Aims New-onset left bundle branch block (LBBB) and complete atrioventricular block (AV3B) frequently occur following transcatheter aortic valve implantation (TAVI). We sought to determine the timing and potential mechanisms of new conduction abnormalities (CAs) during TAVI, using the Medtronic CoreValve System (MCS).
Methods and results Sixty-five consecutive patients underwent TAVI with continuous 12-lead ECG analysis. New CAs were defined by the occurrence of LBBB, RBBB, and/or AV3B after the following pre-defined time points: (i) crossing of valve with stiff wire, (ii) positioning of balloon catheter in the aortic annulus, (iii) balloon valvuloplasty, (iv) positioning of MCS in the left ventricular outflow tract (LVOT), (v) expansion of MCS, (vi) removal of all catheters. A new CA occurred during TAVI in 48 patients (74%) and after TAVI in 5 (8%). Of the 48 patients with procedural CAs, a single new CA occurred in 43 patients (90%) and two types of CAs in 5 (10%). A new LBBB was seen in 40 patients (83%), AV3B in 9 (19%), and RBBB in 4 (8%). The new CA first occurred—in descending order of frequency—after balloon valvuloplasty in 22 patients (46%), MCS expansion in 14 (29%), MCS positioning in 6 (12%), positioning of balloon catheter in 3 (6%), wire-crossing of aortic valve in 2 (4%), and after catheter removal in 1 patient (2%). Patients who developed a new CA during balloon valvuloplasty had a significantly higher balloon/annulus ratio than those who did not (1.10 ± 0.10 vs. 1.03 ± 0.11, P = 0.030). No such relationship was found with the valve/annulus ratio.
Conclusion Transcatheter aortic valve implantation with the MCS was associated with new CAs in 82% of which more than half occurred before the actual valve implantation. It remains to be elucidated by dedicated studies whether new CAs can be reduced by appropriate balloon sizing—a precept that also holds for valve size given the observed directional signal of the valve size/aortic annulus ratio.
Transcatheter aortic valve implantation
New-onset left bundle branch block (LBBB), third-degree atrioventricular block (AV3B), and the need for new permanent pacemaker implantation (PPI) constitute an important clinical problem during transcatheter aortic valve implantation (TAVI). This is in particular true after the implantation of the self-expanding Medtronic CoreValve System (MCS). Following the latter, new LBBB, AV3B, and PPI have been reported to vary between 29 and 65%, 15 and 44%, and 9 and 49%, respectively1–5 and to vary between 6 and 18%, 0 and 27%, and 0 and 27%, respectively, after the implantation of the EDWARDS Sapien valve.6–9
The pathophysiology of new conduction abnormalities (CAs) has not yet been elucidated. A number of studies indicate that both patient- and procedure-related factors such as septal wall thickness, non-coronary cusp thickness, pre-existing right bundle branch block (RBBB), depth of valve implantation within the left ventricular outflow tract (LVOT), post-implant prosthesis expansion, and the type of prosthesis play a role.1–4,8,10–12
Transcatheter aortic valve implantation constitutes a complex and multi-step procedure including crossing of the aortic valve and exchange and manipulation of various guide wires and bulky catheter systems in the LVOT, which may inflict temporary or permanent injury to the conduction system. Hence, procedure-related causes of CAs during TAVI may not necessarily relate to the prosthesis itself but to many other actions inherently associated with TAVI. Therefore, we sought to examine the timing of the occurrence of new CAs in a series of 65 consecutive patients who underwent TAVI with the MCS during six pre-defined time points of the procedure while using continuous ECG analysis and sought to explore potential mechanisms of new CAs. In particular, the relationship between new CAs and the balloon and valve/annulus ratio in addition to markers of inflammation was studied. The latter stems from propositions that the implantation of a bioprosthesis may induce an inflammatory reaction due to trauma inflicted on the LVOT.2,4,8,11,13,14
The study population consisted of 65 consecutive patients with severe symptomatic aortic stenosis who underwent TAVI with the MCS between March 2009 and August 2010. Details of the prosthesis and procedure have been previously published.5 Briefly, all patients were accepted for TAVI by Heart Team consensus between a cardiologist and a cardiac surgeon who agreed that conventional open-heart surgery was associated with either too high or prohibitive risk. The prosthesis consists of a self-expanding nitinol tri-level frame to which is secured a trileaflet bioprosthetic porcine pericardial tissue valve. Currently, the prosthesis is available in sizes of 26 and 29 mm. In case a 26 mm MCS was chosen, pre-dilatation of the aortic valve was performed with a 22 mm nucleus balloon (NuMed, Hopkington, NY, USA). In case of a 29 mm MCS, a 23 mm Z-Med-II balloon was used (NuMed). The procedure was performed with the patient under general anaesthesia, with a temporary pacemaker wire positioned in the right ventricle and with default femoral arterial access through an 18F sheath. Patients were extubated before leaving the catheterization laboratory or within 2 h after arrival in the cardiac care unit. Per TAVI protocol, the temporary pacemaker was maintained for at least 48 h after the procedure or longer if indicated. This study complies with the Declaration of Helsinki.
