Outcomes following implantation of mechanical circulatory support in adults with congenital heart disease: An analysis of the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS)

http://www.jhltonline.org

ORIGINAL CLINICAL SCIENCE

Outcomes following implantation of mechanical circulatory support in adults with congenital heart disease: An analysis of the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) Christina J. VanderPluym, MD,a,1 Ari Cedars, MD,b,1 Pirooz Eghtesady, MD, PhD,c Bryan G. Maxwell, MD, MPH,d Jill M. Gelow, MD, MPH,e Luke J. Burchill, MMBS, PhD,e Simon Maltais, MD, PhD,f Devin A. Koehl, BS,g Ryan S. Cantor, MSPH,g,h and Elizabeth D. Blume, MDa From the aDepartment of Cardiology, Boston Children’s Hospital, Boston, Massachusetts; bDepartment of Cardiology, Baylor University Medical Center, Dallas, Texas; cDepartment of Cardiothoracic Surgery, Washington University School of Medicine, St. Louis, Missouri; dDepartment of Anesthesiology, Randall Children’s Hospital, Portland, Oregon; e Knight Cardiovascular Institute, Oregon Health and Science University, Portland, Oregon; fDepartment of Cardiovascular Surgery, Mayo Clinic School of Medicine, Rochester, Minnesota; gDepartment of Cardiothoracic Surgery, University of Alabama at Birmingham, Birmingham, Alabama; and the hDepartment of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, Alabama.

KEYWORDS: congenital heart disease; ventricular assist device; LVAD; single ventricle; mechanical circulatory support; total artificial heart

BACKGROUND: Adults with congenital heart disease represent an expanding and unique population of patients with heart failure (HF) in whom the use of mechanical circulatory support (MCS) has not been characterized. We sought to describe overall use, patient characteristics, and outcomes of MCS in adult congenital heart disease (ACHD). METHODS: All patients entered into the Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) between June 23, 2006, and December 31, 2015, were included. Patients with ACHD were identified using pre-operative data and stratified by ventricular morphology. Mortality was compared between ACHD and non-ACHD patients, and multivariate analysis was performed to identify predictors of death after device implantation. RESULTS: Of 16,182 patients, 126 with ACHD stratified as follows: systemic morphologic left ventricle (n ¼ 63), systemic morphologic right ventricle (n ¼ 45), and single ventricle (n ¼ 17). ACHD patients were younger (42 years ⫾ 14 vs 56 years ⫾ 13; p o 0.0001) and were more likely to undergo device implantation as bridge to transplant (45% vs 29%; p o 0.0001). A higher proportion of ACHD patients had biventricular assist device (BiVAD)/total artificial heart (TAH) support compared with non-ACHD patients (21% vs 7%; p o 0.0001). More ACHD patients on BiVAD/TAH support were INTERMACS profile 1 compared with patients on systemic left ventricular assist device (LVAD) support (35% vs 15%; p ¼ 0.002). ACHD and non-ACHD patients with LVADs had similar survival; survival was worse

1

These authors have contributed equally to this work. Reprint requests: Christina J. VanderPluym, MD, Department of Cardiology, Boston Children’s Hospital, 300 Longwood Avenue, Boston, MA 02115. Telephone: 617-355-6329. Fax: 617-734-9930. E-mail address: [email protected] 1053-2498/$ - see front matter r 2017 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2017.03.005

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]

2

for patients on BIVAD/TAH support. BiVAD/TAH support was the only variable independently associated with mortality (early phase hazard ratio 4.4; 95% confidence interval, 1.8–11.1; p ¼ 0.001). For ACHD patients receiving MCS, ventricular morphology was not associated with mortality. CONCLUSIONS: ACHD patients with LVADs have survival similar to non-ACHD patients. Mortality is higher for patients requiring BiVAD/TAH support, potentially owing to higher INTERMACS profile. These outcomes suggest a promising role for LVAD support in ACHD patients as part of the armamentarium of therapies for advanced HF. J Heart Lung Transplant ]]]];]:]]]–]]] r 2017 International Society for Heart and Lung Transplantation. All rights reserved.

The prevalence of end-stage heart failure (HF) among adults with congenital heart disease has grown exponentially in the last 2 decades as a result of improved diagnosis and surgical outcomes of congenital heart disease in childhood. In 2010, the population of patients with congenital heart disease surviving into adulthood in the United States was estimated to be 1.1 million, and it is projected to be approximately 1.3 million by 2020 with similar increases throughout North America and worldwide.1–4 Patients with adult congenital heart disease (ACHD) are a heterogeneous group with unique anatomy and long-term sequelae related to chronic cardiac disease. HF is among the most common reasons for hospitalization and death in patients with ACHD, similar to patients with acquired heart disease.5–8 However, the mechanisms underlying HF in ACHD are lesion dependent and differ substantially from the mechanisms responsible for HF in the general population.9 As a result, therapies developed to treat HF in patients with normal cardiac anatomy have frequently failed to produce a benefit in ACHD. Among the novel therapies developed for HF, durable mechanical circulatory support (MCS) in the form of ventricular assist devices (VADs) have become a mainstay of advanced HF management.10 MCS devices decrease mortality in end-stage HF among non-transplant candidates and appear to offer a similar benefit to transplant candidates doing poorly while awaiting transplant.11–13 Although the numbers of heart transplants in patients with ACHD have increased over the past decade, these patients remain more likely to have a prolonged waiting time and to die while on the transplant waitlist.14–17 Despite the potential benefit of MCS in this setting, its use as a bridge to transplant in ACHD has not been as widely adopted as it has in HF with normal anatomy.15,18 There are multiple reasons for this, but one is the limited data describing MCS use in this group.19 In the present study, we sought to better define the role of durable MCS in ACHD by examining its overall use, patient and device characteristics, post-implantation outcomes, and predictors of mortality using a large national database of patients on MCS.

