- Email: [email protected]
Predictors of Permanent Pacemaker Implantation After Transcatheter Aortic Valve Replacement
Predictors of Permanent Pacemaker Implantation After Transcatheter Aortic Valve Replacement Matthew J. Hoyt, MD, Jessica Hathaway, MD, Roseanne Palmer, RN, and Michael Beach, MD, PhD Objective: Determine predictors of permanent pacemaker (PPM) implantation after transcatheter aortic valve replacement (TAVR). Design: A retrospective chart review of patients undergoing TAVR at the authors’ institution. Extracted data included patient demographics, electrocardiogram, procedural, and echocardiographic data. Multivariate regression was performed to identify associations with PPM implantation. Setting: Single-center academic hospital. Participants: Patients undergoing TAVR. Interventions: This study was retrospective. No interventions were performed on patients. Measurements and Main Results: Baseline electrocardiogram, Society of Thoracic Surgeons score, age, and echocardiographic parameters were not predictors of PPM implantation. However, multiple deployments was a risk factor, and degree of paravalvular leak trended toward significance. Ten patients
required placement of a 2nd valve, or valve-in-valve (VIV). Of the 10 patients with VIV, 5 (50%) required a PPM, compared with 8 (14%) of 56 patients with a single valve (OR 6.0, p ¼ 0.02). PPM implantation occurred in 5 (42%) patients with no leak, 8 (19%) patients with trace leak, and no patients with mild or moderate leak (p ¼ 0.085). In patients with no or trace leak, VIV increased the likelihood of PPM from 17.4% to 62.5% (OR 7.9, p ¼ 0.006). For the 42 patients with trace leak, VIV increased the likelihood of PPM from 11.4% to 57.1% (OR 10.33, p ¼ 0.005). Conclusions: The authors found VIV placement, and likely degree of paravalvular leak, to be predictors of PPM placement. VIV and the degree of leak may be useful markers for postoperative prophylactic pacemaker placement. & 2015 Elsevier Inc. All rights reserved.
T
This study combined ECG, procedural, and echocardiographic characteristics to predict PPM implantation. The authors hypothesized that if the need for a PPM is related to conduction system injury, then procedural characteristics may influence the rate of conduction abnormality following the procedure. Many of the previous studies that have compared ECG findings and PPM implantation rate have not been evaluating solely the balloon expandable SAPIEN XT valve, whereas the authors’ study included only patients with SAPIEN XT implantation.
RANSCATHETER AORTIC VALVE replacement (TAVR) is an increasingly common procedure for replacement of the aortic valve in patients with aortic stenosis (AS). In patients older than 65, the overall prevalence of AS is 2%, increasing to 4% by the age of 85.1 Patients with severe aortic stenosis or symptomatic disease are considered for valve replacement. TAVR has found a niche in improving symptoms in patients deemed to be high-risk surgical candidates, with lower overall mortality.2,3,4 Injury to the conduction system that requires permanent pacemaker (PPM) placement is a known complication in both surgical aortic valve replacement and TAVR. This risk is believed generally to result from the proximity of the conducting system to the aortic valve.4–6 It is likely that the risk of PPM placement is secondary to mechanical trauma, inflammation, and tissue edema to the conduction system.7 Previous studies have used pre-procedural electrocardiogram (ECG) data to determine characteristics of patients who required PPM implantation, assuming that patients with pre-existing conduction disturbances would be at higher risk for more severe injury. In these studies, right bundle-branch block (RBBB) was identified most consistently as a risk factor for PPM implantation.8 Other studies have stratified risk by patient comorbidities such as diabetes mellitus, peripheral vascular disease, septal wall thickness, and male gender.8–10 Few studies have examined the effect of procedural characteristics in the prediction of need for PPM, although there is evidence that balloon predilation, using the self-expandable CoreValve device and significant valve oversizing may contribute to PPM requirement.11,12 An additional cofactor that has been shown to increase the risk of PPM placement following TAVR is low implantation of the valve and valve-in-valve procedures.4
From the Dartmouth-Hitchcock Medical Center, Lebanon, NH. Address reprint requests to Matthew J. Hoyt, MD, 1 Medical Center Dr. Lebanon, NH, USA 03766. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.06.001 1162
KEY WORDS: Transcatheter aortic valve replacement, ECG, ECHO, paravalvular leak, pacemaker
METHODS
This study was approved by the Dartmouth-Hitchcock Medical Center Committee for the Protection of Human Subjects and the Institutional Review Board for Dartmouth College. Written informed consent was waived due to the retrospective nature of the study. From February 6, 2012 to May 28, 2014, consecutive patients undergoing TAVR with the SAPIEN XT valve device for severe aortic stenosis formed the study population. This included the authors’ entire experience with the TAVR procedure at their institution at the time the investigation began. Exclusion criteria included perioperative mortality, previous pacemaker placement, and aborted procedure. For the procedure, patients were taken to the catheterization laboratory where standard American Society of Anesthesiologists monitors were applied and an arterial catheter was placed, most commonly in the radial artery. General anesthesia was induced in a controlled fashion followed by endotracheal intubation, transesophageal echocardiography (TEE) probe placement, central venous cannulation, and pulmonary artery catheter placement. The femoral veins and femoral arteries were cannulated by the surgical team. A pacemaker lead was placed in the right ventricle for rapid pacing during valvuloplasty, placement of the valve, and further dilations or deployments. If the approach was transapical, an incision was made in the fourth interspace on the left. Patients were heparinized prior to balloon valvuloplasty. Following this, a pacing run was performed, and the valve was deployed. Final valve placement was confirmed
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1162–1166
1163
PREDICTORS OF PERMANENT PACEMAKER IMPLANTATION
Table 1. Pre-Procedure Characteristics Combined
Mean or % N Pre-procedure RBBB Pre-procedure LBBB Pre-procedure left axis deviation Pre-procedure right axis deviation Pre-procedure intraventricular conduction delay Pre-procedure fascicular hemiblock Pre-procedure rhythm Sinus First-degree block Atrial fibrillation Atrial flutter Second-degree block Junctional LVH on ECG LVH on ECHO Annular calcification 1 2 3 STS Scores Age Pre PR length Pre-procedure heart rate Pre-procedure QRS length Pre-procedure QT Interval Pre-procedure QTc Pre-procedure calculated T-axis Pre-procedure calculated R-axis Pre-procedure calculated P-axis Pre-procedure ejection fraction
No Pacemaker
Pacemaker
SD or Mean or N % N
SD or Mean or N % N
Comparison
SD or Difference or N OR
24.24 10.61 16.67 4.55 9.09
66 66 66 66 66
16 7 11 3 6
24.53 9.43 15.09 3.77 9.43
53 53 53 53 53
13 5 8 2 5
23.08 15.38 23.08 7.69 7.69
13 13 13 13 13
3 2 3 1 1
0.92 1.75 1.69 2.12 0.80
13.64
66
9
11.32
53
6
23.08
13
3
2.35
65.15 4.55 25.76 1.52 1.52 1.52 24.24 13.64
66 66 66 66 66 66 66 66
43 3 17 1 1 1 16 9