Patient demographics and procedural and post-procedural data were prospectively collected and entered in a dedicated database. Endpoints regarding in-hospital outcome were selected and defined according to the Valve Academic Research Consortium (VARC) recommendations, including the 30-day safety endpoint, defined as composite all-cause death, major stroke, major vascular complication, life-threatening bleeding, acute kidney injury—stage 3, peri-procedural myocardial infarction, repeat procedure for valve-related dysfunction.15
All 12-lead surface ECGs immediately before and after the procedure and at discharge were analysed by two senior cardiologists who are not involved in the TAVI procedure and who were blinded to the results of the continuous rhythm analysis during the procedure. These surface ECGs were used to record the heart rate and rhythm, PR interval, and the presence of first-, second-, or third-degree AV block. Left and right fascicular hemiblocks and left and right bundle branch blocks were defined according to the guidelines of World Health Organization and International Society and Federation for Cardiology Task Force.16
During TAVI, an electronic 12-lead ECG was continuously recorded and digitally collected in the catheterization laboratory database for invasive cardiac procedures. These strips were analysed by two independent researchers (postgraduate research fellows, interventional cardiology) for the assessment of new CAs after the following six pre-defined phases of TAVI. Phase 1: crossing of the stenotic valve with a straight wire and exchange for a stiff support wire; phase 2: positioning of a balloon catheter (typical size 22 or 23 mm × 4 cm) within the aortic annulus used for pre-dilatation; phase 3: full inflation of the balloon catheter under rapid ventricular pacing at a rate of 180 or 220b.p.m.; phase 4: positioning of the MCS delivery catheter into the LVOT with the ventricular edge of the frame approximately within 6–8 mm of the lower edge of the non-coronary cusp as identified by contrast aortography; phase 5: complete expansion of the MCS prosthesis; phase 6: retrieval of all catheters and wires.
For this study, the following new CAs were collected during the procedure: LBBB, RBBB, and AV3B. For confirmation purposes, all electronic rhythm strips were printed after each individual phase. New CAs were considered (i) persistent if present during all subsequent phases of the procedure; (ii) intermittent in case of spontaneous appearance and disappearance during the procedure; and (iii) permanent if still present on the ECG at hospital discharge.
To explore the mechanisms of new CAs, a univariate analysis was performed assessing the relationship between the balloon/aortic annulus ratio and new CAs during phase 3 (balloon valvuloplasty) and the valve size/aortic annulus ratio and new CAs during phase 5 (valve expansion). Also, the relationship was studied between markers of inflammation [C-reactive protein and white blood cell count (WBC) at 24 and 72 h after TAVI] and new CAs. The balloon and valve sizes were defined by the nominal size provided by the manufacturer. The aortic annulus was defined and quantified using multi-sliced computed tomography according to the protocol previously described.17 The mean of the minimum and maximum diameter in, respectively, the sagital and coronoral view was used to define the diameter of the aortic annulus.17
Categorical variables are presented as frequencies and percentages, and normal and skewed continuous variables are presented as means (±SD) and medians (IQR), respectively. The normality distribution for continuous data was examined with the Shapiro–Wilk test. Comparison of categorical variables was performed using the two-sided Student's t-test or Wilcoxon rank-sum test, and the χ2 or Fischer's exact tests were used to compare categorical variables, with a two-sided P< 0.05 indicating statistical significance. All analyses were performed with the SPSS software (version 17).