Methods INTERMACS database and study population The Interagency Registry for Mechanically Assisted Circulatory Support (INTERMACS) is a national prospective database of

416,000 patients in the United States supported on durable MCS devices approved by the Food and Drug Administration. The database is a collaboration between the US National Heart, Lung, and Blood Institute, US Food and Drug Administration, US Centers for Medicare and Medicaid Services, industry, and implantation centers. All registry data are monitored by an independent observation study monitoring group appointed by the National Institutes of Health. Between June 23, 2006, and December 31, 2015, 16,182 patients (Z19 years old at time of MCS implantation) were entered into the INTERMACS database, which provided the study cohort.

INTERMACS database audit Patients with congenital heart disease (CHD) were identified by searching the INTERMACS database for the following variables: “cardiac diagnosis, primary”; “cardiac diagnosis, secondary”; “previous cardiac operation—congenital cardiac surgery”; “concomitant surgery”; “intervention within 48 hours of implant”; and “clinical events this hospitalization.” All patients identified with a CHD diagnostic code were reviewed by 2 cardiologists (C.J.V. and E.D.B.) with training in CHD to ensure that the primary diagnoses were consistent with previous surgical procedures. If there were incongruences between codes or missing data, the inputting center was contacted to clarify data. Of the 155 patients identified with diagnostic and surgical codes for CHD, 27 were excluded for erroneous coding, and 2 were excluded because of insufficient information to confirm the diagnosis.

Definitions Patients with CHD were grouped based on ventricular morphology owing to the assumption that complications associated with device implantation would be related predominantly to the type of ventricle into which the device was being implanted. The groups were systemic morphologic left ventricle, systemic morphologic right ventricle, and single ventricle. Cardiac lesions included in the groups are as follows:  Morphologic right ventricle: levo-transposition of the great

arteries (l-TGA, or congenitally corrected l-TGA) unrepaired, dextro-TGA (d-TGA) status post atrial switch procedure (Senning or Mustard)  Morphologic left ventricle: d-TGA status post arterial switch procedure or l-TGA status post double switch procedure, common atrioventricular canal, tetralogy of Fallot, congenital valvar disease (stenosis or regurgitation of the tricuspid, pulmonary, mitral, or aortic valve), ventricular septal defect, atrial septal defect, patent ductus arteriosus)

VanderPluym et al.

MCS in Adults With Congenital Heart Disease

3

 Single ventricle: single ventricle circulations, unbalanced

atrioventricular canal with hypoplastic left or right ventricle, status post Fontan circulation For this article, left ventricular assist device (LVAD) denotes a VAD implanted into the systemic ventricle regardless of the underlying morphology of that ventricle. A right ventricular assist device (RVAD) is defined as a VAD implanted into the subpulmonary ventricle. All standardized definitions of adverse events can be found at https://www.uab.edu/medicine/intermacs/ appendices-4-0/appendix-a-4-0. Early adverse events were defined as occurring o3 months after device implantation, and late adverse events were defined as occurring Z3months after device implantation.

Data collection and follow-up Per usual INTERMACS protocol, data regarding demographics, hemodynamics, and previous surgical procedures were collected before implantation, and clinical data were collected at serial intervals during the follow-up period. Major adverse events, including bleeding, infection, device malfunction, stroke, and death, were recorded when they occurred. Follow-up for all study events was continued through December 31, 2015, and was complete for 495% of enrolled patients. Patients were censored at time of transplantation, explantation for recovery, or death.

Statistical analysis Summary data are presented as mean ⫾ SD or number (percent). Baseline characteristics of patients with and without a diagnosis of CHD were compared using the chi-square test for categorical variables or Fisher’s exact test where appropriate and t-test for continuous variables. Survival after device implantation among groups was compared using Kaplan-Meier survival analysis and log-rank test. Multiphase parametric hazard modeling was used to model the time-varying hazard of death on a device in ACHD patients.20 The effect of dominant ventricle category was evaluated in a series of univariate models. A multivariate model using forward selection was developed to evaluate risk factors for mortality after device implantation; potential covariates included age, sex, pulsatile flow device, INTERMACS patient profile, biventricular support, and ventricular type as classified earlier. Data were analyzed using SAS version 9.4 (SAS Institute Inc., Cary, NC). All statistical tests were 2-sided, and a p-value o 0.05 was considered statistically significant. The authors had full access to the data and take responsibility for its integrity.

Figure 1 Graphic representation of patient selection and categorization into groups.

(range, 236–412 patients) compared with the 41 centers that performed implantations in o2 ACHD patients with a median non-ACHD implantation volume of 156 patients (range, 0–334 patients). Compared with non-ACHD patients, ACHD patients were younger (42 years ⫾ 14 vs 56 years ⫾ 13; p o 0.0001) and more likely to be white (76% vs 68%; p ¼ 0.0004). In addition, ACHD patients were more likely to have depressed right ventricular function (p o 0.0001) and less likely to have depressed left ventricular function or mitral regurgitation (p o 0.0001 for both). There was no difference in era of device implantation or in INTERMACS profile at the time of implantation between the 2 groups. With regard to transplant consideration, ACHD patients were more likely to be allosensitized (7% vs 1%; p ¼ 0.0003) and to have unfavorable chest anatomy (15% vs 4%; p ¼ 0.0003). Complete pre-implantation patient characteristics of ACHD and non-ACHD patients are listed in Table 1.

