Infective Endocarditis Risk After Percutaneous Pulmonary Valve Implantation With the Melody and Sapien Valves

JACC: CARDIOVASCULAR INTERVENTIONS

VOL. 10, NO. 5, 2017

ª 2017 BY THE AMERICAN COLLEGE OF CARDIOLOGY FOUNDATION

ISSN 1936-8798/$36.00

PUBLISHED BY ELSEVIER

http://dx.doi.org/10.1016/j.jcin.2016.12.012

Infective Endocarditis Risk After Percutaneous Pulmonary Valve Implantation With the Melody and Sapien Valves Sebastien Hascoet, MD,a Lucia Mauri, MD,a Caroline Claude, MD,a Emmanuelle Fournier, MD,a Julie Lourtet, MD,b Jean-Yves Riou, MD,c Philippe Brenot, MD,c Jérôme Petit, MDa

ABSTRACT OBJECTIVES This study compared the risk of infective endocarditis (IE) after percutaneous pulmonary valve implantation (PPVI) with the Sapien and Melody valves. BACKGROUND The incidence of IE after PPVI is estimated at 3% per year with the Melody valve. The Sapien valve is a more recently marketed valve used for PPVI. METHODS We retrospectively included consecutive patients who underwent PPVI at a single center between 2008 and 2016. IE was diagnosed using the modified DUKE criteria. RESULTS PPVI was performed in 79 patients (Melody valve, 40.5%; Sapien valve, 59.5%). Median age was 24.9 years (range 18.1 to 34.6). IE occurred in 8 patients (10.1%) at a median of 1.8 years (minimum: 1.0; maximum: 5.6) after surgery. Causative organisms were methicillin-sensitive Staphylococcus aureus (n ¼ 3), Staphylococcus epidermidis (n ¼ 1), Streptococcus mitis (n ¼ 1), Aerococcus viridans (n ¼ 1), Corynebacterium striatum (n ¼ 1), and Haemophilus influenzae (n ¼ 1). All 8 cases occurred after Melody PPVI (25.0% vs. 0.0%). The incidence of IE was 5.7% (95% confidence interval: 2.9% to 11.4%) per person-year after Melody PPVI. The Kaplan-Meier cumulative incidence of IE with Melody PPVI was 24.0% (95% confidence interval: 12.2% to 43.9%) after 4 years and 30.1% (95% confidence interval: 15.8% to 52.5%) after 6 years, compared with 0.0% with the Sapien PPVI after 4 years (p < 0.04 by log-rank test). There was a trend toward a higher incidence of IE in the first 20 patients with Melody PPVI (who received prophylactic antibiotics during the procedure only) and in patients who had percutaneous interventions, dental care, or noncardiac surgery after PPVI. CONCLUSIONS IE after PPVI may be less common with the Sapien compared with the Melody valve. (J Am Coll Cardiol Intv 2017;10:510–7) © 2017 by the American College of Cardiology Foundation.

P

ercutaneous pulmonary valve implantation

certification in 2006 and Food and Drug Administra-

(PPVI) has emerged as an alternative to sur-

tion approval in 2010 for PPVI. The Sapien valve

gery for reconstructing the right ventricular

(Edwards SAPIEN pulmonic transcatheter heart valve,

outflow tract (RVOT). PPVI was first described in

Edwards Lifesciences, Irvine, California) was used

2000 (1) and since then many studies have supported

initially for transcatheter aortic valve replacement

its efficacy (2–13) The Melody valve (Medtronic Inc.,

and subsequently licensed for PPVI (Europe, 2010;

Minneapolis, Minnesota) was the first valve inserted

and Food and Drug Administration, 2016). However,

percutaneously in humans and received European

although

hemodynamic

outcomes

From the aHospital Marie Lannelongue, Congenital Heart Diseases Department, Complex Congenital Heart Diseases M3C Network, Paris-Sud University, Paris-Saclay University, Plessis-Robinson, France; bHospital Saint-Joseph, Department of Microbiology, Paris, France; and the cHospital Marie Lannelongue, Imaging and Interventional Radiology Department, Paris-Sud University, Paris-Saclay University, Plessis-Robinson, France. The Percutaneous Pulmonary Valve Implantation Program was supported by a grant from the French Health Ministry. The authors have reported that they have no relationships relevant to the contents of this paper to disclose. Manuscript received August 2, 2016; revised manuscript received November 11, 2016, accepted December 15, 2016.

and

device

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

durability are promising, reports of infective endocar-

unidirectional, 3-leaflet valve made of bovine

ABBREVIATIONS

ditis (IE) after PPVI are generating considerable

pericardium and surrounded by a poly-

AND ACRONYMS

concern (14–18). The annual incidence of IE after

ethylene terephthalate skirt. It is available in

PPVI is estimated at 3% with the Melody valve

4 diameters (20, 23, 26, and 29 mm; corre-

(14–18). Little information is available about the risk

sponding heights are 13.5, 14.3, 17.2, and 19.1

of IE after Sapien valve PPVI. Transcatheter aortic

mm, respectively). The Retroflex III delivery

Sapien valve replacement is followed by a 1.0% to

system (Edwards Lifesciences) was used until

2.3% incidence of IE over 1 to 3 years (19).