66.04 5.66 24.53 1.89 0.00 1.89 24.53 11.32
53 53 53 53 53 53 53 53
35 3 13 1 0 1 13 6
61.54 0.00 30.77 0.00 7.69 0.00 23.08 23.08
13 13 13 13 13 13 13 13
8 0 4 0 1 0 3 3
0.82 0.00 1.37 0.00 .* 0.00 0.92 2.35
41.54 43.08 15.38 8.77 83.50 192.17 69.79 111.85 432.12 462.09 42.73 1.92 37.47 56.59
65 65 65 60 66 46 66 66 66 66 66 63 45 66
27 28 10 5.23 8.46 55.81 13.86 25.46 44.94 41.82 84.91 49.87 34.60 14.29
44.23 42.31 13.46 8.33 83.06 183.74 70.53 109.40 428.94 461.00 40.02 3.96 34.61 56.81
52 52 52 48 53 38 53 53 53 53 53 50 36 53
23 22 7 4.82 8.97 49.85 13.85 23.30 43.86 41.60 82.69 48.32 35.36 14.31
30.77 46.15 23.08 10.50 85.31 232.25 66.77 121.85 445.08 466.54 53.77 -5.92 48.89 55.69
13 13 13 12 13 8 13 13 13 13 13 13 9 13
4 6 3 6.60 5.85 68.31 13.99 32.00 48.73 44.11 96.23 56.86 30.53 14.74
0.56 1.17 1.93 -2.17 -2.25 -48.51 3.76 -12.45 -16.13 -5.54 -13.75 9.88 -14.28 1.12
95% CI (0.14, (0.15, (0.24, (0.03, (0.02,
4.38) 12.42) 8.76) 43.44) 8.22)
p 0.91 0.53 0.49 0.54 0.84
(0.32, 13.29) 0.27 0.36 (0.20, 3.69) 0.76 (0.00, 5.38) 0.38 (0.26, 5.98) 0.64 (0.00, .) 0.62 (0.00, .) .042 (0.00, .) 0.62 (0.14, 4.38) 0.91 (0.32, 13.29) 0.27 0.57 (0.11, 2.36) 0.38 (0.28, 4.71) 0.80 (0.27, 10.39) 0.39 (-6.50, 2.17) 0.30 (-6.41, 1.91) 0.28 (-106.04, 9.01) 0.089 (-5.28, 12.80) 0.40 (-32.47, 7.57) 0.21 (-47.24, 14.97) 0.29 (-33.85, 22.77) 0.69 (-74.85, 47.35) 0.64 (-26.29, 46.05) 0.57 (-39.30, 10.74) 0.24 (-8.38, 10.62) 0.81
NOTE. For categorical variables, percent with the risk factor and count are shown. For continuous variables, mean and standard deviation are shown. Abbreviations: CI, confidence interval; ECHO, echocardiogram; ECG, electrocardiogram; LBBB, left bundle-branch block; LVH, left ventricular hypertrophy; OR, odds ratio; RBBB, right bundle-branch block; SD, standard deviation; STS, Society of Thoracic Surgeons. *Incalculable data point.
with fluoroscopy and TEE. Grading of the paravalvular leak was done at this time by cardiology. If placement of the valve was complicated by a significant paravalvular leak, a postdeployment re-inflation was performed. If the valve was seated poorly, a second valve was deployed. Data were obtained retrospectively from chart review. Each patient underwent a pre- and post-procedure 12-lead ECG. These ECGs were examined for information, including PR length; QRS length; QT/QTc interval, calculated P, R, and T axes; right- or left-axis deviation; RBBB; left bundle-branch block; interventricular conduction delay; fascicular hemiblock; left ventricular hypertrophy (LVH); and rhythm. Postoperative surgical notes were reviewed for number of balloon valvuloplasties, valve-invalve (VIV) technique, number of post-placement valve inflations, and balloon inflation volume (standard inflation v over- or underinflation). Each echo report was reviewed for LVH, post-valve deployment leak, and annular calcification. Additionally, ejection fraction (EF) was obtained from the initial echocardiogram and when available at 24 hours and 30 days. For calcification, all valves were graded originally as having severe calcification on
initial read. Consequently, the images were reviewed individually by a single researcher, who was board certified in echocardiography, and given a grade of 1-3 (an arbitrary scale denoting an increasing degree of annular calcification within this cohort). Valvular leak was graded using the following nomenclature: none, trace, mild, mild-moderate, moderate, moderate-severe, and severe. This was in agreement with the guidelines developed by the Valve Academic Research Consortium.9 No patient had a degree of leak higher than moderate. A patient was determined to have required PPM placement if they had the procedure performed within 30 days of TAVR. STATISTICAL ANALYSIS
The primary outcome was the requirement for permanent pacemaker after the procedure that the authors modeled as a dichotomous variable. Univariate statistical analyses included unadjusted chi-squared or t-test as appropriate for categorical and continuous data, respectively. Odds ratios (ORs) with 95% confidence intervals were reported. Multivariate exact logistic
1164
HOYT ET AL
regression models were used because of multiple empty cells when VIV or degree of leak was entered as a covariate. The authors reported 95% confidence intervals and considered p o 0.05 to indicate statistical significance. They did not adjust for multiple comparisons. RESULTS
Of these 76 potentially eligible patients, 10 (13%) patients were excluded due to perioperative mortality (3 patients), existing pacemaker (4 patients), or aborted procedure or change to surgical AVR (3 patients). For the remaining 66 patients, the overall incidence of PPM placement was 19.6% (13 of 66). Table 1 lists the preoperative characteristics of the authors’ patient population and their effect on PPM placement. The majority of the patients were originally in sinus rhythm (65.1%), with the next most frequent rhythm being atrial fibrillation (25.8%). The only pre-procedure ECG characteristic associated with PPM placement was second-degree heart block, Mobitz type I (7.7% v 0%, p ¼ 0.042), but this result should be viewed with caution given the large number of pre-procedure variables collected. Sixteen (24.2%) and 9 (13.6%) patients showed signs of LVH on ECG and TEE, respectively, but there was no association with PPM placement. Additionally, no statistically significant differences were found for Society of Thoracic Surgeons score, patient age, or degree of annular calcification. In contrast to pre-procedural data, placement of a second valve was found to be a significant predictor of PPM placement (Table 2). Also, degree of paravalvular leak was found to be either significant (p ¼ 0.03 no leak v trace, mild, or moderate) or very-near significant (p ¼ 0.058 none or trace v mild or moderate) depending on the analysis performed. The apparent association of 2 or more attempts at valvuloplasty with pacemaker placement (p ¼ 0.04) should be viewed with caution given the multiple variables examined. Table 3 details requirements for PPM placement stratified by the degree of leak and implantation of VIV. In this cohort, 10 (15.1%) patients required placement of a second valve (VIV) during the same procedure. Of patients with VIV, 5 (50%) required a PPM compared with 8 (14%) of the 56 patients with a single valve (OR 6.0, 95% CI 1.07-32.07, p ¼ 0.02). In regards to paravalvular leak, a PPM was placed in 5 of 12 (42%) patients with no leak, 8 of 42 (19%) patients with trace leak, and 0/12 patients with either mild or moderate leak (p ¼ 0.085). Among patients with no leak or trace leak, VIV increased the likelihood of receiving a PPM from 17.4% to 62.5% (OR 7.9, 95% CI 1.2-58.7, p ¼ 0.006). For the 42 patients with trace leak (the largest group), VIV increased the likelihood of needing a pacemaker from 11.4% to 57.1% (OR 10.33, 95% CI 1.2-92.7, p ¼ 0.005). Exact logistic regression models are presented in Table 4 and also supported the combined effect of VIV and degree of leak on predicting PPM placement. In models B and C, leak was categorized as none or trace versus mild or moderate. In models D and E, leak was categorized as none versus trace, mild, or moderate. Regardless of degree of leak, VIV increases the odds of pacemaker placement. Degree of leak is significant, depending on how it is categorized. Given the relatively low number of pacemakers placed, the authors were unable to fit models that considered leak as an ordinal variable using all 4 levels.