Biventricular assist device/total artificial heart ACHD subgroup As there was a higher proportion of biventricular assist device (BiVAD)/total artificial heart (TAH) use in ACHD patients compared with non-ACHD patients (21% vs 7%; p o 0.001), we chose to examine the 26 ACHD BiVAD/TAH patients

Results Study population Among 16,182 patients in the INTERMACS database, 126 were identified as having ACHD (Figure 1). Of 158 centers contributing to the INTERMACS registry, only 59 were responsible for all 126 patients, with 470% of centers contributing only 1 or 2 patients (Figure 2). The 6 centers that had the highest volume of ACHD patients on MCS (5 or 6 patients) were high-volume implantation centers with a median non-ACHD implantation volume of 320 patients

Figure 2

INTERMACS hospitals enrolling ACHD patients.

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]

4 Table 1

Pre-Implant Patient Characteristics (INTERMACS June 23, 2006 to December 31, 2015)

Characteristic

ACHD (n ¼ 126)

Non-ACHD (n ¼ 16,056)

p-value

Age, years Race White African American Other Female INTERMACS Profile 1. Critical cardiogenic shock 2. Stable but inotrope dependent 3. Progressive decline 4. Resting symptoms 5. Exertion intolerant 6. Exertion limited 7. Advanced NYHA class III Device strategy Bridge to transplant Bridge to candidacy Destination therapy Bridge to recovery Other Device type LVAD BiVAD TAH RVAD Implant date 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Mitral regurgitation—moderate or severe Tricuspid regurgitation—moderate or severe Aortic regurgitation—moderate or severe LVEF o40% RV dysfunction—moderate or severe Cardiac index Laboratory values Sodium (mEq/liter) BUN (mg/dl) Creatinine (mg/dl) ALT (U/liter) AST (U/liter) Bilirubin (mg/dl) Albumin (g/liter) Pre-albumin (mg/dl) BNP (pg/ml) Pro-BNP (pg/ml) Hemoglobin (g/dl) Platelet count ( 1,000) CRP (mg/liter) Pre-implantation comorbidities a Advanced age

41.9 ⫾ 14.3

56.5 ⫾ 12.9

o0.0001 0.0004

96 12 18 25

10,956 (68) 3,715 (23) 1,385(9) 3,410 (21)

(76) (10) (14) (20)

0.83 0.32

23 (18) 57 (46) 29 (23) 13 (10) 1 (1) 2 (2) 0

2,742 (17) 5,936 (37) 4,614 (29) 2,072 (13) 386 (2) 153 (1) 82 (1)

57 (45) 48 (38) 20 (16) 1 (1) 0

4,601 (29) 5,145 (32) 6,102 (38) 111 (1) 97 (1)

97 (77) 14 (11) 12 (10) 3 (2)

14,889 (93) 819 (5) 332 (2) 16 (1)

1 (1) 4 (3) 4 (3) 3 (2) 10 (8) 17 (14) 14 (11) 22 (18) 27 (21) 24 (19) 32 (27) 50 (42) 13 (11) 88 (74) 56 (47) 2.4 ⫾ 0.9

96 (1) 336 (2) 733 (5) 1,006 (6) 1,642 (10) 1,894 (12) 2,239 (14) 2,668 (17) 2,741 (17) 2,701 (17) 8,245 (54) 6,233 (41) 576 (4) 14,325 (94) 4,870 (32) 2.2 ⫾ 0.9

134.4 ⫾ 4.9 28.8 ⫾ 16.6 1.4 ⫾ 0.7 90.5 ⫾ 371.7 107.8 ⫾ 556.4 1.6 ⫾ 1.8 35.9 ⫾ 7.3 17.5 ⫾ 7.7 1,374.3 ⫾ 1,374.9 8,905.7 ⫾ 11,708 11.1 ⫾ 2.3 188.5 ⫾ 76.5 29.8 ⫾ 45.5

134.8 ⫾ 5.0 29.9 ⫾ 18.6 1.4 ⫾ 0.7 79.6 ⫾ 263.9 71.2 ⫾ 285.4 1.5 ⫾ 2.1 33.9 ⫾ 6.7 18.5 ⫾ 7.5 1,176.2 ⫾ 1,097.6 6,743.9 ⫾ 7,752.3 11.3 ⫾ 2.1 196.6 ⫾ 81.5 20.3 ⫾ 40.6

0.40 0.53 0.70 0.65 0.17 0.53 0.0011 0.30 0.21 0.10 0.19 0.26 0.28

3 (4)

1,556 (16)

0.0008

o0.0001

o0.0001

0.55

o0.0001 0.001 o0.0001 o0.0001 o0.0001 0.11

Continued on page 5

VanderPluym et al.

