Time for A Fib to Lead to Anticoagulation ========================================= The use of TEE has provided new insights into left atrial (LA) thrombus formation in atrial fib (AF) and has implications for anticoagulant treatment. The conventional recommendation has been to anticoagulate for 3 weeks prior to and 4 weeks after a conversion (either electrical or pharmacologic) of AF into NSR. Patients presenting within 48 hours of onset of AF were felt to not require anticoagulation prior to conversion since thrombogenesis required a longer period of time to develop. It has been found that thrombosis, when it occurs in the LA, generally does so in the left atrial appendage (LAA) and not in the body of the atrium proper. It has also been found that although the risk of embolism is greatest when conversion (either spontaneous or induced) occurs from AF to NSR that the formation of the LAA thrombus itself may occur before the first occurence of AF while the patient is still in NSR. Thrombus formation has been documented by TEE to occur, in some cases, within 2 minutes after conversion of AF to NSR. Contrary to conventional teaching, LAA thrombus by TEE has been shown to develop in some cases with the first 3 days of AF in a study by Stoddard. The actual risk of embolism after conversion, by any means, ranges from 0.6 to 5.6%. After cardioversion the LA is frequently in a state of EMD or, at least, greatly diminished contractile function as documented by assesments of mitral valve leaflet excursion by echo. The occurence of spontaneous echo contrast (SCE), a poorly understood phenomenon which has been postulated to reflect changes in blood viscosity within the LA, also correlates with the risk of LAA thrombosis and subsequent embolism. The duration of time that EMD will persist (and thus the theoretical span of time over which LAA thrombosis may occur) is a function of the length of time that AF has persisted prior to conversion. For patients with duration of AF of less than 2 weeks, atrial systolic function rapidly returns to normal within 24 hours of conversion. AF of 2 to 6 weeks duration may require 1 week before EMD resolves and those whose AF has persisted for greater than 6 weeks prior to conversion may require 1 month. When compared to patients who were converted electrically, those converted pharmacologically exhibted a lesser degree of impairment in LA function after return to NSR. This *difference* in the impairment in atrial function between the two modalities for conversion persists up to one week after conversion for patients whose AF has been present less than 5 weeks before conversion. (Have I made myself clear?) If the duration of AF is greater than 4 weeks prior to conversion there is no difference in the extent of impairment in LA function after conversion to NSR in the first 24 hours by either modality. Atrial EMD may also occur after *spontaneous* conversion of AF to NSR and, as has been previously mentioned, occurs but to a lesser degree after pharmacologic as opposed to electrical conversion. A strategy for avoiding the use of anticoagulation *prior* to attempted conversion has been proposed using TEE to assess for the presence of pre-existing LAA thrombus. Nevertheless embolism, immediately after conversion, has been documented to occur in spite of a negative screening TEE. LAA thombosis has also occured after conversion *in spite* of the presence of therapeutic anticoagulation. The above considerations have led some to suggest that anticoagulation should be a *routine* component of the management of acute onset AF reagrdless of its duration. H. Louzon MD (1) Manning et. al. Atrial Anatomy and Function Postccardioversion: Insights from Transthoracic and Transesophageal Echocardiography. Prog in Card Dis 1996;39(1):33-46 (2) Stoddard MF Risk of Thromboembolism in New Onset or Transient Atrial Fibrillation. Prg in Card Dis 1996;39(1):69-80 --------------------- On Fri, 27 Dec 1996, Lombardo F Palma wrote: > Harvey, > > In your review of the day on Anticoagulation for Atrial Fib you wrote: > [snip] > > The actual risk of embolism after conversion, by any means, ranges > > from 0.6 to 5.6%. > > Do you have actual stroke risk rates? > > > The above considerations have led some to suggest that anticoagulation > > should be a *routine* component of the management of acute onset AF > > reagrdless of its duration. > > What about the use of neuroprotective agents? > > First of all the incidence of systemic embolism that I mentioned above was from Stoddard's review (1) which referenced studies done from 1969 to 1989. More recent studies have documented a risk of embolism of about 2% overall after cardioversion without anticoagulation to 0.33% with anticoagulation (2,3). How many of these are cerberal emboli? I don't know but in a study done on chronic AF about 2/3 were cerebral emboli (4). In cases of chronic AF fifty-percent of cerebral emboli occur in patients over the age of 75 and are often of devastating consequence. It has also been recognized that lessor degrees of cerebral ischemia such as 'minor' stroke or TIA may also occur in patients with AF. And although a majority of CVAs in AF are of embolic origin an appreciable minority are thrombotic in nature due to concomitant cerbral vascular disease (5). Whether patients with atrial flutter require anticoagulation prior to conversion or not is a matter of some dispute. What is indisputable is that patients with