Author + information
- Zoltan G. Turi, MD∗ ()
- Division of Cardiology, Department of Medicine, Cooper University Hospital and Cooper Medical School of Rowan University, Camden, New Jersey
- ↵∗Reprint requests and correspondence:
Dr. Zoltan G. Turi, Cooper Vascular Center, Cooper University Hospital D-427, One Cooper Plaza, Camden, New Jersey 08103.
For patients with atrial fibrillation (AF) inappropriate for or refractory to ablation or cardioversion, the choices have been austere: treat with anticoagulation or expose the patient to a 5-fold or greater risk of embolic events including stroke. Anticoagulation remains the standard of care and has been studied extensively in multiple level I trials, most recently as part of the introduction of a new-generation polypharmacy for nonvalvular AF. Anticoagulation nevertheless remains an iatrogenically induced disease with significant associated morbidity and mortality to which patients may be exposed for decades. Mechanical alternatives have been based on the assumption that the left atrial appendage (LAA) is the locus for virtually all clots in nonvalvular AF. Despite a general consensus that elimination of the LAA substantially reduces stroke risk, and decades of ad hoc surgical practice of primarily suturing or stapling the appendage shut, the evidence base is remarkably meager, in particular in the cardiac surgery literature, where only 1 prematurely abandoned and generally unsuccessful randomized pilot trial was reported (1).
Three papers in this issue of the Journal(2–4) explore 3 different percutaneous or minimally invasive approaches to elimination of the LAA as a nidus of clot embolization. This field originated with the invention by Lesh and van der Burg of the PLAATO (Percutaneous Left Atrial Appendage Transcatheter Occlusion) device, a nitinol and fabric mesh combination that was the precursor to the extensively studied WATCHMAN as well as the AMPLATZER Cardiac Plug (ACP). In addition to the ACP, 2 other approaches are described in this issue: suture ligation of the LAA via a percutaneous subxyphoid transpericardial approach and amputation of the LAA via thoracoscopic use of a widely available stapler/cutter. That we have 3 such papers in 1 issue of the Journal(2–4) underscores the extent to which LAA interventions have benefited from a decade of technology development. However, despite substantial enthusiasm on the part of the interventional cardiology and electrophysiology communities, the evidence base for percutaneous approaches has been largely wanting as well, with only 1 high-level randomized controlled trial (5). Although the WATCHMAN and ACP are commercially available and widely distributed outside the United States, access in the United States is limited to investigational sites under rigidly controlled enrollment criteria that largely exclude patients who need the technology most: those with a high or prohibitive risk for anticoagulation.
For now, the best evidence base by far remains the 800-patient PROTECT AF (WATCHMAN Left Atrial Appendage System for Embolic PROTECTion in Patients with Atrial Fibrillation) study (5), which showed the noninferiority of the WATCHMAN to anticoagulation, with strong efficacy data but a high safety event rate, largely learning curve driven. It has been argued that a per-protocol analysis of that study provides proof–of-concept regarding LAA closure; patients with successful exclusion of the LAA who discontinued warfarin therapy had outcomes superior to those who took warfarin alone. The rest of the literature, including the 3 smaller studies in this issue of the Journal(2–4), provide primarily descriptive series: none are randomized; all have limited data and duration of follow-up, without central core laboratories, clinical events committees for adjudication, or data safety monitoring boards. Thus, although the paper on ACP implantation by Urena et al. (3) is a welcome addition to the scant data published on the ACP thus far (6), it is emblematic that 65 patients followed a mean 20 months represents the second largest series and the longest follow-up of the ACP despite a number of years of commercial availability outside the United States. The ACP is a familiar platform for operators accustomed to a variety of AMPLATZER technologies; that the device has been touted as being safe for use without anticoagulation is problematic: there remains a lack of evidence base for this claim. It should be noted that the ACP pivotal trial, like the completed pivotal trials of the WATCHMAN, will enroll only warfarin-eligible patients who will have anticoagulation for at least 45 days after device deployment.
