Objective The onset of new conduction abnormalities requiring permanent pacemaker implantation (PPI) after transcatheter aortic valve implantation (TAVI) is still a relevant adverse event. The main objective of this registry was to identify modifiable procedural risk factors for an improved outcome (lower rate of PPI) after TAVI in patients at high risk of PPI.
Methods Patients from four European centres receiving a balloon-expandable TAVI (Edwards SAPIEN 3/3 Ultra) and considered at high risk of PPI (pre-existing conduction disturbance, heavily calcified left ventricular outflow tract or short membranous septum) were prospectively enrolled into registry.
Results A total of 300 patients were included: 42 (14.0%) required PPI after TAVI and 258 (86.0%) did not. Patients with PPI had a longer intensive care unit plus intermediate care stay (65.7 vs 16.3 hours, p<0.001), general ward care stay (6.9 vs 5.3 days, p=0.004) and later discharge (8.6 vs 5.0 days, p<0.001). Of the baseline variables, only pre-existing right bundle branch block at baseline (OR 6.8, 95% CI 2.5 to 18.1) was significantly associated with PPI in the multivariable analysis. Among procedure-related variables, oversizing had the highest impact on the rate of PPI: higher than manufacturer-recommended sizing, mean area oversizing as well as the use of the 29 mm valve (OR 3.4, 95% CI 1.4 to 8.5, p=0.008) all were significantly associated with PPI. Rates were higher with the SAPIEN 3 (16.1%) vs SAPIEN 3 Ultra (8.5%), although not statistically significant but potentially associated with valve sizing. Implantation depth and postdelivery balloon dilatation also tended to affect PPI rates but without a statistical significance.
Conclusion Valve oversizing is a strong procedure-related risk factor for PPI following TAVI. The clinical impact of the valve type (SAPIEN 3), implantation depth, and postdelivery balloon dilatation did not reach significance and may reflect already refined procedures in the participating centres, giving attention to these avoidable risk factors.
Trial registration number NCT03497611.
- aortic valve stenosis
- transcatheter aortic valve replacement
- pacemaker, artificial
Data availability statement
Data are available upon reasonable request. Data are available from the corresponding author upon request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
WHAT IS ALREADY KNOWN ON THIS TOPIC
While a number of registries have documented the rates of permanent pacemaker implantation (PPI) with different valves, much less evidence has been provided for the procedural variables associated with the need for PPI after transcatheter aortic valve implantation (TAVI).
WHAT THIS STUDY ADDS
Valve oversizing has been found to be strongly associated with PPI following TAVI. The clinical impact of the valve type (SAPIEN 3), implantation depth, and postdelivery balloon dilatation did not reach significance and may indicate improvements in the procedures performed at the participating centres considering these avoidable risk factors.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
Determining factors associated with an increased likelihood of PPI allows the optimisation of preprocedural and intraprocedural care, potentially decreasing the incidence rates of postoperative PPI.