A total of 65 consecutive patients underwent TAVI with the MCS (transfemoral 64, subclavian 1) of which the baseline characteristics and in-hospital clinical results are listed in Tables 1 and 2, respectively. The 30-day event rate was 17% both in patients with (n = 9) and without (n = 2) a new CA (P = 1.0). The in-hospital or 30-day mortality, however, was 11% in patients with a new CA and 0% in those without a new CA (P = 0.35). Two patients died during TAVI (electromechanical dissociation during phase 1 in one patient and LVOT rupture after phase 3 in another), and four deaths occurred during hospital stay [severe paravalvular aortic regurgitation (AR) at day 14 in one patient, pneumonia at day 28 in two patients, and pneumothorax following PPI at day 32 in another patient]. In these four patients, the ECG just before in-hospital death was used to determine the persistence of the CAs eventually seen during TAVI.
In-hospital peri-procedural complications, therapy-specific and echocardiographic results in patients undergoing transcatheter aortic valve implantation (n = 65)
Mortality (30-day or in-hospital), n (%)
Myocardial infarction, n (%)
Peri-procedural (<72 h)
Spontaneous (>72 h)
Cerebrovascular, n (%)
Transient ischaemic attack
Vascular, n (%)
Bleeding, n (%)
Life-threatening or disabling
Life-threatening or disabling
Acute kidney injury, n (%)b
Combined safety endpoint (at 30 days), n (%)c
Valve-in-valve implantation, n (%)
Post-implantation balloon dilatation, n (%)
Unplanned cardiopulmonary bypass use, n (%)
In-hospital re-intervention, n (%)
Aortic valve area (cm2), mean ± SD
1.8 ± 0.8
Left ventricular ejection fraction ≤35%, n (%)
Aortic regurgitation grade ≥III, n (%)
Mitral regurgitation grade ≥III, n (%)
Mutually non-exclusive analysis (one or more events/patient possible).
aIncluding two intraprocedural deaths.
bFour patients with pre-procedural haemodialysis and two patients who died during TAVI were excluded from the analysis of acute kidney injury.
cComposite all-cause mortality, major stroke, major vascular complication, life-threatening bleeding, acute kidney injury—stage III, peri-procedural, myocardial infarction, repeat procedure for valve-related dysfunction.
Details of the type and timing of new CAs are listed in Supplement A. Of the 65 patients, 12 patients (18%) had a pre-existing CA. In 3 out of these 12 patients, the pre-existing LBBB/RBBB progressed to AV3B during TAVI. In another 45 patients, a new CA was seen during TAVI. In five other patients, a new CA occurred after TAVI (as identified on ECG at discharge) but not during the procedure. In all five patients, the new CA consisted of an LBBB except in one who had a pre-existing LBBB and developed an AV3B after the procedure. Therefore, a total of 53 patients (82%) had new peri-procedural CAs: during TAVI in 48 patients (74%) and after TAVI in another 5 patients (8%). Details are summarized in Table 3.
Summary of 53 patients with new conduction abnormalities during and after transcatheter aortic valve implantation
Type of CAs
During TAVI, n (%)
After TAVI, n (%)
AV3B, third-degree atrioventricular block; CAs, conduction abnormalities; LBBB, left bundle branch block; RBBB, right bundle branch block.
aNew LBBB during TAVI changed to AV3B after TAVI in two patients.
bNew RBBB during TAVI changed to LBBB after TAVI in one patient.
In the 48 patients with a new CA during TAVI, a single new CA was seen in 43 (90%) and two types of CAs in 5 (10%). A new LBBB was seen the most (40 patients or 83%), followed by AV3B in 9 (19%) and RBBB in 4 patients (8%). In three patients, the new CAs that occurred during TAVI changed from RBBB to LBBB at discharge in one patient (No. 5) and progressed from LBBB to AV3B in two patients (Nos 16 and 39).
In these 48 patients, the new CAs first occurred—in descending order of frequency—during phase 3 (balloon valvuloplasty) in 22 patients (46%), phase 5 (complete MCS expansion) in 14 patients (29%), phase 4 (positioning of MCS in the LVOT) in 6 patients (12%), phase 2 (positioning of balloon catheter in the LVOT) in 3 patients (6%), phase 1 (crossing of aortic valve with wire) in 2 patients (4%), and phase 6 (removal of catheters from the body—most likely caused by the touching of the cone of the LVOT when removing the delivery catheter out of the left ventricle) in 1 patient (2%) (Figure 1). Hence, 56% of the new CAs occurred during the preparatory phases (phases 1–3) and 44% during and after valve delivery and implantation (phases 4–6).