DISCUSSION
In this retrospective review of TAVR procedures, the authors attempted to identify possible predictors of PPM implantation following TAVR. They identified second-degree heart block, multiple balloon valvuloplasties, degree of paravalvular leak, and VIV procedures as predictors of PPM placement. The authors’ finding of second-degree heart block as a risk factor should be interpreted with some caution as the p value of 0.042 was not adjusted for multiple comparisons. The same should be kept in mind when considering more than 2 valvuloplasties. Interestingly, the authors’ study did not show RBBB as a significant predictor of need for PPM; this has been the most consistently identified conduction abnormality associated with a need for PPM placement.8 To the authors’ knowledge, this was the first study to suggest that the degree of paravalvular regurgitation may be an important predictor of PPM placement. The effect was most apparent for no leak either considered alone (univariate comparison in Table 2) or when compared to any degree of leak when the effect of VIV was included (multivariate model E). There was a trend that did not reach statistical significance when no leak and trace leak were combined into 1 group (multivariate models B and C). Additionally, the authors have shown an increased risk of PPM placement after VIV. Their study supported the increased risk of PPM implantation after VIV procedures, as previously was identified by Makkar et al.13 From the authors’ analysis, patients also were likely to be at increased risk when the 2 events occurred in the same patient. Again, the authors hypothesized that there was some degree of trauma with valve placement and that the placement of multiple valves was likely to cause increased trauma when compared to single valve placement. The authors also suggested that the degree of leak was a correlate of the amount of contact the procedural valve had with the annulus and that more contact created more tissue trauma. Given the findings from the authors’ study, it would be reasonable to consider placing a post-procedure transvenous pacemaker in those patients who have VIV procedures, no or trace leak, or combined events. The exact degree of leak that would suggest pacemaker placement was unclear from the authors’ study. The multiple models they present in Table 4 were not developed a priori. Further research with larger cohorts will need to be done to clarify this, but, with the data presented in their study, there should be some consideration for transvenous pacemaker placement in patients with less leak and other risk factors for PPM placement. Placement easily can be facilitated via the venous access used during the case or a second venous stick at the conclusion of the procedure, taking advantage of the availability of fluoroscopy. Additionally, the authors found that the majority of their PPM placements occurred within the first 1 to 2 postoperative days (mean ¼ 5, median and mode ¼ 2), a reasonable timeline to rely on a transvenous pacemaker. While the authors’ study may appear to suggest that there is a population of patients who benefit from tolerance of a higher degree of aortic insufficiency following TAVR to protect against PPM placement, they are not advocating intentional post-procedure leak. Previous studies14 have shown that
1165
PREDICTORS OF PERMANENT PACEMAKER IMPLANTATION
Table 2. Peri-Procedure and Post-Procedure Data Combined
Post-procedure RBBB Post-procedure LBBB Post-procedure left axis deviation Post-procedure right axis deviation Post-procedure conduction delay Post-procedure fascicular hemi-block Post-procedure rhythm Sinus First-degree block Atrial fibrillation Atrial flutter Second-degree block Paced Leak None Trace Mild Moderate Balloon volume Standard Increase in volume Decrease in volume Pre-deployment valvuloplasty Zero One Two or more Post-deployment re-inflations Zero One Two or more VIV Leak* Length of stay Ejection fraction 24 hours Ejection fraction 30 days
No Pacemaker
Mean or %
N
SD or N
22.73 27.27 33.33
66 66 66
3.03
Pacemaker
Mean or %
N
SD or N
15 18 22
24.53 28.30 32.08
53 53 53
66
2
1.89
15.15
66
10
15.15
66
43.94 18.18 27.27 1.52 3.03 6.06
Comparison
Mean or %
SD or N
Difference or OR
N
13 15 17
15.38 23.08 38.46
13 13 13
2 3 5
0.56 0.76 1.32
(0.05, 3.13) (0.12, 3.55) (0.29, 5.42)
0.48 0.70 0.66
53
1
7.69
13
1
4.33
(0.05, 346.48)
0.27
16.98
53
9
7.69
13
1
0.41
(0.01, 3.53)
0.40
10
16.98
53
9
7.69
13
1
0.41
(0.01, 3.53)
0.40
66 66 66 66 66 66
29 12 18 1 2 4
49.06 18.87 28.30 1.89 1.89 0.00
53 53 53 53 53 53
26 10 15 1 1 0
23.08 15.38 23.08 0.00 7.69 30.77
13 13 13 13 13 13
3 2 3 0 1 4
0.31 0.78 0.76 0.00 4.33 .
(0.05, (0.07, (0.12, (0.00, (0.05, (5.45,
18.18 63.64 13.64 4.55
66 66 66 66
12 42 9 3
13.21 64.15 16.98 5.66
53 53 53 53
7 34 9 3
38.46 61.54 0.00 0.00
13 13 13 13
5 8 0 0
4.11† 0.89 0.00 0.00
(0.80, (0.22, (0.00, (0.00,
65.15 24.24 10.61
66 66 66
43 16 7
66.04 24.53 9.43
53 53 53
35 13 5
61.54 23.08 15.38
13 13 13
8 3 2
0.82 0.92 1.75
(0.20, (0.14, (0.15,
15.15 83.33 1.52
66 66 66
10 55 1
16.98 83.02 0.00
53 53 53
9 44 0
7.69 84.62 7.69
13 13 13
1 11 1
0.41 1.12 .
