acceptable risk of bleeding.31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43 This was nicely demonstrated in the Stroke Prevention and Atrial Fibrillation (SPAF) Study III, in which patients with AF, a mean age of 72 years, and considered at high risk of thromboembolic stroke, were randomized to warfarin with an INR target of 2 to 3 versus warfarin with an INR target of 1.2 to 1.5 plus 325 mg aspirin daily. Individuals randomized to the higher INR target had a significant decrease of 72% in ischemic stroke or systemic thromboembolism compared with patients randomized to treatment with 325 mg oral aspirin daily plus oral warfarin to achieve an INR between 1.2 and 1.5.33 In the general population, oral anticoagulant therapy has proven effective in preventing thromboembolic events both in patients with paroxysmal AF as well as in patients with chronic AF. For example, in ATRIA (Anticoagulation and Risk Factors in Atrial Fibrillation Study), warfarin was associated with significantly lower adjusted thromboembolic rates for both women (60% reduction) and men (40% reduction) with similar annual rates of major bleeding (1.0% and 1.1%, respectively).44 Critically, all of these randomized clinical trials have excluded patients with late-stage CKD.

There are several observational studies examining anticoagulation in populations with CKD. In 1 study of patients with AF and a high proportion of chronic kidney failure undergoing percutaneous coronary intervention with stent implantation, age (P < 0.01) and nonanticoagulation (P = 0.02) were independent predictors of death, acute myocardial infarction, or target lesion revascularization in patients with AF.45 However, there was a borderline significant finding that individuals with chronic renal failure (undefined by the authors) were less likely to be anticoagulated at discharge, making it difficult to draw any conclusions regarding patients with kidney failure from this article.45 In 1 observational study analyzing US Renal Data Service DMMS Wave 2 (Dialysis Morbidity and Mortality Wave 2 Study) data, there were 123 individuals hospitalized for AF and 90 who died during the study period (mean follow up of 2.92 ± 1.14 years). Among these individuals, only use of warfarin and systolic blood pressure greater than 130 mm Hg were associated with increased survival.14

Many physicians consider CKD a condition that confers a high risk of bleeding complications in patients given warfarin for AF and are reluctant to prescribe warfarin for this indication.46 However, this viewpoint has been challenged.47, 48, 49 In 1 small study of hemodialysis patients with mean follow up of 20 months, hemorrhagic events occurred in 31% of 29 patients (7 with AF) receiving oral anticoagulants and in 14% of 211 patients not receiving oral anticoagulants; importantly, no patients receiving oral anticoagulants developed a fatal hemorrhage, an intracranial hemorrhage, or serious clinical sequelae.48 Notably, in a more recent systematic review of warfarin use in hemodialysis patients, Elliott et al noted major bleeding rates ranging from 0.1 to 0.54 events per patient-year of warfarin exposure and stressed that the true bleeding risk associated with the use of warfarin in hemodialysis patients remains unknown, given the reliance on small observational studies with potential confounding by comorbid conditions.50

Studies also need to be performed to investigate whether other clinical factors in dialysis patients would predispose to greater benefit associated with anticoagulation. These factors include the increased risk of pulmonary embolism observed in long-term dialysis patients,51 as well as therapy targeted at hemodialysis patients with detection of left atrial appendage thrombi by transesophageal echocardiography, a condition (noted in 33% of 215 patients on hemodialysis in 1 study52) that identifies a very high-risk group for thromboembolic stroke in the general population.

Quinn et al specifically examined the issue of anticoagulation for AF of hemodialysis patients in a cost-utility analysis. At a threshold of $100,000 per quality-adjusted life-year, the probabilities that no therapy, aspirin, or warfarin was the most efficient therapy were 20%, 23%, and 58%, respectively.53 An editorial accompanying this article recommended that until more data are available, standard warfarin anticoagulation should be considered in hemodialysis patients with AF, adhering to screening and monitoring as in the general population.54

Because of the high prevalence of AF and its association with an increased incidence of thromboembolic events in patients with late-stage CKD, it is essential that prospective, double-blind, placebo-controlled trials be performed in these patients to determine the efficacy of oral anticoagulant therapy in preventing thromboembolic events and the incidence and type of hemorrhagic events. Until these data are available, this author favors treating AF patients on hemodialysis with warfarin on an individual basis, taking into account both the thromboembolism risk as well as the hemorrhagic risk.

