Incidence, Predictors, and Outcome of Conduction Disorders AfterTranscatheter Self-Expandable Aortic Valve Implantation

Chiara Fraccaro, MDa,*, Gianfranco Buja, MDa, Giuseppe Tarantini, MD, PhDa,Valeria Gasparetto, MDa, Loira Leoni, MD, PhDa, Renato Razzolini, MDa,

Domenico Corrado, MD, PhDa, Raffaele Bonato, MDb, Cristina Basso, MD, PhDc,Gaetano Thiene, MDc, Gino Gerosa, MDd, Giambattista Isabella, MDa,

Sabino Iliceto, MDa, and Massimo Napodano, MDa

The aims of the present study were to investigate the incidence and characteristics ofconduction disorders (CDs) after transcatheter aortic valve implantation (TAVI), to ana-lyze the predictors of permanent pacemaker (PPM) implantation, and to evaluate theoutcomes of CDs over time. In particular, we sought to investigate whether the depth ofdeployment and other technical aspects of valve implantation might predict the need forPPM implantation after TAVI. TAVI has been reported to favor the onset or worsening ofCDs often requiring PPM implantation. A total of 70 patients with aortic stenosis due todystrophic calcification underwent TAVI with third-generation CoreValve Revalving Sys-tem from May 2007 to April 2009. We collected electrocardiograms at baseline, duringTAVI, during hospitalization and at the 1-, 3-, 6-, and 12-month follow-up visits thereafter.The clinical, anatomic, and procedural variables were tested to identify the predictors ofPPM implantation. The PPM dependency at follow-up was analyzed. Six patients wereexcluded from the analysis because of a pre-existing PPM. Of the 64 patients, 32 (50%) hadone or more atrioventricular-intraventricular CDs at baseline. TAVI induced a worseningin the CDs in 49 (77%) of the 64 patients, with 25 (39%) requiring in-hospital PPMimplantation. On multivariate analysis, the independent predictors of PPM implantationwere the depth of the prosthesis implantation (p � 0.039) and the pre-existing right bundlebranch block (p � 0.046). A trend in the recovery of the CDs over time was recorded,although 2 patients required PPM implantation 1 month after discharge for late completeatrioventricular block. In conclusion, TAVI often induces or worsens CDs, requiring PPMin more than one third of patients, although a trend in the recovery of CDs during themidterm was recorded. The independent predictors of PPM implantation were the depth ofprosthesis implantation and pre-existing right bundle branch block. © 2011 Elsevier Inc.

All rights reserved. (Am J Cardiol 2011;107:747–754)

The aim of the present study was to investigate theincidence and characteristics of conduction disorders (CDs)in patients undergoing transcatheter aortic valve implanta-tion (TAVI) and the need for subsequent permanent pace-maker (PPM) implantation. In addition, to help identify theclinical, anatomic, and procedural predictors of postopera-tive PPM implantation and the outcome of CDs over time.In particular, we sought to investigate whether the depth ofdeployment and other technical aspects of valve implanta-tion might predict the need for PPM implantation afterTAVI.

aDivision of Cardiology and dDivision of Cardiac Surgery, Departmentof Cardiac, Thoracic, and Vascular Sciences, bInstitute of Anesthesia, andcDepartment of Medical Diagnostic Sciences and Special Therapies, Uni-versity of Padova, Padua, Italy. Manuscript received July 28, 2010; man-uscript received and accepted October 19, 2010.

Dr. Fraccaro is a Ph.D. fellow and was supported in part by a Grant ofthe Italian Society of Cardiology.

*Corresponding author: Tel: (�39) 049-821-1844; fax: (�39) 049-876-1764.

E-mail address: chiara_fraccaro@hotmail.com (C. Fraccaro).

0002-9149/11/$ – see front matter © 2011 Elsevier Inc. All rights reserved.doi:10.1016/j.amjcard.2010.10.054

Methods

The data from 70 consecutive patients with aortic valvestenosis undergoing TAVI at our Department at PadovaUniversity from May 2007 to April 2009 were analyzed. Atotal of 6 patients were excluded from the analysis becausethey already had undergone PPM implantation beforeTAVI. All patients were a part of the multicenter, expandedevaluation registry after conformité européenne mark ap-proval.1 The inclusion and exclusion criteria for TAVI havebeen previously reported.1–3

