Survival and Risk Factors for Death After Cardiac Transplantation in Infants
A Multi-institutional Study
Background Despite the increasing application of cardiac transplantation in infants, reported survival rates vary, and risk factors for death are poorly understood.
Methods and Results To examine early survival and risk factors for death in infants (<1 year of age) undergoing cardiac transplantation, 141 infants (36 <1 month of age) underwent primary cardiac transplantation between January 1, 1993, and January 1, 1995, at 23 centers in the Pediatric Heart Transplant Study (PHTS). Diagnoses were hypoplastic left heart syndrome (66%), other congenital heart disease (17%), cardiomyopathy (14%), and other (3%). Actuarial survival after cardiac transplantation was 84% at 1 month, 70% at 1 year, and 69% at 2 years, with the greatest hazard for death within the first 3 months. The principal cause of death was early graft failure in 20 patients (52% of deaths), infection in 10 (26% of deaths), and rejection in 4 (10%). On the basis of multivariate analysis, risk factors for early mortality were history of previous sternotomy (P=.0003), nonidentical blood type donor (P=.01), recipient non–blood group A (P=.02), and donor cause of death other than closed head trauma (P=.04). Diagnosis at listing, waiting time (mean, 1.3 months), graft ischemic time (mean, 228 minutes; range, 68 to 479 minutes), and recipient ventilatory or inotropic support at listing were not predictive for mortality after transplant.
Conclusions The higher mortality rate observed with infant heart transplantation is due to a higher mortality within the first month after transplantation as a result of early graft failure. Strategies to improve donor heart function at implantation would have the greatest impact on survival after infant cardiac transplantation.
The use of cardiac transplantation as therapy for otherwise untreatable heart disease in infants has occurred for >25 years.1 Within the past 10 years, the routine use by some centers of primary infant transplantation for such congenital lesions such as hypoplastic left heart syndrome has greatly increased the frequency of the procedure.2 3 4 Data from the registries of UNOS5 and ISHLT6 indicate that nearly one third of all heart transplantations performed between the ages of 0 and 18 years occur when the recipient is <12 months of age.
Infant heart transplantation has many different characteristics from transplantation in adults and older children, including (1) a markedly increased proportion of transplantations performed for congenital heart disease,5 6 (2) more frequent use of greater ranges of donor/recipient size mismatch as well as longer graft ischemic times,7 (3) more frequent use of noninvasive techniques (as opposed to endomyocardial biopsy) as the primary surveillance technique for cellular rejection,8 and, unfortunately, (4) an increased mortality rate.5 6 Some centers,2 4 however, report infant survival rates comparable to those of older children and adults. An analysis of risk factors for mortality and morbidity in infant heart transplantation by individual centers has been performed,9 10 but it has been limited due to the small numbers of infant transplantations performed at an individual center.
The PHTS was initiated as a multi-institutional effort to overcome these limitations. The purpose of this study was to examine survival and risk factors for death in infants (age <12 months) undergoing cardiac transplantation.
The PHTS is an event-driven database that was established in January 1993 by investigators from 23 pediatric heart transplant centers in the United States and Canada (see “Appendix 1”). Data are prospectively compiled on the clinical and demographic variables of donors and recipients as well as post-transplantation events such as rejection, infection, death, and retransplantation. Data were collected with coded event forms and forwarded to the PHTS Center at the University of Alabama-Birmingham, where the data were entered into a computer, checked with range checks and other statistical methods, verified, and analyzed. Data from a particular institution were included in a particular analysis only if complete; suspicious data entries were verified and corrected by participating institutions. All deaths were reviewed, and a specific cause of death was assigned for each case.
