Objectives To assess the incidence of conduction disturbances leading to permanent pacemaker implantation (PPI) following isolated aortic valve replacement (AVR) in a large cohort of elderly patients with severe symptomatic aortic stenosis, and to determine the predictive factors and prognostic value of PPI following AVR in such patients.
Methods A total of 780 consecutive elderly patients (age 77±4 years, logistic EuroSCORE 10.4±8.5%, STS score 3.5±1.5%) with severe aortic stenosis and no previous pacemaker were analysed.
Main outcome measures The incidence, clinical indications, timing and predictive factors of PPI within 30 days after AVR and their prognostic value were evaluated.
Results Baseline ECG showed the presence of conduction abnormalities in 37.1% of the patients. Twenty-five patients (3.2%) needed PPI during the index hospitalisation due to the occurrence of complete atrioventricular block (2.6%) or severe bradycardia (0.6%). The presence of preprocedural left bundle branch block (OR 4.65, 95% CI 1.62 to 13.36, p=0.004) or right bundle branch block (OR 4.21, 95% CI 1.47 to 12.03, p=0.007) predicted the need for PPI after AVR. The need for PPI was associated with a longer hospital stay (p<0.0001). Thirty-day mortality rates were similar between patients with and without PPI (4% vs 3.2%, p=0.56). Survival rate at 5-year follow-up was 75%, with no differences between patients with and without PPI (p=0.12).
Conclusions The need for PPI following isolated AVR in elderly patients with severe symptomatic aortic stenosis was low. Pre-existing bundle branch block predicted the need for PPI. PPI determined a longer hospital stay, but had no effect on acute and long-term mortality.
- Aortic stenosis
- coronary intervention (PCI)
- interventional cardiology
- pacemaker implantation
- percutaneous valve therapy
- valve replacement
Statistics from Altmetric.com
- Aortic stenosis
- coronary intervention (PCI)
- interventional cardiology
- pacemaker implantation
- percutaneous valve therapy
- valve replacement
New-onset complete atrioventricular block (CAVB) needing permanent pacemaker implantation (PPI) is a widely recognised complication following cardiac valve surgery, with an incidence ranging from 3% to 8.5% associated with standard aortic valve replacement (AVR).1–10 Most AVR studies evaluating this complication, however, have included patients with a wide range of ages (including young adults), those undergoing coronary artery bypass grafting (CABG) during the same intervention, and those with both aortic stenosis and insufficiency as the predominant underlying aortic valve disease. In addition, the predictive factors and the potential prognostic value of PPI following cardiac surgery have been mostly evaluated either in large cohorts of patients undergoing very different types of cardiac surgery or AVR including both concomitant CABG and aortic insufficiency as the predominant valvular disease. Very few data exist, therefore, on the incidence, predictive factors and prognostic value of PPI in elderly patients undergoing isolated AVR as a result of pure or predominant severe aortic stenosis. Transcatheter aortic valve implantation (TAVI) has emerged as an alternative treatment for those elderly patients with pure or predominant severe aortic stenosis considered either non-candidates or at very high risk for AVR,11 and the potential expansion of this technique to the treatment of a lower-risk population is being evaluated. However, TAVI has been associated with a high rate (up to 30–40%) of PPI,12 13 and several studies have already identified some predictive factors of PPI following TAVI.13–15 A more precise knowledge of the PPI rate in contemporary elderly patients undergoing isolated AVR because of severe symptomatic aortic stenosis and the identification of those patients at higher risk of developing severe conduction disorders leading to PPI would thus be of major clinical relevance in the current AVR versus TAVI era. Therefore, the aims of this study were to assess the incidence of conduction disturbances leading to PPI following AVR in a large cohort of elderly patients with severe symptomatic aortic stenosis, and to determine the predictive factors and prognostic value of PPI after AVR in such patients.