MCS in Adults With Congenital Heart Disease

5

Table 1 (Continued ) Characteristic History of hepatitis History of alcohol abuse History of atrial arrhythmia History of illicit drug use History of smoking Allosensitization Large BMI Limited social support Liver dysfunction Limited cognitive understanding Major stroke Malnutrition/cachexia Musculoskeletal limitation Other comorbidity Other cerebrovascular disease Peripheral vascular disease Pulmonary disease Pulmonary hypertension Pulmonary embolus Non-compliance Severe depression Severe diabetes Chronic infectious concerns Thoracic aortic disease Unfavorable mediastinal anatomy Chronic renal disease Current smoker Frailty Frequent ICD shocks

ACHD (n ¼ 126)

Non-ACHD (n ¼ 16,056)

3 (4) 6 (7) 25 (30) 7 (8) 21 (25) 6 (7) 9 (11) 3 (4) 1 (1) 1 (1) 4 (5) 4 (5) 1 (1) 10 (12) 2 (2) 2 (2) 6 (7) 28 (34) 1 (1) 1 (1) 2 (2) 5 (6) 2 (2) 2 (2) 12 (15) 18 (22) 2 (2) 7 (8) 4 (5)

148 (2) 761 (8) 2,083 (22) 688 (7) 2,903 (30) 101 (1) 1,484 (15) 434 (5) 393 (43) 163 (2) 380 (4) 270 (3) 81 (1) 775 (8) 228 (2) 467 (5) 939 (10) 2,271 (23) 162 (2) 308 (3) 279 (3) 930 (10) 97 (1) 111 (1) 424 (4) 2,189 (23) 463 (5) 619 (6) 542 (6)

p-value 0.14 0.82 0.06 0.65 0.34 0.0003 0.25 1 0.26 1 0.57 0.29 0.51 0.18 0.72 0.44 0.58 0.03 1 0.52 1 0.35 0.21 0.25 0.0003 0.82 0.44 0.50 1

Data expressed as mean ⫾ SD or as number (%). ACHD, adult congenital heart disease; ALT, alanine aminotransferase; AST, aspartate aminotransferase; BiVAD, biventricular assist device; BMI, body mass index; BNP, B-type natriuretic peptide; BUN, blood urea nitrogen; CRP, C-reactive protein; ICD, implantable cardioverter defibrillator; LVAD, left ventricular assist device; LVEF, left ventricular ejection fraction; NYHA, New York Heart Association; RV, right ventricular; RVAD, right ventricular assist device; TAH, total artificial heart. a Pre-implantation comorbidities are reported for patients receiving implants after May 2, 2012.

separately. The BIVAD/TAH cohort had a higher INTERMACS patient profile, with all patients being either patient profile 1 (n ¼ 9 [35%]) or 2 (n ¼ 17 [65%]), compared with ACHD patients supported with LVADs (profile 1, n ¼ 14 [15%]; profile 2, n ¼ 39 [41%]). There was no destination therapy in the BiVAD/TAH ACHD group; rather, these patients were supported as bridge to transplant (n ¼ 13 [50%]) or bridge to candidacy (n ¼ 13 [50%]). Compared with ACHD patients on LVAD support, ACHD patients on BiVAD/TAH support had higher blood urea nitrogen (35 mg/dl ⫾ 27 vs 27 mg/dl ⫾ 13; p ¼ 0.03), higher creatinine (1.7 mg/dl ⫾ 1.0 vs 1.3 mg/dl ⫾ 0.7; p ¼ 0.04), and lower hemoglobin (9.9 g/dl ⫾ 1.5 vs 11.4 g/dl ⫾ 2.4; p ¼ 0.003). There was a trend toward more right ventricular dysfunction in the BIVAD/TAH ACHD group; however, it did not reach statistical significance. Compared with ACHD patients on LVAD support, ACHD patients on BiVAD/TAH support were more likely to have chronic renal disease (41% vs 16%; p ¼ 0.03), pulmonary disease (23% vs 4%; p ¼ 0.009), and unfavorable mediastinal anatomy (35% vs 9%; p ¼ 0.02). Complete

pre-implantation patient characteristics of ACHD patients on BIVAD/TAH support and LVAD support are listed in Table 1.

Device strategy and characteristics Compared with non-ACHD patients, ACHD patients supported on all forms of MCS were more likely to have undergone device implantation as a bridge to transplant (45% vs 29%) than as destination therapy (16% vs 38%; p o 0.0001). There was a higher proportion of BiVAD/ TAH use in ACHD patients compared with non-ACHD patients (21% vs 7%; p o 0.001). As such, ACHD patients were less likely to receive a continuous-flow device (82% vs 92%; p o 0.0001). When examining type of device implanted in the 3 groups of ACHD patients (morphologic left ventricle, morphologic right ventricle, and single ventricle), there was a predominance of continuous-flow axial devices, with continuous-flow centrifugal devices being the second most frequently used device across the 3 groups (Table S4 [available in the online version of this article at www.jhltonline.org]).

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]

6 Table 2

Adverse Event Rates for ACHD vs Non-ACHD Patients Receiving MCS (INTERMACS June 23, 2006 to December 31, 2015). ACHD

Event

Perioda

Arterial non-CNS thromboembolism

Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late Early Late

Bleeding Cardiac arrhythmia Hepatic dysfunction Infection Myocardial infarction Neurologic dysfunction CVA Other SAE Pericardial drainage Psychiatric episode Rehospitalization Renal dysfunction Respiratory failure Venous thromboembolism Wound dehiscence

Non-ACHD AE Rateb

Events

AE Rateb

p-value

1 2 52 26 33 29 18 12 64 85

0.3 0.1 16.1 1.9 10.2 2.1 5.6 0.9 19.8 6.2

5.6 1.5 3.4 0.9 19.2 2.0 1.6

12 6 66 244 34 15 49 27 7

3.7 0.4 20.5 17.9 10.5 1.1 15.2 2.0 2.2

2

0.6

0.4 0.04 19.6 3.3 11.2 1.0 1.6 0.2 16.9 4.0 0.1 0.03 4.6 1.2 2.5 0.7 13.2 1.9 2.2 0.01 2.5 0.3 20.6 15.6 4.3 0.4 8.0 0.5 1.7 0.07 0.5 0.03