In this study, we examine the entire primary heart transplantation experience of the PHTS participating institutions from January 1, 1993, to December 31, 1994. During this period, 141 infants (age <12 months) underwent primary cardiac transplantation. This group represents >50% of all worldwide infant heart transplantations that were performed during this time period.6 The mean age at time of transplantation was 3.1 months, with 121 <6 months old and 36 <1 month old at the time of transplantation. Mean waiting time for an organ was 1.3 months (longest wait, 5.6 months). Cardiac diagnosis leading to transplantation was hypoplastic left heart syndrome in 55%, other congenital heart disease in 27%, cardiomyopathy in 11%, and assorted other diagnoses in 6%. Twenty-eight of the 38 recipients (74%) with congenital heart disease other than hypoplastic left heart syndrome had had previous surgical intervention before transplantation. There is a single UNOS status for infants <6 months of age; of the 20 infants >6 months of age, 10 were status 1 and 10 were status 2 at the time of transplantation. Mean follow-up for this cohort was 11.7 months (range, 0.1 to 24 months).
Donor size was frequently greater than recipient size, with a mean donor-to-recipient weight ratio of 1.65 (range, .65 to 4.2). Long donor ischemic times were also common (mean, 230 minutes; range, 68 to 479 minutes). One hundred two of 141 (73%) and 38 of 141 donor hearts were blood type identical or nonidentical to recipients, respectively. Blood type was not specified in one recipient.
Survival data were analyzed with the use of standard Kaplan-Meier actuarial techniques for estimation of survival probabilities. Parametric methods in the hazard domain were used to estimate time-related probabilities of death.11 Possible risk factors for death were initially examined with simple contingency tables and survival curves. The specific variables examined are listed in “Appendix 2.” Although panel-reactive antibody of HLA antigens was reported from each institution, the institutional variation in measurement technique precluded analysis of this variable for the cohort. A multivariate analysis in the hazard function domain was performed to identify specific risk factors associated with death.11 Variables were retained in the final risk factor analysis if P≤.05.
Survival at 1 month for the entire group of 141 patients was 84% (Fig 1⇓). Survival declined to 73% by 6 months with little change in mortality at 12 (70% survival) and 24 (69% survival) months after transplantation. Hazard function analysis (Fig 1⇓) revealed a single rapidly declining hazard with greatest risk during the first month after transplantation. By 6 months, the instantaneous risk for death was very low, but a further gradual fall in the instantaneous risk of death was seen to 24 months.
Cause of Death
There were a total of 38 deaths in the study population. Twenty-four of these 38 occurred within 1 month after transplantation (Table 1⇓). Early graft failure (11 patients), graft failure believed to be secondary to right ventricular failure from pulmonary hypertension (4 patients), sudden cardiac death on the day of transplantation (2 patients), and death from nonspecific graft failure on the day of transplantation (1 patient) accounted for 75% of these early deaths. For this study, early graft failure was defined as cardiac failure within 1 month after transplantation not attributed to rejection or surgical technical problems. The most common cause of death 1 month after transplantation was infection (8 patients), followed by acute rejection (3 patients).
Hazard function analysis of risk of death from infection peaked at 1.5 months after transplantation, and the instantaneous risk for death from infection was low by 6 months after transplantation.
Autopsies were obtained in 30 of the 38 deaths (79%). Rejection was confirmed by autopsy in all 4 of the deaths attributed to rejection. Nineteen of the 24 infants (79%) whose deaths were attributed to graft failure underwent an autopsy, and rejection was excluded in each of these cases.
Risk Factors for Death
Significant risk factors for death by multivariate analysis (Table 2⇓) included history of previous sternotomy at time of listing (P=.0003), nonidentical blood type donor (P=.01), recipient non–blood group A (P=.02), and donor cause of death other than closed head trauma (P=.05). Thus, most of the demographic variables (time on waiting list, age, gender), donor variables (ischemic time, type of preservation solution, duration of cardiopulmonary resuscitation), recipient variables (diagnosis, need for ventilatory support, need for inotropic support), donor/recipient mismatch variables (weight, age, gender, CMV serology), and institutional variables (transplant volume, rejection diagnosis policy) that were analyzed were not significantly associated with mortality in this cohort of infants undergoing cardiac transplantation.
The effects of previous sternotomy at listing and cause of death of the donor had their main impact on survival within 1 month after transplantation. Fig 2⇓ demonstrates that the 1-month survival rate for an infant without previous sternotomy was 87% versus a survival rate of 57% for an infant with a previous sternotomy at time of listing. Fig 3⇓ shows that survival rates after infant heart transplantation when the donor died of closed head trauma were 90% at 1 month compared with 79% for other causes of donor death.
