Watchman Thrombi 'Not Infrequent,' May Cause Strokes

  Watchman device-related thrombi developed in 3.7% of patients and appeared temporally-related to strokes, meta-analysis of clinical trials and registries showed.

The annualized stroke rate across the 1,739 patients in the PROTECT-AF and PREVAIL trials and two continued access program registries was 1.77%, among which 10.2% occurred in patients found to have device-related thrombi (DRT), Vivek Reddy, MD, of Icahn School of Medicine at Mount Sinai in New York City, reported here at the Heart Rhythm Society (HRS) meeting and online in Circulation.

The expected ischemic stroke rate for the same CHA2DS2-VASc score, based on the literature, would be 6.0% for untreated patients and 1.71% on oral anticoagulants, Reddy noted at the late-breaking clinical trial session.

Antithrombotic use data for the Watchman cohorts was not known, although the FDA trials recommended warfarin and aspirin for the first 6 weeks, followed by clopidogrel (Plavix) and aspirin to 6 months and aspirin alone thereafter.

Clots can form on any foreign body, and Watchman studies have shown only thin endothelialization at 200 days after implantation and the metal screw still visible in cadaver specimens even at 852 days, said HRS discussant David De Lurgio, MD, of Emory Saint Joseph's Hospital in Atlanta, who called that "kind of scary."

A timeline analysis trying to get at causation showed that about half of strokes in Watchman patients did co-localize within 1 month of finding thrombus on the device, "which suggests at least half of patients' DRT is causing stroke," although other stroke risk factors are playing a role too, Reddy said.

When DRT was found on study-mandated repeat transesophageal echocardiography (TEE), the risk of stroke or systemic embolism was 3.55-fold elevated overall and 3.2-fold elevated for ischemic stroke specifically. However, no significant association with mortality emerged.

"So putting this all together I think there are a couple of things: First, it's clear that DRT is not that frequent but it's also very clear that when it occurs it increases the risk of subsequent ischemic stroke," Reddy concluded. "Two: It's an important clinical issue, what do we do with patients who have DRT? ... We don't have any prospective data to make absolute recommendations.

"But minimally, we believe that if DRT is detected, we need to have an aggressive surveillance regimen afterward with transesophageal echo or CTA."

As to the surveillance regimen more broadly, the FDA approval mandated TEE at 45 days and 1 year.

In the meta-analysis, comparing the registries (which mandated TEE at 45 days and 12 months) to the trials (which added a TEE at 6 months) showed a non-significant trend for more DRT capture with the additional TEE but no difference in ischemic stroke or systemic embolism.

The strategy Reddy said he favors would be to delay the first TEE to 3 to 4 months post-implantation, rather than adding a third intermediate procedure.

"There doesn't appear to be anything we know to do right now" to get rid of the problem, such as increased training, De Lurgio noted. "There's a big question, When are we out of the woods?"

Anecdotally, Reddy said his practice is to treat Watchman patients with a non-vitamin K antagonist oral anticoagulant and that DRTs don't usually emerge until after stopping anticoagulation. In the meta-analysis, some DRTs first showed up at the 1 year TEE.

Reddy cautioned that the findings may only apply to patients treated with the specific post-implantation antithrombotic regimen used in the studies and not to other types of left atrial appendage occlusion devices, although "the general concepts raised are likely applicable to other devices."

Register as Visitor to CPhI China 2018!