Embolism Risk Versus Stroke Risk: More Than a Matter of Semantics

To the Editor: In a recent study that enrolled subjects with atrial fibrillation (AF) who had had an acute myocardial infarction, criteria were established to distinguish between an embolic etiology and an atherothrombotic etiology for coronary artery occlusion.1 When ischemic stroke occurs in the presence of nonvalvular AF (NVAF), a similar distinction should be made between cerebrovascular occlusion attributable to cardiogenic embolism and thrombotic occlusion attributable to atherosclerotic disease, given that as many as 29.6% of individuals with NVAF may have coexisting stenotic cerebrovascular disease,2 which is a risk factor for ischemic stroke in its own right. Furthermore, for the purpose of identifying individuals with NVAF who would benefit from thromboprophylaxis against cardiogenic embolism, risk stratification schemes should be based more on parameters indicative of embolic risk, in general, than on those that focus only on risk of stroke. In the context of NVAF-associated stroke, a major shortcoming of the CHADS2 (congestive heart failure, hypertension, aged ≥75, diabetes mellitus, stroke) and CHA2DS2-Vasc risk stratification schemes3 is that parameters such as age, hypertension, and diabetes mellitus may be predictive not only of embolic stroke, but also of ischemic stroke attributable to coexisting cerebrovascular atherosclerosis. For example, in a study in which the prevalence of cerebrovascular atherosclerosis was as high as 29.6% of individuals with NVAF presenting with stroke, a positive correlation (correlation coefficient = 0.187, P < .001) was documented between CHADS2 score and number of arteries with concomitant cerebral atherosclerosis,2 indicating that, in some of those individuals, stroke might have been attributable to cerebrovascular disease rather than cardiogenic embolism. The ability to predict systemic embolism, in general, as opposed to exclusive prediction of stroke, can be enhanced by incorporating parameters of left atrial (LA) and left atrial appendage (LAA) thrombogenicity into risk stratification schemes, given that several studies have shown that parameters of LA blood stasis are highly predictive of cardiogenic embolism.4 A study that compared vitamin K antagonists with aspirin for prevention of systemic embolism in individuals with paroxysmal atrial fibrillation (PAF) exemplified the incorporation of parameters of LA stasis into the CHA2DS2-Vasc risk stratification scheme. In that study, after exclusion of individuals with LA stasis or thrombogenic stigmata in the ascending aorta, 238 (mean age 65) remained who had a mean CHA2DS2-Vasc score of 2.1. In the absence of a prothrombotic state in the left atrium, that study showed no advantage of vitamin K antagonists over aspirin in preventing ischemic stroke,5 presumably because exclusion of individuals with LA stasis was tantamount to exclusion of individuals who were at risk of cardioembolic stroke Accordingly, the superiority of vitamin K antagonists over aspirin for mitigating the risk of NVAF-associated stroke6 may hold true only in the presence of a prothrombotic state in the left atrium and LAA. Furthermore, when systemic embolism (including extracranial embolism) occurs in NVAF despite aspirin therapy,7 this might well be attributable to coexistence of a LA and LAA prothrombotic state, which is not amenable to modification using antiplatelet agents. Likewise, in a study in which a prothrombotic state in the left atrium and LAA had not been excluded, combined clopidogrel and aspirin therapy proved to be just as ineffective as the sole use of aspirin for preventing NVAF-related extracranial systemic embolism, although combined therapy had greater efficacy in preventing ischemic stroke.8 When combined clopidogrel and aspirin therapy proved to be as ineffective as sole use of aspirin in preventing NVAF-related extracranial embolism,8 the failure of both antiplatelet regimes was probably attributable to a coexisting prothrombotic state in the left atrium and LAA that was not amenable even to combined antiplatelet therapy. When combined therapy was superior to sole use of aspirin in preventing strokes,8 the greater efficacy was probably because combined therapy was superior to sole use of aspirin for modifying the natural history of cerebrovascular disease that might coexist with NVAF. Accordingly, a clinical decision aid for the selection of antithrombotic therapy for prevention of stroke due to atrial fibrillation9 should include the evaluation of LA and LAA thrombogenicity to ensure that the use of antiplatelet agents is not implemented in individuals with NVAF who have high thrombogenicity in the left atrium and LAA. Nevertheless, the fact that evaluation of LA/LAA thrombogenicity relies on the use of transesophageal echocardiography, which is invasive, hinders it use, although with the advent of four-dimensional flow magnetic resonance imaging (MRI), LA stasis can be evaluated noninvasively.10 In the MRI study, which compared LA flow in four individuals with persistent AF with 25 individuals in sinus rhythm (including 6 cardioverted from AF), LA blood flow was significantly (P = .005) lower in individuals with persistent AF than in than who had been successfully cardioverted from AF to sinus rhythm. LA blood flow was also significantly (P < .001) lower in individuals with persistent AF than in 19 age-matched volunteers.10 Accordingly, MRI holds promise for incorporation of parameters of LA and LAA thrombogenicity into schemes for stratifying risk of NVAF-related cardiogenic embolism. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the author and has determined that the author has no financial or any other kind of personal conflicts with this paper. Author Contributions: Oscar Jolobe is the sole author of this paper. Sponsor's Role: None.