Abstract: Background. To determine predictive variables for permanent pacemaker (PPM) insertion after transcatheter aortic valve replacement (TAVR) with the CoreValve Revalving System (CRS). Methods and Results. A total of 121 patients with severe aortic stenosis (AS) were recruited from six Asian medical centers between March 2010 and May 2013. Four patients with preexisting PPM were excluded. The mean age of the remaining 117 patients was 81.2 ± 5.1 years. Twenty-three patients (19.7%) required post-TAVR PPM, with a median time-to-insertion of 7 days (interquartile range, 5-13 days). Two variables were identified as independent predictors of PPM: (1) device depth from the non-coronary cusp (NCC) (odds ratio [OR], 1.263; P=.02) determined by aortic root angiography; and (2) the perimeter stretching index (OR, 1.584; P<.001) determined by computed tomography. The predictive cut-off values were as follows: a perimeter stretching index >1.13 (P<.001) and a device depth from the NCC >7.8 mm (P<.001). The diagnostic accuracy of these variables was 93.2% and 71%, respectively. Conclusion. Depth of the device from the NCC and the perimeter stretching index are independent predictors of PPM insertion after CRS-TAVR.
J INVASIVE CARDIOL 2015;27(7):334-340
Key words: aortic stenosis, transcatheter aortic valve replacement, CoreValve, pacemaker
Transcatheter aortic valve replacement (TAVR) is an alternative treatment for patients with inoperable aortic stenosis (AS) or for those categorized as high-risk for surgical aortic valve replacement (AVR).1,2 However, conduction disturbances are relatively common after TAVR, and the occurrence of a complete atrioventricular block (AVB) requires the insertion of a permanent pacemaker (PPM).3 The frequency of post-TAVR PPM insertion ranges from 4.9%-28.9%, depending on the type of TAVR device used.3 A PPM is needed more frequently after insertion of a CoreValve Revalving System (CRS; Medtronic) device than after insertion of the Edward Sapien (Edwards Lifesciences) device.3 Although several baseline clinical and electrocardiographic variables have been suggested as potential AVB risk factors,4 data on the prevention of post-TAVR AVB development are limited. Here, we attempted to determine the factors that allow the prediction of post-TAVR PPM.
Patients. Between March 2010 and May 2013, a total of 121 patients with severe symptomatic AS, whose AS was inoperable or for whom the anticipated mortality or morbidity from surgical AVR was considered too high, were recruited from six Asian medical centers and enrolled in the study. Four patients had a preexisting PPM and so were excluded. The remaining 117 patients were included in this study. The eligibility of each patient for TAVR was established at the centers, based on the consensus of the local multidisciplinary heart team, which included cardiologists, cardiac surgeons, and cardiac anesthesiologists.
Prior to TAVR, all patients underwent echocardiography, iliofemoral angiography, coronary and aortic root angiography, and computed tomography (CT) angiography. All patients were scheduled to receive postenrollment clinical evaluations, including electrocardiography (ECG), echocardiography, and/or CT at 1, 6, and 12 months post procedure and then annually thereafter. The local institutional review board at each participating center approved the protocol. Written informed consent was obtained from all patients.
Implantation procedures. The choice of anesthesia used for the TAVR procedure (general anesthesia or conscious sedation) was made by the heart team during the preoperative meeting. The preoperative imaging tests, which in most cases included angiography and CT angiography, determined the location and method used for arterial access by percutaneous puncture or surgical exposure. For the majority of percutaneous femoral access patients, hemostasis was achieved by means of 2 or 3 Perclose ProGlide (Abbott Vascular).
After the procedure, the majority of patients were managed for at least 1 day in intensive or coronary care units, with continuous ECG monitoring. In all patients, a temporary pacemaker was used for a minimum of 48 hours after TAVR, as a back-up against paroxysmal or persistent advanced AVB. On the cardiology ward, patients were monitored with telemetry until discharged. An ECG was printed daily and added as a source document to the chart of each patient.
ECG and PPM insertion data. ECG tracings obtained before and after treatment were interpreted in the core laboratory, along with those obtained at the 1-month follow-up visit. The ECG records were analyzed for cardiac rhythm, PR and QRS intervals, the presence of left- or right-axis deviation, right bundle-branch block (BBB), left BBB, and the degree of AVB. Periprocedural data were obtained regarding the occurrence of AVB (either transient or persistent), the use of balloon predilation and postdilation of the aortic valve, and the size of the balloon used. PPM insertions, and the number of days that they occurred after TAVR, were also recorded.
