Miba.auc.dk

Antithrombotic Therapy in
for the prevention of stroke in AF have been published Atrial Fibrillation
(Tables 1– 4). The study designs will be briefly described,according to the type of antithrombotic regimen studied.
Gregory W. Albers, MD, Chair; James E. Dalen, MD, MPH;Andreas Laupacis, MD; Warren J. Manning, MD; Oral Anticoagulation vs Control: Six studies7,15–19 were Palle Petersen, MD, DMSc; and Daniel E. Singer, MD randomized, controlled trials comparing oral anticoagula-tion (OAC) with control. In the Canadian Atrial Fibrilla- Abbreviations: ACUTE ϭ Assessment of Cardioversion Using
tion Anticoagulation (CAFA) study18 and the Stroke Pre- Transesophageal Echocardiography; AF ϭ atrial fibrillation; vention in Nonrheumatic Atrial Fibrillation (SPINAF) AFASAK ϭ Atrial Fibrillation Aspirin and Anticoagulation; AFI ϭ Atrial Fibrillation Investigators; BAATAF ϭ Boston Area study,19 assignment to anticoagulation or placebo groups Anticoagulation Trial for Atrial Fibrillation; CAFA ϭ Canadian was double blind, while anticoagulation administration Atrial Fibrillation Anticoagulation; CI ϭ confidence interval; was open labeled in the Stroke Prevention in Atrial DC ϭ direct current; EAFT ϭ European Atrial Fibrillation Trial; Fibrillation (SPAF)-1 study,16 the Atrial Fibrillation Aspi- ESPS ϭ European Stroke Prevention Study; INR ϭ interna- tional normalized ratio; LV ϭ left ventricular; MI ϭ myocardial rin and Anticoagulation (AFASAK)-1 study,15 and the infarction; NNT ϭ number needed to treat for 1 year; OAC ϭ European Atrial Fibrillation Trial (EAFT).7 In the Boston oral anticoagulation; PAF ϭ paroxysmal atrial fibrillation; RR ϭ risk reduction; RRR ϭ relative risk reduction; SIFA ϭ Studio (BAATAF),17 the control group was not administered Italiano Fibrillazione Atriale; SPAF ϭ Stroke Prevention in Atrial anticoagulation but could choose to take aspirin (46% of Fibrillation; SPINAF ϭ Stroke Prevention in Nonrheumatic Atrial Fibrillation; TEE ϭ transesophageal echocardiography; the patient-years in the control group were contributed by patients who were receiving aspirin regularly). Among the (CHEST 2001; 119:194S–206S)
studies, the target international normalized ratio (INR)varied from approximately 1.4 to 2.8 in the SPINAF Atrial fibrillation (AF) is the most common sustained
arrhythmia and is an important independent risk factor for stroke. AF is present in Ͼ 2 million people in the Aspirin vs Placebo or Control: Five studies compared United States.1 Its prevalence begins to increase in both aspirin with control: four studies7,11,12,14–16 were placebo- genders after age 40 years and rises rapidly after age 65 controlled, and one study23 had a nontreatment control.
years.2–5 AF is particularly common in the elderly, reach- The dose of aspirin varied between 325 mg/d16 and 125 mg ing a prevalence of roughly 10% in those Ͼ 80 years old.2–5 The median age of patients with AF is approximately 75years.1 The condition is more prevalent in men than inwomen.3–5 However, because there are more women than OAC vs Aspirin: Five studies7,13,15,22,25 compared OAC men in the older age groups, the absolute number of with aspirin. In SPAF-2,22 patients who had been random- ized to aspirin or warfarin in the SPAF-1 study16 continued The rate of ischemic stroke among patients with AF with their assigned treatment. Patients originally assigned included in clinical trials of primary prevention and not to placebo and 419 new patients were randomized to treated with antithrombotic therapy averages about 5%/yr, warfarin or aspirin. Randomization was stratified accord- with wide, clinically important variation among subpopu- ing to the patient’s age (Ͻ 75 years; Ն 75 years).
lations of AF patients.6–9 AF becomes an increasinglyimportant cause of stroke with advancing age. In the OAC vs Low-Dose OAC and Aspirin: In the SPAF-3 Framingham Heart Study,2 the attributable risk of stroke high-risk study,8 AF patients who had at least one of four in AF patients rose from 1.5% in the 50- to 59-year age thromboembolic risk factors (congestive heart failure or group to 23.5% in the 80- to 89-year age group. In patients left ventricular [LV] fractional shortening Յ 25%, history Ͼ 80 years old, AF was the only cardiovascular condition of a previous thromboembolism, systolic BP Ͼ 160 mm associated with an increased risk of stroke.2 Hg at study entry, or female gender Ͼ 75 years old) were This chapter deals primarily with stroke prevention randomized to either a combination of low-intensity, when AF is not associated with rheumatic mitral valve fixed-dose warfarin (INR 1.2 to 1.5; daily dose of warfarin disease or prosthetic heart valves. These specific condi- Յ 3 mg) plus aspirin (325 mg/d), or to adjusted-dose tions are discussed in the chapters on valvular heart warfarin (target INR 2.0 to 3.0). The AFASAK-2 study13 disease and prosthetic heart valves.
randomized patients to warfarin, 1.25 mg/d, and aspirin, 300mg/d, or to adjusted-dose warfarin (target INR 2.0 to 3.0).
OAC vs Low-Dose Anticoagulation: Three studies13,24,25 have compared adjusted-dose anticoagulation with lower doses of OAC: warfarin, 1.25 mg/d, in two studies,13,24 andwarfarin (target INR 1.1 to 1.6) in the third study.25 During the last decade, many studies7–25 assessing the efficacy and safety of different antithrombotic therapies Other Antiplatelet Agents: The Studio Italiano Fibril- lazione Atriale (SIFA) study10 randomized AF patients Correspondence to: Gregory W. Albers, MD, Stanford StrokeCenter, Building B, Suite 325, 701 Welch Rd, Palo Alto, CA with a recent nondisabling stroke or transient ischemic 94304-1705; e-mail: [email protected] attack (TIA) to therapy for 1 year with either indobufen Sixth ACCP Consensus Conference on Antithrombotic Therapy Table 1—Characteristics of AF Studies*
*S ϭ Ischemic stroke; NSE ϭ Non-CNS systemic embolus; ICB ϭ Intracranial bleed; FB ϭ fatal bleed; VD ϭ vascular death; PE ϭ pulmonary †This represents only the patients in ESPS-2 with AF.
‡Primary outcome not specified; however, sample size calculated using ischemic stroke and intracranial bleed.
(a reversible inhibitor of cyclooxygenase), 200 mg bid, risk factors in the SPAF-3 high-risk study8 (see above), or warfarin (INR 2.0 to 3.5) within 15 days of the were administered aspirin only, 325 mg/d, and followed in qualifying ischemic event. In the second European a nonrandomized, longitudinal, cohort study. This nonran- Stroke Prevention Study (ESPS-2),11,12 patients with a domized study does not provide data regarding the effi- TIA or stroke within the previous 3 months were cacy of aspirin for stroke prevention, but it is useful in randomized to one of four treatments: (1) placebo; (2) determining the risk of stroke in selected patients with AF aspirin, 25 mg/d bid; (3) extended-release dipyridamole, 200 mg/d bid; or (4) aspirin, 25 mg/d bid, and extended-release dipyridamole, 200 mg/d bid.
Aspirin Therapy in Low-Risk Patients: Finally, in the The primary outcome events in each study are listed in SPAF-3 low-risk study,9 AF patients considered to be at Table 1. The data reported herein are the results of the low risk of stroke, based on the absence of any of the four intention-to-treat analyses, although it is not clear if the Table 2—Treatment Arms in AF Studies
Aspirin, 325 mg, plus warfarin (INR 1.2–1.5) *ESPS-2 also included two other treatment groups: (1) extended-release dipyridamole, 200 mg bid; (2) aspirin, 25 mg bid, plus extended-release †Indobufen, 200 mg bid (not aspirin).
‡Prothrombin time ratio-based target range; INR range is estimated.
§Posada and Barriales23 evaluated two doses of aspirin: 125 mg qd and 125 mg every other day.
CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT Table 3—Primary Outcome Events in AF Studies*
*NS ϭ not significant.
†Based on intention-to-treat analysis.
