Author + information
- Received August 26, 2009
- Revision received November 2, 2009
- Accepted November 5, 2009
- Published online February 1, 2010.
- Daniel John, MD,
- Lutz Buellesfeld, MD⁎ (, )
- Seyrani Yuecel, MD,
- Ralf Mueller, MD,
- Georg Latsios, MD,
- Harald Beucher, MD,
- Ulrich Gerckens, MD and
- Eberhard Grube, MD
- ↵⁎Reprint requests and correspondence:
Dr. Lutz Buellesfeld, HELIOS Heart Center Siegburg, Ringstrasse 49, 53721 Siegburg, Germany
Objectives The aim of this study was to assess the influence of amount and distribution of calcifications of the aortic valve and the left ventricular outflow tract on the acute procedural outcome of patients undergoing transcatheter aortic valve implantation (TAVI).
Background Transcatheter aortic valve implantation is a new percutaneous technique especially for elderly, high-risk patients with significant aortic valve stenosis (AS). After TAVI, post-interventional paravalvular aortic regurgitations (PAR) can occur, which is believed to be related partially to valve calcifications.
Methods We prospectively analyzed 100 symptomatic patients with severe AS scheduled for TAVI with the CoreValve ReValving (Medtronic, Minneapolis, Minnesota) prosthesis. In all patients, a native and contrast-enhanced multislice cardiac computed tomography was performed pre-interventionally. Calcification load of the valve and the adjacent outflow tract was estimated by the Agatston Score (AgS), and the amount and distribution of calcification was semi-quantitatively assessed and graded on a 1 to 4 scale (device “landing zone” calcification score [DLZ-CS]). Aortography was performed to evaluate the PAR pre-interventionally, after initial device release (PAR0) and after termination of the procedure (PAR1). Transthoracic echocardiography was performed 2 weeks after implantation (PAR2).
Results The AgS and DLZ-CS showed a significant correlation with the grade of PAR0 (AgS: r = 0.329, p = 0.001; DLZ-CS: r = 0.356, p < 0.001), PAR1 (AgS: r = 0.254, p = 0.011; DLZ-CS: r = 0.240, p = 0.016), and PAR2 (AgS: r = 0.341, p = 0.001; DLZ-CS: r = 0.300, p = 0.002). Both scores (AgS and DLZ-CS) showed a significant positive correlation (r = 0.858, p < 0.001).
Conclusions Calcification in the CoreValve device “landing zone” shows a significant positive correlation to PAR after TAVI. Furthermore, the need for “second maneuvers” (i.e., post-dilation after initial device release) can be predicted by these calcification scores (AgS and DLZ-CS).
Percutaneous transcatheter aortic valve implantation (TAVI) was introduced in 2002 (1) for treatment of severe symptomatic aortic valve stenosis (AS) in patients not eligible for surgical valve replacement. Today, TAVI is a rapidly evolving field with almost exponentially increasing numbers of treated patients as well as centers performing this intervention. Data available so far demonstrate feasibility, safety, and efficacy of this young technology in patients considered high-risk for conventional surgical valve replacement (1–8), although these results are based on nonrandomized registry data and need to be validated in randomized studies.
By now, 2 devices for TAVI have been CE-marked—the balloon-expandable Edwards-Sapien prosthesis (Edwards Lifesciences, Inc., Irvine, California) and the self-expandable CoreValve ReValving System (Medtronic, Minneapolis, Minnesota).
A clinically relevant potential side effect of TAVI is the development of aortic regurgitation (AR) after valve implantation. A mild-to-moderate, post-interventional AR is present in approximately 50% of patients (6). Previously published data from our group demonstrated a “33% rule”: probability of unchanged, improved, and worsened grades of AR is approximately 33% each (8). In 5% to 10% of cases, severe AR occurs immediately after device release, which can lead to serious and life-threatening problems (2,3,6) and which requires additional “maneuvers” to be corrected. If AR after TAVI occurs, it is mainly localized in the paravalvular area (6).