The effects of blood type on survival are summarized in Fig 4⇓. An A blood type recipient receiving an A blood type donor had a good 12-month survival rate of 85%, whereas other identical matches (70%) and nonidentical matches (64%) had similar but lower 12-month survival rates. Thus, the benefit of an identical donor/recipient blood type match in this cohort appears to stem from the enhanced survival of the A blood type–to–A blood type group.
Cardiac transplantation in infants has been performed routinely in some institutions for a decade. Although the survival rates with infant transplantation have matched the survival rates of older children and adults in individual centers,2 4 large multi-center registries have demonstrated a lower survival rate after transplantation in infants. Data for 1988 to 1994 from UNOS5 cite survival rates of 79%, 68%, and 63% at 1 month, 1 year, and 2 years for 531 recipients <1 year of age compared with survival rates of 90%, 80%, and 73% for 6- to 10-year-old recipients and 93%, 84%, and 79% for 18- to 34-year-olds, respectively. Similar results were observed in the registry of ISHLT.6
In both registries, the difference in survival rates is mainly due to increased mortality during the first month after transplantation in the infant group. Our results confirm this discrepancy. Survival rates for infants in this study are 84% at 1 month, 77% at 3 months, 70% at 12 months, and 69% at 24 months. These survival rates are significantly (P<.0001) lower than the survival rates of PHTS children 1 to 18 years of age12 (93% at 1 month, 88% at 3 months, 82% at 12 months, and 81% at 24 months) and adult survival rates (94%, 90%, 85%, and 81%, respectively) from the Transplant Cardiologists Research Database,13 the adult counterpart to the PHTS. However, for the 1-month survivors for the entire group for infancy through adulthood, the 3-, 12-, and 24-month survival rates for infants (92%, 84%, and 82%, respectively) are similar (P=.4) to the survival rates of older children (94%, 88%, and 87%) and adults (96%, 90%, and 86%).
The results from this study indicate that the reason for the increased mortality rates in infants during the first month after transplantation is early failure of the graft, not acute rejection or infection. The newly transplanted heart may fail for a variety of reasons: primary myocardial failure from preoperative myocardial insult in the donor, perioperative preservation, reperfusion injury, and pulmonary hypertension in the recipient. All of these factors, either alone or in conjunction with each other, may contribute to development of early failure of the infant graft. Varying degrees of myocardial damage in infant donors before harvest have been documented by measuring elevations of the cardiac-specific protein cardiac troponin I14 that did not correlate with the echocardiographic appearance of cardiac function in the donor. Although some studies have suggested that graft dysfunction from pulmonary hypertension in the donor is less of a problem in the infant population than in older children,15 the majority of infant heart transplantations are performed for hypoplastic left heart syndrome palliated by keeping the ductus arteriosus widely patent. These patients will generally have systemic pulmonary artery pressures, but measurement of pulmonary pressures and pulmonary vascular resistance, unlike other older transplant candidates, is not routinely performed before transplantation in these infants. Postoperative catheterizations in such patients usually reveal normal pulmonary blood pressures, but there is pathological evidence16 that some of these patients may develop chronic pulmonary arteriopathy or veno-occlusive disease within the first months of life. Thus, these patients have the potential to develop increasing pulmonary vascular resistance during the course of their wait for a transplant, which might adversely affect the function of the graft in the immediate postoperative period.
Several of the risk factors associated with higher mortality in adult heart transplant recipients, such as congenital heart disease, duration of donor ischemic time, and ventilator dependence of the recipient, were not found to be risk factors for death in this infant cohort. Previous studies of infant recipients have not been able to demonstrate specific risk factors for death,9 although there is evidence, similar to findings in adults, that increasing experience with infant heart transplantation leads to increasingly better results.10 Additional experience and time of follow-up within the PHTS are needed to confirm this finding in the PHTS cohort.