0.75 0.04 0.16 0.006 0.62 o0.001 o0.001 o0.001 0.21 o0.001

18 20 11 12 62 27 5

184 86 8,461 7,305 4,814 2,332 687 440 7,300 9,085 55 56 1,997 2,710 1,079 1,623 5,676 4,200 941 29 1,087 679 8,869 34,995 1,845 1,008 3,430 1,133 715 148 225 74

Events

0.43 0.39 0.31 0.50 0.003 0.77 0.44 0.18 0.36 0.97 0.03 o0.001 o0.001 o0.001 o0.001 0.48 0.81

ACHD, adult congenital heart disease; AE, adverse event; CNS, central nervous system; CVA, cerebrovascular accident; SAE, serious adverse event. a Early ¼ within 3 months of device implantation. Late ¼ Z 3 months after device implantation. b Rates per 100 patient-months. ACHD early follow-up, 322.47 months; ACHD late follow-up, 1,364.13 months; non-ACHD early follow-up, 43,119.34 months; non-ACHD late follow-up, 224,374.69 months.

Adverse events Adverse events for all types of MCS devices are presented in Table 2. Compared with non-ACHD patients, ACHD patients were more likely to have both early and late hepatic dysfunction, renal dysfunction, and respiratory failure (p o 0.001 both early and late for all 3). In addition, ACHD patients were more likely to experience late arrhythmias, late infection (p o 0.001 for both), and late hospital readmission (p ¼ 0.03) but less likely to experience late bleeding (p ¼ 0.006).

LVAD adverse events Adverse events for ACHD and non-ACHD patients supported with LVADs are presented in Table S2 (available in the online version of this article at www.jhltonline.org). ACHD patients with LVADs had higher rates of late arrhythmia (p o 0.001), early hepatic dysfunction (p o 0.001), early and late renal dysfunction (p ¼ 0.002, p ¼ 0.01), early and late

respiratory failure (p ¼ 0.02, p o 0.001), and late infection (p ¼ 0.04). ACHD patients with LVADs were less likely to experience early or late bleeding compared with non-ACHD patients with LVADs (p ¼ 0.01, p ¼ 0.005). There was no difference in pump thrombosis, neurologic events, or readmis sion rate between ACHD and non-ACHD patients with LVADs.

BiVAD/TAH adverse events Adverse events for ACHD patients on BiVAD/TAH support are presented in Table S3 (available in the online version of this article at www.jhltonline.org). Compared with ACHD patients with LVADs, ACHD patients on BiVAD/TAH support had a significantly higher rate of early bleeding, late hepatic dysfunction, early and late infection, early and late neurologic dysfunction, early and late renal dysfunction, early and late respiratory failure, early and late device malfunction, and late rehospitalization (p o 0.001 for all).

VanderPluym et al.

MCS in Adults With Congenital Heart Disease

7

Figure 3 (A) Kaplan-Meier survival after MCS implantation for all ACHD compared with all non-ACHD patients. (B) Kaplan-Meier survival after VAD implantation for all ACHD compared with all non-ACHD patients. (C) Kaplan-Meier survival after BiVAD and TAH implantation for all ACHD compared with all non-ACHD patients. All patients were censored at the time of transplant or recovery (MCS explantation).

Mortality and transplant Overall, ACHD patients had a higher mortality after MCS device implantation than non-ACHD patients (p ¼ 0.01) (Figure 3A). When stratified by device type, ACHD and non-ACHD patients had a similar mortality rate after LVAD implantation, whereas the mortality rate after BiVAD/TAH implantation was significantly higher in ACHD patients (p ¼ 0.02) (Figure 3B and C). Similarly, on competing outcomes analysis, a greater proportion of ACHD patients died while supported by MCS compared with non-ACHD

patients (28% vs 19%; p ¼ 0.08) (Figure 4A and B). However, among patients with LVADs, this difference was not significant (21% vs 18%; p ¼ 0.64) (Figure 5A and B).

Predictors of mortality in patients with ACHD supported with MCS Ventricular morphology was not associated with reduced survival among ACHD patients on all forms of MCS or for LVADs alone (Figure 6A–D). On multivariate analysis, only BiVAD/TAH support correlated with mortality (early

Figure 4 Competing outcomes analysis, including alive with device in place, death before transplant, transplant, and explantation to recovery after MCS for all ACHD (A) and all non-ACHD (B) patients.

8

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]

Figure 5 Competing outcomes analysis, including alive with device in place, death before transplant, transplant, and explantation to recovery after LVAD for ACHD (A) and non-ACHD (B) patients.

phase hazard ratio 4.4; 95% confidence interval, 1.8–11.1; p ¼ 0.001), whereas there was a trend toward increased mortality among ACHD patients with an INTERMACS profile of 1 (early phase hazard ratio 2.4; 95% confidence interval 1.0–5.8; p ¼ 0.06).

Discussion In this article, we present an INTERMACS analysis of the utilization practices, patient characteristics, and outcomes of MCS in patients with ACHD. ACHD patients differed most

notably from non-ACHD patients in expected ways, including younger age, greater allosensitization, more right ventricular dysfunction, and unfavorable mediastinal anatomy at time of device implantation. As a group, ACHD patients had higher rates of mortality and adverse events than nonACHD patients; however, the increased mortality rate was exclusively attributable to ACHD patients on BiVAD/TAH support. ACHD and non-ACHD patients with LVADs demonstrated similar survival regardless of cardiac anatomy. Only BiVAD/TAH support predicted mortality after MCS among ACHD patients.

