Influence of permanent pacemaker implantation after transcatheter aortic valve implantation with new-generation devices

Data on the impact of permanent pacemaker implantation after TAVI with the new-generation devices are rare. Available follow-ups on this matter are from patients treated with the early-generation devices. Since survival after TAVI seems to continue to improve with the new-generation devices, but pacemaker implantation rate does not decline, the impact of pacemaker implantation in these patients needs to be further evaluated. We add first one-year follow-up data on outcome after TAVI in patients treated with the new-generation devices, which did not show a disadvantage in outcome for patients with permanent pacemaker implantation.

Within the last decade, transcatheter aortic valve implantation (TAVI) has emerged as the standard care for severe aortic stenosis in inoperable patients or in patients at high risk for surgical valve replacement (SAVR) [1, 2]. In patients at intermediate risk for SAVR, noninferiority of TAVI in comparison with SAVR was shown for the balloon-expandable Edwards Sapien valves (Edwards Lifesciences, Irvine, CA, USA) in the PARTNER 2A randomised trial [3] as well as for the self-expandable Medtronic CoreValve (Medtronic, Minneapolis, MN, USA) in the SURTAVI trial [4]. However, rate of permanent pacemaker implantation (PPMI) after TAVI is significantly higher compared with SAVR [5]. Even though the new-generation transcatheter heart valves have managed to reduce periprocedural complications, such as moderate or severe paravalvular aortic regurgitation or vascular access complications, when compared with the early-generation devices [6], PPMI still remains a common postprocedural complication, with tendencies to higher rates of PPMI compared with the early-generation transcatheter aortic heart valves [7, 8].

Rates of PPMI range from ~17% for the balloon-expandable Edwards Sapien 3 valve (ES3) [5], ~22% for the self-expandable Medtronic CoreValve Evolut R [9] up to 40% for the mechanically expanded Boston Lotus valve (Boston Scientific Corporation, Marlborough, MA) [10]. The impact of PPMI on further patient survival and quality of life is still a matter of debate. There is evidence of an association of chronic right ventricular pacing with the occurrence of heart failure in non-TAVI patients based on the MOST trial and DAVID trial, which is feared to be translated to the outcome of elderly patients with PPMI after TAVI [11, 12]. Data from the FRANCE-2 registry evaluating TAVI revealed PPMI after TAVI as an independent predictor for 3‑year mortality (p = 0.02) [13] , whereas a recently published meta-analysis of 23 studies including 20,287 patients undergoing TAVI did not observe an increased mortality during follow-up [14]. All of these data were generated from studies with the early-generation TAVI devices. Currently, there are no data on the impact of PPMI after TAVI with the new-generation devices.

Therefore, we wanted to analyse the influence of PPMI after TAVI with new-generation devices on mortality as well as on clinical outcome in a large single-centre all-comers patient collective within a 12-month period.

We evaluated 688 consecutive patients undergoing transfemoral TAVI for treatment of symptomatic aortic stenosis (Clinical Trial Registration: NCT02162069). TAVI was performed at the University of Ulm (Germany) between February 2014 and September 2016. Baseline characteristics of patients with (PPMI) and without need for PPMI (no-PPMI) after TAVI were compared as well as clinical outcome within a 12-month period. Patients with pre-existing permanent pacemakers were excluded. Severe aortic stenosis was documented by echocardiography and cardiac catheterisation with an aortic valve area (AVA) ≤1 cm² or an indexed AVA ≤0.6 cm². Patients were at high or intermediate risk for SAVR indicated by the Society of Thoracic Surgeons (STS) score for mortality, were at high risk according to the risk assessment combining STS score for mortality, frailty or major organ system dysfunction according to the 2014 American College of Cardiology/American Heart Association Guidelines for Valvular Heart Disease [15] or had relevant further comorbidities favouring a transcatheter approach. There was a Heart Team decision for a transcatheter valve replacement in all patients. Written consent was obtained in all patients. The study was approved by the local ethics committee. All TAVI procedures were performed via a transfemoral approach under conscious sedation as described elsewhere [16, 17]. Predilatation was performed in all patients. The new-generation TAVI devices used were the balloon-expandable ES3, the mechanically expanded Lotus valve and the self-expandable CoreValve Evolut R. The size of the implanted valve was determined by a pre-procedural 256 multislice computed tomography as described elsewhere [18].

Primary endpoint was all-cause mortality at one year followed by a combined endpoint of major adverse events including all-cause-death, cardiac events (myocardial infarction, new onset of atrial fibrillation, heart failure), valve function related symptoms, stroke/transitory ischaemic attack, bleeding complications or vascular complications occurring after hospital discharge according to the Valve Academic Research Consortium 2 (VARC-2) criteria [19]. Furthermore, we evaluated the rate of rehospitalisation due to cardiac symptoms.

After excluding patients with a permanent pacemaker prior to the TAVI procedure, we carried out the analysis on 612 patients. PPMI after TAVI was needed in 168 patients (27.5% of patients without prior PPMI and 24.4% of the total patient cohort). Pacemaker specific details are displayed in Tab. 1.

Mean population age was 80.4 ± 5.9 years. More than three quarters of patients were severely symptomatic with NYHA class III or IV (77.9%). The STS score for predicted risk of mortality of the cohort was 6.5 ± 4.9%. Mean aortic valve area acquired by echocardiography was 0.77 ± 0.22 cm², mean aortic peak gradient 66.2 ± 23.5 mm Hg, mean LVEF 57.5 ± 15.2%.