(0.01, (0.19, (0.00,
71.21 27.27 1.52 15.15 81.82 8.53 60.22 59.23
66 66 66 66 66 57 63 62
47 18 1 10 54 5.17 12.17 11.53
69.81 28.30 1.89 9.43 77.36 8.43 60.85 59.64
53 53 53 53 53 44 52 50
37 15 1 5 41 5.60 11.63 11.81
76.92 23.08 0.00 38.46 100.00 8.85 57.27 57.50
13 13 13 13 13 13 11 12
10 3 0 5 13 3.51 14.72 10.55
1.44 0.76 0.00 6.00‡ . -0.41 3.57 2.14
(0.31, (0.12, (0.00, (1.07, (0.93, (-3.03, (-6.61, (-5.12,
95% CI
p
o 0.01 1.42) .091 4.55) 0.77 3.55) 0.70 .) 0.62 346.48) 0.27 .) o 0.001 .085 19.46) 0.034 4.00) 0.86 1.55) 0.11 5.38) 0.38 0.82 3.69) 0.76 4.38) 0.91 12.42) 0.53 0.097 3.53) 0.40 12.15) 0.89 .) .042 0.81 9.18) 0.61 3.55) 0.70 .) 0.62 32.07) o 0.01 .) 0.058 2.20) 0.75 13.75) 0.46 9.40) 0.55
NOTE. For categorical variables, percent with the risk factor and count are shown. For continuous variables, mean and standard deviation are shown. Abbreviations: CI, confidence interval; LBBB, left bundle branch block; OR, odds ratio; RBBB, right bundle-branch block; SD, standard deviation; VIV, valve in valve. *Leak categorized as none or trace versus mild or moderate. †P values less than 0.05. ‡P values less than 0.01.
Table 3. Permanent Pacemaker Placement by Degree of Leak and Deployment Attempts Leak
VIV
None
Trace
Mild
Moderate
No Yes
4/11 (36.4%) 1/1 (100%)
4/35 (11.4%) 4/7 (57.1%)
0/8 (0%) 0/1 (0%)
0/2 (0%) 0/1 (0%)
NOTE. Cell entries reflect the number and percentage of patients who required permanent pacemaker placement. Abbreviations: VIV, valve-in-valve.
patients with post-TAVR aortic insufficiency (AI) have decreased survival benefit compared to those patients without AI. The degree of AI also has shown a mortality gradient, with increasing AI having a higher risk of mortality.14 Attempts at improving the degree of leak likely are warranted, despite increased PPM risk, as improved mortality benefit outweighs the harm from PPM implantation. The authors’ study did possess several limitations. In general, their study had a relatively small population with an uncommon event. This does increase the difficulty in
1166
HOYT ET AL
Table 4. Multivariate Exact Logistic Models for Pacemaker Prediction Model
Model A VIV Model B Leak* Model C VIV Leak* Model D Leak† Model E VIV Leak†
OR
95% CI
p Value
5.77
(1.07, 32.07)
0.04
5.05
(0.74, 1)
0.11
7.51 6.03
(1.19, 58.71) (0.83, 1)
0.029 0.082
4.00
(0.80, 19.46)
0.099
6.13‡ 8.51‡
(1.04, 40.09) (1.37, 60.69)
0.045 0.019
NOTE. Analysis includes all 66 patients. Abbreviations: CI, confidence interval; OR, odds ratio; VIV, valve in valve. *Leak is categorized as none or trace versus mild or moderate. †Leak is categorized as none versus trace, mild, or moderate. ‡Indicates I value o 0.01.
identifying differences within the study population. The authors believe that it would be beneficial to continue to build a database to investigate these uncommon outcomes, especially investigating the importance of the degree of paravalvular leak in predicting PPM placement as this may be a significant
contributor to the risk of PPM placement. Since the authors’ study looked at patients in a new TAVR program, there likely will be some change in the rate of complications associated with the procedure. This can be seen in their relatively high rate of PPM implantation in their cohort compared to previous studies. A meta-analysis in 2012 stated a PPM placement rate of 5.9% for the Edwards Sapien valve.15 There is evidence that as teams gain experience placing the valves, they increase their proficiency, suggesting a learning curve to the procedure.16 This, in part, could be due to better placement of the valve with fewer low implantations as experience with the procedure improves. The authors’ high rate of PPM placement was also surprising given that the Edwards Sapient XT is associated with a lower incidence of PPM placement.15 CONCLUSION
PPM implantation is a known complication of TAVR, and the authors’ data suggest that patients with VIV and those with a less significant degree of paravalvular leak are at a higher risk for PPM placement following the procedure. Based on the outcomes of the authors’ study, they believe that this finding supports the placement of temporary transvenous pacemakers in the highest risk group to avoid complications of bradycardia postoperatively.
REFERENCES 1. Carabello BA, Paulus WJ: Aortic stenosis. Lancet 373:956-966, 2009 2. Leon MB, Smith CR, Mack M, et al: Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med 363:1597-1607, 2010 3. Makkar RR, Fontana GP, Leon MB, et al: Transcatheter aorticvalve replacement for inoperable severe aortic stenosis. N Engl J Med 366:1697-1704, 2012 4. Binder R, Webb J, Toggweiler S, et al: Impact of postimplantation sapien XT geometry and position on conduction disturbances, hemodynamic performance, and paravalvular regurgitation. JACC Cardiovasc Interv 6:462-468, 2013 5. Piazza N, Onuma Y, Jesserun E, et al: Early and persistent intraventricular conduction abnormalities and requirements for pacemaking after percutaneous replacement of the aortic valve. JACC Cardiovascular Intervention 1:310-316, 2008 6. Piaxxa N, Grube E, Greckens U, et al: Procedural and 30-day outcomes following transcatheter aortic valve implantation using the third generation (18Fr) CoreValve Revalving System: Results from the Multicenter, Expanded Evaluation Registry 1 year following CE mark approval. Euro Intervention 4:242-249, 2008 7. Klein AA, Skubas NJ, Ender J: Contoversies and complications in the perioperative management of transcatheter aortic valve replacement. Anesth Analg 119:784-798, 2014 8. Siontis GCM, Juni P, Pilgrim T, et al: Predictors of permanent pacemaker implantation in patients with severe aortic stenosis undergoing TAVR. J Am Coll Cardiol 64:129-140, 2014 9. Leon MB, Piazza N, Nikolsky, et al: Standardized endpoint definitions for transcatheter aortic valve implantation trials:
A consensus report from the Valve Academic Research Consortium. J Am Coll Cardiol 57:253-269, 2011 10. Siontis GC, Juni P, Pilgrim T, et al: Predictors of permanent pacemaker implantation in patients with severe aortic stenosis undergoing TAVR: A meta-analysis. J Am Coll Cardiol 64:129-140, 2014 11. Khawaja MZ, Rajani R, Cook A, et al: Permanent pacemaker insertion after CoreValve transcatheter aortic valve implantation: Incidence and contributing factors (the UK CoreValve Collaborative). Circulation 123:951-960, 2011 12. Bleiziffer S, Ruge H, Horer J, et al: Predictors for new-onset complete heart block after transcatheter aortic valve implantation. JACC Cardiovascular Intervention 3:524-530, 2013 13. Makkar RR, Jilaihawi H, Chakravarty T, et al: Determinants and outcomes of acute transcatheter valve-in-valve therapy of embolization: A study of multiple valve implants in the U.S. Partner Trial (Placement of AoRTic TraNscathetER Valve Trial Edwards Sapien Transcatheter Heart Valve). Jo Am Coll Cardiol 62:418-430, 2013 14. Kodali SK, Williams MR, Smith CR, et al: PARTNER Trial Investigators. Two-year outcomes after transcatheter or surgical aorticvalve replacement. N Engl J Med 366:1686-1695, 2012 15. Jilaihawi H, Chakravarty, Weiss RE, et al: Meta-analysis of complications in aortic valve replacement: Comparison of MedtronicCorevalve, Edwards-Sapien, and Surgical Aortic Valve Replacement in 8,536 patients. Cath Cardiovasc Interv 80:128-138, 2012 16. Alli O, Booker JD, Lennon RJ, et al: Transcatheter aortic valve implantation assessing the learning curve. JACC Cardiovasc Interv 5: 72-79, 2012
required placement of a 2nd valve, or valve-in-valve (VIV). Of the 10 patients with VIV, 5 (50%) required a PPM, compared with 8 (14%) of 56 patients with a single valve (OR 6.0, p ¼ 0.02). PPM implantation occurred in 5 (42%) patients with no leak, 8 (19%) patients with trace leak, and no patients with mild or moderate leak (p ¼ 0.085). In patients with no or trace leak, VIV increased the likelihood of PPM from 17.4% to 62.5% (OR 7.9, p ¼ 0.006). For the 42 patients with trace leak, VIV increased the likelihood of PPM from 11.4% to 57.1% (OR 10.33, p ¼ 0.005). Conclusions: The authors found VIV placement, and likely degree of paravalvular leak, to be predictors of PPM placement. VIV and the degree of leak may be useful markers for postoperative prophylactic pacemaker placement. & 2015 Elsevier Inc. All rights reserved.
