Incidence and risk factors for pacemaker implantation following aortic valve replacement

Abstract

Our aim was to identify the predictive factors for permanent pacemaker (PM) implantation in patients undergoing isolated aortic valve replacement (AVR). A total of 3534 patients received an AVR between January 1990 and December 2003 in our institution. Permanent PM implantation was performed in 234 (6.6%) patients, over median time of three days (range one to 24 days). This patient population was compared to a random sample of 191 patients undergoing AVR without permanent PM implantation. The overall mean age was 63.5 years (±14.2) and 261 patients (62%) were male. Univariate and multivariate logistic regression analysis of pre- and perioperative data were performed. Overall the 30 days mortality was 4.2% (10/234) in patients with PM and 1% (2/191) in the control group (P=0.046). Patients with PMs were older (P<0.001), had more additional coronary artery bypass grafting (CABG) surgery or mitral valve replacement (MVR) (P<0.001), complete right bundle branch block (RBBB) prior to surgery, and more frequently underwent re-operations compared to patients without PMs (P<0.001). The multivariate logistic regression model with PM implantation as the dependent variable demonstrated that older age was not independently associated with PM implantation. As independent predictors concomitant severe mitral valve insufficiency, CABG, subaortic stenosis (SAS) or re-do operations were identified.

1. Introduction

Aortic valvular stenosis and regurgitation are associated with abnormalities of conduction, including higher degrees of atrioventricular (AV) block. Aortic valve replacement (AVR) can result in the development of further conduction abnormalities, which may be associated with an increased risk of sudden death [1].

The association of intraventricular conduction disorders and aortic valve surgery is frequently seen as a transient episode. However, definitive conduction lesions, which result in a permanent pacemaker (PM) implantation, are reported to occur in between 2% and 7% of patients undergoing AVR [2, 3].

Since the first description was made of intraventricular conduction defects as a complication of AVR in 1972, there has been a remarkably wide variation in the literature in regard to timing of PM implantation following AVR. The recommended waiting period ranges from three to 21 days after the onset of a high-degree AV block prior to a permanent PM implantation [3, 4]. Mechanical trauma to the conduction system in the vicinity of the aortic valve is thought to be the most common contributing factor for a permanent PM implantation following AVR. Additionally surgical procedures, such as myectomy for hypertrophic obstructive cardiomyopathy, or the repair of a concomitant ventricular septal defect (VSD) [4] were postulated as predictive factors for PM implantation. However, risk factors, such as patient age, preoperative conduction disturbances, preoperative medication, strong calcified aortic root or extracorporeal circulation (ECC) time and balance, are less well studied as potential contributing risk factors for a PM implantation after AVR.

Identifying risk factors responsible for developing conduction system abnormalities that are likely to require postoperative permanent PM would be of substantial clinical benefit to facilitate the planning of postoperative care. In the presented study, we sought to identify clinical criteria that might predict the need for early postoperative permanent PM implantation following AVR.

2. Materials and methods

2.1. Patient population

Between January 1990 and December 2003, a total of 3534 patients underwent AVR in the Department of Cardiovascular Surgery at the University Hospital of Zurich.

Permanent PM implantation, due to a high degree of AV block (AV block II–III), was performed in 234 patients (6.5%) during their hospital stay. As a control group a random sample of 191 patients undergoing AVR without PM implantation was chosen. Consequently, the study population consisted of 425 patients (262 were male with a mean age of 65 years). Exclusion criteria were the presence of a preoperative permanent PM and evidence of AV block in preoperative ECG.

The major indications for AVR were predominant valvular stenosis in both investigations' subpopulations, 167 (71.5%, 167/234) in PM and 145 (75.9%, 145/191) in control group. Comprehensive perioperative data were collected from the database and from surgical or anesthetic notes. Mechanical prosthesis was implanted in 115 (49%) in the PM group, in control group 118 (61.7%) patients received a mechanical valve. Baseline demographic details are shown in Tables 1 and 2 .