‡ESPS-2 also had two other treatment arms: dipyridamole, 200 mg bid (annual stroke rate, 15.1%), and dipyridamole, 200 mg bid and aspirin 25 data in the study by Posada and Barriales23 were analyzed control, there was a decrease in the rate of primary according to the intent-to-treat principle. All studies con- outcome events in adjusted-dose anticoagulation-treated sidered stroke a primary event, and some studies also patients compared with control patients, which reached or included other vascular events as primary events. The exceeded conventional statistical significance in all studies definition of major bleeding varied slightly among studies.
except the CAFA study.18 The CAFA study18 was stopped In general, bleeding was classified as major if transfusion early because of the results of the other trials (Table 3).
was required, if the patient was hospitalized, or if the Pooling the results of all of these trials except the EAFT7 bleeding occurred in a critical anatomic location (eg, in an intention-to-treat analysis revealed an annual stroke rate intracranial, perispinal). The criteria used by the BAATAF of 4.5% for the control patients and 1.4% for the adjusted- investigators17 were different: intracranial bleeding, fatal dose warfarin patients (relative risk reduction [RRR] ϭ 68%; bleeding, or bleeding leading to transfusion of Ն 4 U of 95% confidence interval [CI], 50 to 79%; number needed to treat for 1 year [NNT] ϭ 32).6 The percentage of strokes that were classified as moderate, severe, or fatal ranged between43% and 64%. Anticoagulation was effective for preventing The primary results of the studies are summarized in strokes of all severities; there was no evidence that the strokes occurring in anticoagulated patients were more severe. In theEAFT,7 which enrolled only patients with a TIA or stroke OAC vs Control: In all randomized studies comparing within the previous 3 months, the RRR was virtually identi- adjusted-dose warfarin anticoagulation with placebo or cal, although the absolute risk of stroke was higher; the Sixth ACCP Consensus Conference on Antithrombotic Therapy Table 4 —Major Bleeding in AF Studies*
*OAC ϭ Oral anticoagulant; ASA: Aspirin; NA ϭ not available.
†Major bleeds include intracranial hemorrhages and major systemic bleeds. Intracranial bleeds include both intraparenchymal hemorrhages and subdural hematomas. BAATAF criteria for serious bleeding were different from those used in the other trials (see text).
‡One fatal hemorrhagic stroke in ASA 125 mg qd group, but nonfatal ICH and major non-CNS bleeds not reported.
annual rate of stroke in control patients was 12% vs 4% in the SPAF-116 showed a statistically significant RRR of anticoagulated patients (RRR ϭ 66%; 95% CI, 43 to 80%; 42%. In the SPAF-1,16 the efficacy of aspirin was apparent in only one of the two component subtrials. When the data There was no significant increase in major bleeding from the AFASAK-1 study,15 the EAFT,7 and the SPAF- events in adjusted-dose anticoagulation-treated patients in 116 were combined in an individual-patient analysis, aspi- these randomized trials (Table 4). In five of the studies rin therapy was associated with a 21% reduction in the risk (the EAFT7 was excluded), anticoagulation lowered the of ischemic stroke (annual stroke rate, 8.1% in control death rate by 33% (95% CI, 9 to 51%) and lowered the patients and 6.3% in aspirin-treated patients; p ϭ 0.05; combined outcome of stroke, systemic embolism, and 95% CI, 0 to 38%).26 One meta-analysis27 combining all four published trials as well as a small unpublished studyfound a virtually identical 22% reduction in the risk of Aspirin vs Placebo or Control: The evidence supporting stroke. A second meta-analysis28 concluded the aspirin the superiority of aspirin to placebo is less robust than the results were heterogeneous, resulting in a substantially evidence for warfarin. In the AFASAK-1 study,15 the broader CI: RRR ϭ 24% (range, Ϫ 33% to ϩ 66%).
EAFT,7 the ESPS-2,11 and the study by Posada andBarriales,23 the relative reduction in the stroke rate was Adjusted-Dose Anticoagulation vs Aspirin: In the generally small and not statistically significant. In contrast, AFASAK-1 study15 and the EAFT,7 adjusted-dose OAC CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT decreased the risk of primary events by 48% and 40%, receiving aspirin, 300 mg, with fixed-dose warfarin, 1.25 respectively, compared with aspirin, 300 mg/d (both re- sults were statistically significant). The results of theSPAF-2 study22 were reported separately for patients Յ 75 Adjusted-Dose OAC vs Low-Dose Anticoagulation: In years of age (mean age, 65 years) and for patients Ͼ 75 the studies13,24 comparing adjusted-dose warfarin with years (mean age, 80 years; Table 3). In the younger group, warfarin, 1.25 mg/d, the risk of stroke was reduced by 13% adjusted-dose warfarin therapy decreased the rate of and 42% in the adjusted-dose anticoagulation groups, stroke by 33%, compared with a 27% reduction in the respectively, both not statistically significant. In another older patients (both differences were not statistically recent study,25 the risk of stroke was slightly lower in significant). However, in SPAF-2,22 many of the strokes patients randomized to a target INR of 1.1 to 1.6, com- occurred in individuals who had discontinued treatment pared with OAC with a target INR of 2.5 to 3.5 with OACs. The AFASAK-2 study13 was stopped about (RRR ϭ 14%), although this difference is likely due to midway through the planned enrollment; therefore, it did chance. Combining the results from all three trials in a not have substantial power to detect a difference between meta-analysis27 yielded an RRR of 38% (95% CI, 20 to the two drugs. In the AFASAK-2 study,13 the annual risk 68%) in favor of adjusted-dose OAC, which was not of primary events was increased slightly in adjusted-dose warfarin-treated patients compared with those receivingaspirin (3.4% vs 2.7%), although the difference was not OAC vs Other Antiplatelet Agents: In the one random- statistically significant. The study by Hellemons et al25 ized trial10 comparing adjusted-dose warfarin with in- reported a 19% RRR of stroke with OAC, which was not dobufen, there was no significant difference in the inci- statistically significant. Finally, the SPAF-3 high-risk dence of primary events (stroke, myocardial infarction study8 found a marked superiority of adjusted-dose war- [MI], pulmonary embolism, or vascular death) between farin (INR 2.0 to 3.0) over low-dose warfarin plus aspirin the two groups (12% in indobufen group vs 10% in (see next paragraph). Over all, these results suggest that warfarin group; p ϭ 0.47). There were four major GI the RRR associated with adjusted-dose warfarin is consid- hemorrhages in the warfarin group compared with none in erably greater than that provided by aspirin. A recent the indobufen group. The frequency of major bleeding meta-analysis27 of these five studies reported a 36% RRR episodes was 0.9% in the warfarin group and 0% in the (95% CI, 14 to 52%) of all stroke with adjusted-dose OAC indobufen group. Indobufen is not currently available in compared with aspirin, and a 46% reduction (95% CI, 27 North America. However, the SIFA study10 results suggest to 60%) in the risk of ischemic stroke. The difference that additional studies of this agent may be warranted.
between the two analyses was largely due to the increased For a discussion of when to begin anticoagulation after rate of intracranial hemorrhage in the SPAF-2 study.22 Of a stroke in AF patients, please refer to the chapter on note, the target INR range (2.0 to 4.5) in the SPAF-2 “Antithrombotic and Thrombolytic Therapy for Ischemic study22 extended above currently recommended intensities.