2015

and

the

Novaflex

delivery

IE = infective endocarditis PPVI = percutaneous pulmonary valve implantation

RVOT = right ventricular outflow tract

system

The primary objective of this study was to compare

(Edwards Lifesciences) thereafter. With Novaflex, the

the incidence of IE after Melody and Sapien PPVI. We

valve is crimped below the balloon. The valve is un-

also sought to identify risk factors for IE.

covered in the inferior vena cava, advanced onto the balloon, and pushed gently toward the landing zone,

SEE PAGE 518

where it is implanted.

METHODS PROCEDURE. Any STUDY

DESIGN. An

observational,

foci of dental infection were

retrospective,

treated routinely before PPVI. One of 3 physicians

single-center cohort design was used. Consecutive

(J.P., J.-Y.R., S.H.) performed the PPVI procedures,

patients who underwent PPVI between 2008 and 2016

under general anesthesia with biplane fluoroscopic

in a French referral center for complex congenital

guidance, via the transfemoral route, using a stan-

heart defects (M3C network, Marie Lannelongue

dardized method, in 1 or 2 stages as most appropriate

Hospital) were included. Patients were considered for

for each individual patient. For patients with a large

PPVI if they had a history of RVOT surgery for

RVOT tract and native valve, the landing zone was

congenital heart disease and met current criteria for

stented to improve its stability and PPVI was per-

pulmonary valve replacement (20). Patients were not

formed a few months later. Other patients had one-

eligible for PPVI if they had any of the following: body

stage PPVI.

weight of <30 kg, pregnancy, active infection, unfavorable

morphology,

coronary

Direct invasive hemodynamic measurements and

artery

angiography were performed routinely before and

compression during balloon testing, or history of a

after valve deployment. Coronary artery compression

failed PPVI. A history of IE was not a contraindication

during balloon inflation in the RVOT was usually

to PPVI.

sought before landing zone stenting. A clinical and

ETHICAL

RVOT

CONSIDERATIONS. The

study database

was reported to the French Data Protection Authority (CNIL, #1837880 v 0, February 26, 2015). The study complied with the Declaration of Helsinki and was approved by our institutional review board (HML 2016). Written informed consent to PPVI and to a follow-up phone interview to determine the incidence of valve-related events and death was obtained from all patients before study inclusion.

laboratory workup was performed to rule out active infection at the time of PPVI. All patients received prophylactic antibiotic therapy (1.5 g intravenous cefazolin bolus or 1 g intravenous vancomycin bolus in patients allergic to penicillin) and anticoagulation (100 IU/kg intravenous heparin) at the beginning of the procedure. The antibiotic and heparin injections were repeated if the PPVI was prolonged. Starting in April 2010, the antibiotic and heparin were continued for 48 hours after PPVI. After PPVI, all patients took

VALVES. Melody and Sapien valves were used. Mel-

oral aspirin (100 mg once daily) for 6 months. Pro-

ody is a bovine jugular vein valve (Contegra Pulmo-

phylactic antibiotic therapy and nonspecific preven-

nary Valved Conduit, Medtronic) sutured into a

tion measures were used as recommended (21).

Cheatham-Platinum stent (NuMed, Hopkinton, Hopkinton, New York). A single size was available at the time of the study. A dedicated balloon-in-balloon (NuMED) catheter delivery system was used for valve implantation (Ensemble Transcatheter Delivery System, Medtronic). The valve was crimped manually over the inner balloon then covered by a retractable sheath. The

511

Endocarditis After PPVI: Melody Versus Sapien

outer

balloon

was

available

in

3

diameters (18, 20, and 22 mm).

DATA ANALYSIS. For each patient who underwent

successful PPVI, we collected demographic characteristics and pre-procedural hemodynamic data. Procedural data, technical details, results, and complications were retrieved from a prospective database of all catheterizations performed at our center. CRITERIA FOR IE. Cases of IE or other infections

were reviewed meticulously and classified as definite,

The Sapien valve is a radiopaque, stainless steel,

possible, or refuted, on the basis of the European

balloon-expandable frame containing an integrated,

Society of Cardiology modified Duke algorithm (21).

