Author + information
- Received January 17, 2017
- Revision received June 9, 2017
- Accepted June 15, 2017
- Published online October 16, 2017.
- Oliver Husser, MD, PhDa,
- Won-Keun Kim, MDb,
- Costanza Pellegrini, MDa,
- Andreas Holzamer, MDc,
- Thomas Walther, MDd,
- Patrick N. Mayr, MDe,
- Michael Joner, MDa,
- Albert M. Kasel, MDa,
- Teresa Trenkwalder, MDa,
- Jonathan Michel, MBBS, BSca,
- Tobias Rheude, MDa,
- Adnan Kastrati, MDa,f,
- Heribert Schunkert, MDa,f,
- Christof Burgdorf, MDa,g,
- Michael Hilker, MDc,
- Helge Möllmann, MDb and
- Christian Hengstenberg, MDa,f,∗ ()
- aKlinik für Herz und Kreislauferkrankungen, Deutsches Herzzentrum München, Technical University Munich, Munich, Germany
- bDepartment of Cardiology, Kerckhoff Heart and Lung Center, Bad Nauheim, Germany
- cKlinik für Herz, Thorax, und Herznahe Gefäßchirurgie, University of Regensburg Medical Center, Regensburg, Germany
- dDepartment of Cardiovascular Surgery, Kerckhoff Heart and Lung Center, Bad Nauheim, Germany
- eInstitut für Anästhesiologie, Deutsches Herzzentrum München, Technical University Munich, Munich, Germany
- fDeutsches Zentrum für Herz- und Kreislauf-Forschung (German Center for Cardiovascular Research), Partner Site Munich Heart Alliance, Munich, Germany
- gHerz und Gefäßzentrum Bad Bevensen, Klinik für Kardiologie, Bad Bevensen, Germany
- ↵∗Address for correspondence:
Prof. Dr. med Christian Hengstenberg, Klinik für Herz- und Kreislauferkrankungen–Deutsches Herzzentrum München, Technische Universität München, Lazarettstrasse 36, 80636 Munich, Germany.
Objectives This study sought to compare 2 next-generation transcatheter heart valves (THV), the self-expanding ACURATE neo (NEO) and the balloon-expandable SAPIEN 3 (S3), in terms of device failure and early safety at 30 days.
Background Deployment of these THV showed promising initial clinical results. However, no comparative data are available.
Methods Of 1,121 treated patients at 3 centers, a 1-to-2 nearest neighbor matching was performed to identify 2 patients treated with S3 (n = 622) for each patient treated with NEO (n = 311).
Results In-hospital complications were comparable between NEO and S3, including stroke (1.9% vs. 2.4%; p = 0.64), major vascular complications (10.3% vs. 8.5%; p = 0.38), or life-threatening bleeding (4.2% vs. 3.7%; p = 0.72). Device failure with NEO was comparable with S3 (10.9% vs. 9.6%; odds ratio: 1.09 [95% confidence interval: 0.69 to 1.73]; p = 0.71) with more paravalvular leakage (PVL II+, 4.8% vs. 1.8%; p = 0.01), but less elevated gradients (≥20 mm Hg, 3.2% vs. 6.9%; p = 0.02) and pacemaker implantations (9.9% vs. 15.5%; p = 0.02). Thirty-day mortality (2.3% vs. 1.9%; p = 0.74) and the early safety composite endpoint (15.8% vs. 15.6%; hazard ratio: 0.97 [95% confidence interval: 0.68 to 1.39]; p = 0.88) were similar with NEO and S3.
Conclusions Very high success rates were achieved for both valves, and the clinical and procedural results were comparable. Compared with S3, NEO was associated with less new pacemaker implantations and less elevated gradients, but with more paravalvular leakage.