Implanted depth of the device. The distances from the lower edge of the non-coronary cusp (NCC) and from the left coronary cusp (LCC) to the basal skirt of the device were measured using quantitative angiographic techniques (CASS II; Pie Medical) (Figure 1A).
Echocardiographic and CT parameters. The echocardiographic aortic valve area, the interventricular septal diameter, and the aortic annulus size were measured in all patients, along with the routine measurement of ejection fraction, valvular function, dimensions, and pulmonary hypertension.
CT parameters were collected on three different planes: coronal, sagittal, and double oblique (basal plane) during mid-systole. CT data were used for the curved multiplanar reconstructions, which were performed by tracing a line through the center point of the proximal ascending aorta, the aortic valve, annulus, and the left ventricular outflow tract. The basal plane was defined as a plane perpendicular to the curved multiplanar reconstruction line at the ventricular aspect to the point at which all three leaflets were seen to disappear. This approximated the nadir of the three leaflets and generated an image defined as the annular plane.5
Calibrated images from the curved multiplanar reconstruction annular plane were generated using Vitrea software (Vitrea 6.3; Vital Images), and a polygonal line circumscribing this plane was traced to determine its perimeter. Based on this measurement, the “perimeter stretching index” was calculated, and was defined as the calculated perimeter of the device divided by the measured CT-measured perimeter of the annulus (Figure 1B). The “stretching index diameter” is the diameter of the device divided by the CT-measured mean annulus diameter. The aortoseptal angles, and the amount of calcium in the individual cusps of the aortic and mitral valves, were also calculated using Vitrea software.
Postprocedural electrophysiology. Postprocedural electrophysiological studies were performed in 18 patients immediately after CRS implantation. Multiple electrocardiographic leads (lead I, aVF, and V1) and intracardiac bipolar electrograms (filtered between 30 and 500 Hz) were simultaneously displayed and recorded using a digital electrophysiological recording system (CardioLab). Three quadripolar catheters were positioned in the high right atrium, right ventricular apex, and the His bundle position. The AH- and HV-intervals were assessed in these 18 patients: an AH interval of 60 to 125 ms and an HV interval of 35 to 55 ms were considered normal.
Statistical analysis. Continuous variables are expressed as the mean ± standard deviation, or as the median and interquartile range. Categorical variables are expressed as number or percentage. The Wilcoxon rank sum test for continuous variables and the χ2 test or Fisher’s exact tests for categorical variables were performed, as appropriate. The distributions of the QRS duration measurements were skewed to the right; therefore, natural logarithms were used to provide a more even distribution for analysis. Logistic regression models were used to identify variables that predicted PPM. Unadjusted models for each covariate were fitted, and multivariable models were built, using backward elimination, with an inclusion criterion of P<.40.
Receiver operator characteristic curves were created using MedCalc (MedCalc Software), and were analyzed to assess the best cut-off values for the perimeter stretching index and the depth of device from the NCC that would minimize the distance between the curves and the upper corners. The sensitivities, specificities, positive predictive values, and negative predictive values, with 95% confidence intervals (CIs), were then obtained. Using the best cut-off values of the perimeter stretching index and the device depth, we constructed four groups and measured their discrimination abilities by calculating the area under the curve (AUC).
A P-value <.05 was considered significant. All data were processed using the Statistical Package for Social Sciences, version 18 (SPSS, Inc).
Baseline clinical data. Table 1 shows the baseline clinical characteristics of the patients. The mean age was 81.2 ± 5.1 years, and the mean logistic European System for Cardiac Operative Risk Evaluation was 19.2 ± 12.2%.
Procedural results and clinical outcomes (Table 1). The CRS devices were implanted successfully in 91.5% of procedures. A two-valve CRS implantation was performed in 8.6% of patients. Balloon predilation and postdilation were used in 58.9% and 35.9% of procedures, respectively. The three CoreValve device sizes (26, 29, and 31 mm) were used in 46.2%, 43.6%, and 10.3% of procedures, respectively. The access site was transfemoral for 92% of patients (n = 108); direct aortic and subclavian access sites were used in 5% (n = 6) and 2.5% (n = 3) of patients, respectively (data not shown). For all patients, the implantation depths of the CoreValve devices were measured on the postprocedural angiogram; the average depths from the NCC and LCC were 6.3 ± 3.6 mm and 8.4 ± 3.9 mm, respectively.
Postprocedural ECG changes (Table 1). The mean baseline PR interval of the patients was 170 ± 40.1 ms. In this cohort, 14.5% of patients had baseline right BBB, but there was no baseline left BBB. There was a significant postprocedural increase in the duration of the QRS complex from 101.4 ± 21.7 ms to 134.3 ± 32.6 ms (P<.001 by Wilcoxon rank sum test). New-onset left BBB developed in 38.5% of patients, but there was no new-onset right BBB. In total, 59.8% of patients (n = 70) developed AVB during the procedure, either before or after balloon dilatation, or during CRS implantation, and 31.5% of them underwent PPM insertion. Of the patients with new-onset left BBB, 29.3% underwent PPM insertion (data not shown).