All procedures were performed using the third-genera-tion self-expanding CoreValve Revalving System (Medtronic,Minneapolis, Minnesota), as previously described,1,2 usinga transfemoral or transubclavian approach, according to theanatomy of the iliac and femoral arteries.3 Implantationsuccess was defined as the correct positioning and perfor-mance of the prosthesis. Procedural success was defined asthe success of implantation, with the patient leaving thecatheterization laboratory alive. The good performance ofthe bioprosthesis was defined as a reduction in the mean

transaortic gradient to �20 mm Hg and aortic regurgitation

www.ajconline.org

CCCPPNLPHA

ctive tis

748 The American Journal of Cardiology (www.ajconline.org)

to �2�/4, as evaluated on an aortic angiogram or echocar-diogram.1

The degree of aortic valve calcium was scored accordingto the presence and extent of cusp calcification as it ap-peared on the aortic angiogram. The grading was as follows:grade 1, no calcification; grade 2, mild calcification appear-ing as a thin marginal rim in one or more cusps; grade 3,moderate calcification characterized by a thick rim occupy-ing the entire surface of one or more cusps; and grade 4,severe calcification, defined as the presence of heavy calci-fication of all cusps or bulky calcification.

The depth of bioprosthesis implantation was measured inthe right anterior oblique projection as the distance (inmillimeters) of the aortic prosthesis within the left ventric-ular outflow tract, from the lower edge of the noncoronarycusp (D1) and from the lower edge of the left coronary cusp(D2) to the ventricular end of the prosthesis frame usingquantitative angiographic digital techniques (Allura, Philips

Table 1Baseline characteristics

Variable Total(n � 64)

Age (years)Mean � SD 80.97 � 6.55Range 55–91

Men 29 (45%)Logistic EuroSCORE

Mean � SD 23.64 � 14.72Range 3–71

New York Heart AssociationI 9 (14%)II 19 (30%)III 32 (50%)IV 4 (6%)

Canadian cardiovascular society0 39 (61%)1 2 (3%)2 6 (9%)3 10 (16%)4 7 (11%)

Calcium score�2 17 (27%)3 30 (47%)4 17 (27%)

Coronary artery disease 40 (63%)Congestive heart failure 30 (47%)

erebral vascular accident 7 (11%)hronic kidney disease 35 (55%)hronic obstructive pulmonary disease 14 (22%)eripheral vascular disease 22 (34.4%)revious cardiac surgery 16 (25.0%)eurologic dysfunction 10 (15.6%)iver cirrhosis 5 (7.8%)orcelain aorta* 15 (23.4%)ostile thorax† 9 (14.1%)ortic valve area (cm2) 0.78 � 0.21

Ejection fraction (%) 52.32 � 13.24

Data are presented as mean � SD or n (%), unless otherwise noted.* Defined as an aorta with diffuse, circumferential, plate-like calcifica

cross-clamping.† Included patients with a severe deformity of the thorax, severe conne

Medical System, Best, The Netherlands).4 The difference

between D2 and D1 was calculated as the coaxial index.Prosthesis implantation was defined as coaxial when thecoaxial index ranged from �1.0 mm to �1.0 mm andnoncoaxial when the coaxial index was ��1.0 mm or��1.0 mm. The ratio between the prosthesis nominal di-ameter and native annulus size was calculated as the pros-thesis/annulus ratio. The ratio between the diameter of thedeployed prosthesis measured at the level of the aorticannulus and the native aortic annulus was calculated as theprosthesis expansion index.

All patients underwent standard 12-lead electrocardiog-raphy before the procedure. To assess intraoperative CDs,3-lead continuous electrocardiographic monitoring was re-corded and electronically stored throughout the procedure.After the procedure, continuous monitoring was routinelyperformed in all patients during the hospital stay. Postop-eratively, 12-lead electrocardiography was performed dailyduring hospitalization and at 1-, 3-, 6-, and 12-month fol-

PM After TAVI(n � 25)

No PPM After TAVI(n � 39)

p Value

0.56781.56 � 5.10 80.59 � 7.37

15 (60%) 14 (36%) 0.0580.464

25.47 � 15.70 22.66 � 14.00

0.7192 (8%) 7 (18%)8 (32%) 11 (28%)

13 (52%) 19 (49%)2 (8%) 2 (5%)

0.62917 (68%) 22 (56%)1 (4%) 1 (3%)3 (12%) 3 (8%)2 (8%) 8 (21%)2 (8%) 5 (13%)

0.5958 (32%) 9 (23%)