The deployment of transcatheter heart valves (THV) in transcatheter aortic valve replacement (TAVR) is performed by self-expanding or balloon-expandable technologies in most available devices. Both technologies have been used in large clinical registries with very good clinical outcomes (1,2). Potential advantages of self-expanding technology include the possibility to reposition and to resheath, the option to retrieve, and a higher conformability to the individual aortic annular anatomy. Balloon-expandable THV are not repositionable, but may have advantages in calcified anatomy because of a higher radial force. The only randomized comparison between both technologies in earlier-generation devices showed superiority for device success with balloon-expandable over self-expanding THV (3); however, clinical outcome at 1 year was equivalent (4).
Since then, a considerable development of balloon-expandable and self-expanding THV has taken place to address limitations of earlier-generation devices, such as paravalvular leakage (PVL), new permanent pacemaker implantations (PPI), and vascular complications. Among these novel devices are the balloon-expandable SAPIEN 3 (Edwards Lifesciences, Irvine, California) and the self-expanding ACURATE neo (SYMETIS S.A., Ecublens, Switzerland). Initial clinical results for SAPIEN 3 (5,6) and ACURATE neo (7) are promising, but no comparative data of both THV are available.
Therefore, in this multicenter study from 3 centers in Germany, we performed a propensity-matched comparison of ACURATE neo and SAPIEN 3 in terms of device failure and early safety at 30 days according to the updated criteria of the Valve Academic Research Consortium (VARC-2) (8).
Between January 2014 and January 2016, a total of 1,121 consecutive patients with symptomatic, severe stenosis of the native aortic valve were treated with transfemoral TAVR using ACURATE neo (n = 311) or SAPIEN 3 (n = 810) at 3 centers in Germany (Department of Cardiology, Deutsches Herzzentrum München, Munich; Kerckhoff Heart and Lung Center, Bad Nauheim; and University of Regensburg Medical Center, Regensburg) (Figure 1, Online Figure 1). The interdisciplinary heart team discussed all cases and consensus was achieved regarding the therapeutic strategy. All patients provided written informed consent for the procedures.
Multislice computed tomography data analysis
Multislice computed tomography was performed as part of the standard pre-procedural screening protocol. Aortic annulus measurements were assessed in multiple plane reconstructions according to the guidelines of the Society of Cardiovascular Computed Tomography (9). In short, area and perimeter of the virtual aortic annulus were obtained by direct planimetry and the minimum and maximum diameters were assessed. The eccentricity index was calculated and an eccentric annulus was assumed for an eccentricity index >0.25 (3). Calcification of the valvular apparatus was visually graded and dichotomized as mild/moderate versus severe. The final decision on prosthesis type and size was left at the discretion of the treating physician.
The ACURATE neo/TF (Online Figure 2A) is available in 3 sizes (small, medium, and large) and the technical features have been described elsewhere (7). The device consists of a self-expanding nitinol frame with a porcine pericardial leaflet valve in a supra-annular position and a pericardial sealing-skirt on the outer and inner surface of the stent body. The ACURATE neo is transfemorally delivered using the ACURATE neo/TF Delivery System compatible with a 15-F to 18-F catheter sheath (internal diameter).
The balloon-expandable SAPIEN 3 (Online Figure 2B) consists of a cobalt chromium alloy frame with bovine pericardial leaflets and is delivered with the Commander delivery system. At the time of the study period, the SAPIEN 3 is available in 23-, 26-, and 29-mm sizes and features an external polyethylene terephthalate fabric seal to reduce PVL, as has been previously described in detail (10,11). The technical features of the respective THV, associated delivery systems, sheath dimensions, and sizing recommendations are summarized in Online Table 1.
Definition of endpoints and follow-up
The endpoints of this study were device failure and the early safety composite endpoint at 30 days according to the VARC-2 criteria (8). Follow-up to 30 days was prospectively collected at each of the participating sites in the outpatient clinic, contacting the primary care physician, or by direct contact with the patient. Transthoracic echocardiography was performed at baseline, before discharge, and at 30 days.
Continuous variables are expressed as the mean with the SD or the median with the interquartile range and compared using Student t test or the Mann-Whitney U test, respectively.
