On the basis of outcomes, there was no clear winner from a trial that directly compared two modern devices used in patients undergoing percutaneous left atrial appendage (LAA) closure.
But the devices were not interchangeable for rates of complications or leaks, according to results of the open-label SWISS APERO trial, which compared the Amplatzer Amulet to the Watchman FLX device at eight participating centers in Europe.
At 45 days, the overall rates of leaks and the clinical outcomes in the two randomized groups were not significantly different, but there were differences in secondary endpoints, such as rates of peridevice leak (PDL), which were lower in the Amulet device group, and procedural complications, which were higher, Roberto Galea, MD, reported at the Transcatheter Cardiovascular Therapeutics annual meeting, held virtually and live in Orlando.
LAA closure devices were developed as an alternative to oral anticoagulation in patients with nonvalvular atrial fibrillation. Although a similar comparison of LAA closure devices, called Amulet IDE, was recently published, that trial compared Amulet to Watchman 2.5, an earlier generation device.
Started later, SWISS APERO was also a planned comparison of Amulet and the Watchman 2.5, but the comparison switched to the Watchman FLX, when it was released in March of 2019.
First randomized comparison with Watchman FLX
“This is the first multicenter randomized controlled trial to include the Watchman FLX,” said Dr. Galea, a clinical investigator in the department of cardiology, Bern (Switzerland) University Hospital. He noted that Watchman FLX included some adjustments in design with the potential to reduce leak rates.
After preprocedural transesophageal echocardiography confirmed that patients had suitable anatomy to receive either device, the 221 patients who qualified for SWISS APERO were randomized. The primary endpoint was a composite of a justified crossover to a device other than the one to which they were assigned or residual patency detected by coronary computed tomography angiography (CCTA) at 45 days.
The primary endpoint was reached by 67.6% of patients randomized to the Amulet device and 70% of those randomized to Watchman Flex, a statistically nonsignificant difference (P = .71).
Because only one patient in the Amulet group and none in the Watchman group had a justified crossover to a nonrandomized device, most of the differences in the 45-day CCTA involved patency, defined as LAA density of at least 100 Hounsfield units. While the proportion of patients with leaks was similar, the types of leaks, which were stratified by underlying leak mechanism into PDL, mixed leaks (including incomplete side sealing), intradevice leaks, and leaks of unclear origin, were different.
Peridevice leaks twofold greater with Watchman
Those randomized to the Watchman device were more than twice as likely to have PDL (27.5% vs. 13.7%; P = .02), although no visible leak exceeded 5 mm in size. They were also more likely to have mixed leaks (14% vs. 3.8%; P = .01) and patency with no visible leak (21.0% vs. 9.5%; P = .02). There were also more device-related thrombi in the Watchman group even though the difference did not reach statistical significance (9.9% vs. 3.7%; P = .08).
Intradevice leaks (44.8% vs. 23.0%; P = .001) were the only type of patency significantly more common among patients randomized to Amulet, but the difference was relatively large. In addition, procedural complications of any type (32.4% vs. 19.1%; P = .023) were higher in the Amulet group. Most of these involved non–clinically relevant pericardial effusions, Dr. Galea said at the meeting, sponsored by the Cardiovascular Research Foundation.
The proportion of patients with adverse outcomes by 45 days was similar, but the types of complications differed. Of the six deaths, two occurred in the Amulet group as a result of periprocedural complications (one stemming from an air embolism and the other from a series of events following pericardial effusion). Three of the four deaths in the Watchman group were due to fatal bleeding. The fourth was a sudden death that occurred 30 days after the procedure.
Amulet IDE trial generates similar data
The much larger Amulet IDE trial, which compared Amulet to the Watchman 2.5 device, produced generally similar results. Again, the proportion of patients reaching the composite primary endpoints was similar.
The primary safety endpoint, which included death and major bleeding within 12 months of randomization, occurred in 14.5% and 14.7% of the Amulet and Watchman patients, respectively (P < .001 for noninferiority). The primary efficacy endpoint, which included stroke or systemic embolism within 18 months of randomization, occurred in 2.8% of patients in both groups.
As in SWISS APERO, the 1,878-patient Amulet IDE trial showed that the devices are similarly effective and safe but not necessarily interchangeable. Ultimately, the rate of LAA occlusion was higher for Amulet than the older generation Watchman (98.9% vs. 96.8%; P = .003) but procedural complication occurred more frequently among those randomized to the Amulet device (4.5% vs. 2.5%).
“The closure mechanisms are not the same, which might explain why we see differences in some secondary outcomes even when they perform similarly on the primary outcomes,” said Dhanunjaya R. Lakkireddy, MD, executive medical director, Kansas City (Kansas) Heart Rhythm Institute.
Dr. Dhanunjaya R. Lakkireddy
The lead investigator of the Amulet IDE trial, Dr. Lakkireddy was referring to both the AMULET IDE and the SWISS APERO study when he said that the currently available data do not allow one device to be considered superior to the other. He did suggest that differences between devices might still be considered meaningful in specific clinical situations or to specific clinicians.
Without studies to show objective differences, Dr. Lakkireddy suggested that training and experience is probably the most important variable in achieving treatment goals. “Operator comfort is certainly important,” he said.
Dr. Galea reports no significant financial relationships. The investigator-initiated study received funding from Abbott, the manufacturer of the Amulet device. Dr. Lakkireddy has financial relationships with Abbott, AltaThera, Medtronic, Biotronik, and Boston Scientific, which makes the Watchman device.