By Michael H. Crawford, MD, Editor

SYNOPSIS: A study of nonrheumatic patients with atrial fibrillation who were inadequately or not anticoagulated undergoing transesophageal echocardiography before cardioversion showed moderate-to-severe mitral valve regurgitation attenuates the risk of finding left atrial thrombi or spontaneous echo contrast and could be considered a thromboembolic risk modifier.

SOURCE: Van Laer SL, Verreyen S, Winkler KM, et al. Effect of mitral regurgitation on thrombotic risk in patients with nonrheumatic atrial fibrillation: A new CHA2DS2-VASc score risk modifier? Am J Cardiol 2021;145:69-76.

The presence of mitral valve regurgitation (MR) has been shown to reduce thromboembolic events (TE) in atrial fibrillation (AF) patients with rheumatic mitral valve disease by about half. However, the effect of MR in non-rheumatic mitral valve disease is unclear. Van Laer et al conducted this single-center, observational study of 795 consecutive patients undergoing transesophageal echocardiography (TEE) before AF cardioversion in patients with non-rheumatic AF. These patients either were not on anticoagulants or were thought to be inadequately anticoagulated for the three weeks before cardioversion.

Patients with mitral stenosis, mitral valve surgery, left atrial appendage (LAA) ligation/closure, or poor echocardiographic images were excluded. MR was graded as mild, moderate, or severe by standard methods. CHA2DS2-VASc scores of 0-1 were categorized as low risk, 2-3 as intermediate, and greater than 3 as high. Left atrial volume (LAV) was calculated from a recent transthoracic echocardiogram, and left ventricular ejection fraction (EF) was calculated or estimated from the TEE. The primary endpoint was left atrial thrombus (LAT) or spontaneous echocardiographic contrast (SEC) graded at > 2+ out of 4. Of the final population of 686 patients, the mean age was 67 years, 72% were men, and 46% were not on anticoagulants.

LAT or SEC > 2+ was present in 17% of patients. It was 11% in those with a low CHA2DS2-VASc score, 15% with an intermediate score, and 24% with a high score (P = 0.001). The increasing incidence with higher scores was caused mainly by more SEC. The incidence of LAT/SEC decreased with increasing severity of MR: 20% with no MR, 15% with moderate, and 12% with severe (P = 0.03). In a multivariate analysis, there were five independent predictors of LAT/SEC: odds ratio 0.51 for moderate MR, 0.24 for severe MR, 1.25 for CHA2DS2-VASc score, 1.9 for LAV, and 4.08 for EF. When MR grade, LAV, and EF were added to the CHA2DS2-VASc score, the C statistic rose from 0.62 to 0.75. In those with a high CHA2DS2-VASc score, if moderate-to-severe MR is considered, the risk of LAT/SEC decreases to that of an intermediate CHA2DS2-VASc score. If MR is considered in those with an intermediate score, the risk drops to that of a low score. The authors concluded MR attenuates the TE risk in non-rheumatic AF and could be considered as a CHA2DS2-VASc score risk modifier when considering anticoagulant therapy.


The CHA2DS2-VASc score is used widely to determine the risk of TE in patients with atrial fibrillation, but its predictive accuracy is modest (C statistic = 0.6). In addition to the clinical factors in the CHA2DS2-VASc score, echocardiographic parameters, such as LAV and EF, influence the risk of TE. Also, it is thought that moderate-to-severe MR can wash out the LA and prevent stasis. This concept has been demonstrated for rheumatic AF patients, but not in non-rheumatic AF patients. Thus, the Van Laer et al study is of interest.

This was not a clinical outcomes study, but since the risk of TE is strongly related to LAT and SEC, their abrogation in more severe MR is compelling. Also, the authors showed this reduction in risk is mainly caused by less SEC rather than LAT. This makes sense considering the jet of MR does not often reach into the left atrial appendage, where almost all LAT reside. One strength of this study is that it involved patients either not on anticoagulants or inadequately anticoagulated. In anticoagulated patients, it is likely the effect of MR would be attenuated. Another strength is the authors demonstrated the value of considering echocardiographic parameters in addition to the CHA2DS2-VASc score. I have never been a fan of using one clinical score alone to make patient care decisions.

There were limitations to this study. Its retrospective design and medium-sized population precluded the evaluation of TE. Still, considering patients with LAT/SEC almost always would be anticoagulated, the prevention of events would be difficult to study. Also, the authors did not investigate persistent or permanent AF, but there is no reason to suspect the results would be different in such patients. In addition, Van Laer et al could not assess the effect of MR chronicity. With long-standing moderate-to-severe MR, LA dilation could attenuate the washout effect of MR. However, MR was an independent risk factor of LAV and EF for predicting LAT/SEC in these patients inadequately anticoagulated or not anticoagulated. At this point, MR should be considered a TE risk modifier, in addition to LAV and EF, in patients with AF.