Outcomes of Transcatheter Tricuspid Valve-in-Valve Implantation in Patients With Ebstein Anomaly

ARTICLE IN PRESS Outcomes of Transcatheter Tricuspid Valve-in-Valve Implantation in Patients With Ebstein Anomaly Nathaniel W. Taggart, MDa,*, Allison K. Cabalka, MDa, Andreas Eicken, MD, PhDb, Jamil A. Aboulhosn, MDc, John D.R. Thomson, MDd, Brian Whisenant, MDe, Martin L. Bocks, MDf, Stephan Schubert, MDg, Thomas K. Jones, MDh, Jeremy D. Asnes, MDi, Thomas E. Fagan, MDj, Jeffery Meadows, MDk, Mark Hoyer, MDl, Mary H. Martin, MDm, Frank F. Ing, MDn, Daniel R. Turner, MDo, Azeem Latib, MBBChp, Aphrodite Tzifa, MDq, Stephan Windecker, MDr, Bryan H. Goldstein, MDs, Jeffrey W. Delaney, MDt, James A. Kuo, MDu, Susan Foerster, MDv, Matthew Gillespie, MDw, Gianfranco Butera, MD, PhDx, Shabana Shahanavaz, MDy, Eric Horlick, MDz, Younes Boudjemline, MD, PhDaa, Daniel Dvir, MDbb, and Doff B. McElhinney, MDcc for the VIVID Registry We sought to describe the acute results and short- to medium-term durability of transcatheter tricuspid valve-in-valve (TVIV) implantation within surgical bioprostheses among patients with Ebstein anomaly (EA). Cases were identified from a voluntary, multicenter, international registry of 29 institutions that perform TVIV. Demographic, clinical, procedural, and follow-up data were analyzed. Eighty-one patients with EA underwent TVIV from 2008 to 2016. Thirty-four patients (42%) were New York Heart Association (NYHA) class 3/4 at time of TVIV. The most common indication for TVIV was the presence of moderate or severe tricuspid regurgitation (40%). Most patients received a Melody valve (64%). TVIV was ultimately successful in all patients, and there was no procedural mortality. Four patients (5%) developed acute valve thrombosis, 4 patients (5%) developed endocarditis, and 9 patients (11%) developed valve dysfunction not related to thrombosis or endocarditis. Eight patients (10%) underwent reintervention (2 transcatheter, 6 surgical) due to thrombosis (3), endocarditis (2), other valve dysfunction (2), and patient-prosthesis mismatch without valve dysfunction (1). Among 69 patients who were alive without reintervention at latest follow-up, 96% of those with NYHA status reported were class 1/2, a significant improvement from baseline (62% NYHA class 1/2, p <0.001). In conclusion, transcatheter TVIV offers a low-risk, minimally invasive alternative to surgical tricuspid valve rereplacement in patients with EA and a failing tricuspid valve bioprosthesis. © 2017 Elsevier Inc. All rights reserved. (Am J Cardiol 2017;■■:■■–■■)

a Department of Pediatrics and Adolescent Medicine, Division of Pediatric Cardiology, Mayo Clinic, Rochester, Minnesota; bDivision of Pediatric Cardiology and Congenital Heart Disease, German Heart Center Munich, Munich, Germany; cDivision of Cardiology, University of California-Los Angeles, Los Angeles, California; dDivision of Pediatric and Adult Congenital Cardiology, Leeds General Infirmary, Leeds, United Kingdom; eIntermountain Heart Institute, Murray, Utah; fDepartment of Pediatrics, Division of Pediatric Cardiology, UH Rainbow Babies and Children’s Hospital, Cleveland, Ohio; gDivision of Pediatric Cardiology, Deutsches Herzzentrum Berlin, Berlin, Germany; hDivision of Pediatric Cardiology, Seattle Children’s Hospital, Seattle, Washington; iSection of Pediatric Cardiology, Yale University, New Haven, Connecticut; jDivision of Pediatric Cardiology, University of Tennessee Health Science Center, Memphis, Tennessee; kDivision of Pediatric Cardiology University of California-San Francisco Medical Center, San Francisco, California; l Division of Pediatric Cardiology, Riley Hospital for Children, Indianapolis, Indianapolis; mDivision of Pediatric Cardiology, Primary Children’s Hospital, Salt Lake City, Utah; nDivision of Cardiology, Children’s Hospital of Los Angeles, Los Angeles, California; oDivision of Pediatric Cardiology, Children’s Hospital of Michigan, Detroit, Michigan; pCardothoracic and Vascular Department, San Raffaele Hospital, Milan, Italy; qDepartment of Pediatric Cardiology, Mitera Children’s Hospital, Athens, Greece; rDepartment of