Adjusted-Dose Anticoagulation vs Low-Dose Anticoag- ulation Plus Aspirin: The SPAF-3 high-risk study8 wasterminated early at the suggestion of the External Safety Intracranial hemorrhage is the most feared complica- Monitoring Committee because of a substantially in- tion of anticoagulant therapy because it is frequently fatal creased rate of primary outcome events in patients receiv- or permanently disabling. Observational studies29,30 from ing combination therapy with fixed-dose, low-intensity large anticoagulation clinics demonstrate that the risk of warfarin (INR 1.2 to 1.5; maximum daily dose, 3 mg) plus intracranial hemorrhage rises dramatically at INR values aspirin, 325 mg/d (7.9%/yr) compared with those receiving Ͼ 4.0 to 5.0. Overall, the initial randomized trials compar- adjusted-dose warfarin with a target INR of 2.0 to 3.0 ing anticoagulation with control or placebo for AF were (1.9%/yr). The absolute difference in stroke rate of 6%/yr reassuring about the rate of intracranial hemorrhage (Ta- translates into a NNT of 17. The high stroke rate in the ble 4). However, a substantially higher rate of intracranial combination therapy arm of this trial8 suggests that the hemorrhage was observed in the SPAF-2 study.22 In low-intensity anticoagulation selected for this study was particular, seven intracranial hemorrhages were observed ineffective in these high-risk AF patients. In addition, no among patients Ͼ 75 years old, for an annualized rate of evidence of a synergistic effect of the low-dose warfarin/ 1.8%, compared with 0.8% in patients receiving aspirin. In aspirin combination could be detected. No significant contrast, taken together, the earlier primary prevention differences in the rates of major hemorrhage were de- trials observed a rate of intracranial hemorrhage of only tected between the two groups (Table 4). The smaller 0.3%/yr among patients Ͼ 75 years old, one sixth of that AFASAK-2 study13 of moderate-risk patients (excluded seen in the SPAF-2 study.31 In the secondary prevention were patients Ͻ 60 years old with lone AF and those with EAFT study,7,32 the average age at entry was 71 years and a history of stroke or TIA in the past 6 months or BPs no intracranial hemorrhages were reported, although a CT Ͼ 180/100 mm Hg) was stopped prematurely following scan was not done in all patients with symptoms of stroke.
the publication of the SPAF-3 data.8 Analysis of their data In the high-risk arm of SPAF-38 (mean age, 71 years; demonstrated no differences, with an annual rate of mean INR, 2.4; target INR, 2.0 to 3.0), the rate of primary events of 3.4% in patients receiving adjusted-dose intracranial hemorrhage was 0.5%/yr compared to a rate of warfarin (INR 2.0 to 3.0) compared with 3.2% in patients 0.9%/yr in the aspirin plus low-dose warfarin arm. The Sixth ACCP Consensus Conference on Antithrombotic Therapy AFASAK-2 study14 recently reported two intracranial risk of intracranial hemorrhage was fairly low at INR hemorrhages in the INR 2.0 to 3.0 arm for an annual rate values Ͻ 4.0 but was sharply higher at greater INR levels.
of 0.6%, compared to 0 to 0.3%/yr rates in the three other Several studies29,33,37–39 have shown that the risk of bleed- ing while receiving oral anticoagulants increased among The reasons for the unusually high intracranial hemor- older patients. The risk of ischemic stroke is low down to rhage rate in the SPAF-2 trial33 in patients Ͼ 75 years old INR values of 2.0. Since randomized trials have success- as compared with the other studies are not entirely clear, fully used INR targets of 2.0 to 3.0, this target range seems although the patients were older than in any other AF an appropriate standard. There is currently no evidence trial, and the target anticoagulation intensity was high about whether this range should be changed for the very (INR 2.0 to 4.5). The importance of high INR levels in elderly (patients Ͼ 75 years old), who have both a higher increasing the risk of intracranial hemorrhage was further risk of stroke and bleeding while receiving oral anticoagu- reinforced by the SPIRIT trial,34 a non-AF secondary lants than younger patients.29,33,37–39 Suggested opinions stroke prevention trial that used an INR target intensity of from the literature for anticoagulation of very elderly 3.0 to 4.5. In the SPIRIT trial,34 the annual rate of patients include aiming for a target INR of 2.5 (range, 2.0 intracranial hemorrhage was Ͼ 3% among patients treated to 3.0) with especially close monitoring35 (which is consis- with anticoagulants. This rate was strongly related to INR tent with our recommendation) or a target INR of 2.0 Optimal Level of Anticoagulation for AF Risk Stratification in Patients With AF Only limited data are available directly comparing Numerous studies have demonstrated that OAC is very different intensities of OAC in patients with AF.8 How- effective in decreasing the risk of stroke in patients with ever, the results of the randomized trials and of observa- AF and that it is considerably more effective than daily tional studies of clinical practice provide fairly consistent aspirin. It is also clear that OAC is associated with a higher evidence about the optimal level of anticoagulation for AF.
frequency of hemorrhage and is more inconvenient than The initial set of randomized trials of OAC vs control aspirin. Each individual AF patient’s risk of stroke and employed a range of target intensities, both prothrombin hemorrhage must be considered when making the deci- time ratio-based and INR-based. The BAATAF study17 sion about the best antithrombotic preventive therapy.
and the SPINAF study19 used the lowest target intensity, The risk of stroke among AF patients not receiving prothrombin time ratio 1.2 to 1.5, corresponding roughly anticoagulants has been studied in subjects participating in to an INR range of 1.5 to 2.7. Anticoagulation appeared several of the randomized trials of antithrombotic thera- just as effective at preventing strokes in these trials as in py.6,42–45 The Atrial Fibrillation Investigators (AFI) group6 the others using a higher target intensity. A target INR of analyzed the data from the pooled control groups of the 1.2 to 1.5 was ineffective in the high-risk SPAF-3 trial,8 first five primary prevention trials and found the following even when combined with aspirin, 325 mg/d. There were independent risk factors for stroke in AF: prior stroke or too few patients in the AFASAK-2 study14 to reliably TIA (relative risk [RR] ϭ 2.5), age (RR ϭ 1.6/decade), determine the efficacy of low-dose warfarin (1.25 mg/d) or history of hypertension (RR ϭ 1.6), and diagnosis of dia- low-dose warfarin combined with aspirin (325 mg/d) com- betes mellitus (RR ϭ 1.7). In addition, patients Ͻ 80 years pared with warfarin (INR 2.0 to 3.0; annual event rates of of age whose only stroke risk factor was coronary artery 3.9%, 3.2%, and 3.4%, respectively). To our knowledge, disease (previous MI or angina) had stroke rates of 4.6%/yr no trials have compared target intensities between an INR if not receiving anticoagulants. In essence, patients Ͼ 65 of 1.5 to 2.0 with an INR between 2.0 and 3.0 in a years old and/or those with any of these risk factors faced randomized fashion. One trial25 compared an INR range a substantial annual risk of stroke. This risk was lowered to of 1.1 to 1.6 with an INR range of 2.5 to 3.5. No difference about 1.5%/yr with adjusted-dose anticoagulant therapy. A in efficacy was detected; however, the low event rates in subsequent AFI analysis43 of echocardiograms done in this study limit the power to detect a difference. The three of the original trials found that moderate-to-severe EAFT32 found a decrease in efficacy below an INR of 2.0, LV dysfunction was an additional strong risk factor but the trial could not assess gradations in INR Ͻ 2.0. A (RR ϭ 2.5). Left atrial diameter was not related to risk of case-control study35 based in a large anticoagulation unit found that INR levels Ͼ 2.0 added little efficacy, while the The AFI analyses6 included data from the untreated risk of stroke increased at INR levels Ͻ 2.0. For example, control group of the SPAF-1 study.16 The SPAF Investi- the odds of stroke doubled at an INR of 1.7 and tripled at gators recently published44 an analysis of risk factors for an INR of 1.5 compared to an INR of 2.0, and increased stroke among the 2,012 patients allocated to the aspirin even more dramatically if the INR was Ͻ 1.5. A second arms of the SPAF-1, SPAF-2, and SPAF-3 randomized hospital-based case-control study36 also found a sharp trials (in SPAF-3, aspirin was combined with very-low- increase in risk of stroke among AF patients with INR intensity anticoagulation) and the SPAF-3 aspirin cohort study. Six features were found to be significant indepen- The optimal level of anticoagulation in AF is that level dent risk factors: prior stroke or TIA (RR ϭ 2.9), age that preserves efficacy in preventing ischemic strokes (RR ϭ 1.8/decade), history of hypertension (RR ϭ 2.0), while minimally increasing the risk of major hemorrhage, systolic BP Ͼ 160 mm Hg (RR ϭ 2.3), female gender especially intracranial hemorrhage. In two studies,29,30 the (RR ϭ 1.6), and alcohol consumption of Ն14 drinks/wk CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT (RR ϭ 0.4, ie, protective). When patients with a prior controlled for, clinical trial data suggest that PAF confers stroke or TIA were excluded from the analysis, female an RR of stroke similar to constant AF.6,49 Patients with gender was no longer significant, but the other features PAF tend to be younger and have a lower incidence of remained significant. Diabetes was a univariate risk factor associated cardiovascular disorders than those with con- (RR ϭ 1.6) that dropped out of the multivariable model.