Ventricular Rate Control 

Rate control may be accomplished either through rhythm control (with conversion back to sinus rhythm) or through use of medications that affect the rate of transmission across the atrioventricular (AV) node. Emergent rhythm control through direct-current (DC) cardioversion should be performed immediately in patients who have paroxysmal AF with a very fast ventricular rate associated with an acute myocardial infarction, chest pain caused by myocardial ischemia, hypotension, severe congestive heart failure, syncope, or pre-excitation syndromes.47, 55 In the acute setting, medications including intravenous β-blockers,47, 55 calcium channel blockers,47, 55 other anti-arrhythmic agents (specifically amiodarone), and cardiac glycosides (specifically digoxin) have all been used with varying results to reduce a very rapid ventricular rate associated with AF except in patients with pre-excitation syndromes. These medications are discussed below and advantages, disadvantages, and special considerations for medications in each of these classes are summarized in Table 4.

Table 4. Selected Agents Frequently Used for Rate Control in Atrial Fibrillation

	Agent	Mechanism	Acute	Chronic	Removal by Dialysis		Comments
					HD	PD
	β-Blockers⁎
	Atenolol	β1-selective	NA	50 mg orally, 1x/d	75%	53%	Active metabolites may accumulate in ESRD, predisposing to heart block Consider dose reduction or increase in dosing interval in kidney failure
	Carvedilol	Nonselective β1 antagonism with additional α1 antagonism	NA	6.25-25 mg orally, 2x/d	None	None	Lipophilic Hepatic metabolism Metabolite, R-carvedilol, has prolonged half-life with reduced kidney function but minimal activity Significant reduction in morbidity and mortality in dialysis patients with dilated cardiomyopathy in 1 clinical trial
	Esmolol	β1-selective	500 μg/kg IV over 1 min followed by 60 μg/kg/min IV	NA	None	None	IV only Marked prolongation in half life of primary metabolite (ASL-8123, very weakly active)
	Metoprolol	β1-selective	2.5-5 mg IV bolus over 2 min; up to 3 doses	25-100 mg orally, 2x/d	High	Unknown	Hepatic metabolism via CYP2D6 Less active metabolite, α-hydroxymetoprolol, has prolonged half life in kidney failure
	Calcium Channel Blockers†
	Diltiazem		0.25 mg/kg IV over 2 min	60-90 mg orally, 4x/d	Low	Low	Metabolized by the liver and excreted in the kidneys and bile; metabolites have partial activity Consider dose reduction or increase in dosing interval in kidney failure Likely better option than verapamil
	Verapamil		0.075 mg/kg IV over 2 min	80 mg orally, 3x/d	Low	Yes	Extensive hepatic metabolism with 70% kidney excretion (3%-4% as unchanged drug) Consider dose reduction or increase in dosing interval in kidney failure
	Other Agents
	Amiodarone	Class III anti-arrhythmic agent with additional non-selective α and β antagonism, as well as class I and class IV properties	150 mg IV over 10 min followed by 0.5-1 mg/min IV	200 mg daily following loading doses	None	None	Follow INR and digoxin levels after initiating due to interactions; reduce dose of warfarin Thyroid toxicity; baseline thyroid function tests recommended Lung toxicity; baseline pulmonary function tests recommended
	Digoxin	Inhibition of the Na/K ATPase pump which acts to increase the intracellular Na-Ca exchange to increase intracellular Ca	0.375-0.75 mg load (divided over several doses every 2-6 h) followed by 0.125-0.25 mg every 48-72 h	0.0625-0.125 mg every 48-96 h	0-5%	0-5%	High risk of digoxin toxicity Follow levels, with target levels below 1 Avoid dramatic K shifts (eg, avoid dialysate bath concentration of 1 mmol/L K) if possible

Note: Compiled from package inserts, the KDOQI Workgroup,57 and Fuster et al.47 Abbreviations: AV, atrioventricular; ESRD, end-stage renal disease; INR, international normalized ratio; IV, intravenous; NA, not applicable; NA/K ATPase, sodium/potassium transporting adenosine triphosphatase.