A total of 900 consecutive elderly (≥70-year-old) patients underwent isolated AVR for severe symptomatic aortic stenosis (as a predominant lesion) at the Quebec Heart and Lung Institute (Quebec City, Quebec, Canada) between 2002 and 2010. Of these, 120 patients were excluded for the following reasons: previous AVR (n=38), previous pacemaker, defibrillator or resynchronisations devices (n=65) and concomitant ascending aorta replacement procedures (n=17), leading to a study population of 780 patients. All patients were entered prospectively into a dedicated database, and all clinical, echocardiographic, procedural and postprocedural data were prospectively gathered.
AVRs were performed through midline sternotomy, using standard surgical techniques. Temporary epicardial pacemaker leads were inserted at the completion of the surgery in all patients, and patients were placed on continuous ECG monitoring until hospital discharge. The occurrence and timing of postoperative conduction disturbances managed by temporary pacing as well as those conduction disturbances leading to PPI were prospectively recorded. The indications for PPI were in accordance with the American College of Cardiology/American Heart Association guidelines,16 ie, PPI was indicated for third-degree and advanced second-degree AVB at any anatomical level associated with postoperative AVB that was not expected to resolve after cardiac surgery and for sinus node dysfunction with documented symptomatic bradycardia, including frequent sinus pauses that produced symptoms.
For the purpose of the present study, the baseline ECG obtained before AVR were analysed by two investigators unaware of the clinical data, and the presence of conduction abnormalities as defined by the criteria of the American Heart Association Electrocardiography and Arrhythmias Committee, Council on Clinical Cardiology, the American College of Cardiology Foundation and the Heart Rhythm Society17 were recorded.
For the follow-up data, the medical insurance numbers of the patients were linked to medico-administrative data concerning hospitalisations and revascularisation procedures as well as to Quebec's death registry. The study protocol was performed in accordance with the institutional ethics committee and patients provided a signed informed consent for the procedures and the anonymous use of their data for research purposes.
Qualitative variables were expressed as percentages and quantitative variables as mean±SD or median (25–75th IQR) depending on variable distribution. The normality distribution for continuous data was examined with the Shapiro–Wilk test. Comparison of numerical variables was performed using the two-sided Student's t test or Wilcoxon rank sum test, and the χ2 or Fisher's exact tests were used to compare qualitative variables. A stepwise logistic regression analysis including all variables with p value less than 0.10 in the univariate analysis was used to determine the predictive factors of PPI and hospital stay. The following variables were included in the model for the prediction of PPI: gender, preprocedural left bundle branch block (LBBB) and right bundle branch block (RBBB). The following variables were included in the model for the prediction of hospitalisation length: age, chronic atrial fibrillation, history of cerebrovascular disease, chronic obstructive pulmonary disease, estimated glomerular filtration rate, STS score, logistic EuroSCORE (European System for Cardiac Operative Risk Evaluation), left ventricular ejection fraction, aortic valve area, conduction disturbances, cardiopulmonary bypass and aortic cross-clamp time, PPI, postprocedural stroke, the need for haemodialysis, sepsis, re-operation for bleeding and intubation greater than 48 h. Survival curves were constructed according to the Kaplan–Meier method and compared by the log-rank test. Differences were considered statistically significant at p values of less than 0.05. The data were analysed using SAS statistical software V.9.1.3.
Baseline clinical, ECG, echocardiographic findings and the main procedural and 30-day outcomes of the study population are shown in table 1. The echocardiography performed at hospital discharge showed a significant reduction in mean aortic gradient and an increase in aortic valve area (from 47±15 mm Hg to 15±6 mm Hg, and from 0.65±0.17 cm2 to 1.39±0.38 cm2, p<0.0001).
