http://www.jhltonline.org

ORIGINAL CLINICAL SCIENCE

Ventricular assist devices as a bridge-to-transplant improve early post-transplant outcomes in children Ryan R. Davies, MD,a,b Shylah Haldeman, RN,a Michael A. McCulloch, MD,a,c and Christian Pizarro, MDa,b From the aNemours Cardiac Center, Nemours/A.I. duPont Hospital for Children, Wilmington, Delaware; and the bDepartment of Surgery and the cDepartment of Pediatrics, Thomas Jefferson University, Philadelphia, Pennsylvania.

KEYWORDS: heart transplantation; pediatrics; mechanical circulatory support; ventricular assist device; extracoporeal membrane oxygenation; survival

BACKGROUND: The use of ventricular assist devices (VADs) to bridge pediatric patients to transplant or recovery has been expanding. There are few current pediatric data assessing the impact of VAD support on post-transplant survival. METHODS: We performed a retrospective review of all pediatric (r18 years old, n ¼ 4,028) transplants performed between 1995 and 2011 and contained within the United Network for Organ Sharing data set. Transplants were divided into three eras: early (1995 to 2002, n ¼ 1,450); intermediate (2003 to 2007, n ¼ 1,138); and recent (2008 to 2011, n ¼ 1,440). VADs were present at transplant in 398 patients (9.8%). Outcomes among patients with and without VADs were assessed and compared across eras. RESULTS: The use of VADs for bridge to transplant has increased (early 1.1%, intermediate 10.5%, recent 17.9%; p o 0.0001). Mean weight among VAD-supported patients (early 63.5 kg, intermediate 42.3 kg, recent 28.8 kg; p o 0.0001) has decreased during this period. VAD patients o10 kg had an increased risk of stroke (odds ratio [OR] ¼ 4.9, 95% confidence interval [CI] 2.1 to 10.8) compared with non-mechanical support patients. In multivariable analyses, extracorporeal VADs were the only type of VAD associated with higher post-transplant mortality (OR ¼ 3.0, 95% CI 0.8 to 10.6). Other types of VAD had lower mortality (OR ¼ 0.5, 95% CI 0.2 to 1.0). Long-term survival was unaffected by the use of a VAD pre-transplant. CONCLUSIONS: Pediatric patients bridged to transplantation with VADs are increasingly younger and smaller. Complication rates remain high among patients o10 kg. Early post-transplant survival among intracorporeal and paracorporeal VAD patients is excellent and better when compared with unsupported patients. The use of short-term support devices is associated with higher post-transplant mortality. Longterm survival is unaffected by VAD use. J Heart Lung Transplant ]]]];]:]]]–]]] r 2014 International Society for Heart and Lung Transplantation. All rights reserved.

Use of ventricular assist devices (VADs) is rapidly becoming a frequent management strategy for children with end-stage heart failure. This increasing use of VADs has resulted in part due to the progressive miniaturization of mechanical circulatory support (MCS) devices, including Reprint requests: Ryan R. Davies, MD, Nemours Cardiac Center, Nemours/A.I. duPont Hospital for Children, 1600 Rockland Road, Wilmington, DE 19803. Telephone: 302-651-6600. Fax: 302-651-5345. E-mail address: [email protected]

devices directed specifically at children as well as smaller implantable adult devices. However, aside from the issue of appropriately sized devices, management challenges in smaller patients continue to influence outcomes. It is not clear that the excellent post-transplant results seen in adults bridged with VADs can be extrapolated to children. Potential differences include physiologic and anatomic variation, a dissimilar immunologic response to blood transfusions, and a higher thromboembolic risk due to an immature coagulation cascade.1,2

1053-2498/$ - see front matter r 2014 International Society for Heart and Lung Transplantation. All rights reserved. http://dx.doi.org/10.1016/j.healun.2014.02.010

2

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

Previous multi-institutional studies of VAD use in children have been limited primarily to older patients and earlier generation devices.3,4 A more contemporary analysis—including the increasing use of MCS in younger children—is necessary to begin to understand the long-term post-transplant outcomes of children in all age groups using current technology.

Methods Data collection The United Network for Organ Sharing (UNOS) provided deidentified patient-level data (Source #12/01/2011-2, data as of September 30, 2012). Use of these data is consistent with the regulations of the Nemours Institutional Review Board.

Study population Heart transplantation was performed in 4,028 pediatric (r18 years old at listing) patients listed Status 1A/1B between 1995 and 2012. Patients with intra-aortic balloon pumps were excluded because numbers were limited. Patients were stratified based on year of transplant into three eras: early (1995 to 2002, n ¼ 1,450); intermediate (2003 to 2007, n ¼ 1,138); and recent (2008 to 2011, n ¼ 1,440). Outcomes were analyzed according to type of support and era of transplantation and were compared with patients on extracorporeal membrane oxygenation (ECMO) and those without mechanical circulatory support (nMCS).