In this cohort, a history of previous sternotomy at the time of listing was a strong risk factor for death after transplantation and has been demonstrated to be a risk factor for death in the adult Transplant Cardiologists Research Database Group cohort.13 Previous sternotomy in itself likely does not increase the risk of death after transplantation but rather is a marker for a previous unsuccessful attempt for surgical palliation or repair of the original cardiac lesion. Anatomic and/or hemodynamic problems leading to the failure of the initial procedure, as well as critical illness in an infant after a recent surgical procedure, could individually or collectively increase the risk of death after transplantation because of the status of the recipient as well as the tendency to loosen the criteria for an acceptable donor for a transplant candidate thought to be near death. The presence of a 43% 1-month mortality rate in infants with a history of previous sternotomy suggests that the risks of cardiac transplantation in an infant after previous, unsuccessful open heart surgery must be given careful consideration, especially given the long waiting times and substantial mortality that occur while waiting in the infant population17 as a whole.
The reason for the apparent advantage of having an A blood type recipient receive an A blood type donor is unclear. The UNOS registry6 has found A blood type recipients to have better survival. However, in the adult Transplant Cardiologists Research Database Group cohort, O blood type donor to O blood type recipient carried a better chance for survival.13 It is possible that specific blood group compatibility is not a factor in improved survival but rather is only a marker for a group of patients who come to transplantation less ill and/or receive healthier grafts.
Closed head trauma as a donor cause of death was associated with a greater chance of survival in this study. Previous work14 has found a relationship in infants between degree of donor myocardial injury as determined by cardiac troponin I levels and cause of death in the donor. In that study, sudden infant death syndrome was associated with higher levels of cardiac troponin I in the donor before harvest and subsequent mortality from primary graft failure in the recipient.
This study has the intrinsic limitations of a multicenter data registry: specifically, the lack of standardization in what can be diverse and varied methods of pretransplantation, perioperative, and post-transplantation management among the participating centers of infant transplantation. Although this cohort represents the majority of infants who received a transplant around the world during the study period, its size remains relatively small. Therefore, with more time and increasing numbers of infant heart recipients enrolled, the PHTS may identify other factors that influence the short- and intermediate-term survival rates after infant heart transplantation.
In summary, this study has demonstrated that the higher mortality associated with heart transplantation in infants compared with older children and adults is the result of a higher prevalence of death from early graft failure in this group. It appears that the ongoing mortality rate for 1-month survivors of infant heart transplantation is similar to that of older children and adults. Interestingly, the looser acceptable criteria for infant donors, such as the routine use of grafts with long (>4 hours) ischemic times, prolonged cardiopulmonary resuscitation, and high (>2/1) donor/recipient size mismatches compared with adults, were not associated with an increased risk of mortality, confirming earlier studies.9 10 Further investigation will be needed to design strategies to improve early donor heart function in infant recipients and thus improve overall survival.
Selected Abbreviations and Acronyms
|ISHLT||=||International Society of Heart and Lung Transplantation|
|PHTS||=||Pediatric Heart Transplant Study|
|UNOS||=||United Network of Organ Sharing|
The Pediatric Heart Transplant Study consists of the following participants: Elizabeth Frazier, MD, Arkansas Children’s Hospital (Little Rock); Robert Scott Appleton, MD, Cardinal Glennon Children’s Hospital, St Louis, Mo; W. Robert Morrow, MD, Children’s Hospital of Michigan (Detroit); Claudio Ramaciotti, MD, Children’s Hospital of Philadelphia, Philadelphia, Pa; F. Jay Fricker, MD, Children’s Hospital of Pittsburgh, Pittsburgh, Pa; Vincent R. Zales, MD, Children’s Memorial Hospital, Chicago, Ill; Richard Sterba, MD, Larry Latson, MD, and Charles Fraser, MD, Cleveland Clinic Foundation, Cleveland, Ohio; Linda J. Addonizio, MD, Columbia Presbyterian Medical Center, New York, NY; Randall L. Caldwell, MD, Indiana University (Indianapolis); Richard E. Chinnock, MD, Loma Linda University Medical Center, Loma Linda, Calif; William Moskowitz, MD, Medical College of Virginia (Richmond); Anthony Rossi, MD, Mount Sinai Medical Center, New York, NY; Robert Shaddy, MD, Primary Children’s Medical Center and the University of Utah (Salt Lake City); Stephen Bash, MD, St Francis Medical Center, Peoria, Ill; Charles E. Canter, MD, St Louis Children’s Hospital, St Louis, Mo; Daniel Bernstein, MD, Stanford University Medical Center, Palo Alto, Calif; Jeffrey Towbin, MD, Texas Children’s Hospital and Baylor College of Medicine, Houston, Tex; David C. McGiffin, MD, University of Alabama at Birmingham; Juan Alejos, MD, University of California at Los Angeles; Mark M. Boucek, MD, University of Colorado (Denver); Mitchell Madden, MD, University of Florida/Shands Hospital (Gainesville); W. Steve Ring, MD, University of Texas Southwestern Medical School (Dallas); Debra A. Dodd, MD, Vanderbilt University Medical Center, Nashville, Tenn.
