Mid-Term Outcomes Following Implantation of Watchman or Amulet: A Meta-Analysis of Observational Studies
Arav Jhand, MBBS, Mayo Clinic, Rochester, MN
Arav Jhand, MBBS1, Arslan Ahmed, MD2, Rahul Dhawan, MD3, Jacqueline Saw, M.D., FSCAI4, Mohamad Alkhouli, M.D., FSCAI5, Matthew James Daniels, M.D., Ph.D., FSCAI6, Megan Coylewright, M.D., FSCAI7, Mauricio G. Cohen, M.D., FSCAI8, John W Schleifer, M.D.3 and Dr. Andrew Michael Goldsweig, MD, MS, FSCAI9, (1)Mayo Clinic, Rochester, MN, (2)Creighton University School of Medicine, Omaha, NE, (3)University of Nebraska Medical Center, Omaha, NE, (4)Vancouver General Hospital, Vancouver, BC, Canada, (5)Mayo Clinic - School of Medicine, Rochester, MN, (6)Manchester University Foundation NHS Trust, Altrincham, Manchester, United Kingdom, (7)Erlanger Health System, Chattanooga, TN, (8)Cleveland Clinic Florida, Pinecrest, FL, (9)University of Massachusetts - Baystate Medical Center, West Hartford, CT
Keywords: Left Atrial Appendage Closure (LAAC) and Structural Heart Disease (SHD)
Background:
Watchman (WM) is the only left atrial appendage occlusion (LAAO) device approved in the United States. We conducted a meta-analysis of observational studies comparing mid-term outcomes of WM vs. Amulet (Amplatzer Cardiac Plug), the second most popular device worldwide.
Methods:
PubMed, Scopus and Google Scholar were searched systematically for studies comparing >3-month outcomes of WM and Amulet. Patient characteristics and outcomes of study-defined severe peri-device leak (PDL), device-related thrombosis (DRT), stroke, death and major bleeding were collected. Imaging for PDL/DRT was performed at 2-6 months post-implant. Risk ratios (RRs) and 95% confidence intervals (CIs) were calculated using a random effects model.
Results:
11 studies with 2145 patients (WM 970, Amulet 1175) met inclusion criteria. Baseline characteristics including age, sex, CHA2DS2VASc (WM 4.2 ± 0.5, Amulet 4.2 ± 0.4, p=0.91) and HAS-BLED scores (WM 3.2 ± 0.4, Amulet 3.3 ± 0.5, p=0.86) were comparable (Table 1). Mean follow-up duration was 13.4 ± 7.3 months. WM was associated with a higher risk of severe PDL (RR: 2.80, 95% CI 1.16 - 6.71, p=0.02). DRT, stroke, death and major bleeding were similar between the 2 devices.
Conclusions:
WM was associated with a higher incidence of severe PDL than Amulet without significant differences in clinical events at mid-term follow-up.
Table 1:
| Patient Characteristics |
| | Watchman (n = 970) | Amulet (n = 1175) | p-value |
| Age (mean ± SD), years | 74.2 ± 4.9 | 74.1 ± 5.3 | 0.96 |
| Female Sex, n (%) | 361 (37%) | 440 (37%) | 0.66 |
| CHA2DS2VASc Score (mean ± SD) | 4.2 ± 0.5 | 4.2 ± 0.4 | 0.91 |
| HAS-BLED Score (mean ± SD) | 3.2 ± 0.4 | 3.3 ± 0.5 | 0.86 |
| Procedural Outcomes |
| Outcome | Number of Studies Reporting the Outcome | Watchman | Amulet | RR | 95% CI | p-value |
| Severe Peri-Device Leak* | 7/11 | 3.5 % | 1.3 % | 2.80 | 1.16 – 6.71 | 0.02 |
| Device-Related Thrombosis | 9/11 | 3.1 % | 1.0 % | 2.40 | 0.94 – 6.15 | 0.07 |
| Stroke | 11/11 | 1.7 % | 2.2 % | 0.82 | 0.42 – 1.61 | 0.57 |
| Death | 10/11 | 6.3 % | 9.1 % | 0.80 | 0.59 – 1.08 | 0.14 |
| Major Bleeding | 11/11 | 2.5 % | 2.6 % | 0.79 | 0.47 – 1.34 | 0.39 |
*Severe Peri-Device Leak was defined as > 3 mm in 4 out of 7 studies (Cruz-Gonzalez et al, Figini et al, Kim et al and Chen et al) and >5 mm in 3 studies (Fastner et al, Cheung et al and Davtyan et al); CI = confidence interval, RR = risk ratio, SD = standard deviation.