Within the last decade, transcatheter aortic valve implantation (TAVI) has established itself to be a standard therapy in moderate-risk and high-risk patients with symptomatic aortic stenosis (AS) and recently even non-inferior to surgical aortic valve replacement in low-risk patients.1–6 A new generation of balloon-expandable transcatheter heart valves (THVs), such as Edwards SAPIEN 3/3 Ultra, has been shown to reduce periprocedural complications (reduced moderate or severe paravalvular regurgitation and major vascular complications) and improve 30-day and 12-month outcomes.7 However, despite technological advances, permanent pacemaker implantation (PPI) remains a potential postprocedural complication after TAVI, leading to reduced survival, increased costs, and prolonged hospital stay.8
To date, several reports on PPI rates and predictors associated with the use of the Edwards SAPIEN 3 THV have been published.9–17 Pre-existing conduction disturbance, aortic valve calcification, greater implantation depth/oversizing, and several other anatomical and procedural factors, including heavily calcified left ventricular outflow tract (LVOT), complete right bundle branch block (RBBB), prolonged QRS duration, and short membranous septum, have been associated with the need for PPI after SAPIEN 3 THV TAVI. Maeno et al used these data to develop a scoring algorithm to predict PPI post TAVI, including non-coronary cusp (NCC) device-landing zone calcium volume, RBBB, short membranous septum length and ventricular implantation depth as predictive factors.16 In a recent study examining PPI rates of balloon‐expandable and self-expanding THV in a large cohort of patients with AS, both RBBB and left bundle branch block were identified as predictors for PPI after TAVI, in addition to age, sex, history of hypertension, obesity, diabetes mellitus, and myocardial infarction, among others.17
Nevertheless, most of the current evidence is derived from single-centre series with a lack of a consistent definition of potentially associated variables and the unexplained differences in the number and type of identified variables. In the previously published retrospective analysis by our group, several baseline characteristics, including older age, pulmonary hypertension, prolonged QRS duration, bradycardia, first-degree atrioventricular (AV) block, left anterior hemiblock (LAHB), RBBB as well as higher calcium volume in LVOT were significantly associated with PPI following TAVI.18
In the present prospective, multicentre analysis of patients at high risk of PPI based on their disease characteristics, we aimed to identify potentially modifiable procedural variables that can reduce the risk of PPI in high-risk patients.
CONDUCT is a prospective, multicentre, observational registry of patients undergoing balloon-expandable TAVI at four European institutions: (1) Heart and Diabetes Centre Nordrhein-Westfalen, Germany; (2) Academic Medical Center (AMC), Amsterdam University, The Netherlands; (3) University Hospital Tübingen, Germany; and (4) Linköping University Hospital, Sweden. Site selection was based on the site’s prior experience with TAVI and the recommendations of the steering committee.
The primary objective was to identify procedural variables that can be adjusted to result in an improved outcome (lower rate of PPI) of TAVI in patients with a high risk of procedure-related PPI. The secondary objectives were to verify the results for increased risk of PPI developed in the retrospective phase and to determine PPI-related and overall clinical outcomes in the high-risk cohort of patients.
A high-risk cohort of patients was formed based on the results from the retrospective analysis of patients undergoing TAVI at our institutions.18 Patients undergoing transfemoral TAVI with a balloon-expandable device due to AS and having at least one of the previously identified risk factors for PPI, including pre-existing conduction disturbance (RBBB, LAHB, atrioventricular (AV) block, prolonged QRS duration and bradycardia), heavily calcified LVOT or short membranous septum, were prospectively included in our registry. Patients with a prior pacemaker, indication for PPI prior to TAVI, and valve-in-valve implantation were excluded from the registry. The decision to perform TAVI was made by the local heart team and conducted according to the local protocol.
Documentation and endpoints
Patient demographics, medical history, symptoms, surgical risk scores (EuroScore II and Society of Thoracic Surgeons risk score), and echocardiographic and ECG parameters were recorded at baseline. Periprocedural details and postprocedural outcomes, including device success and complications, were also reported. In addition, data on the time until pacemaker implantation, underlying rhythm disturbances, and in-hospital outcomes were collected. All data were subject to automatic checks for plausibility and completeness.
A routine multislice CT scan performed prior to TAVI was obtained for each patient. The type of scanner, transverse slice thickness, acquisition time, electrographic R–R interval threshold for scanning initiation, and tube current/voltage were specific to each site. Contrast agent was used at all centres and calcium quantification was performed by a single designated, experienced core lab in Tübingen (Germany) as described previously.18
Implantation depth was assessed relative to the annular plane. Data were classified as the percentage of valve being below the annulus.14 15 Sizing classifications were determined by the physiological aortic annulus area in relation to the nominal area range of the implanted valve declared by the manufacturer. Oversizing was defined as physiological annulus area smaller than minimum area range of the valve size. Undersizing was defined as the physiological annulus area bigger than the maximum area range of the implanted valve size.
Pseudonymised data and CT scans were sent to the Institute for Pharmacology and Preventive Medicine (Cloppenburg, Germany). Missing data were requested but were not imputed in case of no provision. The analysis was based on available data only.