stant AF; therefore, their absolute stroke rate is lower. The The SPAF analysis provided an additional provocative RRR provided by warfarin appears to be similar for finding. Among women in the SPAF-3 studies,8 hormone patients with both PAF and constant AF. This conclusion, replacement therapy was found to be a powerful indepen- however, is limited by the relatively small number of dent correlate of stroke risk (RR ϭ 3.2). On the basis of patients (about 12% in the first five randomized trials6) these analyses, the SPAF Investigators44 proposed strati- with PAF participating in the trials. Analyses of PAF are fying patients with AF into categories of high, moderate, complicated by the fact that PAF patients differ greatly in and low risk of stroke. Overall, high-risk patients faced a the frequency and length of AF episodes. Studies of PAF Ͼ 7%/yr risk of stroke; moderate-risk patients, 2.5%/yr; are also limited by significant differences in patient aware- and low-risk patients, about 1%/yr. The features qualifying ness of their episodes of AF. The risk-benefit ratio for for these three risk strata are as follows: (1) high risk (any anticoagulation therapy in patients with PAF therefore of the following: prior stroke or TIA, women Ͼ 75 years remains imprecise. In patients with very infrequent and old with a history of hypertension; or systolic BP Ͼ 160 brief episodes of AF, the benefits of warfarin therapy may mm Hg at any age); (2) moderate risk (history of hyper- be offset by inconvenience and bleeding risks. In patients tension and age Յ 75 years, or diabetes); (3) low risk (no with frequent or prolonged paroxysms of AF, particularly high-risk or moderate-risk features). Patients with multi- those with stroke risk factors, warfarin therapy should be ple risk factors appear to be at substantially higher stroke risk than those with a single risk factor.44,45 The risk of stroke in patients with atrial flutter may be It is clear that the AFI and SPAF risk stratification higher than previously assumed, as suggested in a retro- schemes are largely consistent with each other. Prior spective analysis50 of 100 patients with atrial flutter. This stroke or TIA, older age, hypertension, and diabetes are assumption is also supported by the results of a study51 considered by both analyses to be risk factors for stroke in that evaluated the risk of thromboembolism in 191 con- AF. Unlike the AFI analysis, the SPAF scheme empha- secutive unselected patients referred for treatment of sizes the impact of age in women and separates the effect atrial flutter, and documented an embolic event rate of 7% of hypertension into an effect associated with the diagnosis during 26 months of follow-up. These studies differ from itself and an effect due to elevated systolic BP at exami- earlier reports52 that found no risk of stroke or thrombo- nation. There is, as well, a difference in the observed embolism related to atrial flutter. To our knowledge, the absolute risks of stroke. For patients without a history of role of anticoagulation therapy for patients with atrial stroke or TIA, the annual risk of stroke in the AFI data was flutter has not been evaluated in clinical trials; however, 4.0% vs 2.7% in the SPAF data. This difference may be because these patients have a significant risk of developing the result of differences in patient populations, chance, or AF, it may be reasonable to use similar antithrombotic a therapeutic benefit of aspirin among the SPAF patients.
Such small differences can affect the decision to use AF develops in 10 to 15% of patients with thyrotoxicosis anticoagulants in apparently lower-risk patients. The dif- and is most common in patients Ն 60 years of age, ferent impact of age in the AFI and SPAF risk schema presumably reflecting an age-related reduction in the probably affects the greatest percentage of AF patients. In threshold for developing AF.47 The prevalence of thyro- particular, the AFI scheme would view all patients Ն 65 toxicosis in patients with AF is 2 to 5%.47 Some studies53–57 years old as at high risk for stroke, including those without have reported a high frequency of stroke and systemic any other risk factor for stroke. By contrast, the SPAF embolism in patients with thyrotoxic AF, although one scheme would view women with AF Յ 75 years old and study58 did not find a statistically significant difference men of any age, without other risk factors, as at low risk of when AF patients were compared to age- and sex-matched stroke. The resulting uncertainty about the risk faced by patients with normal sinus rhythm. Some of these studies AF patients aged 65 to 75 years and men of any age have methodologic problems, which complicate interpre- without other risk factors applies to roughly 20% of the tation of the results.47 Accordingly, available studies do not confirm that thyrotoxic AF is a more potent risk factor for A recurrent clinical concern is whether patients with stroke than other causes of AF. Since the incidence of paroxysmal, or intermittent, AF (PAF) face the same risk thromboembolic events in patients with thyrotoxic AF of stroke as those with sustained AF. Periods of sinus appears to be similar to other etiologies of AF,47 anti- rhythm should lessen stroke risk, yet transitions from AF thrombotic therapies should probably be chosen based on to sinus rhythm may acutely heighten risk in a manner associated risk factors (see “Recommendations” section).
similar to the increase in risk caused by cardioversion (see Left atrial size can be adequately assessed by transtho- below). Retrospective studies47,48 suggest that PAF is racic echocardiography, but other abnormalities of the left associated with a lower risk of stroke than chronic AF.
atrium can be seen via transesophageal echocardiography These epidemiologic data suggest that PAF has an inter- (TEE). While this modestly invasive approach is com- mediate risk of stroke between constant AF and sinus monly used as an adjunct to elective cardioversion, it has rhythm. However, when associated stroke risk factors are also been applied to studies of outpatients with chronic Sixth ACCP Consensus Conference on Antithrombotic Therapy AF.59,60 Spontaneous echo contrast (a marker of stasis) and frank thrombi in the left atrium appear to confer a twofold to fourfold increase in risk of subsequent stroke. The vastmajority (Ͼ 90%) of these thrombi involve or are confined Synchronized capacitor discharge was introduced by to the left atrial appendage. Patients with TEE-detected Lown and coworkers65 for the rapid termination of atrial aortic plaques with complex features (mobile, peduncu- and ventricular tachyarrhythmias. Systemic embolism is lated, ulcerated, or Ն 4 mm in diameter) had extremely the most serious complication of cardioversion and may high stroke rates in the SPAF-3 study.
follow direct current (DC), pharmacologic, and spontane- no clear evidence that TEE findings add independently torisk stratification when clinical and transthoracic echocar-diographic risk factors are considered.
Finally, studies have shown that AF patients with prosthetic heart valves (both mechanical and tissue valves) Bjerkelund and Orning66 performed a prospective co- or rheumatic mitral valve disease are at high risk of stroke hort study in which cardioversion without anticoagulants (see the chapters on valvular heart disease and prosthetic resulted in a 5.3% incidence of clinical thromboembolism, valves) and should be treated with adjusted-dose warfarin.
whereas a 0.8% incidence of thromboembolism was noted The purpose of risk classification schemes is to identify in patients receiving OACs. Although this was not a subgroups of patients with different risks of stroke: those randomized study, the results are compelling because the in whom the risk of stroke is so high that warfarin is clearly patients receiving anticoagulants were also at higher risk indicated unless their risk of bleeding is very high, and than those who were not. Several authors of case se- those in whom the risk of stroke is sufficiently low that ries52,67–70 also favor the use of adjusted-dose anticoagula- warfarin need not be used. Although there are groups of tion before cardioversion. Although sometimes occurring patients who clearly fall into these categories, there are up to Ն 10 days after cardioversion, the majority of these also patients for whom the choice of warfarin vs aspirin is adverse events occur during the first 72 h after cardiover- more difficult. Patients with AF who have at least one of sion and are presumed to be the result of thrombi present the following risk factors are at high risk of stroke and within the left atrium at the time of cardioversion.71 New should be offered OAC unless their risk of bleeding is thrombus may develop after DC cardioversion and high- high: previous stroke or TIA or systemic embolism, age lights the importance of periconversion anticoagulation Ͼ 75 years old, history of hypertension, prosthetic heart (see below). The duration of anticoagulation before car- valve (mechanical or tissue valve), or rheumatic mitral dioversion is not clearly defined, as the majority of these valvular disease. Patients with poor LV systolic function studies were retrospective analyses, but specific recom- also appear to be at high risk. The risk factor status is less mendations of 3 to 4 weeks of prophylactic adjusted-dose secure in those age 65 to 75 years, in those with diabetes warfarin therapy before and after have been made by mellitus, and in those with coronary artery disease in the many investigators.72,73 In the recommendations that fol- absence of LV dysfunction. However, we recommend low, clinical observations and the data from several of anticoagulation if more than one of these “less severe” risk factors are present. Patients without cardiovascular disease The vast majority of data on cardioversion-related or risk factors who are Ͻ 65 years old are at such low risk thromboembolism are based on electrical cardioversion.
of stroke that they should be treated with aspirin alone.