Hascoet et al.

512

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

Endocarditis After PPVI: Melody Versus Sapien

T A B L E 1 Patient Demographics, Procedural Data, and Post-Procedural Outcomes

and from phone calls to the patients and to their cardiologists and general practitioners. For cases of

PPVI With Melody Valve (n ¼ 32)

PPVI With Sapien Valve (n ¼ 47)

Standardized Difference

19.9 (15.8–28.9)

26.3 (18.9–39.9)

0.58*

56.5  13.5

65.8  17.6

0.59*

Male (%)

53.1

66.0

0.26

Genetic syndrome (%)

18.8

10.6

-0.23

performed using Stata 11.2 software (StataCorp, Col-

History of severe infectious disease (%)

9.4

8.5

-0.03

lege Station, Texas). Continuous data were described

History of endocarditis (%)

6.3

2.1

-0.20

Pacemaker/defibrillator (%)

6.3

10.6

0.16

Conotruncal malformation

81.3

68.1

Ross procedure

9.4

21.3

were performed to compare variables between the

TGA

3.1

0.0

two valve types and between patients with versus

PA-IVS/PVS

3.1

4.3

without IE during follow-up.

DORV

3.1

6.4

Age (yrs) Weight (kg)

IE, every effort was made to obtain information on the Duke criteria, clinical and microbiological details, medical and surgical strategies, and outcome. STATISTICAL ANALYSIS. Statistical analyses were

as mean  SD if normally distributed and as median

Congenital heart diseases (%)

(interquartile range [IQR]) otherwise. Categorical variables were described as number (%). Bivariate analyses with calculation of standardized differences

Kaplan-Meier curves of the cumulative IE inci-

RVOT (%)

dence were plotted using the date of PPVI as the entry

Native RVOT

3.1

25.5

Bioprosthesis

9.4

23.4

Homograft

25.0

31.9

Conduits

62.5

19.2

Stenosis

84.4

50.0

the log-rank test. The Kaplan-Meier method was also

Regurgitation

0.0

35.7

used to assess the cumulative incidences of pulmo-

Mixed

15.6

14.3

Pre-stenting (%)

90.6

93.6

0.11

1-stage PPVI (%)

87.5

87.2

0.01

0

9.4

6.4

1

90.6

85.1

end of follow-up. Differences in incidence were

2

0.0

6.4

determined using the log-rank test.

4

0.0

2.1

62.1

68.2

date and the time since PPVI as the time scale. The right censor was the date of IE, valve replacement, heart transplantation, death, or follow-up completion. Differences in incidence were evaluated using

RVOT lesion (%)

nary valve replacement and of death, with the date of

No. of stents (%)

PPVI as the entry date and the time since PPVI as the time scale and with the right censor set as the date of valve replacement, heart transplantation, death, or

RESULTS

Type of stent (%) EV3 Intrastent LD MAX Andrastent XXL

3.5

6.8

Cheatham-Platinum stent

34.5

22.7

Sinus XL

0.0

2.3

POPULATION. We included 79 patients, 48 men and

31 women, with a median age of 25.0 years (IQR: 18.1 to 35.0 years) (Table 1). Among them, 3 (3.8%) had a

Valve diameter (%) 18 mm

21.9

20 mm

53.1

22 mm

25.0

history of IE before PPVI (Melody valve: n ¼ 2; and Sapien valve: n ¼ 1). The most common diagnoses

4.2

were conotruncal defect (73.4%) and congenital aortic

23 mm

2.1

valve disease with Ross surgery (16.5%). A right-

26 mm

63.8

ventricle-to-pulmonary-artery tube had been inser-

29 mm

25.5

Valve post-dilation (%) Procedure duration (min) Fluoroscopy time (min) Dose–length product (mGy$m2)

46.9

8.5

-0.94*

83 (71–118)

95 (79–121)

0.21

ted in 36.7% and a homograft or bioprosthesis implanted in 29.1% and 17.7% of patients, respec-

20.4 (17.4–28.8)

30 (22.7–40.9)

0.45

tively. PPVI was performed on the native patched

7,895 (3,573–13,535)

11,647 (7,353–26,749)

0.55

RVOT in 16.5% of patients.

0.0

4.3

-0.30

Age was younger and body weight lower at PPVI in

Severe procedural complications (%)

Continued on the next page

the Melody group than in the Sapien group. The Sapien group had a higher proportion of patients with their native RVOT compared with the Melody group (25.5% and 3.1%, respectively). Implantation dates

Cases of IE within 1 year after PPVI were classified as

were December 2008 to July 2014 for Melody and

early IE and other cases as late IE (21).

