The first 100 patients treated with the ES at our institution were included in the study. In all cases, the indication was established by the heart team, with the participation of clinical cardiologists, interventional cardiologists and cardiac surgeons. Briefly, patients had symptomatic severe aortic stenosis (valve area <1 cm2) with high surgical risk and an estimated survival >1 year. Initially, a EuroSCORE >20% was required, but subsequently patients with EuroSCORE <20% and with other situations (e.g., patent left internal mammary artery grafts and porcelain aorta were accepted).

Of the 100 patients, 78 underwent TAVI by transfemoral access and 22 by transapical access. The transfemoral approach was the first choice, but transapical TAVI was performed when the iliac anatomy did not allow a safe procedure by a transfemoral approach.

In all cases, the procedure was performed under three-dimensional transesophageal monitoring and general anesthesia.

A femoral vein was punctured to advance a temporary pacemaker into the right ventricle, and a femoral artery was used to advance a pigtail catheter into the ascending aorta for angiographic monitoring of the procedure. After the aortic valve was crossed with the guidewire, balloon valvuloplasty under rapid pacing was performed in most cases. The prosthetic valve was also implanted under rapid pacing (180–220 bpm).

The temporary pacemaker was withdrawn 24 hours after the procedure in the absence of conduction abnormalities.

The statistical analysis was performed using SPSS (SPSS Inc., Chicago, IL). Continuous variables were expressed as mean ± standard deviation, and compared with the Student's t test. Discrete variables were expressed as percentages (proportions), and compared with the chi-square test (with Fisher's correction when necessary). Associations were considered statistically significant with p<0.05. The incidence of PM implantation was estimated with Kaplan-Meier survival analysis.

A new permanent PM was indicated before hospital discharge in eight patients. The PM was implanted between 48 and 72 hours in six cases, and in 3–5 days in two. However, 12 out of the 100 patients had a permanent PM before TAVI. Thus, the rate of new PM implantation was 9.1% (eight of 88 patients without previous PM). The indication for PM was atrial fibrillation with heart rate <50 bpm in four patients, sinus rhythm with complete AV block in three, and sinus rhythm with high-degree AV block in one.

After discharge, another four patients required a permanent PM, at 42 days, three months, 18 months, and 30 months. The cumulative rate of PM implantation was 9.1±0.3%, 12.1±3.6%, 12.1±3.6%, 15.8±5.0% and 24.2±9.2% at 30 days, six months, one year, two years and three years, respectively (Figure 1).

Figure 1.

Kaplan-Meier curve of cumulative incidence of new permanent pacemaker implantation. PM: permanent pacemaker.

(0.06MB).

Important variables such as age, aortic annulus diameter, prosthetic valve diameter, syncope previous to the procedure, and left ventricular ejection fraction were not associated with the need for a new PM during the first 30 days or during follow-up (Tables 1–3). Interestingly, some parameters of renal function were associated with the need for PM implantation in both the short and long term. The presence of right bundle branch block was associated with the need for PM during the first month after TAVI.

Previous dialysis and right bundle branch block at baseline showed a statistically significant association with need for PM using the chi-square test (Table 1). However, after performing Fisher's correction, only right bundle branch block showed a tendency for a greater need for PM.

The presence of right bundle branch block at baseline was not significantly related to the need for PM more than one month after TAVI (Table 2). Nevertheless, creatinine plasma level before the procedure was associated with the need for new PM beyond the first month: 2.5±1.7 mg/dl and 1.3±0.6 mg/dl in patients with and without need for PM, respectively (p=0.001). On the other hand, a history of previous myocardial infarction was statistically associated with the need for late PM.

Considering both the first month after TAVI and long-term follow-up, there were only two characteristics associated with the need for a new PM: plasma creatinine at baseline and dialysis previous to TAVI (Table 3).

Receiver operating characteristic (ROC) curves for plasma creatinine showed an area under the curve (AUC) for need for PM during hospitalization of 0.706 (95% confidence interval [CI]: 0.500–0.912) (Figure 2A). The ROC curve showed an AUC for need for PM throughout follow-up of 0.777 (95% CI: 0.624–0.930) (Figure 2B).

Figure 2.

Receiver operating characteristic curves of baseline plasma creatinine value for predicting need for pacemaker implantation during the entire follow-up (A) and during hospitalization or in the first month after transcatheter aortic valve implantation (B).

(0.1MB).