There are limited clinical data that have examined the For patients who do not meet the high-risk or low-risk issue of embolization after pharmacologic or spontaneous criteria, the absolute benefit of warfarin therapy is likely to cardioversion of AF to sinus rhythm. Goldman74 reported be small. Treatment decisions should be individualized that embolism occurred in 1.5% of 400 patients treated and consideration given to patient preferences and risk with quinidine for reversion of AF to sinus rhythm. This was similar to the 1.2% incidence of embolization that Anticoagulation is a potentially risky therapy that im- Lown68 reported in 450 electrical cardioversions in pa- poses a variety of lifestyle constraints on patients. As a tients not receiving anticoagulants. Therefore, it seems result, patient education and involvement in the anticoag- prudent to administer anticoagulants to individuals under- ulation decision is important. Many AF patients have a going pharmacologic cardioversion in a similar manner to great fear of suffering a stroke and wish to take warfarin those undergoing electrical cardioversion.
for a relatively small decrease in the risk of stroke,61 while The mechanism of benefit conveyed by the month of others who are at relatively low risk for stroke will want to warfarin treatment prior to elective cardioversion had avoid the burdens and risks of anticoagulation and opt for previously been ascribed to the promotion of thrombus aspirin.62–64 The safe use of anticoagulants depends on organization and adherence to the atrial wall.74 More patient cooperation and a monitoring system that can recently, serial TEE studies75–77 of those presenting with achieve INR targets on a regular basis. The AFASAK-2 new-onset AF and atrial thrombi on initial TEE have study13 demonstrates that anticoagulation at an INR of 2.0 demonstrated resolution of the atrial thrombi after 1 to 3.0 can be quite safe even for elderly patients, and the month of warfarin treatment in the majority of subjects. It study by Palareti et al38 demonstrates that low hemorrhage thus appears that the month of warfarin treatment may rates can be duplicated in clinical practice outside of trials, also facilitate “silent” thrombus resolution.
particularly if anticoagulation clinics are involved.
The immediate postcardioversion period is associated CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT with increased risk for thrombus formation. Utilizing Stroke has been described among patients who did not TEE, further depression of atrial appendage velocities, receive anticoagulation at the time of TEE or continued more intense left atrial spontaneous echocardiographic anticoagulation for a full month after cardioversion despite contrast, and even new thrombus formation have been the absence of left atrial appendage thrombi on TEE.89–93 described after external DC, internal DC, and even These adverse events may have occurred because the spontaneous cardioversion.78–81 These data underscore sensitivity of TEE for small atrial appendage thrombus is the importance of therapeutic anticoagulation during the not 100%, development of new thrombus because of pericardioversion period. Following restoration of normal transient atrial dysfunction during the postcardioversion atrial electrical activity on the surface ECG, the mechan- period, or other mechanisms. Because of uncertainty ical contraction of the body of the left atrium may remain regarding the role of TEE in guiding anticoagulant ther- dysfunctional for as long as 2 to 4 weeks after cardiover- apy at the time of electrical cardioversion, a large (Ͼ 1,000 sion.82–84 For this reason, adjusted-dose anticoagulation patients) randomized multicenter international study, As- should be continued for 1 month after cardioversion. In sessment of Cardioversion Using Transesophageal Echo- addition to prophylaxis against new thrombus formation cardiography (ACUTE), comparing conventional vs the during recovery of atrial mechanical activity, warfarin alsoserves as prophylaxis against thrombus formation should novel TEE approach is currently underway. The results of the ACUTE pilot study75 comparing TEE-guided cardio- Therefore, for patients with AF, the following are version with standard management of cardioversion in AF recommended: (1) therapeutic warfarin (target INR 2.5; patients have been reported. Sixty-two of 126 patients who range, 2.0 to 3.0) anticoagulation should be given for 3 had AF lasting Ͼ 48 h were randomly selected to receive weeks before elective cardioversion; (2) anticoagulation TEE-guided cardioversion. TEE was performed in 56 should be continued for 4 weeks after successful cardio- patients, and atrial thrombi were found in 7 patients.
version because it will decrease the likelihood that a fresh Cardioversion was successful in 38 of 45 patients who had thrombus will form in the noncontractile left atrial ap- early cardioversion. There were no embolic events in the pendage if the resumption of mechanical contraction is patients who were free of left atrial thrombus. There was delayed, and it will decrease the formation of thrombus if one embolic event (1.6%) occurring 3 days after cardiover- AF recurs soon after successful cardioversion. For patients sion in a patient randomized to the conventional manage- presenting with their first episode of AF, long-term anti- ment group. Though cardioversion occurred earlier in the coagulation beyond the first 4 weeks after cardioversion TEE-guided group, there was no difference in the likeli- may be indicated if the patient has high clinical risk factors hood of sinus rhythm at 8 weeks after cardioversion.
for stroke or is at high risk for recurrent AF (enlarged left For AF of short duration (Ͻ 48 h), the usual clinical atrium, significant LV dysfunction). If AF recurs, long- practice is to perform cardioversion without TEE or term (after 1 month) anticoagulation decisions should be prolonged precardioversion anticoagulation. This practice based on the previously described clinical and echocardio- was called into question when a study94 reported a 13% graphic criteria for chronic or paroxysmal AF.
prevalence of atrial thrombi on TEE among patients with Over the past decade, an alternative strategy has been AF of Ͻ 72 h duration. Subsequently, however, data were suggested for cardioversion of patients with AF of Ͼ 2 reported from a study95 of 357 patients who had a days or of unknown duration. Among patients with AF, thevast majority (Ͼ 90%) of thrombi are located within, or symptomatic duration of AF for Ͻ 48 h. Two hundred fifty involve, the left atrial appendage.75,76,81,84,85 While the patients converted spontaneously, and 107 underwent detection of left atrial appendage thrombi is unreliable pharmacologic or electrical cardioversion, all without utilizing conventional transthoracic echocardiography, bi- screening TEE or a month of warfarin treatment prior to plane and multiplane TEE have demonstrated very high cardioversion. Clinical thromboembolism occurred in accuracy86,87 and therefore offer the opportunity to per- three subjects (Ͻ 1%), all of whom were elderly woman form early cardioversion for those in whom no atrial without a history of prior AF and with normal LV systolic appendage thrombi are observed. Systemic anticoagula- function. Preliminary data from the Canadian AF Regis- tion with IV heparin and/or warfarin should still be try94,96 also suggest a very low incidence of adverse events employed at the time of TEE and cardioversion because of if these patients undergo early cardioversion. Although the concern that new thrombus may form during the safe in these studies, it may be prudent to perform TEE or pericardioversion or postcardioversion period. Data from delay cardioversion for 1 month for very high-risk patients several studies75,76,81,84,85 currently suggest rates of throm- (eg, patients with a history of prior stroke/thromboembo- boembolism that are similar to those associated with standard therapy, with the advantages of an earlier recov- While patients with short-duration AF (Ͻ 48 h) may not ery of atrial mechanical function, ease of anticoagulation require TEE or a month of prolonged warfarin treatment management, elimination of the need for hospital read- prior to cardioversion, it may be prudent to initiate mission for elective cardioversion, and of cost-effective- heparin anticoagulation at presentation. Many of these ness if performed expeditiously and without a somewhat patients will require anticoagulation after cardioversion, redundant transthoracic echocardiographic examination.88 and the use of heparin will further decrease the likelihood Limitations of the TEE approach include patient discom- of new thrombus formation during the pericardioversion fort, rare procedural complications, and limited availability Sixth ACCP Consensus Conference on Antithrombotic Therapy Anticoagulation for Emergency Cardioversion of patients with one moderate-risk factor, we recom- mend aspirin, 325 mg/d, or warfarin (target INR 2.5; Emergency cardioversion is performed to terminate range, 2.0 to 3.0). For patients with no high-risk atrial tachyarrhythmias with a rapid ventricular response factors and no moderate-risk factors, we recommend causing angina, heart failure, hypotension, or syncope. In individuals with impaired ventricular function, clinicaldeterioration may occur within minutes or hours of the onset of the arrhythmia, and urgent electrical or pharma-cologic cardioversion is indicated. The role of anticoagu- High-risk factors include prior stroke/TIA or sys- lation in these circumstances remains controversial, but temic embolus, history of hypertension, poor LV heparin therapy at the time of cardioversion may be useful systolic function, age Ͼ 75 years, rheumatic mitral to prevent thrombi from forming due to further atrial valve disease, and prosthetic heart valve. Moderate- appendage dysfunction after cardioversion.