To reduce imbalance in patient baseline characteristics and the effect of a potential selection bias on both endpoints for comparing ACURATE neo with SAPIEN 3, propensity matching was performed using R version 3.2.3 (The R Foundation for Statistical Computing, Vienna, Austria) and the package “MatchIt” (12). In short, a 1-to-2 nearest neighbor matching was used to identify 2 control cases treated with SAPIEN 3 (n = 622) for each case treated with ACURATE neo (n = 311). A 1-to-2 matching was chosen to minimize play of chance in the control group and thus to increase statistical power among the control subjects. Baseline, electrocardiogram, and multislice computed tomography characteristics with known prognostic impact or showing significant univariate differences between both groups were included in the matching algorithm. Figure 1 summarizes the study flow and variables used for propensity matching.
The occurrence of VARC-2 defined in-hospital complications was calculated for each group. The odds ratios with 95% confidence intervals (CIs) for device failure were computed using logistic regression analysis. Follow-up at 30 days was complete for 99% (310 of 311) of patients treated with ACURATE neo and for 99% (803 of 810) of patients treated with SAPIEN 3. The 30-day event rates were estimated using the Kaplan-Meier method and compared with the log-rank test. A Cox proportional regression with computation of hazard ratios and the 95% CI for the risk of the early safety composite endpoint was performed. To correct for a center-specific influence, all analyses were stratified by center and the interaction between THV and center was tested.
All analyses were conducted in the matched population and in the subgroups of heavily calcified anatomy and eccentric annulus. An analysis of the entire unmatched population is provided in the Online Appendix, because results were comparable with the matched sample.
A 2-sided p value of <0.05 was considered statistically significant for all analyses. STATA version 13.0 (STATA Corp., College Station, Texas) and R version 3.2.3 (The R Foundation) were used for analyses.
Baseline characteristics and propensity matching
Baseline characteristics of the unmatched population are displayed in Table 1. Compared with patients treated with SAPIEN 3, patients treated with ACURATE neo were more frequently female (60.8% vs. 45.4%; p < 0.001), were more symptomatic (New York Heart Association functional class III/IV: 82.3% vs. 74.8%; p = 0.008), and had a lower prevalence of depressed left ventricular function (<35%; 5.8% vs. 9.8%; p = 0.034). Patients treated with SAPIEN 3 had significantly larger aortic annular anatomy and a higher percentage of severely calcified aortic annulus compared with patients treated with ACURATE neo.
After 1-to-2 matching for variables summarized in Figure 1, no baseline differences between the 2 groups were present (Table 1) except for aortic annular area remaining statistically larger in patients treated with SAPIEN 3 (4.5 ± 0.8 cm2 vs. 4.4 ± 0.6 cm2; p = 0.003). There were no unmatched patients treated with ACURATE neo (n = 311) and 622 control cases treated with SAPIEN 3. Online Figure 3 provides further information on the distribution of the propensity score across treatment and control cases before and after matching.
Procedural data and in-hospital outcome
Procedural characteristics and in-hospital complications are displayed in Table 2 (see Online Table 2 for unmatched population). Approximately one-half of both groups underwent TAVR in conscious sedation (ACURATE neo 47.3% vs. SAPIEN 3 46.0%; p = 0.710). Pre-dilatation was more frequently performed with ACURATE neo (95.8% vs. 74.3%; p < 0.001). The small, medium, and large sizes of ACURATE neo were used in 30.9%, 40.2%, and 28.9% of the cases, whereas the 23 mm, 26 mm, and 29 mm of SAPIEN 3 were used in 43.9%, 41.6%, and 14.5%, respectively. Fluoroscopy time was shorter and more contrast was used with ACURATE neo compared with SAPIEN 3. Post-dilatation was more often performed with ACURATE neo compared with SAPIEN 3 (42.1% vs. 23.8%; p < 0.001).
There was no difference in stroke, major vascular complications, life-threatening bleeding, renal failure, or myocardial infarction between both groups (Table 2). Patients treated with SAPIEN 3 had a higher rate of new PPI (15.5% vs. 9.9%; p = 0.024).