Figure 6 (A) Kaplan-Meier survival after MCS implantation for all ACHD patients divided by ventricular morphology. (B) KaplanMeier survival after LVAD implantation for patients with a systemic morphologic left ventricle. (C) Kaplan-Meier survival after LVAD implantation for patients with a systemic morphologic right ventricle. (D) Kaplan-Meier survival after LVAD implantation for patients with a single ventricle. All patients were censored at the time of transplant or recovery (MCS explantation).

VanderPluym et al.

MCS in Adults With Congenital Heart Disease

These results provide new insight into the outcomes after durable MCS in patients with ACHD. Existing published data on the use and outcomes of MCS in patients with ACHD are limited to case reports and case series from various adult and pediatric centers, all of which highlight the difficulties imposed by heterogeneous anatomy when using MCS in this population. Among patients with a systemic right ventricle, variable location of the systemic ventricle and the highly trabeculated right ventricular anatomy have made standardization of operative approach for device implantation impossible given limited data.21–24 Similarly, MCS for patients with a Fontan circulation, in whom anatomic heterogeneity in many cases proves an even greater obstacle,25,26 is poorly studied with limited case reports in adolescent and adult patients with a Fontan circulation describing systemic ventricular and/or cavopulmonary support.27–29 Despite these impediments to optimizing the operative approach, the present study identified only a slightly higher early hazard for mortality among ACHD patients with morphologic right ventricle supported with LVADs, which appeared to resolve after the perioperative period. Similarly, although numbers were limited, we found that single ventricle anatomy was not associated with reduced survival among ACHD patients after MCS overall. Furthermore, we anticipate that these outcomes will improve with time, given the fact that experience with MCS in ACHD patients is largely limited, with 470% of centers having performed implantations in only 1 or 2 ACHD patients in the present analysis. Overall, current use of MCS in ACHD patients remains low, at o1% of the total INTERMACS population, and there has been no appreciable increase over the last decade.15,18 Anecdotal evidence suggests that the lack of experience with MCS in ACHD combined with concern over creating additional surgical difficulty and allosensitization with subsequent heart transplantation drives physicians to use MCS only as a “last resort” in these patients. It has been shown that 2-month waitlist outcomes are worse in transplant-listed ACHD patients who undergo MCS implantation.16 However, this finding was attributable in part to increased prevalence of risk factors for adverse events in these patients18 and in part to early post-operative mortality well described in ACHD patients after transplant30–32 and demonstrated to be present after device implantation in the present study. Furthermore, the heightened threshold for MCS utilization has the potential to delay implantation until elevated INTERMACS profiles are reached, creating a selffulfilling prophecy and confirming a perception of increased risk. In the present study, ACHD patients receiving BiVAD/ TAH support had higher INTERMACS patient profiles at the time of implantation and had evidence of pre-implantation renal and pulmonary disease. We can postulate that earlier consideration of MCS in these patients before development of end organ dysfunction may ultimately have improved postimplantation morbidity and mortality. This hypothesis seems supported by our finding that elevated INTERMACS profile tended to be associated with increased mortality risk. In view of these considerations, the present data suggest a potential benefit for increased MCS use as a bridge to

9

transplant in select ACHD patients. ACHD patients spend more time awaiting transplant and have a higher mortality rate on the waitlist than their counterparts with normal anatomy.16,17 Given the benefit of MCS among patients with normal cardiac anatomy,33 one would anticipate that MCS should be appealing in transplant-listed ACHD patients. In support of this belief, Maxwell et al18 demonstrated that transplant outcomes after MCS in ACHD patients are equivalent to outcomes without MCS despite the additional sternotomy and potential for device-related adverse events. Possibly more important, from the practical standpoint, current allocation practices in the United States provide 30 days of 1A listing time if a patient is supported on MCS, and MCS-supported patients can have additional 1A time for MCS-related complications. The low use of MCS in patients with ACHD is therefore a disadvantage in terms of listing priority. This disadvantage is hinted at in work by Everitt et al,17 who found that patients with ACHD were less likely to have a VAD (5% vs 14%; p o 0.01) and were more likely to be listed at a lower priority status compared with patients without CHD. As such, fewer ACHD patients underwent transplantation than non-ACHD patients (53% vs 65%; p o 0.001), and patients with ACHD were more likely to experience cardiovascular death (60% vs 40%; p ¼ 0.03). Furthermore, the elevated waitlist mortality among ACHD patients appears to be predominantly attributable to patients listed as status 1A, raising the possibility that the “stable” 1A time ACHD patients miss out on because of less MCS support may be a factor contributing to increased waitlist mortality.34 The present data also suggest a possible role for increased LVAD use as destination therapy in patients with ACHD. Only 16% of ACHD patients who underwent device implantation did so as destination therapy; this is likely due to the significantly lower median age of ACHD patients. However, as previously mentioned, mortality rates for LVAD were no different between ACHD patients and non-ACHD patients. The heterogeneity of ACHD and lack of a standardized definition of end-stage HF in processes as diverse as systemic right ventricle vs single ventricle with a Fontan circulation make extrapolation from the REMATCH trial difficult.11 Nevertheless, similar post-LVAD survival among ACHD patients at high enough risk to merit MCS device implantation despite limited data suggests that LVAD as destination therapy in non-transplant candidates might be at least as beneficial as among patients with normal cardiac anatomy. As the ACHD population ages and LVAD use in ACHD expands, this issue requires further exploration. The incidence of adverse events in patients with ACHD was higher than in patients with normal cardiac anatomy. However, the adverse events were not entirely unanticipated and may be as much related to the sequelae of ACHD as to failure of MCS to benefit ACHD patients. We found that patients with ACHD were more likely to experience early hepatic and renal dysfunction, respiratory failure, and other non-specific early events. Presumably the increased risk of hepatic dysfunction is related to patients with a Fontan