Despite the clinical relevance and its frequency, there are only limited data available on this important potential problem. However, recent observations indicate that the degree of post-interventional paravalvular aortic regurgitation (PAR) depends on aortic valve calcifications (AVC), which prevents a sufficient apposition of the prosthesis to the native annulus (7).
Early developed calcification scores from the pre-TAVI era confirmed the aortic valve as the interesting area in surgical valve replacement (9,10) and were recently transferred to the collective of high-risk patients designed for TAVI (11,12) without consideration of the lower part of the CoreValve device landing zone (DLZ) (i.e., the left ventricular outflow tract [LVOT]). Therefore, we turned our attention on the aortic valve and the LVOT, which both are part of the CoreValve DLZ.
The aim of this study was to assess the influence of calcifications of the device “landing zone” on the procedural success of the CoreValve prosthesis in patients with hemodynamically significant AS. Furthermore, a semi-quantitative (i.e., visual 4-step) calcification score describing calcifications in the DLZ was established and implemented to predict the potential interventional outcome (i.e., incidence of significant paravalvular leakage and need for post-dilation).
A prospective single-center nonrandomized study was performed to assess the influence of the total amount and distribution of calcification in the DLZ on the short-term device outcome with regard to PAR in patients undergoing TAVI with the third generation of the CoreValve prosthesis. Table 1 provides an overview on the examination steps. All patients underwent computed tomography (CT) screening before the procedure as described in the following text. The AR during the procedure were assessed by repeated supra-aortic angiograms immediately after device deployment (PAR0) as well as at the end of the procedure (PAR1) on the basis of the established angiographic 0 to 4+ scale according to Baim and Grossman (13).
Two weeks after the procedure, transthoracic echocardiography was performed to assess short time outcome of the hemodynamic valve status with particular evaluation of the regurgitation grade (PAR2). The procedural result concerning PAR (PAR0, PAR1, and PAR2) was analyzed by the operators, who were blinded to the CT data concerning the amount and distribution of calcification in the DLZ.
Clinical follow-up included 30 days after the procedure. All patients signed an informed written consent. This study was approved by the local medical ethics committee.
We included 100 symptomatic high-risk patients (age 82.10 ± 6.25 years; 43 men) with severe AS (aortic valve area: 0.61 ± 0.17 cm2; mean transvalvular aortic pressure gradient: 41.4 ± 14.6 mm Hg). Additional AR was present at baseline in 79 cases (79%) (Table 2). Patient characteristics and comorbidities are presented in Tables 2 and 3.⇓ A patient was considered high-risk for conventional surgery when excessive morbidity and/or mortality (logistic EuroSCORE >20%, Society of Thoracic Surgeons score >10%) was predicted. Mean logistic EuroScore was 24.8 ± 18.8%, and mean Society of Thoracic Surgeons score was 21.6 ± 9.9%.
Baseline cardiac CT evaluation
In all patients, a native and contrast-enhanced multislice computed tomography (MSCT) (Philips Mx8000 IDT [Philips Medical Systems, Best, the Netherlands]; 16 × 0.75 mm; 120 kV; 160 mAs) was performed before the procedure. The analysis of the native and contrast-enhanced images was separately accomplished with a commercially available computer workstation (Philips Extended Brilliance Workstation, Philips Medical Systems) by 2 experienced investigators (S.Y., D.J.). Both investigators were blinded to clinical data.
The native MSCT was performed for the quantitative assessment of the amount of calcification in the DLZ with the Agatston score (AgS) as described elsewhere (14). The area of calcification in the DLZ was defined as at least 4 adjacent pixels with a density of ≥130 Hounsfield Units measured by native electrocardiography-gated cardiac MSCT. The DLZ was defined as the area including the aortic valve (i.e., the aortic annulus and valvular cusps) and the LVOT (until the junction point of the anterior mitral leaflet). The measurement of the amount of calcification via AgS was performed in transverse source images in a standardized way and was averaged for statistical analysis.
The minimum requirements for the cardiac-CT CoreValve screening process (i.e., the technical data for a sufficient data acquisition for the contrast-enhanced MSCT) are shown in Table 4.