risk factors (factors for stroke that have been identi-fied in AF patients in various studies but are not asstrong or consistent as the high-risk factors listedabove) include age 65 to 75 years, diabetes mellitus, and coronary artery disease with preserved LV sys- In several published series97–100 of patients who underwent cardioversion, all three arrhythmias (AF, atrial flutter, and supraventricular tachycardia) were 1.1. We recommend the use of adjusted-dose pooled together when the data were analyzed. There- warfarin anticoagulation (target INR 2.5; range 2.0 to fore, it is difficult to estimate the risk of embolism 3.0) rather than aspirin in patients with AF at high during cardioversion for atrial flutter. However, there risk for ischemic stroke because it markedly de- have been several reports50,51,97–99,101 of embolization creases the risk of ischemic stroke in patients with AF after cardioversion of patients with pure atrial flutter.
Patients at particularly high risk include those with 1.2. For high-risk patients, we recommend that valvular heart disease, prior thromboembolism, conges- clinicians offer aspirin therapy if adjusted-dose war- tive heart failure, and LV systolic dysfunction. Whether farin is contraindicated or declined by the patient and these patients had unrecognized episodes of AF or if there are no contraindications to aspirin (grade 1A).
spontaneous echo contrast is unknown.99 Similar to AF, 1.3. We recommend that clinicians do not use delayed restoration of atrial function after cardioversion aspirin plus low-fixed-dose warfarin therapy (grade from atrial flutter has been described.102 These findings raise concern that patients with atrial flutter are at 1.4. Although to our knowledge no randomized increased risk of embolization at the time of cardiover- trials of OAC have been undertaken in AF patients sion. Consideration should be given to treating patients with rheumatic mitral valve disease or prosthetic with atrial flutter in the same manner as patients with heart valves (mechanical or tissue valves), we recom- AF at the time of cardioversion, especially those with a mend that clinicians use OAC in these patients history of prior AF or thromboembolism, or LV systolic dysfunction.101,103 Although some retrospective stud-ies50,101 have suggested an increased risk of stroke andthromboembolism in patients with sustained or inter- mittent atrial flutter, more information is required 1.6. We recommend that patients with AF who are before a firm recommendation can be made about Ͻ 65 years with no clinical or echocardiographic long-term OAC therapy in these patients.
evidence of cardiovascular disease should be treatedwith aspirin (grade 2C).
1.7. Some AF patients will have a risk of stroke that is between that of the high-risk and low-risk groups mentioned. For these patients, the absolute strokeRR of warfarin vs aspirin is likely to be small. We For patients with any high-risk factor or more than recommend the use of either OAC or aspirin for one moderate-risk factor, we recommend warfarin patients with one of these moderate risk factors (target INR 2.5; range, 2.0 to 3.0). See chapter (grade 1A in comparison to no treatment).
“Antithrombotic Therapy in Patients With Mechani- 1.8. Patients with more than one of these moder- cal and Biological Prosthetic Heart Valves” for target ate-risk factors are at higher risk of stroke than are INRs in patients with mechanical heart valves. For those with only one risk factor, and we recommend to CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 5 Furberg CD, Psaty BM, Manolio TA, et al. Prevalence of atrial treat these patients in the same manner as high-risk fibrillation in elderly subjects (The Cardiovascular Health Study).
The ultimate choice of therapy depends on many 6 Atrial Fibrillation Investigators. Risk factors for stroke and effi- factors, including the clinician’s assessment of the cacy of anti-thrombotic therapy in atrial fibrillation: analysis of magnitude of the patient’s risk (eg, whether the patient pooled data from five randomized controlled trials. Arch Intern has single or multiple risk factors), the ability to provide high-quality monitoring of the intensity of OAC, the 7 Secondary prevention in non-rheumatic atrial fibrillation after patient’s risk of bleeding with OAC, and patient transient ischaemic attack or minor stroke: EAFT (EuropeanAtrial Fibrillation Trial) Study Group. Lancet 1993; 342:1255– 8 Adjusted-dose warfarin versus low-intensity, fixed-dose warfarin plus aspirin for high-risk patients with atrial fibrillation: Stroke Prevention in Atrial Fibrillation III randomised clinical trial.
Lancet 1996; 348:633– 638 9 SPAF III Writing Committee for the Stroke Prevention in Atrial Fibrillation Investigators. Patients with nonvalvular atrial fibrilla- 2.1.1. We recommend that clinicians administer tion at low risk of stroke during treatment with aspirin: Stroke oral anticoagulant therapy (target INR 2.5; range 2.0 Prevention in Atrial Fibrillation III study. JAMA 1998; 279:1273– to 3.0) for 3 weeks before and at least 4 weeks after 10 Morocutti C, Amabile G, Fattapposta F, et al. Indobufen versus elective DC cardioversion of AF patients (grade warfarin in the secondary prevention of major vascular events in nonrheumatic atrial fibrillation: SIFA (Studio Italiano Fibrillazi- 2.1.2. Alternatively, we recommend that AF pa- one Atriale) Investigators. Stroke 1997; 28:1015–1021 tients undergo anticoagulation then undergo TEE, 11 Diener H, Cunha L, Forbes C, et al. European Stroke Prevention and have cardioversion performed without delay if no Study 2: dipyridamole and acetylsalicylic acid in the prevention of thrombi are seen (grade 1C). For these patients, 12 Diener HC, Lowenthal A. Letter to the editor. J Neurol Sci 1997; adjusted-dose warfarin therapy should still be contin- ued until normal sinus rhythm has been maintained 13 Gullov AL, Koefoed BG, Petersen P, et al. Fixed mini-dose warfarin and aspirin alone and in combination versus adjusted- 2.1.3. Although data are limited, the risk of embo- dose warfarin for stroke prevention in atrial fibrillation: Second lism following cardioversion in patients who have Copenhagen Atrial Fibrillation, Aspirin, and Anticoagulation been in AF for Ͻ 48 h appears to be low. However, Study (the AFASAK-2 study). Arch Intern Med 1998; 158:1513–1521 we recommend the use of anticoagulation during the 14 Gullov AL, Koefoed BG, Petersen P. Bleeding during warfarin pericardioversion period (grade 2C).