Postoperative conduction disturbances and pacemaker implantation following AVR
A total of 25 patients (3.2%) had PPI within 30 days following AVR. The indications for PPI were CAVB (20 patients, 2.6%) and severe symptomatic bradycardia (five patients, 0.6%). Conduction disturbances leading to PPI occurred at a median of 12 h (25–75th IQR 0–96 h) following AVR. CAVB occurred at a median of 6 h (25–75th IQR 0–36 h) following AVR, and severe symptomatic bradycardia occurred at a median of 5 days (25–75th IQR 4–5 days) following AVR. Most (60%, 75% for CAVB) conduction disturbances occurred either during AVR or within the 24 h after the intervention, and 40% (25% for CAVB and 100% for severe symptomatic bradycardia) occurred between days 2 and 14 after AVR. All these conduction disturbances were initially managed with temporary pacing during a median time of 4 days (25–75th IQR 3–8 days) until PPI. The median time of temporary pacing until PPI was 4 days (25–75th IQR 3–8 days) for those patients who had complications with CAVB, and 7 days (25–75th IQR 4–9 days) for those presenting with severe symptomatic bradycardia. PPI was performed at a median time of 8 days (25–75th IQR 5–9 days) following AVR, 6 days (25–75th IQR 4–8 days) for those patients who had CAVB and 12 days (25–75th IQR 9–14 days) for those who had severe symptomatic bradycardia. The main individual characteristics of the 25 patients who received PPI after AVR are shown in table 2.
Thirty-three patients (4.2%) complicated with transient CAVB (n=16, 2.0%) or severe symptomatic bradycardia (n=17, 2.2%) and were managed exclusively with temporary pacing, with no need for PPI. The median time from AVR to these transient conduction abnormalities was 24 h (25–75th IQR 24–48 h) and their median duration was 48 h (25–75th IQR 5–72 h).
Predictive factors of PPI following AVR
Baseline, ECG and procedural variables grouped according to the need for PPI following AVR are shown in table 3. Patients who needed PPI post-AVR tended to be more frequently men (p=0.06), and their preprocedural ECG more frequently exhibited intraventricular conduction abnormalities such as LBBB (p=0.02) and RBBB (p=0.03). The results of the multivariate analysis showed that preprocedural LBBB and RBBB were the two independent predictors of PPI following AVR (OR 4.65, 95% CI 1.62 to 13.36, p=0.004 for LBBB; OR 4.21, 95% CI 1.47 to 12.03, p=0.007 for RBBB). The need for PPI stratified according to preprocedural ECG conduction abnormalities is shown in figure 1.
Prognostic value of PPI
Procedural variables, postprocedural complications and the length of hospital stay depending on the need for PPI implantation are shown in table 3. The median hospitalisation stay for the entire study population was 8 days (25–75th IQR 6–10 days), and was significantly longer in those patients who required PPI (10 days, 25–75th IQR 9–15 days) compared with those with no need for PPI (7 days, 25–75th IQR 6–10 days) p=0.0001. To further evaluate the predictive factors associated with a longer hospital stay the patients were classified according to the length of hospital stay above or below the median hospitalisation time (8 days), table 4. In the multivariate analysis, the independent predictors of a longer hospital stay were age (OR 1.08 for each increase of 1 year, 95% CI 1.04 to 1.12, p<0.0001), chronic atrial fibrillation (OR 1.76, 95% CI 1.15 to 2.68, p=0.009), chronic obstructive pulmonary disease (OR 1.68, 95% CI 1.11 to 2.51, p=0.01), cardiopulmonary bypass time (OR 1.05 for each increase of 5 min, 95% CI 1.02 to 1.09, p=0.001), intubation greater than 48 h (OR 4.29, 95% CI 1.65 to 11.20, p=0.003), need for re-operation for bleeding (OR 2.29, 95% CI 1.14 to 4.59, p=0.02), PPI (OR 4.84, 95% CI 1.78 to 13.14, p=0.002), postprocedural stroke (OR 3.94, 95% CI 1.76 to 8.80, p=0.0009) and the need for haemodialysis (OR 5.24, 95% CI 1.07 to 25.75, p=0.04).