Data analysis Data were analyzed using SAS 9.2 for AIX (SAS Institute, Cary, NC). The primary outcome was operative mortality (death within 30 days of transplant or before hospital discharge). Secondary outcomes included patient and graft survival analyzed by the lifetable method and the incidence of post-transplant complications (as defined by UNOS: dialysis; drug-treated infection; stroke; and permanent pacemaker implantation). Continuous variables were compared using Student’s t-test (2-tailed) and 1-way analysis of variance. Categorical variables were compared by chi-square test. Kaplan–Meier analysis (logrank test with Sidak correction, p o 0.05) and Cox regression (backward selection, p o 0.2) were used for time-to-event analysis; the assumption of hazards proportionality was tested by introducing terms of interaction with log(time) in the Cox model. Center-clustered data were fitted by the marginal model using the robust sandwich estimate. For regression modeling, missing variables were imputed using the multiple imputation approach (10 imputations, Markov chain Monte Carlo procedure).5 The percentage of missing variables and the handling of missing data within the UNOS data set have been reported elsewhere.6

Results Baseline demographics at transplantation and incidence of MCS use There were 1,780 female and 2,248 male patients; median age was 3 years (interquartile range [IQR] 0 to 12 years). Of the 4,028 patients, 710 (17.6%) were receiving MCS at the time of transplant—either VAD (398 patients, 9.9%) or

ECMO (317 patients 7.9%). Baseline demographics and clinical data are shown in Table 1. As shown in Figure 1, there was an increased use of VAD support at transplantation in the more recent era (17.9%; vs early 1.1%, intermediate 10.5%; p o 0.0001). Over time, patients successfully bridged to transplant have been progressively younger, and smaller (Table 2). Even among the smallest patients (o10 kg), in the recent era, the frequency of VAD support has remained similar to ECMO (10.3% vs 9.3%).

Types of VAD Before 2003, limited data are available on the type of VADs used. Comparing the recent (2008 to 2012) to the intermediate (2003 to 2007) eras, the type of VAD used has remained constant (Table 3; p ¼ not statistically significant [NS]). The use of the Berlin Heart EXCOR device has increased over time (early: 2, 11.8%; intermediate: 48, 38.4%; recent: 160, 56.9%; p o 0.0001) whereas use of the Thoratec paracorporeal VAD has decreased. The overall frequency of biventricular VAD support did not change over the study period (early 19.1%, intermediate 36.0%, late 31.2%; p ¼ NS). Compared with patients supported on other VADs, EXCOR patients were younger (odds ratio [OR] ¼ 3.7, 95% confidence interval [CI] 3.0 to 4.4 vs OR ¼ 14.0, 95% CI 13.4 to 14.6 years; p o 0.0001), and more likely to have congenital heart disease (23.7% vs 7.2%; p o 0.0001), require mechanical ventilation (24.2% vs 15.9%; p ¼ 0.037) or be in the ICU (80.0% vs 63.0%; p ¼ 0.0002).

In-hospital post-transplant complications Compared with ECMO, patients on VADs had a lower incidence of post-operative dialysis (5.4% vs 22.0%; OR ¼ 0.2, 95% CI 0.03 to 0.3), and drug-treated post-operative infection (33.0% vs 49.1%; OR ¼ 0.5, 95% CI 0.3 to 0.9). Stroke risk (4.9% vs 6.9%; OR ¼ 0.7, 95% CI 0.4 to 1.3) and need for permanent pacemaker (0.8% vs 2.3%; OR ¼ 0.3, 95% CI 0.09 to 1.3) were similar. The risk of stroke was higher in the VAD cohort than in nMCS patients (4.9% vs 2.3%; OR ¼ 2.2, 95% CI 1.3 to 3.7). When stratified by patient size, this was confined to the smallest (o10 kg) patients supported with VADs (Table 4). In contrast, ECMO patients had higher rates of stroke and post-operative dialysis across all patient sizes (Table 4). Complications, including renal dysfunction requiring dialysis (OR ¼ 15.9, 95% CI 12.0 to 21.0), post-operative infection (OR ¼ 3.1, 95% CI 2.3 to 4.1) and stroke (OR ¼ 4.0, 95% CI 2.6 to 6.2), were associated with higher post-transplant mortality.

Early mortality All patients Unadjusted post-transplant mortality was lowest in the VAD group (5.0%; OR ¼ 0.5, 95% CI 0.3 to 0.8) and highest in the ECMO group (27.4%; OR ¼ 5.1, 95% CI 3.8 to 6.7)

Davies et al. Table 1

VAD Use in Children

3

Baseline Pre-transplant Variables by Need for and Type of Mechanical Circulatory Support

Variable

Overall (n ¼ 4,028)

nMCS (n ¼3,318)

VAD (n ¼ 393)

ECMO (n ¼ 317)

p-value

Gender (male) Race White (non-Hispanic) African-American Hispanic Other Socioeconomic status (low) Waiting list Days on waiting list Days as Status 1/1A/1B Age at transplant (years) Neonate Infant 2–5 years 6–12 years Z13 years Weight at transplant (kg) Height at transplant (cm) BSA at transplant (m2) Etiology of heart failure Congenital heart disease Dilated cardiomyopathy Retransplant Other or unspecified Clinical status at listing Albumin o3.5 g/dl Creatinine clearance o40 ml/min Inotropic support Ventilator-dependent In intensive care unit Dialysis VAD at listing ECMO at listing Clinical status at transplant Creatinine clearance o40 ml/min Inotropic support Ventilator-dependent In intensive care unit Dialysis Donor/match variables Donor age (years) o13 years 13–50 years Z50 years Ischemic time (hours) Donor to recipient distance (miles)

