Author + information
- Received December 17, 2014
- Revision received April 29, 2015
- Accepted May 5, 2015
- Published online August 17, 2015.
- Ashraf Hamdan, MD∗,†∗ (, )
- Victor Guetta, MD∗,
- Robert Klempfner, MD∗,
- Eli Konen, MD†,
- Ehud Raanani, MD‡,
- Michael Glikson, MD∗,
- Orly Goitein, MD†,
- Amit Segev, MD∗,
- Israel Barbash, MD∗,
- Paul Fefer, MD∗,
- Dan Spiegelstein, MD‡,
- Ilan Goldenberg, MD∗ and
- Ehud Schwammenthal, MD, PhD∗
- ∗Heart Institute, Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel
- †Department of Diagnostic Imaging, Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel
- ‡Department of Cardiac Surgery, Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine, Tel-Aviv University, Tel Aviv, Israel
- ↵∗Reprint requests and correspondence:
Dr. Ashraf Hamdan, Heart Center, Chaim Sheba Medical Center, Tel Hashomer, Sackler, Faculty of Medicine, Tel-Aviv University, Tel Hashomer 52621, Israel.
Objectives This study sought to examine whether imaging of the atrioventricular (AV) membranous septum (MS) by computed tomography (CT) can be used to identify patient-specific anatomic risk of high-degree AV block and permanent pacemaker (PPM) implantation before transcatheter aortic valve implantation (TAVI) with self-expandable valves.
Background MS length represents an anatomic surrogate of the distance between the aortic annulus and the bundle of His and may therefore be inversely related to the risk of conduction system abnormalities after TAVI.
Methods Seventy-three consecutive patients with severe aortic stenosis underwent contrast-enhanced CT before TAVI. The aortic annulus, aortic valve, and AV junction were assessed, and MS length was measured in the coronal view.
Results In 13 patients (18%), high-degree AV block developed, and 21 patients (29%) received a PPM. Multivariable logistic regression analysis revealed MS length as the most powerful pre-procedural independent predictor of high-degree AV block (odds ratio [OR]: 1.35, 95% confidence interval [CI]: 1.1 to 1.7, p = 0.01) and PPM implantation (OR: 1.43, 95% CI: 1.1 to 1.8, p = 0.002). When taking into account pre- and post-procedural parameters, the difference between MS length and implantation depth emerged as the most powerful independent predictor of high-degree AV block (OR: 1.4, 95% CI: 1.2 to 1.7, p < 0.001), whereas the difference between MS length and implantation depth and calcification in the basal septum were the most powerful independent predictors of PPM implantation (OR: 1.39, 95% CI: 1.2 to 1.7, p < 0.001 and OR: 4.9, 95% CI: 1.2 to 20.5, p = 0.03; respectively).
Conclusions Short MS, insufficient difference between MS length and implantation depth, and the presence of calcification in the basal septum, factors that may all facilitate mechanical compression of the conduction tissue by the implanted valve, predict conduction abnormalities after TAVI with self-expandable valves. CT assessment of membranous septal anatomy provides unique pre-procedural information about the patient-specific propensity for the risk of AV block.
The aortic valve is the heart’s centerpiece; it interfaces along its circumference with all cardiac cavities and is placed at the atrioventricular (AV) junction through which the conduction system runs (1,2). Transcatheter aortic valve implantation (TAVI) can therefore affect the conduction tissue at various potential levels, resulting in conduction abnormalities and the need for permanent pacemaker (PPM) implantation (3–5), but is most likely to do so where the penetrating bundle of His emerges at the surface of the left ventricular outflow tract. The more remote from the annulus the His bundle emerges at the left ventricular surface, the less likely an implanted transcatheter valve frame will impinge on it.
Because the His bundle surfaces sandwiched between the AV membranous septum (MS) and the posterior crest of the muscular septum (1,2) (Central Illustration), the lower end of the MS constitutes an anatomic landmark for the left ventricular exit point of the His bundle, with the length of the MS equaling the aortic annulus-to-His distance.
