Study Population

Patients admitted to the emergency department for persistent, non-self-terminating atrial tachyarrhythmia who were candidates for CV were eligible for this study. Inclusion criteria were age ≥ 18 years and arrhythmia duration > 2 days. Patients were informed of the risks and benefits of the two treatment strategies (i.e., ≥3 weeks of oral anticoagulation followed by CV or expedited CV using transesophageal echocardiographic guidance). Patients who gave written consent for TEE-guided CV were prospectively enrolled in the study. Exclusion criteria were onset time > 1 year, underlying acute coronary syndrome, pulmonary embolism or thyrotoxicosis, and the presence of contraindications to TEE. Hemodynamic instability was not a contraindication to enrollment. All patients received anticoagulation with intravenous sodium heparin, low–molecular weight heparin, or an oral anticoagulant (warfarin) at therapeutic doses. Patients free of thrombi on TEE were candidates for CV. Population characteristics are summarized in Table 1 . Heart failure was defined as the presence of a clinical scenario of decompensated heart failure at admission with the need for diuretic therapy. Excluding mild degrees of valvular dysfunction, 124 patients were affected by significant valvular heart disease: in 28 cases, prosthetic valves were present, and among patients with native valvulopathy, there were 39 with mitral insufficiency, 13 with mitral stenoses, 34 with aortic stenoses, and 10 with aortic insufficiency. Mitral stenosis, defined as valve area < 2 cm, included rheumatic (nine cases) and degenerative (four cases) forms. The average duration of arrhythmias was 18 days (range, 1–360 days). Informed consent was obtained from all patients.

TEE

All patients underwent transthoracic echocardiography and multiplane TEE <24 hours before the scheduled CV using an Acuson Sequoia 512 ultrasound system (Siemens Healthcare, Erlangen, Germany) from 1998 to 2001 and a Vivid 7 ultrasound system (GE Healthcare, Milan, Italy) from 2002 to 2013. Echocardiographic examinations were performed by experienced operators certified by the Italian Society of Cardiovascular Echography, with competence level III according to the American Society of Echocardiography, performing about 100 transesophageal echocardiographic examinations per year. Studies were digitally recorded in cine loops and reviewed by two independent operators. Interobserver discrepancies occurred in six cases, and accord was reached after review of the images with another well-trained observer not involved in the study. Artifacts were considered when location and echogenicity suggestive of pectinate muscles or of reverberations (in particular the interatrial septum or the terminal crest for the RAA and the left upper pulmonary vein for the LAA) were obtained at different transesophageal rotational angles.

TEE was performed using a 5-MHz multiplane probe. The probe was inserted into the esophagus after topical anesthesia with lidocaine spray. Viscous lidocaine solution was used on the probe to facilitate introduction. Intravenous midazolam hydrochloride 2.5 to 5 mg was used as needed. No significant adverse effects occurred during TEE. Images of the LAA were acquired at 0°, 60°, 90°, and 130° to explore all of its lobes entirely. The RAA was examined in a midesophageal section at 90° and 130°, and additional views were used as needed: upper esophagus transverse 0° with a clockwise rotation of the probe and transgastric right bicameral view at 130° ( Figure 1 ). Generally, the RAA was fully displayed until its bottom at 130 ± 10° and the LAA at 60 ± 10°. However, to achieve a complete exploration and differentiate thrombus from stacked pectinate muscles and reverberations, at least two different views were visualized. To obtain a good pulsed Doppler velocity spectrum, the best view was chosen. The presence of spontaneous echocardiographic contrast (SEC) was classified according to Fatkin et al . Dense persistent and very low mobile SEC (Fatkin class 4) was differentiated from true thrombus, defined as a separate mass with a clear, well-defined, “nonsmoking” edge. LAA flow velocities were measured by placing the pulsed Doppler sample volume in the middle part of the LAA orifice, 1 cm inside the cavity. RAA flow velocities were measured in the same modality ( Figure 2 ). In all cases, the highest positive velocities within each of five consecutive cycles were measured and then averaged. When measurements using different views were discordant, the highest value was retained. A cutoff of 25 cm/sec was used to define low emptying velocity.

RAA transesophageal echocardiographic views. ( Right ) Midesophageal bicaval view at 90°. ( Left ) From this position, with a rotation of the probe until 130°, the RAA bottom can be entirely explored. IVC , Inferior vena cava; LA , left atrium; RA , right atrium; SVC , superior vena cava; TV , tricuspid valve. Arrows indicate RAA.

RAA color Doppler study. (A) Emptying flow is represented by the red color inside the RAA during its contraction. (B) Pulsed Doppler sampling inside the right appendage in a normal patient in sinus rhythm, orange arrowhead indicates the point of peak ejection velocity measurement.

Statistical Analysis

All data are expressed as mean ± SD. For anatomic and functional comparisons between groups, continuous variables of mean LAA and RAA data were compared by using independent-samples Student’s t tests. Categorical variables between groups were compared by using χ ² tests. All comparisons were two tailed, with significance set at P ≤ .05. Statistical analysis was performed on the entire population and then on the two subgroups, AF and AFL. Differences in the incidence of appendage THR and in peak emptying velocities were investigated in the AF and AFL groups.

Follow-Up

On warfarin, at a median 2-month transesophageal echocardiographic follow-up, THR resolution was documented in all RAA cases, and no clinical signs of pulmonary embolism were detected. In the 96 patients with atrial THR, warfarin was prescribed, with a target international normalized ratio of 2.5 to 3. Among the 887 THR-free patients, CV was not performed in 142 (16%), for several reasons, such as the presence of severe left atrial enlargement, severe SEC or “sludge,” low appendage velocities, or severe valve disease, suggesting a low probability of restoration and maintenance of stable sinus rhythm. Seven hundred forty-five patients underwent CV within 2 days (range, 0–6 days), and successful restoration of sinus rhythm was obtained in 611 (82%). Among 96 patients with THR, TEE was repeated in 62 (65%), after an average anticoagulation period of 48 days, and complete resolution of THR was found in 33 (53%).

At 3-month follow-up, no cases of evident pulmonary embolism were observed in the group of patients who underwent CV, while three clinically manifest cerebral embolic events (0.4%) occurred. None of the seven patients with RAA THR experienced clinically relevant pulmonary embolism; among this group, we did not observe manifest systemic embolic events either, but no one met the criteria for atrial septal aneurysm, and only one case of patent foramen ovale was found. This is an important issue, because paradoxical systemic embolism might occur in cases of interatrial shunt.