Baseline clinical characteristics of patients, according to the pacemaker status after TAVI, did not show any statistically significant differences (Tab. 2). However, there was a significant difference in pre-existing conduction disorders between patients with and patients without PPMI after TAVI. The rate of a pre-existing first-degree atrioventricular block was significantly higher in patients with PPMI after TAVI (30.0% vs. 11.8%, p < 0.001) The same applies to the rate of a pre-existing complete right bundle branch block (22.2% vs. 2.9%, p < 0.001). Rate of complete left bundle branch block at baseline (10.2% vs. 10.5%, p = 0.94) did not differ between groups, neither did history of atrial fibrillation (39.3% vs. 34.3%, p = 0.30).

Implanted valves were the ES3 valve in 360 patients (58.8%), the Boston Lotus valve in 218 patients (35.6%), the Boston Lotus Edge valve in 7 patients (1.1%) and the Metronic CoreValve Evolut R in 27 patients (4.4%). In patients with PPMI after TAVI, the Boston Lotus valve had the highest proportion within the valve distribution, whereas the ES3 valve had the highest proportion in the patients without PPMI after TAVI.

Major or life-threatening bleeding complications after TAVI according to the VARC-2 criteria were significantly higher in patients after PPMI (8.9% vs. 3.8%, p = 0.01), driven by a significantly higher rate of major bleeding events (5.4% vs. 2.5%, p = 0.01). However, the higher proportion of patients with major bleedings events did not occur due to the PPMI (only 2 patients of all patients with major bleeding events). Rate of life-threatening bleeding events was low with 3.0% (PPMI) and 1.8% (no-PPMI) (p = 0.37), with 3 patients having a pericardial effusion after PPMI. Further complications after PPMI were re-operation due to lead dislocation in 3 patients and haematoma/bleeding event at the site of the pacemaker in 4 patients. There was no pneumothorax. Ventricular pacing burden at pacemaker check 1 to 4 days after the implantation was 68.0 ± 43.3% (median 98%, interquartile range [IQR] 11–100%), with 59.9% of patients with a ventricular pacing burden ≥90% and 8.1% of patients with a ventricular pacing burden of 0%.

This study evaluates the impact of PPMI after TAVI on clinical outcome and mortality with new-generation TAVI devices. Our study with a PPMI rate of 24.4% did not reveal a worsened 30-day or one-year outcome for these patients with low total all-cause mortality rates of 1.3% and 12.4% respectively, even for patients with reduced LVEF. The low all-cause mortality rates we experienced with the new-generation devices are in line with recently published data on the ES3 with 1.9% and 11.8% respectively for a transfemoral access [20] and the Evolut R with survival rates of 99.0% and 85.8% for 30 days and one year [9].

However, there have been conflicting findings regarding the impact of PPMI on the outcome after TAVI with the early-generation devices. In the PARTNER trial the presence of a new permanent pacemaker was associated with higher one-year mortality (hazard ratio [HR] 1.38, 1.00 to 1.89, p = 0.05, 26.3% vs. 20.0%), as well as worsened clinical and echocardiographic outcomes for these patients in comparison with patients without pacemaker and left bundle branch block. They had significantly higher rates of rehospitalisation (42% vs 31.9%, p = 0.03), and significantly lower LVEF at one year after TAVI (55.4 ± 9.9% vs. 57.6 ± 8.2%, p = 0.01) [21]. However, in that same patient collective the impact of PPMI on mortality did not persist when compared with the whole population of patients without PPMI after TAVI, including those with left bundle branch block (one-year mortality 26.3% vs. 20.8%, p = 0.08) [22]. Recent data from the FRANCE 2 registry with 4,201 patients covering a follow-up of 3 years identified PPMI after TAVI with the self-expandable CoreValve and the balloon-expandable Edwards Sapien XT as an independent predictor for all-cause mortality. This stands in contrast to previous data from the FRANCE 2 registry (covering a shorter follow-up period of one year and only patients treated with the CoreValve device) concluding that PPMI did not have an impact on survival [12, 23]. Yet, there are several further studies that did not indicate PPMI to be independently associated with mortality after TAVI, including a meta-analysis of >20,000 patients. However, none of these studies covered the period of the latest data from the FRANCE 2 registry [24, 25]. Studies that demonstrated a negative effect on survival and occurrence of heart failure with PPMI included long periods of chronic right ventricular pacing [26, 27]. Therefore, the time periods for the impact of PPMI on the outcome of patients with TAVI that have been analysed up to now are most likely not long enough yet to reach clarity on this matter. Furthermore, there is evidence suggesting that pacing burden decreases in patients with PPMI after TAVI [28], a fact that we also experienced in our cohort, and there are data suggesting that worsening LVEF and higher rates of heart failure are related to higher ventricular pacing burdens [29]. Therefore, further studies evaluating these aspects on outcome will be necessary.

The rate of the combined endpoint of adverse events after hospital discharge in our study did not prove to be significantly higher for patients with PPMI. With the new-generation devices, we experienced low rates of cardiac and cerebral events. This is consistent with publications concerning the new-generation TAVI devices, especially the ES3 [20]. There is limited data on outcome >30 days with the Evolut R or Lotus valve. There seems to be no relation between these events and the presence of a pacemaker after TAVI.

In addition, the need for PPMI did not have an impact on rehospitalisation due to cardiac symptoms within 1 year after TAVI.

Like all registry data, this study in its design is prone to effects or possible confounders and bias. Furthermore, this is a single-centre study and therefore limited in its sample size.

In this large all-comers TAVI population with new-generation devices the need for postprocedural PPMI did not show a statistically significant impact on survival or on combined endpoint of major adverse events within 12 months of follow-up.