We therefore hypothesized that the length of the AV MS, readily assessable by computed tomography (CT), is a powerful anatomic predictor of the risk of high-degree AV block and the need for PPM implantation associated with TAVI, independent of procedural outcome or device-related factors. Furthermore, we hypothesized that the impact of MS length on the risk of conduction abnormalities requiring PPM implantation may be modified by tissue calcifications, also detectable by CT.
These hypotheses were studied in a consecutive group of patients with severe aortic stenosis who underwent CT as part of the routine pre-procedural evaluation before TAVI.
Ninety-two consecutive patients with symptomatic severe aortic stenosis were evaluated with CT before TAVI using self-expandable valves. Nine patients who had a previously implanted PPM, 5 patients who underwent valve-in-valve implantation, 3 patients who required a second valve implantation in the same procedure due to valve migration, 1 patient who ultimately underwent surgical aortic valve replacement, and 1 patient with a left ventricular assist device were excluded from the analysis. Thus, 73 patients represented the final study group (Figure 1). Baseline clinical, electrocardiographic, anatomic, and procedural parameters were collected for each patient, as well as information regarding the development of conduction system abnormalities and the need for PPM implantation after TAVI (Table 1). The study was approved by the local institutional review board.
CT acquisition protocol
All patients underwent native and contrast-enhanced CT using a 256-slice system (Brilliance iCT, Philips Healthcare, Cleveland, Ohio) before TAVI.
The native scan was acquired with collimation of 112 × 0.625 mm. The tube current was 377 mA at 120 kV, and the slice thickness was 2 mm. The scan direction was craniocaudal, and the scan volume ranged from the carina to the diaphragmatic face of the heart.
The contrast-enhanced CT scan was acquired with a collimation of 96 × 0.625 mm and a gantry rotation time of 330 ms. Tube current was 485 mA at 100 kV, pitch value was 0.2, and scan direction was craniocaudal. Intravenous injection of 70 to 85 ml of nonionic contrast agent (Iomeron 350, Bracco, Milan, Italy) at a flow rate of 3.5 ml/s was followed by a 30-ml saline chase bolus (5 ml/s). Automated peak enhancement detection in the descending aorta was used for timing of the scan, and the data acquisition was automatically initiated at a threshold level of 100 Hounsfield units. Acquisition was performed during an inspiratory breath-hold while the electrocardiogram was recorded simultaneously to allow retrospective gating of the data. All images were reconstructed with a slice thickness of 0.67 mm and a slice increment of 0.34 mm.
CT data analysis
The native and contrast-enhanced CT datasets were transmitted to a dedicated CT workstation (IntelliSpace Portal, version 6, Philips) to allow for multiplanar reformations. The standard coronal and sagittal views of the native CT scan were used for initial orientation at the level of the aortic valve to generate double-oblique axial images for quantitative assessment of the left, right, and noncoronary aortic cusp's calcification (Figure 2A). In addition, the basal ventricular septum was scanned for the presence of calcification (0 = no calcification, 1 = presence of calcification) in the reformatted coronal view (Figures 2B and 2C).
The contrast-enhanced CT data were reconstructed at 10% increments of the R-R interval, and the systolic phase (at 40% of the R-R interval) was used for data analysis. Because measurements of MS length in the coronal plane and double oblique plane did not show significant differences in an initially evaluated subgroup of 15 patients, MS length was determined in the nonreformatted standard coronal view, for simplicity, as shown in Figure 2D. The aortic annulus perimeter was measured after reconstruction of the aortic annulus from the sagittal and the coronal planes using double oblique multiplane reformations, as described previously (6–8).
Transcatheter aortic valve implantation
Patients underwent TAVI after a careful evaluation and discussion by the heart team. Prostheses used in the study included Medtronic CoreValve 26-, 29-, and 31-mm and Medtronic Engager 23- and 26-mm (Medtronic, Minneapolis, Minnesota). Size selection was based on perimeter measurements of the aortic annulus obtained by CT per vendor recommendations and supplemented by echocardiographic and angiographic information. The Medtronic CoreValve was implanted in 67 patients (92%), transfemorally in 59 patients and transaxillary in 8 patients, and the Medtronic Engager in 6 patients (8%) transapically.