0002-9149/© 2017 Elsevier Inc. All rights reserved. https://doi.org/10.1016/j.amjcard.2017.10.017

Cardiovascular Diseases, Swiss Cardiovascular Center, Bern, Switzerland; sDivision of Pediatric Cardiology, Cincinnati Children’s Hospital, Cincinnati, Ohio; tDivision of Cardiology, Children’s Hospital and Medical Center of Omaha, Omaha, Nebraska; uDivision of Cardiology, Cook Children’s Medical Center, Ft. Worth, Texas; vHerma Heart Center, Children’s Hospital of Wisconsin, Milwaukee, Wisconsin; wCardiac Center, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania; xDivision of Pediatric and Adult Congenital Cardiology, IRCCS Policlinico San Donato, Milan, Italy; yDivision of Pediatric Cardiology, St. Louis Children’s Hospital, St. Louis, Missouri; zPeter Munk Cardiac Centre, Toronto General Hospital, Toronto, Canada; aaDivision of Pediatric Cardiology, Necker Enfants Malades Hospital, Paris, France; bbDivision of Pediatric Cardiology, St. Paul’s Hospital, Vancouver, Canada; and ccDivision of Pediatric Cardiac Surgery, Division of Pediatric Cardiology, Stanford University, Palo Alto, California. Manuscript received August 14, 2017; revised manuscript received and accepted October 4, 2017. See page •• for disclosure information. Funding sources: The authors report no external funding related to the publication of this manuscript. *Corresponding author: Tel: (507) 266 0676; fax (507) 284 3968. E-mail address: [email protected] (N.W. Taggart). www.ajconline.org

ARTICLE IN PRESS 2

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

Although surgical valve repair is the preferred treatment for tricuspid regurgitation (TR) in patients with Ebstein anomaly (EA), a significant proportion of patients may undergo tricuspid valve replacement (TVR) as primary therapy or following initial repair.1 Bioprosthetic valves are typically favored in the tricuspid valve (TV) position, particularly in patients with EA who have depressed right ventricular (RV) function.2 Over time, bioprosthetic valves develop dysfunction and may again require replacement.2–6 Historically, surgical valve rereplacement has been the only therapeutic option in such situations. Transcatheter tricuspid valve-in-valve (TVIV) implantation for surgical bioprosthesis dysfunction has since been described using available transcatheter valves in an offlabel fashion.7–14 Recently, the Valve-in-Valve International Database (VIVID) reported excellent procedural results and short- to medium-term outcomes of patients who underwent attempted TVIV.15 However, there was significant diversity relating to the type and severity of underlying heart disease in this patient cohort. Although TV dysfunction is often the most striking feature in EA patients, EA also affects RV systolic and diastolic function. Thus, EA presents unique considerations relative to other causes of TV pathology, and the durability of surgical TVR or transcatheter TVIV may vary in these patients. For these reasons, examination of patients with EA who have undergone TVIV may provide more nuanced insight than a heterogeneous cohort. Methods The TVIV registry is a voluntary international registry. Operators were originally solicited through the VIVID Registry or invitation to those performing transcatheter valve procedures. For this study, we confined the data analysis to patients with an underlying diagnosis of EA and a preexisting TV bioprosthesis in whom TVIV implant was performed. Patientspecific information collected was summarized in a previous report.15 The institutional review board or ethics committee of participating institutions approved the submission of data. TV function was assessed before and after TVIV by echocardiography, and, if reported, by invasive hemodynamic measurements, according to methods described previously15 and in the Supplemental Material. TVIV was performed at the discretion of the implanting physician using either a Melody (Medtronic, Inc, Minneapolis, Minnesota) or Sapien (Edwards Lifesciences, Irvine, California) valve. Procedural variables collected are summarized in the Supplemental Material. Procedural success was defined as a transcatheter valve implanted in the intended (TV) position. Post-TVIV antiplatelet/anticoagulation treatment at time of hospital discharge was reported. Treatment was nonstandardized and directed by the implanting cardiologist. Follow-up data collected included death, cause of death, reintervention, endocarditis, valve thrombosis, or other valve dysfunction. Primary outcomes assessed were TVIV reintervention, dysfunction, and other valve-related adverse outcomes. Secondary outcomes included technical success, acute hemodynamic results, adverse events, New York Heart Association (NYHA) class, TV function over time, and survival. Time-dependent outcomes were reported using Kaplan-Meier curves and correlated with potential risk factors using Cox regression analysis.

