A Need for Closure

Despite delays, percutaneous left atrial appendage closure devices seem closer than ever to FDA approval.

Prevention of stroke in patients with atrial fibrillation is a growing and continuous dilemma in the cardiology community. Through the years, a number of strategies for patients with nonvalvular AF have been unveiled. In particular, percutaneous left atrial appendage closure has been touted as an alternative to long-term anticoagulant therapy for stroke risk reduction in this patient population.

“Percutaneous left atrial appendage closure is an interventional procedure that has shown to be at least competitive to anti-vitamin K oral anticoagulation and probably also to non–anti-vitamin K oral anticoagulation in terms of stroke prevention. Moreover, it avoids most of the bleeding risk germane to anticoagulation,” Bernhard Meier, MD, professor of cardiology and chair of the department of cardiology at the Swiss Cardiovascular Center, Bern University Hospital, Switzerland, told Cardiology Today’s Intervention.

“However, because it is a preventive procedure and because it is one of the more intricate interventional procedures, it is only reluctantly adopted by interventional cardiologists or electrophysiologists,” Meier said.

The path to US regulatory approval of left atrial appendage (LAA) closure devices has been difficult, but experts said continued data and follow-up support a role for LAA closure as an alternative to anticoagulation in selected patients.

The Evolution of LAA Closure

The concept of mechanical occlusion of the LAA is not new, but has been refined over the years, David R. Holmes Jr, MD, from the department of cardiovascular diseases and internal medicine at the Mayo Clinic in Rochester, Minn., told Cardiology Today’s Intervention.

“The history dates back to the late 1940s, when the first surgical experience of tying off the left atrial appendage or excluding it some way was reported in the literature,” Holmes said. “If you fast forward to the last decade and a half, people have increasingly recognized the relationship between increasing stroke and increasing atrial fibrillation as the patient population ages. Given the mortality and morbidity of stroke, we have focused a lot on this group of patients with nonvalvular atrial fibrillation.”

Holmes said in the 1990s research into the causes of stroke determined that in patients with nonvalvular AF who develop stroke, the stroke typically originates in a thrombus in the LAA and subsequently embolizes to the brain or other vital organ.

“So then the idea came that because the left atrial appendage does not have a huge function in the setting of atrial fibrillation, that perhaps a mechanical approach to occluding the left atrial appendage might prevent stroke from happening,” Holmes said. “That is the genesis of mechanical closure of the left atrial appendage, which is now being done in surgery as well as in the catheterization laboratory by both electrophysiologists and invasive interventional cardiologists. “

Holmes said interest in LAA closure has continued to increase over the years, and is driven by a clear clinical need.

“We know that anticoagulants such as warfarin, and now the new oral anticoagulants, are used for stroke prevention and are quite effective,” he said. “The problem is that by virtue of side effects or concerns about bleeding, about half of the patients, particularly those at highest risk for stroke, are not treated with drugs. We have a therapy that works, but is underutilized in the patients who need it most.”

The Watchman Device

The Watchman Left Atrial Appendage Closure Device (Atritech/Boston Scientific) is backed by strong data from the PROTECT AF and PREVAIL trials and has been the subject of three FDA advisory committee meetings. The device was approved for use in Europe in 2005 and is currently available in more than 70 countries, according to Boston Scientific. It remains an investigational device in the United States.

The Watchman device is designed to block off the LAA, considered to be the origin of more than 90% of stroke-causing thromboembolisms in patients with nonvalvular AF. The device consists of a self-expanding nitinol frame, available in 21 mm, 24 mm, 27 mm, 30 mm and 33 mm, and a polyethylene terephthalate cap intended to block emboli and facilitate healing at 45 days. It is preloaded with a delivery catheter and expands upon deployment into the LAA. During a catheter procedure, the device is permanently placed within or slightly distal to the LAA ostium. Once implanted, the device is anchored in place through fixation barbs on the frame. On average, the procedure lasts less than 1 hour, and patients stay in the hospital for 24 hours.

This device is appropriate for patients who do not tolerate or whose lifestyle or personal preferences rule out anticoagulation, Vivek Y. Reddy, MD, professor of medicine at Mount Sinai School of Medicine and co-principal investigator for the PROTECT AF trial, told Cardiology Today’s Intervention.

“In patients who are doing well on oral anticoagulation, there is no compelling reason to make any changes. On the other hand, for patients who are not good candidates for oral anticoagulation, LAA closure with the Watchman device is a good option,” he said.