Implantation depth was determined fluoroscopically in the implantation projection, pre-determined using CT (implanter’s view) to verify orthogonality with respect to the aortic annulus. Implantation depth was defined as the average distance from the native aortic annulus plane (on the side of the noncoronary or right cusp and on the side of the left coronary cusp), to the most proximal edge of the implanted self-expandable valve (deepest level in the left ventricle) (8,9). The perimeter of the implanted prosthesis was calculated by multiplying the nominal diameter × π. In addition, the difference between MS length (obtained from CT) and implantation depth (obtained from fluoroscopy) was calculated to define the difference between MS length and implantation depth (ΔMSID).
Continuous variables are expressed as mean ± SD, and results are presented in box plot format where appropriate. The baseline characteristics of the study patients, stratified by the presence or absence of subsequent high-degree AV block, were compared using the Mann-Whitney–Wilcoxon test for continuous variables, and the chi-square test for categorical variables. The clinical, electrocardiographic, anatomic, and post-procedural binary covariates, that were regarded as candidate risk factors for AV block after TAVI, based either on reports in the literature or on the study hypothesis, are listed in Table 1. Variables with p values <0.1 on univariate analysis were entered into 2 multivariate logistic regression models: ante-factum prediction model (only pre-procedural predictors) and post-factum prediction model (pre-procedural and post-procedural predictors). The primary outcome measures of the present study were high-degree AV block, defined as third-degree AV block or Mobitz type II second-degree AV block, and the need for PPM within 30 days after TAVI. Receiver-operating characteristic curves were constructed to illustrate and analyze the performance of varying decision thresholds, and the optimal threshold (cutoff value) was identified on a decision plot as the crossover of the sensitivity and specificity curves (10). To facilitate convenient use in varying clinical scenarios, additional weighted cutoff values were provided, accounting for differences in importance of false-negative and false-positive results (11). Statistical analysis was performed with a statistical software package (SPSS, version 17.0, SPSS, Chicago, Illinois).
Of the 73 included patients, high-degree AV block developed in 13 (18%) within 1.2 ± 1.1 days after TAVI, and 21 (29%) underwent PPM implantation within 2.2 ± 2.1 days. No pacemaker was implanted beyond day 8, and no additional higher degree conduction abnormality, sudden death, or syncope were documented until day 30. The indications for PPM were: 1) new-onset complete AV block in 9 patients; 2) new-onset Mobitz type II AV block in 4 patients; 3) new left bundle branch block with PR interval prolongation in 3 patients; 4) new left bundle branch block with atrial fibrillation with slow ventricular response in 3 patients; and 5) temporary asystole during the procedure in 2 patients.
Table 2 shows the results of the univariate and multivariate analyses of predictors of high-degree AV block after TAVI. On univariate analysis, the odds ratio (OR) for high-degree AV block was increased in female patients and in those with previous myocardial infarction (borderline significance). The OR for AV block significantly increased with shorter MS length (p = 0.01) and greater implantation depth (p = 0.006); consequently, the OR increased significantly with decreasing ΔMSID (p < 0.001). Figures 3A and 3B show the distribution of MS length and ΔMSID in patients with and without post-procedural AV block in box plot format.
Table 3 shows the results of the univariate and multivariate analyses of predictors of subsequent PPM implantation. ORs for PPM implantation were higher in patients with baseline right bundle branch block and a larger prosthesis perimeter, but only at borderline statistical significance. The presence of calcification in the basal septum significantly increased the odds of post-procedural PPM implantation (p = 0.04). The OR significantly increased with shorter MS length (p = 0.002), and greater implantation depth (p = 0.01); consequently, the OR of PPM implantation increased significantly with smaller ΔMSID (p < 0.001). Figures 3C and 3D show the distribution of MS length and ΔMSID in patients with and without subsequent PPM implantation in box plot format.
