Anticoagulation in the Elderly

Abhay J. Dhond, MD, MPH, Hector I. Michelena, MD, Michael D. Ezekowitz, MD, PhD


Am J Geriatr Cardiol. 2003;12(4) 

In This Article

Treatment Strategies for AF

The first step in management is to determine if hemodynamic compromise is present and then institute therapy directed at removing reversible causes of AF, controlling the heart rate, and preventing or treating heart failure and ischemia.[18] If the patient is hemodynamically compromised, direct-current cardioversion should be performed with the patient heparinized. Intravenous calcium channel blockers or β blockers should be used initially to control the rapid ventricular response in hemodynamically stable patients. Digoxin, though widely used, is less effective than calcium channel blockers or β blockers in controlling the ventricular rate. Care must be taken in patients with sick sinus syndrome (tachycardia-bradycardia syndrome) as they may need a pacemaker in combination with rate controlling drugs.

It is well known that cardioversion restores sinus rhythm in 90% of cases. However, it is estimated that only 50% of patients remain in sinus rhythm 12 months after successful cardioversion.[19]

The rationale for maintenance of sinus rhythm is based on the observation that cardiac output and stroke volume, the primary measures of cardiac performance, are reduced in patients with AF.[20] Patients with AF show a significant increase in exercise tolerance following restoration of sinus rhythm.[21,22] While optimal anticoagulation significantly reduces the occurrence of stroke in patients with AF, it does not eliminate it.[2] Finally, anticoagulation may be contraindicated in up to 14%-16% of patients older than 75 years with AF.[23]

The question of whether maintenance of sinus rhythm with antiarrhythmic drugs (+ anticoagulation) is superior to the use of rate-controlling drugs (+ anticoagulation) was addressed by the Atrial Fibrillation Follow-up Investigation of Rhythm Management (AFFIRM) and the Comparison of Rate Control and Rhythm Control in Patients with Recurrent Persistent Atrial Fibrillation (RACE) trials.[24,25] In the AFFIRM trial[24] a total of 4060 patients with AF and other high-risk features for stroke, with a mean age of 69.7±9 years, were randomized to either treatment stategy. There was a nonsignificant trend toward increased mortality in the sinus maintenance group (5-year mortality of 23.8% vs. 21.3%; p =0.08). The majority of strokes in both groups occurred after warfarin was stopped or when the international normalized ratio (INR) was subtherapeutic. The authors concluded that the sinus maintenance strategy does not offer a survival benefit over rate-control strategy. The RACE trial[25] randomized 522 patients, with a mean age of 68±8 years, who had persistent AF after previous electrical cardioversion to receive treatment aimed at rate control or rhythm control. Patients in the rate-control group received oral anticoagulants and rate slowing medication. Patients in the rhythm-control group underwent serial cardioversions and received antiar-rhythmic drugs and oral anticoagulation. The primary end point (composite of death from cardiovascular causes, heart failure, thromboembolic complications, bleeding, the need for implantation of pacemaker, or severe adverse effects of antiarrhythmic drugs) showed a trend in favor of the rate-control group (occurred in 17.2% of participants in the rate-control group and 22.6% in the rhythm-control group). However, the absolute difference of -5.4% (90% CI, -11.0 to 0.4) did not meet the predetermined criteria for superiority (>10% difference). Thromboembolic complications were seen in 5.5% of the rate-control group and 7.9% of the rhythm-control group, respectively (absolute difference, 2.4%; 90% CI, -6.0 to 1.2). The authors concluded that rate control is not inferior to rhythm control for the prevention of death and morbidity from cardiovascular causes and may be appropriate therapy in patients with a recurrence of persistent AF after electrical cardioversion. We conclude that regardless of the strategy used, anticoagulation must be continued in patients at high risk for stroke. These studies included patients over 75 years of age.

There is compelling evidence for the important role of anticoagulation in preventing strokes in patients with AF for both primary and secondary prevention.[15,16,17,26,27,28] Warfarin was more effective than aspirin; either one was better than placebo. In the primary prevention setting, the pooled analysis of trials from the AFI revealed a 68% relative risk reduction[13] (absolute risk reduced from 4.5% to 1.4% per year) in stroke risk with warfarin anticoagulation compared with placebo. Overall, the relative risk reduction of stroke afforded by aspirin was 36%.13 In these trials most patients were between 66 and 75 years of age, and 24.2% were >75 years of age.

Regarding secondary prevention, the European Atrial Fibrillation Study[28] randomized patients with previous TIA or minor stroke to anticoagulation or treatment with either placebo or aspirin. This study showed a 66% relative risk reduction (absolute risk reduced from 12% to 4%) of subsequent stroke. The mean age of the patients was 71±7 years.