Reddy noted that the efficacy of this device has not been proven in certain patient populations. “The Watchman device was not studied in patients with severely reduced ejection fraction [<30%)]. this is a distinction from oral anticoagulants, which can be used in [patients with] any ejection fraction,” he said.

Path to US Approval

The clearing of US regulatory hurdles has progressed in fits and starts for the Watchman device.

In October, the device was held to a third FDA Circulatory System Devices advisory panel meeting and revote. Overall, the panel concluded by a 6-5 vote that the benefits outweighed the risks, but were undecided regarding efficacy. The panel voted 12-0 that the device is safe, but voted 6-7, with the panel chair providing the tie-breaking vote, that it is not effective.

According to an FDA executive summary, updated follow-up data from the PREVAIL trial raised concerns about risk for ischemic stroke with the device. The data, which were locked on Jan. 30, 2014, and included an additional year of follow-up beyond what was reviewed at the December 2013 advisory committee meeting, revealed 24 primary endpoint events among patients assigned the Watchman device vs. nine events among patients assigned warfarin.

Despite concerns raised by those data, many experts said they believe that previous studies, particularly PROTECT AF, provide strong support for the safety and efficacy of the catheter-delivered Watchman device. Additionally, some have questioned the applicability of the PREVAIL data to the FDA’s recent review.

“It is not appropriate to weigh an interim analysis of a trial higher than the data of a finished trial with 5 years’ follow-up, which has shown clearly a superiority of the Watchman [device] regarding the primary endpoint of stroke, systemic embolism and death. Even if you look at CV death only, Watchman was superior [to warfarin],” Horst Sievert, MD, from the CardioVascular Center Frankfurt, Sankt Katharinen, Germany, said in an interview.

PROTECT AF

The longest-term trial of LAA closure to date, PROTECT AF enrolled 707 patients with nonvalvular AF who were randomly assigned 2:1 to the Watchman device (n=463) or warfarin (n=244).

Primary analyses supported a significant finding of noninferiority for the Watchman group relative to the warfarin group for the primary efficacy endpoint of stroke, systemic embolism or CV death. Initial clinical trial results with 1,065 patient-years of follow-up revealed that the Watchman device was associated with a 38% reduction in primary endpoint events vs. warfarin; the device was successfully implanted in 91% of patients; 87% of patients were able to discontinue warfarin 45 days after successful implant; and patients who were successfully implanted and in whom warfarin was discontinued had a 60% reduction in the rate of the composite endpoint of all stroke, systemic embolism and CV death. An increased risk for procedural pericardial effusion emerged, but risk declined during the course of the trial as investigators gained more experience with the device and procedure.

“What this showed was, for the primary endpoint, the Watchman device was superior to warfarin,” Reddy said.

Further, “if you look specifically at individual endpoints, there were more ischemic strokes in the Watchman group compared with warfarin, but fewer hemorrhagic strokes in the Watchman group, giving an overall 80% to 85% reduction in hemorrhagic stroke,” he said. “If you look at CV death, which is arguably the most important endpoint, there were fewer CV deaths in the Watchman group compared with warfarin, and fewer all-cause deaths. It was a strikingly positive trial for the Watchman device.”

According to Meier, the slightly higher number of ischemic strokes in the Watchman group should be weighed against a decreased bleeding risk with the device vs. warfarin.

In November, 3.8-year follow-up data from PROTECT AF were published in JAMA demonstrating that LAA closure met the standards for superiority vs. warfarin for the prevention of the composite primary endpoint of stroke, systemic embolism and CV or unexplained death. At follow-up, researchers reported 39 (8.4%) events in the Watchman group vs. 34 (13.9%) in the warfarin group. The primary event rate was 2.3 events per 100 patient-years with the device vs. 3.8 events per 100 patient-years with warfarin (RR=0.6; 95% CI, 0.41-1.05). Criteria for both noninferiority and superiority were met. The device also was associated with lower rates of CV mortality (one event per 100 patient-years vs. 2.4 events per 100 patient-years; HR=0.4; P=.005) and all-cause mortality (3.2 events per 100 patient-years vs. 4.8 events per 100 patient years; HR=0.66; P=.04).

“However, it should be noted that these mortality endpoints are secondary endpoints, and due to multiplicity of data analysis, there is some uncertainty in the confidence of this conclusion,” Reddy and colleagues wrote in JAMA. “Although the device implantation procedure was associated with early complications, the accumulation of complications related to chronic anticoagulation resulted in similar safety profiles for the two modalities.”

