It is believed that 15 percent of all strokes and 40 percent of embolic strokes in the U.S. are the result of untreated atrial fibrillation (AFib). For many years, warfarin has been the primary therapy for stroke prevention in patients with AFib. However, it is estimated that 45 percent of patients eligible for anticoagulation are inadequately treated due to intolerance, poor adherence, or as a result of physician decision-making – a staggering number. With over 33.5 million individuals affected by AFib worldwide and its prevalence expected to rise as the proportion of individuals age >65 years increases, reduction of AFib-associated strokes is a major global health care concern.

Data suggest that the left atrial appendage (LAA) is the source of systemic thromboemboli in approximately 90 percent of AFib-related stroke cases. This has provided the mechanistic rationale for development of percutaneous technologies aimed at closure of the LAA in hopes of obviating need for anticoagulation. While currently there are three devices in clinical use for LAA closure, only one has been approved thus far for this indication by the U.S. Food and Drug Administration (FDA).

In January 2016, my first year of practice, I was approached by my department to assist in starting a new LAA closure program at our institution utilizing the WATCHMAN device (Boston Scientific, Marlborough, MA). Although it seemed like an intriguing opportunity and was presented by my employer as a potential part of our hospital's growth strategy, I was certainly aware of the device's long road before its eventual approval by the FDA.

For those in need of a quick recap: the first randomized control trial (RCT), PROTECT-AF, randomized non-valvular AF patients suitable for long-term oral anticoagulation to either the WATCHMAN device or warfarin treatment in 2:1 fashion. The trial's enrollment started in 2005, prior to the era of novel oral anticoagulants (NOACs), and met the non-inferior hypothesis for the primary composite efficacy endpoint, but failed to meet the safety endpoint with a complication rate of nearly 8 percent. It was troubling that the rate of ischemic strokes was significantly higher in the WATCHMAN arm, leading many to question the adequacy of LAA occlusion even if technically successful.

The FDA mandated a second RCT be performed called PREVAIL, which was Bayesian in design. While the study population was smaller than PROTECT-AF, the mean CHADS2-VASc scores were higher (3.8 vs. 3.2 percent). The trial failed to meet the primary efficacy endpoint for non-inferiority, notable for higher-than-usual efficacy and performance of the warfarin group. A key finding of the trial was a substantial decrease in the peri-procedural complication rate (2.2 percent) despite an inclusion of new operators suggesting improvement in procedural operator learning curve.

The device was approved on its third review by the FDA; however, approval was modeled after indications used in the RCTs, namely restricting use only to patients eligible for long-term anticoagulation, limiting the device's clinical scope. In February 2016, the Centers for Medicare and Medicaid (CMS) made WATCHMAN the first reimbursable LAAC device, but made the device implantation accessible to patients not suitable for long-term anticoagulation. Other requirements by CMS included documentation of shared decision-making and participation in a national registry database for each case.

So where do things stand now? A quick search of the Internet will show that even with expert consensus guidance, the role of the WATCHMAN device in clinical care remains a controversial topic and seems likely to remain an impediment to its widespread adoption. While pertinent questions remain such as how the Watchman device would fair against NOACs, no plans for another large RCT are on the horizon (at least with this particular device).

Although RCTs are considered the gold standard for testing treatment efficacy, they are often limited in their generalizability to real practice. For now, we'll have to be content with observational longitudinal data in the form of industry-collected post-trial and commercial registries. Results from the international registry EWOLUTION appear reassuring from a procedural success and operator safety standpoint, and notably 72 percent of patients for whom the device was implanted were contraindicated to take anticoagulation. Caution should be exercised in interpreting long-term efficacy data from these registries due to their observational nature.

Ultimately, we have decided to move forward with starting a LAAC program at our institution based on several factors: 1) clinical need based upon a relative high number of patients with contraindications for anticoagulation, 2) commitment from our hospital administration for capital support and additional staff to support the program, and 3) interest from other members of our institution and potential referrers in the community. While there have certainly been the expected challenges, overall the program has been beneficial and has allowed delivery of better care to our patients.

This article was authored by Henry D. Huang, MD, FACC, clinical assistant professor at the University of Illinois Chicago, and director of cardiac electrophysiology at the Jesse Brown VA Medical Center, Chicago, IL.