10

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]]

circulation and to the relatively high incidence of right ventricular failure found in the ACHD patients in the present study. The increased incidence of renal dysfunction, respiratory failure, and non-specific early events may relate to greater operative difficulty in patients with ACHD compared with patients with normal anatomy, although the present data are inadequate to determine this definitively. Not surprisingly, ACHD patients requiring BiVAD/TAH support had significantly higher rates of adverse events compared with ACHD patients with LVAD support. Compared with ACHD patients with LVADs, ACHD patients with BiVAD/TAH support had an increased rate of early bleeding, late device malfunction, early and late hepatic dysfunction, infection, and neurologic dysfunction. Nevertheless, the increased incidence of adverse events in ACHD patients was similar even when considering only patients who received an LVAD. More accurate delineation of operative duration and underlying anatomy would likely be necessary to identify risk factors for these events and strategies to prevent them, meriting further study. The present study has multiple limitations related to the nature of the data source. The INTERMACS registry includes only patients from participating centers who agree to their data being tracked. The accuracy with which all data were captured depends on the originating center, and there is likely center-to-center variability in the accuracy of the data. In particular, underlying anatomy was recorded in a nonstandardized fashion and was therefore difficult to establish and may be inaccurate. Finally, the number of patients with ACHD and centers implanting devices in these patients are limited compared with patients without congenital heart disease and may limit the conclusions derived from the comparison of the 2 groups. There are variables that are not collected within the INTERMACS database (e.g., center or surgeon experience with ACHD patients, number of ACHD heart transplants performed) that may be important factors in MCS outcomes in ACHD patients. We did not evaluate center specificity (pediatric vs adult vs combined), which may be relevant to the present analysis, but which was outside the scope of the analysis performed. This is an important consideration that we plan to address in future research. INTERMACS collects data only from patients who actually received a device. As such, this study reports only ACHD patients who underwent device implantation and not patients who were evaluated for possible device implantation but were deemed too sick or too well or were never evaluated at all. In addition, we cannot comment on ACHD patients with HF who died or underwent transplantation without using MCS. In conclusion, survival after LVAD implantation in ACHD patients is similar to survival of non-ACHD patients, regardless of ACHD lesion. ACHD patients undergoing BiVAD/TAH support have worse survival compared with non-ACHD patients; this is likely due to worse INTERMACS profile and greater comorbid burden. These data suggest that broader use of LVADs in ACHD patients with end-stage HF may result in benefits similar to those demonstrated in non-ACHD patients.

Disclosure statement None of the authors has a financial relationship with a commercial entity that has an interest in the subject of the presented manuscript or other conflicts of interest to disclose. This study was funded in part by support from the Cardiac Transplant and Education Fund at Boston Children’s Hospital.

Supplementary data Supplementary data are available online at www.jhltonline. org.

References 1. Marelli AJ, Ionescu-Ittu R, Mackie AS, Guo L, Dendukuri N, Kaouache M. Lifetime prevalence of congenital heart disease in the general population from 2000 to 2010. Circulation 2014;130:749-56. 2. van der Linde D, Konings EEM, Slager MA, et al. Birth prevalence of congenital heart disease worldwide: a systematic review and metaanalysis. J Am Coll Cardiol 2011;58:2241-7. 3. Marelli A, Gilboa S, Devine O, et al. Estimating the congenital heart disease population in the United States in 2010—what are the numbers? J Am Coll Cardiol 2012;59:E787. 4. Benziger CP, Stout K, Zaragoza-Macias E, Bertozzi-Villa A, Flaxman AD. Projected growth of the adult congenital heart disease population in the United States to 2050: an integrative systems modeling approach. Popul Health Metr 2015;13:29. 5. Opotowsky AR, Siddiqi OK, Webb GD. Trends in hospitalizations for adults with congenital heart disease in the U.S. J Am Coll Cardiol 2009;54:460-7. 6. Verheugt CL, Uiterwaal CSPM, van der Velde ET, et al. Mortality in adult congenital heart disease. Eur Heart J 2010;31:1220-9. 7. Oechslin EN, Harrison DA, Connelly MS, Webb GD, Siu SC. Mode of death in adults with congenital heart disease. Am J Cardiol 2000; 86:1111-6. 8. Diller G-P, Kempny A, Alonso-Gonzalez R, et al. Survival prospects and circumstances of death in contemporary adult congenital heart disease patients under follow-up at a large tertiary centre. Circulation 2015;132:2118-25. 9. Budts W, Roos-Hesselink J, Rädle-Hurst T, et al. Treatment of heart failure in adult congenital heart disease: a position paper of the Working Group of Grown-Up Congenital Heart Disease and the Heart Failure Association of the European Society of Cardiology. Eur Heart J 2016;37:1419-27. 10. Kirklin JK, Naftel DC, Pagani FD, et al. Seventh INTERMACS annual report: 15,000 patients and counting. J Heart Lung Transplant 2015;34: 1495-504. 11. Rose EA, Gelijns AC, Moskowitz AJ, et al. Long-term use of a left ventricular assist device for end-stage heart failure. N Engl J Med 2001;345:1435-43. 12. Attisani M, Centofanti P, La Torre M, et al. Advanced heart failure in critical patients (INTERMACS 1 and 2 levels): ventricular assist devices or emergency transplantation? Interact Cardiovasc Thorac Surg 2012;15:678-84. 13. Pamboukian SV, Tallaj JA, Brown RN, et al. Comparison of observed survival after ventricular assist device placement versus predicted survival without assist device using the Seattle heart failure model. ASAIO J 2012;58:93-7. 14. Karamlou T, Hirsch J, Welke K, et al. A United Network for Organ Sharing analysis of heart transplantation in adults with congenital heart disease: outcomes and factors associated with mortality and retransplantation. J Thorac Cardiovasc Surg 2010;140:161-8. 15. Gelow JM, Song HK, Weiss JB, Mudd JO, Broberg CS. Organ allocation in adults with congenital heart disease listed for heart transplant: impact of ventricular assist devices. J Heart Lung Transplant 2013;32:1059-64.