Aspirin and warfarin inhibit thrombosis by different mechanisms and theoretically coadministration of the two drugs may have a synergistic effect. The combination of low-intensity, fixed-dose warfarin plus aspirin has shown no synergistic effect and seems to be ineffective when compared with full anticoagulation for the prevention of stroke in nonvalvular AF. The SPAF III study[14] evaluated 1044 patients with AF, with a mean age of 72 (±9), who had at least one high-risk factor for thromboembolism. Subjects were randomized to a combination therapy group of low fixed-dose warfarin (titrated to an INR of 1.2-1.5 for initial dose adjustment) plus aspirin 325 mg/day or monotherapy with adjusted-dose warfarin (target INR, 2.0-3.0). The trial was terminated after a mean follow-up of 1.1 years when the rate of ischemic stroke and systemic embolization in the combination therapy group was 7.9% per year compared with 1.9% per year in the dose-adjusted warfarin group (p <0.001). The use of this combination for stroke prevention in AF is currently not recommended.[29]

The Combination Hemotherapy and Mortality Prevention (CHAMP) study[30] tested the hypothesis that aspirin and warfarin, when combined, would be more effective than aspirin monotherapy in patients with myocardial infarction (MI). The trial randomized 5059 patients, with a mean age of 64 (±10), to warfarin (target INR, 1.5-2.5) plus aspirin 81 mg/day vs. monotherapy with aspirin 162 mg/day. The death rate, recurrent MI rate, and strokes were comparable in the two groups. However, the rate of major bleeds was higher in the combination therapy group (1.28 vs. 0.72 per 100 person years of follow-up; p <0.001) and increased significantly with advancing age ( Table II ). However, the Warfarin, Aspirin, Reinfarction Study (WARIS II) trial[31] showed a significant (4%-5%) absolute risk reduction favoring anticoagulation in the composite end point of death, nonfatal reinfarction, and thromboembolic stroke in post-MI patients treated with warfarin plus aspirin or warfarin alone vs. aspirin alone. It is important to note that this study excluded patients >75 years of age and that the bleeding risk was three times higher in the anti-coagulation arms. The Antithrombotics in the Prevention of Reocclusion In Coronary Thrombolysis (APRICOT)-2 trial[32] studied 308 patients with MI who had patent infarct-related artery by angiography performed <48 hours after fibrinolytic therapy and were receiving aspirin and intravenous heparin. Patients were randomized to standard heparinization and continuation of aspirin alone or a 3-month combination of aspirin with moderate-intensity coumarin (target INR, 2.0-3.0), including heparinization till target INR was achieved. Reocclusion was observed in 15% of patients receiving the combination of coumarin and aspirin compared with 28% of those receiving aspirin alone (95% CI, 0.33-0.90; p <0.02). Survival rates free from reinfarction and revascularization were 86% and 66%, respectively ( p <0.01). Bleeding (Thrombolysis in Myocardial Infarction [TIMI] major and minor) was infrequent and the difference between the two groups was nonsignificant. The investigators concluded that a 3-month regimen of moderate intensity coumarin as an adjunctive to aspirin markedly reduced reocclusion and recurrent event rates after successful fibrinolysis. The Antithrombotics in the Secondary Prevention of Events in Coronary Thrombosis-2 (ASPECT-2) study[33] randomized, and followed-up for a maximum of 26 months, 999 patients who were admitted with MI or unstable angina to three arms for therapy with low-dose aspirin (80 mg/day), high-intensity oral anticoagulation (INR, 3.0-4.0), or combined low-dose aspirin (80 mg/day) and moderate-intensity oral anticoagulation (INR, 2.0-2.5). The primary end point (MI, stroke, or death) was reached in 9% of patients on aspirin, 5% on anticoagulants (95% CI, 0.30-1.00; p=0.0479), and 5% on combination therapy (95% CI, 0.27-0.92; p=0.03). Major bleeding was infrequent and there was no statistically significant difference between the three groups. However, minor bleeding was statistically higher in the combination therapy group. It should be noted that in both APRICOT-2 and ASPECT-2 the age of participants averaged a relatively young 58-60 years. Thus, at present the combination of aspirin and warfarin in routine management of elderly post-MI patients aged ≥75 needs further study.