2.2. Surgical technique

Implantation technique was identical in all patients during the study period. AVR was performed on ECC with a moderate hypothermia (26–28 °C) and a mean pressure of 50–70 mmHg. Myocardial protection was achieved by retrograde cold blood cardioplegia. Repetitive dosages were given every 10 min. The left ventricle (LV) was vented through the right superior pulmonary vein. The valve prosthesis was implanted with single pledged stitches.

Postoperatively, a 12 lead ECG was recorded on a daily basis until the fourth postoperative day and then every second day until discharge of the patient.

Operative mortality was defined as death occurring within 30 days after the operation. The studied parameters were divided in pre-, intra-, and postoperative data blocks and are shown by Tables 2–4 .

2.3. Statistical analysis

Standard demographic and clinical variables were recorded on standardized data forms. Statistical analysis was performed using JMP IN^® (SAS Campus Drive, Cary, NC 27513, USA; version 5.1). Univariate relations of all perioperative factors requiring PM implantation after AVR were analysed by using χ²-test, Fisher's exact test and the Mann–Whitney test. To identify significant differences in numerical variables the Mann–Whitney test was used. The χ²-test and Fisher's exact test were performed to analyse categorical variables. Differences were considered significant at a P-value of less than 0.05. Combinations of risk factors were evaluated with multiple logistic regression models. Factors showing a value of P<0.05 in univariate analyses were entered into multivariate logistic regression analysis.

3. Results

A permanent PM was implanted in 234 (6.5%) out of 3534 patients, the mean period of implantation was 4.4±3.8  days (range one to 24 days). Indications for PM implantations were as follows, 227 of 234 (97%) developed a higher degree of AV block, 99 patients (42%) developed a complete or high degree AV block intraoperatively. During hospitalisation 128 patients (55%) also developed higher AV of higher degree. The onset for the postoperative AV block in this patient population appeared in an average time of 1.96 days (range: one to nine days). Persistent symptomatic sinus bradycardia and junctional rhythms was found in seven patients (3.8%) this was also an indication for permanent PM implantation (two of these patients underwent a superior trans-septal approach to the mitral valve). In this investigation group (PM implantation) 140 (59.8%) patients received an additional surgical procedure, such as coronary artery bypass grafting (CABG) (99/234), mitral valve replacement (MVR) (13/234) and repair (11/234) (Table 2).

In the control group, the onset of transient AV block was found to occur between postoperative days two to five (average of 3.8 days) in six patients.

Patients with PM implantation were found to be older (P<0.001) and to have had a myocardial infarction (MI) (P<0.01) and a complete right bundle branch block (RBBB) (P<0.006) more frequently prior to aortic valve surgery. Furthermore, patients in the PM group underwent more additional cardiac procedures, such as CABG surgery (P<0.0001), MVR (P<0.001), subaortic stenosis (SAS) resection (P<0.001), or re-operations (P<0.001) when compared to patients without PM implantation (Tables 2–4).

The overall in-hospital mortality rate was 4.2% (10/234) for patient group undergoing PM implantation, 1% (2/191) in the control group (P=0.046) (Tables 5 and 6 ). Mortality for isolated AVR with PM implantation was 1.4% (1/73).

The multivariate logistic regression model with PM implantation as the dependent variable demonstrated that older age was not independently associated with PM implantation (Table 6).

Furthermore, a small aortic annular size, a prolonged ECC duration, and a higher ECC balance were found to present minor additional risk factors for PM implantation following AVR (Tables 1 and 3).

4. Discussion

Rhythm disturbances and transient conduction disorders are frequently seen following cardiac procedures, and most of them present a sinus node dysfunction, various fascicular blocks, or higher AV blocks. Fortunately, the majority are self-limited and vanish before hospital discharge of the patient. A few, however, are permanent and result in the implantation of a permanent PM [5–8].