VanderPluym et al.

MCS in Adults With Congenital Heart Disease

16. Davies RR, Russo MJ, Yang J, Quaegebeur JM, Mosca RS, Chen JM. Listing and transplanting adults with congenital heart disease. Circulation 2011;123:759-67. 17. Everitt MD, Donaldson AE, Stehlik J, et al. Would access to device therapies improve transplant outcomes for adults with congenital heart disease? Analysis of the United Network for Organ Sharing (UNOS). J Heart Lung Transplant 2011;30:395-401. 18. Maxwell BG, Wong JK, Sheikh AY, Lee PHU, Lobato RL. Heart transplantation with or without prior mechanical circulatory support in adults with congenital heart disease. Eur J Cardiothorac Surg 2014; 45:842-6. 19. Ross HJ, Law Y, Book WM, et al. Transplantation and mechanical circulatory support in congenital heart disease: a scientific statement from the American Heart Association. Circulation 2016;133:802-20. 20. Blackstone EH, Naftel DC, Turner ME. The decomposition of timevarying hazard into phases, each incorporating a separate stream of concomitant information. J Am Stat Assoc 1986;81:615-24. 21. George RS, Birks EJ, Radley-Smith RC, Khaghani A, Yacoub M. Bridge to transplantation with a left ventricular assist device for systemic ventricular failure after Mustard procedure. Ann Thorac Surg 2007;83:306-8. 22. Stewart AS, Gorman RC, Pocchetino A, Rosengard BR, Acker MA. Left ventricular assist device for right side assistance in patients with transposition. Ann Thorac Surg 2002;74:912-4. 23. Wiklund L, Svensson S, Berggren H. Implantation of a left ventricular assist device, back-to-front, in an adolescent with a failing mustard procedure. J Thorac Cardiovasc Surg 1999;118:755-6. 24. Joyce DL, Crow SS, John R, St Louis JD, Braunlin EA, Pyles LA, et al. Mechanical circulatory support in patients with heart failure secondary to transposition of the great arteries. J Heart Lung Transplant 2010;29:1302-5. 25. Tchantchaleishvili V, Hallinan W, Schwarz KQ, Massey HT. Long-term total cardiac support in a Fontan-type circulation with HeartMate II

26.

27.

28.

29.

30.

31.

32.

33.

34.

11 left ventricular assist device. Interact Cardiovasc Thorac Surg 2016; 22:692-4. Russo P, Wheeler A, Russo J, Tobias JD. Use of a ventricular assist device as a bridge to transplantation in a patient with single ventricle physiology and total cavopulmonary anastomosis. Paediatr Anaesth 2008;18:320-4. Sinha P, Deutsch N, Ratnayaka K, et al. Effect of mechanical assistance of the systemic ventricle in single ventricle circulation with cavopulmonary connection. J Thorac Cardiovasc Surg 2014;147: 1271-5. Giridharan GA, Ising M, Sobieski MA, et al. Cavopulmonary assist for the failing Fontan circulation: impact of ventricular function on mechanical support strategy. ASAIO J 2014;60:707-15. Throckmorton AL, Lopez-Isaza S, Downs EA, Chopski SG, Gangemi JJ, Moskowitz W. A viable therapeutic option: mechanical circulatory support of the failing Fontan physiology. Pediatr Cardiol 2013;34: 1357-65. Chen JM, Davies RR, Mital SR, et al. Trends and outcomes in transplantation for complex congenital heart disease: 1984 to 2004. Ann Thorac Surg 2004;78:1352-61: discussion 1361. Patel ND, Weiss ES, Allen JG, et al. Heart transplantation for adults with congenital heart disease: analysis of the United network for organ sharing database. Ann Thorac Surg 2009;88:814-21: discussion 821-2. Lietz K, Miller LW. Improved survival of patients with end-stage heart failure listed for heart transplantation: analysis of organ procurement and transplantation network/U.S. United Network of Organ Sharing data, 1990 to 2005. J Am Coll Cardiol 2007;50:1282-90. Karamlou T, Gelow J, Diggs BS, et al. Mechanical circulatory support pathways that maximize post-heart transplant survival. Ann Thorac Surg 2013;95:480-5: discussion 485. Alshawabkeh LI, Hu N, Carter KD, et al. Wait-list outcomes for adults with congenital heart disease listed for heart transplantation in the U.S. J Am Coll Cardiol 2016;68:908-17.