Predictive factors for pacemaker requirement after transcatheter aortic valve implantation

In our retrospective analysis, between January 2007 and January 2008 45 patients without previously known pacemaker who underwent TAVI using the third-generation percutaneous self-expanding CoreValve prosthesis (Medtronic Inc; Minneapolis, USA) were identified for this analysis. The criteria for inclusion and exclusion to the TAVI procedure have been described elsewhere [1‐4]. In brief, patients were included with echocardiographic measurements of severe native valvular stenosis and an area < 1 cm², or < 0.6 cm²/m² regardless of adjunct valvular regurgitation; a diameter of basal orifice of the stenosed valve between 20 and 27 mm; and a diameter at the sinutubular junction ≤ 43 mm. Most importantly, all patients were considered unfit for open surgery with an EuroSCORE ≥ 20% [7]. TAVI was suggested in agreement between a cardiac surgeon and both, a clinical and interventional cardiologist; patient`s or referring physician`s preference was not relevant [8, 9]. Treatment strategy was in compliance with the Helsinki Declaration. Our local ethics committee approved this treatment strategy. Patients gave informed consent prior TAVI. Pacemaker implantation at follow-up was considered indicated in case of complete AV-block, type II second-degree AV block, and in presence of new LBBB in combination with HV prolongation ≥ 75 msec. We compared patients requiring a new pacemaker within 7 days with those not requiring any pacemaker and no significant conduction delay after TAVI.

Details of the implantation technique have been described elsewhere [1‐4]. In brief, all procedures were performed in an interventional hybrid suite under general anaesthesia to assure stable hemodynamic and minimize patient movement. TAVI was performed via femoral access under fluoroscopy; the aortic valve was initially dilated using a standard valvuloplasty balloon with a nominal diameter similar to the aortic valve and followed by CoreValve (Medtronic Inc.) insertion [1, 3‐6]. In each patient, prior to TAVI and aside a right ventricular bipolar pacing lead (Pacel™, St. Jude Medical, St Paul, MN, USA), a 6F quadripole electrode catheter with ring electrodes (5-mm interpole distance) (Webster D™, Biosense Webster, Diamond, USA) was introduced and advanced to the His bundle. A 6F quadripolar electrode catheter (Soloist™, Medtronic, Minneapolis, MN, USA) was advanced to the right atrium and another to the right ventricle to record a bipolar electrogram and for programmed atrial and ventricular stimulation. The access sites for all electrophysiologic electrode catheters were the femoral veins. With such instrumentation the sinus node recovery time (SNRT), corrected sinus node recovery time (c-SNRT), antegrade and retrograde effective refractory period (ERP) of the AV node, as well as intracardiac conduction times (atrium to His and His to ventricle time; AH and HV interval) were assessed. The rational to measure SNRT was the observations of sinus node arrest in a single patient receiving a CoreValve prior this study. Thus, we wanted to avoid oversee sinus node pathology in this elderly patients suffering from a high comorbidity index. All measurements were done on an Axiom Sensis™ (Siemens, Erlangen, Germany) electrophysiology workstation prior to the valve implantation, immediately after valve implantation and at a follow-up of 7 days. The AV nodal ERP was measured by introducing a single extrastimulus (S2) after a drive train of 8 stimuli at a fixed rate (S1) (600ms), at which time the S1-S2 interval is decreased until the S2 impulse does not conduct to the His bundle. To assess conduction disorders, patients were attached to uninterrupted ECG monitoring using the Philips monitoring system (IntelliVue™, Best, The Netherlands) that is installed at our ICU/IMC unit. All patients were prophylactically given a temporary pacemaker via the existing femoral venous access; with VVI mode the active pacing was 60 bpm for at least 24 hours.

All data were stored and analysed using SPSS statistical package 17.0 (Chicago, Illinois, USA). The statistics computed included mean and standard deviations of continuous variables and are presented as means ± SD, frequencies and relative frequencies of categorical factors. The Fisher exact test and chi²-test were used to compare proportions. A normal distribution of differences was confirmed by the Kolmogorov-Smirnov test; in presence of normally distributed data the t-test was used while with non-normal distribution the Mann–Whitney U-test was applied. The logistic regression model was used to assess the interdependence of high-grade AV block (type II second-degree and third-degree) at 7 day follow-up from prognostic factors. First, univariate analyses were performed to reveal unadjusted significant associations between prognostic variables and high-grade AV block. Thereafter, variables yielding p-values ≤ 0.25 with univariate analysis were entered in the multivariate model to assess adjusted associations between outcomes and covariates which were univariate and of borderline significance. All p-values resulted from two-sided statistical tests and ≤ 0.05 was considered to be significant.

The present study suffers from the limitation of being a retrospective analysis. Additionally, the low number of patients may influence the results. However, our study showed that there might be no further information from intracardiac measurements as the ECG changes predominantly present as LBBB, first grade and complete AV block. The differentiation of AV block in supra-His and infra-His is not essential in these frailty patients with high comorbidity index. Thus, there will always be a somewhat “liberal” indication for pacemaker implantation. To date, there is a need for further larger scale prospective trials as well as registries with longer follow-up for predicting factors for pacemaker requirement and defining the role and duration of ECG changes.