Distribution of first occurrence of new CAs (LBBB, RBBB, and AV3B) per phase of TAVI and associations with permanent change as identified on discharge ECG among a total of 48 patients with new CAs during TAVI. The frequencies of first appearance of new CAs (=LBBB, RBBB, and/or AV3B) on the continuous ECG analysis are presented per phase of the TAVI procedure as well as the association with a permanent change as identified on the discharge ECG among 48 patients who developed a new CA during TAVI. AV3B, third-degree atrioventricular block; CAs, conduction abnormalities; LBBB, left bundle branch block; RBBB, right bundle branch block; TAVI, transcatheter aortic valve implantation.
In 70% of the patients in whom the new CA first occurred before the actual valve implantation (phases 1–3), the CA was still present on the discharge ECG. It was 62% in the patients in whom the new CA first occurred during the actual valve implantation (phases 4–6). Overall, the new CAs were intermittent in 12 (25%) and persistent in 36 patients (75%) out of the total of 48 patients in whom a new CA was observed during TAVI. In 31 (65%) out of these 48 patients, the new CA was permanent (still present on the ECG at discharge).
In 14 out of the 65 patients (22%), a new permanent pacemaker after TAVI was implanted because of new-onset AV3B in 10 patients, persisting bradycardia in 3, and brachy-tachy-syndrome in 1 patient (Supplement B). Among those with AV3B, the diagnosis was made during the procedure in seven patients and after the procedure in three patients (two at day 2 and one at day 5).
Table 4 summarizes potential determinants of new CAs during balloon valvuloplasty (phase 3) and during valve implantation (phase 5). Patients who developed a new CA during balloon valvuloplasty had a significantly higher balloon/annulus ratio than those who did not (1.10 ± 0.10 vs. 1.03 ± 0.11, P = 0.030). No such relationship was found with the valve size/annulus ratio. Patients who developed new CAs during valve expansion (phase 5) had a higher WBC at 24 and 72 h after TAVI than those who did not develop a new CA.
Technical and inflammatory associations with new conduction abnormality occurrences during phases 3 and 5 in patients undergoing transcatheter aortic valve implantation
Phase 3 new CAs
Phase 3 no CAs
Phase 5 new CAs
Phase 5 no CAs
Balloon size—minimal annulus diameter ratio, mean ± SD
1.10 ± 0.10
1.03 ± 0.11
1.06 ± 0.12
1.06 ± 0.11
Balloon size—maximal annulus diameter ratio, mean ± SD
0.85 ± 0.07
0.84 ± 0.08
0.84 ± 0.07
0.85 ± 0.08
Valve size—minimal annulus diameter ratio, mean ± SD
1.36 ± 0.11
1.30 ± 0.12
1.30 ± 0.17
1.33 ± 0.10
Valve size—maximal annulus diameter ratio, mean ± SD
1.04 ± 0.07
1.06 ± 0.07
1.04 ± 0.09
1.05 ± 0.07
Depth of implantation from non-coronary cusp, mean ± SD
9.01 ± 3.64
8.01 ± 3.15
7.76 ± 3.05
8.66 ± 3.47
Depth of implantation from left coronary cusp, mean ± SD
9.68 ± 4.06
8.50 ± 3.51
8.30 ± 3.87
9.22 ± 3.73
Leucocyte count <24 h (× 109/L), mean ± SD
11.25 ± 3.48
11.43 ± 4.24
14.07 ± 5.24
10.39 ± 2.80
Leucocyte count <72 h (× 109/L), mean ± SD
12.71 ± 4.42
11.78 ± 4.18
14.97 ± 5.26
11.09 ± 3.32
C-reactive protein <24 h, mean ± SD
64 ± 90
64 ± 55
71 ± 64
62 ± 73
C-reactive protein <72 h, mean ± SD
84 ± 113
75 ± 61
85 ± 66
76 ± 92
CAs, conduction abnormalities.
In this study in which 65 consecutive patients underwent TAVI using the MCS, we found that peri-procedural new CAs occurred in 82% of the patients. The majority of these new CAs occurred during the procedure (91%) of which 56% occurred before the actual valve implantation and most often consisted of a new LBBB (83%). A higher balloon/annulus ratio was associated with a new CA during balloon valvuloplasty. We did not find a relationship between the valve size/annulus ratio and new CAs.
