Complications Spiked, Then Fell, After Initial Introduction of Watchman LAA Closure Device
All device manufacturers need to focus on a tightly controlled rollout when moving from clinical trials to real-world use, a researcher says.
When the Watchman left atrial appendage (LAA) closure device (Boston Scientific) came on the market, the number of complications reported to the US Food and Drug Administration (FDA) shot up initially but then started to come back down, a new analysis shows.
The lesson, according to senior author Dhanunjaya Lakkireddy, MD (University of Kansas Hospital & Medical Center, Kansas City), is not that there is a problem with the Watchman device itself, but that there is an adjustment period for technology in general when moving from clinical trials to real-world use.
“I was not at all surprised and, in fact, [the findings] actually fit in very well with our hypothesis that when you really roll out something new the learning curve oftentimes impacts the safety performance of most of these devices in the real world,” Lakkireddy told TCTMD.
He noted that his team has since extended follow-up from the current study, finding that the number of complications reported to FDA has continued to decline after the initial spike.
“That’s a very, very encouraging sign. That tells you [about] the natural adjustment that typically happens as people gain more experience and become more comfortable with the procedure,” Lakkireddy said. “I think this is a universal message to all the device manufacturers to focus on a very [tightly controlled] rollout plan in the transition from a trial to routine clinical use in the community.”
Christoph Nienaber, MD, PhD (Royal Brompton & Harefield NHS Foundation Trust, London, England), who was not involved in the study, also said he was not surprised by the sharp increase in complications reported shortly after Watchman was approved in March 2015 and agreed that the issue was more about how new devices are introduced to the medical community and less about problems with any particular device.
“Once a new device or a new technology is available to anybody who considers himself an interventional cardiologist or a cardiologist with skills in catheter technique, the complication rate initially goes up,” he said. “That is only an issue related to learning curve and to new operators that are not completely familiar with the device yet. It’s a typical finding that you can see in other technologies as well.”
With improved training and more careful selection of operators and centers during the initial introduction, “this steep rise would probably not occur,” he added. “Industry should be more careful and not sell these new devices to everybody right away. Proper training and a very careful market release is, I think, a better option.”
Safety of LAA Closure
Though oral anticoagulation is the standard approach to managing stroke risk in patients with nonvalvular A-fib, some patients have contraindications to or are otherwise not good candidates for the therapy.
LAA closure devices emerged as an alternative for that group, mimicking surgical exclusion, which is regularly performed—at least by some surgeons—as an add-on procedure during cardiac surgery. There are currently two devices available in the United States used for excluding the LAA: the Lariat suture device (SentreHEART), which has 510(k) approval for soft tissue ligation, and the Watchman device, which was cleared through the premarket approval pathway for patients with A-fib who have an increased stroke risk and are deemed suitable for warfarin, but for whom there is an appropriate reason to seek an alternative to warfarin.
Evidence supporting either percutaneous or surgical exclusion of the LAA, however, has come under fire in the past, with critics decrying the dearth of randomized controlled trials that unequivocally support the techniques. Of the two trials on which FDA clearance was based, the first—PROTECT-AF—met its primary noninferiority endpoint for efficacy, but not for safety, while PREVAIL met just two of its three primary endpoints.
In the current study, published online October 8, 2017, ahead of print in the Journal of Cardiovascular Electrophysiology, Lakkireddy, lead author Mohammad-Ali Jazayeri, MD (University of Kansas Hospital & Medical Center), and colleagues evaluated the safety of Lariat and Watchman using the FDA’s Manufacturer and User Facility Device Experience (MAUDE) database. They examined reports submitted between May 1, 2006, and May 1, 2016. There were 167 reports related to Lariat and 445 to Watchman, but the analysis was restricted to 356 unique and relevant reports with adequate data.
The number of total procedures performed during the study period was obtained from the manufacturer in the case of Lariat (4,889) and from trial and registry populations in the case of Watchman (2,027 prior to FDA approval and 3,822 in the postmarketing experience).
The cumulative rate of complications received by the FDA was 2.7% with Lariat and 8.1% with Watchman. The most frequent complications with the former device were pericardial effusion (0.94%), need for cardiac surgery (0.78%), and pericardiocentesis (0.47%) and with the latter device were malfunction (1.7%), pericardial effusion (1.4%), pericardiocentesis (0.97%), and intracardiac thrombus (0.84%).
Comparing the two devices, the rate of major complications (stroke/TIA/ pericardiocentesis, need for cardiac surgery, or death) was higher with Watchman (1.93% vs 1.15%; P = 0.001), mostly due to differences in stroke/TIA and pericardiocentesis. There was no significant difference in mortality.
Looking only at the Watchman device, there was an increase in the rates of various complications from the preapproval to the postmarketing period, characterized by an initial large increase followed by declining rates.
Reports of complications also increased for the Lariat device, with a peak in 2013 followed by a gradual decline. Lakkireddy attributed that trend to refinements in technique, which included use of a micropuncture needle for pericardial access and periprocedural colchicine to limit pericardial effusions.
More Careful Real-World Introduction Important
Both Lakkireddy and Nienaber underscored the necessity of introducing new technologies to the medical community in a controlled fashion, and Lakkireddy said industry has already gotten better about doing just that.
Careful introduction of new devices should include, Lakkireddy said, an honest assessment of the required skill set, selection of well-equipped operators and centers for the initial rollout, and a mandate for robust training that enhances understanding of anatomy and imaging techniques and includes exposure to procedural videos of complex cases and discussion with experienced operators.
“How do we really make the learning experience quicker and better?” he said. “I think that’s where all these device companies should spend some time and resources, and FDA should actually mandate an action plan from them that is more robust and foolproof.”
Lakkireddy et al also point out that improved methods for uncovering safety signals with new devices are needed. They cite several limitations of the MAUDE database, including the fact that reporting is not mandatory for healthcare professionals, resulting in underreporting of potential issues.
“Passive surveillance mechanisms such as the FDA MAUDE database are inadequate for nimble, effective postapproval monitoring,” they write. “The process stands to be greatly improved with the advent of more rigorous postapproval studies, registry participation, operator training, and the integration of multiple streams of relevant data to improve detection of important safety signals and facilitate a more robust feedback mechanism.”
Photo Credit: Boston Scientific. Adapted from http://www.watchman.com/hcp/home.html
Jazayeri M-A, Vuddanda V, Turagam MK, et al. Safety profiles of percutaneous left atrial appendage closure devices: an analysis of the Food and Drug Administration Manufacturer and User Facility Device Experience (MAUDE) database from 2009-2016. J Cardiovasc Electrophysiol. 2017;Epub ahead of print.
- Jazayeri, Lakkireddy, and Nienaber report no relevant conflicts of interest.