Intradevice Leak on Late Follow-Up after Watchman Implantation.

Introduction

Currently the Watchman device (Boston Scientific, Natick, MA) is approved for use in patients with nonvalvular atrial fibrillation at increased risk for stroke on the basis of CHADS2 or CHA2DS2-VASc score, suitable for warfarin, and in whom there is justification to use the device as an alternative to warfarin when weighing the safety and effectiveness of the two treatments. Although small peridevice leaks are commonly noted, intradevice leaks are rare and were not reported in the initial trials.2, 3, 4

Case Presentation

The patient was an 88-year-old man who underwent left atrial appendage closure with a Watchman device for atrial fibrillation in the setting of a CHA2DS2-VASc score of 5 and had an increased bleeding risk from frequent falls due to peripheral neuropathy. He also had a medical history significant for coronary artery disease with prior coronary artery bypass graft surgery, ischemic cardiomyopathy, diabetes mellitus, and complete heart block with biventricular implantable cardioverter defibrillator placement. At the time of Watchman implantation and on 45-day transesophageal echocardiography (TEE), the device was well seated, without peri- or intradevice leak in any of the standard views obtained from the 0°, 45°, 90°, and 135° angles at the midesophageal level (Figure 1, Figure 2, Figure 3, Figure 4). Immediately following device implantation, the patient was continued on anticoagulation with apixaban. According to the protocol used in the Watchman trials, anticoagulation, in this case apixaban, was stopped after 45-day TEE, and clopidogrel was started and taken with aspirin for 6 months. After 6 months, the patient was continued on aspirin 81 mg/day indefinitely. One-year follow-up TEE demonstrated a 2-mm intradevice leak (Figure 5, Figure 6, Figure 7, Figure 8, Figure 9, Figure 10, Videos 1–4). Because of the small leak size and the patient's increased risk for bleeding because of frequent falls, anticoagulation was not resumed, and he has been maintained on low-dose aspirin alone without bleeding or neurologic complications.

Figure 1

At the time of Watchman implantation, transesophageal echocardiogram imaging at 0⁰ with two-dimensional (2D) images with and without color flow in the right and left panels, respectively, during systole without evidence of high-velocity flow either around or through the Watchman device. The Watchman struts (arrows) are shown to be well opposed to the wall of the left atrial appendage (LAA). CR, Coumadin ridge; LA, left atrium; PV, pulmonary vein.

Figure 2

At the time of Watchman implantation, transesophageal echocardiogram imaging at ∼135 (146)⁰ with two-dimensional (2D) images with and without color flow in the right and left panels, respectively, during diastole without evidence of high-velocity flow either around or through the Watchman device. In this case, the image was obtained prior to release of the device. Again, the Watchman struts (arrows) are shown to be well opposed to the wall of the left atrial appendage. ∗, Delivery system; CR, Coumadin ridge; LA, left atrium.

Figure 3

At 45-day follow up after Watchman implantation, transesophageal echocardiogram imaging at ∼45 (62)⁰ with two-dimensional (2D) images with color flow, during systole without evidence of high-velocity flow either around or through the Watchman device. The Watchman struts (arrows) are shown to be well opposed to the wall of the left atrial appendage. CR, Coumadin ridge; LA, left atrium; LV, left ventricle; MV, mitral valve.

Figure 4

At 45-day follow up after Watchman implantation, transesophageal echocardiogram imaging at ∼90 (94)⁰ with two-dimensional (2D) images with color flow, during diastole without evidence of high-velocity flow either around or through the Watchman device. The Watchman device is partially enface obscuring the struts adjacent to the Coumadin ridge, otherwise the Watchman struts (arrows) are shown to be well opposed to the wall of the left atrial appendage. CR, Coumadin ridge.

Figure 5

At 1-year follow up after Watchman implantation, transesophageal echocardiogram imaging at 0⁰ with two-dimensional (2D) images with color flow, during diastole (A) and systole (B). During diastole (A), there is high-velocity flow (white arrow) seen entering the Watchman device inside of the struts (yellow arrows). During systole (B), there is high-velocity flow (white arrow) seen exiting the Watchman device from inside of the struts (yellow arrows). CR, Coumadin ridge; LA, left atrium; LAA, left atrial appendage; PV, pulmonary veins.