10 (40%) 20 (51%)17 (28%) 10 (26%)17 (71%) 23 (61%) 0.58613 (52%) 17 (44%) 0.5112 (8%) 5 (13%) 0.547

17 (68%) 18 (46%) 0.0877 (29%) 7 (18%) 0.2988 (32.0%) 14 (35.9%) 0.7499 (36%) 7 (17.9%) 0.1043 (12%) 7 (17.9%) 0.5231 (4%) 4 (10.3%) 0.3633 (12%) 12 (30.8%) 0.0843 (12.0%) 6 (15.4%) 0.704

0.80 � 0.20 0.77 � 0.22 0.54451.68 � 12.95 52.74 � 13.58 0.759

volving the whole proximal ascending aorta, precluding cannulation or

sue disease, and previous mediastinal/thorax radiotherapy.

P

tion in

low-up visits thereafter to detect any modifications in the

PoAtB

PH

TC

P

749Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy

atrioventricular (AV) and intraventricular conduction. Theanalyses of the records were performed by an experiencedelectrophysiologist. The presence of CDs at any time wasdefined by the presence of at least one of the followingabnormalities: first-, second-, or third-degree AV block, leftbundle branch block (BBB), right BBB, and/or left anterioror posterior hemiblock. The currently accepted criteria wereused to code for each of these CD.5,6 The requirement for

PM implantation was determined by the attending cardi-logist according to the standardized criteria from themerican College of Cardiology/American Heart Associa-

ion/Heart Rhythm Society 2008 Guidelines for Device-ased Therapy of Cardiac Rhythm Abnormalities.7 All

systems were implanted using a transvenous subclavian

Table 2Procedural data

Variable Total(n � 64)

Procedural success 61 (95%)General anesthesia 15 (23%)Double-lumen intubation 8 (13%)Transesophageal echocardiography 13 (20%)Access

Transfemoral 60 (94%)Transubclavian 4 (6%)

Prosthesis size (mm)26 36 (56%)29 28 (44%)

Predilation 62 (97%)Postdilation 19 (30%)Valve-in-valve 5 (8%)Prosthesis/annulus diameter ratio 1.21 � 8.89Depth of implantation (mm)

D1 10.25 � 3.39D2 11.41 � 3.27

Coaxial index (mm) 1.15 � 1.59Noncoaxial alignment 44 (69%)Prosthesis expansion index 0.93 � 0.11

rocedural duration (minutes) 72.7 � 37.6ospital stay (days) 12.34 � 8.37

Data are presented as n (%) or mean � SD.

able 3hange in conduction parameters over time

Variable Before TAVI After TAVI Last(6.0

Atrial fibrillation 10/64 (16%) 11/64 (17%) 10/5Left bundle branch block 9/64 (14%) 37/64 (58%) 25/5Right bundle branch block 8/64 (13%) 3/64 (5%) 2/5Anterior hemiblock 11/64 (17%) 3/64 (5%) 4/5Posterior hemiblock 2/64 (3%) 2/64 (3%) 1/5

R interval (ms) 182.9 � 5.7 211.1 � 5.5 182QRS width (ms) 103.6 � 4.1 144.3 � 3.6 125QT (ms) 406.2 � 45.7 424.4 � 47.7 411

Data are presented as n (%) or mean � SD.* Three cases of periprocedural mortality and four with paced rhythm

analysis was not applicable.

approach.

Follow-up data were collected at 1, 3, 6, and 12 monthsand yearly thereafter. Periprocedural death was defined asany death within 30 days after the procedure or any deathbefore discharge. The clinical follow-up events includeddeath from all causes, cardiac death (including all unex-plained deaths), acute myocardial infarction, stroke, cardiacheart failure requiring rehospitalization, and PPM implan-tation. Moreover, at each temporal step, a 12-lead electro-cardiogram was collected in all patients, to record modifi-cations in AV and intraventricular conduction. In patientswith a PPM, the percentage of ventricular pacing was de-tected by PPM interrogation. Moreover, to evaluate thePPM dependency in patients with a paced baseline electro-cardiogram, the pacemaker was programmed to VVI at the

After TAVI(n � 25)

No PPM After TAVI(n � 39)

p Value

23 (92%) 38 (97%) 0.3158 (33%) 7 (18%) 0.3603 (12%) 5 (13%) 0.4196 (27%) 7 (21%) 0.604

22 (88%) 38 (97%) 0.1283 (12%) 1 (3%)