T
This study combined ECG, procedural, and echocardiographic characteristics to predict PPM implantation. The authors hypothesized that if the need for a PPM is related to conduction system injury, then procedural characteristics may influence the rate of conduction abnormality following the procedure. Many of the previous studies that have compared ECG findings and PPM implantation rate have not been evaluating solely the balloon expandable SAPIEN XT valve, whereas the authors’ study included only patients with SAPIEN XT implantation.
RANSCATHETER AORTIC VALVE replacement (TAVR) is an increasingly common procedure for replacement of the aortic valve in patients with aortic stenosis (AS). In patients older than 65, the overall prevalence of AS is 2%, increasing to 4% by the age of 85.1 Patients with severe aortic stenosis or symptomatic disease are considered for valve replacement. TAVR has found a niche in improving symptoms in patients deemed to be high-risk surgical candidates, with lower overall mortality.2,3,4 Injury to the conduction system that requires permanent pacemaker (PPM) placement is a known complication in both surgical aortic valve replacement and TAVR. This risk is believed generally to result from the proximity of the conducting system to the aortic valve.4–6 It is likely that the risk of PPM placement is secondary to mechanical trauma, inflammation, and tissue edema to the conduction system.7 Previous studies have used pre-procedural electrocardiogram (ECG) data to determine characteristics of patients who required PPM implantation, assuming that patients with pre-existing conduction disturbances would be at higher risk for more severe injury. In these studies, right bundle-branch block (RBBB) was identified most consistently as a risk factor for PPM implantation.8 Other studies have stratified risk by patient comorbidities such as diabetes mellitus, peripheral vascular disease, septal wall thickness, and male gender.8–10 Few studies have examined the effect of procedural characteristics in the prediction of need for PPM, although there is evidence that balloon predilation, using the self-expandable CoreValve device and significant valve oversizing may contribute to PPM requirement.11,12 An additional cofactor that has been shown to increase the risk of PPM placement following TAVR is low implantation of the valve and valve-in-valve procedures.4
From the Dartmouth-Hitchcock Medical Center, Lebanon, NH. Address reprint requests to Matthew J. Hoyt, MD, 1 Medical Center Dr. Lebanon, NH, USA 03766. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.06.001 1162
KEY WORDS: Transcatheter aortic valve replacement, ECG, ECHO, paravalvular leak, pacemaker
METHODS
This study was approved by the Dartmouth-Hitchcock Medical Center Committee for the Protection of Human Subjects and the Institutional Review Board for Dartmouth College. Written informed consent was waived due to the retrospective nature of the study. From February 6, 2012 to May 28, 2014, consecutive patients undergoing TAVR with the SAPIEN XT valve device for severe aortic stenosis formed the study population. This included the authors’ entire experience with the TAVR procedure at their institution at the time the investigation began. Exclusion criteria included perioperative mortality, previous pacemaker placement, and aborted procedure. For the procedure, patients were taken to the catheterization laboratory where standard American Society of Anesthesiologists monitors were applied and an arterial catheter was placed, most commonly in the radial artery. General anesthesia was induced in a controlled fashion followed by endotracheal intubation, transesophageal echocardiography (TEE) probe placement, central venous cannulation, and pulmonary artery catheter placement. The femoral veins and femoral arteries were cannulated by the surgical team. A pacemaker lead was placed in the right ventricle for rapid pacing during valvuloplasty, placement of the valve, and further dilations or deployments. If the approach was transapical, an incision was made in the fourth interspace on the left. Patients were heparinized prior to balloon valvuloplasty. Following this, a pacing run was performed, and the valve was deployed. Final valve placement was confirmed
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1162–1166
1163
PREDICTORS OF PERMANENT PACEMAKER IMPLANTATION
Table 1. Pre-Procedure Characteristics Combined
Mean or % N Pre-procedure RBBB Pre-procedure LBBB Pre-procedure left axis deviation Pre-procedure right axis deviation Pre-procedure intraventricular conduction delay Pre-procedure fascicular hemiblock Pre-procedure rhythm Sinus First-degree block Atrial fibrillation Atrial flutter Second-degree block Junctional LVH on ECG LVH on ECHO Annular calcification 1 2 3 STS Scores Age Pre PR length Pre-procedure heart rate Pre-procedure QRS length Pre-procedure QT Interval Pre-procedure QTc Pre-procedure calculated T-axis Pre-procedure calculated R-axis Pre-procedure calculated P-axis Pre-procedure ejection fraction
No Pacemaker
Pacemaker
SD or Mean or N % N
SD or Mean or N % N
Comparison
SD or Difference or N OR
24.24 10.61 16.67 4.55 9.09
66 66 66 66 66
16 7 11 3 6
24.53 9.43 15.09 3.77 9.43
53 53 53 53 53
13 5 8 2 5
23.08 15.38 23.08 7.69 7.69
13 13 13 13 13
3 2 3 1 1
0.92 1.75 1.69 2.12 0.80
13.64
66
9
11.32
53
6
23.08
13
3
2.35
65.15 4.55 25.76 1.52 1.52 1.52 24.24 13.64
66 66 66 66 66 66 66 66
43 3 17 1 1 1 16 9
66.04 5.66 24.53 1.89 0.00 1.89 24.53 11.32
53 53 53 53 53 53 53 53
35 3 13 1 0 1 13 6
61.54 0.00 30.77 0.00 7.69 0.00 23.08 23.08
13 13 13 13 13 13 13 13
8 0 4 0 1 0 3 3
0.82 0.00 1.37 0.00 .* 0.00 0.92 2.35
41.54 43.08 15.38 8.77 83.50 192.17 69.79 111.85 432.12 462.09 42.73 1.92 37.47 56.59
65 65 65 60 66 46 66 66 66 66 66 63 45 66
27 28 10 5.23 8.46 55.81 13.86 25.46 44.94 41.82 84.91 49.87 34.60 14.29
44.23 42.31 13.46 8.33 83.06 183.74 70.53 109.40 428.94 461.00 40.02 3.96 34.61 56.81
52 52 52 48 53 38 53 53 53 53 53 50 36 53
23 22 7 4.82 8.97 49.85 13.85 23.30 43.86 41.60 82.69 48.32 35.36 14.31
30.77 46.15 23.08 10.50 85.31 232.25 66.77 121.85 445.08 466.54 53.77 -5.92 48.89 55.69
13 13 13 12 13 8 13 13 13 13 13 13 9 13
4 6 3 6.60 5.85 68.31 13.99 32.00 48.73 44.11 96.23 56.86 30.53 14.74
0.56 1.17 1.93 -2.17 -2.25 -48.51 3.76 -12.45 -16.13 -5.54 -13.75 9.88 -14.28 1.12
95% CI (0.14, (0.15, (0.24, (0.03, (0.02,
4.38) 12.42) 8.76) 43.44) 8.22)
p 0.91 0.53 0.49 0.54 0.84