⁎ β-blockers reduce ventricular rate by depressing AV node conduction. All β-blockers may reduce blood pressure, reduce contractility in the acute setting, and cause bronchospasm. † Calcium channel blockers inhibit calcium ion from entering the “slow channels” during depolarization, slowing AV node conduction. All nondihydropyridine calcium channel blockers may reduce blood pressure and reduce contractility leading to heart failure.

For long-term management, while there are multiple options, oral lipophilic β-blockers such as metoprolol, carvedilol, or propranolol are likely preferable as first-line agents to slow the ventricular rate in patients with AF.55, 56 In settings where β-blockers may be contraindicated, such as severe bronchospasm, oral diltiazem would be the preferred agent assuming that the LVEF is 40% or higher.55 Of note, several medications commonly used in the general population or even in later stages of CKD may be associated with greater risk of adverse effects, most notably heart block, in a dialysis population. This primarily reflects decreased clearance of the active drug or active metabolites, and examples include atenolol, verapamil, and digoxin.58 With atenolol and verapamil in particular, other drugs in the same class may have a better safety profile, although in select cases the prolonged half-life can be advantageous as it may allow for observed thrice-weekly therapy at dialysis as was done in 1 study of atenolol.59 Verapamil has been associated with third-degree atrioventricular block in dialysis patients.60 Digoxin, while commonly used, is ineffective in reducing a rapid ventricular rate in any condition associated with increased sympathetic tone and has an extremely narrow therapeutic window in dialysis patients, particularly given the rapid fluctuations in potassium balance induced by hemodialysis.58, 61 Given their negative inotropic effects, the nondihydropyridine calcium channel blockers (diltiazem and verapamil) should be avoided if the patient has a LVEF less than 40%.62, 63 Sotalol, an infrequently used medication for rate control in AF, is contraindicated in patients with severe kidney disease.58

Amiodarone is an effective drug for reducing a rapid ventricular rate in AF as well as for long-term rate control.47, 55, 64 The noncompetitive β-receptor inhibition and calcium channel blockade associated with amiodarone represent powerful AV nodal conduction depressants. However, the adverse side-effect profile of amiodarone may limit its routine use in the treatment of AF. This adverse effect profile was demonstrated in the CASCADE (Cardiac Arrest in Seattle: Conventional Versus Amiodarone Drug Evaluation) Study, where the incidence of pulmonary toxicity was 10% at 2 years in patients receiving amiodarone at a mean dose of 158 mg daily.65 The incidence of adverse effects from amiodarone may approach 90% after 5 years of therapy.66 If intravenous β-blockers and diltiazem are unable to be used in the acute or long-term treatment of AF with a very rapid ventricular rate, intravenous or oral amiodarone, respectively, could be considered (Table 4).

Ventricular Rate Control Versus Maintenance of Sinus Rhythm 

DC cardioversion of AF has a higher success rate in converting AF to sinus rhythm and a lower incidence of cardiac adverse effects than anti-arrhythmic drug therapies.47, 55 On the basis of numerous studies,67, 68, 69, 70, 71 this author prefers ventricular rate control over maintenance of sinus rhythm facilitated with the use of available anti-arrhythmic agents. Evidence supporting this position comes from the AFFIRM (Atrial Fibrillation Follow-up Investigation of Rhythm Management) Study, which randomized 4,060 patients (mean age 70 years and 39% women) with paroxysmal or chronic AF of less than 6 months' duration at high risk for stroke to either maintenance of AF with ventricular rate control or to an attempt to maintain sinus rhythm with anti-arrhythmic drugs following successful cardioversion.72 Patients in both arms of this study were treated with warfarin. All-cause mortality at 5 years was nonsignificantly increased 15% in the maintenance of sinus rhythm group compared with the ventricular rate control group (24% v 21%, P = 0.08).72 There was a non–statistically significant reduction in the rate of thromboembolic stroke in the ventricular rate control group (5.5% v 7.1%), and all-cause hospitalization was significantly lower in the ventricular rate control group (73% v 80%, P < 0.001).72 In both groups, the majority of strokes occurred after warfarin was stopped or when the INR was subtherapeutic. Importantly, there was no significant difference in quality of life or functional status between the 2 treatment groups.72 In post hoc subgroup analyses of AFFIRM data, rhythm control did not improve mortality, hospitalization, or New York Heart Association class in patients with LVEF of 40% to 49%, 30% to 39%, or less than 30%.67