12 (48%) 24 (62%) 0.28713 (52%) 15 (39%)24 (96%) 38 (97%) 0.747

6 (24%) 13 (33%) 0.4253 (12%) 2 (5%) 0.318

.20 � 1.05 1.21 � 7.80 0.561

.34 � 3.62 9.50 � 3.04 0.031

.16 � 3.25 10.89 � 3.23 0.108

.81 � 1.47 1.39 � 1.64 0.15515 (58%) 29 (76%) 0.114.93 � 0.11 0.93 � 0.12 0.727.52 � 28.33 72.74 � 42.79 0.982.27 � 11.06 10.24 � 5.24 0.022

Upo)

p Value

Before VersusAfter TAVI

After TAVI VersusLast Follow-Up

Before TAVI VersusLast Follow-Up

) 0.705 0.655 0.655) �0.0001 0.108 �0.0001

) 0.059 0.317 0.046) 0.005 0.317 0.034) 1.000 0.317 1.000.9 �0.0001 �0.0001 0.761.3 �0.0001 �0.0001 �0.0001.0 0.028 0.039 0.486

t adequate ventricular response during PPM inhibition excluded because

PPM

1

11120

07215

Follow-� 4.2 m

7* (18%7* (44%7* (4%7* (7%7* (2%.7 � 29.3 � 26.5 � 30

withou

lowest rate possible and the underlying rhythm was ob-

ricularl valve

750 The American Journal of Cardiology (www.ajconline.org)

tained. The patients were considered pacemaker dependentif they continued to be paced or had complete AV block oratrial fibrillation with inadequate ventricular response. Thepatients were considered as nonpacemaker dependent ifthey had sinus rhythm or atrial fibrillation with an adequateventricular response.

The categorical data are expressed as numbers and per-centages and compared by Fisher’s or chi-square exact testas appropriate. The continuous variables are expressed asthe mean � SD and compared using Student’s t test. Thepreoperative clinical variables, anatomic characteristics, andprocedural data thought likely to influence the conductingsystem were tested by univariate logistic regression analysisto determine the predictors of postoperative PPM implan-tation. This model included all the variables with a biolog-ically relevant correlation to the onset of CD: age, gender,anatomic characteristics (e.g., aortic valve area, calciumscore, aortic regurgitation, left ventricular mass index), ef-fects of drugs (type of anesthesia), technical aspects that

Figure 1. Views of 85-year old woman who had developed complete AV bl(A) Electrocardiogram showing right BBB and 60 mm Hg transaortic gradafter TAVI. (C) Computed tomography scan of heart explanted at autooverlapping membranous septum (dotted circle) and crest of interventriculNote, expansion of prosthesis frames in subaortic region compressing ventlines). Ao, aorta; LV, left ventricle; MS, membranous septum; MV, mitra

might mechanically effect the conduction system (e.g., val-

vuloplasty balloon diameter, prosthesis size, prosthesis/an-nulus diameter ratio, need for postdilation, depth of implan-tation, prosthesis expansion index, and valve-in-valve), all

Figure 2. Incidence of PPM implantation according to baseline CD. RightBBB was the only baseline CD significantly related to PPM implantation.

uiring PPM implantation after TAVI who died 13 days after the procedure.aseline. (B) Electrocardiogram showing complete AV block immediatelyte, deep positioning of CoreValve within left ventricular outflow tract,m. (D) Gross anatomic view of left ventricular outflow seen from below.septum and overlapping proximal branching of left bundle branch (dotted; RV, right ventricle; VS, ventricular septum.

ock reqient at bpsy. Noar septu

pre-existing CDs (AV block I, left BBB, anterior hemi-

Bh

mC

tbtTtathbcpOoa

lg4lBrtr(TtterCrai

ALAVPPPDPVFLAPRB

751Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy

block, posterior hemiblock, right BBB, bifascicular block).Multiple stepwise logistic regression analyses of those sig-nificant variables (p �0.10) on univariate analysis wereperformed to identify independent predictors of PPM im-plantation. Univariate and multivariate analyses were alsoperformed considering the same variables to identify pre-dictors of worsening in CDs. The odds ratios and theircorresponding 95% confidence intervals are provided. A pvalue of �0.05 with a 2-tailed test was considered statisti-cally significant. Statistical analysis was performed usingthe statistical software Statistical Package for Social Sci-ences, version 17.0, for Windows (SPSS, Chicago, Illinois).

