In the year and a half after its approval the Watchman left atrial appendage closure device was successfully implanted 95.6% of the time, or more than 3,600 times out of a total of 3,822 consecutive cases.
The finding comes from procedural data collected by Boston Scientific and presented at TCT by Vivek Reddy (Mt. Sinai) and published simultaneously in the Journal of the American College of Cardiology.
The findings are reassuring, say the company and several cardiologists who have analyzed the data, but others point out that the data was not collected independently and does not look at long term effectiveness or safety.
Boston Scientific clinical specialists, who are present at every Watchman implantation, recorded procedural data on standardized forms. When the Watchman was approved in March 2015 there was no existing independent registry. In August 2016 the independent NCDR LAAO registry was established.
In addition to the 95.6% rate of successful implantation, Reddy and colleagues reported a median procedure time of 50 minutes. About half of the procedures were performed by physicians who were not involved in Watchman clinical trials. There were 39 cases of pericardial tamponade, three procedure-related strokes, 9 device embolizations, and 3 procedure-related deaths.
The authors concluded that “despite a large fraction of previously-inexperienced operators, in the real-world post-FDA approval experience of Watchman LAAC, procedural success was high and complication rates low.
In an accompanying editorial, Jacqueline Saw (Vancouver General Hospital) and Matthew Price (Scripps Clinic) write that the “results are remarkably favorable for the earliest phase of widespread dissemination of this technology.” But interpretation of the results is hindered by the study’s limitations, they argue. “This study is not a prospective registry with patient informed consent.” Further, important peri-procedural complications, including vascular and bleeding events which were included in previous trials, were not included in this study. In addition, they note, since the clinical specialist was only required to be present during the procedure, events such as late pericardial effusions or tamponade that occurred in the following week could easily have been missed.
The Ongoing Debate Over Watchman
The discussion over procedural safety overlooks a larger controversy about the efficacy of Watchman. Reddy et al write that the trials “have been concordant in demonstrating non-inferiority of Watchman to warfarin for stroke or systemic embolism, and superiority in reducing hemorrhagic stroke, cardiovascular mortality and non-procedure related bleeding.” But, according to Sanjay Kaul (Cedars-Sinai), the statement is incorrect. It is based on early published data that did not include complete patient followup of patients. The FDA review of the data concluded that only the safety endpoint was met in PREVAIL and both efficacy endpoints failed to meet noninferiority. “Collectively,” said Kaul, “the evidence supports reasonable assurance of safety but whether it meets reasonable assurance of efficacy is debatable. Hence, the FDA has required a comprehensive menu of post approval studies to clarify the benefit-risk profile of Watchman device.”
Sanket Dhruva (Yale University) said it was “too bad that Watchman dissemination was allowed to occur before implementation of the NCDR LAAO registry. While it’s great that device representatives are monitoring and reporting AEs, there is an inherent conflict of interest which is hard to disentangle. This reporting is not as reliable as an unbiased third party.”
Dhruva pointed out that the FDA’s Jeff Shuren recently stated that hospitals don’t always report adverse events and that other investigations have uncovered manufacturer delays in reporting AEs. “While I doubt such things are happening since the LAAO is under a microscope, we clearly need more rigorous registry data. This is all the more important for the Watchman device which was turned twice down twice by FDA because of various safety and efficacy concerns, and approved only on the third attempt despite an overall uncertain FDA advisory panel vote. Also, the registry focuses on just procedural success – we need longer-term real-world outcome data. Just placing a device with 95.6% success does not mean that it is safe and effective and helping patients.”
Dhruva also observed that in its original proposal CMS had called for a registry with a control arm but that this was eventually removed, thus “leaving patients and physicians with greater uncertainty about the device’s effectiveness.”