December 2011 to May 2016 for Sapien.

OUTCOMES. Outcomes were assessed in June 2016.

PROCEDURES. Most patients underwent 1-stage PPVI

Information was obtained from our local database

(69 of 79; 87.4%) and pre-stenting of the landing zone

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

(73 of 79; 92.4%). A single stent was sufficient in all Melody group patients, whereas in the Sapien group 4

T A B L E 1 Continued

PPVI With Melody Valve (n ¼ 32)

PPVI With Sapien Valve (n ¼ 47)

Standardized Difference

Infective endocarditis during follow-up (%)

25.0

0.0

-0.80*

Pulmonary valve replacement during follow-up (%)

25.0

4.3

-0.59*

patients required 2 to 4 stents. Pre-stenting was not performed in 6 patients with valve-in-valve implantation. Balloon post-dilation was performed in 19 of 79 patients (24.1%) and was more common in the Melody group (46.9% vs. 6.5%, respectively). Procedure duration, fluoroscopy time, and irradiation were higher in the Sapien valve group. Severe procedural complications occurred in 2 pa-

513

Endocarditis After PPVI: Melody Versus Sapien

Percutaneous Surgical Death during follow-up (%)

3.1

2.1

-0.06

21.9

2.1

-0.63*

3.1

2.1

-0.06

tients (2.5%). One patient died of massive hemothorax due to perforation of a distal pulmonary branch during Sapien valve PPVI over a Lunderquist guidewire. In the other patient, who had a mechanical

Values are median [interquartile range] or %. Standardized difference computed as the difference in means or proportions divided by the SE. *Significant imbalance. DORV ¼ double-outlet right ventricle; PA-IVS ¼ pulmonary atresia with intact ventricular septum; PPVI ¼ percutaneous pulmonary valve implantation; PVS ¼ pulmonary valve stenosis; RVOT ¼ right ventricle outflow tract; TGA ¼ transposition of the great arteries.

aortic valve, a large femoral hematoma developed

F I G U R E 1 Kaplan-Meier Cumulative Incidences of Death and Pulmonary Valve Replacement and Kaplan-Meier Cumulative Incidences of Infective Endocarditis

Kaplan-Meier cumulative incidences of death and pulmonary valve replacement among patients with versus without infective endocarditis during follow-up (A) and among patients without infective endocarditis (IE) during follow-up (B). Kaplan-Meier cumulative incidence of IE in patients with the Melody and Sapien valves (C). Kaplan-Meier cumulative incidence of IE in patients with the Melody valve stratified per period of implantation (D) (period 1, before April 2010; period 2, after April 2010). PPVI ¼ percutaneous pulmonary valve implantation; pulm. ¼ pulmonary.

514

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

Endocarditis After PPVI: Melody Versus Sapien

F I G U R E 2 Pictures of Infected Melody Valve Removed Surgically

(A) Melody valve removed surgically 1 week after the onset of infective endocarditis. Complete valve obstruction by vegetations is visible. (B) Melody valve removed surgically 2 years after the onset of infective endocarditis. Cultures of the valve were negative. Severe deterioration of the valve is visible, with budding on the leaflets.

upon switching from heparin to oral anticoagulation

Duke criteria, the IE diagnosis was considered defi-

after PPVI; the outcome was favorable after blood

nite in 5 patients and possible in 3 patients. Never-

transfusions and vasoactive drug therapy.

theless, 2 of the 3 patients with possible IE had

POST-PROCEDURAL OUTCOMES. Median follow-up

was 1.9 years (IQR: 0.5 to 4.1 years) overall and was longer in the Melody group than in the Sapien group (4.9 years [IQR: 2.2 to 6.5 years] vs. 1.0 years [IQR: 0.2 to 2.6 years], respectively). One patient (1.2%), who was a resident of a foreign country, was lost to followup after Melody PPVI. IE occurred in 8 patients (10.1%; 95% confidence interval [CI]: 4.5% to 19.0%). Table 1 compares the post-procedural outcomes in the two groups. The Kaplan-Meier cumulative incidence of pulmonary valve replacement was 15.5% over 40 months, with no difference between groups (log-rank test, p ¼ 0.3). The Kaplan-Meier cumulative incidences of death and pulmonary valve replacement among patients with versus without IE during follow-up were 65.0% (95% CI: 30.7% to 95.1%) and 5.8% (95% CI: 1.9% to 17.2%) 40 months after PPVI (log-rank test; p < 0.0001) (Figure 1A). The Kaplan-Meier cumulative incidences of death and pulmonary valve replacement in patients without IE during follow-up were not different between the Melody and Sapien groups (4.3%; 95% CI: 0.6% to 26.6%; and 8.4%; 95% CI: 2.1% to 30.1%) 40 months after PPVI (Figure 1B).