2,248 (55.8%)

1,843 (55.6%)

228 (58.0%)

317 (55.8%)

NS o0.0001

2,297 808 328 595 1,170

1,902 646 263 507 960

183 112 50 48 126

212 50 15 40 84

(57.7%) (20.3%) (8.2%) (14.8%) (31.1%)

31 (11–70) 28 (10–62) 446 1055 837 697 993 13 92 0.57

(11.1%) (26.2%) (20.8%) (17.3%) (24.7%) (5.8–41) (62–152) (0.30–1.35)

1,711 1,833 185 299

(42.5%) (45.5%) (4.6%) (7.4%)

1,364 335 2,452 1,028 1,790 71 264 322

(58.1%) (19.7%) (8.0%) (15.3%) (30.9%)

31 (12–72) 29 (11–64) 379 918 664 560 797 12 89 0.54

(46.6%) (28.5%) (12.7%) (12.2%) (24.7%)

47 (23–95) 41 (19–80)

(66.9%) (15.8%) (4.7%) (12.6%) (28.9%)

11 (5–28) 10 (4–28)

NS o0.0001 o0.0001 o0.0001

(11.4%) (27.7%) (20.0%) (16.9%) (24.0%) (5.6–40.0) (61–150) (0.29–1.32)

3 52 92 85 161 32 142 1.15

(0.8%) (13.2%) (23.4%) (21.6%) (41.0%) (12–59) (85–165) (0.50–1.68)

64 85 81 52 35 9 76.5 0.42

(20.2%) (26.8%) (25.6%) (16.4%) (11.0%) (4.6–22) (56–120) (0.26–0.88)

1,494 1,404 160 260

(45.0%) (42.3%) (4.8%) (7.9%)

60 308 6 19

(15.3%) (78.4%) (1.5%) (4.8%)

157 121 19 20

(49.5%) (38.2%) (10.3%) (2.0%)

(50.1%) (11.1%) (60.9%) (25.5%) (77.9%) (1.8%) (6.6%) (8.0%)

1,032 245 2,005 714 1,529 37 85 75

(48.1%) (10.2%) (60.5%) (21.5%) (76.0%) (1.1%) (2.6%) (2.3%)

176 26 224 117 76 14 166 37

(50.4%) (7.0%) (57.0%) (28.8%) (85.4%) (3.6%) (62.9%) (9.4%)

156 64 222 197 185 20 13 210

(67.8%) (26.0%) (70.0%) (62.2%) (94.4%) (6.6%) (4.1%) (66.3%)

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

371 2,302 921 2,554 125

(10.1%) (57.8%) (23.1%) (64.3%) (3.2%)

269 1,922 612 1,960 60

(8.9%) (58.7%) (18.7%) (60.1%) (1.9%)

30 143 79 281 24

(8.0%) (36.4%) (20.1%) (71.5%) (6.2%)

72 237 230 313 41

(24.8%) (74.8%) (72.6%) (98.7%) (13.4%)

o0.0001 o0.0001 o0.0001 o0.0001 o0.0001

2,746 1,273 10 3.5 304

(69.0%) (30.8%) (0.2%) (2.8–4.3) (99–492)

2,308 1,004 6 3.6 304

(70.6%) (28.2%) (0.2%) (2.8–4.3) (96–495)

191 200 2 3.4 280

(48.6%) (50.9%) (0.5%) (2.8–4.2) (126–412)

247 68 2 3.5 340

(77.9%) (21.5%) (0.6) (2.8–4.2) (156–557)

o0.0001 o0.0001 o0.0001 o0.0001

o0.0001

NS o0.0001

For continuous variables, medians are given with interquartile range in parentheses; otherwise data are presented as number of subjects with parentheses values indicating incidence among patients with non-missing values. BSA, body surface area; ECMO, extracorporeal membrane oxygenation; nMCS, no mechanical circulatory support; NS, not statistically significant; VAD, ventricular assist device.

compared with patients who did not receive MCS (7.3%). In the multivariable analysis, VAD use overall was protective (OR ¼ 0.5, 95% CI 0.2 to 1.0), although extracorporeal VADs were associated with higher mortality (OR ¼ 3.0, 95% CI 0.8 to 10.6) (Table 5). Patients on VAD at transplant. Among VAD patients, those supported with extracorporeal devices had higher unadjusted post-transplant mortality (OR ¼ 3.8, 95% CI 1.2 to 12.5), whereas those with intracorporeal devices had a

trend toward lower mortality (1.2% vs 5.9%; OR ¼ 0.2, 95% CI 0.03 to 1.5). Multivariable analysis confirmed the negative impact of both extracorporeal devices (OR ¼ 5.9, 95% CI 1.5 to 23.7) and biventricular support (OR ¼ 2.9, 95% CI 0.9 to 8.6). Male gender was protective (OR ¼ 0.2, 95% CI 0.1 to 0.9). Other predictors of higher post-operative mortality in the model included: congenital heart disease (OR ¼ 2.7, 95% CI 0.9 to 8.3); need for ECMO in addition to a VAD (OR ¼ 6.0, 95% CI 1.3 to 28.1); creatinine

