Risk Stratification for Stroke in Atrial Fibrillation: A Critique

Ammar M. Killu; Christopher B. Granger; Bernard J. Gersh


Eur Heart J. 2019;40(16):1294-1302. 

In This Article

Biomarkers and Atrial Fibrillation

Biomarkers representing myocardial injury (cardiac troponin) and myocardial stress (NT-proBNP) are strongly associated with stroke in AF, even after accounting for the clinical characteristics incorporated into traditional risk scores. Credence to this is provided by an ARTISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in AF) trial sub-study that demonstrated even low-risk patients with elevated high-sensitivity (hs)-troponin are at much greater risk of thromboembolism in addition to bleeding and cardiac death (Figure 4),[46] albeit in an anticoagulated population. NT-proBNP is also associated with adverse events in patients with AF irrespective of CHA2DS2-VASc score.[47] In fact, a single baseline value of NT-proBNP had more prognostic ability than the entire CHADS2 score. Given the suboptimal C-statistic for current scoring systems, use of biomarkers could help refine risk assessment in AF patients and novel stroke risk scores will likely incorporate these parameters. One example is the ABC (age, biomarkers, and clinical history) score though this only had a modestly improved C-statistic over current scores,[48,49] which may diminish over long-term follow-up.[50] Given the association of elevated biomarkers with adverse outcomes, certain patients may be candidates for anticoagulation based on these alone. In recent years, it has been shown that elevated troponin levels are independently associated with an increased risk of stroke.[51,52] Furthermore, the risk appears to be linearly related to the degree of elevation. The reasons behind the association are undefined; however, it may be related to oxidative stress, oxygen supply/demand mismatch, inflammation, myocyte fibrosis, microvascular dysfunction, and micro-emboli. NT-proBNP is also associated with adverse events in AF[53,54] and the risk may be more than that associated with troponin.[47] Again, mechanisms are undefined, though may be related to myocyte stress. At present, the use of biomarkers in AF is not of much clinical use. However, they are likely to improve understanding regarding the pathophysiology of AF itself and associated stroke risk; in addition, the elevated risk of adverse outcomes including stroke associated with increased biomarker values suggests that they are useful in identifying patients unlikely to fully respond to anticoagulants, placing further emphasis on risk factor modification to reduce stroke risk. Furthermore, they provide the promise for early disease detection (e.g. screening of patients with elevated biomarker levels) and novel methods of treatment (e.g. drug targets). Other important biomarkers associated with increased stroke and other adverse events are those demonstrating renal dysfunction (glomerular filtration rate and Cystatin C),[55] coagulation abnormalities (D-dimer),[56] and inflammation (IL-6 and CRP).[57] While some of these biomarkers may simply be markers of risk for future events, others may be associated with the underlying pathophysiology of both AF and cardiovascular events (Figure 5). Though biomarkers may provide a (marginal) improvement in prediction, the associated cost and practicality of measuring multiple biomarkers needs to be considered before these are routinely used.[58] In addition, biomarkers are non-specific and also predict hospitalization, myocardial infarction and overall mortality. Finally, biomarker data in non-anticoagulated patients with AF are required. Whether or not anticoagulation is beneficial in individuals conventionally deemed 'low risk' on the basis of elevated biomarkers is unclear.

Figure 4.

The importance of high-sensitivity troponin levels in stroke and systemic embolism occurrence irrespective of CHA2DS2-VASc score is demonstrated.

Figure 5.

The potential role of biomarkers in atrial fibrillation and risk of stroke is demonstrated. AF, atrial fibrillation; BNP, brain-natriuretic peptide; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; LA, left atrial; LAA, left atrial appendage; LVH, left ventricular hypertrophy; vWF, von Willebrand factor.