severe pulmonary valve obstruction (Figure 2A). The pulmonary valve gradient was increased at IE diagnosis in 7 patients. Micro-organisms were identified in 7 patients: methicillin-susceptible Staphylococcus aureus in 3, Aerococcus viridans in 1, Streptococcus mitis in 1, Haemophilus influenzae in 1, and both Staphylococcus

epidermidis

and

Corynebacterium

striatum in 1. One patient died with septic shock on the day after the onset of fever and 4 patients required

elective

surgical

valve

explantation.

Delayed surgical valve explantation was required in 2 patients for severe pulmonary valve dysfunction (Figure 2B). The Online Appendix reports the features of patients with IE. Table 2 compares patients with and without IE. The overall incidence of IE was 3.9% per person-year (95% CI: 1.9% to 7.7%) after PPVI. All 8 patients with IE were in the Melody group; the incidence of IE was 5.7% per person-year (95% CI: 2.9% to 11.4%) in the Melody group compared with 0.0% in the Sapien group (p < 0.05). In the Melody group, the Kaplan-Meier cumulative incidence of IE was 17.0% (95% CI: 7.5% to 36.2%) after 24 months, 24.0% (95% CI: 12.2% to 43.8%) after 40 months, and 29.8% (95% CI: 15.7% to 51.9%) after 72 months (Figure 1C). The

INFECTIVE ENDOCARDITIS (n [ 8). IE occurred in 8

Kaplan-Meier cumulative incidence after 40 months

patients. The median time from PPVI to IE was 1.8

tended to be higher in the patients who had Melody

years (range 12.2 months to 5.6 years). Acute IE onset

valve implantation before versus after April 2010

was observed in 5 cases. According to the revised

(33.3% [95% CI: 16.6% to 59.7%] vs. 8.7%, [95% CI:

515

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

Endocarditis After PPVI: Melody Versus Sapien

1.3% to 49.2%]; p < 0.15; log-rank test) (Figure 1D); the incidence of IE was 7.8% per person-year (95% CI:

T A B L E 2 Case-Control Analysis

3.7% to 16.4%) before and 2.0% per person-year (95% CI: 0.3% to 14.0%) after April 2010 (p ¼ 0.18). Factors associated with trends toward a higher rate

Patients Without Endocarditis During Follow-Up (n ¼ 71)

Standardized Difference

19.6 (16.9–20.7)

25.5 (18.1–35.3)

-0.86

of IE were balloon post-dilation, smaller RVOT, con-

Male

50.0

62.0

-0.23

duits, stenotic lesions, history of IE, antiplatelet or

Genetic syndrome

25.0

12.7

0.31

anticoagulant therapy discontinuation, and percuta-

History of severe infectious disease

12.5

8.5

0.13

neous intervention, dental care, or noncardiac surgery after PPVI.

Age at PPVI (yrs)

Patients With Endocarditis During Follow-Up (n ¼ 8)

History of endocarditis

12.5

2.8

0.35

Pacemaker/implantable cardiac defibrillator

12.5

8.5

0.13

100.0

70.4

0.91

Congenital heart diseases

DISCUSSION

Conotruncal malformation

In this single-center cohort of 79 patients with PPVI,

RVOT Native or patched RVOT

0.0

18.3

Bioprosthesis

12.5

18.3

compared with the Sapien valve. Good procedural

Homograft

25.0

29.6

results have been reported with both valves (2–13,22).

Conduits

62.5

33.8

57.1  12.7

62.6  17.0

IE was more common with the Melody valve

However, IE after Melody valve implantation is an

Weight (kg)

emerging

RVOT lesion

concern

(14–17,23).

The

Sapien

valve

became available for PPVI only very recently (2–5,22) and is now the valve most often used for PPVI at our center, despite limited data on the risk of IE

Stenosis

87.5

62.1

Regurgitation

12.5

28.8

Mixed Implantation before April 2010

compared with the Melody valve. Cases of IE after

Pre-stenting

transaortic Sapien valve implantation have been

One-stage PPVI

reported (19,24–26), as well as a few cases after PPVI

Melody valve

(23,27,28). It has been argued that the apparently

Valve diameter (mm)

lower frequency of IE with the Sapien valve is related

Valve post-dilation

only to the shorter follow-up, due to the more recent

Procedure duration (min)

0.0

9.1

87.5

16.9

1.93*

100.0

91.6

0.43 -0.01

87.5

87.3

100.0

33.8

-1.97*

20 (20–20)