atrial flutter have a considerably lower risk of embolism than do those with AF. Some have suggested that anticoagulation is unecessary in these cases (3) and others, pointing to case reports, have argued that it might be prudent to do so nevertheless. A nice summary of consensus panel recommendations on the treatment of AF is to be found in (5). Finally, I'm not sure what you mean by the use of neuroprotective agents in this setting. H. Louzon MD (1) Stoddard MF Risk of Thromboembolism in New Onset or Transient Atrial Fibrillation. Prg in Card Dis 1996;39(1):69-80 (2) Mugge A -Anticoagulation in patients with atrial fibrillation- Herz 1996 Feb;21(1):28-36 Nonvalvular atrial fibrillation is a frequent finding in elderly patients; the risk of thromboembolic complications is comparable to that reported in patients with rheumatic atrial fibrillation. Recently, 6 multicenter clinical trials (5 primary prevention, 1 secondary prevention trail) have been published which demonstrate equivocally that oral anticoagulation therapy significantly reduces the embolic risk in patients with nonvalvular atrial fibrillation by about 67% to 87%. The target INR of anticoagulation with warfarin in 2 of these trials was 1.4 to 2.8 ("low-dose" warfarin); interestingly, the magnitude of risk reduction was similar to these 2 studies with "low-dose" warfarin as it has been reported by the others using full-dose warfarin with an INR target between 2.0 and 4.5. Side effects of oral anticoagulation (severe bleeding complications) were low in these trials. Thus, the benefit-risk ratio in these 6 clinical trials encourage the use of oral anticoagulation in patients with nonvalvular atrial fibrillation. It will be a challenge to transfer these results to clinical practice, and to define in more detail the risk-benefit ratios for subgroups of patients with atrial fibrillation, e.g. patients > 75 years of age, or patients with "lone" or paroxysmal atrial fibrillation. It is well established that patients with chronic atrial fibrillation undergoing medical or DC-cardioversion are at risk for thromboembolic complications. In previous studies, this risk appears to be in the range of 2% without concomitant anticoagulation, but only 0.33% in those patients with concomitant anticoagulation. Thus, it is widely accepted that patients should be anticoagulated for at least 2 weeks prior and after planned cardioversion. Recently, an alternative concept has been proposed omitting anticoagulation before cardioversion; instead, transesophageal echocardiography is used to exclude intracardiac thrombi. Because it is known that mechanical function of the left atrium and appendage is still impaired after cardioversion, this concept of echocardiographic-guided cardioversion does not assign the necessity of anticoagulation at the day of cardioversion, and 2 weeks afterwards. The safety aspects of this concept of echocardiographic-guided cardioversion is under current investigation. (3) Arnold AZ, Mick MJ, Mazurek RP, Loop FD, Trohman RG Role of prophylactic anticoagulation for direct current cardioversion in patients with atrial fibrillation or atrial flutter [see comments] J Am Coll Cardiol 1992 Mar 15;19(4):851-5 The need for prophylactic anticoagulation to prevent embolism before direct current cardioversion is performed for atrial fibrillation or atrial flutter is controversial. To examine this issue further, a retrospective review was undertaken to assess the incidence of embolic complications after cardioversion. The review involved 454 elective direct current cardioversions performed for atrial fibrillation or atrial flutter over a 7 year period. The incidence rate of embolic complications was 1.32% (six patients); the complications ranged from minor visual disturbances to a fatal cerebrovascular event. All six patients had atrial fibrillation, and none had been on anticoagulant therapy (p = 0.026). The duration of atrial fibrillation was less than 1 week in five of the six patients who had embolic complications. Baseline characteristics of patients with a postcardioversion embolic event are compared with those of patients who did not have an embolic event. There was no difference in the prevalence of hypertension, diabetes mellitus or prior stroke between the two groups, and there was no difference in the number of patients who were postoperative or had poor left ventricular function. Left atrial size was similar between the two groups. No patient in the embolic group had valvular disease. No patient with atrial flutter had an embolic event regardless of anticoagulant status; therefore, anticoagulation is not recommended for patients with atrial flutter undergoing cardioversion. Prophylactic anticoagulation is pivotal in patients undergoing elective direct current cardioversion for atrial fibrillation, even those with atrial fibrillation of less than 1 week's duration. (4) Roy D, Marchand E, Gagne P, Chabot M, Cartier R Usefulness of anticoagulant therapy in the prevention of embolic complications of atrial fibrillation. Am Heart J 1986 Nov;112(5):1039-43 The medical records of 254 patients with atrial fibrillation were reviewed to determine the incidence of embolic events in relation to type of cardiovascular disease, duration of atrial fibrillation, and use of anticoagulants. During a total follow-up of 833 patient-years in atrial fibrillation, there were 32 instances of systemic embolism: 21 involved the cerebral