Results

The baseline characteristics were similar between thosepatients who required PPM implantation and those who didnot (Table 1). Implantation success was achieved in 62(97%) of the 64 patients, with procedural success in 61(95%) of 64. The procedural data were similar betweenthose who required PPM implantation and those who didnot, except for the depth of prosthesis implantation mea-sured from the lower edge of the noncoronary cusp (D1),which was significantly deeper (i.e., more ventricular) in thepatients who underwent PPM implantation than in thosewho did not. Also, the hospital stay was longer in patientswho underwent PPM implantation (Table 2).

Of the 64 patients, 32 (50%) had one or more degrees ofAV-intraventricular CDs before TAVI, including first de-gree AV block (n � 15), complete left BBB (n � 9), right

BB (n � 8), anterior hemiblock (n � 11), and posterioremiblock (n � 2; Table 3).

After TAVI, worsening or new-onset CD appeared inost patients (77%). Left BBB was the most frequent new

Table 4Predictors of permanent pacemaker (PPM) implantation after transcathete

Variable Univa

OR 9

Age 0.992 0.919Male gender 3.077 1.092Aortic valve area 2.530 0.226Calcium score 0.836 0.433

ortic regurgitation 0.773 0.431eft ventricular mass index 0.992 0.968nesthesia type 1.265 0.495alvuloplasty balloon diameter 1.312 0.935rosthesis size 0.500 0.181rosthesis/annulus diameter ratio 0.211 0.001ostdilation 0.577 0.186epth of implantation 1.200 1.006rosthesis expansion index 1.371 0.015alve-in-valve 6.727 0.706irst-degree atrioventricular block at baseline 1.968 0.612eft bundle branch block at baseline 0.696 0.157nterior hemiblock at baseline 1.230 0.332osterior hemiblock at baseline 1.480 0.088ight bundle branch block at baseline 5.400 0.995ifascicular block at baseline 1.522 0.200

CI, confidence interval; OR, odds ratio; NS, not significant.

D, appearing in 28 patients (44%), isolated or associated

o new first-degree AV block (Table 3). A complete AVlock appeared in 19 patients (Figure 1). During the hospi-alization, 25 patients (39%) underwent PPM implantation.he indications for PPM implantation were permanent or

ransient complete AV block in 16, second-degree AV blockssociated with left BBB in 6, sick sinus syndrome in 1, andrifascicular block in 1 patient. Indeed, 1 patient who did notave any CD before TAVI underwent PPM implantationecause of the development of first-degree AV block asso-iated with new left BBB after TAVI, with progressiverolongation of PR and QRS intervals during hospital stay.f the 25 patients requiring PPM implantation, 9 had no AVr intraventricular CD before TAVI, and 16 patients showedt least one CD (Figure 2).

On univariate analysis, male gender, right BBB at base-ine, and the depth of prosthesis implantation resulted in areater prevalence of PPM implantation after TAVI (Table). After adjustment by multivariate analysis for the base-ine clinical, anatomic, and operative characteristics, rightBB at baseline, and the depth of prosthesis implantation

emained the only independent predictors of PPM implan-ation (Table 4). After TAVI, 6 (75%) of 8 patients withight BBB at baseline required PPM implantation versus 1934%) of 56 patients, who had not had right BBB beforeAVI (p � 0.026). Right BBB was the only baseline CD

hat significantly affected the occurrence of PPM implanta-ion after TAVI (Figure 2 and Table 4). However, consid-ring as a dependent variable the worsening in the CD,ather than the need for PPM implantation, the predictors ofD worsening remained the depth of implantation (odds

atio 1.30, 95% confidence interval 1.00 to 1.69, p � 0.05)nd pre-existing left BBB (odds ratio 0.07, 95% confidencenterval 0.01 to 0.49, p � 0.007).

valve implantation (TAVI)

alysis Multivariate Analysis

p Value OR 95% CI p Value

0.828 — — NS0.033 1.182 0.982–1.423 0.127

3 0.451 — — NS0.594 — — NS0.388 — — NS0.507 — — NS0.624 — — NS0.116 — — NS0.180 — — NS

9 0.596 — — NS0.341 — — NS0.042 1.210 1.010–1.449 0.039

70 0.892 — — NS9 0.097 2.768 0.226–33.843 0.425

0.256 — — NS0.632 — — NS0.757 — — NS

7 0.785 — — NS7 0.051 6.132 1.030–36.519 0.0469 0.685 — — NS

r aortic

riate An

5% CI

–1.070–8.671–28.31–1.615–1.387–1.016–3.230–1.840–1.379–66.53–1.790–1.431–128.3–64.07–6.335–3.074–4.558–24.77–29.29–11.57