For the analysis, patients were stratified into those receiving PPI versus those not receiving PPI after TAVI. Data were analysed using descriptive statistics, with categorical variables presented as absolute values and frequencies (%) and the continuous variables presented as means±SD and/or median (IQR). Group comparisons were carried out using t-test or Mann-Whitney U test for continuous variables as appropriate, depending on distribution, and Fisher exact or χ2 test for categorical variables. Test for normal distribution was carried using Kolmogorov-Smirnov test.
ORs were calculated by logistic regression. Multivariable logistic regression analysis was performed to identify baseline and procedural parameters being associated with the outcome (PPI after TAVI). In the multivariable analysis, all other baseline variables of potential interest (p<0.05; diabetes, chronic obstructive pulmonary disease (COPD), QRS ≥110 ms, AF, RBBB complete, distance to left coronary artery (LCA) <median, aortic annulus calcification NCC >median) were included. Despite a p-value lower than 0.05, pulmonary artery pressure (PAP) was not included due to several missing data. For the multivariable model for the procedural parameters, all baseline variables plus as procedural variable of interest with p<0.05, valve type SAPIEN 3, were included.
All statistical analyses were performed using SPSS V.26.0. A p-value of <0.05 was considered significant.
Between January 2019 and November 2021, a total of 300 patients (171 in Germany, 80 in The Netherlands and 49 in Sweden) were documented: 42 (14.0%) of patients required PPI after TAVI and 258 (86.0%) patients did not (figure 1). In a total of 42 patients who had PPI, 36 (85.7%) had it before discharge and 6 (14.3%) after discharge with a median of 5 (IQR 1.8–7.0) days after TAVI. The major reason for PPI was AV block grade III (n=34, 80.9%), followed by AV block grade II (type Mobitz) (n=2, 4.8%) and sinoatrial block (n=2, 4.8%) (table 1).
Identification of patient baseline characteristics associated with PPI
Patients had a mean age of 79.9±6.3 years and 75.0% were female. In comparison to patients without PPI after TAVI, patients undergoing PPI had higher rates of diabetes (42.9% vs 26.0%, p=0.024) and COPD (16.7% vs 6.2%, p=0.028) (table 2). Further, patients with PPI had lower left ventricular ejection fraction (47.5% vs 51.2%, p=0.014) and higher systolic PAP (48.8 vs 39.5 mm Hg, p=0.010).
Differences in ECG-based characteristics included higher rates of atrial fibrillation (33.3 vs 19.5%, p=0.042), longer QRS duration (141.1 vs 116.8 ms, p<0.001) and presence of a complete RBBB (64.3 vs 17.5%, p<0.001) (table 3). In addition, results from CT data showed that pacemaker-dependent patients had a significantly shorter distance from the aortic annulus plane to the LCA (12.6 vs 14.0 mm, p=0.035) and a lower calcium volume of the NCC (289 vs 433 mm3, p=0.044) than patients without a pacemaker (table 4).
In the univariable regression analysis, diabetes (OR 2.1, 95% CI 1.1 to 4.2, p=0.027), COPD (OR 3.0, 95% CI 1.2 to 7.9, p=0.023), atrial fibrillation (OR 2.1, 95% CI 1.0 to 4.2, p=0.045), QRS duration ≥110 ms (OR 5.2, 95% CI 2.1 to 12.7, p<0.001), complete RBBB (OR 8.4, 95% CI 4.1 to 17.1, p<0.001) and aortic calcium volume of NCC >median (OR 0.4, 95% CI 0.2 to 0.9, p=0.030) were independently associated with PPI after TAVI (table 5). However, after including these parameters in the multivariable analysis, only complete RBBB (OR 6.8, 95% CI 2.5 to 18.1, p<0.001) emerged as an independent baseline predictor of PPI (table 5).