(0.32, 13.29) 0.27 0.36 (0.20, 3.69) 0.76 (0.00, 5.38) 0.38 (0.26, 5.98) 0.64 (0.00, .) 0.62 (0.00, .) .042 (0.00, .) 0.62 (0.14, 4.38) 0.91 (0.32, 13.29) 0.27 0.57 (0.11, 2.36) 0.38 (0.28, 4.71) 0.80 (0.27, 10.39) 0.39 (-6.50, 2.17) 0.30 (-6.41, 1.91) 0.28 (-106.04, 9.01) 0.089 (-5.28, 12.80) 0.40 (-32.47, 7.57) 0.21 (-47.24, 14.97) 0.29 (-33.85, 22.77) 0.69 (-74.85, 47.35) 0.64 (-26.29, 46.05) 0.57 (-39.30, 10.74) 0.24 (-8.38, 10.62) 0.81
NOTE. For categorical variables, percent with the risk factor and count are shown. For continuous variables, mean and standard deviation are shown. Abbreviations: CI, confidence interval; ECHO, echocardiogram; ECG, electrocardiogram; LBBB, left bundle-branch block; LVH, left ventricular hypertrophy; OR, odds ratio; RBBB, right bundle-branch block; SD, standard deviation; STS, Society of Thoracic Surgeons. *Incalculable data point.
with fluoroscopy and TEE. Grading of the paravalvular leak was done at this time by cardiology. If placement of the valve was complicated by a significant paravalvular leak, a postdeployment re-inflation was performed. If the valve was seated poorly, a second valve was deployed. Data were obtained retrospectively from chart review. Each patient underwent a pre- and post-procedure 12-lead ECG. These ECGs were examined for information, including PR length; QRS length; QT/QTc interval, calculated P, R, and T axes; right- or left-axis deviation; RBBB; left bundle-branch block; interventricular conduction delay; fascicular hemiblock; left ventricular hypertrophy (LVH); and rhythm. Postoperative surgical notes were reviewed for number of balloon valvuloplasties, valve-invalve (VIV) technique, number of post-placement valve inflations, and balloon inflation volume (standard inflation v over- or underinflation). Each echo report was reviewed for LVH, post-valve deployment leak, and annular calcification. Additionally, ejection fraction (EF) was obtained from the initial echocardiogram and when available at 24 hours and 30 days. For calcification, all valves were graded originally as having severe calcification on
initial read. Consequently, the images were reviewed individually by a single researcher, who was board certified in echocardiography, and given a grade of 1-3 (an arbitrary scale denoting an increasing degree of annular calcification within this cohort). Valvular leak was graded using the following nomenclature: none, trace, mild, mild-moderate, moderate, moderate-severe, and severe. This was in agreement with the guidelines developed by the Valve Academic Research Consortium.9 No patient had a degree of leak higher than moderate. A patient was determined to have required PPM placement if they had the procedure performed within 30 days of TAVR. STATISTICAL ANALYSIS
The primary outcome was the requirement for permanent pacemaker after the procedure that the authors modeled as a dichotomous variable. Univariate statistical analyses included unadjusted chi-squared or t-test as appropriate for categorical and continuous data, respectively. Odds ratios (ORs) with 95% confidence intervals were reported. Multivariate exact logistic
1164
HOYT ET AL
regression models were used because of multiple empty cells when VIV or degree of leak was entered as a covariate. The authors reported 95% confidence intervals and considered p o 0.05 to indicate statistical significance. They did not adjust for multiple comparisons. RESULTS
Of these 76 potentially eligible patients, 10 (13%) patients were excluded due to perioperative mortality (3 patients), existing pacemaker (4 patients), or aborted procedure or change to surgical AVR (3 patients). For the remaining 66 patients, the overall incidence of PPM placement was 19.6% (13 of 66). Table 1 lists the preoperative characteristics of the authors’ patient population and their effect on PPM placement. The majority of the patients were originally in sinus rhythm (65.1%), with the next most frequent rhythm being atrial fibrillation (25.8%). The only pre-procedure ECG characteristic associated with PPM placement was second-degree heart block, Mobitz type I (7.7% v 0%, p ¼ 0.042), but this result should be viewed with caution given the large number of pre-procedure variables collected. Sixteen (24.2%) and 9 (13.6%) patients showed signs of LVH on ECG and TEE, respectively, but there was no association with PPM placement. Additionally, no statistically significant differences were found for Society of Thoracic Surgeons score, patient age, or degree of annular calcification. In contrast to pre-procedural data, placement of a second valve was found to be a significant predictor of PPM placement (Table 2). Also, degree of paravalvular leak was found to be either significant (p ¼ 0.03 no leak v trace, mild, or moderate) or very-near significant (p ¼ 0.058 none or trace v mild or moderate) depending on the analysis performed. The apparent association of 2 or more attempts at valvuloplasty with pacemaker placement (p ¼ 0.04) should be viewed with caution given the multiple variables examined. Table 3 details requirements for PPM placement stratified by the degree of leak and implantation of VIV. In this cohort, 10 (15.1%) patients required placement of a second valve (VIV) during the same procedure. Of patients with VIV, 5 (50%) required a PPM compared with 8 (14%) of the 56 patients with a single valve (OR 6.0, 95% CI 1.07-32.07, p ¼ 0.02). In regards to paravalvular leak, a PPM was placed in 5 of 12 (42%) patients with no leak, 8 of 42 (19%) patients with trace leak, and 0/12 patients with either mild or moderate leak (p ¼ 0.085). Among patients with no leak or trace leak, VIV increased the likelihood of receiving a PPM from 17.4% to 62.5% (OR 7.9, 95% CI 1.2-58.7, p ¼ 0.006). For the 42 patients with trace leak (the largest group), VIV increased the likelihood of needing a pacemaker from 11.4% to 57.1% (OR 10.33, 95% CI 1.2-92.7, p ¼ 0.005). Exact logistic regression models are presented in Table 4 and also supported the combined effect of VIV and degree of leak on predicting PPM placement. In models B and C, leak was categorized as none or trace versus mild or moderate. In models D and E, leak was categorized as none versus trace, mild, or moderate. Regardless of degree of leak, VIV increases the odds of pacemaker placement. Degree of leak is significant, depending on how it is categorized. Given the relatively low number of pacemakers placed, the authors were unable to fit models that considered leak as an ordinal variable using all 4 levels.