23 (21–23)

-1.37*

62.5

19.7

0.92*

96 (76–134)

90 (75–118)

0.22

28.8 (20.1–76.5)

27.6(21.4–37.1)

0.45

10,885 (7,439–27,562)

10,850 (5,969–24,127)

-0.14

At-risk procedures after PPVI

50.0

21.4

0.60

Right heart catheterization

12.5

11.4

-0.08

Electrophysiological intervention

0.0

5.7

-0.35

Noncardiac surgery

25.0

5.7

Dental or ENT intervention

12.5

5.7

0.23

bovine jugular vein valve sewn into a Cheatham-

Aspirin or anticoagulant agent discontinued

25.0

12.7

-0.31

Platinum stent (20) and the bovine jugular vein is

Outcome

introduction of this valve (23). Our study, however, strongly supports a lower risk of IE with the Sapien valve than with the Melody valve for PPVI. Although the same implantation method was used in both groups, the Sapien group had no cases of IE, whereas the cumulative incidence of IE was up to 24% after 40 months with the Melody valve. The Melody valve is a

Fluoroscopy time (min)

-0.36

Dose–length product (mGy$m2)

0.53*

used also as a surgically implanted conduit (Con-

Pulmonary valve replacement

75.0

5.6

2.60*

tegra). Several recent observational studies suggest

Death

12.5

1.4

0.42

that the risk of IE is similar between the Melody valve and Contegra conduit but higher with these materials than with other surgical implants (15,16,18). More-

Values are median [interquartile range] or %. Standardized difference computed as difference in means or proportions divided by SE. *Significant imbalance. ENT ¼ ear nose throat; other abbreviations as in Table 1.

over, an in vitro study demonstrated greater bacterial adhesion to the bovine jugular vein wall and leaflets compared with the bovine pericardial leaflets used in

No cases of IE occurred in patients with their native

the Sapien valve (29).

RVOT, who may thus be at lower risk. Aspirin

However, the incidence of IE in our Melody group

discontinuation and right heart catheterization were

was higher than previously reported (17). The num-

also associated with a higher risk of IE in a previous

ber of patients with IE was too small to allow the

study (17). Our results support life-long aspirin

identification of risk factors. Nevertheless, trends

therapy after Melody PPVI (17). IE was more common

toward a higher IE risk were found for antith-

early in our experience, with 7 cases among the first

rombotic therapy discontinuation, history of IE,

19 patients and a single case among the remaining

smaller valves, conduits, stenotic lesions, balloon

70 patients. The implantation method was not

post-dilation, noncardiac surgery, and dental care.

changed

between

the

2

periods.

In

contrast,

516

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

Endocarditis After PPVI: Melody Versus Sapien

prophylactic antibiotic therapy was administered

characteristics. They also show that introduction of

only during the procedure until April 2010 and was

the Sapien valve has contributed to extend the

extended to 48 h after the procedure thereafter.

feasibility of PPVI to larger conduits (5). Finally, as

However, the time from PPVI to IE was consistently

demonstrated by the longer fluoroscopy time in our

longer than 1 year. Thus, the relationship between IE

study, Sapien valve implantation is more chal-

and the procedure remains unclear. The expanding

lenging, due to the uncovered and rigid delivery

published data on IE after PPVI may have prompted

system. Neither the Retroflex nor the Novaflex de-

more painstaking selection of patients with no risk

livery systems are designed for PPVI, and the risk of

factors for IE and greater attention among pediatric

tricuspid injury is a matter of concern when per-

cardiologists to compliance with IE prevention rules.

forming the procedure. The Sapien 3 valves and

Finally, the longer follow-up with the Melody valve

Commander delivery system may yield improved

increased the opportunities for IE to occur. Never-

outcomes, given their smaller profile and greater

theless, even since 2010, the incidence of IE was

flexibility. However, the development of a covered

2.0% per person-year with Melody, a value similar to

delivery system is awaited.

previous reports and higher than the 0.3% to 1.2% incidence per person-year reported with prosthetic

CONCLUSIONS

valves in any position (21). Our data confirm the severity of IE after Melody

The incidence of IE after PPVI was lower with the

valve PPVI (17), with 1 of 8 patients dying of septic

Sapien valve compared with the Melody valve. Given

shock

valve

its efficacy and decreased risk of IE, the Sapien valve

explantation. IE modifies the morphology and there-

may deserve to be recommended as the valve of first

fore the durability of the Melody valve (15).

choice for PPVI in patients with a suitable RVOT.

and

4

patients

requiring

elective

STUDY LIMITATIONS. First, the retrospective design

may have biased the collection of periprocedural data and contributed to decrease the capacity for identifying IE risk factors other than valve type. Second, a chronological bias may have occurred, because the Sapien valve was introduced more recently than the Melody valve. Thus, follow-up was longer in the Melody group. Nevertheless, a steady increase in cases of IE occurred over the first 40 months with Melody, contrasting with the complete

Prospective multicenter comparative studies with longer follow-ups are needed to further assess this issue. ADDRESS FOR CORRESPONDENCE: Dr. Sebastien

Hascoet, Hopital Marie Lannelongue, 133 avenue de la résistance,

92350,

Le

Plessis-Robinson,

France.