Identification of procedural variables associated with PPI (adjusted for differences in patient characteristics)
Patients with PPI post-TAVI received larger sized valves (27.3%±2.2% vs 26.6%±2.1%, p=0.027) than patients without PPI with 14.3% vs 17.1% receiving 23 mm, 28.6% vs 47.3% receiving 26 mm, and 57.1% vs 35.7% receiving 29 mm, respectively (p=0.026) (table 6). Mean area oversizing was 4.0%±19.6% in patients with PPI compared with 0.0%±17.3% in those without PPI (p=0.255). Based on the manufacturer’s (Edwards) sizing recommendations, the prosthesis was oversized in 17.6% vs 11.7%, within normal sizing range in 61.8% vs 47.7%, and undersized in 20.6% vs 40.5% patients with and without a need for PPI, respectively, with a significant difference between the three sizing groups (p=0.029) (table 6). Notably, only 8.5% (7/82) of patients receiving SAPIEN 3 Ultra THV needed PPI compared with 16.1% (35/218) of patients receiving SAPIEN 3. It is important to note, however, that the distribution of valve sizes was different; more patients with SAPIEN 3 received a 29 mm valve than those with SAPIEN 3 Ultra (47.8% vs 14.6%). The depth of implantation, which is known to be strongly associated with PPI, did not reach significance in our analysis (mean NCC, right coronary cusp, LCC >median; 63.9% vs 47.3%, p=0.068). Patients with PPI tended to have a longer duration of the procedure (62.3 vs 49.9 min, p=0.065), although the difference was not significant. The difference in postdelivery balloon dilatation rates was also not significant between patients with and without PPI in our cohort (4.8% vs 2.0%, p=0.264).
The univariable analysis of procedure-related variables indicated only the valve size 29 mm (OR 2.4, 95% CI 1.2 to 4.7, p=0.009) to be an independent predictor of PPI after TAVI. In the multivariable analysis, valve size of 29 mm (OR 3.4, 95% CI 1.4 to 8.5, p=0.008) remained the only independent predictor of PPI.
Postprocedural outcomes and hospitalisation
Differences in postprocedural outcomes, such as device success or severe periprocedural complications, did not reach statistical significance (table 7). Patients undergoing PPI had a delayed discharge (7 days vs 5 days, p<0.001), longer intensive care unit plus intermediate care stay (46 hours vs 2 hours, p<0.001) and general ward care stay (7 days vs 5 days, p=0.004) than those who did not receive PPI (table 7).
In the current study, the major findings are as follows: (1) larger valve size is an independent procedural predictor of PPI after TAVI in high-risk patients with AS; (2) valve type (SAPIEN 3), higher implantation depth and postdelivery balloon dilatation may be potentially associated with PPI and should be carefully considered; (3) pre-existing RBBB is an independent patient-related variable associated with PPI after TAVI; and (4) patients undergoing PPI tend to have worse hospitalisation characteristics, including longer hospital stay and delayed discharge.
The association between the onset of new AV conduction disturbance requiring PPI after TAVI is well known, although the rates have decreased with new-generation balloon-expandable prostheses.19 In the present population of 300 patients undergoing transfemoral TAVI due to AS and identified for being at a high risk of PPI, the incidence of PPI after TAVI was 14.0%, which is within the range reported for patients undergoing TAVI with the new-generation SAPIEN 3 device (4 to 20%).19 However, since we included only TAVI patients who had at least one risk factor for PPI, we expected the incidence of PPI to be considerably higher. Such a low number of patients with PPI post TAVI may indicate the procedural and technical improvements in TAVI. Furthermore, fewer patients with SAPIEN 3 Ultra THV required PPI after TAVI, which could possibly indicate its superiority compared with the SAPIEN 3 THV.