Following aortic valve surgery, the onset of bradycardia is usually caused by a persistent, complete AV block, which requires permanent pacing in between 2% and 7% of patients [9–11]. Similar to the data in the recent literature in our study population series the incidence of the permanent PM implantation was 6.5%. This includes the patients undergoing AVR associated with surgical procedures, such as CABG and additional valve replacement and/or reconstructions. Additionally in the patient subpopulation where only isolated AVR was performed, 2% (73/3534) of them received a permanent PM implantation. These data are compared to the recent literature, where the incidence of the PM implantation following isolated ARV ranges between 3.2% and 4.6% [3, 8, 12].

A more detailed analysis of our results revealed that of the additional surgical procedures, such as MVR, SAS resection, VSD closure might be postulated as predicted factors for the occurrence of postoperative AV conduction disturbances. These findings are not surprising given the well-recognized association between surgical manipulations at the fibrous skeleton of the aortic valve and the immediate anatomical vicinity of the AV node and the proximal conduction bundle. The necessity for de-calcification of a heavily calcified aortic annulus is chiefly assumed to be a major contributing factor for the occurrence of a partial or complete AV block postoperatively [2, 3, 8]. Surprisingly, we were unable to find a significant correlation between the severity of aortic annular calcifications and the occurrence of a permanent PM implantation, as well as the occurrence of a postoperative left bundle branch block (LBBB). However, the assessment of the degree of valvular calcifications was subjective and based on the visual impressions of the surgeon. Although the finding of preoperative RBBB as a significant risk factor for post-AVR PM implantation is surprising, its cause is not clear.

Further surgery-related factors, which significantly increased the risk for PM implantation, are the performance of concomitant CABG, mitral valve surgery and re-operations. These findings are also reflected by the fact, that a longer ECC time and balance were revealed as risk factors for a PM implantation following AVR. AVR was performed with moderate haemodilution and myocardial protection, which was achieved as retrograde cold blood cardioplegia. In the literature some evidence exists to suggest that blood cardioplegia may increase the risk of permanent PM implantation [13]. Additionally in the literature it was suggested that that a significant factor in the pathogenesis of conduction blocks is cold-related injury [14].

Further the retrograde cardioplegia protects only the left coronary artery supply systems, in case of right origin of the atrio-ventricular bundle artery, this is not protected adequately during the surgical procedure, and as thus may be also an origin of postoperative conduction disturbance. The occurrence of ischaemic injury to the conduction system and the development of myocardial oedema following prolonged periods of ECC might explain these findings [15].

The analysis of the preoperative risk factors showed that older age, MI, and a complete RBBB prior to AVR present risk factors for a postoperative PM implantation and were confirmed by multivariate analysis. The findings of the preoperative significance of severe mitral valve insufficiency as a predictive factor is reflected in the additional procedure on the mitral valve that was performed parallel to the AVR. However, age did not continue to present a risk factor for PM implantation following AVR when undergoing multivariate analysis and this finding stands in contrast to several reports [4, 10, 13].

An important aspect of postoperative PM implantation after AVR is in the timing of the implantation procedure in order to reduce periods of intensive care unit (ICU) and hospital stays and the subsequent costs. Remarkable is the wide variation of this timing reported in the literature, which ranges from three to 21 days and does not allow for a clear answer to the question of how long one should wait for the recovery of the conduction system prior to the implantation of a permanent PM [4, 5].

In this study, complete AV block occurred in 38% (88) of patients intraoperatively and in the majority of patients (55%) within one to two days after AVR. The onset for the AV block appeared in an average time of 1.96 days (range: one to nine days). In our control group, there was an onset of transient AV block following AVR that was found to occur between postoperative days two to five (average of 3.8 days). These findings are in accordance with other reports in the literature. Keefe et al. reported transient complete heart block in 18% of patients receiving isolated AVR [15].

Given these findings, it appears that the practise of an early PM implantation is justified after the development of permanent higher AV block intraoperatively or in the first 24 hours following AVR. A waiting time of two to three days postoperatively appears to be appropriate, and by following such an early implantation practise, ICU and hospital stays and subsequent costs could be reduced.

We gratefully acknowledge Ursi Wacker for her expert assistance in data acquisition.

References