Device failure according to VARC-2
Table 3 and Figure 2A show the rate of device failure for ACURATE neo and SAPIEN 3 (see Online Table 3 and Online Figure 4 for unmatched population). There was no significant difference between both groups (10.9% vs. 9.6%; odds ratio: 1.09; 95% CI: 0.69 to 1.73; p = 0.708) and for the subgroups of heavily calcified or eccentric annuli. There was no significant interaction with THV type and center for device failure (p for interaction 0.196). Procedural mortality was 1.0% for ACURATE neo and 0.3% for SAPIEN 3 (p = 0.340). Although the incidence of PVL II+ was higher with ACURATE neo (4.8% vs. 1.8%; p = 0.008), the rate of elevated gradients (≥20 mm Hg) was lower (3.2% vs. 6.9%; p = 0.021) resulting in a similar intended performance of both THVs.
Thirty-day outcome and early safety composite endpoint according to VARC-2
Table 4 and Figures 2B and 3 show comparable rates of the early safety composite endpoint at 30 days for ACURATE neo and SAPIEN 3 (15.8% vs. 15.6%; hazard ratio: 0.97; 95% CI: 0.68 to 1.39; p = 0.879) (see Online Table 4 and Online Figure 5 for unmatched population). There was no difference in the individual contributors to the early safety composite endpoint at 30 days. The 30-day all-cause mortality was 2.3% for ACURATE neo and 1.9% for SAPIEN 3 (p = 0.742). The rate of new PPI at 30 days was significantly higher with SAPIEN 3 compared with ACURATE neo (16.4% vs. 10.2%; p = 0.018). Figure 2B shows early safety at 30 days with both THV and the hazard ratio according to the use of ACURATE neo in the matched population and in patients with severely calcified or eccentric annuli. In the case of eccentric annuli use of ACURATE neo was associated with a significantly higher hazard for the early safety composite endpoint at 30 days compared with SAPIEN 3 (hazard ratio: 2.24; 95% CI: 1.02 to 4.93; p = 0.044). There was no significant interaction with THV type and center for the early safety composite endpoint (p for interaction = 0.498).
Mean transvalvular gradients decreased after TAVR with both THVs, but were significantly lower with ACURATE neo compared with SAPIEN 3, before discharge (9 ± 5 mm Hg vs. 13 ± 5 mm Hg; p < 0.001) and at 30 days (8 ± 4 mm Hg vs. 12 ± 5 mm Hg; p < 0.001) (Figure 4A). Lower gradients with ACURATE neo were also observed across different prosthesis sizes of each THV (Figures 4B to 4D).
Of a total of 1,121 patients undergoing TAVR from 3 large German centers, we performed a propensity-matched comparison of the novel balloon-expandable SAPIEN 3 with the self-expanding ACURATE neo in terms of device success and 30-day outcome. The main finding is that there was no difference of both endpoints between both THV. Although ACURATE neo was associated with a higher rate of PVL compared with SAPIEN 3, the incidence of elevated gradients and new PPI was significantly lower.
Device failure according to VARC-2
Device success is an important measure of procedural success and proper prosthesis function in TAVR (8). With ACURATE neo, device success rates have been reported in 95% (7). In the present study, in 311 patients treated with NEO, we observed a device success rate of 89%, which was comparable with SAPIEN 3 (90%). Data on device success with SAPIEN 3 are scarce, because the PARTNER II SAPIEN 3 trial (5) did not report this measure. A single-center study using SAPIEN 3 reported device success rates up to 97.6% (6).
Residual PVL after TAVR is associated with long-term mortality (13). Both SAPIEN 3 and ACURATE neo feature sealing skirts to minimize PVL. Consistent with previous studies using the SAPIEN 3 reporting PVL II+ in about 3.5% (5,6,11), we found a very low incidence of 1.8%. Although in the present study the incidence of more than mild PVL was significantly higher with ACURATE neo than with SAPIEN 3, the rate of elevated gradients was higher with SAPIEN 3, resulting in comparable rates of device success.