The close anatomical relationship between the aortic valvar complex and the conduction tissue explains the high frequency of new CAs during TAVI with the MCS.18 The herein reported incidence of new CAs is in accordance with the observations made by others with both the MCS and the EDWARDS valve although that the incidence of new LBBB and AV3B is higher after the self expanding MCS (29–65% and 15–44%, respectively) than after the balloon expandable EDWARDS valve (6–18% and 0–27%, respectively).1–4,7–9 Moreover, transapical aortic valve implantation may be associated with few CAs and new PPI most likely as a result of less manipulations and trauma to the LVOT during the procedure. The rate of AV3B and new PPI following transapical TAVI are both reported to vary between 0 and 20%.2,8,19
Of note, we found that a new CA may occur not only during but also at some time after the procedure, which was the case in five patients in our study who were free of new CAs during the procedure. In all patients, it concerned a new LBBB except one in whom a pre-existing LBBB progressed to a complete heart block. In addition, a progression of procedural new CAs to complete heart block after TAVI was seen in three other patients. Whether the late new CAs are caused by injury or oedema of the conduction tissue by the continuous radial expansive force of the self-expanding nitinol frame of the MCS needs to be elucidated. This clinical observation underscores the importance of careful monitoring of patients who undergo TAVI by means of continuous telemonitoring similar to the surgical practice.
More than half of the new CAs in our series occurred before the actual valve implantation. A minority of previous studies reported new CAs following balloon valvuloplasty prior to the valve implantation, which may be explained by the fact that in these studies no continuous ECG recordings were used to determine the occurrence of CAs during the procedure.1–4,7,8,10–12,14,20–22 Our findings are, moreover, in accordance with the incidence of new CAs reported after isolated aortic balloon valvuloplasty.23–25
In terms of mechanisms of new CAs, Bleiziffer et al.12 recently reported an association between balloon size and the occurrence of new-onset AV3B requiring PPI after TAVI. In the present study, we found a significantly higher balloon/annulus ratio in patients who developed a new CA during balloon valvuloplasty in comparison with those who did not (1.10 ± 0.10 vs. 1.03 ± 0.11, P = 0.030).
Given the preponderance of new CAs during balloon valvuloplasty and its relationship with the balloon/annulus ratio, the findings of this study suggest that new CAs (and potentially new PPI) may be reduced by using a balloon/annulus ratio close to 1.0. This is independent of the valve technology itself and the access to the aortic valve (transfemoral, transapical, subclavian, direct access via the ascending aorta) since pre-dilatation of the stenotic aortic valve is a standard step in all procedures. Yet, the observational nature of this study does not allow to draw firm conclusions. This needs to be demonstrated by appropriately designed studies in which one should also acknowledge that differences in the physical properties of the frame between a self-expanding and a balloon expandable prosthesis (i.e. continuous radial force vs. plastic deformation without continuous radial force) and the technique of implantation in addition to shape and height of the frame may result in a difference in the incidence of new CAs during the actual valve implantation, which in turn may explain a disparity in the overall incidence of new CAs during TAVI between these two technologies.
We acknowledge that the overlap in balloon/aortic annulus between the two groups in this series is considerable. Therefore, the proposal of balloon sizing needs to be examined in larger series allowing a more precise cutoff value and needs to be validated in prospective clinical research projects. One should also bear in mind that the use of smaller balloons may result in suboptimal pre-dilatation of the native valve, leading to a higher incidence of paravalvular AR after TAVI which in turn may induce CAs due to increased wall tension and stretch of the conduction tissue.26,27
At variance with Gutiérrez et al.,8 who studied 33 patients who underwent transapical TAVI, we found no relationship between the valve size/aortic annulus ratio and new CAs. Yet, the data of this study nevertheless indicate a higher risk of new CAs in case of a higher ratio. We most likely would have found such a relationship in case of a more disperse distribution of the data, thereby allowing a proposal of sizing. The present data indicate, however, that a ratio of approximately 1.30 (when using the minimal annulus dimension) and a ratio of approximately 1.05 (when using the maximum annulus dimension) are safe and may be recommended to avoid new CAs. Similar to the proposal of balloon size selection, proposals of valve size selection need to be confirmed by more in-depth analysis in larger cohorts of patients allowing multivariate analysis and need subsequently to be validated in prospective research projects. At present, only two sizes of valves are available. The issue will be even more pertinent when four sizes become available.