circulation and 11 were extracerebral. Thirty of these events occurred during 549 patient-years of follow-up without anticoagulation therapy (5.46 of 100 patient-years), while only two embolic events occurred during 284 patient-years on anticoagulants (0.7 of 100 patient-years). Thus, the incidence of embolism was eight times more frequent during the unanticoagulated period of observation in atrial fibrillation (p less than 0.002). The incidence of embolism during follow-up without anticoagulants was the same regardless of the presence or absence of mitral valve disease and regardless of whether atrial fibrillation was chronic or paroxysmal. The rate of serious hemorrhagic complications on anticoagulants was acceptably low (2.11 of 100 patient-years). We conclude that in this study population anticoagulant therapy reduced the risk of embolic complications of atrial fibrillation. The results also indicate that the use of anticoagulants should not be limited to patients with atrial fibrillation due to mitral valve disease. (5) Prystowsky et. al. Management of Patients With Atrial Fibrillation: A Statement for Healthcare Professionals From the Subcommittee on Electrocardiography and Electrophysiology, American Heart Association. Circulation 1996;93:1262-1277 --------------------------------- > Harvey, > > Thanks for your interesting points on atrial fib management. > I asked a cardiologist about cardioverting a patient presenting with > lone afib in the ED. Let's say someone presents with a clear history of > sudden onset of palpitations, and they are found to be in afib in the > ED. Can you envision any scenarios whereby it would be reasonably safe > to chemically or electrically cardiovert them in the ED and follow them > as an outpatient? The cardiologist says no, because of the risk of > thromboembolism. I wonder though if there might be a subset of > patients in which it could be done (a la the Italian or Candian? model) Of course lone AF is a diagnosis of exclusion and can only be made after you have ruled out hypertensive, ischemic and valvular heart disease. Not to mention rarer etiologies such as myocarditis and infiltrative cardiomyopathies. Usually this requires an ECHO and, in some cases, a TMT and perhaps further work up. Patients with true lone AF who are under the age of 60 have a very low risk of embolism under any circumstances. Nevertheless this is not a common diagnosis. In several series 90 to 95% of patients presenting with AF were found to have an underlying cause most commonly hypertensive or ischemic heart disease. The same has been found in autopsy studies (which I hope we are not contributing to in the ED). When patients were stratified as to risk of thromboembolism from chronic AF the studies that did not strictly exclude those with hypertension from qualifying for a diagnosis of lone AF had a higher rate than those who did. But I will certainly grant you that we occasionally see patients in the ED who are young and who have no *apparant* cause of their AF including thryrotoxicosis and alcohol ingestion. (Incidentally, at least 3 studies have document the occurence of embolism in patients with thyrotoxicosis and AF and the majority of them have been to the cerebral circulation.) How safe is it to chemically or electrically convert these people (with lone AF) and send them home? For my money I would never perform elective cardioversion in the ED. Setting aside the necessity of making provisions for sedation and airway control, cadioversion has a small but finite risk of complications including asystole. If I have to cardiovert someone in the ED for any indication they automatically get admitted. Performance of cardioversion results in atrial stunning whose duration is a function of the length of time that AF has been present as I indicated yesterday. Even new onset (< 48 hours) AF will result in atrial EMD after cardioversion for the first 24 hours as the review article by Manning that I cited yesterday points out. What about chemically converting them? This we often at least attempt to do provided the onset of AF is within 48 hours of presentation AND they have no other risk factors for thromboembolism such as mitral valve disease. Just 3 days ago we sucessfully converted a new onset AF with IV procainamide in the ED. Case reports of thrombolembolism occuring after conversion of recent onset AF have been described (3). This has occurred even in the face of a negative screening TEE (2). The reivew by Stoddard that I cited yesterday describes LAA thrombosis occuring literally within minutes of cardioversion. Information concerning the risks of atrial stunning (and thus of thromboembolism) after chemical or spontaneous conversion, as opposed to cardioversion, are limited but this does occur. Information about the precise risks of conversion by any means within the first 48 hours is also limited although this is a generally accepted procedure. You might be interested to read the exchange of letters in Acad EM 1996;3(8):819-821 concerning an earlier study by Mulcahy et. al. where a proposal was made to manage AF on an outpatient basis. This suggestion resulted in 2 letters from authors who were alarmed at the prospect of outpatient management and Mulcahy's reply. The risk of embolism in 3 series of patients with AF who were converted to NSR within 48 hours of onset was 0.9%. Is it really worth the risk? Particularly if Class IA drugs have to be prescribed on an