4

The Journal of Heart and Lung Transplantation, Vol ], No ], Month ]]]] differ and the relationship between the type of MCS and outcomes did not change across eras (Figure 2). In the two most recent eras, VAD patients exhibited a more significant decline in survival starting approximately 5 years posttransplant (Figure 2). Unadjusted post-transplant survival within the VAD group has not changed over time (Figure 2; p ¼ NS). In addition, there was no difference in unadjusted long-term post-transplant survival based on type of VAD used (intra-, para- or extracorporeal). Conditional long-term mortality among patients surviving to hospital discharge was not strongly affected by peritransplant clinical variables (e.g., renal insufficiency, mechanical ventilation) or the era of transplantation (Table 6).

Causes of death Primary causes of death among patients requiring VADs did not differ from those not requiring mechanical support (rejection 26.4%, graft failure 28.2%, infection 8.3%, cardiac arrest 12.8%, other cardiovascular causes 9.1%, multiple-organ failure 5.1%, non-compliance 1.3%).

Discussion

Figure 1 Percentage of transplanted patients supported with either (A) a ventricular assist device or (B) ECMO. Columns indicate percentage of all transplants, either overall (left) or in each era (right), by weight of recipient (white: o10 kg; gray: 10 to 30 kg; black: 430 kg). VAD use has increased over time in all weight categories (p o 0.0001). Within each era, increasing patient weight was associated with increasing VAD use (p o 0.0001).

clearance o40 ml/min (OR ¼ 7.7, 95% CI 2.0 to 28.7); and weight o10 kg (OR ¼ 5.0, 95% CI 1.6 to 15.7).

Patients o10 kg Neither the use of a VAD nor VAD type was a significant predictor of early mortality among patients o10 kg. In this group, the strongest predictors of early mortality included: a donor:recipient weight ratio o0.7 (OR 3.1, 95% CI 1.3 to 7.3); ischemic time 46 hours (OR ¼ 3.1, 95% CI 1.7 to 5.5); creatinine clearance o40 ml/min (OR ¼ 3.2, 95% CI 2.3 to 4.4); ECMO (OR ¼ 2.9, 95% CI 2.0 to 4.3); need for mechanical ventilation (OR ¼ 2.1, 95% CI 1.5 to 2.9); and a previous sternotomy (OR ¼ 2.0, 95% CI 1.1 to 3.6) (refer to Table S1 in supplementary material, available online).

Long-term mortality ECMO patients had decreased graft survival primarily due to higher early failure (Figure 2). Late graft failure did not

In this study we have demonstrated that children supported by a VAD have better post-transplant survival than patients on ECMO or those without mechanical support. This represents an improvement over earlier results where VAD and nMCS patient survival rates were equivalent.3,4 The threshold for VAD implantation in children has traditionally been high due to the heightened thromboembolic risk as well as a lack of familiarity with mechanical support among pediatric practitioners. Although these results do not account for the important risks due to device implantation, they suggest that lowering the threshold for VAD implantation may improve post-transplant survival. Traditionally, VADs have been compared with ECMO regarding their effectiveness both in bridging patients to transplantation and altering rates of post-transplant survival.7 Our data reaffirm a growing body of literature demonstrating that VAD support is superior to ECMO for several possible reasons, including: the need for sedation and mechanical ventilation; the need for higher levels of anti-coagulation; the persistent systemic inflammatory response; and possibly the deleterious effects of non-pulsatile flow on renal perfusion in small children.1,4,7 All of these factors likely contribute to the higher incidence of renal failure among ECMO patients, the lower likelihood that patients initially listed on ECMO will have resolution of either renal insufficiency or respiratory failure,8 and the higher incidence of post-operative renal failure and stroke among patients bridged on ECMO. Post-transplant complication rates are also higher among ECMO patients across all sizes (although at smaller sizes, the risk of stroke and PPM implantation were similar). In particular, the risks of stroke and post-operative dialysis are exceedingly high, corroborating the concept that ECMO provides inadequate support for end-organ function and that ECMO-supported patients are not optimized before

Davies et al. Table 2

VAD Use in Children

5

Baseline Pre-transplant Variables by Era Among Patients Supported by VADs at Transplant Overall (n ¼ 398)

Variable Gender (male) Race White (non-Hispanic) African-American Hispanic Other Socioeconomic status Low Middle High Status at transplant 1A Age at transplant (years) Neonate Infant 2–5 years 6–12 years Z13 years Weight at transplant (kg) Height at transplant (cm) BSA at transplant (m2) Etiology of heart failure Congenital heart disease Dilated cardiomyopathy Retransplant Other or unspecified Clinical status at transplant Creatinine clearance o40 ml/ min Inotropic support Ventilator-dependent In intensive care unit Dialysis Donor/match variables Donor age (years) o13 years 13–50 years Z50 years Ischemic time (hours)

