Transesophageal Echocardiographic-Guided Cardioversion

Abstract

The incidence and prevalence of atrial fibrillation (AF) continues to grow especially as the population ages. The uncoordinated atrial activation and subsequent ineffective atrial contraction is a strong stimulus for left atrial (LA) thrombus formation especially in the LA appendage. AF is associated with increased risk of stroke, heart failure, and all-cause mortality. The presence of LA thrombus in AF portends a poor prognosis. Direct current cardioversion (DCCV) is the most effective method of restoring sinus rhythm and, with it there is relief of symptoms, improved LV filling, reversed atrial remodeling, and possibly reduced cardio-embolic risk. However, there are significant risks of systemic embolization following DCCV if LA thrombus is present. Transesophageal echocardiography (TEE) is an ideal non-invasive imaging modality to detect thrombus in the LA and especially in the LA appendage. Its proper use can lead to earlier DCCV for AF and improve safety of DCCV in AF. A key part of this strategy is proper use of anticoagulation therapy. This chapter illustrates the important role of TEE in patient evaluation and risk stratification prior to cardioversion.

TEE-Guided Direct Current Cardioversion Cases

Case 1: Thrombus (Fig. 1.5a–g)

A 70 year-old woman presents to the emergency department with shortness of breath. She has a past medical history significant for hypertension, heart failure with preserved ejection fraction, and apical variant hypertrophic cardiomyopathy. She states that she was in her usual state of health when she drifted off to sleep while watching TV. She woke up short of breath but denies any chest pain, fevers, or cough. She intermittently takes her heart failure medications, including her diuretic. She was taken by a family member to the emergency room, where on arrival, she was in respiratory distress, hypertensive, and tachycardic, with a blood pressure of 182/90 mmHg, heart rate of 126 beats per minute, and respiratory rate of 44 breaths per minute, with a oxygen saturation of 100 % on room air. Her initial exam as noted for irregular, tachycardic heart rate, bilateral crackles over her lung fields, and mild lower extremity edema. Her chest x-ray was consistent with pulmonary edema and her laboratory work-up was unremarkable, including negative cardiac enzymes, except for an elevated beta-naturetic pepitide level. Her ECG showed atrial fibrillation, which was a new diagnosis. She was started on a nitroglycerin infusion, placed on a non-invasive positive pressure ventilator (NIPPV), and given IV furosemide. With diuresis, she was quickly weaned off NIPPV and the nitroglycerin drip less than 24 h after presentation. Given her new diagnosis of atrial fibrillation, she was also started on IV heparin with plan for TEE-guided DCCV.

Her TEE revealed severe SEC in her LA and LAA, which had a large primary lobe and smaller adjacent lobe. There was a large, mobile echodensity in the LAA, which prolapsed into the left atrium, which was consistent with thrombus. Her LAA peak emptying velocity was 38 cm/s. Of note, a TTE performed prior to her TEE showed spontaneous echocontrast but did not detect any thrombus in the LA.

Fig. 1.5

(a–g) Case of LAA thrombus (Case 1) (a–d) demonstrate a large mobile echodensity within the primary lobe of the LAA that prolapses in the LA. (b) Depicts the simultaneous, multiplanar view of the thrombus. (e) Shows a 3D reconstruction of the thrombus. (f) Shows color flow around the thrombus, offering additional evidence that the echodensity was indeed thrombus and not an echo artifact. (g) Shows the erratic emptying velocities of the LAA in atrial fibrillation, with an emptying velocity of 38 cm/s

Her elective DCCV was canceled and she was bridged on IV heparin and started on warfarin therapy until her INR was >2. She was discharged with outpatient follow-up; at her clinic visit, she was noted to have self-converted into normal sinus rhythm. There were no obvious clinical manifestations of thromboembolism on history or physical exam.

Clinical Pearl

Regarding utilizing TTE to detect thrombus, the Comprehensive Left Atrial Appendage Optimization of Thrombus (CLOTS) multicenter pilot study demonstrated that TTE may be useful in detecting thrombus when utilizing harmonic imaging combined with IV echocontrast. In that study, LAA wall tissue Doppler velocities also correlated to the presence of sludge or thrombus (apical E velocity ≤9.7 cm/s, anterior S velocity ≤5.2 cm/s) [120]. TEE, however, remains the gold standard for LAA thrombus detection. While a majority of patients have a multilobular appendage, which has been shown to be independently associated with thrombus formation, 20 % of subjects in an autopsy study had a single lobe [34, 121]. Simultaneous multiplanar imaging can be essential in determine the lobular structure of the LAA and detecting possible thrombus.