Paired data were compared using a paired t test, Wilcoxon signed-rank test, McNemar test, or Fisher’s exact test. Data were compared between age groups using a Kruskal-Wallis or Cochran-Armitage trend test. Nominal and ordinal data are presented as frequency (%); continuous data are presented as median (interquartile range) or as mean ± standard deviation. p Values <0.05 (2-sided) were considered significant. Statistical analysis was performed using JMP Pro 12.2.0 (SAS, Cary, North Carolina). Results Eighty-one patients with EA underwent transcatheter TVIV implantation at 29 centers between 2008 and 2016. Patients were separated into 3 age groups: <19, 19 to 40, and >40 years old. Demographics and preimplantation clinical characteristics are listed in Table 1. Stated bioprosthetic valve size ranged from 20 to 35 mm. There was no difference in baseline mean echocardiographic TV gradient according to surgical prosthesis size. Procedural data are listed in Table 2. Younger patients were more likely to receive a Melody valve (p = 0.007). One patient initially had moderate TR due to distal implant of the first Sapien valve, so a second valve was implanted during the same procedure. There were no other malpositioned or embolized valves. One adult patient developed a retroperitoneal hematoma that did not require treatment. There were no other serious adverse events and no procedural mortalities. Both catheter-measured (7 vs. 2 mmHg) and Doppler estimates of mean TV inflow gradient (8 vs. 3 mmHg) decreased significantly after TVIV (all p <0.001). All patients had mild or less TR immediately following implant. Fifteen patients (19%) had a bidirectional cavopulmonary anastomosis (BCPA) reducing TV inflow. Patients with BCPA had smaller TV bioprostheses than those without [median 25 (25 to 29) vs. 29 (27 to 33) mm; p = 0.003] and were more likely to receive a Melody valve (93% vs. 58%; p = 0.008); only 1 patient with BCPA received a Sapien valve. There were no differences in age (mean 24.5 vs. 30.5 years, p = 0.25), weight (61.5 vs. 67.5 kg, p = 0.36), and pre- or early postTVIV Doppler or catheter-derived TV mean gradient between patients who had a BCPA and those who did not. Postprocedure anticoagulation strategies are listed in Table 2. There was no difference in anticoagulation use between those who received a Melody valve and those who received a Sapien valve (38% vs. 45%, p = 0.64). Patients with a history of atrial fibrillation/flutter were no more likely to be treated with anticoagulation than those with no such history (p = 0.11). During a median follow-up of 13 months (5 to 27 months), there were 4 deaths (Figure 1). Two patients, both hospitalized with right heart failure, renal insufficiency, and ascites before TVIV, died of progressive decline and multiorgan failure 0.3 and 15 months post-TVIV. One died of a ruptured cerebral aneurysm 2.6 months post-TVIV. All 3 of these patients were in the oldest cohort (age 49 to 77 years) and were inpatients in NYHA class 3 or 4 before intervention. A fourth patient who was 11 years old at time of TVIV died suddenly 40 months later. Autopsy was not performed, but there was a history of intra-atrial reentrant tachycardia treated with ablation. Although the small number of deaths precluded robust

ARTICLE IN PRESS Congenital Heart Disease/Transcatheter Tricuspid Valve-in-Valve in Ebstein Anomaly

3

Table 1 Baseline patient characteristics Variable

Male Age (years) Weight (kg) BCPA TV prosthesis ≥ 29 mm Time TV bioprosthesis in place (years) Prior CV operations PPM Transvenous Epicardial NYHA Class 3/4 (n = 80) Prior endocarditis Any comorbidity* Baseline TV mean echo gradient (mmHg) Baseline TS ≥10 mmHg Baseline moderate/severe TR