DISCUSSION
In this retrospective review of TAVR procedures, the authors attempted to identify possible predictors of PPM implantation following TAVR. They identified second-degree heart block, multiple balloon valvuloplasties, degree of paravalvular leak, and VIV procedures as predictors of PPM placement. The authors’ finding of second-degree heart block as a risk factor should be interpreted with some caution as the p value of 0.042 was not adjusted for multiple comparisons. The same should be kept in mind when considering more than 2 valvuloplasties. Interestingly, the authors’ study did not show RBBB as a significant predictor of need for PPM; this has been the most consistently identified conduction abnormality associated with a need for PPM placement.8 To the authors’ knowledge, this was the first study to suggest that the degree of paravalvular regurgitation may be an important predictor of PPM placement. The effect was most apparent for no leak either considered alone (univariate comparison in Table 2) or when compared to any degree of leak when the effect of VIV was included (multivariate model E). There was a trend that did not reach statistical significance when no leak and trace leak were combined into 1 group (multivariate models B and C). Additionally, the authors have shown an increased risk of PPM placement after VIV. Their study supported the increased risk of PPM implantation after VIV procedures, as previously was identified by Makkar et al.13 From the authors’ analysis, patients also were likely to be at increased risk when the 2 events occurred in the same patient. Again, the authors hypothesized that there was some degree of trauma with valve placement and that the placement of multiple valves was likely to cause increased trauma when compared to single valve placement. The authors also suggested that the degree of leak was a correlate of the amount of contact the procedural valve had with the annulus and that more contact created more tissue trauma. Given the findings from the authors’ study, it would be reasonable to consider placing a post-procedure transvenous pacemaker in those patients who have VIV procedures, no or trace leak, or combined events. The exact degree of leak that would suggest pacemaker placement was unclear from the authors’ study. The multiple models they present in Table 4 were not developed a priori. Further research with larger cohorts will need to be done to clarify this, but, with the data presented in their study, there should be some consideration for transvenous pacemaker placement in patients with less leak and other risk factors for PPM placement. Placement easily can be facilitated via the venous access used during the case or a second venous stick at the conclusion of the procedure, taking advantage of the availability of fluoroscopy. Additionally, the authors found that the majority of their PPM placements occurred within the first 1 to 2 postoperative days (mean ¼ 5, median and mode ¼ 2), a reasonable timeline to rely on a transvenous pacemaker. While the authors’ study may appear to suggest that there is a population of patients who benefit from tolerance of a higher degree of aortic insufficiency following TAVR to protect against PPM placement, they are not advocating intentional post-procedure leak. Previous studies14 have shown that
1165
PREDICTORS OF PERMANENT PACEMAKER IMPLANTATION
Table 2. Peri-Procedure and Post-Procedure Data Combined
Post-procedure RBBB Post-procedure LBBB Post-procedure left axis deviation Post-procedure right axis deviation Post-procedure conduction delay Post-procedure fascicular hemi-block Post-procedure rhythm Sinus First-degree block Atrial fibrillation Atrial flutter Second-degree block Paced Leak None Trace Mild Moderate Balloon volume Standard Increase in volume Decrease in volume Pre-deployment valvuloplasty Zero One Two or more Post-deployment re-inflations Zero One Two or more VIV Leak* Length of stay Ejection fraction 24 hours Ejection fraction 30 days
No Pacemaker
Mean or %
N
SD or N
22.73 27.27 33.33
66 66 66
3.03
Pacemaker
Mean or %
N
SD or N
15 18 22
24.53 28.30 32.08
53 53 53
66
2
1.89
15.15
66
10
15.15
66
43.94 18.18 27.27 1.52 3.03 6.06
Comparison
Mean or %
SD or N
Difference or OR
N
13 15 17
15.38 23.08 38.46
13 13 13
2 3 5
0.56 0.76 1.32
(0.05, 3.13) (0.12, 3.55) (0.29, 5.42)
0.48 0.70 0.66
53
1
7.69
13
1
4.33
(0.05, 346.48)
0.27
16.98
53
9
7.69
13
1
0.41
(0.01, 3.53)
0.40
10
16.98
53
9
7.69
13
1
0.41
(0.01, 3.53)
0.40
66 66 66 66 66 66
29 12 18 1 2 4
49.06 18.87 28.30 1.89 1.89 0.00
53 53 53 53 53 53
26 10 15 1 1 0
23.08 15.38 23.08 0.00 7.69 30.77
13 13 13 13 13 13
3 2 3 0 1 4
0.31 0.78 0.76 0.00 4.33 .
(0.05, (0.07, (0.12, (0.00, (0.05, (5.45,
18.18 63.64 13.64 4.55
66 66 66 66
12 42 9 3
13.21 64.15 16.98 5.66
53 53 53 53
7 34 9 3
38.46 61.54 0.00 0.00
13 13 13 13
5 8 0 0
4.11† 0.89 0.00 0.00
(0.80, (0.22, (0.00, (0.00,
65.15 24.24 10.61
66 66 66
43 16 7
66.04 24.53 9.43
53 53 53
35 13 5
61.54 23.08 15.38
13 13 13
8 3 2
0.82 0.92 1.75
(0.20, (0.14, (0.15,
15.15 83.33 1.52
66 66 66
10 55 1
16.98 83.02 0.00
53 53 53
9 44 0
7.69 84.62 7.69
13 13 13
1 11 1
0.41 1.12 .
(0.01, (0.19, (0.00,
71.21 27.27 1.52 15.15 81.82 8.53 60.22 59.23
66 66 66 66 66 57 63 62
47 18 1 10 54 5.17 12.17 11.53
69.81 28.30 1.89 9.43 77.36 8.43 60.85 59.64
53 53 53 53 53 44 52 50
37 15 1 5 41 5.60 11.63 11.81
76.92 23.08 0.00 38.46 100.00 8.85 57.27 57.50
13 13 13 13 13 13 11 12
10 3 0 5 13 3.51 14.72 10.55
1.44 0.76 0.00 6.00‡ . -0.41 3.57 2.14
(0.31, (0.12, (0.00, (1.07, (0.93, (-3.03, (-6.61, (-5.12,
95% CI
p
o 0.01 1.42) .091 4.55) 0.77 3.55) 0.70 .) 0.62 346.48) 0.27 .) o 0.001 .085 19.46) 0.034 4.00) 0.86 1.55) 0.11 5.38) 0.38 0.82 3.69) 0.76 4.38) 0.91 12.42) 0.53 0.097 3.53) 0.40 12.15) 0.89 .) .042 0.81 9.18) 0.61 3.55) 0.70 .) 0.62 32.07) o 0.01 .) 0.058 2.20) 0.75 13.75) 0.46 9.40) 0.55
NOTE. For categorical variables, percent with the risk factor and count are shown. For continuous variables, mean and standard deviation are shown. Abbreviations: CI, confidence interval; LBBB, left bundle branch block; OR, odds ratio; RBBB, right bundle-branch block; SD, standard deviation; VIV, valve in valve. *Leak categorized as none or trace versus mild or moderate. †P values less than 0.05. ‡P values less than 0.01.