Case 2: SEC/Sludge/Thrombus (Fig. 1.6)

A 69 year-old man presents to outpatient clinic with increasing fatigue. He has a past medical history significant for atrial fibrillation, hypertension, hyperlipidemia, sleep apnea, congestive heart failure, and diabetes. He has a known non-ischemic cardiomyopathy with left ventricular systolic ejection fraction of 20 %. He has an ICD that was placed for primary prevention of sudden cardiac death. He has been compliant with his medications, including his warfarin for CVA prevention (CHADS₂ risk score = 3, 5.9 %/year risk of thromboembolic event, CHA₂DS₂-VASc risk score = 4, 9.3 %/year risk of stroke) as well as his heart failure regimen. He was diagnosed with atrial fibrillation 6 months prior to presentation and underwent an uneventful TEE-guided cardioversion and was placed on dofetilide. However, his antiarrhythmic regimen was changed to amiodarone due to patient preference several months later. On exam, his vital signs were acceptable and he clinically appeared euvolemic with flat neck veins and no peripheral edema. His heart sounds were notable for an irregular rhythm; atrial fibrillation was confirmed on ECG. His INR was therapeutic at 2.3. On interrogation of his ICD, it was found that his atrial fibrillation burden was 2 %. There was concern that the loss of atrial kick or heart rate changes secondary to atrial fibrillation was worsening his heart failure symptoms, thus it was decided that the patient should have a TEE-guided DCCV.

His TEE revealed no left atrial appendage thrombus but the presence of severe spontaneous echo contrast (Fig. 1.6a–d).

Fig. 1.6

(a–d) Spontaneous echocontrast (Case 2) (a–c) demonstrate severe echocontrast within the LAA but no LAA thrombus. Simultaneous multiplanar imaging was used to evaluate for possible thrombus. (d) Shows the SEC that appears within the lumen of the LAA in (d). (e–h) Sludge (Case 2) (e–f) demonstrate severe echocontrast within the LAA as well as layered sludge. (g) Shows that color flow Doppler to the apex of the LAA. (h) Shows reduced LAA emptying velocities. (i–j) Thrombus (Case 2) (i–j) demonstrate formation of a thrombus in the apex of the LAA. (j) Shows marking of the thrombus by the sonographer. Overall, the Figures in Case 2 highlight the progression of SEC to sludge to thrombus formation in a patient with LAA and low emptying velocities

He underwent successful DCCV and his heart failure was well controlled for 2 years; however, he again returned to clinic with increasing fatigue and was found to be back in atrial fibrillation. His INR again was therapeutic and he underwent a TEE for possible DCCV.

His TEE this time showed again showed severe SEC in the LAA with a peak emptying velocity of 16 cm/s. There is no thrombus, but a layer of sludge in the LAA (Fig. 1.6e–h).

Given the presence of sludge, the elective DCCV was canceled and patient was maintained on anticoagulation but subsequently was admitted to the hospital 3 weeks later in decompensated heart failure. Following stabilization of his volume status, he underwent interval TEE to evaluate whether he could again be converted to sinus rhythm.

His interval TEE now showed the presence of a LAA thrombus despite documented adequate anticoagulation between the echocardiographic studies. His elective DCCV was canceled and he remained on anticoagulation (Fig. 1.6i–j).

Clinical Pearl

SEC is seen in 50 % of patients with AF and in 2 % of patients with NSR. Furthermore, SEC was present in 90 % of patients with LAA thrombus. In a subgroup analysis of the Stroke Prevention in Atrial Fibrillation (SPAF-III) trial, dense SEC was an independent predictor of thromboembolic events [122]. Furthermore, despite anticoagulant therapy, patients with AF and dense SEC have a high likelihood of cerebral embolism and/or death [123]. In a separate study, the presence of sludge has also been independently associated with thromboembolic events and all-cause mortality [92]. The decision to DCCV in the setting of dense SEC or sludge, however, is controversial and remains highly debated.