For the semiquantitative (i.e., visual) estimation of the amount and distribution of calcification in the DLZ, contrast-enhanced cardiac CT images were used. Thus, for evaluation of calcification in the DLZ in a 3-dimensional way, multiplanar reformations of the heart (including the LVOT and the aortic valve) and the ascending aorta were used. For this, 2 projections (i.e., coronal and single oblique sagittal view) were orthogonal orientated. The correct orientation was then confirmed by inspection of the so-called double oblique transverse projection (Fig. 1). With this projection, we assessed amount of calcification on the level of the aortic valve, cusp accentuation, calcification-caused commissural fusion (all via transversal view) and calcification clumps outreaching the annulus level toward the LVOT, the latter via coronal single oblique sagittal view (Fig. 1). Atherosclerosis of the upper aortic root (distal to the annulus level) of the proximal coronary arteries and the mitral annulus were excluded from the semi-quantitative assessment of calcification.
The total amount of calcification in the DLZ was assessed semi-quantitatively and rated on the basis of the following so-called DLZ calcification score (DLZ-CS): grade 1 = mild calcification, 2 = moderate calcification, 3 = heavy calcification (mostly associated with commissural fusion), and 4 = massive calcification including big calcification clumps outreaching the annulus level. Examples of each grade are displayed in Figure 2. In addition to this scoring system, the distribution of aortic cusp calcification was assessed as symmetrical versus asymmetrical distribution. In case of asymmetrical distribution, the most calcified cusp was identified and fed into the analysis to detect calcification patterns in the overall study population.
The quantitative amount of calcification in the DLZ was also rated on a 1 to 4 scale on the basis of the AgS (grade 1: AgS ≤1,000 AU; grade 2: AgS 1,001 to 3,000 AU; grade 3: AgS 3,001 to 5,000 AU; grade 4: AgS >5,000 AU) and compared with the DLZ-CS.
CoreValve device description and procedure
The current third-generation 18-F CoreValve aortic valve prosthesis consists of a trileaflet bioprosthetic valve made of porcine pericardial tissue, which is mounted and sutured in a self-expanding nitinol frame. The prosthetic frame is manufactured by laser-cutting of a nitinol metal tube with a total length of 50 mm. The upper portion of this prosthesis is flared to anchor the stent in the ascending aorta and coaxial alignment. The middle portion is constrained to avoid direct coverage of the coronary arteries. The lower part carries the tissue valve and expands with high radial force within the native valve annulus.
Applied strategies in case of post-deployment regurgitation
Invasive aortography was performed and analyzed by the operators to evaluate pre-procedural AR, PAR after initial release (PAR0), and acute post-procedural AR (PAR1) and was used as the basis for the decision for further strategies. Interventional options in case of more-than-moderate post-deployment regurgitation (PAR0 grade >2+) included post-dilation, device repositioning with a snare catheter and/or implantation of a 2nd CoreValve prosthesis (valve-in-valve replacement/2 valves “in series”). Application of these escalating measures was at the operator's discretion.
Device success was defined as stable device placement and adequate function in the first attempt as assessed by angiography. Acute procedural success was defined as device success with absence of periprocedural major adverse cardiovascular and cerebral events including cardiac tamponade in the first 24 h after device implantation. Major adverse cardiovascular and cerebral events consisted of death from any cause, myocardial infarction (creatine kinase-myocardial band more than 2 times the upper limit of normal), and stroke (as assessed by routine neurological assessment before and after procedure and before hospital discharge). Clinical adverse events were adjudicated by an independent clinical events committee.
All data were analyzed with the SPSS statistical utility software (SPSS, Inc., Chicago, Illinois). Categorical variables are presented as frequencies; continuous variables are presented as mean ± SD. Differences were assessed with a paired-sample t test for normally distributed data. For correlations, Pearson bivariate analysis with a 2-tailed test for significance was used. A p value < 0.05 was considered significant.