“PROTECT AF has shown that LAA closure with the Watchman device, despite periprocedural complications, is as safe as anticoagulation, even after only 4 years. This is very encouraging because device- or procedure-related complications are extremely rare beyond 30 days, whereas on the other side anticoagulants continue to cause complications over time,” Sievert said.

PREVAIL

PREVAIL is a smaller study aimed at corroborating the results from PROTECT AF and evaluating the safety of LAA closure. The trial enrolled 407 patients who were randomly assigned 2:1 to the Watchman device or warfarin.

Eighteen-month results demonstrated that the co-primary composite endpoint of stroke, systemic embolism and CV or unexplained death was 0.064 in the Watchman group vs. 0.063 in the warfarin group (RR=1.07; 95% CI, 0.57-1.89). The Watchman device failed to reach the prespecified noninferiority criteria. Also at 18 months, the Watchman group yielded a rate of 0.0253 for the second co-primary endpoint of stroke or systemic embolism more than 7 days after hospitalization vs. 0.02 in the warfarin group, which demonstrated noninferiority.

According to Reddy, the safety endpoint of PREVAIL also was important, as it was designed to address early safety events that were reported in PROTECT AF. “There were a lot of safety events early in [PROTECT AF] and fewer later in the trial. The hypothesis was that, with experience, there would be fewer safety events. So, one of the questions of PREVAIL was: Is this true?”

The device was associated with a 2.2% rate of early safety events, significantly lower than in the initial PROTECT AF study; this outcome satisfied the prespecified performance goal. When a more inclusive definition of adverse events was used, findings from the PREVAIL trial demonstrated lower event rates than in PROTECT AF (4.2% vs. 8.7%; P=.004). A decrease in pericardial effusions requiring surgical repair was reported, from 1.6% to 0.4% (P=.027). A decrease from 2.9% to 1.5% (P=.36) was reported for those requiring pericardiocentesis.

“Although noninferiority was not achieved for overall efficacy, event rates were low and numerically comparable in both arms. This trial provides additional data that LAA occlusion is a reasonable alternative to warfarin therapy for stroke prevention in patients with [nonvalvular] AF who do not have an absolute contraindication to short-term warfarin therapy,” Holmes and colleagues wrote in the study.

Reddy, also a co-principal investigator for PREVAIL, said there was another interesting endpoint in this trial. The trial “mandated that a certain number of devices had to be implanted by new operators, meaning operators who had never done the procedure before,” he said. “And indeed, approximately 40% of the patients were implanted by new operators. Importantly, the complication rate between the new operators and experienced operators was not significantly different.”

Reddy said this finding is important because it suggests that the learning curve for this procedure and device may not be a significant obstacle to its adoption.

Meier noted that one potential flaw of the PROTECT AF and PREVAIL trials is the exclusion of patients not eligible for warfarin.

“In the real world, such patients are the mainstay of LAA closure,” he said. “Notwithstanding, PROTECT AF shows a significant overall benefit of LAA closure over warfarin, which includes improved mortality at 4 to 5 years.”

The Watchman device has also been studied in the ASAP study and the CAP and CAP2 registries.

According to Meier, the combined data provide compelling evidence in favor of LAA closure as a first-line recommendation for patients with nonvalvular AF. However, “in the guidelines, LAA closure only gets a class II indication for patients not suitable for oral anticoagulation,” he said.

Alternative Devices

Several other percutaneous LAA closure devices are under clinical investigation, some of which are available for use in countries outside the United States. Such devices include the Amplatzer Cardiac Plug (St. Jude Medical), the Amplatzer Amulet Left Atrial Appendage Occluder (St. Jude Medical), the WaveCrest Left Atrial Appendage Occlusion System (Coherex Medical), the Transcatheter Patch (Custom Medical Devices) and the Lariat Suture Delivery Device (SentreHeart).

“There are many other percutaneous devices in clinical trials and the results seem comparable to the Watchman device,” Sievert said. “There are technical and design issues that make some devices more or less suitable for specific anatomies.”

Meier has been involved in the development of the Amplatzer technique, and said he has implanted about 500 of the devices during LAA closure procedures.

Follow-up data published recently in EuroIntervention demonstrated that LAA closure with the Amplatzer Cardiac Plug was safe and effective for the prevention of stroke in a study of 134 high-risk patients with nonvalvular AF with long-term oral anticoagulant contraindication.

“The Amplatzer technique has been used in close to a million people worldwide, and is therefore highly seasoned. The Amplatzer devices also have proven not to be thrombosis-prone, particularly in patients with atrial shunts,” Meier said. He added, however, that the likelihood of thrombosis with the Amplatzer device, or any percutaneous LAA closure device, also depends on the condition for which it is used.