Case 3: NOAC Case (Fig. 1.7)

A 50-year old man with palpitations presented to clinic for evaluation. He has a past medical history for hypertension and diabetes. He denies any chest pain but describes an irregular heart rhythm most notable with activity but can occur at rest. He had a negative exercise stress test earlier in the year as part of an insurance physical exam. On physical exam, he has an irregular heart rhythm and no cardiac murmurs. An ECG confirms atrial fibrillation with a controlled rate. A TTE was performed that showed a mildly enlarged left atrium but no significant valvular abnormalities. He also had no metabolic dearrangement on laboratory testing, including normal renal function. He dislikes frequent blood tests; thus, after discussion of the risks and benefits with his physician, he is started on dabigatran for stroke prophylaxis. Following 3-weeks of anticoagulation, he presented for TEE-guided cardioversion for trial of sinus rhythm restoration.

Fig. 1.7

(a–c) NOACs (Case 3) (a–c) demonstrate a bi-lobed LAA with no SEC, no sludge, and no thrombus after 3 weeks of anticoagulation with dabigatran, a NOAC. (c) Shows high LAA emptying velocities in AF, 61 cm/s

His TEE revealed a bilobed LAA with no SEC, no sludge, and no thrombus. Doppler of the LAA orifice revealed a fibrillatory patter with high emptying velocities (61 cm/s). He underwent successful DCCV without any immediate complications.

Clinical Pearl

A number of new non-vitamin K oral anticoagulants (NOAC) have been developed and approved for thromboembolic prophylaxis. Direct thrombin inhibitor, dabigatran through the RE-LY trial [124], and factor Xa inhibitors: rivaroxaban through the ROCKET AF trial [125], apixaban through the ARISTOTLE trial [126], and edoxaban through the ENGAGE AF-TIMI 48 trial [127], have been shown to be non-inferior to conventional warfarin therapy for the prevention of embolic events in the setting of atrial fibrillation. Post-hoc, subgroup analyses of RE-LY, ROCKET AF, and ARISTOTLE, have similar shown at least non-inferiority of dabigatran [128], rivaroxaban [129], and apixaban [130] for embolic events compared to warfarin for subjects who underwent cardioversion. There is a prospective, randomized, open-label, blinded endpoint evaluation (PROBE) trial currently enrolling that compares edoxaban to warfarin in subjects undergoing DCCV for AF [131]. Given the lack of direct head-to-head comparisons of the NOACs, it is not clear that a particular NOAC is superior. There have been several meta-analyses pooling the data from the NOAC-trials that have shown similar or lower rate of embolic events, major bleeding, and/or all-cause mortality with NOACs compared to warfarin [132–137]. The results of these analyses and the convenience of no required monitoring (i.e. unlike blood draws for INR with warfarin) must be weighed against the lack of reversal agents and safety data to come with long-term use of these newer agents.

The current AHA/ACC/HRS guidelines recommend at least 3 weeks of anticoagulation before DCCV for AF ≥48 h in duration (or when duration is unknown); may proceed with DCCV sooner than 3 weeks of therapeutic anticoagulation if TEE is performed prior to DCCV to rule out LA thrombus [8]. In the cardioversion subgroup analyses of RE-LY (dabigatran), ROCKET-AF (rivaroxaban), and ARISTOTLE (apixaban), the patients had received long-term anti-coagulation (>3 weeks) at time of DCCV. In the Rivaroxaban vs. Vitamin K antagonists (VKAs) for Cardioversion in Atrial Fibrillation (X-VeRT) trial, the investigators prospectively evaluated the efficacy and safety of rivaroxaban versus VKAs undergoing early (1–5 days of anticoagulation) versus delayed (3–8 weeks of anticoagulation) DCCV, with or without TEE-guidance [138]. While underpowered for statistical significance, the study showed similar incidence of adverse events in those receiving rivaroxaban versus VKAs in both the early and delayed cardioversion arms, suggesting that early DCCV with rivaroxaban may be comparable to the standard VKA paradigm. The current AHA/ACC/HRS guidelines do not specify the type of anticoagulation needed if proceeding with TEE-guided DCCV prior to achieving 3 weeks of anticoagulation.