The 30-day mortality rate for the entire study population was 3.2%, with no differences between those patients who needed PPI after AVR and those who did not (4% vs 3.2%, p=0.56). A total of 206 patients (26%) had died at a median follow-up time of 40 months. The Kaplan–Meier survival curves at 5-year follow-up for the entire study population and for those patients with and without PPI after AVR are shown in figure 2. The survival rate at 5-year follow-up among patients who underwent AVR and received PPI following the procedure was 96% (95% CI 89% to 100%) compared with 74% (95% CI 70% to 78%) among those patients who did not need PPI following AVR (p=0.12).
The rate of new PPI in a large contemporary cohort of consecutive elderly patients with severe symptomatic aortic stenosis undergoing isolated AVR was 3.2%. The main indication for PPI was CAVB (2.6%) followed by severe symptomatic bradycardia (0.6%). Most conduction disturbances occurred either during or within the 24 h following the procedure and were initially managed by temporary pacing, with PPI occurring at a median time of 8 days after AVR. The presence of preprocedural conduction intraventricular abnormalities such as LBBB or RBBB increased approximately fourfold the risk of PPI after AVR. PPI determined a longer hospital stay but had no negative effect on acute and long-term mortality in this subset of patients. Previous studies in patients undergoing isolated AVR reported an incidence of PPI between 4.1% and 7.6%.2 8 9 However, those studies included patients with a wide age range as well as aortic stenosis and regurgitation as the main underlying aortic diseases. It is well known that aortic regurgitation is a predictive factor of PPI following AVR. Limongelli et al6 showed that the need for PPI among patients undergoing AVR as a result of aortic stenosis was as low as 1.1% compared with 2.2% among those patients undergoing AVR because of aortic regurgitation. The present study, which is the first to evaluate the occurrence of conduction disturbances and the need for PPI in a large contemporary series of elderly patients undergoing isolated AVR due to severe symptomatic aortic stenosis, showed a PPI rate of 3.2% following the intervention. Two small studies conducted in the 1990s including elderly patients who had undergone AVR with or without CABG reported a PPI rate of 6%.18 19 Unlike our study, those studies included patients with both predominant aortic stenosis and regurgitation, and a significant number of patients had undergone concomitant CABG, which can also be associated with conduction disturbances.20 Also, the improvements in surgical techniques and myocardial protection during the past decade might also have contributed to the lower rate of PPI observed in our study. Consistent with our results, Thourani et al21 more recently reported an incidence of PPI of 3.6% following isolated AVR in 309 elderly patients undergoing AVR, with a rate as low as 2.3% in the group of patients more than 80 years old. Atrioventricular conduction disturbances generally originate in a triangular area bounded posteriorly by the anterior attachment of the non-coronary cusp, anteriorly by the posterior margin of the right coronary cusp and inferiorly by the junction of the membranous and muscular portions of the interventricular septum.22–24 The area below the attachment of the non-coronary cusp to the aortic wall corresponds to the anterior extremity of the His bundle before it gives off the left bundle branch (figure 3). Consequently, any direct trauma or mechanical stress to the His bundle at the region of the membranous septum and right trigone beneath the non-coronary/right coronary cusps might explain the occurrence of new conduction abnormalities leading to CAVB following AVR. Direct trauma of the conduction system during AVR might occur during debridement of a severely calcified aortic valve or during the placement of sutures after extensive valve leaflet debridement.4 8 22 Also, some surgical techniques such as the ‘continuous’ suture technique (as opposed to the ‘interrupted’ suture technique) have been associated with a higher incidence of PPI.5 The fact that most conduction abnormalities, in particular CAVB, occurred during or within the few hours following the procedure support the hypothesis of a mechanical stress on the conduction system as the main mechanism leading to conduction abnormalities and PPI. It is therefore not surprising that the presence of previous intraventricular conduction abnormalities such as LBBB or RBBB determined a higher rate of CAVB and PPI following AVR. Koplan et al7 had already shown that preprocedural RBBB was the strongest predictor of PPI following cardiac valve surgery, and Erdogan et al8 demonstrated that LBBB or RBBB were predictors of PPI following isolated AVR in young adults with aortic stenosis and/or insufficiency. The present study extends these findings to the group of older patients with pure or predominant aortic stenosis. The rate of PPI among those patients with preprocedural RBBB or LBBB was as high as 9.6% and 8.8%, respectively, compared with less than 3% among those patients without any of these intraventricular conduction abnormalities.