We also found that the new CA occurrence during valve implantation (phase 5) was associated with increased levels of leucocyte count after TAVI (14.07 vs. 10.39 × 109/L, P = 0.001). It is unclear whether this concerns a causal relationship (e.g. more trauma and/or oedema of the conduction tissue during TAVI) or whether the increased leucocyte count is caused by post-TAVI conditions (e.g. more frequent pacing). In case of the former, all measures should be taken to limit injury and, thus, inflammation. In this respect, more direct access to the aortic valve that is achieved by transapical, subclavian, and direct access of the ascending aorta may play a role as they may be associated with less contact and injury of the LVOT.28–30 The information currently available on PPI rates after transfemoral and transapical implantation of the EDWARDS valve, however, does not reveal a difference. It varies between 2–27% and 0–20%, respectively.8,9,19,31 Also, better control of the positioning and release of the valve may help to reduce injury to the tissue of the LVOT during the procedure. This may be achieved by software allowing online definition of annulus and base of frame during implantation and/or by novel delivery systems with improved ergonomics and enhanced control of catheter stability during release and the eventual retrieval of the valve.32
Although it concerns a prospective study in which two independent researchers continuously monitored the electrocardiographic recordings during the procedure, some electrocardiographic changes may have remained undetected, leading to an underestimation of the reported frequency of new CAs during TAVI. In addition, post-procedural onset of CAs as identified on continuous telemetry recordings was less intensively monitored and was most likely only detected in the case of more evident CAs. Also, the duration of analysis was limited to the hospital stay, and, therefore, the occurrence of late new CAs as well as late disappearance of TAVI-induced CAs remains uncertain although they are unlikely to occur.4 Considering the observational nature of the current study, further research is needed to elucidate whether the association between balloon/annulus ratio and new CAs represents a causal relationship and if modification of the sizing will reduce the frequency of new CAs. In addition, the study lacks the power to provide a comprehensive analysis of the mechanisms or determinants of new CAs. Many potential determinants may have remained undetected.
Transcatheter aortic valve implantation with the MCS was associated with peri-procedural new CAs in 82% of the patients. More than half of these new CAs occurred before the actual valve implantation, and two-thirds of the new CAs were still present on the ECG at discharge. It remains to be elucidated by dedicated studies whether appropriate balloon and valve sizing will reduce new CAs.
. Factors associated with cardiac conduction disorders and permanent pacemaker implantation after percutaneous aortic valve implantation with the CoreValve prosthesis. Am Heart J 2010;159:497-503. doi:10.1016/j.ahj.2009.12.009.
. Procedural and 30-day outcomes following transcatheter aortic valve implantation using the third generation (18 Fr) CoreValve revalving system: results from the multicentre, expanded evaluation registry 1-year following CE mark approval. EuroIntervention 2008;4:242-249. doi:10.4244/EIJV4I2A43.
. Early and persistent intraventricular conduction abnormalities and requirements for pacemaking after percutaneous replacement of the aortic valve. JACC Cardiovasc Interv 2008;1:310-316. doi:10.1016/j.jcin.2008.04.007.
. Criteria for intraventricular conduction disturbances and pre-excitation. World Health Organizational/International Society and Federation for Cardiology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261-1275. doi:10.1016/S0735-1097(85)80335-1.
. Three dimensional evaluation of the aortic annulus using multislice computer tomography: are manufacturer's guidelines for sizing for percutaneous aortic valve replacement helpful? Eur Heart J 2010;31:849-856. doi:10.1093/eurheartj/ehp534.
. Outcomes after transcatheter aortic valve implantation with both Edwards-SAPIEN and CoreValve devices in a single center: the Milan experience. JACC Cardiovasc Interv 2010;3:1110-1121. doi:10.1016/j.jcin.2010.09.012.
. Safety and efficacy of the subclavian approach for transcatheter aortic valve implantation with the CoreValve revalving system. Circ Cardiovasc Interv 2010;3:359-366. doi:10.1161/CIRCINTERVENTIONS.109.930453.
. Six-month results of a repositionable and retrievable pericardial valve for transcatheter aortic valve replacement: the Direct Flow Medical aortic valve. J Thorac Cardiovasc Surg 2010;140:897-903. doi:10.1016/j.jtcvs.2010.01.017.
Rutger-JanNuis, Nicolas M.Van Mieghem, Carl J.Schultz, ApostolosTzikas, Robert M.Van der Boon, Anne-MarieMaugenest, JinCheng, NicoloPiazza, Ron T.van Domburg, Patrick W.Serruys, Peter P.de JaegereEur Heart J(2011)32 (16):
2067-2074DOI: http://dx.doi.org/10.1093/eurheartj/ehr110First published online: 28 May 2011 (8 pages)