Management of the Patient Presented 

The patient presented in the introductory case has AF with a very rapid ventricular rate that precipitated congestive heart failure with pulmonary congestion. This woman requires immediate direct current cardioversion to sinus rhythm; evidence extrapolated from the general population suggests that heparin should be administered at the time of cardioversion and oral warfarin therapy should be initiated.47 With improvement in ventricular filling that can be seen with rate control and restoration of sinus rhythm, pulmonary congestion may subside without emergent dialysis.73 In nondialysis patients in the setting of heart failure associated with an abnormal LVEF, evidence favors treatment with a lipophilic β-blocker such as metoprolol or carvedilol as well as with an angiotensin-converting enzyme (ACE) inhibitor.73 Despite a lack of conclusive evidence, I would favor the long-term use of an ACE inhibitor (or an angiotensin receptor blocker) in this patient given the reduced ejection fraction if the blood pressure is sufficiently high to allow this therapy.

Should AF recur, the β-blocker may be effective in preventing an accompanying rapid ventricular rate. β-Blockers reduce mortality in patients with congestive heart failure and abnormal LVEF due to numerous mechanisms including antagonizing neurohormonal systems that cause myocyte apoptosis, myocyte necrosis, myocyte hypertrophy, extracellular matrix alterations, and β-receptor uncoupling. These mechanisms are discussed in detail elsewhere.74 Diltiazem and verapamil, both negative inotropes, should be avoided in this patient because the LVEF is abnormal, and these drugs will increase mortality in patients with congestive heart failure and an abnormal LVEF.62, 63 Digoxin should also be avoided due to a lack of efficacy for rate and rhythm control as well as a narrow therapeutic window in patients with end-stage renal disease.58 Because of its adverse effect profile,65, 66 long-term amiodarone use should be undertaken with caution in patients with AF and CKD. Additionally, amiodarone interacts with numerous drugs58: it may cause atrioventricular block in patients treated with β-blockers, it increases serum digoxin levels, it increases the risk of myopathy and rhabdomyplysis in patients treated with simvastatin or lovastatin, and it potentiates the effect of warfarin, increasing the risk of bleeding. If the β-blocker was ineffective in controlling a rapid ventricular rate in this patient, complete atrioventricular block produced by radiofrequency catheter ablation followed by permanent pacemaker implantation could be considered.75

Blood pressure goals remain poorly defined in hemodialysis patients, with current guidelines suggesting predialysis and postdialysis blood pressure goals of less than 140/90 mm Hg and less than 130/80 mm Hg, respectively, assuming that orthostatic changes can be avoided.57, 76 Again extrapolating from the general population, I would favor continuing the statin,77, 78, 79 although there currently is a lack of evidence for statin use for primary or secondary cardiovascular outcome prevention in dialysis patients.80

Based on a CHADS2 score of 6, this patient theoretically has a 18.2% predicted annual risk of thromboembolic stroke.21 On the basis of this information, again extrapolating from the general population, I favor treating her with oral warfarin to maintain her INR between 2.0 to 3.0. I would educate her on further dietary changes that will promote stable warfarin metabolism, and, given the inherent bleeding risk with warfarin, would encourage her to avoid other medications that may increase her bleeding risk, in particular nonsteroidal anti-inflammatory drugs.58