Early, 1995–2002 (n ¼ 20)

232 (58.3%)

18 (90.0%) 9 4 0 7

68 (56.7%)

146 (56.6%)

50 39 9 22

127 69 41 21

186 112 50 50

(46.7%) (28.1%) (12.6%) (12.6%)

129 150 79 382 10 3 52 94 85 166 33.5 143 1.18

(35.1%) 7 (36.8%) 37 (35.2%) (43.5%) 10 (52.6%) 45 (42.9%) (21.5%) 2 (10.5%) 23 (21.9%) (96.0%) 16 (80.0%) 114 (95.0%) (2–15) 16 (12.5–18) 12 (2–16) (0.8%) 0 (0.0%) 0 (0.0%) (13.1%) 0 (0.0%) 12 (10.0%) (23.9%) 1 (5.0%) 28 (23.3%) (21.4%) 4 (20.0%) 22 (18.3%) (35.7%) 15 (75.0%) 59 (49.1%) (12.0–59.0) 63.5 (41.4–71.0) 42.3 (12.1–63.0) (86.4–176.0) 169.4 (157.0–175.0) 151.0 (88.0–170.2) (0.51–1.68) 1.72 (1.40–1.84) 1.32 (0.51–1.77)

60 311 6 21

(15.1%) (78.1%) (1.5%) (5.3%)

1 16 0 4

30 (7.9%)

(45.0%) (20.0%) (0.0%) (35.0%)

Intermediate, 2003–2007 Recent, 2008–2012 (n ¼ 120) (n ¼ 258) p-value

(5.0%) (80.0%) (0.0%) (15.0%)

2 (13.3%)

14 98 3 7

(41.7%) (32.5%) (7.5%) (18.3%)

(11.7%) (81.7%) (2.5%) (5.8%)

7 (6.0%)

85 105 54 252 9 3 40 66 59 92 28.8 130.5 1.03 45 197 3 15

0.06 NS

(49.2%) (26.7%) (15.9%) (8.1%) NS (35.8%) (43.0%) (22.1%) (97.7%) 0.0001 (1–14) o0.0001 (1.2%) (15.5%) (25.6%) (22.9%) (35.6%) (11.3–57.8) o0.0001 (77.9–162.6) 0.0003 (0.47–1.61) 0.0008 0.13 (17.4%) (76.4%) (1.2%) (5.8%)

21 (8.5%)

148 79 283 24

(36.7%) (19.9%) (71.1%) (6.1%)

9 3 15 0

(45.0%) (15.0%) (75.0%) (0.0%)

47 26 86 6

(39.2%) (21.7%) (71.7%) (5.0%)

90 50 182 18

(34.9%) (19.4%) (70.5%) (7.0%)

191 205 2 3.4

(48.0%) (51.5%) (0.5%) (2.8–4.2)

3 16 1 3.8

(15.0%) (80.0%) (5.0%) (2.5–4.6)

49 71 0 3.3

(40.8%) (59.2%) (0.0%) (2.7–4.0)

139 118 1 3.5

(53.4%) (46.2%) (0.4%) (2.8–4.2)

NS NS NS NS NS 0.002

NS

For continuous variables, medians are given with interquartile ranges in parentheses; otherwise data represent number with values in parentheses indicating incidence among patients with non-missing values. BSA, body surface area.

transplantation. Although the use of ECMO as a bridge to transplant continues to be necessary in selected cases, especially those with congenital anatomy not amenable to VAD support, every effort should be made to avoid its use when other methods of mechanical support are possible. Table 3 Types of Ventricular Assist Device (VAD) Implanted in Pediatric Patients at Transplant by Era

Type of VAD

Intermediate era, 2003–2007

Recent era, 2008–2012

Extracorporeal Paracorporeal Intracorporeal Total artificial heart

6 95 20 2

19 199 58 1

(4.9%) (77.2%) (16.3%) (1.6%)

(6.9%) (71.8%) (20.9%) (0.4%)

Our data also demonstrate a less favorable outcome with the use of extracorporeal VADs and suggest that conversion to other modes of support may improve post-transplant outcomes. Extracorporeal VADs are often implanted for stabilization before implantation of a more permanent device, but they have been used for more extended periods in smaller patients.1 Although the relative impact of patient condition versus device limitations is unclear, waitlist mortality is as high as 40%.1 Consistent with adult data,9 we found that the poor survival with extracorporeal VADs extends to the post-transplant period. This may be related to both the pre-support clinical condition and the inherent limitations of the support provided by extracorporeal devices, including the more frequent need for an open chest and mechanical ventilation, as well as a limited ability to encourage ambulation and physical rehabilitation. In either

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

6

Table 4 Incidence of Post-operative Complications by Patient Weight and Type of Mechanical Circulatory Support at Transplantation, Odds Ratio for Complication is Given For VAD and ECMO Patients Using nMCS Patients as Reference o10 kg In-hospital complications Dialysis Drug-treated Infection PPM Stroke

