Device Related Thrombus after Left Atrial Appendage Occlusion: The Villian of the Piece

Left atrial appendage occlusion (LAAO) is an important treatment option for patients with atrial fibrillation (AF) who are not good candidates for long-term oral anticoagulation(1) and in select patients at high-risk for bleeding.(2) Use of LAAO has increased significantly. While LAAO provides an option for stroke prevention in patients who cannot tolerate long-term oral anticoagulation, as with any interventional procedure, it does carry risk. Device-related thrombosis (DRT) is among the most-feared complications of LAAO as it increases risk of the very outcome it is designed to prevent, namely embolic stroke. While DRT is an infrequent event, its true incidence, risk factors and mechanisms, and optimal measures of prevention are incompletely understood.

In this issue of JACC: Clinical Electrophysiology, Schmidt and colleagues report on the incidence, predictors, and outcomes of DRT in the pivotal Amulet investigational device exemption trial (IDE). In the trial, patients with atrial fibrillation (AF) and a CHADS₂ score >=2 or CHA₂DS₂-VASc score >=3 were randomized 1:1 to the Amulet or Watchman device. Amulet patients were treated with dual antiplatelet therapy (DAPT) for 6 months and Watchman patients were treated with aspirin and oral anticoagulation (OAC) for 45 days followed by DAPT for 6 months and then aspirin. The trial protocol mandated transesophageal echocardiography (TEE) at 45 days and 12 months. Additionally, patients with a peri-device leak of >5 mm at 45 days were required to have a TEE at 6 months and a TEE was required after any confirmed ischemic stroke. The rate of evaluable TEEs was 88% in the Amulet group and 83% in the Watchman group.

In the trial, the rate of DRT was 3.9% through 18 months of follow-up, including 3.4% (n=30/903) in the Amulet group and 4.8% (n=40/885) in the Watchman group. Most of the DRT events in the Amulet patients (61%) occurred within 45 days, whereas most DRT events in the Watchman group occurred after 45 days (74%). Surprisingly, only 46% of the Amulet patients and only 68% of the Watchman patients were switched to OAC after the diagnosis of DRT. Predictors of DRT included the presence of AF (versus sinus rhythm), increasing age, and female sex. Finally, DRT was associated with a significantly increased risk of cardiovascular death (8.7 vs 3.9%, p=0.04).

The Amulet IDE analysis has many important strengths. First and foremost, it is the first large multicenter randomized trial to compare two distinct LAAO devices with mandated TEE imaging both in 45-day and 1-year follow-up. Second, all TEEs and clinical events were adjudicated by a core laboratory and clinical events committee. However, there are also some notable limitations of the analysis that should be kept in mind. First, the trial evaluated the Watchman 2.5 device and not the Watchman FLX device, which has several efficacy and safety advantages over the Watchman 2.5 device. Moreover, despite the large size of the trial, given the infrequency of the DRT events and ischemic stroke, there was limited power to compare key outcomes, including ischemic stroke. Finally, post procedure anticoagulation after Watchman 2.5 implantation was most commonly with warfarin and aspirin and this combination is infrequently used in clinical practice.(3)

The findings from Schmidt and colleagues have several implications. While the rates of DRT were not significantly different between the devices, the timing of DRT events was different and consistent with prior studies of DRT.(4) DRT events occurred earlier with Amulet and later with the Watchman devices. The observed temporal differences could be due to mechanisms of DRT that are unique to each device, differences in the rate of endothelialization, differences in drug therapy, or some combination of these factors. Among patients treated with the Amulet device who develop DRT, the ridge at the left superior pulmonary vein has been found to be uncovered in 82% of those cases.(5) In the case of the Watchman device, most DRT appear to be associated with the central screw/threaded insert.(6) The changes in design (reduced metal exposure) with the Watchman FLX device appear to have led to a lower rate of DRT given the 1.75% (n=7/400) rate of DRT in the PINNACLE FLX trial.(7) It is also worth emphasizing that deep implantation (>10 mm) has also been associated with DRT and should be avoided whenever possible.(4)

The use of OAC early after Watchman implant raises the possibility that antiplatelet therapy leaves patients more vulnerable to DRT. A recent retrospective analysis from Della Rocca and colleagues demonstrated that half-dose direct acting OAC (DOAC) may be superior to standard antithrombotic therapy (DOAC + aspirin x 45 days, aspirin & clopidogrel x 6-months, and aspirin indefinitely) after Watchman implant.(8) When surveilling with 3 TEE studies after implant (45, 180, and 365 days), the investigators found that DRT occurred in 3.4% (n=12/357) of patients treated with standard antithrombotic therapy versus no DRT in those treated with half-dose DOAC (3.4% vs 0.0% at mean follow-up 13 months; p=0.009). The half-dose DOAC strategy was also associated with a lower occurrence of the composite endpoint of DRT, stroke or systemic embolism, or major bleeding. In contrast, a large meta-analysis that included data from 83 studies and 12,326 patients found no differences in the occurrence of DRT between those treated with short-term OAC versus antiplatelet therapy after LAAO (RR 1.33, 95% CI 0.74–2.39).(9)

One of the more surprising findings form Schmidt and colleagues was the absence of a clear association between DRT and the occurrence of ischemic stroke. As the Amulet IDE investigators emphasize, this finding should be viewed with caution. While there was no difference in ischemic stroke, CV death was more than two-fold greater in those with DRT. In the global study of DRT after Amulet implantation, DRT was associated with a 5-fold increase in the risk of stroke (5.27, 95% CI 1.58–17.55).(5) Findings from a large multicenter DRT registry(4) and a pooled analyses of the prospective Watchman trials(10) have also identified significantly increased risks of ischemic stroke in those who have DRT. Once DRT is identified, prompt anticoagulation is indicated, even in the presence of significant relative contraindications to OAC.(1) It is surprising that only 46% of Amulet patients and 68% of Watchman patients with DRT were switched to OAC in the Amulet IDE study. These findings highlight an opportunity to improve care for patients with DRT.

Putting it all together, what are the key take-home messages from the insightful analysis by the Amulet IDE investigators? First, clinicians must remain vigilant for DRT. Clinicians should have a high index of suspicion for DRT in the first 45 days after implant in those treated with Amulet and >45 days in those treated with Watchman. Older patients, women, and those in AF are at high-risk of developing DRT. Imaging in late follow-up is important and should not be neglected. We can’t diagnose DRT if we don’t look, and early detection is crucial to prevent ischemic stroke. However, the ideal timing to conduct follow-up imaging (6 months, 12 months?) remains uncertain and may vary by post-procedure antithrombotic regimen. Once DRT is identified, prompt initiation of OAC is critical to avoid ischemic stroke. Finally, despite the availability of LAAO for more than 7 years we have not identified the best drug regiment to prevent DRT in the acute and longer-term periods after implant. This may be the most important unanswered question in left atrial appendage closure management.