Total

37 (46%) 23 (16–40) 67 (52–82) 15 (19%) 39 (51%) 8.6 (5.7–13.4) 2 (1–3) 30 (37%) 12 (15%) 18 (22%) 34 (42%) 1 (1%) 25 (30%) 8 (6–10) 25 (32%) 61 (75%)

Ages (Years)

p-value

<19 (n = 28)

19–40 (n = 33)

>40 (n = 20)

14 (50%) 14 (10–17) 50.5 (33.0–65.5) 8 (29%) 4 (15%) 7.1 (5.2–9.0) 2 (1–3) 9 (32%) 2 (7%) 7 (25%) 7 (26%) 0 (0%) 3 (11%) 8 (6–11) 11 (41%) 25 (89%)

17 (52%) 26 (22–32) 71.5 (55.8–83.5) 5 (15%) 21 (66%) 12.5 (6.1–14.7) 2 (1–3) 11 (33%) 5 (15%) 6 (18%) 13 (39%) 1 (3%) 10 (30%) 7.5 (6–10) 9 (28%) 24 (73%)

7 (35%) 52 (46–68) 76.1 (64.1–92.8) 2 (10%) 14 (74%) 9.3 (6.8–19.0) 2 (1–3) 10 (50%) 5 (25%) 5 (25%) 14 (70%) 0 (0%) 11 (55%) 8 (5–10) 5 (25%) 12 (60%)

0.33 − <0.001 0.09 <0.001 0.01 0.50 0.23 0.09 0.94 0.003 0.90 0.001 0.83 0.23 0.019

BCPA = bidirectional cavopulmonary anastomosis; TV = tricuspid valve; CV = cardiovascular; PPM = permanent pacemaker; NYHA = New York Heart Association; TS = tricuspid stenosis; TR = tricuspid regurgitation. * Includes history of atrial fibrillation/flutter, coronary artery disease, renal failure, chronic obstructive pulmonary disease, or liver disease. Continuous variables are reported as median (IQR).

Table 2 Procedural/discharge data Variable

General Anesthesia Vascular approach Femoral Internal jugular Transthoracic/RA TVIV Type Melody Sapien Additional procedures (N = 9) EP study/ablation TPVI Closure left SVC to LA Hospital stay, days† Anticoagulation Antiplatelet only Warfarin/NOAC only Antiplatelet + warfarin/NOAC Other

Total

Ages (Years)

p-value*

<19 (n = 28)

19–40 (n = 33)

>40 (n = 20)

75 (93%)

27 (96%)

31 (94%)

17 (85%)

0.15

68 (84%) 12 (15%) 1 (1%)

24 (86%) 4 (14%) 0 (0%)

27 (82%) 6 (18%) 0 (0%)

17 (85%) 2 (10%) 1 (5%)

0.91 0.74 0.15

52 (64%) 29 (36%)

22 (79%) 6 (21%)

22 (67%) 11 (33%)

8 (40%) 12 (60%)

0.007 —

1 (4%) 1 (4%) 1 (4%) 1 (1–4)

3 (9%) 2 (6%) 0 (0%) 1 (0–6)

1 (5%) 0 (0%) 0 (0%) 4 (1–33)

0.77 0.59 0.23 <0.001

21 (75%) 1 (4%) 6 (21%) 0 (0%)

19 (58%) 3 (9%) 10 (30%) 1 (3%)

7 (35%) 5 (25%) 8 (40%) 0 (0%)

0.006 0.024 0.16 0.90

5 (6%) 3 (4%) 1 (1%) 1 (0–33) 47 (58%) 9 (11%) 24 (30%) 1 (1%)

RA = right atrium; TVIV = tricuspid valve-in-valve; EP = electrophysiology; TPVI = transcatheter pulmonary valve implantation; SVC = superior vena cava; LA = left atrium; NOAC = novel oral anti-coagulant. * Three-way comparison between age groups. † Reported as median (min-max).

statistical identification of risk factors, none of these patients had evidence of valve dysfunction at the most recent evaluation before death. Eight patients underwent reintervention—balloon valvuloplasty of the TVIV (n = 1), repeat TVIV (n = 1), or surgical TVR (n = 6)—for valve dysfunction related to thrombosis (n = 3, see below), endocarditis (n = 2), recurrent TR or TR/

tricuspid stenosis (TS) (n = 2), or a TVR bioprosthesis that was believed to be too small for the patient’s size without clear valve dysfunction (n = 1). Four patients (2 Melody, 2 Sapien) who underwent TVIV at a median age of 38 years (10 to 50 years), developed proven or suspected TVIV thrombosis, all within the first 1.5 months after TVIV. Three of these patients had significant RV

ARTICLE IN PRESS 4

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

Figure 1. This flow diagram describes the outcomes of 81 patients with Ebstein anomaly after transcatheter tricuspid valve-in-valve (TVIV). TS, tricuspid stenosis; TR, tricuspid regurgitation.