VAD (n ¼ 91) 1.0 0.7 8.6 4.9

430 kg

10–30 kg

b

(0.4–2.6) (0.2–2.2)c (1.5–47.3)a (2.1–10.8)a

ECMO (n ¼ 182) 5.4 2.0 6.4 2.7

a,b

(3.6–8.2) (1.4–3.1)a,c (1.4–28.9)a (1.2–5.7)a

VAD (n ¼ 97)

ECMO (n ¼ 72)

c

a,c

1.2 0.9 0 2.0

(0.5–3.0) (0.2–4.1)c (NA)c (0.8–5.1)

3.2 3.2 2.1 3.3

(1.5–6.7) (1.7–6.0)a,c (0.4–9.6)c (1.3–7.9)a

VAD (n ¼ 205) 0.8 2.9 0.3 1.2

b

(0.4–1.6) (1.6–5.0) (0.0–2.0)c (0.5–3.5)b

ECMO (n ¼ 63) 5.5 3.6 1.8 4.5

(2.9–10.2)a,b (1.8–7.4)a (0.4–7.8)c (1.6–12.4)a,b

Data are shown as odds ratio (95% confidence interval), except where the incidence of a complication is zero and odds ratios cannot be calculated. NA, not available; PPM, permanent pacemaker implantation. a p o 0.05 (vs nMCS patients). b p o 0.05 (ECMO vs VAD patients). c p o 0.10 (ECMO vs VAD patients).

case, based on this observation, the use of extracorporeal devices to achieve stabilization seems appropriate, but it should be followed by timely conversion to an implantable device to improve outcomes. Furthermore, the poor outcomes among VAD patients with renal insufficiency suggests that additional time on the appropriate device can provide the necessary conditions to improve perfusion and resolve end-organ dysfunction before transplantation.10 Mechanical support should improve these parameters, and mitigate the high post-transplant risk associated with pre-transplant renal insufficiency.10,11 This Table 5 Patients

is especially important given the risk for worsening renal failure after initiation of calcineurin inhibitors.12 Although early mortality after transplantation is lower among VAD patients, reoperative sternotomy is associated with increased mortality. Thus, VAD implantation is not without risk. Explantation of paracorporeal devices in small patients presents major technical challenges during reoperative sternotomy. The cannula size relative to the pericardial space is larger than among older patients with an implantable, continuous-flow LVAD and smaller outflow grafts. Similar issues as well as fundamental differences in

Logistic Regression Model Predicting Post-transplant Mortality (Within 30 Days or Before Hospital Discharge) Among All

Variables Gender male Etiology of congenital heart disease Etiology of restrictive cardiomyopathy Etiology of dilated cardiomyopathy ABO type A ABO type B ABO type O ABO type AB Mechanical ventilation at transplant VAD at transplant ECMO at transplant Reoperative sternotomy Creatinine clearance o40 ml/min Total bilirubin o2.0 mg/dl Dialysis at transplant Status 1A at transplant Donor:recipient weight ratio 0.8–2.0 Donor:recipient weight ratio 42.0 Ischemic time o2 hours Ischemic time 4–6 hours Ischemic time 46 hours Intracorporeal VAD Extracorporeal VAD Isolated LVAD VAD support Biventricular VAD support Weight 430 kg Weight 10–30 kg Early era (1995–2002) Late era (2008–2012)

Odds ratio 0.77 1.86 2.04 1.00 2.94 2.86 3.20 1.00 2.22 0.46 2.73 1.70 2.57 1.66 1.62 0.74 1.00 2.20 1.00 1.67 3.19 1.00 2.99 1.00 2.36 0.70 1.00 1.39 1.00

95% CI

p-value

0.58–1.00 1.31–2.65 0.91–4.56 Reference 1.24–7.01 1.17–7.01 1.29–7.91 Reference 1.64–2.98 0.21–1.02 1.89–3.95 1.11–2.63 1.77–3.73 1.23–2.24 0.92–2.86 0.52–1.06 Reference 0.97–4.96 Reference 1.25–2.24 1.90–5.38 Reference 0.84–10.59 reference 0.89–6.26 0.49–1.01 Reference 0.97–1.98 Reference

0.0532 0.0006 0.0821 0.0148 0.0215 0.0118 o0.0001 0.0563 o0.0001 0.016 o0.0001 0.001 0.0972 0.1012 0.0579 0.0006 o0.0001 0.0898 0.0844 0.0544 0.07

Davies et al.