Echocardiography and CT (Table 1). In this cohort, the post-TAVI transaortic mean and peak pressure gradients were significantly lower than baseline (reduced from 58.1 ± 18.6 mm Hg to 10.9 ± 5.9 mm Hg, and from 100.5 ± 31.5 mm Hg to 23.0 ± 8.8 mm Hg, respectively; P<.001 [data not shown]). The mean baseline annular size, measured by transesophageal echocardiography, was 22.2 ± 2.3 mm. CT using the coronal and sagittal views showed the annulus size to be 22.6 ± 2.2 mm and 19.9 ± 2.0 mm, respectively. The mean perimeter at baseline was 78.6 ± 6.9 mm, and the perimeter stretching index was 1.12 ± 0.06.
Univariate and multivariate analysis to determine PPM insertion predictors. There was no significant difference in the need for PPM among the centers (P=.41 by Fisher’s exact test; data not shown). One patient received a PPM on the same day as the index procedure. The median time-to-insertion for the 23 patients (19.7%) who received a PPM was 7 days (interquartile range, 5-13 days) (Table 1).
Table 2 shows the predictors of PPM insertion after CRS implantation according to univariate and multivariate analyses. Univariate logistic regression showed that PPM insertion was significantly correlated to the presence of left-axis deviation (P=.04), peri-implantation AVB (P=.04), and new-onset left BBB (P=.03). PPM insertion was also significantly correlated to the device depth, and to the perimeter stretching index. There was no relationship between PPM insertion and the other baseline characteristics, ECG, or procedural findings. Multivariate regression analysis showed that the strongest independent indicators of PPM insertion were the perimeter stretching index (OR, 1.548; 95% CI, 1.239-1.935; P<.001) and the device depth from the NCC (OR, 1.263; 95% CI, 1.034-1.543; P=.02).
Using these two independent PPM predictors, the cut-off values that best predicted the need for PPM insertion were calculated. The cut-off value for the perimeter stretching index was 1.13 (87% sensitivity; 95% specificity; AUC = 0.91; 95% CI, 0.82-0.992; P<.001), with an overall diagnostic accuracy of 93.2% (Figure 2). The cut-off value of the device depth from the NCC was 7.8 mm (61% sensitivity; 74.5% specificity; AUC = 0.70; 95% CI, 0.593-0.806; P<.001), with an overall diagnostic accuracy of 71% (Figure 2).
Postprocedural electrophysiology. Eighteen patients were electrophysiologically evaluated immediately post procedure; of these, 15 were in sinus rhythm. With the exception of 1 patient who exhibited a complete AVB and H-V block (H-V interval, 70 ms), the remaining patients showed no prolongation of their A-H and H-V intervals; however, 6 patients who initially exhibited normal H-V intervals received PPM within 30 days (Table 3).
Impact of persistent left bundle-branch block following CoreValve implantation. Baseline clinical and procedural characteristics of the patients, grouped according to the occurrence of persistent left BBB (vs no left BBB) following the TAVR procedure are shown in Supplemental Table 1 (available online at www.invasivecardiology.com). In patients with left BBB, the widening of the QRS was remarkable, reaching a mean duration of 161.6 ms. In the same group, the implantation depth of the devices was deeper than in the group of no left BBB. However, there was no difference in the rate of PPM insertion between the two groups.
The major findings of this study are as follows: (1) in this cohort, 19.7% of patients received a PPM following CRS implantation; (2) multivariate analysis showed that the perimeter stretching index determined by CT and the device depth from the NCC are the best predictors of the need for PPM insertion; and (3) the optimum values of these variables required to prevent PPM insertion are a stretching index of <1.13 and a device depth of <7.8 mm.
The number of TAVR patients is increasing: currently, more than 65,000 patients have been implanted with the CoreValve device worldwide (September 2014, data from Medtronic, Inc). However, subsequent requirement for a PPM is a complication of this procedure. A meta-analysis showed that TAVR patients using the CRS have a higher incidence of subsequent PPM than those receiving the Edwards valve.2,3
The baseline variables that best predict the need for a PPM after CoreValve implantation include female gender, low ejection fraction,6 baseline right BBB,8 presence of left-axis deviation,7,9 thickness of the interventricular septum,4,7 thickness of the native NCC,4,7,8 and the amount of calcification.6 Periprocedural predictors include larger device, periprocedural AVB, balloon predilation,4 and the device depth.7,9,10 Many of these predictive variables are baseline risk factors; others, such as the use of balloon predilation or a larger device, are based on decisions that are tailored to individual circumstances. Therefore, we focused on the procedural factors that can be modified to prevent postprocedural PPM implantation.