outpaient basis? When the group of patients who might conceivebly be candidates for this management is so small? As I indicated yesterday some authors are now proposing routine anticoagulation for ALL patients with new onset AF irrespective of time of onset. Obviously we are a long way from being able to do this on an outpatient basis, although I agree that this is not the standard of care (yet). Hopefully the currently ongoing AFFIRM trial will settle the issue of the best approach to the management of AF once and for all. So the short answer to your question is no. H. Louzon MD (1) Missault L, Jordaens L, Gheeraert P, Adang L, Clement D Embolic stroke after unanticoagulated cardioversion despite prior exclusion of atrial thrombi by transoesophageal echocardiography. Eur Heart J 1994 Sep;15(9):1279-80 Recent studies in patients with atrial fibrillation, not on anticoagulation, suggest that if transoesophageal echocardiography (TEE) excludes the presence of thrombi, early cardioversion can be performed safely without the need for anticoagulation before the procedure. Immediately after successful cardioversion, however, left atrium or left atrial appendage stunning may be present, potentially carrying a risk for de novo thrombus formation. Furthermore, the presence of spontaneous contrast is considered as a contraindication for unanticoagulated cardioversion since it has been associated with postcardioversion thromboembolism. We present a case in which stroke developed in relation to unanticoagulated cardioversion regardless of careful prior evaluation with TEE. (2) Conti CR Atrial fibrillation, transesophageal echo, electrical cardioversion, and anticoagulation [editorial] [see comments] Clin Cardiol 1994 Dec;17(12):639-40 Although successful electrical cardioversion is accomplished in most cases without any evidence of embolic stroke, a few patients have experienced this catastrophe. The current thinking is that when electrical energy is applied to the chest wall, the atrium, although it returns to sinus rhythm, is stunned. It is not known how long this stunning lasts in the individual patient nor whether high energy produces stunning and low energy does not. Nor is it known whether chemical conversion of atrial fibrillation to sinus rhythm affects the atrium in the same way. However, the atrium seems to recover more quickly in patients with a short duration of atrial fibrillation and these patients may not require the usual four weeks of postcardioversion anticoagulation. Based on what we know, or more precisely what we don't know, it seems reasonable to ensure that every patient with atrial fibrillation is anticoagulated during and after DC cardioversion to sinus rhythm. Of course, this is easy to do with intravenous heparin, but that requires hospitalization. Perhaps subcutaneous heparin in high doses would suffice until the patient can be anticoagulated with coumadin. From the research perspective it might be interesting to perform serial echo/Doppler studies on these patients to identify when the individual patient's atrial function returns to normal. This might provide a clinical rationale for discontinuing anticoagulation. Comparing the time to return of normal atrial function (as measured by Doppler echo) between patients undergoing pharmacologic cardioversion versus electrical cardioversion and studying the relationship of the amount of electrical energy required for cardioversion versus the duration of stunning would be clinical research projects of interest to clinicians. (3) Arnold AZ, Mick MJ, Mazurek RP, Loop FD, Trohman RG Role of prophylactic anticoagulation for direct current cardioversion in patients with atrial fibrillation or atrial flutter [see comments] J Am Coll Cardiol 1992 Mar 15;19(4):851-5 The need for prophylactic anticoagulation to prevent embolism before direct current cardioversion is performed for atrial fibrillation or atrial flutter is controversial. To examine this issue further, a retrospective review was undertaken to assess the incidence of embolic complications after cardioversion. The review involved 454 elective direct current cardioversions performed for atrial fibrillation or atrial flutter over a 7 year period. The incidence rate of embolic complications was 1.32% (six patients); the complications ranged from minor visual disturbances to a fatal cerebrovascular event. All six patients had atrial fibrillation, and none had been on anticoagulant therapy (p = 0.026). The duration of atrial fibrillation was less than 1 week in five of the six patients who had embolic complications. Baseline characteristics of patients with a postcardioversion embolic event are compared with those of patients who did not have an embolic event. There was no difference in the prevalence of hypertension, diabetes mellitus or prior stroke between the two groups, and there was no difference in the number of patients who were postoperative or had poor left ventricular function. Left atrial size was similar between the two groups. No patient in the embolic group had valvular disease. No patient with atrial flutter had an embolic event regardless of anticoagulant status; therefore, anticoagulation is not recommended for patients with atrial flutter undergoing cardioversion. Prophylactic anticoagulation is pivotal in patients undergoing elective direct current cardioversion for atrial fibrillation, even those with atrial fibrillation of less than 1 week's duration.