Figure 2. Kaplan-Meier survival curve defining freedom from reintervention or valve dysfunction after tricuspid valve-in-valve (TVIV).

dysfunction; however, data regarding RV dysfunction were not systematically reported, limiting our ability to analyze it as a risk factor for thrombosis. The fourth patient had undergone implant of 2 Sapien valves at the initial procedure (see previous discussion). Details are provided in the Supplemental Material. Four patients were diagnosed with endocarditis 0 to 14 months after TVIV (1 Sapien, 3 Melody) with cultures positive for Staphylococcus aureus (n = 2), Enterococcus faecalis (n = 1), or Candida albicans (n = 1). Two of these patients were treated with surgical TVR (see below), 1 improved with

medical treatment alone, and 1 had been diagnosed just before database closure for analysis. Freedom from reintervention or valve dysfunction at latest follow-up is shown in Figure 2. Among patients who were alive, reintervention-free, and did not have valve dysfunction (n = 62), the mean Doppler inflow gradient at most recent follow-up was a median of 4 mmHg,3–6 which is lower than pre-TVIV (p <0.001). Relative to the immediate postimplant Doppler gradient (median 3 mmHg2–4), there was a small increase at most recent follow-up (p = 0.007). All but 2 of the patients who were alive and reintervention-free were in NYHA

ARTICLE IN PRESS Congenital Heart Disease/Transcatheter Tricuspid Valve-in-Valve in Ebstein Anomaly

5

Table 3 Potential risk factors for valve dysfunction or reintervention Variable Male Age 19–40 years BCPA TV prosthesis <29 mm* Melody Valve Antiplatelet only

Dysfunction or Reintervention (n = 15)

No Dysfunction or Reintervention (n = 66)

Odds Ratio

p-value

7 (47%) 12 (80%) 1 (7%) 6 (40%)

30 (45%) 21 (32%) 14 (21%) 32 (52%)

1.06 8.02 0.16 1.06

0.91 0.001 0.057 0.92

11 (73%) 10 (67%)

41 (62%) 37 (56%)

1.12 1.47

0.86 0.49

BCPA = bidirectional cavopulmonary anastomosis; TV = tricuspid valve. * Valve sizes were not known or reported for 4 patients

class 1 or 2, an improvement over baseline (96% vs. 62%, p <0.001). Nine patients, all young adults (22 to 38 years old), developed valve dysfunction (4 TR, 4 TS, 1 mixed TR/TS) not attributed to acute thrombosis or endocarditis, diagnosed in the first month after TVIV in 4 patients and within 1 year of TVIV in 8. Two of those patients underwent reintervention (1 surgical TVR, 1 repeat TVIV), and both developed early dysfunction of the subsequent transcatheter or surgical valve. Patients 19 to 40 years old had shorter freedom from reintervention or valve dysfunction (p = 0.001, Table 3). Discussion In our cohort of 81 patients with EA, we observed excellent acute outcomes of TVIV with no procedural mortality or acute complications requiring surgery. Mean TV gradient and degree of TR improved significantly. This improvement was generally maintained over the duration of follow-up. We observed significant improvement in functional status among patients who were symptomatic before TVIV. Before TVIV, older patients had more severe functional limitation, likely related to pre-TVIV co-morbidities. Patients in the youngest group were less likely to have significant TS before TVIV than older patients. This finding may be partly explained by selection bias in that patients who are not full-grown may be more likely to be referred for surgical TVR rather than TVIV. Notably, none of the 9 patients who developed nonthrombotic valve dysfunction were in the youngest group, suggesting that even growing pediatric patients may experience sustained benefit from TVIV within appropriately sized surgical prostheses. Similarly, although patients with a BCPA generally had smaller surgical bioprosthetic TVs, there was no difference in TV gradient or increased risk of valve thrombosis or dysfunction after TVIV. Thus, initial TVIV function and short to medium-term durability appear to be unaffected by the smaller TV size in this small subgroup of patients. Overall mortality during follow-up was low and unrelated to the TVIV procedure or valve-related complications. With only 4 deaths during follow-up, we are unable to draw any conclusions about risk factors for mortality. Transcatheter valve thrombosis is a recognized but uncommon complication in patients who underwent transcatheter aortic valve replacement.16 Thrombosis has also been reported in patients with transcatheter valves in the pulmonary

