Author + information
- Vinod H. Thourani, MD∗ (, )
- Lillian Tsai, AB and
- Hanna Jensen, MD, PhD
- Structural Heart and Valve Center, Division of Cardiothoracic Surgery, Joseph B. Whitehead Department of Surgery, Emory University School of Medicine, Atlanta, Georgia
- ↵∗Reprint requests and correspondence:
Dr. Vinod H. Thourani, Emory University Hospital Midtown, 550 Peachtree Street NE, 6th Floor, Medical Office Tower, Atlanta, Georgia 30308.
- aortic stenosis
- self-expanding transcatheter aortic valve
- transcatheter aortic valve replacement
- transient ischemia attack
In this issue of the Journal, Bosmans et al. (1) report much-awaited stroke outcomes from the ADVANCE study, a multicenter, prospective, nonrandomized cohort of patients undergoing transcatheter aortic valve replacement (TAVR) with the CoreValve prosthesis (Medtronic Inc., Minneapolis, Minnesota) at 44 mostly European sites between March 2010 and July 2011. In this real-world cohort of patients with severe aortic stenosis, the investigators observed stroke rates of 3.0% at 30 days post-TAVR and 5.6% at 2 years. They reported no significant predictors of periprocedural stroke or transient ischemic attack (TIA) occurring within the first postoperative day but demonstrated an association between acute kidney injury, female sex, or major vascular complication and 30-day stroke. The only predictor of long-term stroke was a prior coronary artery bypass grafting operation.
These data come at an intriguing time. The field is currently seeing reports from randomized controlled trials in which a survival benefit and a reduction in the rate of major adverse cardiovascular and cerebrovascular events in TAVR patients is suggested, for the first time, compared with surgical aortic valve replacement (SAVR) (2). Third-generation balloon-expandable valve systems are evolving to an extent at which the rate of moderate or greater paravalvular leak is almost insignificant (3–5). Clinical trials are recruiting intermediate-risk patients to undergo a treatment that only 5 short years ago was reserved for those with no other options. Rare innovations in cardiac surgery have had the momentum of TAVR, equally complemented and driven by the constant, fluid evolution of valve systems and delivery devices. Robust, structured, and monitored research must sprint to keep up.
When the results of the randomized controlled landmark PARTNER (Placement of Aortic Transcatheter Valves) trial that led to U.S. Food and Drug Administration approval of the balloon-expandable valve were published, there was cause for alarm in terms of neurological outcomes: An increased risk of operative stroke and TIA was observed in TAVR patients compared with SAVR patients (6,7). Although cardiopulmonary bypass and aortic cross-clamping, the traditional nemeses of cerebral circulation, were avoided in this novel transcatheter intervention, 30-day stroke rates were quoted to be as high as 5.0% for inoperable patients and 3.8% for patients with high surgical risk (6,7). Subsequent studies speculated that temporary circulatory disruptions during balloon valvuloplasty or rapid ventricular pacing may have led to periods of suboptimal cerebral perfusion and ischemia (8), or that valvular calcium deposits, tissue factor, and thrombin reserves from the diseased native valve were released into the circulation, triggering coagulation cascades that led to increased thrombogenicity (9). It is very likely that the larger first-generation devices that crossed the aortic arch also played a significant role in the seemingly high stroke rate with TAVR.
However, more contemporary comparisons have largely mitigated the difference in stroke risk in TAVR compared with SAVR. In a recent study that specifically examined neurological outcomes in patients with aortic stenosis, there were no differences in overall neurological injury, cerebral embolic load, ischemic lesions, or oxygen desaturation in patients undergoing SAVR or TAVR (10). It is likely that the technology development that has led to smaller devices manipulated in the arch has neutralized, at least in part, the risk profile between the interventions. The Society of Thoracic Surgeons (STS) and American College of Cardiology Transcatheter Valve Therapies registry, which documents the commercial TAVR experience in the U.S. post–Food and Drug Administration approval, quoted a stroke rate as low as 2.0% at 30 days (11) and 4.1% at 1 year after TAVR (12), with these data gathered approximately 1 year after the recruitment of the ADVANCE study patients.
Although it fortunately appears that stroke rates in TAVR are diminishing, it is undeniable that permanent neurological injury is the single most significant and most feared morbidity of cardiac interventions. A debilitating stroke is viewed by many as the greater evil, even compared with operative death. As the target audience for TAVR expands, the identification of any procedural risk factors for neurological complications is imperative, an effort to which the current paper by Bosmans et al. (1) commendably contributes. Nevertheless, a few words of caution, recognized by the authors but further emphasized in this editorial, are in order.
The patients in the ADVANCE study were healthier (mean STS score <9%) than patients enrolled in previous randomized controlled trials. This, combined with the fact that TAVR procedures were undertaken in experienced centers, led to a fairly low stroke rate, which although fortunate clinically, limits the relevance of especially the multivariable analysis in the presence of very few events. The interpretation of these data must be done while the possibility of type II error is kept in mind; although a large cohort was enrolled, the study may be a victim of the investigators’ success in avoiding neurological complications, and thus, some important risk factors may have been masked.
