The recently published report by Parikh et al reinforces the thromboembolic nature of atrial flutter. The described prevalence of left atrial (LA) thrombi and spontaneous echocardiographic contrast (SEC) in this population presenting with a mean CHADS ₂ score of 1.62 ± 1.18 and a mean CHA ₂ DS ₂ -VASc score of 2.67 ± 1.55 was 5.3% and 25.9%, respectively.

In a recently published investigation by my group in patients with nonrheumatic atrial fibrillation with slightly higher clinical risk (mean CHADS ₂ score 2.21 ± 1.18, mean CHA ₂ DS ₂ -VASc score 3.89 ± 1.71), when assessed by transesophageal echocardiographic (TEE) imaging, a similar prevalence of LA thrombus and SEC (9.8% and 22.1%, respectively) to that described by Parikh et al was observed. Moreover, Parikh et al found a higher sensitivity of CHA ₂ DS ₂ -VASc score for the detection of LA thrombus and SEC at the expense of lower specificity compared to CHADS ₂ score. This also applies to what was found in our atrial fibrillation sample. Therefore, as far as risk stratification is concerned, CHADS ₂ and CHA ₂ DS ₂ -VASc scores displayed similar performance for the atrial fibrillation and atrial flutter populations.

Unlike in our report, in which a zero prevalence of LA thrombus and SEC was found in subjects with CHA ₂ DS ₂ -VASc scores of 0, in subjects with atrial flutter, a prevalence of 6.7% for LA thrombus and 10% for SEC was observed in the group presenting with similar risk as assessed by CHA ₂ DS ₂ -VASc score. Because the study by Parikh et al was retrospective and included some patients who underwent TEE imaging performed in 2002, we hypothesize that some information concerning clinical risk stratification necessary for the calculation of CHADS ₂ and CHA ₂ DS ₂ -VASc scores may have been missing, which may have resulted in an underestimation of risk and the presence of TEE markers of LA stasis in subjects misclassified as having CHA ₂ DS ₂ -VASc scores of 0. The presence of LA thrombus and SEC does not necessarily mean that these patients will sustain thromboembolic events but has been associated with increased risk for stroke and thromboembolism: relative risk of 2.5 (p = 0.04) for LA thrombus and 3.7 (p <0.001) for SEC. Moreover, in the CHA ₂ DS ₂ -VASc validation cohort, a score of 0 was defined as truly low risk, because no events were detected in these patients during follow-up. Therefore, the presence of these TEE changes in subjects classified as having CHA ₂ DS ₂ -VASc scores of 0 is different from what we would expect.

The observed overlap of the remaining TEE findings in patients with atrial fibrillation and atrial flutter reinforces that thromboembolic risk in atrial flutter is by no means inferior to the risk observed in atrial fibrillation, despite the slightly more preserved LA motility. In this way, a common link may be considered between atrial flutter and atrial fibrillation as far as thromboembolic nature is concerned (as assessed using these surrogate TEE markers of an increased risk for stroke). In this way, we think that risk assessment and prevention of thromboembolism in nonrheumatic atrial fibrillation and atrial flutter should be addressed using exactly the same approach.