and aspirin therapy in patients with atrial fibrillation: theAFASAK-2 Study. Arch Intern Med 1999; 159:1322–1328 2.2. Atrial Flutter and Supraventricular 15 Petersen P, Boysen G, Godtfredsen J, et al. Placebo-controlled, randomised trial of warfarin and aspirin for prevention of throm-boembolic complications in chronic atrial fibrillation: the Copen- 2.2.1. We recommend that clinicians manage OAC hagen AFASAK study. Lancet 1989; 1:175–178 at the time of cardioversion in patients with atrial 16 Stroke Prevention in Atrial Fibrillation Investigators. Stroke flutter in a manner similar to that used for AF (grade Prevention in Atrial Fibrillation Study: final results. Circulation1991; 84:527–539 17 Boston Area Anticoagulation Trial for Atrial Fibrillation Investi- 2.2.2. In the absence of prior thromboembolism, gators. The effect of low-dose warfarin on the risk of stroke in we do not recommend antithrombotic therapy for patients with nonrheumatic atrial fibrillation. N Engl J Med 1990; cardioversion of supraventricular tachycardia (grade 18 Connolly SJ, Laupacis A, Gent M, et al. Canadian Atrial Fibril- Treatment of potential precipitants of AF (ie, lation Anticoagulation (CAFA) Study. J Am Coll Cardiol 1991; thyrotoxicosis, pneumonia, congestive heart failure) 19 Ezekowitz MD, Bridgers SL, James KE, et al. Warfarin in the should be completed prior to attempting elective DC prevention of stroke associated with nonrheumatic atrial fibrilla- tion: Veterans Affairs Stroke Prevention in Nonrheumatic AtrialFibrillation Investigators. N Engl J Med 1992; 327:1406 –1412 20 Ezekowitz MD, James KE, Nazarian SM, et al. Silent cerebral infarction in patients with nonrheumatic atrial fibrillation. Circu- 1 Feinberg WM, Blackshear JL, Laupacis A, et al. The prevalence of atrial fibrillation: analysis and implications. Arch Intern Med 21 Albers GW, Easton JD, Sacco RL, et al. Antithrombotic and thrombolytic therapy for ischemic stroke. Chest 1998; 114(5 2 Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: the Framingham study. Stroke 22 Stroke Prevention in Atrial Fibrillation Investigators. Warfarin versus aspirin for prevention of thromboembolism in atrial fibril- 3 Lake FR, McCall MG, Cullen KJ, et al. Atrial fibrillation and lation: Stroke Prevention in Atrial Fibrillation II Study. Lancet mortality in an elderly population. Aust N Z J Med 1989; 23 Posada IS, Barriales V. Alternate-day dosing of aspirin in atrial 4 Phillips SJ, Whisnant JP, O’Fallon WM, et al. Prevalence of fibrillation: LASAF Pilot Study Group. Am Heart J 1999; 138: cardiovascular disease and diabetes in residents of Rochester, Minnesota. Mayo Clin Proc 1990; 65:344 –359 24 Pengo V, Zasso A, Barbero F, et al. Effectiveness of fixed Sixth ACCP Consensus Conference on Antithrombotic Therapy minidose warfarin in the prevention of thromboembolism and 46 Go AS, Hylek EM, Henault LE, et al. Implications of different vascular death in nonrheumatic atrial fibrillation. Am J Cardiol stroke risk criteria on anticoagulation decision in atrial fibrillation [abstract]. J Gen Intern Med 1999; 14(suppl 2):99 25 Hellemons BSP, Lanbenberg M, Lodder J, et al. Primary preven- 47 Petersen P. Thromboembolic complications in atrial fibrillation.
tion of arterial thromboembolism in non-rheumatic atrial fibrilla- tion in primary care: randomised controlled trial comparing two 48 Brand FN, Abbott RD, Kanle WB, et al. Characteristics and intensities of coumarin with aspirin. BMJ 1999; 319:958 –964 prognosis of lone atrial fibrillation: 30 year follow-up in the 26 Atrial Fibrillation Investigators. The efficacy of aspirin in patients Framingham study. J Am Coll Cardiol 1985; 254:3449 –3453 with atrial fibrillation: analysis of pooled data from three randomized 49 Hart RG, Pearce LA, Rothbart RM, et al. Stroke with intermittent trials. Arch Intern Med 1997; 157:1237–1240 atrial fibrillation: incidence and predictors during aspirin therapy.
27 Hart RG, Benavente O, McBride R, et al. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta- 50 Lanzarotti CJ, Olshansky B. Thromboembolism in chronic atrial analysis. Ann Intern Med 1999; 131:492–501 flutter: is the risk underestimated? J Am Coll Cardiol 1997; 28 Segal JB, McNamara RL, Miller MR, et al. Prevention of thromboembolism in atrial fibrillation: a meta-analysis of trials of 51 Seidl K, Hauer B, Schwick NG, et al. Risk of thromboembolic anticoagulants and antiplatelet drugs. J Gen Intern Med 2000; events in patients with atrial flutter. Am J Cardiol 1998; 82:580 – 29 Hylek EM, Singer DE. Risk factors for intracranial hemorrhage in 52 Zieler A, Do A, Mich MJ, et al. Role of prophylactic anticoagu- outpatients taking warfarin. Ann Intern Med 1994; 120:897–902 lation for direct current cardioversion in patients with atrial 30 Cannegieter SC, Rosendaal FR, Wintzen AR, et al. Optimal oral fibrillation or atrial flutter. J Am Coll Cardiol 1992; 19:851– 855 anticoagulant therapy in patients with mechanical heart valves.
53 Presti CF, Hart RG. Thyrotoxicosis, atrial fibrillation, and embo- lism, revisited. Am Heart J 1989; 117:976 –977 31 Connolly S. Stroke Prevention in Atrial Fibrillation II Study 54 Staffurth JS, Gibberd MC, Ng Tang Fui S. Arterial embolism in thyrotoxicosis with atrial fibrillation. BMJ 1977; 2:688 – 690 32 European Atrial Fibrillation Trial Study Group. Optimal oral 55 Yuen RWM, Gutteridge DH, Thompson PL, et al. Embolism in anticoagulation therapy in patients with nonrheumatic atrial thyrotoxic atrial fibrillation. Med J Aust 1979; 1:630 – 631 fibrillation and recent cerebral ischemia. N Engl J Med 1995; 56 Hurley DM, Hunter AN, Hewett MJ, et al. Atrial fibrillation and arterial embolism in hyperthyroidism. Aust NZ J Med 1981; 33 Stroke Prevention in Atrial Fibrillation Investigators. Bleeding during antithrombotic therapy in patients with atrial fibrillation.
57 Bar-Sela S, Ehrenfeld M, Eliakim M. Arterial embolism in thyrotoxicosis with atrial fibrillation. Arch Intern Med 1981; 34 The Stroke Prevention in Reversible Ischemia Trial (SPIRIT) Study Group. A randomized trial of anticoagulants versus aspirin 58 Petersen P, Hansen JM. Stroke in thyrotoxicosis with atrial after cerebral ischemia of presumed arterial origin. Ann Neurol 59 Stollberger C, Chnupa P, Kronik G, et al. Transesophageal 35 Hylek EM, Skates SJ, Sheehan MA, et al. An analysis of the lowest echocardiography to assess embolic risk in patients with atrial effective intensity of prophylactic anticoagulation for patients with fibrillation. Ann Intern Med 1998; 128:630 – 638 nonrheumatic atrial fibrillation. N Engl J Med 1996; 335:540 –546 60 Stroke Prevention in Atrial Fibrillation Investigators Committee 36 Brass LM, Krumholz HM, Scinto JM, et al. Warfarin use among on Echocardiography. Transesophageal echocardiographic corre- patients with atrial fibrillation. Stroke 1997; 28:2382–2389 lates of thromboembolism in high-risk patients with nonvalvular 37 van der Meer FJM, Rosendaal FR, Vandenbroucke JP, et al.
atrial fibrillation. Ann Intern Med 1998; 128:639 – 647 Bleeding complications in oral anticoagulant therapy: an analysis 61 Man-Son-Hing M, Laupacis A, O’Connor A, et al. Warfarin for of risk factors. Arch Intern Med 1993; 153:1557–1562 atrial fibrillation: the patient’s perspective. Arch Intern Med 38 Palareti G, Leali N, Coccheri S, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collabora- 62 Man-Son-Hing M, Laupacis A, O’Connor A, et al. A patient tive study (ISCOAT). Lancet 1996; 348:423–428 decision aid regarding antithrombotic therapy for stroke preven- 39 Fihn SD, Callahan CM, Martin DC, et al, for the National tion in atrial fibrillation: a randomized controlled trial. JAMA Consortium of Anticoagulation Clinics. The risk for and severity of bleeding complications in elderly patients treated with warfa- 63 Man-Song-Hing M, Laupacis A, O’Connor A, et al. Warfarin for atrial fibrillation: the patient’s perspective. Arch Intern Med 40 Hart RG. Intensity of anticoagulation to prevent stroke in patients with atrial fibrillation [letter]. Ann Intern Med 1998; 128:408 64 Gauge BF, Cardinalli AB, Owens DK. The effect of stroke and 41 Sheffield JVL, Larson EB. Intensity of anticoagulation to prevent stroke prophylaxis with aspirin or warfarin on quality of life. Arch stroke in patients with atrial fibrillation [letter]. Ann Intern Med 65 Lown B, Amarasingham R, Neuman J. New method for termi- 42 Stroke Prevention in Atrial Fibrillation Investigators. The Stroke nating cardiac arrhythmias. JAMA 1962; 182:548 –555 Prevention in Atrial Fibrillation III Study: rationale, design, and 66 Bjerkelund C, Orning O. The efficacy of anticoagulant therapy in patient features. J Stroke Cerebrovasc Disord 1997; 6:1–13 preventing embolism related to DC electrical conversion of atrial 43 Atrial Fibrillation Investigators. Echocardiographic predictors of fibrillation. Am J Cardiol 1969; 23:208 –216 stroke in patients with atrial fibrillation: a prospective study of 67 Morris JM, Peter RH, McIntosh HD. Electrical conversion of 1,066 patients from three clinical trials. Arch Intern Med 1998; atrial fibrillation: immediate and long-term results and selection of patients. Ann Intern Med 1966; 65:216 –231 44 Hart RG, Pearce LA, McBride R, et al. Factors associated with 68 Lown B. Electrical reversion of cardiac arrhythmias. Br Heart J ischemic stroke during aspirin therapy in atrial fibrillation: anal- ysis of 2012 participants in the SPAF I-III clinical trials: Stroke 69 Resnekov L, McDonald L. Complication in 220 patients with Prevention in Atrial Fibrillation (SPAF) Investigators. Stroke cardiac dysrhythmias treated by phased DC shock and indications for electroversion. Br Heart J 1967; 29:926 –936 45 Stroke Prevention in Atrial Fibrillation Investigators. Predictors 70 McCarthy C, Varghese PJ, Baritt DW. Prognosis of atrial arrhyth- of thromboembolism in atrial fibrillation: I. Clinical features of mias treated by electrical countershock therapy. Br Heart J 1969; patients at risk. Ann Intern Med 1992; 116:1–5 CHEST / 119 / 1 / JANUARY, 2001 SUPPLEMENT 71 Berger M, Schweitzer P. Timing of thromboembolic events after esophageal echocardiography in detecting left atrial thrombus.