The procedural data are shown in Table 6. In 97 of 100 patients a trans-femoral approach was used without general anesthesia. Device implantation was successful in 99 patients. One patient died on day 1 during surgical pericardiocentesis for treatment of periprocedural pericardial tamponade caused by right ventricular perforation due to the temporary pacemaker wire. Periprocedural major adverse cardiovascular and cerebral events rate was 2%, including death (1%) and stroke (0%). The 30-day all-cause mortality rate was 7%.
Procedural hemodynamic valve performance
The hemodynamic results are presented in Table 7. The TAVI procedure resulted in a decrease in mean pressure gradient from 41 ± 15 mm Hg to 9 ± 4 mm Hg and a decrease of the peak pressure gradient from 70 ± 22 mm Hg to 17 ± 8 mm Hg.
The AR increased from 1.1 ± 0.7 before procedure (AR 0) to 1.3 ± 0.9 immediately after deployment of the CoreValve prosthesis (PAR0) in the overall population. In 34 patients (34%), additional maneuvers after device deployment were necessary due to more-than-moderate ARs. All of these patients underwent post-dilation. In 4 of these patients a repositioning maneuver with a snare catheter was attempted, due to an additional so-called “deep position,” which caused suboptimal results after post-dilation. This snare maneuver improved the AR in 1 patient; the remaining 3 patients underwent a second implantation of a CoreValve prosthesis (valve-in-valve replacement or 2 valves “in series”), which improved the regurgitation grade in all cases (Table 6).
Final regurgitation grade after correction measures was 0.9 ± 0.8 (PAR1) in the overall population. This result remained constant after a 2-week period, with a PAR of 0.9 ± 0.6 (PAR2), taking into consideration that acute PAR after termination of the procedure (PAR1) was determined by invasive aortography whereas PAR2 was determined by transthoracic echocardiography. None of these changes were statistically significant (Table 7, Fig. 3). However, in patients with the need for post-dilation, grade of PAR was significantly decreased from PAR0 of 2.2 ± 0.5 to PAR1 of 1.3 ± 0.3 (p < 0.001).
CT data and correlation analysis
The data derived from cardiac CT are shown in Table 3. The amount of calcification, measured semiquantitatively by the (DLZ-CS) was 2.25 ± 0.09. The amount of calcification, measured quantitatively by the AgS was 3,355 ± 1,773 AU, ranging from 790 to 11,007 AU. Both, AgS and DLZ-CS correlated significantly in a positive fashion (r = 0.858, p < 0.001) (Fig. 4). The DLZ-CS grade 1 (mild calcification) equates to an AgS of ≤1,000 AU, DLZ-CS grade 2 (moderate calcification) equates to an AgS of 1,001 to 3,000 AU, DLZ-CS grade 3 (heavy calcification) equates to an AgS 3,001 to 5,000 AU, and DLZ-CS grade 4 (massive calcification with calcification clumps outreaching the annulus level) equates to an AgS >5,000 AU (Table 8).
In 47% of the aortic valves the noncoronary cusp was—regarding the distribution of AVC—the most calcified, followed by the left-coronary (7%) and right-coronary cusp (5%). Forty-one percent showed a symmetric valve calcification without cusp accentuation.
The amount of calcification in the DLZ, measured by the conventional AgS, showed a weak although statistically significant positive correlation with the grade of PAR0 (r = 0.329, p = 0.001) (Fig. 5), PAR1 (r = 0.254, p = 0.011) (Fig. 6), and PAR2 (r = 0.341, p = 0.001). The amount of calcification in the DLZ, visually assessed by the DLZ-CS, also showed a weak statistically significant positive correlation with the grade of PAR0 (r = 0.356, p < 0.001), PAR1 (r = 0.240, p = 0.016), and PAR2 (r = 0.300, p = 0.002).
Furthermore, these 2 scoring systems (AgS and DLZ-CS) showed a weak but significant correlation with the need for post-dilation of the CoreValve prosthesis after initial device release (r = 0.297, p = 0.003 for AgS; r = 0.300, p = 0.002 for DLZ-CS).
However, at least in our population, there was no outcome correlation with the distribution pattern of valve calcifications (Fig. 7).