The major complication of anticoagulant therapy is bleeding. There is a strong and proven relationship between the risk of bleeding and the intensity of anticoagulation. Randomized clinical trials show that high-intensity warfarin therapy (INR >3.0) doubles the risk of major hemorrhage as compared with warfarin therapy with INR between 2.0-3.0. The risk of intracranial hemorrhage increases dramatically with INR >4.0.[34]

The major determinants of oral anticoagulantinduced bleeding are the intensity of anticoagulation, patient characteristics, the concomitant use of drugs that interfere with hemostasis, poorly controlled hypertension, and the duration of therapy. Increased variation in the INR is associated with an increased frequency of hemorrhage independent of the mean INR.[34] The SPAF II[35] study revealed an increased risk of anticoagulation-induced intracranial bleeding in patients >75 years of age (1.8% per year), probably due to the fact that the range of INR used in this study was 2.0-4.5.

In the five randomized trials analyzed by AFI, virtually no increase in the frequency of major bleeding (1.3%, 1.0%, and 1.0% per year for warfarin, aspirin, and placebo groups respectively) was found. This difference was not statistically significant. The INR ranges of these studies were 2.8-4.2 for the Copenhagen Atrial Fibrillation, Aspirin, and Anticoagulant Therapy Study (AFASAK) study,[26] 2.0-4.5 for SPAF,[15] 1.5-2.7 for BAATAF,[16] 2.0-3.0 for the Canadian Atrial Fibrillation Anticoagulation (CAFA) study,[27] and 1.4-2.8 for SPINAF.[17] The observed rate of intracranial hemorrhage in patients >75 years old was only 0.8% per year.[29] In the high-risk arm of SPAF III,[14] patients had a mean age of 71 years and an INR range of 2.0-3.0 and the rate of intracranial hemorrhage in this trial was only 0.5% per year. In the AFASAK 2 study,[36] the rate of bleeding events associated with the incidence of thromboembolic events in patients receiving fixed mini-dose warfarin, fixed mini-dose warfarin plus aspirin, aspirin and adjusted-dose warfarin therapy (INR, 2.0-3.0) was studied in 677 patients, and no significant influence of age on the risk of bleeding was found.

Regarding high blood pressure as a risk factor for anticoagulation-induced intracranial bleed, it is critical to understand that reducing systolic blood pressure is as important as reducing diastolic blood pressure as was shown in the Systolic Hypertension in the Elderly Program (SHEP) trial.[37] We recommend that hypertension be vigorously treated so that systolic blood pressure is maintained below 140 mm Hg and diastolic blood pressure below 90 mm Hg.

Patients aged >75 may have many years to live. Treatment goals should not only prevent death and prolong life, but also prevent disability and preserve quality of life. In patients with AF, the high risk of thromboembolism and of bleeding when anticoagulated, calls for individual patient-tailored decisions.[38]

Eckman et al.[38] proposed an interesting decision-analysis model that evaluates two choices: choice one, administer anticoagulant therapy; choice two, withhold anticoagulant therapy. This model takes into account risks of both choices, temporary and permanent morbidity of both choices, mortality of both choices, and most importantly, it considers the quality of the patient's life (expressed quantitatively) as a major factor in clinical decision making. The clinician can calculate a patient-specific risk for both ischemic stroke and major bleeding events using Table III , which is based on the AFI analysis[13] for the stroke risk, and on a retrospective review of 565 patients starting outpatient treatment with warfarin[39] for the bleeding risk. Once risk is calculated for both events by adding each value, refer to the Figure to determine whether the combination of risk weights falls within the "anticoagulate" or the "do not anticoagulate" range.[38] For example: An 81-year-old patient with AF, hypertension, and past TIA has an ischemic stroke risk of 2.50 (1.02+0.49+0.99) and a major bleed risk of 1.87 (1.03+0.84). This combination of risk weights, when taken to the Figure, clearly falls in the anticoagulate region.

Two-way sensitivity analysis of relation between the risk of thromboembolic and bleeding events calibrated for use with Table III. The sum of the risk weights for ischemic stroke appears on the horizontal axis, while the sum of the risk weights for major bleeds appears on the vertical axis. A second horizontal axis depicts the probability of ischemic stroke. Two diagonal lines separate the graph into regions in which anticoagulation or no anticoagulation are preferred. The upper diagonal line represents the lower bound for the baseline bleeding risk (0.5% per year) for patients in atrial fibrillation without additional risk factors, and the lower line represents the upper bound for the baseline rate of bleedin (1.3% per year) in low-risk patients.Modified with permission from Chest. 1998;114:699S-714S.[38]

Patients over age 75 with AF have both a higher risk of stroke and a slightly increased incidence of bleeding while receiving oral anticoagulation. However, there is currently no evidence in favor of withholding anticoagulation in these patients (unless otherwise contraindicated). There is also no evidence that the recommended range of INR 2.0-3.0 should be changed for them either. Age >75 years is a high-risk factor for stroke in AF and standard oral anticoagulation for these patients is a grade 1A indication.[29]