position, although the true incidence is not known.17 In our cohort, 4 patients developed acute TV thrombosis, all early after TVIV. By comparison, the earlier VIVID database report of TVIV reported thrombosis in only 2 of 151 patients (1%) who underwent TVIV.15 The fact that those 2 patients were included in our 4 patients with TV thrombosis suggests that patients with EA may have a greater risk of TV thrombosis after TVIV than those without EA and highlights the importance of monitoring for valve dysfunction in this patient population. Thrombus occurred in patients with both Melody and Sapien valves, suggesting that this issue is not a function of TVIV valve type. Three of the 4 had RV dysfunction, and we speculate that right atrium and RV enlargement and abnormal RV diastolic hemodynamics may predispose EA patients to acute valve thrombosis. However, we did not have consistently reported, quantifiable data on RV size or function for the entire cohort, so it was not possible to determine whether there were RV factors that predisposed to thrombus. There are no formal recommendations for thromboprophylaxis after TVIV. One study estimated the risk of bioprosthetic valve thrombosis in the TV position to be higher than in other positions.18 The risk of thrombosis after TVIV is probably even higher than after TVR.15 Published guidelines recommend short-term (3 to 6 month) anticoagulation immediately after bioprosthetic aortic or mitral valve replacement and aspirin (75 to 100 mg daily) indefinitely thereafter.19 Considering the potential for poor TV inflow hemodynamics due to RV dysfunction in patients with EA, thromboprophylaxis with both anticoagulant and antiplatelet therapy for at least the first 3 to 6 months after TVIV may be reasonable. We speculate that anticoagulation rather than antiplatelet therapy alone after TVIV may improve valve durability in patients with EA. Early valve dysfunction or a sudden increase in gradient should raise concerns for thrombosis and prompt more aggressive anticoagulant therapy. Nonthrombotic dysfunction of transcatheter TVIV was previously described in a small case series20—all of the patients in that series were also included in the present registry cohort. For comparison, a study of surgical TVR reported 89% survival free from reoperation at 5 years after placement of a bioprosthetic valve among a population of patients with EA.2 Based upon our limited follow-up data, survival free from reintervention after TVIV may fall short of that reported for surgical TVR. However, this is not a straightforward comparison. A number of the patients in our cohort were considered

