Patients presenting with severe aortic stenosis and large aortic annuli are challenging to treat because of the size limitations of available transcatheter heart valves. In this study, we aimed to determine clinical and hemodynamic outcomes in patients presenting with large aortic annuli who underwent transcatheter aortic valve implantation (TAVI). Patients from the German Aortic Valve Registry who underwent TAVI either with the Edwards Sapien (ES) or Medtronic CoreValve (MCV) systems from 2011 to 2017 were included. They were further stratified into a large (aortic annulus diameter 26 to 29 mm for ES; 26 to 30 mm for MCV) and extra-large (aortic annulus diameter >29 mm for ES; >30 mm for MCV) group and analyzed using propensity score adjustment. Extra-large was set beyond the sizing limitations according to the manufacturer’s instructions for use. Patients in the large (n = 5,628) and extra-large (n = 509) groups were predominantly male (large: 92.6% vs extra-large: 91.9%). The 30-day mortality was comparable (large: 3.9% vs extra-large: 5.0%, p = 0.458). Procedure duration (large: 78.9 minutes ± 0.82 vs extra-large: 86.4 minutes ± 1.9, p <0.001) was longer in the extra-large group. Likewise, vascular complications (large: 6.2% vs extra-large: 12%, p = 0.002) and the need for a permanent pacemaker implantation (large: 18.8% vs extra-large: 26.0%, p = 0.027) were more often present in the extra-large group. Aortic regurgitation ≥II after valve implantation was numerically higher (large: 3.0% vs extra-large: 5.3%, p = 0.082) in patients with extra-large anatomy. In conclusion , patients with large and extra-large aortic annulus diameters who underwent TAVI have comparable 30-day mortality. Beyond the recommended annulus range, there is a higher risk for vascular complications and permanent pacemaker implantation.
Transcatheter aortic valve implantation (TAVI) has evolved to a cornerstone of cardiovascular care delivery for patients with severe aortic stenosis. , Patients presenting with a large aortic annulus diameter may pose a special task to the interventional team because of the size limitations of transcatheter heart valves. In a respectable number of patients, the aortic annulus diameter is out of range for commercially available valves, and some centers perform off-label valve implantation if surgery is not an option. Although valve hemodynamics and clinical outcomes after TAVI in small aortic annuli have been well researched, , there is only scarce data , on patients presenting with large anatomy who underwent TAVI. Therefore, we aimed to determine the hemodynamic and clinical outcomes in patients presenting either with an extra-large (aortic annulus diameter >29 mm for Edwards Sapien [ES], Edwards Lifesciences, Irvine, California, USA ; >30 mm for Medtronic CoreValve [MCV], Medtronic, Dublin, Ireland) or large (aortic annulus diameter 26 to 29 mm for ES; 26 to 30 mm for MCV) anatomy. Extra-large was set beyond the recommended annulus range for each prosthesis type according to the instructions for use.
Methods
The German Aortic Valve Registry (GARY) is a large, prospective, multicenter (78 sides) registry that monitors quality and outcomes in interventional and surgical aortic valve procedures all across Germany. Initiated in 2010, the all-comers registry collects data up to 5 years after the index procedure. Patient participation is voluntary, and the study design has been described in detail previously. Standardized electronic case report forms are used to collect baseline information and procedural and follow-up data. Patients, participating in the GARY, who underwent either transfemoral or transapical TAVI from January 2011 to December 2017, with an aortic annulus diameter ≥26 mm were found eligible. Valve model and size were at the discretion of the implanting physician. To facilitate comparability, only patients receiving the ES or the MCV system were included in the current analysis. Participants were further stratified into large anatomy within the recommended annulus range (aortic annulus diameter 26 to 29 mm for ES; 26 to 30 mm for MCV) and an extra-large anatomy beyond the recommended annulus range for each prosthesis type (aortic annulus diameter >29 mm for ES; >30 mm for MCV). Propensity score matching with the variables ejection fraction, cardiac decompensation, cardiogenic shock, and aortic valve calcification was performed. Hemodynamic and clinical outcomes of 5,628 patients with large and 509 with extra-large aortic annulus diameter were compared. The patient selection process is also depicted in Figure 1 . In a subanalysis, we investigated the mortality, procedure-related, clinical, and hemodynamic characteristics of patients in the extra-large group with regard to the implanted valve type. For continuous variables, we examined differences with the Mann-Whitney U test or weighted nonparametric rank test and for categorical variables with the chi-square test or univariate weighted logistic regression respectively. To describe the data, we present continuous variables as mean ± SE and categorical variables as frequencies and percentages. Because TAVI valves for patients with large and extra-large anatomy were not randomly assigned to the patients, a weighted propensity score model was used to adjust the comparison of outcome parameters according to baseline characteristics. Variables included were ejection fraction, cardiac decompensation, cardiogenic shock, and aortic valve calcification was performed. Propensity score weighting was done for patients with large anatomy and extra-large anatomy. In addition, a propensity score weighting model was performed for the Edwards and Medtronic valves within the large anatomy and extra-large anatomy. A 2-sided p ≤0.05 was considered statistically significant. Statistical analysis was performed with SAS statistical software version 9.4 (SAS Institute, Cary, North Carolina) and R Software (R Foundation for Statistical Computing, Vienna, Austria). The R packages “twang” and “survey” were used for calculation propensity score weights and the corresponding weighted analysis. All proper ethical oversight was secured. The study is in line with the Declaration of Helsinki. Patients provided written informed consent before the procedure. Initial ethical approval for the GARY registry was obtained from the Freiburg International Ethics Committee and GARY is registered at ClinicalTrials.gov (NCT01165827).