VAD Use in Children

7

Figure 2 Kaplan–Meier survival estimates after transplantation among (A) all patients, and then separated by era: (B) early, 1995 to 2002; (C) intermediate, 2003 to 2007; and (D) recent, 2008 to 2012. In all groups, ECMO patients had poorer survival than nMCS patients and VAD patients. Survival was equivalent between VAD and nMCS patients.

patients requiring an isolated LVAD versus biventricular support may contribute to the increased mortality among patients with biventricular VADs.13–16 Table 6

Notably, despite these technical challenges, the donor organ ischemic times do not differ between VAD and nMCS patients. Travel distances among VAD-supported patients

Cox Proportional Hazards Regression of Long-term Mortality Among Patients Surviving to Post-transplant Hospital Discharge

Variable Race African-American Race white ABO type O ABO type AB Age infant (2–12 months) Age 2–5 years Age 6–12 years Age 13–18 years Creatinine clearance o40 ml/min Donor age o13 years Donor age 13–50 years Donor age Z50 years Donor gender male Donor cause of death stroke Donor:recipient weight ratio o0.7 Donor:recipient weight ratio 0.7–2.0

Hazard ratio 2.1 1.0 1.1 1.0 1.2 1.3 0.7 1.0 1.2 0.4 1.0 2.1 1.1 1.3 1.5 1.0

95% CI 1.7–2.4 Reference 1.0–1.3 Reference 0.9–1.5 1.0–1.6 0.6–0.9 Reference 0.9–1.6 0.4–0.7 Reference 0.8–5.8 1.0–1.3 1.0–1.7 1.0–2.4 Reference

p-value o0.0001 0.08 0.17 0.08 0.01 0.12 o0.0001 0.14 0.15 0.04 0.06

8

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

are shorter, suggesting that transplant teams recognize the risks of longer ischemic times and are reluctant to accept organs from distant centers. In addition, because VAD patients are relatively stabilized, centers can afford to wait for more optimal donor offers. Consistent with previous findings,17 patients o10 kg exhibit the highest incidence of stroke. These patients are most likely to be supported with the Berlin Heart EXCOR. This may be related to technical challenges at the time of explantation, but is most likely due to thromboembolic events that are unrecognized or undocumented before transplantation, or the result of inaccurate coding of strokes in the UNOS data set. Results from the Berlin Heart EXCOR trials suggest that the risk of stroke is highest early after implantation, and that explantation and transplantation are not associated with stroke.18,19 Among patients o10 kg, VAD and ECMO support is associated with a higher incidence of requiring a permanent pacemaker after transplantation, although the absolute risk in VAD patients is only 2.3%. Plausible explanations for this include longer donor organ ischemic times and older donor age, but neither of these is borne out in our data. The need for post-operative MCS is not documented within the UNOS data set, but could be higher in patients requiring pre-operative support, adding the potential for mechanical injury to the sino-atrial node or conduction system. Further evaluation of more complete data sets is required to understand the association between pre-transplant MCS and the need for permanent pacemaker implantation. The present analysis confirms a late decrement in survival among VAD patients approximately 5 years after transplantation.4 We observed this difference in all eras, although it only reached statistical significance in the intermediate-era group. We previously speculated that this could be related to the older age of VAD patients and the well-described increased risk of non-compliance and graft loss among teenage recipients.4 However, higher mortality beyond 5 years is present in all age groups. This may reflect the greater degree of allosensitization seen in VAD patients,20 and a consequent higher risk of rejection, although, in multivariable analyses, VAD has not been associated with long-term graft loss. Alternatively, the late decrement in survival may reflect a combination of underlying factors (including recipient and donor age) known to influence long-term outcomes. Ongoing evaluation of late outcomes after successful mechanical bridge to transplantation is needed to understand the long-term consequences of VAD implantation.

details of the transplant procedure, and whether there is an increased risk of post-transplant MCS. Although posttransplant outcomes are improved among supported patients, complete evaluation of the use of VADs as a bridge to transplant also requires an understanding of the peri-implant complications and of waitlist survival with and without MCS. A complete analysis of waitlist survival is beyond the scope of the present study, which focused on post-transplant outcomes, and peri-implant complications are not available within the UNOS data set. Individual institutional series or a combination of multiple data sets may be required to assess these issues completely. Without details related to the duration of mechanical support and the clinical status at implantation, it remains difficult to identify which patients are most likely to benefit from MCS. This evaluation of the current status of patients transplanted after institution of mechanical support demonstrates that patients bridged to transplantation on VADs are increasingly younger and smaller and are being supported with a changing spectrum of devices. Early post-transplant outcomes among VAD patients are excellent, continue to improve in the current era, and are better than the outcomes in unsupported patients. However, devices used for temporary support, including both ECMO and extracorporeal VADs, are associated with a higher early mortality, suggesting these patients may benefit from transition to a longer term device. The impact of bridging to transplantation using VADs on long-term outcomes beyond 5 years remains less clear. Further investigations addressing long-term impact, including allosensitization and rejection, will be needed to fully understand the impact of VAD implantation. Both size and age of VAD-supported patients continue to decrease, but complication rates among patients o10 kg remains high and an optimal strategy for mechanically supporting the smallest patients remains elusive.

Disclosure statement The authors have no conflicts of interest to disclose. We thank UNOS for supplying the data for our analysis. The content of this article is the responsibility of the authors alone and does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products or organizations imply endorsement by the U.S. Government. This work was funded in part by the Health Resources and Services Administration (Contract No. 231-00-0115) and departmental funding sources.