Transcatheter aortic valve implantation is an established procedure in elderly, high-risk patients with significant AS (1–8).
Data currently available for TAVI suggest that this method is feasible and provides hemodynamic and, more importantly, clinical improvement for up to 2 years in high-risk patients suffering from severe AS (6).
Development of post-procedural PAR is one of the potential problems associated with TAVI. In theory, incomplete expansion of the CoreValve nitinol frame, mismatch of valve annulus and prosthesis diameter sizes, or suboptimal device positioning—too low or too high compared with the annulus level—can result in AR after device deployment. Previous studies have shown that the incidence of a more-than-moderate PAR after CoreValve implantation is approximately 5% to 10% (2,3,6). Approximately one-third of patients undergoing TAVI experience a worsening of their pre-procedural grade of AR (8). Incomplete device expansion due to calcifications is believed to be one of the major contributing factors, but up to now, there have been no scientific data available supporting this theory.
We therefore conducted this study, comparing calcifications in the DLZ assessed by MSCT and post-procedural incidence of paravalvular leakages. The ability to quantify valve calcification by cardiac MSCT has been shown previously and has been validated by histomorphometric and pathological analysis (15–19). The AgS (14) is known as a valid measure of objective noninvasive definition of in vivo calcification. Even more, an increased progression rate of AS has been observed in patients with aggravated AVC (20), and the degree of AVC is an independent predictor for adverse clinical outcomes (9,21,22). The pre-surgical knowledge of the extent of AVC is of importance to the cardiac surgeon, because the surgical aortic valve replacement outcome is influenced by this parameter (23–27).
Therefore, MSCT in patients with AS already plays an important role. Early developed calcification scores from the pre-TAVI era focused on the aortic valve as the interesting area in surgical valve replacement in patients with AS (9,10). Furthermore, the semiquantitative AVC 4-step scoring system developed by Rosenhek et al. (9) contains grades with a lack of calcification (i.e., “grade 1” [no calcification]). More recently available studies, however, suggest that, in patients with hemodynamically increased transvalvular aortic gradients due to aortic valve stenosis, AVC is constantly present (15,16), which was affirmed by Willmann et al. (10)—who found no patients with “grade 1” AVC with this early calcification score in patients with severe AS. These findings are in line with our data. The semiquantitative AVC score developed by Rosenhek et al. (10) has been recently transferred to the collective of high-risk patients designated for TAVI (11,12) without consideration of the lower part of the CoreValve DLZ (i.e., the LVOT). For the collective of patients undergoing TAVI we turned our attention to the aortic valve and the LVOT, which are both part of the CoreValve DLZ because, in patients with severe AS, calcification is often not limited to the aortic valve level. Our data demonstrate that the amount of calcification of the aortic valve and the LVOT shows a significant positive correlation to PAR after TAVI. Therefore, calcification in the DLZ impairs the short-term device outcome of the CoreValve prosthesis in patients with severe AS due to incomplete apposition of the TAVI prosthesis within the native annulus and the wall of the LVOT. After CoreValve implantation, the calcium of the native valve and the LVOT is sandwiched between the nitinol frame of the prosthesis and the aortic wall, respectively. This factor induces gaps that in turn cause several diastolic paravalvular regurgitation jets that add up to a noticeable regurgitation.
The cusp accentuation of these calcifications is more than 50% asymmetrical—as shown in our study. In nearly 50% of our collective, the noncoronary cusp was the most calcified one. In our patients, the distribution of calcification of the aortic valve cusps did not influence the post-procedural degree of PAR. Whether this will prove true in a larger cohort of patients remains to be studied.
Balloon post-dilation in cases with more-than-moderate PAR is a safe and often practiced maneuver, as shown in our study, taking place in approximately one-third of cases. In none of the 34 cases in which this maneuver was performed did a patient-related (myocardial infarction, stroke) or device-related (rupture, valve dysfunction, or migration) complication occur. With this technique, the post-interventional result concerning PAR can be remarkably improved as demonstrated.