Some patients undergoing TAVI need a permanent PM because of severe AV conduction disturbances,1–12,15 which appear to be related to device-induced trauma of the interventricular septum (IVS).14 During and after TAVI, new left bundle branch block is relatively frequent, secondary to conduction system damage during the procedure.16 As a result, patients with conduction abnormalities before the procedure are at higher risk of needing a permanent PM after TAVI.

Although the rate is 2.5–5 times lower with the ES than with the CV,2,5,9 4–11% of patients treated with the ES need a permanent PM.1–12 With both systems, this complication is more common than with surgical aortic valve replacement.17 In our series, 9.1% of patients treated with the ES needed a permanent PM during initial hospitalization, which is similar to the 8.8% incidence in the PARTNER trial and other studies.1–3,12,15

Beyond the first month after TAVI, some patients also needed PM implantation, but probably in these cases it was not directly due to the valve procedure. The need for PM beyond the first month after TAVI is probably related to the type of patients that are candidates for TAVI (advanced age, and high prevalence of renal failure), irrespective of the valve intervention.

The need for PM after TAVI is not associated with higher mortality after the procedure,3,9 but this complication prolongs hospital stay and can lead to complications such as pneumothorax, bleeding and infections. Moreover, PM rhythm may be associated with impaired ventricular function18 and a higher incidence of cardiac failure and cerebrovascular events during follow-up.19,20

Most of the conduction abnormalities associated with TAVI develop during the procedure (sometimes even before valve implantation)21 or soon after, but may also occur beyond the first 48 hours.22 These conduction disturbances are usually persistent, but some studies have shown that a significant proportion of patients undergoing permanent PM implantation after TAVI were not PM-dependent months after the procedure.8,11,7 This could justify more restrictive indications for permanent PM after TAVI, precluding unnecessary PM implantation by delaying the decision in some patients.23 However, our opinion is rather not to delay the indication for PM, because in most cases conduction abnormalities after TAVI are irreversible, and prolonged temporary transvenous pacing may also be associated with complications.24 Moreover, in some studies of patients treated with TAVI, those discharged with a permanent PM had a lower risk of sudden death after discharge.3

As the risk of permanent PM after TAVI is higher with the CV system,2,5,9 a number of characteristics have been identified associating this complication with the use of this device1,2,4,6,9,15,25,26: depth of implantation, use of first-generation devices, IVS thickness, left ventricular end-diastolic diameter, conduction abnormalities before the procedure, valve oversizing, age over 75, heart rate below 65 bpm before the procedure, atrial fibrillation, porcelain aorta, balloon pre-dilatation, and non-coronary cusp thickness.

In our series of patients treated with the ES, we found that both right bundle branch block and terminal chronic renal failure (on dialysis) were associated with the need for early permanent PM (during hospitalization or during the first month). Other variables, such as access route (transfemoral or transapical), age, valve diameter, aortic annulus diameter, or body mass index, were not associated with this complication.

Right bundle branch block has been associated with higher frequency of PM implantation in several series of TAVI with the CV.4,10,27,28 The AV node and left bundle branch are located near the aortic root and native aortic valve,13 while the non-coronary cusp is adjacent to the membranous portion of the IVS, which extends into a triangle bounded by the non-coronary and the right coronary cusps, where the His bundle is located. The left branch emerges in the membranous septum, crosses the fibrous trigone and passes along the left surface of the muscular septum. This is why previous right bundle branch block increases the risk of complete AV block if there is damage to the left ventricular outflow tract during TAVI.

The association between renal function and the need for PM after TAVI is less well understood. Renal failure is associated with electrolyte abnormalities that can induce rhythm and conduction disturbances. Most importantly, chronic renal failure is associated with deposition of extracellular matrix and collagen and development of fibrosis in the myocardium that may result in conduction abnormalities, arrhythmias and systolic and diastolic dysfunction.29 More severe renal dysfunction is associated with a higher prevalence of left bundle branch block and advanced AV block.30 Renal failure is also a predictor of progression to complete AV block in patients with bifascicular block,31 which is consistent with the greater need for permanent PM in longer-term follow-up that we observed late after TAVI.

Finally, in our study previous infarction was associated with the need for permanent PM more than one month after TAVI. Myocardial dysfunction occurring late after myocardial infarction may induce conduction abnormalities. Previous myocardial infarction was not, however, associated with the need for PM during the first month after TAVI.

The authors declare that no experiments were performed on humans or animals for this study.

The authors declare that no patient data appear in this article.

The authors declare that no patient data appear in this article.