Ongoing Experience Improves Safety for LAA Closure with Watchman Device

Percutaneous closure of the left atrial appendage (LAA) with the Watchman device in patients with atrial fibrillation (A-fib) is becoming a much safer procedure as operator experience increases, according to a comparison of clinical trial and registry data published online January 17, 2011, ahead of print in Circulation.

Building the Case for Safety

The PROTECT AF trial, published in August 2009 in the Lancet, randomized patients with nonvalvular A-fib to percutaneous LAA closure with the Watchman device (Atritech, Plymouth, MN; n = 463) or ongoing warfarin treatment alone (n = 244). Although the device was found to be noninferior to warfarin in terms of preventing stroke, systemic embolism, and cardiovascular death, its use was accompanied by a significantly higher risk of complications, mostly pericardial effusion and procedural stroke.

A US Food and Drug Administration (FDA) advisory panel initially recommended approval of the Watchman after findings from PROTECT AF were presented in March 2009 at the American College of Cardiology Annual Scientific Session/i2 Summit; but 1 year later, the agency asked Atritech to confirm the device’s safety and efficacy by conducting a new randomized trial.

In the midst of this discussion, Vivek Y. Reddy, MD, of the Mount Sinai School of Medicine (New York, NY), and colleagues performed an analysis to assess how the Watchman’s safety might have changed over time. They looked at data from 2 sources:

  • The device arm of the PROTECT AF trial (n = 542 attempted implantations)
  • Patients enrolled in the subsequent CAP (Continued Access Protocol) registry from August 2008 to April 2010 (n = 460 attempted implantations)

Persistence Pays Off

At baseline, the PROTECT AF patients were younger (72 ± 9 years vs. 74 ± 8 years) and had lower CHADS2 scores (2.2 ± 1.2 vs. 2.4 ± 1.2) than the registry patients (P < 0.001 for both comparisons).

Procedure time, implant success, and safety outcomes all were significantly better in the CAP registry than in PROTECT AF trial. The primary composite safety endpoint included both procedure-related safety events (pericardial effusion/tamponade, stroke, or device embolization) and excessive bleeding events (intracranial or gastrointestinal bleeding requiring transfusion; table 1).

Table 1. Outcomes in PROTECT AF and CAP Registry


(n = 542)

(n = 460)

P Value

Procedure Time, min

62 ± 34

50 ± 21

< 0.001

Implant Success




Composite Safety Endpoint ≤ 7 Days




Serious Pericardial Effusion ≤ 7 Days




Procedure-Related Stroke




To address the fact that “not all safety events have the same clinical impact on the patient,” the investigators conducted a separate analysis of Watchman and warfarin patients from PROTECT AF. Subjects in the trial’s device arm were significantly less likely to experience a safety event that resulted in significant disability (defined as an increase in the modified Rankin score) or death compared with those in the drug arm. The incidence of safety events that had such effects ranged from 1.4 to 1.8% with Watchman and 3.3 to 4.3% with warfarin.

Sheath Management, Vigilance Key

Two main improvements occurred over the course of the Watchman experience, Dr. Reddy told TCTMD in a telephone interview.

First, sheath management was optimized, leading to a marked reduction in procedure-related strokes, which were “a huge issue” in PROTECT AF, he said. “That was actually the difference between the Watchman proving superior vs. what it turned out to be, which is simply noninferior.”

In addition, the rate of tamponade—which can arise from a variety of sources including initial transseptal puncture and device deployment—decreased with experience. Although that complication still needs to be addressed, Dr. Reddy noted, “The good news is that in both of the studies, the functional significance of the tamponade ended up not being very important. The reason is that the procedure is done under [transesophageal echocardiographic] guidance, so you’re constantly looking for complications, which then allows you to act quickly and deal with them.”

Speaking in telephone interviews with TCTMD, Robert J. Sommer, MD, of Columbia University Medical Center (New York, NY), and Ziyad M. Hijazi, MD, MPH, of Rush University Medical Center (Chicago, IL), both agreed that these improvements are noteworthy.

In particular, the transseptal puncture technique can be challenging to new operators, said Dr. Hijazi, adding, “I think once you select investigators who are savvy in transseptal puncture techniques, your complication rate should be low.” Dr. Reddy pointed out that the newer iterations of the Watchman “are less traumatic as you implant them in the appendage and have less chance of causing a tear or effusion/tamponade.”

Big Picture Is Looking Good

All 3 physicians said that the learning curve with the Watchman is to be expected, given that it is a first-in-class device.

“The data are basically comparable to what we do with devices all the time,” Dr. Hijazi said. “If you are a good doctor, the more you do it, the better you get at it. And this group of doctors is good.”

Dr. Hijazi pointed out 2 drawbacks to the current analysis: The CAP registry does not include all investigators from PROTECT AF, so it is a select group, and no new investigators were added during the registry study. He suggested that the data could be analyzed in a different way, documenting the outcomes of individual researchers as they proceeded through both the randomized trial and registry phases. “So the question is,” he asked, “when the device gets approved, are we going to go through the same commotion with a new learning curve or are we going to be able to eliminate [that process]?”

PREVAIL, the randomized trial recently begun at the FDA’s request, aims to answer this question, said Dr. Sommer, who is serving as an investigator for the study at Columbia. The study will only involve participating centers that were not part of PROTECT AF.

“One of the things the FDA wanted to see is, with the modifications to the device and the delivery system, whether or not your average cath lab could perform this procedure,” Dr. Sommer commented. “Yes, a learning curve is inherent in every procedure, but if the FDA approves this, everybody and his uncle are going to have access to these devices. What are the complication rates going to be like if you open it up to an uninitiated population of physicians? And what kind of training is going to be required? I think those are all critical questions.”

Dr. Sommer also reported that AGA Medical (Plymouth, MN) has a new device for LAA closure called the Amplatzer Cardiac Plug. The company is just finishing up a feasibility trial in the United States and will soon start a nationwide pivotal trial, he said.

Percutaneous treatment of A-fib is an evolving field, Dr. Sommer observed. “Over time, there are going to be a lot of smart people out there coming up with new ways to trap the mouse, so to speak, and there will be improvements in the technology, in the delivery systems, in everything,” he noted. “But the important thing is that we’re proving with these data that the concept of closing the left atrial appendage is one that has merit.”

Dr. Hijazi expressed similar enthusiasm. “I think this is a phenomenal time now, because we really need solutions for these patients with A-fib who do not tolerate [warfarin] or even patients who do,” he said, relating his own experience treating a 43-year-old woman taking warfarin for A-fib. “She fell at home and injured her head. She had a brain bleed that required multiple surgeries, and within a week she died. . . . She was otherwise doing extremely well.

“So if there is a way that we can stop all these medications that have hazards like this and perform a safe procedure, then everybody should be for this technology,” he said.


Reddy VY, Holmes D, Doshi SK, et al. Safety of the percutaneous left atrial appendage closure: Results from the Watchman left atrial appendage system for embolic protection in patients with AF (PROTECT AF) clinical trial and the Continued Access registry. Circulation. 2011;123:417-424.



Related Stories:

  • Both the PROTECT AF trial and the CAP registry were supported by Atritech.
  • Dr. Reddy reports having received research grant support from Atritech.
  • Drs. Hijazi and Sommer report no relevant conflicts of interest.