Case 4: Artifact (Fig. 1.8)

A 72 year old woman was referred for another opinion regarding a left atrial appendage thrombus. She has a past medical history significant for severe aortic stenosis, atrial fibrillation s/p DCCV in the past and on chronic anticoagulation, hypertension, hyperlipidemia, CAD, and a TIA. She is symptomatic from her severe AS with significant shortness of breath performing activities of daily living; however, due to her comorbidities, she was felt to be too high risk to undergo open heart surgery. She subsequently underwent a TEE for preoperative planning prior to possible transcatheter aortic valve replacement (TAVR) when she was reported to have a LAA thrombus. Her TAVR planning was halted, her anticoagulation was continued, and she was referred to another medical center for further evaluation.

Fig. 1.8

(a–h) Artifact (Case 4) (a–b) demonstrate an echodensity in the middle of the LAA orifice that is concerning for possible thrombus. Of note, the echodensity is two times the distance from the warfarin ridge. (c) Demonstrates that simultaneous multiplanar imaging through the echodensity does not reveal any density in the perpendicular plane. (d) Shows that color flow does not respect the echodensity seen on 2D imaging. (e) Shows another multiplanar image of the same LAA from different angles does not reveal the same echodensity. (f) Shows a typical quadriphasic LAA emptying pattern in sinus rhythm, although the late diastolic emptying velocity is low at 21 cm/s. (g–h) Show the use of echocontrast to confirm that the echodensity is reverberation artifact from the warfarin ridge due to the lack of a filling defect

She underwent a repeat TEE for further assessment of her left atrial appendage. An echodensity was visualized within the LAA between the membranous band of the left upper pulmonary vein and the LAA wall without any clear connection to the wall. It had a fully dependent motion with respect to the LAA and heart motion. Simultaneous multiplanar imaging, in addition to echocontrast, was utilized to confirm suspicions that the visualized density was an artifact from the warfarin ridge.

She underwent successful TAVR; of note, TEE at time of TAVR again confirmed the presence of a reverberation artifact within the LAA.

Clinical Pearl

Reverberation artifacts and pectinate muscles are often mistaken for LAA thrombi during TEE. In a series of patients undergoing TEE-guided DCCV, 23 % of patients had LAA artifacts. Reverberation artifacts from the warfarin ridge are located twice the distance from the transducer to the ridge, have a fully dependent motion with respect to the heart, and are morphologically consistent with the ridge. Thrombi have an independent motion, often-rounded morphology, attached to the LAA, often associated with SEC, and are distinct in echogenicity from the surrounding walls. Pectinate muscles are confined to the body of the LAA, have a fully dependent motion, and have the same echogenicity as the LAA walls. The use of echocontrast, simultaneous multiplanar imaging, and 3D techniques can help distinguish between true LAA thrombus, artifact, and intrinsic structures.

Case 5: LAA Exclusion (Fig. 1.9)

A 74 year old woman presented for TEE prior to anti-arrhythmic drug load (sotalol) for atrial fibrillation. She had a past medical history significant for hypertension, hyperlipidemia, interstitial lung disease, and atrial fibrillation. At the time of thoracic surgery for recurrent pleural effusions, her left atrial appendage was surgically oversewn with suture. She continued on anticoagulation therapy for stroke prophylaxis. She was admitted for anti-arrhythmic therapy due to symptomatic palpitations that was felt to be due to atrial fibrillation.

Fig. 1.9

(a–d) LAA surgical oversewing (Case 5) (a–b) demonstrate formation of a thrombus (marked with an arrow) within a LAA that has been surgical oversewn. (c–d) Demonstrate residual communication between the LAA and the LA. There is suggestion of thrombus formation in the LA where the LAA was surgical oversewn (marked with a + in a, c, d)

Her TEE showed a large mobile thrombus within the LAA with residual communication from the appendage to the LA. She was not started on anti-rhythmic therapy and a rate control strategy was adopted with continued anticoagulation.

Clinical Pearl

While the LAA is often surgically excised or excluded at time of concomitant thoracic surgery, multiple TEE studies have demonstrated flow between the LA and residual LAA [76–78]. A study performed at the Cleveland Clinic reviewing 137 patients who had LAA closure and subsequent TEE evaluation, showed that only 55 of 137 (40 %) of closures were successful and that LAA thrombus was found in 41 % of patients with patent communication between the LA and LAA. From the study, it appeared that of the various LAA interventional techniques, surgical excision was more successful than exclusion (suture or stapler) [77]. With the advent to newer percutaneous LAA closure techniques, TEE evaluation of residual flow and/or thrombus formation should not be overlooked.