Identified patient characteristics associated with PPI
Several studies have found pre-existing RBBB to be the strongest patient-related predictor for PPI after TAVI,8 11 14 20 which was confirmed by the results of our study. In the retrospective analysis for the same group of centres, RBBB was also one of the patient-related variables significantly associated with PPI.18 Patients with a high burden of PPI following TAVI in our study had a higher prevalence of several comorbidities before the procedure, including COPD. The prevalence of COPD at baseline in patients undergoing TAVI has been reported to be between 12.5% and 43.4%, and COPD has been associated with an increased risk of respiratory complications and pneumonia following the procedure.21 22 However, there has been no association between COPD and cardiovascular postoperative complications, which may require PPI.21 Although both study groups had a high baseline systolic PAP, the rate was higher in the PPI group and is classified as mild-to-moderate pulmonary hypertension (systolic PAP 40–59 mm Hg).23 Since RBBB is commonly found in patients with pulmonary hypertension, which is a complication of COPD, this may explain the difference in COPD prevalence between PPI and non-PPI groups in our study.
Compared with the retrospective data from this centre group, pulmonary hypertension, QRS >110 ms, first-degree AV block and LAHB were not associated with PPI by multivariable analysis in the present study.18 Based on the previous reports, calcification of the LVOT, especially in the NCC zone, is one of the anatomical factors that are related to the pacemaker dependency after TAVI.16 24 The results of the retrospective analysis supported this finding.18 However, the analysis in the current study demonstrated no difference in the calcification of the LVOT between patients receiving PPI following TAVI and those who did not. A possible explanation could be the overall low LVOT calcification grade in the present cohort (only 25 mm3), which could impact statistical power. One can speculate that high LVOT calcification may be less common today due to the fact that TAVI has extended to low-risk and intermediate-risk patients, who typically present with a non-calcified or less calcified LVOT.
Identified procedure characteristics associated with PPI
Patients with PPI post TAVI in the present analysis had a larger implanted valve size, compared with patients without PPI. A large diameter valve has been previously found to be an independent risk factor for PPI in several other studies, including the retrospective study of this registry.18 25 26 It is, however, unsurprising that patients requiring a larger valve size tend to have poorer outcomes post TAVI as they present with anatomical differences at baseline, such as more calcified annulus, which potentially impact the procedural factors, including larger balloon size and higher implantation depth. Furthermore, a higher number of patients with PPI in our study had an oversized prosthesis. Prosthesis oversizing to a certain degree is recommended for SAPIEN 3 to achieve device success and to avoid paravalvular leak, whereas excessive oversizing increases the risk of PPI due to added stress on the membranous septum, aortic annulus and LVOT.27 Leber et al demonstrated that the rate of postprocedural PPI tended to be lower in patients with <15% oversizing compared with those with >25% oversizing for the Edwards SAPIEN XT.28 Husser et al also showed that prosthesis oversizing was a predictor of PPI using the SAPIEN 3, suggesting avoidance of extreme oversizing.11 On the other hand, Gonska et al concluded that oversizing had no significant effect on the PPI rate.15 The rates of clinically relevant (moderate/severe) paravalvular regurgitation after TAVI were not significantly different between our study groups, despite the higher proportion of patients with PPI receiving the 29 mm valve. Moreover, the prevalence of paravalvular regurgitation after TAVI was low in the total patient population in our study, which may be due to the novel outer polyethylene terephthalate sealing cuff in SAPIEN 3 that provides a tighter seal and reduces the risk of leakage.29
We previously reported a valve implantation depth of >30% to be associated with an increased risk of PPI in the retrospective cohort of patients.18 Although implantation depth did not reach statistical significance in the present analysis (p=0.068), the numerical difference in rates of NCC, RCC and LCC >median between patients with and without PPI was substantial (63.9 vs 47.3%), and a higher number of cases could potentially further prove this association. Schwerg et al compared the PPI rate in ‘low implantation’ with ‘high implantation’ independently from the patients’ pre-existing conduction disturbances and suggested choosing a higher implantation technique with the central marker 2 mm or more over the annular plane to minimise the risk of PPI.13 Furthermore, Mauri et al identified implantation depth as an independent predictor for PPI, proposing an implantation height of <25.5%.14 Contrary to these findings, there was no impact of implantation depth of the prosthesis on the need of PPI post TAVI in our study. Furthermore, the valve was implanted higher in our cohort compared with the previous studies with a similar definition of implantation depth.12 14 This may be explained by the fact that higher implantation depth has been established as a strong predictor of PPI, and many physicians tend to implement an implantation technique that results in a high final prosthesis position.