The 30-day mortality rate was 5% (3 of 64), with no

we

rniw

o

752 The American Journal of Cardiology (www.ajconline.org)

difference between those patients who required PPM im-plantation and those who did not (4% vs 5%; p � 0.83). Thecauses of death were left ventricular perforation, pneumo-nia, and cerebral edema owing to electrolyte disturbances.During hospitalization, the rate of stroke and myocardialinfarction was 2% (1 of 64) and 2% (1 of 64), respectively,with no differences between the groups. At a mean fol-low-up of 6.0 � 4.2 months (range 30 days to 13.3months), the mortality rate was 18% (11 of 61); 29% ofpatients who needed PPM versus 11% of patients who didnot (p � 0.07). It was a cardiac death in only 1 case. Of61 patients, myocardial infarction occurred in 1 (2%),stroke in 1 (2%), heart failure in 7 (12%), with nodifferences between the 2 groups (p � 0.39, p � 0.39,and p � 0.31, respectively). During the follow-up, thePR, QRS, and QT intervals decreased significantly withinthe first month after the procedure (Table 3). Two pa-tients, discharged with first-degree AV block plus leftBBB and with left BBB alone, respectively, underwentPPM implantation 1 month after TAVI for late onset ofcomplete AV block. In these patients, the prosthesis/annulus was similar to those of patients with early PPMimplantation or no PPM implantation (1.10 vs 1.20 and1.21, respectively). No sudden death occurred. Analyzingthe patient’s pacemaker dependency at follow-up, of the17 PPM patients who were alive, 12 presented with aspontaneous rhythm with a mean ventricular pacing per-centage of 19% (range 0.2% to 62%; 7 patients �20%).One patient underwent pacing at baseline and had anadequate ventricular response during pacemaker inhibi-tion, with a ventricular pacing percentage of 4%. Finally,4 patients underwent pacing at baseline electrocardiog-raphy and presented with an inadequate ventricular re-sponse at the lowest rate programmable (ventricular pac-

Figure 3. Impact of CoreValve positioning on electrical conduction systemand conduction system. A deeper prosthesis implantation into left ventricune (A).

ing percentage �95% in all cases). o

Discussion

Transcatheter revalving therapy has been reported asvaluable alternative strategy to aortic valve replacement inhigh-risk patients with severe symptomatic aortic valve ste-nosis.1,8,9 However, TAVI has often been associated with

orsening or new-onset CD, particularly when the self-xpandable CoreValve device is used,1,4 as confirmed by

our data showing PPM implantation in more than one thirdof cases. The geometry and design of this device, based ona 53- to 55-mm-long stent cage extending from ascendingaorta to left ventricular outflow, might lead to a variableamount of prosthesis frames, pushing aside the interventric-ular septum and the underlying conduction tissue. In par-ticular, the lower one third of the prosthesis stent frames,characterized by high radial forces for secure anchoring ofthe stent against the native annulus and outflow septum,might account for CDs due to compression on the leftbundle branch, which runs superficially just below the en-docardium in the uppermost part of the leftward ventricularseptum (Figures 1 and 3). In this scenario, it is reasonablethat the deeper the prosthesis has been implanted into theleft ventricular outflow tract, the greater the risk of com-pression by the prosthesis against the left bundle branchand, consequently, of the development of severe CD requir-ing PPM implantation (Figures 1 and 3). These observationsare consistent with the greater rate of PPM implantationreported with CoreValve device4,10,11 than with the Ed-wards Sapien prosthesis (Edwards Lifesciences, Irvine, Cal-ifornia), which extends just a few millimeters below theannular plane.12,13 Piazza et al4 described a significant cor-elation between the depth of prosthesis implantation andew-onset left BBB after CoreValve implantation, suggest-ng that deploying the prosthesis in a more superior positionithin the left ventricular outflow tract might limit the risk

am illustrating anatomic relation between prosthesis across aortic annulusow tract (B) might effect electrical conduction system more than a higher

. Diagrlar outfl

f AV block and PPM implantation. In light of this hypoth-

plipsfiddfadvmtpce

pfp

waPev

cbmPst

dbpisdms

AB

753Valvular Heart Disease/Conduction Defects After Aortic Revalving Therapy

esis, we performed a multivariate analysis to identify inde-pendent predictors of PPM implantation after TAVI. Ourstudy has confirmed this preliminary hypothesis, identifyingthe depth of prosthesis implantation into the left ventricularoutflow tract as an independent predictor of PPM implan-tation after CoreValve revalving therapy. In particular, themost relevant measurement predicting PPM implantationwas the depth of prosthesis at level of noncoronary cusp,and neither the depth at the level of the left coronary cuspnor the coaxial implantation of prosthesis were related toPPM requirement.