Supplementary data Limitations The limitations of the UNOS data set in terms of missing data and the yearly (rather than event-driven) follow-up data have been described previously.21,22 Of particular relevance to this analysis is the lack of data collection outside of listing, transplant and post-transplant follow-up. Had this information been available, it would have been valuable to understand the timing of VAD implantation, technical

Supplementary data associated with this article can be found in the online version at www.jhltonline.org/.

References 1. Chen JM, Richmond ME, Charette K, et al. A decade of pediatric mechanical circulatory support before and after cardiac transplantation. J Thorac Cardiovasc Surg 2012;143:344-51.

Davies et al.

VAD Use in Children

2. Brancaccio G, Amodeo A, Ricci Z, et al. Mechanical assist device as a bridge to heart transplantation in children less than 10 kilograms. Ann Thorac Surg 2010;90:58-62. 3. Blume ED, Naftel DC, Bastardi HJ, et al. Outcomes of children bridged to heart transplantation with ventricular assist devices: a multiinstitutional study. Circulation 2006;113:2313-9. 4. Davies RR, Russo MJ, Hong KN, et al. The use of mechanical circulatory support as a bridge to transplantation in pediatric patients: an analysis of the United Network for Organ Sharing database. J Thorac Cardiovasc Surg 2008;135:421-7. 5. Ambler G, Omar RZ, Royston P. A comparison of imputation techniques for handling missing predictor values in a risk model with a binary outcome. Stat Methods Med Res 2007;16:277-98. 6. Davies RR, Russo MJ, Morgan JA, et al. Standard versus bicaval techniques for orthotopic heart transplantation: an analysis of the United Network for Organ Sharing database. J Thorac Cardiovasc Surg 2010;140:700-8. 7. Fraser CD, Jaquiss RDB, Rosenthal DN, et al. Prospective trial of a pediatric ventricular assist device. N Engl J Med 2012;367:532-41. 8. Davies RR, McCulloch MA, Haldeman S, Pizarro C. Improving outcomes in children requiring mechanical bridge-to-transplantation (BTT) in the current era. Paper presented at: 33rd Annual Meeting of the International Society for Heart and Lung Transplantation; 2013 Apr 24-27; Montreal, Canada. 9. Russo MJ, Hong KN, Davies RR, et al. Posttransplant survival is not diminished in heart transplant recipients bridged with implantable left ventricular assist devices. J Thorac Cardiovasc Surg 2009;138: 1425-32: e1-3. 10. Sahney S, Chinnock R. Management of infants and young children with combined heart and kidney failure. Pediatr Transplant 2006;10:408-12. 11. Davies RR, Russo MJ, Mital S, et al. Predicting survival among highrisk pediatric cardiac transplant recipients: an analysis of the United Network for Organ Sharing database. J Thorac Cardiovasc Surg 2008;135:147-55.

9 12. Di Filippo S, Cochat P, Bozio A. The challenge of renal function in heart transplant children. Pediatr Nephrol 2007;22:333-42. 13. Holman WL, Kormos RL, Naftel DC, et al. Predictors of death and transplant in patients with a mechanical circulatory support device: a multi-institutional study. J Heart Lung Transplant 2009;28:44-50. 14. Morales DLS, Almond CSD, Jaquiss RDB, et al. Bridging children of all sizes to cardiac transplantation: the initial multicenter North American experience with the Berlin Heart EXCOR ventricular assist device. J Heart Lung Transplant 2011;30:1-8. 15. Hetzer R, Potapov EV, Stiller B, et al. Improvement in survival after mechanical circulatory support with pneumatic pulsatile ventricular assist devices in pediatric patients. Ann Thorac Surg 2006;82:917-24. 16. Fan Y, Weng Y-G, Huebler M, et al. Predictors of in-hospital mortality in children after long-term ventricular assist device insertion. J Am Coll Cardiol 2011;58:1183-90. 17. Polito A, Netto R, Soldati M, et al. Neurological complications during pulsatile ventricular assistance with the Berlin Heart EXCOR in children: incidence and risk factors. Artificial Organs 2013;37: 851-6. 18. Almond CS, Morales DL, Blackstone EH, et al. Berlin Heart EXCOR pediatric ventricular assist device for bridge to heart transplantation in US children. Circulation 2013;127:1702-11. 19. Byrnes JW, Prodhan P, Williams BA, et al. Incremental reduction in the incidence of stroke in children supported with the Berlin EXCOR ventricular assist device. Ann Thorac Surg 2013;96:1727-33. 20. Mahle WT, Tresler MA, Edens RE, et al. Allosensitization and outcomes in pediatric heart transplantation. J Heart Lung Transplant 2011;30:1221-7. 21. Davies RR, Russo MJ, Yang J, et al. Listing and transplanting adults with congenital heart disease. Circulation 2011;123:759-67. 22. Davies RR, Russo MJ, Reinhartz O, et al. Lower socioeconomic status is associated with worse outcomes after both listing and transplanting children with heart failure. Pediatr Transplant 2013;17:573.

Ventricular assist devices as a bridge-to-transplant improve early post-transplant outcomes in children.

The use of ventricular assist devices (VADs) to bridge pediatric patients to transplant or recovery has been expanding. There are few current pediatri...
579KB Sizes 0 Downloads 3 Views