Multivariable logistic regression for the ante-factum prediction model (pre-procedural predictors alone) revealed MS length as the most powerful independent predictor of high-degree AV block (OR: 1.35, 95% confidence interval [CI]: 1.1 to 1.7, p = 0.011) and PPM implantation (OR: 1.43, 95% CI: 1.1 to 1.8, p = 0.002) (Tables 2 and 3). In the post-factum prediction model (pre-procedural and post-procedural predictors), ΔMSID was the most powerful independent predictor of high-degree AV block (OR: 1.4, 95% CI: 1.2 to 1.7, p < 0.001), and ΔMSID and calcification in the basal septum were the most powerful independent predictors of PPM implantation (OR: 1.39, 95% CI: 1.2 to 1.7, p < 0.001 and OR: 4.9, 95% CI: 1.2 to 20.5, p = 0.031; respectively) (Tables 2 and 3).
Figure 4 shows the inverse relationship between MS length and ΔMSID and the likelihood of high-degree AV block. Patients in the lowest quartile of MS length (≤6.8 mm) and ΔMSID (≤−1 mm) had the highest likelihood of high-degree AV block. The OR was 4.7 for MS length (95% CI: 1.3 to 16.4), and 11.3 (95% CI: 2.9 to 43.8) for the ΔMSID, respectively. Sensitivity/specificity decision plots yielded an MS length of 7.8 mm and a ΔMSID of 0.4 mm, respectively, as the optimal cutoff points for prediction of high-degree AV block, with negative predictive values >95% and close to 97%, respectively (Figure 5, Table 4). For prediction of PPM implantation, an MS length of 7.4 and a ΔMSID of 0.4 mm were the optimal cutoff points; negative predictive values were 86% and 89%, respectively (Figure 5, Table 4).
Interobserver and intraobserver variability for MS length was 0.23 ± 0.37 (95% CI: −0.5 to 0.97) and 0.15 ± 0.59 (95% CI: −1.03 to 1.33), respectively. The intraclass correlation coefficient for MS length was 0.97 and 0.97 for inter- and intraobserver measurements.
The proximity of the aortic annulus to the conduction system led early on to the suspicion that deeper valve implantations may be associated with a higher risk of AV block (1,3,9,12,13). Subsequent studies indeed confirmed a correlation between implantation depth and pacemaker rate (14,15). Implantation depth, however, does not reflect patient propensity for conduction abnormalities; it is a procedural outcome and thus only a predictor post-factum. Rather than just assessing the risk associated with device penetration into the outflow tract post-procedurally, the present study investigated whether individual patient susceptibility for conduction abnormalities after TAVI can be determined and quantified pre-procedurally by measuring the distance between an anatomic landmark for the location of the penetrating bundle of His (the membranous-muscular septal junction) and the aortic annulus. That distance is given by the length of the AV MS and can be measured by CT (Central Illustration). The results demonstrate that MS length varies among patients with aortic stenosis referred for TAVI and that this anatomic variation is indeed clearly associated with the risk of AV block and PPM implantation: the shorter the MS length below which the His bundle emerges, the higher the risk of conduction abnormalities and PPM implantation (Figure 6, top). A longer MS length, on the other hand, indicating a longer distance from the annulus to the His bundle, may allow accommodation of more device penetration without necessarily causing conduction abnormalities (Figure 6, bottom).
The present study, conducted prospectively in a consecutive cohort of patients who underwent TAVI with self-expandable frames, has for the first time demonstrated the value of cardiac imaging for assessing a key determinant of the variable topographic anatomy of the conduction system in the individual patient. Two key observations strongly support the pathoanatomic core concept of the study: 1) MS length emerged as the single most powerful independent pre-procedural predictor, not just of pacemaker implantation but of higher degree AV block and complete AV block, outcomes directly linked to the anatomic-physiological basis of the initial hypothesis. In fact, taking into account MS length remained a crucial value even for post-procedural risk assessment when implantation depth was known. Consequently, ΔMSID emerged, in addition to the presence of calcification in the basal septum, as the most powerful determinant for severe conduction abnormalities, integrating both pre-procedural propensity and procedural result. 2) A threshold effect could be detected for both MS length and ΔMSID, dichotomically separating patients with minimal risk (or even no risk) from the rest of the group at a cutoff close to the median; this is best explained by the concept that the junction between the AV MS and the muscular septum is indeed an anatomic landmark for the penetrating bundle of His.