Table 3. Permanent Pacemaker Placement by Degree of Leak and Deployment Attempts Leak
VIV
None
Trace
Mild
Moderate
No Yes
4/11 (36.4%) 1/1 (100%)
4/35 (11.4%) 4/7 (57.1%)
0/8 (0%) 0/1 (0%)
0/2 (0%) 0/1 (0%)
NOTE. Cell entries reflect the number and percentage of patients who required permanent pacemaker placement. Abbreviations: VIV, valve-in-valve.
patients with post-TAVR aortic insufficiency (AI) have decreased survival benefit compared to those patients without AI. The degree of AI also has shown a mortality gradient, with increasing AI having a higher risk of mortality.14 Attempts at improving the degree of leak likely are warranted, despite increased PPM risk, as improved mortality benefit outweighs the harm from PPM implantation. The authors’ study did possess several limitations. In general, their study had a relatively small population with an uncommon event. This does increase the difficulty in
1166
HOYT ET AL
Table 4. Multivariate Exact Logistic Models for Pacemaker Prediction Model
Model A VIV Model B Leak* Model C VIV Leak* Model D Leak† Model E VIV Leak†
OR
95% CI
p Value
5.77
(1.07, 32.07)
0.04
5.05
(0.74, 1)
0.11
7.51 6.03
(1.19, 58.71) (0.83, 1)
0.029 0.082
4.00
(0.80, 19.46)
0.099
6.13‡ 8.51‡
(1.04, 40.09) (1.37, 60.69)
0.045 0.019
NOTE. Analysis includes all 66 patients. Abbreviations: CI, confidence interval; OR, odds ratio; VIV, valve in valve. *Leak is categorized as none or trace versus mild or moderate. †Leak is categorized as none versus trace, mild, or moderate. ‡Indicates I value o 0.01.
identifying differences within the study population. The authors believe that it would be beneficial to continue to build a database to investigate these uncommon outcomes, especially investigating the importance of the degree of paravalvular leak in predicting PPM placement as this may be a significant
contributor to the risk of PPM placement. Since the authors’ study looked at patients in a new TAVR program, there likely will be some change in the rate of complications associated with the procedure. This can be seen in their relatively high rate of PPM implantation in their cohort compared to previous studies. A meta-analysis in 2012 stated a PPM placement rate of 5.9% for the Edwards Sapien valve.15 There is evidence that as teams gain experience placing the valves, they increase their proficiency, suggesting a learning curve to the procedure.16 This, in part, could be due to better placement of the valve with fewer low implantations as experience with the procedure improves. The authors’ high rate of PPM placement was also surprising given that the Edwards Sapient XT is associated with a lower incidence of PPM placement.15 CONCLUSION
PPM implantation is a known complication of TAVR, and the authors’ data suggest that patients with VIV and those with a less significant degree of paravalvular leak are at a higher risk for PPM placement following the procedure. Based on the outcomes of the authors’ study, they believe that this finding supports the placement of temporary transvenous pacemakers in the highest risk group to avoid complications of bradycardia postoperatively.
REFERENCES 1. Carabello BA, Paulus WJ: Aortic stenosis. Lancet 373:956-966, 2009 2. Leon MB, Smith CR, Mack M, et al: Transcatheter aortic-valve implantation for aortic stenosis in patients who cannot undergo surgery. N Engl J Med 363:1597-1607, 2010 3. Makkar RR, Fontana GP, Leon MB, et al: Transcatheter aorticvalve replacement for inoperable severe aortic stenosis. N Engl J Med 366:1697-1704, 2012 4. Binder R, Webb J, Toggweiler S, et al: Impact of postimplantation sapien XT geometry and position on conduction disturbances, hemodynamic performance, and paravalvular regurgitation. JACC Cardiovasc Interv 6:462-468, 2013 5. Piazza N, Onuma Y, Jesserun E, et al: Early and persistent intraventricular conduction abnormalities and requirements for pacemaking after percutaneous replacement of the aortic valve. JACC Cardiovascular Intervention 1:310-316, 2008 6. Piaxxa N, Grube E, Greckens U, et al: Procedural and 30-day outcomes following transcatheter aortic valve implantation using the third generation (18Fr) CoreValve Revalving System: Results from the Multicenter, Expanded Evaluation Registry 1 year following CE mark approval. Euro Intervention 4:242-249, 2008 7. Klein AA, Skubas NJ, Ender J: Contoversies and complications in the perioperative management of transcatheter aortic valve replacement. Anesth Analg 119:784-798, 2014 8. Siontis GCM, Juni P, Pilgrim T, et al: Predictors of permanent pacemaker implantation in patients with severe aortic stenosis undergoing TAVR. J Am Coll Cardiol 64:129-140, 2014 9. Leon MB, Piazza N, Nikolsky, et al: Standardized endpoint definitions for transcatheter aortic valve implantation trials:
A consensus report from the Valve Academic Research Consortium. J Am Coll Cardiol 57:253-269, 2011 10. Siontis GC, Juni P, Pilgrim T, et al: Predictors of permanent pacemaker implantation in patients with severe aortic stenosis undergoing TAVR: A meta-analysis. J Am Coll Cardiol 64:129-140, 2014 11. Khawaja MZ, Rajani R, Cook A, et al: Permanent pacemaker insertion after CoreValve transcatheter aortic valve implantation: Incidence and contributing factors (the UK CoreValve Collaborative). Circulation 123:951-960, 2011 12. Bleiziffer S, Ruge H, Horer J, et al: Predictors for new-onset complete heart block after transcatheter aortic valve implantation. JACC Cardiovascular Intervention 3:524-530, 2013 13. Makkar RR, Jilaihawi H, Chakravarty T, et al: Determinants and outcomes of acute transcatheter valve-in-valve therapy of embolization: A study of multiple valve implants in the U.S. Partner Trial (Placement of AoRTic TraNscathetER Valve Trial Edwards Sapien Transcatheter Heart Valve). Jo Am Coll Cardiol 62:418-430, 2013 14. Kodali SK, Williams MR, Smith CR, et al: PARTNER Trial Investigators. Two-year outcomes after transcatheter or surgical aorticvalve replacement. N Engl J Med 366:1686-1695, 2012 15. Jilaihawi H, Chakravarty, Weiss RE, et al: Meta-analysis of complications in aortic valve replacement: Comparison of MedtronicCorevalve, Edwards-Sapien, and Surgical Aortic Valve Replacement in 8,536 patients. Cath Cardiovasc Interv 80:128-138, 2012 16. Alli O, Booker JD, Lennon RJ, et al: Transcatheter aortic valve implantation assessing the learning curve. JACC Cardiovasc Interv 5: 72-79, 2012