Prevention of Thromboembolism in Patients with New Onset or Recently Discovered Atrial Fibrillation

Abstract

New or newly discovered atrial ~brillation (AF) is encountered in three common settings: symptomatic presentation to of~ce or hospital, incidental discovery either by electrocardiography or detection of pulse irregularity which leads to electrocardiographic documentation, and during hospitalization, often following cardiac or non-cardiac surgery, or during severe medical illness. The timing of onset of AF may be suspected from the onset of symptoms in the ~rst case, but may be unreliable (a ratio of 12 asymptomatic to 1 symptomatic sustained episodes of arrhythmia have been documented in paroxysmal AF patients).1 Onset is precisely known in monitored hospitalized patients, and is unknown in incidentally discovered AF, and so new or newly discovered AF patients are a mixture of acute, subacute, and chronic AF. In three recent reviews which discuss recent onset AF, one suggested that anticoagulation was not indicated,2 another suggested “ . . . impetus for early, full anticoagulation . . . ,”3 while the American College of Chest Physicians (ACCP) Consensus Conference on Antithrombotic Therapy declined to make speci~c recommendations for antithrombotic therapy in this common clinical setting, citing inadequate data.4 The latter authors did note that the usual practice is to withhold anticoagulation for patients with AF of #48 hours duration. While acknowledging the absence of randomized trial data upon which to base decisions, recent studies have provided a useful perspective from which to make provisional recommendations in favor of early anticoagulation.3,5 This perspective includes four basic ~ndings: (1) Stroke risk in AF patients is highly variable and dependent upon easily de~nable patient characteristics; (2) atrial thrombus is prevalent even in AF of brief duration; (3) cardioversion (spontaneous, pharmacologic, or electrical) is frequent in new or newly discovered AF; and (4) cardioversion (especially electrical) is associated with transient increased risk of atrial thrombosis and stroke in AF patients. AF patients at highest risk (annual thromboembolism rate .10%/yr) have prosthetic cardiac valves, rheumatic mitral stenosis, or nonvalvular AF with a history of prior thromboembolism. Those at moderate risk (annual thromboembolism rate 5–10%/yr) include patients with a history of hypertension, diabetes, recent congestive heart failure or moderate to severe left ventricular dysfunction. The lowest risk group are patients without mitral stenosis or prosthetic valves, hypertension, diabetes, congestive heart failure or history of prior thromboembolism in whom a ,5%/yr stroke rate is expected.4,5 Recent studies employing transesophageal echocardiography (TEE) have documented left atrial thrombi in 10–13% of AF patients with known AF onset of ,3 days and no incident episode of thromboembolism.6,7 AF isolated to the post-coronary bypass period has been associated with higher stroke rates vs. non-AF in 5 studies,8–12 a three-fold increase in stroke risk in a sixth study,13 and no increased risk of stroke in a seventh study (Table 1).14 Spontaneous conversion (no therapy or rate control medications only) to sinus rhythm occurs in 40–70% of patients with recent onset AF,15 and elective pharmacologic or electrical cardioversion is a consideration for many others. New or increased spontaneous echo contrast, and diminished atrial appendage velocities, both indicating increased atrial stasis, new thrombi, and thromboembolic events have been reported after cardioversion when no precardioversion atrial thrombus was noted by TEE.3 A vulnerable period of augmented thromboembolic risk appears to follow cardioversion, especially electrical cardioversion, but also in some proportion of AF patients who convert spontaneously or after antiarrhythmic therapy. For electrical cardioversion, atrial stunning may be present in at least 31–41%,3 and atrial contractile function may take 3 weeks or more to recover. Electrical cardioversion without anticoagulation is associated with a 4–7% rate of thromboembolism and cannot be advocated, while either the conventional anticoagulation for cardioversion4 or TEE facilitated approaches appear associated with a low (0–2%) rate of thromboembolism.3 These two strategies are being compared in a clinical trial.16 Atrial function appears to recover more rapidly with pharmacologic vs. electrical cardioversion,17 and with AF of short vs. long duration,18 such that pharmacological conversion of AF of #5 weeks duration results in near normal Doppler-assessed atrial ~lling velocities and fractions measured 24 hours after conversion.