electrical cardioversion of atrial fibrillation or flutter: a retrospec- tive analysis. Am J Cardiol 1998; 82:1545–1547 88 Seto TB, Taira DA, Tsevat J, et al. Cost-effectiveness of trans- 72 DeSilva RA, Graboys TB, Podrid PJ, et al. Cardioversion and esophageal echocardiography-guided cardioversion for hospital- defibrillation. Am Heart J 1980; 100:881– 895 ized patients with atrial fibrillation. J Am Coll Cardiol 1997; 73 Mancini GBJ, Goldberger AI. Cardioversion of atrial fibrillation: consideration of embolization, anticoagulation, prophylactic pace- 89 Alka S, Saeian K, Sagar KB. Cerebral thromboembolization after maker and long-term success. Am Heart J 1982; 104:617– 621 cardioversion of atrial fibrillation in patients without transesoph- 74 Goldman MJ. The management of chronic atrial fibrillation: ageal echocardiographic findings of left atrial thrombus. AmHeart J 1993; 126:722–724 indications and method of conversion to sinus rhythm. Prog 90 Lack IW, Fatkin D, Sagar KB, et al. Exclusion of atrial thrombus by transesophageal echocardiography does not preclude embo- 75 Klein AL, Grimm RA, Block LW, et al. Cardioversion guided by lism after cardioversion of atrial fibrillation: a multicenter study.
transesophageal echocardiography: the ACUTE pilot study. Ann 91 Lack IW, Hopkins AP, Lee LCL, et al. Evaluation of transesoph- 76 Corrado G, Tadeo G, Beretta S, et al. Atrial thrombi resolution ageal echocardiography before cardioversion of atrial fibrillation after prolonged anticoagulation in patients with atrial fibrillation: and flutter in nonanticoagulated patients. Am Heart J 1993; a transesophageal echocardiographic study. Chest 1999; 115:140 – 92 Moreyra E, Finkelhor RS, Cebul RD. Limitations of transesoph- 77 Collins LJ, Silverman DI, Douglas PS, et al. Cardioversion of ageal echocardiography in the risk assessment of patients before non-rheumatic atrial fibrillation: reduced thromboembolic com- nonanticoagulated cardioversion from atrial fibrillation and flut- plications with 4 weeks of pre-cardioversion anticoagulation are ter: an analysis of pooled trials. Am Heart J 1995; 129:71–75 related to atrial thrombus resolution. Circulation 1995; 92:160 – 93 Black IW, Fatkin D, Sagar KB, et al. Exclusion of atrial thrombus by transesophageal echocardiography does not preclude embo- 78 Grimm RA, Stewart WJ, Maloney JD, et al. Impact of electrical lism after cardioversion of atrial fibrillation: a multicenter study.
cardioversion of atrial fibrillation on left atrial appendage function and spontaneous echo contrast: characterization by simultaneous 94 Mitchell MA, Hughes GS, Ellenbogen KE, et al. Cardioversion- transesophageal echocardiography. J Am Coll Cardiol 1993; 22: related stroke rates in atrial fibrillation and atrial flutter. Circu- 79 Omran H, Jung W, Rabahieh R, et al. Left atrial chamber and 95 Weigner MJ, Caulfield TA, Danias PG, et al. Risk for clinical appendage function after internal atrial defibrillation: a prospec- thromboembolism associated with conversion to sinus rhythm in tive and serial transesophageal echocardiographic study. J Am patients with atrial fibrillation lasting less than 48 hours. Ann 80 Grimm RA, Leung DY, Black IW, et al. Left atrial appendage 96 Stoddard MF, Dawkins PR, Prince CR, et al. Left atrial append- “stunning” after spontaneous conversion of atrial fibrillation age thrombus is not uncommon in patients with acute atrial demonstrated by transesophageal Doppler echocardiography. Am fibrillation and a recent embolic event: a transesophageal echo- cardiographic study. J Am Coll Cardiol 1995; 25:452– 459 81 Stoddard MF, Dawkins P, Prince CR, et al. Transesophageal 97 Black IW, Hopkins AP, Lee LCL, et al. Thromboembolic risk of echocardiographic guidance of cardioversion in patients with atrial flutter [abstract]. J Am Coll Cardiol 1992; 19:314A atrial fibrillation. Am Heart J 1995; 129:1204 –1215 98 Santiago D, Warshofsky M, Mandri G, et al. Left atrial appendage 82 Manning WJ, Leeman DE, Gotch PJ, et al. Pulsed Doppler function and thrombus formation in atrial fibrillation-flutter: a evaluation of atrial mechanical function after electrical cardiover- transesophageal echocardiography study. J Am Coll Cardiol 1994; sion of atrial fibrillation. J Am Coll Cardiol 1989; 13:617– 623 83 Padraig GO, Puleo PR, Bolli R, et al. Return of atrial mechanical 99 Mehta D, Baruch L. Thromboembolism following cardioversion of function following electrical cardioversion of atrial dysrhythmias.
“common” atrial flutter: risk factors and limitations of transesopha- geal echocardiography. Chest 1996; 110:1001–1003 84 Manning WJ, Silverman DI, Gordon SPF, et al. Cardioversion 100 Selzer A, Kelly JJ Jr, Johnson RB, et al. Immediate and from atrial fibrillation without prolonged anticoagulation with long-term results of electrical reversion of arrhythmias. Prog use of transesophageal echocardiography to exclude the pres- ence of atrial thrombi. N Engl J Med 1993; 328:750 –755 101 Wood KA, Eisenberg SJ, Kalman JM, et al. Risk of thromboem- 85 Manning WJ, Silverman DI, Keighly CS, et al. Transesophageal bolism in chronic atrial flutter. Am J Cardiol 1997; 79:1043–1047 echocardiography facilitated early cardioversion from atrial fibril- 102 Jordaens L, Missault L, Germonpre E, et al. Delayed restoration lation using short-term anticoagulation: final results of a prospec- of atrial function after cardioversion of atrial flutter by pacing or tive 4.5 year study. J Am Coll Cardiol 1995; 25:1354 –1361 electrical cardioversion. Am J Cardiol 1993; 71:63– 67 86 Manning WJ, Weintraub RM, Waksmonski CA, et al. Accuracy of 103 Irani WN, Grayburn PA, Afridi I. Prevalence of thrombus, transesophageal echocardiography for identifying left atrial spontaneous echo contrast, and atrial stunning in patients un- thrombi: a prospective, intraoperative study. Ann Intern Med dergoing cardioversion of atrial flutter: a prospective study using transesophageal echocardiography. Circulation 1997; 95:962– 87 Fatkin D, Scalia G, Jacobs N, et al. Accuracy of biplane trans- Sixth ACCP Consensus Conference on Antithrombotic Therapy

Source: http://www.miba.auc.dk/~04gr690/vedhaeftede_filer/Antithrombotic%20therapy%20in%20atrial%20fibrillation.pdf