Letter Regarding Article by Gage et al, “Selecting Patients With Atrial Fibrillation for Anticoagulation: Stroke Risk Stratification in Patients Taking Aspirin”
To the Editor:
The use of meta-analysis to refine stroke risk stratification in atrial fibrillation has enabled the separation of the stroke incidences in patients with pure arrhythmic syndromes from patients with ≥1 additional significant cardiovascular comorbidities. Accordingly, the new report by Gage et al1 reiterates that patients with nonvalvular atrial fibrillation designated “low risk” by virtue of the absence of any major cardiovascular comorbidities exhibit very low stroke incidence. Stroke incidence in the “low-risk” subgroups defined by several methods ranged from 0.5 to 1.4 events per 100 patient-years, which closely parallels the findings of previous independent patient-level meta-analyses from the same authors. Importantly, this range of stroke incidence is comparable to but not higher than the normal baseline stroke incidence of age- and sex-matched standard populations.2 What the authors have not discussed, and the medical community has not confronted, is whether these demonstrations of exceedingly low stroke incidence in the pure arrhythmic syndrome refute a cause–effect relationship between atrial fibrillation and stroke.
If atrial fibrillation is an important cause of stroke, then stroke incidence should be markedly elevated in patients with isolated idiopathic atrial fibrillation. The demonstrated absence of any excess stroke attributable to the arrhythmia itself should refute any significant cause–effect relationship. It is especially noteworthy that this key evidence negating a cause–effect relationship is obtained from clinical trials that are themselves premised on the operation of a substantial cause–effect relationship between atrial fibrillation and stroke. The medical community should urgently engage in testing the validity of this paradigm. Part and parcel of this crucial task would be to articulate a fair threshold value or range of values below which observed stroke incidences in idiopathic atrial fibrillation would refute the hypothesis that atrial fibrillation is an important cause of stroke.
We thank Drs Boyajian and Otis for asking the interesting question whether the association between atrial fibrillation and stroke in our study1 and another2 may be a consequence of confounding factors rather than causal. In support of this hypothesis, they highlight the low stroke rates in patients who had atrial fibrillation without other additional stroke risk factors. For example, in our study, the stroke rate was only 0.8 stroke per 100 patient-years of aspirin therapy in the 469 participants who had nonvalvular atrial fibrillation and none of the following additional stroke risk factors: congestive heart failure, hypertension, age >75 years, diabetes mellitus, or previous stroke/transient ischemic attack.
Without atrial fibrillation, these patients would have had a lower risk of stroke. For example, in Framingham studies from 25 years ago, when stroke rates were slightly greater, there was 0.5 stroke per 100 patient-years in men who were free of atrial fibrillation and several other stroke risk factors.2 These investigators found that nonvalvular atrial fibrillation increased the risk of stroke 5- to 6-fold, after controlling for other comorbid conditions. More recently, participants in the Women’s Health Study who were free of major comorbid conditions had fewer than 0.1 strokes per 100 patient-years of aspirin therapy.3
In patients with atrial fibrillation, stasis-precipitated thrombi in the left atrial appendage can embolize to the brain (and elsewhere).4 A powerful argument favoring this mechanism is the striking reduction in ischemic stroke by warfarin over aspirin in atrial fibrillation patients5 but not in other patients. Why, then, do patients with lone atrial fibrillation have a lower stroke risk than others with atrial fibrillation? It is a matter of degree: minimal stasis with lone atrial fibrillation but increasingly sluggish (often with appendage thrombi) in patients with additional stroke risk factors.6 In short, there is a range of stasis severity in atrial fibrillation that correlates with the number of stroke risk factors and risk of causing stroke.
Dr Petersen is a member of the executive steering committee of the Stroke Prevention by ORal Thrombin Inhibitor in atrial Fibrillation (SPORTIF) trials. Dr Gage has served on the scientific advisory board of ev3. Dr Pearce has served as a consultant to the Secondary Prevention of Small Subcortical Stroke.
Gage BF, van Walraven C, Pearce L, Hart RG, Koudstaal PJ, Boode BS, Petersen P. Selecting patients with atrial fibrillation for anticoagulation: stroke risk stratification in patients taking aspirin. Circulation. 2004; 110: 2287–2292.
Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham Study. Stroke. 1991; 22: 312–318.
Zabalgoitia M, Halperin JL, Pearce LA, Blackshear JL, Asinger RW, Hart RG. Transesophageal echocardiographic correlates of clinical risk of thromboembolism in nonvalvular atrial fibrillation. Stroke Prevention in Atrial Fibrillation III Investigators. J Am Coll Cardiol. 1998; 31: 1622–1626.