LAUSR

Atrial fibrillation (AF) presents a five-fold increase of the risk for ischaemic stroke and 20–30% of ischaemic strokes appear to be directly associated with AF. Currently, the need for anticoagulation for nonvalvular AF is assessed on the basis of clinical parameters using the CHAD2DS2-VASc score. 2 Conventional echocardiographic parameters such as left atrial (LA) dilatation, LA appendage thrombus > sludge > smoke, and reduced LA appendage velocities (<20 cm/s) can further risk stratify AF patients who would most benefit from anticoagulation. However, most of these parameters require transoesophageal echocardiography and do not evaluate the full risk spectrum related to LA functional impairment in AF. Moreover, whether they remain predictive of stroke after adjustment for comorbid factors (e.g. age, hypertension, diabetes, and previous stroke), which are the most critical upstream concomitant conditions promoting atrial remodelling, has not been well examined. Beyond the usual methods, including transmitral flow and changes in LA area and volume, novel techniques such as tissue Doppler imaging (TDI) and strain imaging provide more accurate estimation of LA function. LA strain is correlated with LA fibrosis in patients with AF, is significantly reduced in those with prior stroke, and predicts subsequent stroke. In their study, Leung et al. further extended this observation in a large registry cohort of 1361 patients with first diagnosis of AF. At baseline echocardiography, LA volumes, LA reservoir strain, P-wave to A’ duration (PA)-TDI (tissue Doppler imaging), and left ventricular (LV) global longitudinal strain were evaluated in patients with and without stroke. During follow-up (mean: 7.9 years), 100 patients (7%) developed an ischaemic stroke, representing an annualized stroke rate of 0.9% (1.2% in patients who had a CHA2DS2-VASc score of <_1). LA reservoir strain was reduced while PA-TDI was lengthened in the stroke compared with non-stroke group, and emerged as independently associated with risk of stroke in a model including CHA2DS2-VASc score, age, and anticoagulant use. Notably, these two parameters fitting with the prediction of ischaemic stroke have a deep physiopathological meaning, as they are both an expression of LA function. PA-TDI represents LA electromechanical delay and can be obtained by calculating the time duration difference between the onset of the echocardiogram-derived P-wave and the peak of the A’ wave on TDI. PA-TDI reflects the total LA activation time, and is likely a surrogate measurement of myocardial fibrosis and LA remodelling. PA-TDI has already been identified as an independent predictor of new-onset, post-operative, and recurrent AF after electrical cardioversion. Moreover, in a large population of 279 patients free of AF recurrence after successful catheter ablation, it was shown that the longer the LA electromechanical delay, the higher the risk of stroke. LA reservoir strain, i.e. the collection of flow from pulmonary veins during LV systole, can be quantified by two-dimensional speckle-tracking echocardiography as the peak positive longitudinal strain during LV systole. LA reservoir strain gives additional insights beyond traditional measures of LA remodelling corresponding to LA enlargement, and its alteration is associated with a number of cardiac diseases including AF. In recent studies, LA reservoir strain has shown to provide incremental value for thromboembolic risk stratification over CHAD2DS2-VASc score, 12