E-mail: [email protected]. PERSPECTIVES

absence of IE with the Sapien. Third, the populations in the 2 groups were not identical. A history of IE

WHAT IS KNOWN? IE after percutaneous pulmo-

was more common in the Melody group. Sapien

nary Melody valve implantation is a growing concern,

valves—available in diameters of 20, 23, 26, and

with an estimated annual incidence of about 3%.

29 mm—were implanted in conduits as well as in native RVOTs up to 28 mm and Melody valves in conduits up to 24 mm. Thus, patients in the Sapien group were older, had larger regurgitating outflow tracts, and more often had their native RVOT. Melody valve implantation occurred in smaller conduits and more

often required balloon post-dilation.

Trauma to the implanted valve may increase the risk of IE (30). These considerations suggest that the IE risk difference between the 2 valve types may be

WHAT IS NEW? In this retrospective, single-center, comparative study, IE after PPVI was less common with the Sapien valve, recently approved for this procedure, than with the Melody valve. WHAT IS NEXT? Our results need to be tested in a prospective, multicenter clinical trial but already suggest a need for caution regarding use of the Melody valve.

related in part to differences in outflow tract

REFERENCES 1. Bonhoeffer P, Boudjemline Y, Saliba Z, et al. Percu-

2. Haas NA, Moysich A, Neudorf U, et al. Percu-

3. Kenny D, Hijazi ZM, Kar S, et al. Percutaneous

taneous replacement of pulmonary valve in a rightventricle to pulmonary-artery prosthetic conduit with valve dysfunction. Lancet 2000;356:1403–5.

taneous implantation of the Edwards SAPIEN(TM) pulmonic valve: initial results in the first 22 patients. Clin Res Cardiol 2013;102:119–28.

implantation of the Edwards SAPIEN transcatheter heart valve for conduit failure in the pulmonary position: early phase 1 results from an

Hascoet et al.

JACC: CARDIOVASCULAR INTERVENTIONS VOL. 10, NO. 5, 2017 MARCH 13, 2017:510–7

international multicenter clinical trial. J Am Coll Cardiol 2011;58:2248–56. 4. Odemis E, Guzeltas A, Saygi M, Ozyilmaz I, Momenah T, Bakir I. Percutaneous pulmonary valve implantation using Edwards SAPIEN transcatheter heart valve in different types of conduits: initial results of a single center experience. Congenit Heart Dis 2013;8:411–7. 5. Wilson WM, Benson LN, Osten MD, Shah A, Horlick EM. Transcatheter pulmonary valve replacement with the Edwards Sapien System: the Toronto experience. J Am Coll Cardiol Intv 2015;8:1819–27.

Endocarditis After PPVI: Melody Versus Sapien

14. McElhinney DB, Benson LN, Eicken A, Kreutzer J, Padera RF, Zahn EM. Infective endocarditis after transcatheter pulmonary valve replacement using the Melody valve: combined results of 3 prospective North American and European studies. Circulation Cardiovasc Interv 2013;6:292–300. 15. Malekzadeh-Milani S, Ladouceur M, Iserin L, Bonnet D, Boudjemline Y. Incidence and outcomes of right-sided endocarditis in patients with congenital heart disease after surgical or transcatheter pulmonary valve implantation. J Thorac Cardiovasc Surg 2014;148:2253–9.

valve endocarditis: incidence, prevention and management. Arch Cardiovasc Dis 2014;107:615–24. 24. Puls M, Eiffert H, Hunlich M, et al. Prosthetic valve endocarditis after transcatheter aortic valve implantation: the incidence in a single-centre cohort and reflections on clinical, echocardiographic and prognostic features. EuroIntervention 2013;8:1407–18. 25. Spartera M, Schiavo Lena M, Sanvito F, Colombo A. Early degeneration and endocarditis in a trans-catheter heart valve. Eur Heart J 2016;37: 2289.

6. Fraisse A, Aldebert P, Malekzadeh-Milani S, et al. Melody (R) transcatheter pulmonary valve implantation: results from a French registry. Arch Cardiovasc Dis 2014;107:607–14.