Second, neurological events were not actively sought; they were reported when identified. A concern from the clinical team led to a review by an on-site neurologist, followed by an independent neurologist, and finally, the event was retrospectively adjudicated by the clinical events committee of the study using Valve Academic Research Consortium criteria. Thus, although there is little doubt that the reported events were thoroughly investigated and appropriately reported, in the absence of rigorous neurological review before and after the procedure, the possibility of neurological events having gone undetected exists.
Indeed, in a previous prospective 200-patient cohort undergoing SAVR, when patients underwent thorough review by neurologists preoperatively and on postoperative days 1, 3, and 7 accompanied by routine postoperative cerebral magnetic resonance imaging, the clinical stroke rate was a whopping 17%, much higher than in any previously reported contemporary studies (13). At the same time, the reported stroke rate in the STS database for this same patient population was only 7%, which highlights the difference between stroke that is reported rather than stroke that is actively sought. Furthermore, in this same patient cohort, subclinical infarct on neuroimaging was detected in more than one-half of the patients, an important reminder that there is no room for complacency in the quest for effective neuroprotection in cardiac interventions. Although the level of neurological surveillance described in the study by Messé et al. (13) would be challenging for large national registries such as the STS or the Transcatheter Valve Therapies registry, any future prospective study that aims to assess neurological outcomes and risk factors should follow this rigorous example as a matter of course.
The risk factors the ADVANCE study identified for 30-day stroke were female sex, acute kidney injury, and major vascular complications. As the authors have speculated, perhaps female patients with smaller vessels were predisposed to vascular complications, which in turn led to hypotensive episodes and more maneuvering within the atherosclerotic vessels. Renal failure may have been a surrogate marker for a sicker subset of patients, or an underlying kidney-brain interaction may have contributed to the pathophysiology of stroke, as suggested by some previous data (14). The relationship between prior coronary artery bypass grafting and late stroke was likely characterized by more extensive atherosclerotic disease. Patients with atrial fibrillation were at higher risk of a neurological event if stroke and TIA were combined as an outcome. Because patients in this study, as routinely in all TAVR programs, received anticoagulation therapy with a double regimen of aspirin and clopidogrel for 6 months post-TAVR, the question of whether hypotensive episodes played a role in the relationship between AF and neurological events should be considered.
In the quickly developing arena of TAVR, we must also remember that even a year can make a difference, because not only do devices and technology evolve, patient selection by TAVR centers becomes more refined, and alternative TAVR access routes are identified and used with increasing safety and feasibility (15). It will also be extremely interesting to learn how embolic protection devices influence neurological outcomes. Pre-clinical and early clinical data have shown considerable promise for these devices, which aim to reduce the cerebral burden of embolic debris by either capturing or deflecting embolized material from supra-aortic cerebral trunks. With 3 different embolic protection devices already in the clinical phase, we should expect results from larger trials soon (16).
In conclusion, the results from the ADVANCE trial are encouraging. Stroke rates were low, and no particular procedure-related factors appeared to predispose patients to neurological events within the perioperative period. It was clear that the heaviest 30-day stroke burden stemmed from patient-specific factors that have been highlighted by previous studies and are on the clinical radar when a patient’s eligibility for aortic valve interventions is being assessed. However, neurological challenges still form a significant portion of morbidity and mortality after TAVR and SAVR, and their more subtle forms will go undetected if not specifically sought. Future studies, like the evolution of TAVR itself, must maintain a continuous, stringent drive to identify, prevent, and treat neurological dysfunction in this susceptible and often vulnerable patient population.
↵∗ 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. Thourani has served on advisory boards for St. Jude Medical, Edwards Lifesciences, Boston Scientific, and Abbott Medical; and has received research support from Medtronic, Edwards Lifesciences, St. Jude Medical, Direct Flow Medical, and Sorin Medical. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2015 American College of Cardiology Foundation
- Bosmans J.,
- Bleiziffer S.,
- Gerckens U.,
- et al.,
- for the ADVANCE Study Investigators
- Yang T.H.,
- Webb J.G.,
- Blanke P.,
- et al.
- Webb J.,
- Gerosa G.,
- Lefèvre T.,
- et al.
- Messé S.R.,
- Acker M.A.,
- Kasner S.E.,
- et al.,
- for the Determining Neurologic Outcomes from Valve Operations (DeNOVO) Investigators
- Wu V.C.,
- Wu P.C.,
- Wu C.H.,
- et al.,
- for the National Taiwan University Study Group on Acute Renal Failure (NSARF) Group
- Fanning J.P.,
- Walters D.L.,
- Platts D.G.,
- Eeles E.,
- Bellapart J.,
- Fraser J.F.