Randomized data comparing self-expanding and balloon-expandable technology only exist for older-generation devices (Medtronic CoreValve [Minneapolis, Minnesota] vs. Edwards Lifesciences SAPIEN XT [Irvine, California]) used in the CHOICE trial (3). In this trial, device success was significantly lower for the self-expanding device, namely 77% versus 96%. With next-generation THV we found no significant difference in device success between self-expanding and balloon-expandable devices in the matched population or in the subgroups with heavily calcified and eccentric aortic annuli. This finding may argue against a common perception of self-expanding THV being more effective in anatomically challenging patients or inferior in heavy calcification.
VARC-2-defined early safety composite endpoint at 30 days
The early safety composite endpoint at 30 days has been proposed by the VARC for the assessment of patient safety in TAVR summarizing important measures of complications, prosthesis function, and mortality. Two recent meta-analyses in earlier-generation self-expanding and balloon-expandable devices reported freedom from the early safety composite endpoint at 30 days in about 75% of the cases (14). However, data regarding this endpoint especially for newer-generation THV are limited.
In the present propensity-matched comparison, clinical results with ACURATE neo were comparable with SAPIEN 3, with freedom of events at 30 days in 84% of the cases in both groups. In the ACURATE neo CE mark study and SAVI registry, freedom from the composite endpoint was 84% and 91%, respectively (15). Regarding the SAPIEN 3, data on early safety are scarce. Earlier results from a single European center have reported freedom of events in 90% of patients treated with SAPIEN 3 (6). U.S. data on this endpoint are not available, because this endpoint was not reported in the recently published PARTNER II SAPIEN 3 trial (5).
With the development of next-generation transfemoral THV, delivery systems have been optimized to minimize vascular complications and bleeding. Major vascular complications and bleedings have been reported in 3.8% and 1.5% with ACURATE neo (15) and in about 4% to 6% and 4% with SAPIEN 3 (6,16). ACURATE neo is compatible with delivery systems of 15F to 18F catheter inner diameter and a corresponding outer diameter of 22-F catheter. These dimensions are comparable with the introducer set of the SAPIEN 3 that uses a 14-F to 16-F catheter inner diameter sheath expanding to 24-F to 27-F catheter during passing of the crimped prosthesis. Accordingly, we found no difference between ACURATE neo and SAPIEN 3 for major vascular complications and for life-threatening bleeding.
In the present study, median hospital stay after TAVR was 7 days (5 to 10 days), which is in line with recent studies reporting similar duration of stay after TAVR with next-generation THV (16,17). However, these data stand in contrast to results from the PARTNER 2 trial where median hospital stay was 3 days (2 to 6 days) (5). One explanation for this observation may be that in Germany health care providers’ regulations offer little incentive for early discharge.
The incidence of 30-day clinically apparent major stroke after TAVR ranges around 3% (18). Stroke has been reported in 2% of cases with ACURATE neo (15,19) and in 2.7% with SAPIEN 3 (20). In the present study, we observed a rate of 2.3% for ACURATE neo with no difference in the matched SAPIEN 3 population (3.1%).
New PPI after TAVR using earlier-generation ranged between 5% and 12% for balloon-expandable and 28% with self-expanding devices (21). In the case of SAPIEN 3, recent studies have reported rates of new PPI in 11.6% to 16% (5,20,22) and up to 19.7% at 1 year (22). The present study shows comparable rates with the SAPIEN 3 of 16.4% in the matched population. With ACURATE neo, new PPI was reported in 8.2% of the cases (15). In this study using ACURATE neo, this rate was 10.2%, which was significantly lower compared with SAPIEN 3. This stands in contrast to randomized data of earlier-generation devices, where the rate of new PPI was significantly higher with self-expanding compared with balloon-expandable devices (3). A possible explanation may be that ACURATE neo uses a supra-annular design with potentially less interference with the cardiac conduction system.