16. Van Dijck I, Budts W, Cools B, et al. Infective endocarditis of a transcatheter pulmonary valve in comparison with surgical implants. Heart 2015; 101:788–93.

26. Amat-Santos IJ, Messika-Zeitoun D, Eltchaninoff H, et al. Infective endocarditis after transcatheter aortic valve implantation: results from a large multicenter registry. Circulation 2015;131:1566–74.

7. Khambadkone S, Coats L, Taylor A, et al.

17. Malekzadeh-Milani S, Ladouceur M, Patel M, et al. Incidence and predictors of Melody valve endocarditis: a prospective study. Arch Cardiovasc Dis 2015;108:97–106.

27. Vollroth M, Daehnert I, Kostelka M, Wagner R.

Percutaneous pulmonary valve implantation in humans: results in 59 consecutive patients. Circulation 2005;112:1189–97. 8. Lurz P, Coats L, Khambadkone S, et al. Percutaneous pulmonary valve implantation: impact of evolving technology and learning curve on clinical outcome. Circulation 2008;117:1964–72.

18. O’Donnell C, Holloway R, Tilton E, Stirling J, Finucane K, Wilson N. Infective endocarditis following Melody valve implantation: comparison with a surgical cohort. Cardiol Young 2016:1–8.

9. Zahn EM, Hellenbrand WE, Lock JE, McElhinney DB. Implantation of the melody transcatheter pulmonary valve in patients with a

19. Eisen A, Shapira Y, Sagie A, Kornowski R. Infective endocarditis in the transcatheter aortic valve replacement era: comprehensive review of a rare complication. Clin Cardiol 2012;35:E1–5.

dysfunctional right ventricular outflow tract conduit early results from the U.S. clinical trial. J Am Coll Cardiol 2009;54:1722–9. 10. Asoh K, Walsh M, Hickey E, et al. Percutaneous pulmonary valve implantation within bioprosthetic valves. Eur Heart J 2010;31:1404–9. 11. McElhinney DB, Hellenbrand WE, Zahn EM, et al. Short- and medium-term outcomes after transcatheter pulmonary valve placement in the expanded multicenter US Melody valve trial. Circulation 2010;122:507–16. 12. Vezmar M, Chaturvedi R, Lee KJ, et al. Percutaneous pulmonary valve implantation in the young 2-year follow-up. J Am Coll Cardiol Intv 2010;3:439–48. 13. Eicken A, Ewert P, Hager A, et al. Percutaneous pulmonary valve implantation: two-centre experience with more than 100 patients. Eur Heart J 2011;32:1260–5.

20. Hascoet S, Acar P, Boudjemline Y. Transcatheter pulmonary valvulation: current indications and available devices. Arch Cardiovasc Dis 2014;107:625–34. 21. Habib G, Lancellotti P, Antunes MJ, et al. 2015 ESC Guidelines for the management of infective endocarditis: The Task Force for the Management of Infective Endocarditis of the European Society of Cardiology (ESC). Endorsed by: European Association for Cardio-Thoracic Surgery (EACTS), the European Association of Nuclear Medicine (EANM). Eur Heart J 2015;36:3075–128. 22. Boone RH, Webb JG, Horlick E, et al. Transcatheter pulmonary valve implantation using the

First case of blood-culture proven Staphylococcus aureus endocarditis of a Sapien XT valve after percutaneous pulmonary valve implantation. Eur J Cardiothorac Surg 2015;48: e124–5. 28. Hascoet S, Jalal Z, Baruteau A, et al. Stenting in paediatric and adult congenital heart diseases: A French multicentre study in the current era. Arch Cardiovasc Dis 2015;108: 650–60. 29. Jalal Z, Galmiche L, Lebeaux D, et al. Selective propensity of bovine jugular vein material to bacterial adhesions: An in-vitro study. Int J Cardiol 2015;198:201–5. 30. Jalal Z, Galmiche L, Beloin C, Boudjemline Y. Impact of percutaneous pulmonary valve implantation procedural steps on leaflets histology and mechanical behaviour: an in vitro study. Arch Cardiovasc Dis 2016;109:465–75.

KEY WORDS cardiac valve, congenital heart defects, infective endocarditis, Melody valve, percutaneous pulmonary valve implantation, Sapien valve

Edwards SAPIEN transcatheter heart valve. Catheter Cardiovasc Interv 2010;75:286–94. 23. Patel M, Malekzadeh-Milani S, Ladouceur M, Iserin L, Boudjemline Y. Percutaneous pulmonary

A PP END IX For supplemental materials, please see the online version of this article.

517