To quantify and qualify calcifications in the DLZ in patients undergoing TAVI, no adequate calcification scoring system has been validated. We developed a semi-quantitative 4-step calcification scoring system (DLZ-CS) for symptomatic patients with hemodynamically significant AS that provides a semi-quantitative scale with consideration of the calcification pattern. Both calcification scores (DLZ-CS and AgS) showed a significant positive correlation in our analysis, demonstrating that these scores are applicable.
We found that a DLZ-CS ≥3 and/or an AgS >3,000 AU predicts a relevant PAR (PAR ≥2+) after initial release of the CoreValve prosthesis (PAR0) and, moreover, the need for “second maneuvers” (i.e., post-dilation after initial release of the CoreValve prosthesis).
Consecutively, before TAVI, a pre-interventional cardiac CT imaging including the determination of calcification in the DLZ is recommended to specify the likely interventional outcome.
For the estimation of the AgS, a native MSCT is requested. The DLZ-CS has the advantage that the contrast-enhanced CT, which is performed before intervention as part of the regular TAVI screening process, suffices as assessment of the amount and distribution of calcification in the DLZ.
The evaluation of AR at the 4 time points was performed with different methods. The pre-interventional AR (AR 0) and PAR after device release (PAR0) and after termination of the procedure (PAR1) were determined by invasive aortography, whereas the PAR 2 weeks afterward (PAR2) was determined by transthoracic echocardiography. Therefore, the comparison of the acute post-procedural PAR (PAR1) and short-term PAR (PAR2) is limited.
It must be taken into consideration that the cause of acute post-procedural AR can also be due to other correctable factors: 1) the prosthesis size does not match the anatomical needs; this problem can be overcome by meticulous pre-procedural screening by contrast-enhanced cardiac MSCT with multiplanar reconstruction demonstrated in Figure 1; and 2) the prosthesis sits too deep in the LVOT, the so-called “deep position”; careful pulling and re-positioning by means of a snare catheter, either through the femoral or the brachial approach, can correct this problem.
However, with an increasing amount of calcification, the occurrence of PAR and therefore the need for post-dilation increases. Further parameters are required to predict the procedural outcome. Furthermore, experienced operators are necessary to control possible complications during the CoreValve procedure.
Study devices have been provided by CoreValve. Drs. Gerckens and Grube are proctors for Medtronic/CoreValve. The first 2 authors contributed equally to this work.
- Abbreviations and Acronyms
- Agatston score
- aortic regurgitation
- aortic valve stenosis
- aortic valve calcification
- computed tomography
- device landing zone
- device landing zone calcification score
- left ventricular outflow tract
- multislice computed tomography
- paravalvular aortic regurgitation
- angiographic aortic regurgitation assessment immediately after device deployment
- angiographic aortic regurgitation assessment at the end of the procedure
- transthoracic echocardiography evaluation of the regurgitation grade 2 weeks after the procedure
- transcatheter aortic valve implantation
- Received August 26, 2009.
- Revision received November 2, 2009.
- Accepted November 5, 2009.
- American College of Cardiology Foundation
- Cribier A.,
- Eltchaninoff H.,
- Bash A.,
- et al.
- Grube E.,
- Laborde J.C.,
- Gerckens U.,
- et al.
- Grube E.,
- Schuler G.,
- Buellesfeld L.,
- et al.
- Webb J.G.,
- Pasupati S.,
- Humphries K.,
- et al.
- Webb J.G.
- Vahanian A.,
- Alfieri O.,
- Al-Attar N.,
- et al.,
- European Association of Cardio-Thoracic Surgery; European Society of Cardiology; European Association of Percutaneous Cardiovascular Interventions
- Gerckens U.,
- Buellesfeld L.,
- Grube E.
- Tops L.F.,
- Wood D.A.,
- Delgado V.,
- et al.
- Baim D.,
- Grossman W.
- Agatston A.S.,
- Janowitz W.R.,
- Hildner F.J.,
- et al.
- Messika-Zeitoun D.,
- Aubry M.C.,
- Detaint D.,
- et al.
- Woodring J.H.,
- West J.W.
- Davies S.W.,
- Gershlick A.H.,
- Balcon R.