Fig. 2.1
Patient flow chart in Watchman PROTECT-AF: many patients dropped out of the study after screening due to anticipated issues with complete LAA closure in these early days of the technique [1]
The PROTECT-AF trial included a low-risk population of patients with an average CHADs score of 2.1; V. Reddy and colleagues published the four year results in 2014. Early publications had found the initial rate of periprocedural, serious complication rate to be around 5%. In particular, pericardial effusion and periprocedural ischemic stroke emerged as an issue in these early days (Fig. 2.2). Standardization and modifications of the implantation technique have lead to a reduction of serious, non-transient complications to 0.5% even though morbidity of the patients was much higher in the recent EWOLUTION registry (CHA2DS2-VASc average 4.3). The non-inferiority (primary study hypothesis) of LAA closure was confirmed early on and after four years in PROTECT-AF (Fig. 2.3). Surprisingly, patients with the Watchman had an even lower cardiovascular and all-cause mortality than the warfarin group after four years [3]. The combined primary safety and efficacy endpoints (Fig. 2.4) confirmed the primary hypothesis of this trial, namely non-inferiority of Watchman LAA closure compared to warfarin long-term treatment in patients with atrial fibrillation in need of stroke protection (CHADs ≥1). Strokes in the Watchman LAA closure group appeared to be clinically less significant than in the warfarin group with only 0.5% of disabling strokes in the device group compared to 1.2% in the warfarin group (Fig. 2.5). Interestingly, quality of life assessment in the PROTECT-AF trial showed favourable results for the device group [4]. Bleeding outcomes are also favourable for the device group [5].
Fig. 2.2
Primary safety endpoint of the Watchman PROTECT-AF study after 4 years of follow-up. Pericardial effusion and bleeding even under the less intense anticoagulation regime of the trial (45 days of warfarin followed by dual antiplatelet therapy until 6 months; aspirin life long) had the highest event rate in the device group while major bleeding and devastating hemorrhagic stroke occurred in the warfarin group [2, 3]
Fig. 2.3
Primary efficacy endpoints ((a) ischemic stroke; (b) cardiovascular mortality; (c) All-cause mortality) of the Watchman PROTECT-AF study after 4 years of follow-up. The study was designed as a non-inferiority trial; the results not only show non-inferiority but also superiority regarding all-cause and cardiovascular mortality in comparison to warfarin. The event rate of ischemic stroke was low as expected; patients in the trial had a mean CHADs score of 2.1 [3]
Fig. 2.4
Primary efficacy (a) and safety (b) endpoints of the Watchman PROTECT-AF study after 4 years of follow-up. Primary efficacy was defined as composite of stroke, systemic embolization and cardiovascular death. Primary safety outcome was defined as composite of major bleedings and procedure-related complications. HR indicates hazard ratio [3]
Fig. 2.5
Intention-to-Treat primary efficacy and safety outcomes of the Watchman PROTECT-AF study after 4 years of follow-up. Observed event rate is calculated as 100 patient-years with a 95% credible interval. Primary safety outcome was defined as composite of major bleedings and procedure-related complications. Disabling or fatal stroke were defined as those with a modified Rankin score of 3–6. Note that the strokes observed in the LAA closure group were less harmful than in the warfarin group. HR indicates hazard ratio [3]
The PREVAIL trial was started on request by the FDA to prove periprocedural complications to be reduced with training and the evolved implantation technique by 2009; the data provided by the PROTECT-AF and the continued access protocol registry data (CAP 1 & 2) were regarded insufficient to allow FDA approval of the Watchman device. The study included 269 patients randomized 2:1 at higher risk (mean CHADS2 2.6, see Fig. 2.6) and finished recruitment by 2012 [6]. Safety events, particular pericardial effusion, were significantly reduced to 1.7% of procedures; all but one of these effusions could be treated with pericardiocentesis (Fig. 2.7). One patient had to be operated but could be stabilized. During follow-up eight ischemic strokes in the Watchman group to none in the warfarin group were observed; the latter result was quite surprising and outside the expected range for this patient cohort. The trial was also not powered to observe efficacy of the procedure; furthermore strokes in the Watchman group were clinically less significant and rarely disabling. Further data with larger patient numbers are needed to clarify this point.
Fig. 2.6
Baseline patient data of the Watchman PREVAIL trial (V. Reddy, TCT 2014)
Fig. 2.7
Safety and efficacy of Watchman LAA closure in the PREVAIL trial. The primary study hypothesis, namely reduction of periprocedural safety events with a particular focus on pericardial effusion was met. Efficacy endpoints, particular ischemic stroke, was numerically higher in the device group yet particularly low in the warfarin group
Following FDA approval of the Watchman occluder in spring 2015 further randomized trials are in planning as of summer 2016; particularly the ASAP-TOO study will provide further evidence regarding safety and efficacy of LAA closure with the Watchman device in patients ineligible to warfarin therapy. Patients are randomized to platelet inhibition vs. LAA closure with the Watchman device.
2.2 Prospective Registry Data on LAA Closure with the Watchman Device from the US
Following PROTECT-AF the participating centres were allowed to continue implanting Watchman devices within the continued access protocol (CAP) with further prospective data collection. The registries known as CAP 1 (2008–2010) and CAP 2 (2012–2014) included another 1150 patients (Fig. 2.8); they largely confirmed the results of PROTECT-AF regarding efficacy (Fig. 2.9). All the trials showed an effect of the procedural learning curve with periprocedural complications to be reduced to around 1%. During follow up, the rate of hemorrhagic stroke was significantly reduced in the Watchman group as was the rate of unexplained or cardiovascular mortality compared to the expected event rate [7]. Data were also analyzed for cost-effectiveness showing Watchman LAA closure to be effective after around 5 years in comparison to both warfarin including costs for monitoring and NOAC [5, 8].
Fig. 2.8
Summary of currently available data on the Watchman device from the US. Combination of two randomized trials (PROTECT-AF and PREVAIL) together with the continued access protocol (CAP 1: continued access to PROTECT-AF; CAP 2: continued access to PREVAIL) following the studies [4]
Fig. 2.9
Combined analysis of the US randomized trials and registries: significant benefits of the Watchman group can be found for the endpoints ischemic stroke or systemic embolization, hemorrhagic stroke, cardiovascular or unexplained death and non-procedure related major bleeding [4]
2.3 Prospective Registry Data on LAA Closure with the Watchman Device from Europe
In Europe LAA closure emerged in 2009 and was already part of the ESC guidelines on atrial fibrillation of 2012 based on the PROTECT-AF data as well as the ASAP prospective registry performed by four centres in Germany between January 2009 and November 2011; patients with a relative or absolute contraindication for oral anticoagulation had a IIb evidence level B recommendation for the procedure [9]. The same recommendation is given in the new 2016 guidelines further supported by the new randomized trials and registries since the 2012 update; the guidelines ask for a randomized trial comparing LAA closure to NOAC therapy in order to give LAA closure a class I recommendation [10].
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