Piazza et al8 reported that the device depth significantly differed between patients who developed left BBB and those who did not; left BBB was not observed in patients with implant depths of <6.7 mm. Subsequent work has postulated a cut-off of 6.0 mm as an independent predictor of the development of left BBB and the requirement for a PPM.12 The present study also found that device depth from the NCC was a significant predictor of subsequent PPM insertion (OR, 1.263; 95% CI, 1.034-1.543; P=.02). This result is similar to that of Muñoz-García et al;10 however, a larger and more recent study by Khawaja et al found no such association.4
Precise annulus sizing is fundamental to the prevention of serious complications and successful device implantation. The best aortic imaging methods for annulus measurements are still a matter for debate. Echocardiography measurement underestimates the size of the annulus and so is of limited use.13 Three-dimensional CT, which is not limited by planar imaging, provides fine anatomical details and accurate assessment of the complex anatomical shapes of the aortic valve and the annulus.14 The area-derived diameter and the perimeter of the basal ring of the annulus are the most reproducible of the CT annulus measurement parameters.3,14
In the present study, we suggested that the perimeter stretching index, defined as the perimeter of the device divided by the CT-measured perimeter of the annulus, could be used to predict the need for PPM. The stretching index focuses on the native valve, and the injury it sustains by being stretched by the device. Nitinol in the self-expanding CoreValve frame provides a constant outward pressure on and below the aortic annulus, and is highly resistant to inward pressure and crushing.15 These constant pressures on the native annulus may cause inflammation and subsequent edema, leading to impaired function of the atrioventricular node and of the His bundle and its branches.8,16
In this study, we performed post-TAVR electrophysiology on 18 patients in order to observe the stretching effect of prosthesis during hospitalization (Table 3). Apart from 1 patient, who had a complete AV and H-V block, the remaining patients showed no prolongation of the H-V interval immediately after the procedure. However, 6 patients developed complete AVBs and received subsequent PPM. The median time-to-insertion of PPM was 7.0 days (interquartile range, 5-13 days). These results suggest that the effects of sustained pressure on the aortic annulus are responsible for the impaired function of the conduction system.
Study limitations. First, since data were obtained from six Asian medical centers, it was not possible to obtain clear, retrospective indications for PPM insertion for all patients. The attending physicians at each center determined the requirement for PPM insertion according to the Guidelines for Device-Based Therapy of Cardiac Rhythm Abnormalities.17 We did not recommend prophylactic insertion6,11 in asymptomatic patients who did not meet the PPM insertion criteria; instead, we waited approximately 1 week with a temporary pacemaker in place. Also, based on the decisions of the attending physicians, some patients without high-grade AVB underwent PPM insertion due to the detection of H-V blocks by electrophysiology testing. However, despite the use of this new technology, we did not find any differences in the PPM insertion rates among the medical centers. This is possibly due to the extensive proctoring program implemented by CoreValve Medtronic. Second, we were only able to examine a small number of patients using electrophysiology. This type of evaluation should be performed on a large number of patients to properly evaluate the conduction disturbances that might be caused by use of the CRS. Third, the sample sizes included in this study were relatively small. Therefore, our findings should be confirmed or rebutted by large, prospective studies.
In this population, the CoreValve device was associated with a post-TAVR PPM insertion rate of 19.7%. There was an increased rate of PPM insertion in patients in whom the device depth from the NCC was large, and in those whose native aortic valve had been overstretched; the overall median time-to-implantation was 7 days (interquartile range, 5-13 days). To reduce the need for post-TAVR PPM, care must be taken when selecting the device size to prevent overstretching of the native valve (perimeter stretching index, <1.13), and the device should not be implanted too deep (device depth from the NCC, <7.8 mm).
Acknowledgment. The authors thank Dr Gi-Byung Nam for contributing electrophysiology studies to this report.
- Smith CR, Leon MB, Mack MJ, et al. Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med. 2011;364:2187-2198.
- Leon MB, Smith CR, Mack M, et al. Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med. 2010;363:1597-1607.
- Jilaihawi H, Chakravarty T, Weiss RE, Fontana GP, Forrester J, Makkar RR. Meta-analysis of complications in aortic valve replacement: comparison of Medtronic-CoreValve, Edwards-Sapien and surgical aortic valve replacement in 8,536 patients. Catheter Cardiovasc Interv. 2012;80:128-138.