Figure 6

At 1-year follow up after Watchman implantation, transesophageal echocardiogram imaging at ∼45 (40)⁰ with two-dimensional (2D) images with color flow, during diastole (A) and systole (B). During diastole (A) there is high-velocity flow (white arrow) seen entering the Watchman device inside of the struts (yellow arrows) with a PISA visible outside of the device. During systole (B) there is high-velocity flow (white arrow) seen exiting the Watchman device from inside of the struts (yellow arrows) with a PISA visible within the device. CR, Coumadin ridge; LA, left atrium; LV, left ventricle; MV, mitral valve; PV, pulmonary vein.

Figure 7

At 1-year follow up after Watchman implantation, transesophageal echocardiogram imaging at ∼90 (89)⁰ with two-dimensional (2D) images with color flow, during diastole (A) and systole (B). During diastole (A), a high-velocity jet (white arrow) is seen in the anterior aspect of the Watchman device within the struts (yellow arrows), which is not seen during systole as the jet has moved out of the imaging plane (B). CR, Coumadin ridge; PV, pulmonary vein.

Figure 8

At 1-year follow up after Watchman implantation, transesophageal echocardiogram imaging at ∼135 (133)⁰ with two-dimensional (2D) images with color flow, during diastole (A) and systole (B). During systole (B), a high-velocity jet (white arrow) is seen within the Watchman struts (yellow arrows), which is not seen during diastole as the jet has moved out of the imaging plane (A). CR, Coumadin ridge.

Figure 9

At 1-year follow up after Watchman implantation, transesophageal echocardiogram three-dimensional imaging with color flow during systole demonstrating a high-velocity jet (arrow) exiting the Watchman device from the surface of the device. CR, Coumadin ridge; MV, mitral valve.

Figure 10

At 1-year follow up after Watchman implantation, transesophageal echocardiogram three-dimensional imaging with color flow and multiplanar reconstruction during systole demonstrating a high-velocity jet (arrow) exiting the Watchman device from the surface of the device. CR, Coumadin ridge.

Discussion

Initial reports of minimal peridevice leaks (jet width ≤ 5 mm) 1 year after Watchman implantation were as high as 32% in PROTECT-AF, but more recent studies (PREVAIL and EWOLUTION) reported <10% incidence3, 4; however, these leaks were around, not through, the device. Additionally, the absence or presence of only a minimal peridevice leak of jet width ≤ 5 mm was considered part of the initial definition of procedural success leading to release of the device.2, 3 These leaks are often small and, as seen in the figures and videos in this case, may move in and out of the two-dimensional imaging plane during the cardiac cycle; therefore, it is imperative that the entire device be examined throughout the cardiac cycle from multiple angles. Further studies showed that minimal peridevice leaks have not been associated with increased risk for stroke or systemic embolism and that there was no significant difference in risk for thromboembolism on the basis of the size of the minimal defect. Subsequent evaluation of peridevice leaks in this study found that these leaks either remained the same size or decreased in size, in some cases resolving, over time. In the trials evaluating the Watchman device, warfarin was studied as the anticoagulant after implantation2, 3, 4; however, a recent study demonstrated that there was no difference in bleeding or thromboembolic events for patients on nonwarfarin oral anticoagulants, such as apixaban, compared with warfarin during Watchman implantation. Intradevice leaks were not commented on in the original clinical trials, and thus their clinical implications are unknown.2, 3, 4 Proposed mechanisms of intradevice leak development include fabric fatigue, nitinol strut fracture, or damage to the device's fabric membrane during device manipulation or deployment, which became apparent with improved atrial appendage contractility or growth in the size of the defect over time.

Conclusion

It is important for clinicians to recognize that intradevice leaks may occur late after device implantation, even when previously absent at the time of deployment or early 45-day follow-up. For this reason, though this is an exceedingly rare event, it is important for TEE operators to recognize that device leaks may be intradevice in addition to peridevice and therefore require thorough imaging from all angles throughout the cardiac cycle. At this time the clinical implications of an intradevice are not clear.