ARTICLE IN PRESS 6

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

high risk for surgical TVR, so the case mixes may not be similar. Because transcatheter TVIV is likely associated with lower risk and morbidity than surgical TVR, which carries a reported mortality of 5% to 10%,21 the risk-benefit equation is different. Because the number of events in this study was small, and the causes of valve dysfunction were not always obvious, it is unclear whether there were preventable causes, suboptimal patient selection, or other factors. In addition, the aforementioned surgical report included patients who underwent first-time TVR, whereas our patients by definition had undergone at least 1 surgical TVR, which then became dysfunctional before TVIV. To this point, the study cited previously reported 89% 5-year survival free of reoperation among early survivors of bioprosthetic TVR, whereas 20% of our patients had undergone surgical TVR within 5 years before referral for TVIV, suggesting either a bias toward earlier intervention or an intrinsic predisposition to early surgical or transcatheter valve dysfunction in a portion of our patient population. Endocarditis is a known risk of surgical and transcatheter valve replacement in all locations. Four patients in our cohort developed endocarditis after TVIV. Although the overall number of patients and the number of endocarditis cases are too few to perform robust analysis, these observations indicate that it will be just as important to evaluate for and counsel about the risk of endocarditis after TVIV as after other valve replacement procedures. As with thrombus, the occurrence in patients with both Melody and Sapien valves indicates that the risk of this outcome is not valve-specific. This study is limited by its noncontrolled, single-cohort registry design. Institutions participating in the registry were self-selected. Although echocardiographic data were provided by submitting centers, images were not accessible and thus not reviewed by a core laboratory for accuracy or interinstitutional consistency. Our reported cumulative experience may not reflect that of the entire interventional community. We recognize that there are factors related to transcatheter valve selection that were not captured by our study. We also acknowledge that follow-up is short to medium-term, so we are unable to draw conclusions about long-term outcomes. In conclusion, transcatheter TVIV offers a low-risk, minimally invasive alternative to surgical TV re-replacement in patients with EA and a failing TV bioprosthesis. There may be a higher risk of early valve thrombosis than with surgical TVR or with TVIV in patients who do not have EA, so consideration of aggressive thromboprophylaxis and ongoing surveillance of valve function may be warranted. Acknowledgment: The authors would like to thank Dr. Kent R. Bailey, PhD, for his statistical review of their manuscript. Disclosures Consultant/Proctor: Medtronic (Minneapolis, Minnesota): YB, DD, MG, EH, TKJ, AL, and SS; Edwards Lifesciences (Irvine, California): JAA, JDA, MLB, DD, BHG, EH, and BW; W. L. Gore (Scottsdale, Arizona): BHG, MH, and JDRT; St. Jude Medical/Abbott (Abbott Park, Illinois): BHG, MH, and JDRT; Boston Scientific (Marlborough, Massachusetts): BW and SW; Philips Medical Systems (Bothell, Washington): BHG; B Braun Medical (Bethlehem, Pennsylvania): BHG.

Grant/Research Funding: Edwards Lifesciences: SW and MLB; Medtronic: SW and TKJ; Biotronik (Lake Oswego, Oregon): SW. Honoraria: Medtronic: JAA. Supplementary Data Supplementary data associated with this article can be found, in the online version, https://doi.org/10.1016/ j.amjcard.2017.10.017. 1. Brown ML, Dearani JA, Danielson GK, Cetta F, Connolly HM, Warnes CA, Li Z, Hodge DO, Driscoll DJ. The outcomes of operations for 539 patients with Ebstein anomaly. J Thorac Cardiovasc Surg 2008;135:1120– 1136. 2. Brown ML, Dearani JA, Danielson GK, Cetta F, Connolly HM, Warnes CA, Li Z, Hodge DO, Driscoll DJ. Comparison of the outcome of porcine bioprosthetic versus mechanical prosthetic replacement of the tricuspid valve in the Ebstein anomaly. Am J Cardiol 2009;103:555–561. 3. Filsoufi F, Anyanwu AC, Salzberg SP, Frankel T, Cohn LH, Adams DH. Long-term outcomes of tricuspid valve replacement in the current era. Ann Thorac Surg 2005;80:845–850. 4. Chang BC, Lim SH, Yi G, Hong YS, Lee S, Yoo KJ, Kang MS, Cho BK. Long-term clinical results of tricuspid valve replacement. Ann Thorac Surg 2006;81:1317–1323. 5. Moraca RJ, Moon MR, Lawton JS, Guthrie TJ, Aubuchon KA, Moazami N, Pasque MK, Damiano RJ. Outcomes of tricuspid valve repair and replacement: a propensity analysis. Ann Thorac Surg 2009;87:83–88. 6. van Slooten YJ, Freling HG, van Melle JP, Mulder BJ, Jongbloed MR, Ebels T, Voors AA, Pieper PG. Long-term tricuspid valve prosthesisrelated complications in patients with congenital heart disease. Eur J Cardiothorac Surg 2014;45:83–89. 7. Ruparelia N, Mangieri A, Ancona M, Giannini F, Agricola E, Montorfano M, De Bonis M, Alfieri O, Colombo A, Latib A. Percutaneous transcatheter treatment for tricuspid bioprosthesis failure. Catheter Cardiovasc Interv 2016;88:994–1001. 8. Eicken A, Schubert S, Hager A, Hörer J, McElhinney DB, Hess J, Ewert P, Berger F. Percutaneous tricuspid valve implantation: two-center experience with midterm results. Circ Cardiovasc Interv 2015;8:e002155. 9. Tzifa A, Momenah T, Al Sahari A, Al Khalaf K, Papagiannis J, Qureshi SA. Transcatheter valve-in-valve implantation in the tricuspid position. EuroIntervention 2014;10:995–999. 10. Cunnington C, Hoschtitzky JA, Hasan R, Clarke B, Mahadevan VS. Percutaneous tricuspid valve-in-valve implantation in Ebstein’s anomaly: one-year follow-up of valve function. Int J Cardiol 2014;174:e77– e78. 11. Hanna BM, Rodes-Cabau J, Dahdah N. Percutaneous transcatheter valvein-valve implantation with the balloon-expandable valve for the treatment of a dysfunctional tricuspid bioprosthetic valve: a pediatric case report. J Invasive Cardiol 2013;25:310–312. 12. Cullen MW, Cabalka AK, Alli OO, Pislaru SV, Sorajja P, Nkomo VT, Malouf JF, Cetta F, Hagler DJ, Rihal CS. Transvenous, antegrade Melody valve-in-valve implantation for bioprosthetic mitral and tricuspid valve dysfunction: a case series in children and adults. JACC Cardiovasc Interv 2013;6:598–605. 13. Van Garsse LA, Ter Bekke RM, van Ommen VG. Percutaneous transcatheter valve-in-valve implantation in stenosed tricuspid valve bioprosthesis. Circulation 2011;123:e219–e221. 14. Roberts PA, Boudjemline Y, Cheatham JP, Eicken A, Ewert P, McElhinney DB, Hill SL, Berger F, Khan D, Schranz D, Hess J, Ezekowitz MD, Celermajer D, Zahn E. Percutaneous tricuspid valve replacement in congenital and acquired heart disease. J Am Coll Cardiol 2011;58:117–122. 15. McElhinney DB, Cabalka AK, Aboulhosn JA, Eicken A, Boudjemline Y, Schubert S, Himbert D, Asnes JD, Salizzoni S, Bocks ML, Cheatham JP, Momenah TS, Kim DW, Schranz D, Meadows J, Thomson JD, Goldstein BH, Crittendon I 3rd, Fagan TE, Webb JG, Horlick E, Delaney JW, Jones TK, Shahanavaz S, Moretti C, Hainstock MR, Kenny DP, Berger F, Rihal CS, Dvir D. Transcatheter tricuspid valve-in-valve implantation for the treatment of dysfunctional surgical bioprosthetic valves: an international, multicenter registry study. Circulation 2016;133:1582– 1593.