Of note, although the presence of calcification in the basal septum was predictive of PPM implantation, it was not specifically associated with AV block and may therefore be related to lesser degrees of conduction abnormalities.
The observations of this study have caused us to be mindful of the AV MS length during the procedure and to take special care not to implant below its measured range. The information is also helpful to discuss with the patient the individual risk of AV block instead of relying solely on generally reported pacemaker rates.
The decision to implant a PPM was ultimately at the discretion of the attending physician. Except for Class I indications (16), the threshold for choosing to implant a PPM may differ among physicians and even more widely among institutions. To account for the impact of such variations on the outcome parameter, we selected not just the PPM rate as the dependent variable but also higher degree AV block, for which there is no physician- or institution-dependent variation and which was pre-specified as complete AV block, intermittent third-degree AV block, or Mobitz type II second-degree AV block.
The risk of severe conduction abnormalities and/or PPM implantation can also be influenced by pre-existing conduction abnormalities, in particular, pre-existing right bundle branch block. However, the anatomic-geometric parameters that can be assessed by imaging emerged clearly as the most important ones. It is likely that a CT measurement of implantation depth would have been more accurate than a fluoroscopic one and hence would have further improved the predictive value of the difference between MS length and implantation depth; however, there was no clinical indication for performing a post-deployment CT.
To eliminate confounding factors, such as metal type and mechanical properties, the study deliberately focused on self-expandable valve frames, mostly transfemorally implanted from their upstream to their downstream end. These valves may also show more variable implantation depths than balloon-expandable valves and are thus more relevant to the clinical question at hand. However, when including measurements of screened patients who underwent TAVI with balloon-expandable valves during the same study period, the relationship between MS length and ΔMSID on the one hand, and AV block or PPM implantation on the other, remained unchanged.
MS length, as assessed by CT, is the single most powerful pre-procedural predictor of AV block and the need for PPM implantation after TAVI with self-expandable valves. ΔMSID is the single most powerful post-factum predictor of AV block and—together with calcification in the basal septum—the most powerful post-procedural predictor of PPM.
WHAT IS KNOWN? Due to the proximity of the aortic valve to the AV conduction system, TAVI can result in the need for pacemaker implantation, a risk that is increasing with implantation depth.
WHAT IS NEW? The present study demonstrates that this risk is also inversely related to the length of the AV MS (an anatomic landmark for the position of the His bundle), which varies widely among patients. CT assessment of MS length allows quantifying a patient’s anatomic risk of developing serious AV conduction abnormalities after TAVI.
WHAT IS NEXT? It should be prospectively tested whether procedural use of this information can translate into better outcomes.
Dr. Segev is a paid consultant of Edwards Lifesciences and Medtronic. Dr. Schwammenthal is a Proctor for Medtronic Inc. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Drs. Hamdan and Guetta contributed equally to this work.
- Abbreviations and Acronyms
- confidence interval
- computed tomography
- membranous septum
- difference between membranous septum length and implantation depth
- odds ratio
- permanent pacemaker
- transcatheter aortic valve implantation
- Received December 17, 2014.
- Revision received April 29, 2015.
- Accepted May 5, 2015.
- American College of Cardiology Foundation
- Piazza N.,
- de Jaegere P.,
- Schultz C.,
- Becker A.E.,
- Serruys P.W.,
- Anderson R.H.
- Yen Ho S.,
- Ernst S.
- Piazza N.,
- Onuma Y.,
- Jesserun E.,
- et al.
- Bleiziffer S.,
- Ruge H.,
- Hörer J.,
- et al.
- Hamdan A.,
- Guetta V.,
- Konen E.,
- et al.
- Achenbach S.,
- Delgado V.,
- Hausleiter J.,
- Schoenhagen P.,
- Min J.K.,
- Leipsic J.A.
- Urena M.,
- Mok M.,
- Serra V.,
- et al.
- Sinhal A.,
- Altwegg L.,
- Pasupati S.,
- et al.
- Brignole M.,
- Auricchio A.,
- Baron-Esquivias G.,
- et al.