Similarly to the implantation depth, the difference in postdelivery balloon dilatation rates between patients with and without PPI was not statistically significant in our cohort (p=0.264), yet the numerical difference was present (4.8% vs 2.0%), indicating a distinct clinical association. In a recent meta-analysis on predictors of PPI after TAVR, postimplant balloon dilatation was among the 14 notable risk factors for PPI.30 Therefore, the clinical role of postdelivery balloon dilatation in PPI rates after TAVI may be significant in larger cohorts.
The findings in this study are subject to several important limitations. Although the overall sample size in our study was large in comparison to other studies, the number of patients in the group undergoing PPI following TAVI was relatively small. The observational design of this study allowed an evaluation of TAVI patients in a real-world setting, yet there was a higher potential for missing data, and data on some variables in our study were incomplete. All patients in our registry received the Edwards SAPIEN 3/Ultra valves, though standard treatment protocols, post-treatment pathways and individual country and centre healthcare systems may vary, which might have influenced the presented data.
Clinical implications and outlook
Identification of patients at an increased risk of PPI after TAVI is of great clinical importance to prevent patient complications and to reduce the length of hospitalisation and treatment costs. Larger valve size and pre-existing RBBB remain the major predictive factors associated with pacemaker dependency following TAVI. Patient-related baseline markers may be challenging to address but should be thoroughly assessed in the preprocedural planning. The role of pacemaker placement, the timing of placement and prognosis of patients who require PPI are still unexplored and need to be addressed in future studies. There is an ongoing 1-year follow-up of this registry to further validate the reported results and to judge the long-term dependency of the patients on the pacemaker.
The overall incidence of postprocedural PPI was low, considering the strict inclusion of high-risk patients for PPI, which demonstrates the improvement of technical and procedural aspects of TAVI. On this basis, only valve sizing persisted to be a major, avoidable risk factor for PPI. The valve type, implantation depth and postdelivery balloon dilatation all affected PPI rates, but without statistical significance, potentially already reflecting the refined implantation techniques in the participating centres. For further investigation into the role of modifiable risk factors, a study with a much larger sample size or a large database analysis is required. Nonetheless, the data confirm that careful clinical decision making before and after the intervention is key to achieving an uneventful postinterventional outcome.
Data availability statement
Data are available upon reasonable request. Data are available from the corresponding author upon request.
Patient consent for publication
This study involves human participants and was approved by the ethics committee responsible for each site. Participants gave informed consent to participate in the study before taking part. The registry was conducted in accordance with the Declaration of Helsinki and complied with local laws and regulations.
Data were captured using the s4trials Software provided by Software for Trials Europe GmbH, Berlin, Germany. The tremendous help of Dr Karin Bramlage with the statistics (IPPMed, Cloppenburg, Germany) is acknowledged.
TR and MD contributed equally.
Contributors TG is fully accountable for the overall content as guarantor. TR, JBaa, PB, JK, JBar, MD and TG were involved in the conception and design of the registry. MD, TR, JBaa, N-EN, JBar, LH and TG contributed patient data. MD, PB and VH drafted the manuscript. All the other authors revised the article for important intellectual content. All authors gave the final approval of the version to be published.
Funding This work was funded by a research grant from Edwards Lifesciences (Nyon, Switzerland) to the Institute for Pharmacology and Preventive Medicine (IPPMed, Cloppenburg, Germany).
Competing interests MD received travel honoraria by Medtronic; TR, JBaa, N-EN, JBar, PB, TR and TG received speaker fees and/or research grants and/or travel honoraria from Edwards Lifesciences. JK is an Edwards Lifesciences employee. The institution of PB and VH received funding for performing the study.
Provenance and peer review Not commissioned; externally peer reviewed.