TEE and cardioversion

MILAN, ITALY. Electrical cardioversion in the presence of a blood clot (thrombus) in the left atrium is dangerous in that the electric shock may dislodge the clot and send it on its way to the brain where it may cause an ischemic stroke. Thus, cardioversion is usually not attempted unless the patient has been in afib for less than 48 hours or has been treated with warfarin for at least 3 weeks prior to the attempted cardioversion. As an alternative to waiting 3 weeks patients may undergo a TEE (transesophageal echocardiogram) and if this shows no evidence of thrombi the cardioversion may safely proceed. A TEE cannot only detect thrombi, but can also detect the presence of so-called spontaneous echocontrast (SEC) which is considered a forerunner for thrombi. SEC is seen as a swirling pattern (fog) on the TEE and the denser the pattern the more likely it is that a clot will eventually develop. The most dense pattern, presumably just prior to clot formation, is called “sludge”.

A group of Italian researchers recently conducted a study to determine the effectiveness of various anticoagulation protocols in preventing SEC and also to see at what densities of SEC it would still be safe to proceed with the cardioversion. Their study involved 1104 persistent atrial fibrillation patients whose current episodes had lasted an average of 87 days. Ten percent of the group had lone AF, 20% had heart disease, and 42% had hypertension. The patients had undergone several different anticoagulation protocols prior to their cardioversion.

  • Group 1 – 3 weeks of warfarin with an INR consistently above 2
  • Group 2 – short-term (less than 4 days) anticoagulation with low-molecular weight heparin (Enoxaparin)
  • Group 3 – 3 weeks of warfarin with at least one INR measurement below 2
  • Group 4 – less than 3 weeks of warfarin with an INR consistently above 2

All study participants underwent a TEE within 24 hours of their scheduled cardioversion. Sixty-five patients (5.9%) were found to have thrombi, mostly in the left atrial appendage (LAA), and had their cardioversion cancelled.

The majority of the remaining patients (75.8%) were found to have some SEC with 20% having moderate or severe SEC and 5% having sludge. The severity of SEC was significantly correlated with LAA emptying and filling velocity or, to put it another way, the more sluggish the blood flowed in and out of the LAA the more serious the SEC. Patients with no SEC had an average outflow velocity of 40 cm/s, while those with sludge had an average flow velocity of 17 cm/s. It is well established that LAA in- and out-flow is almost entirely dependent on left ventricular ejection fraction. Those with a normal fraction have good flow in and out of the LAA and thus by inference little or no SEC. Conversely, an impaired left ejection fraction is associated with poor LAA blood exchange and thus a greater incidence of severe SEC, sludge and thrombi.

Surprisingly, there was no evidence that the group with 3 weeks of perfect INR control (on warfarin) had fewer thrombi or less severe SEC than groups 2 and 3. The cardioversion was postponed in 65 patients with thrombi and in 87% patients with severe SEC or sludge. However, the attending physicians decided to proceed in the case of 23 patients with severe SEC and 4 with sludge. Cardioversion was successful in all but one of these patients without thromboembolic events.

Maltagliati, A, et al. Incidence of spontaneous echocontrast, ‘sludge’ and thrombi before cardioversion in patients with atrial fibrillation: new insights into the role of transesophageal echocardiography. Journal of Cardiovascular Medicine, Vol. 10, No. 7, 2009, pp. 523-28

Editor’s comment: This large study again proves that the risk of clot and sludge formation in patients with normal left ventricular ejection fraction, such as would normally be found in lone afibbers, is indeed extremely low. It, unfortunately, also shows that even perfect anticoagulation for 3 weeks prior to cardioversion is no guarantee that clots will not be present at the time of the procedure.