This is an observational study without core-laboratory analysis of procedural results and center-independent adjudication of outcomes. Despite propensity matching, the influence of unknown confounders cannot be excluded. Additionally, this registry-based study includes patients treated with the ACURATE neo and SAPIEN 3 according to the heart teams’ decision and not according to predefined selection criteria as in a randomized comparison.
The clinical experience of 3 German high-volume centers with 2 latest-generation balloon-expandable and self-expanding THV was examined. In a propensity-matched comparison, we found equivalent rates of device failure and the early safety composite endpoint with ACURATE neo and SAPIEN 3. Use of ACURATE neo may be associated with a higher rate of PVL compared with SAPIEN 3; however, the incidence of elevated gradients and new PPIs may be significantly lower. Future studies are required to study the true effectiveness of ACURATE neo versus balloon-expandable THV, such as the SAPIEN 3, and other self-expanding devices, such as the Evolut R (Medtronic).
WHAT IS KNOWN? To date, no data exist comparing the latest-generation self-expanding Symetis ACURATE neo and the balloon-expandable Edwards SAPIEN 3 transcatheter heart valves.
WHAT IS NEW? In this multicenter, propensity-matched comparison, we found equivalent rates of device failure and the VARC-2 defined early safety composite endpoint. Use of ACURATE neo was associated with a higher rate of paravalvular leakage compared with SAPIEN 3. However, the incidence of elevated gradients and new PPIs was significantly lower.
WHAT IS NEXT? The SCOPE I trial, a prospective, multicenter, randomized clinical trial, comparing the ACURATE neo and the SAPIEN 3 is ongoing and will determine the true effectiveness of each THV.
Dr. Husser has received minor travel grants from Edwards Lifesciences and Symetis S.A.; and proctor fees from Symetis S.A. Dr. Kim is a proctor for and has received minor lecturing fees from Symetis S.A. and St. Jude Medical. Dr. Pellegrini has received minor travel grants from Edwards Lifesciences. Dr. Kasel has received proctor fees and speaker honoraria from Edwards Lifesciences. Dr. Burgdorf has received proctor fees from Symetis S.A. Dr. Hilker has received proctor fees from Symetis S.A. Dr. Möllmann has received proctor fees and speaker honoraria from Edwards Lifesciences, Symetis S.A., and St. Jude Medical. Dr. Hengstenberg has received proctor fees and speaker honoraria from Edwards Lifesciences and Symetis S.A. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- confidence interval
- permanent pacemaker implantations
- paravalvular leakage
- transcatheter aortic valve replacement
- transcatheter heart valve
- Valve Academic Research Consortium
- Received January 17, 2017.
- Revision received June 9, 2017.
- Accepted June 15, 2017.
- 2017 American College of Cardiology Foundation
- Abdel-Wahab M.,
- Neumann F.-J.,
- Mehilli J.,
- et al.
- Husser O.,
- Pellegrini C.,
- Kessler T.,
- et al.
- Kappetein A.P.,
- Head S.J.,
- Généreux P.,
- et al.
- Achenbach S.,
- Delgado V.,
- Hausleiter J.,
- Schoenhagen P.,
- Min J.K.,
- Leipsic J.A.
- Binder R.K.,
- Rodés-Cabau J.,
- Wood D.A.,
- et al.
- Webb J.,
- Gerosa G.,
- Lefèvre T.,
- et al.
- Ho D.,
- King G.,
- Imai K.,
- Stuart E.
- Généreux P.,
- Head S.J.,
- Van Mieghem N.M.,
- et al.
- Möllmann H.
- Pilgrim T.,
- Stortecky S.,
- Nietlispach F.,
- et al.
- Wendler O.,
- Schymik G.,
- Treede H.,
- et al.
- Siontis G.C.M.,
- Jüni P.,
- Pilgrim T.,
- et al.
- Herrmann H.C.,
- Thourani V.H.,
- Kodali S.K.,
- et al.