“LAA closure is usually done in patients with atrial fibrillation, a much more thrombosis-prone environment for a foreign device. Consequently, thrombi are found on Amplatzer devices in about 3% to 5% of cases in comparison to the same devices being placed in sinus rhythm at the atrial septum, where the thrombosis rate is 10 times lower. However, similar thrombus rates have been found with all competitive devices,” Meier said.

In addition, some of the new devices can be implanted without the use of echocardiographic guidance.

“The Amplatzer device lends itself ideally to implantation under fluoroscopy only, although the Watchman device and the WaveCrest device can also be implanted without echocardiographic guidance,” Meier said. “The Amplatzer device has the advantage of featuring a disk attached to the anchoring plug. The disk is meant to cover the ostium of the left atrial appendage, not unlike the pacifier plate outside a baby’s mouth.” This enables a more complete occlusion with the Amplatzer devices than with the Watchman and WaveCrest devices, he said.

“Looking at the clinical results, this does not seem to be of great importance, but it is certainly an optically appealing feature,” he said. “On the other hand, the disk exerts some continuous tug on the plug, which, rarely, may engender device embolization.”

Meier cited data showing embolization rates of more than 1% with the Amplatzer technique vs. less than 1% with the Watchman technique.

“However, the Watchman device needs to be implanted fairly deeply into the LAA to achieve such low embolization rates,” he said. “This leaves a good part of the LAA neck unplugged. Moreover, there are few, if any, anatomical reasons not to attempt implantation of an Amplatzer device into the LAA, while about 20% of patients are deemed unsuitable for a Watchman device.”

Data from the ongoing Amplatzer Cardiac Plug (ACP) trial are expected to produce more robust clinical evidence.

The second-generation percutaneous transcatheter Amplatzer Amulet device leverages the design of the Amplatzer Cardiac Plug with additional features allowing treatment of a wider range of appendage anatomies in patients with nonvalvular AF. The device is offered in eight sizes to accommodate varying anatomies, according to St. Jude Medical.

The WAVECREST I clinical trial is being conducted to establish the safety and efficacy of the WaveCrest device for LAA closure in patients with nonvalvular AF.

The Lariat device is currently the only FDA-approved device that has been used in the United States for percutaneous catheter-based LAA ligation; however, this device has not been FDA approved or rigorously studied for the prevention of stroke in AF.

Future of LAA Closure in US

Although the split vote from the 2014 FDA panel meeting would appear to suggest a lack of regulator confidence in the Watchman device, the main concerns pertained to the broadness of the indications.

According to a previous report published in Cardiology Today’s Intervention, FDA advisory panel member David Slotwiner, MD, stated that although he voted “no” on effectiveness and benefits outweighing the risks, he would consider a revision that recommends LAA closure as a second-line therapy. “I don’t feel comfortable with the broad indication as labeled. But … I would feel comfortable with … rewording, and clearly stating in the indications that this is not equivalent to warfarin, [and is] a second-line therapy,” he said.

Reddy said he felt that these reservations were representative of many of the panel members’ concerns.

“If you listen to the panel’s final comments, the majority of the panel members would have actually voted ‘yes’ on overall safety and effectiveness if the indications for use statement had been a little bit different,” he said.

Reddy said he believes that the device is important and necessary, and will be approved soon.

Sievert added that regarding which of the two modalities should be given first-line therapy status, he feels that it is ultimately more important for patients to be permitted to make an informed decision between LAA closure or anticoagulation.

“It should not be the task of a regulatory agency, or the task of doctors, to dictate the destiny of individual patients,” he said. “Patients should be informed as well as possible about the advantages and disadvantages of different treatment strategies, and then be allowed to make their own decisions.”

Holmes said he believes that the need for the Watchman device will drive its approval, and that the Amplatzer Amulet will likely be next in line.

“The unmet clinical need is going to drive the field so that we will have the approval of a device — or multiple devices — that will have been proven to be safe and effective in terms of risk-benefit ratio for stroke prevention in selected patients with atrial fibrillation,” he said. – Jennifer Byrne

References:

www.henryford.com/body.cfm?id=57694. Accessed on Dec. 14, 2014.

Disclosures: Holmes reports that he and the Mayo Clinic have a financial interest in technology related to this research; that technology has been licensed to Boston Scientific. Meier reports financial disclosures with several device and pharmaceutical manufacturers. Reddy reports receiving grant support and consultant fees from Boston Scientific. Sievert reports financial disclosures with several device and pharmaceutical manufacturers.