Results
During the observation period from 2011 to 2017, valve implantations in extra-large anatomies increased steadily. Overall, a balloon-expandable ES valve was chosen in 311 patients (61.1%), and 198 patients (38.9%) received an MCV ( Figure 2 ). Patients were in their early eighties (large: 80.4 years vs extra-large: 80.3 years, p = 0.787) and predominantly male (large: 92.6% vs extra-large: 91.9%, p = 0.564). Baseline characteristics after propensity score weighting were comparable, except for the presence of moderate to severe mitral regurgitation (large: 25.7% vs extra-large: 31.9%, p = 0.029) and neurologic dysfunction (large: 17% vs extra-large: 13.2%, p = 0.026) and are listed in Table 1 . In both groups, most of the procedures were performed in an elective setting (large 85.7% vs extra-large: 81.5%, p = 0.070). A balloon-expandable valve was more often implanted in both groups (large: ES 75.2% and MCV 24.8% p <0.001 vs extra-large: ES 61.1% and MCV 38.9%, p <0.001). The transfemoral access route was most often favored (large: 80.8% vs extra-large: 89.6%, p <0.001), followed by a transapical approach (large: 17% vs extra-large: 8.4%, p <0.001). In adjusted analysis, the procedure duration (large: 78.9 minutes ± 0.82 vs extra-large: 86.4 minutes ± 1.9, p <0.001) and radiation time (large: 15.5 minutes ± 0.4 vs extra-large: 18.8 minutes ± 0.8, p <0.001) ( Figure 3) were higher in the extra-large group. Conversion to open heart surgery was necessary in very few cases (large: 0.3% vs extra-large: 0.4%, p = 0.735). Procedure-related complications such as vascular-access site complications (large: 6.2% vs extra-large: 12%, p = 0.002) or the need for a permanent pacemaker implantation (large: 18.8% vs extra-large: 26.8%, p = 0.027) ( Figure 3 ) were more frequent in the extra-large group whereas stroke (large: 0.7% vs extra-large: 1.8%, p = 0.122) or the occurrence of acute kidney injury with new onset of dialysis (large: 4.5% vs extra-large: 5.7%, p = 0.299) did not significantly differ. A mean gradient >14 mm Hg after TAVI was observed in 9.9% of patients in the large group, and 9.5% of the extra-large group (p = 0.858). Moderate to severe aortic regurgitation after valve implantation was numerically more common in the extra-large group (large: 3% vs extra-large: 5.3%, p = 0.084; Figure 3 ). Additional procedures were necessary in 3.2% of patients in the large group and 2.2% of the extra-large group. The additional procedures comprised recapture maneuvers (large: 0.37% vs extra-large: 0.58%, p = 0.311), valve-in-valve procedures (large: 1.15% vs extra-large: 0.98%, p = 0.676) or interventional closure of a paravalvular leakage (large: 0.24% vs extra-large: 0%, p = 0.997). Patients in the extra-large group spend more time in the intensive care unit (large: 2.8 days ± 0.09 vs extra-large: 3.2 days ± 0.2, p = 0.001). The 30-day mortality was comparable (large: 3.9% vs extra-large: 5.0%, p = 0.458). Additional periprocedural, in-hospital, and follow-up data are provided in Table 2 .
| Variable | Large anatomy (n = 5.628) | Extra-large anatomy (n = 509) | p-Value |
|---|---|---|---|
| Body mass index (kg/m²) | 27.63 (0.06) | 27.55 (0.20) | 0.728 |
| Euro-score (%) | 22.32 (0.22) | 22.10 (0.74) | 0.776 |
| STS-score (%) | 5.46 (0.06) | 5.29 (0.20) | 0.422 |
| Previous percutaneous coronary intervention | 1,993 (35.4%) | 196 (38.5%) | 0.163 |
| Previous cardiac surgery | 1,246 (22.1%) | 100 (19.7%) | 0.200 |
| Ney York Heart Association class III + IV | 4,761 (84.6%) | 409 (80.4%) | 0.012 |
| Diabetes on insulin | 772 (41.1%) | 72 (44.7%) | 0.369 |
| Peripheral vascular disease | 1,267 (22.5%) | 104 (20.4%) | 0.278 |
| Chronic renal replacement therapy | 223 (4%) | 15 (2.9%) | 0.256 |
| Previous pacer | 741 (13.7%) | 74 (15%) | 0.393 |
| Previous implanted cardioverter-defibrillator | 201 (4.1%) | 17 (3.7%) | 0.836 |
| Atrial fibrillation | 1,975 (35.1%) | 177 (34.8%) | 0.885 |
| Mean pressure gradient (mm Hg) | 39.847 (0.28) | 39.461 (0.72) | 0.339 |
| Aortic valve area (cm 2 ) | 0.767 (0.00) | 0.777 (0.01) | 0.275 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