The inhospital mortality rate in this large contemporary series of elderly patients undergoing isolated AVR because of severe symptomatic aortic stenosis was 3.2%, which was slightly lower than the mean mortality rate predicted by the STS score (3.5%) and much lower than that predicted by the logistic EuroSCORE (10.4%). Although age has been associated with a higher risk of operative mortality,25 recent series of isolated AVR in elderly patients have reported inhospital mortality rates ranging from 2.5% to 5.2%, suggesting that continued improvements in patient selection, cardiac anaesthesia, surgical techniques, myocardial protection and postoperative management have probably led to better results in this challenging population.21 26 27 Also, the survival rate of 75% at 5-year follow-up was similar to that observed in recent studies.21 Importantly, this survival rate is substantially higher than that observed in elderly patients with severe aortic stenosis receiving medical treatment.28
The present study confirms that the need for PPI had no effect on acute and long-term mortality. However, PPI was associated with a much longer hospital stay. Previous studies suggested a longer hospitalisation in patients requiring PPI following cardiac surgery,3 4 6 but this is the first study to show that PPI is an independent predictor of a longer hospital stay, which in turn would significantly increase the costs of AVR in these patients. The use of temporary pacing for a median of 4 days following the occurrence of conduction disturbances might partly explain the longer hospital stay in these patients. Interestingly, 4.2% of the patients required temporary pacing but not PPI following AVR, and all of them recovered normal rhythm within 3 days, suggesting that the median delay of 4 days to PPI is probably adequate. Kim et al29 showed that if CAVB was present after aortic and mitral valve surgery within the first 24 h after surgery and persisted for 48 h, it was unlikely to resolve within the next 1–2 weeks, clearly suggesting that in patients otherwise ready for discharge, PPI may be performed before the previously recommended postoperative day 7.30 Our findings further support avoiding the extension beyond 3 days of the temporary pacing time in patients with CAVB following AVR in order to avoid unnecessary delays and further increase hospitalisation costs.
Although baseline and procedural data had been prospectively collected, one of the main limitations of the study lies in the retrospective nature of the analysis. Also, some aspects of the surgical technique that might have played a role on the occurrence of conduction disturbances such as the extensive use of surgical debridement were not taken into account. In the case of postoperative conduction disturbances, the time of temporary pacing until PPI was left at the discretion of the cardiac surgeon responsible for the patient and the electrophysiologist. This might partly explain the differences in the time of temporary pacing until PPI between patients.
The need for PPI following isolated AVR in elderly patients with severe symptomatic aortic stenosis was low. Pre-existing conduction abnormalities, and especially RBBB and LBBB, best predicted the need for PPI in such patients. PPI was associated with a longer hospital stay but had no effect on acute and long-term mortality. TAVI has become an alternative to AVR for the treatment of patients with aortic stenosis considered to be at very high or prohibitive surgical risk. However, the high PPI rate associated with these procedures, especially with the use of self-expandable valves, remains a concern. The morbidity and costs associated with this complication might potentially jeopardise the expansion of this technology towards a lower-risk population, and efforts should be made to reduce the incidence of PPI following TAVI procedures. Meanwhile, the results of the present study obtained in a very large contemporary cohort of elderly patients undergoing AVR should be incorporated into the clinical decision-making process for patients with severe aortic stenosis who can undergo either AVR or TAVI.
The authors would like to thank Ms. Stephanie Dionne for her outstanding work in database management and Dr Christian Couture for the anatomopathological images.
Competing interests None.
Patient consent Obtained.
Ethics approval The study protocol was performed in accordance with the institutional ethics committee and patients provided a signed informed consent for the procedures and the anonymous use of their data for research purposes.
Provenance and peer review Not commissioned; externally peer reviewed.
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.