Moreover, it is important to emphasize that, as previ-ously reported by Calvi et al,11 the most frequent CD re-corded after TAVI in our series was new-onset left BBB.This CD was likely related to the close anatomic relationbetween the left bundle branch and the aortic valve appa-ratus and might favor the development of complete AVblock when a right BBB is present before TAVI. Consistentwith this assumption, the presence of right BBB beforeprocedure appeared the most powerful predictor of PPMimplantation in our analysis, confirming previous observa-tions.4 A lower rate of PPM was associated with a lowprevalence of right BBB at baseline in other series.14

A greater prevalence of PPM implantation was reportedin our study compared to other series.11 We decided toerform early PPM implantation to minimize the risk re-ated to transvenous temporary pacing support, such asnfection and ventricular perforation, coupled with potentialroblems related to the immobilization and long hospitaltay for an elderly patient. However, our results have con-rmed that a recovery in intraventricular conduction (i.e., aecrease in the frequency of left BBB and PPM depen-ency) can occur over the time. The possible explanationsor this include transient inflammation, edema, ischemia,nd mechanical trauma with subsequent recovery of con-uction. Postmortem specimens from patients who had de-eloped new AV block after TAVI have demonstratedicroscopic evidence of myocardial injury in the interven-

ricular septum, as well as localized hematoma at the site ofrosthesis expansion, which might account for mechanicalompression of the conduction system coursing in the sub-ndocardium close to the membranous septum.15 Balloon

aortic valvuloplasty itself has been associated in the pastwith the occurrence of CD.16 Similarly, the balloon used toredilate the native aortic valve before TAVI could accountor reversible mechanical or ischemic effects, explaining theossible conduction recovery with time after TAVI.

In contrast, in our series, 2 patients who were dischargedith first-degree AV block plus left BBB and with left BBB

lone, respectively, experienced late complete AV block.rogressive degeneration of the conduction system, accel-rated by mechanical injury and fibrosis in the upper inter-entricular septum, might have occurred in these patients.

Some limitations of our study should be taken into ac-ount to place our findings in the proper perspective. First,ecause we decided to perform early PPM implantation toaximize the patient’s safety, the relatively greater rate ofPM resulting from this might have influenced our analy-es. However, even though most CDs recovered over time,

he PPM dependency was still present in most patients

uring follow-up. In addition, to understand the relationetween prosthesis size and aortic annulus, we assessed therosthesis dimensions by angiography at the end of themplantation procedure. However, because the full expan-ion of the nitinol stent frames seems to be reached someays after deployment,17 this might have led to underesti-ating the maximal dimensions of the prosthesis and con-

equently of its effect on annulus stretching.

cknowledgment: The authors are indebted to Mr. Claudioellini for his valuable support in graphic elaborations.

1. Piazza N, Grube E, Gerckens U, den Heijer P, Linke A, Luha O,Ramondo A, Ussia G, Laborde JC, de Jaegere P, Serruys PW. Proce-dural and 30-day outcomes following transcatheter aortic valve im-plantation using the third generation (18 Fr) CoreValve revalvingsystem: results from the multicentre, expanded evaluation registry1-year following CE mark approval. EuroIntervention 2008;4:242–249.

2. Napodano M, Cutolo A, Fraccaro C, Tarantini G, Bonato R, Bianco R,Gerosa G, Iliceto S, Ramondo A. Totally percutaneous valve replace-ment for severe aortic regurgitation in a degenerating bioprosthesis.J Thorac Cardiovasc Surg 2009;138:1027–1028.

3. Fraccaro C, Napodano M, Tarantini G, Gasparetto V, Gerosa G,Bianco R, Bonato R, Pittarello D, Isabella G, Iliceto S, Ramondo A.Expanding the eligibility for transcatheter aortic valve implantationthe trans-subclavian retrograde approach using: the III generationCoreValve revalving system. J Am Coll Cardiol Interv 2009;2:828–833.