ARTICLE IN PRESS Congenital Heart Disease/Transcatheter Tricuspid Valve-in-Valve in Ebstein Anomaly 16. Latib A, Naganuma T, Abdel-Wahab M, Danenberg H, Cota L, Barbanti M, Baumgartner H, Finkelstein A, Legrand V, de Lezo JS, Kefer J, Messika-Zeitoun D, Richardt G, Stabile E, Kaleschke G, Vahanian A, Laborde JC, Leon MB, Webb JG, Panoulas VF, Maisano F, Alfieri O, Colombo A. Treatment and clinical outcomes of transcatheter heart valve thrombosis. Circ Cardiovasc Interv 2015;8:e001779. 17. Schneider AE, Delaney JW, Cabalka AK. Non-infectious thrombosis of the melody(R) valve: A tale of two cities. Catheter Cardiovasc Interv 2016;88:600–604. 18. Egbe AC, Pislaru SV, Pellikka PA, Poterucha JT, Schaff HV, Maleszewski JJ, Connolly HM. Bioprosthetic valve thrombosis versus structural failure: clinical and echocardiographic predictors. J Am Coll Cardiol 2015;66:2285–2294.

7

19. Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP 3rd, Guyton RA, O’Gara PT, Ruiz CE, Skubas NJ, Sorajja P, Sundt TM. 2014 AHA/ACC Guideline for the Management of Patients With Valvular Heart Disease: a report of the American College of Cardiology/ American Heart Association Task Force on Practice Guidelines. Circulation 2014;129:e521–e643. 20. Bentham J, Qureshi S, Eicken A, Gibbs J, Ballard G, Thomson J. Early percutaneous valve failure within bioprosthetic tricuspid tissue valve replacements. Catheter Cardiovasc Interv 2013;82:428– 435. 21. Jeganathan R, Armstrong S, Al-Alao B, David T. The risk and outcomes of reoperative tricuspid valve surgery. Ann Thorac Surg 2013;95:119–124.