PHTS Variables for 1993 to 1994: Outcomes After Transplantation in Infants
Variables examined in multivariate hazard analyses for death after transplantation and logistic multivariate analyses for early graft failure include the following:
Age, gender, race, height, weight, body surface area, months on waiting list.
Previous sternotomy, number of previous sternotomies, previous thoracotomy, number of previous thoracotomies.
Age, gender, race, height, weight, body surface area, ischemic time, chest compressions (CPR), CMV serology, cause of death, mechanism of death, circumstances of death, echocardiogram abnormalities, hypertension, diabetes, inotropes/pressors, blood type, cardioplegia/myocardial protection solution.
Cytomegalovirus serology, race, gender, age, body surface area, weight, blood type.
Etiology, status at listing and at transplantation, details of status at listing and at transplantation (inotropes, intra-aortic balloon pumping, ventilator, prostaglandin infusion–dependent extracorporeal membrane oxygenator), CMV serology, failure to thrive, history of arrhythmia, history of ventricular tachycardia, creatinine at listing and at transplantation, blood type.
Cardiac index, pulmonary capillary wedge pressure, pulmonary vascular resistance, systemic vascular resistance.
Transplant volume, rejection diagnosis policy.
- Received September 23, 1996.
- Revision received January 9, 1997.
- Accepted February 2, 1997.
- Copyright © 1997 by American Heart Association
Canter CE, Moorhead S, Huddleston CB, Spray TL. Restrictive atrial septal communication as a determinant of outcome of cardiac transplantation for hypoplastic left heart syndrome. Circulation. 1993;88(suppl II):II-456-II-460.
1995 Annual Report of the US Scientific Registry for Transplant Recipients and the Organ Procurement and Transplantation Network: Transplant Data: 1988-1994. Richmond, Va: UNOS/Rockville, Md: Division of Transplantation, Bureau of Health Resources Development, Health Resources and Services Administration, US Department of Health and Human Services; 1995.
Chinnock RE, Larsen RL, Emery JR, Bailey LL. Pretransplant risk factors and causes of death or graft loss after heart transplantation during early infancy. Circulation. 1995;92(suppl II):II-206-II-209.
Shaddy RE, Naftel DC, Kirklin JK, Boyle G, McGiffin DC, Towbin J, Ring WS, Pearce B, Addonezio L, Morrow WR. Outcome of cardiac transplantation in children: survival in a contemporary multi-institutional experience. Circulation. 1996;94(pt 2):II-69-II-73.
Bourge RC, Naftel DC, Costanzo-Nordin MR, Kirklin JK, Young JB, Kubo SH, Oliveri MT, Kasper EK. Pretransplantation risk factors for death after heart transplantation: a multi-institutional study: the Transplant Cardiologists Research Database Group. J Heart Lung Transplant.. 1993;12:549-562.
Grant JW, Canter CE, Spray TL, Landt Y, Saffitz JE, Ladenson JH, Jaffe AS. Elevated donor cardiac troponin I: a marker of acute graft failure in infant heart recipients. Circulation.. 1994;90:2618-2621.
Morrow R, Naftel D, Chinnock R, Canter C, Boucek M, Zales V, McGiffen D, Kirklin J, and the Pediatric Heart Transplant Study (PHTS) Group. Outcome of listing for cardiac transplantation in infants (age <6 months): predictors of death and interval to transplant. J Heart Lung Transplant.. 1995;14:558.