As for the LARIAT device, it is an exceptionally clever technology that has followed a curious regulatory pathway, approved via a 510K process widely seen as a back door by the interventional community. The official approval (for tissue approximation) was gained without the clinical trials or regulatory processes through which the WATCHMAN has been and the ACP is being put. The LARIAT study raises additional concerns: it is a single-site study, with a significant role played by skilled operators involved with the device's development; registry data from the rapidly increasing number of sites and operators are needed. Although the limited short-term results look favorable, here too there are caveats: the patients appear low risk, with 45% having a CHADS2 score of 1, and <5% had a score >3. Surprisingly, 55% were on anticoagulation therapy at 1 year, despite enrollment criteria requiring that patients be poor candidates or ineligible for warfarin; this may reflect a concern (although I am not aware of any data) regarding clot formation in any residual stump. The safety profile in this limited cohort looks favorable (in the absence of a clinical events committee), but many interventional cardiologists believe that pericardiocentesis risk is inversely proportional to effusion size and will find puncture into a dry pericardium frightening. From my limited experience, the LARIAT placement approach does appear relatively safe. In fact, skills learned by LARIAT operators may result in generally safer diagnostic and therapeutic pericardial entry.
Overall, the LARIAT approach is one of the more ingenious of the new structural heart disease interventions, but its widespread introduction with only a rudimentary evidence base turns upside down our concept of the regulatory and clinical environment. Although many see it as liberating to be able to offer patients the opportunity to have this procedure without research and regulatory hurdles (which sometimes limit technology availability for years after widespread adoption outside the United States), it also seems as though a valuable safety net has been removed; its wisdom needs careful consideration. The frequency of adverse events, including trauma to the coronary arteries, puncture of cardiac chambers, tamponade, potential tearing of the LAA, and a host of other at least theoretical possibilities needs to be established in multicenter registries and sufficiently powered clinical trials. In the interim, LARIAT operators need to reflect on the not-yet established risk-benefit ratio; in our own institution, where we have just begun using the device, we have felt a need to create rump guidelines: high CHADS2 scores in patients with an unacceptably high risk of anticoagulation, with a compelling need such as previous embolic events.
Finally, the off-the-shelf cut-and-staple technology used for thoracoscopic appendectomy in a small number of patients by Ohtsuka et al. (4) was designed as a minimally invasive way to address the persistent problem of residual stumps and peridevice leaks inherent in both surgical and percutaneous LAA closure alternatives. Although this approach looks potentially useful, the single site, small number of patients, limited data, and limited follow-up do not allow any claim of safety or efficacy. The method requires left lung deflation, and in my opinion has a risk of catastrophic uncontrolled bleeding into a closed chest. It has the appeal of simplicity and ready availability of a generally inexpensive technology, but the safety in particular needs to be established.
Putting all this into perspective, the increasing number of tools available for the assault on the LAA makes it likely that these procedures will increase substantially. However, there are a number of yet-to-be explored ramifications, including effects other than embolus prevention, such as on the release of natriuretic peptides and on lipid metabolism. The full results of the PREVAIL (Randomized Trial of LAA Closure vs Warfarin for Stroke/Thromboembolic Prevention in Patients with Non-valvular Atrial Fibrillation) trial (a 407-patient extension of PROTECT AF) should be published later this year, and a randomized trial of surgical LAA exclusion is being planned. In the interim, despite the frustrations of waiting out our usual regulatory processes, we should exercise considerable prudence in patient selection for each of the three devices described in this issue of the Journal until a more solid evidence base is available. Until then, although long-term anticoagulation is cumbersome and has its own significant (but well established) toxicities, for those who are appropriate candidates, it remains the best studied approach as well as the standard of care.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Dr. Turi has received grant support from Boston Scientific and St. Jude Medical.
- 2013 American College of Cardiology Foundation
- Bartus K.,
- Han F.T.,
- Bednarek J.,
- et al.
- Urena M.,
- Rodés-Cabau J.,
- Freixa X.,
- et al.
- Ohtsuka T.,
- Ninomiya M.,
- Nonaka T.,
- Hisagi M.,
- Ota T.,
- Mizutani T.