4. Piazza N, Onuma Y, Jesserun E, Kint PP, Maugenest AM, AndersonRH, de Jaegere PP, Serruys PW. Early and persistent intraventricularconduction abnormalities and requirements for pacemaking after per-cutaneous replacement of the aortic valve. J Am Coll Cardiol Interv2008;1:310–316.

5. Willems JL, Robles de Medina EO, Bernard R, Coumel P, Fisch C,Krikler D, Mogensen L, Moret P, Mogensen L, Moret P. Criteria forintraventricular conduction disturbances and pre-excitation: WorldHealth Organizational/International Society and Federation for Cardi-ology Task Force Ad Hoc. J Am Coll Cardiol 1985;5:1261–1275.

6. Elizari MV, Acunzo RS, Ferreiro M. Hemiblocks revisited. Circula-tion 2007;115:1154–1163.

7. Writing Committee to Revise the ACC/AHA/NASPE 2002 GuidelineUpdate for Implantation of Cardiac Pacemakers and AntiarrhythmiaDevices. ACC/AHA/HRS 2008 Guidelines for Device-Based Therapyof Cardiac Rhythm Abnormalities: a report of the American College ofCardiology/American Heart Association Task Force on PracticeGuidelines: developed in collaboration with the American Associationfor Thoracic Surgery and Society of Thoracic Surgeons. Circulation2008;117:e350–e408.

8. Cribier A, Eltchaninoff H, Tron C, Bauer F, Agatiello C, Nercolini D,Tapiero S, Bessou JP, Babaliaros V. Treatment of calcific aorticstenosis with the percutaneous heart valve: mid-term follow-up fromthe initial feasibility studies: the French experience. J Am Coll Cardiol2006;47:1214–1223.

9. Grube E, Schuler G, Buellesfeld L, Gerckens U, Linke A, WenaweserP, Walther T, Zickmann B, Felderhoff T, Cartier R, Bonan R. Percu-taneous aortic valve replacement for severe aortic stenosis in high-riskpatients using the second- and current third-generation self-expandingCoreValve prosthesis: device success and 30-day clinical outcome.J Am Coll Cardiol 2007;50:69–76.

10. Jilaihawi H, Chin D, Vasa-Nicotera M, Jeilan M, Spyt T, Ng GA,Bence J, Logtens E, Kovac J. Predictors for permanent pacemakerrequirement after transcatheter aortic valve implantation with the Cor-eValve bioprosthesis. Am Heart J 2009;157:860–866.

11. Calvi V, Puzzangara E, Pruiti GP, Conti S, Di Grazia A, Ussia GP,

Capodanno D, Tamburino C. Early conduction disorders following

754 The American Journal of Cardiology (www.ajconline.org)

percutaneous aortic valve replacement. Pacing Clin Electrophysiol2009;32:S126–S130.

12. Sinhal A, Altwegg L, Pasupati S, Humphries KH, Allard M, Martin P,Kerr C, Lichtenstein SV, Webb JG. Atrioventricular block after trans-catheter balloon expandable aortic valve implantation. J Am CollCardiol Interv 2008;1:305–309.

13. Godin M, Eltchaninoff H, Furuta A, Tron C, Anselme F, Bejar K,Sanchez-Giron C, Bauer F, Litzler PY, Bessou JP, Cribier A. Fre-quency of conduction disturbances after transcatheter implantation ofan Edwards Sapien aortic valve prosthesis. Am J Cardiol 2010;106:707–712.

14. Gutiérrez M, Rodés-Cabau J, Bagur R, Doyle D, DeLarochellière R,Bergeron S, Bertrand OF, Pibarot P, Dumont E. Electrocardiographic

changes and clinical outcomes after transapical aortic valve implan-tation. Am Heart J 2009;158:302–308.

15. Moreno R, Dobarro D, Lopez de Sa E, Prieto M, Morales C, Calvo OrbeL, Galeote G, Jiménez-Valero S, Lopez-Sendon JL. Cause of completeatrioventricular block after percutaneous aortic valve implantation: in-sights from a necropsy study. Circulation 2009;120:e29–e30.

16. Serruys PW, Luijten HE, Beatt KJ, Di Mario C, de Feyter PJ, EssedCE, Roelandt JR, van den Brand M. Percutaneous balloon valvulo-plasty for calcific aortic stenosis. A treatment “sine cure?” Eur Heart J1988;9:782–794.

17. Napodano M, Tarantini G, Ramondo A. Is it reasonable to treat all

calcified stenotic valves with a valve stent? Probably yes if we get afull stent expansion. J Am Coll Cardiol 2009;53:219.