- Khawaja MZ, Rajani R, Cook A, et al. Permanent pacemaker insertion after CoreValve transcatheter aortic valve implantation: incidence and contributing factors (the UK CoreValve Collaborative). Circulation. 2011;123:951-960.
- Jilaihawi H, Kashif M, Fontana G, et al. Cross-sectional computed tomographic assessment improves accuracy of aortic annular sizing for transcatheter aortic valve replacement and reduces the incidence of paravalvular aortic regurgitation. J Am Coll Cardiol. 2012;59:1275-1286.
- Latsios G, Gerckens U, Buellesfeld L, et al. “Device landing zone” calcification, assessed by MSCT, as a predictive factor for pacemaker implantation after TAVI. Catheter Cardiovasc Interv. 2010;76:431-439.
- Jilaihawi H, Chin D, Vasa-Nicotera M, et al. Predictors for permanent pacemaker requirement after transcatheter aortic valve implantation with the CoreValve bioprosthesis. Am Heart J. 2009;157:860-866.
- Piazza N, Onuma Y, Jesserun E, et al. Early and persistent intraventricular conduction abnormalities and requirements for pacemaking after percutaneous replacement of the aortic valve. JACC Cardiovasc Interv. 2008;1:310-316.
- Bleiziffer S, Ruge H, Horer J, et al. Predictors for new-onset complete heart block after transcatheter aortic valve implantation. JACC Cardiovasc Interv. 2010;3:524-530.
- Muñoz-García AJ, Hernández-García JM, Jiménez-Navarro MF, et al. Factors predicting and having an impact on the need for a permanent pacemaker after CoreValve prosthesis Implantation using the new Accutrak Delivery Catheter System. JACC Cardiovasc Interv. 2012;5:533-539.
- Fraccaro C, Buja G, Tarantini G, et al. Incidence, predictors, and outcome of conduction disorders after transcatheter self-expandable aortic valve implantation. Am J Cardiol. 2011;107:747-754.
- Piazza N. Data presented at I CoreValve Proctors Meeting. Hamburg, Germany: 2009.
- Willson AB, Webb JG, LaBounty TM, et al. 3-dimensional aortic annular assessment by multidetector computed tomography predicts moderate or severe paravalvular regurgitation after transcatheter aortic valve replacement: a multicenter retrospective analysis. J Am Coll Cardiol. 2012;59:1287-1294.
- Gurvitch R, Webb JG, Yuan R, et al. Aortic annulus diameter determination by multidetector computed tomography: reproducibility, applicability, and implications for transcatheter aortic valve implantation. JACC Cardiovasc Interv. 2011;4:1235-1245.
- Rieu R, Barragan P, Masson C, et al. Radial force of coronary stents: a comparative analysis. Catheter Cardiovasc Interv. 1999;46:380-391.
- Piazza N, Nuis RJ, Tzikas A, et al. Persistent conduction abnormalities and requirements for pacemaking six months after transcatheter aortic valve implantation. EuroIntervention. 2010;6:475-484.
- Members WC, Epstein AE, DiMarco JP, et al. ACC/AHA/HRS 2008 guidelines for device-based therapy of cardiac rhythm abnormalities: a report of the American College of Cardiology/American Heart Association task force on practice guidelines (writing committee to revise the ACC/AHA/NASPE 2002 guideline update for implantation of cardiac pacemakers and antiarrhythmia devices): developed in collaboration with the American Association for Thoracic Surgery and Society of Thoracic Surgeons. Circulation, 2008;117:e350-e408.
*Joint first authors.
From the 1Asan Medical Center, University of Ulsan College of Medicine, Seoul, Korea; 2CHA Bundang Medical Center, CHA University School of Medicine, Gyeonggi-do, Korea; 3Chinese General Hospital and Medical Center, Manila, Philippines; 4St. Luke Hospital, Manila, Philippines; 5Queen Elizabeth Hospital, Hong Kong, China; 6Seoul National University Hospital, Seoul, Korea; 7Severance Hospital, Seoul, Korea; and 8University Hospital Bonn, Bonn, Germany.
Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. The authors report no conflicts of interest regarding the content herein.
Manuscript submitted July 8, 2014, provisional acceptance given August 25, 2014, final version accepted October 27, 2014.
Address for correspondence: Seung-Jung Park, MD, PhD, Department of Cardiology, University of Ulsan College of Medicine, Cardiac Center, Asan Medical Center, 388-1 Poongnap-dong, Songpa-gu, Seoul, 138-736 Korea. Email: sjpark@ amc.seoul.kr