HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Arman Kilic George J. Arnaoutakis Duke E. Cameron Luca A. Vricella Eric S. Weiss Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kilic, A. Articles by Weiss, E. S. Search for Related Content PubMed PubMed Citation Articles by Kilic, A. Articles by Weiss, E. S. Related Collections Transplantation - heart

---

Original Articles: Adult Cardiac

Adult-Derived Index for Mortality Prediction After Cardiac Transplantation (IMPACT) Risk Score Predicts Short-Term Mortality After Pediatric Heart Transplantation

^a Division of Cardiac Surgery, Department of Surgery, The Johns Hopkins Medical Institutions, Baltimore, Maryland
^b Division of Cardiothoracic Surgery, Department of Surgery, Columbia University College of Physicians and Surgeons, New York, New York

Accepted for publication December 12, 2011.

^_* Address correspondence to Dr Weiss, Division of Cardiothoracic Surgery, Columbia University College of Physicians and Surgeons, 177 Ft Washington Ave MG, New York, NY 10032 (Email: esw9006{at}columbia.edu).

Presented at the Fifty-eighth Annual Meeting of the Southern Thoracic Surgical Association, San Antonio, TX, Nov 9–12, 2011.

	Abstract

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
Background: We have previously derived and validated the Index for Mortality Prediction After Cardiac Transplantation (IMPACT), which predicts short-term mortality after adult orthotopic heart transplantation (OHT). This study evaluated the predictive accuracy of the IMPACT score in pediatric OHT.

Methods: The United Network for Organ Sharing registry was used to identify pediatric (< 18 years) OHT patients from 2000 to 2008. The IMPACT score was calculated for each patient. The association of IMPACT score with 1-year mortality was evaluated with univariate and multivariable logistic regression analysis. The correlation coefficient between predicted and actual 1-year mortality was determined for each IMPACT score. Kaplan-Meier survival estimates were calculated and stratified by IMPACT score.

Results: We identified 2,518 eligible pediatric OHT patients (1,128 girls [44.8%]). Mean IMPACT score was 10.3 ± 6.3 (range, 0 to 38). A total of 297 patients (11.8%) died within the first year after OHT. Each point increase in the IMPACT score increased the odds of 1-year mortality by 13% (odds ratio, 1.13; 95% confidence interval, 1.11 to 1.15; p < 0.001). The correlation coefficient between predicted and actual 1-year mortality was 0.93 (p < 0.001). One-year survival by disjoint categories of the IMPACT score was 0 to 4 (96.7%), 5 to 9 (92.9%), 10 to 14 (87.6%), 15 to 19 (81.3%), and 20 or more (64.2%; p < 0.001).

Conclusions: In this large-cohort analysis, the IMPACT score accurately predicted mortality following pediatric OHT. The IMPACT score could therefore be useful to providers for organ allocation and prognostication in this patient population.

	Introduction

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
Advances in surgical technique and immunosuppressive therapy have resulted in improvements in survival after orthotopic heart transplantation (OHT) in the pediatric population [1]. Despite this, OHT continues to be limited by the number of available donor organs. An accurate assessment of a potential candidate's risk of death after OHT is therefore crucial to ensure optimal utilization of the limited donor organ supply. Although risk indices have been used in other solid organ transplants, it was not until recently that a risk score for OHT was developed. Moreover, we have recently derived and validated the Index for Mortality Predication After Cardiac Transplantation (IMPACT), a risk score based on pretransplant recipient characteristics, that is highly predictive of short-term death after adult OHT (Table 1) [2]. It is unknown whether the IMPACT score also applies to the pediatric population, where donor organ shortages also exist and where risk indices would also be useful in guiding organ allocation and in prognostication. This study evaluated the predictive capability of the IMPACT score in assessing the short-term death risk after pediatric OHT.

	Patients and Methods

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

Data Source
The United Network for Organ Sharing (UNOS) database was used for this study. This registry contains recipient, donor, and transplant information for all thoracic organ transplantations performed in the United States. The database does not include any patient or center identifiers.

Study Design
All pediatric (< 18 years) OHTs performed during 2000 to 2008 were identified in the UNOS registry. The analysis excluded patients undergoing multivisceral transplantation and those undergoing retransplantation.

The individual components of the IMPACT score were tabulated in this pediatric cohort, and the IMPACT score was calculated for each recipient. Calculations for each of the covariates was similar to the adult population, except that the revised Schwartz equation was used to calculate creatinine clearance, and the Mosteller equation was used for calculating body surface area in pediatric patients [3–5]. The primary end point was all-cause mortality within 1 year of OHT.

Data Analysis
The IMPACT score was calculated for each eligible pediatric OHT recipient. The ability of the IMPACT score to predict death at 1 year was evaluated using logistic regression analysis, with the IMPACT score as a continuous variable as well as a categoric variable. We wanted to ensure that the IMPACT score was applicable in different recipient age cohorts within the pediatric population. Therefore, we secondarily stratified patients in our study sample into neonates or infants aged younger than 1 year and children older than 1 year, and evaluated the effect of the IMPACT score in these separate cohorts.

To ensure that the IMPACT score, which is based on recipient characteristics, was applicable across various donor strata, we secondarily evaluated its predictive capability across various combinations of donor age and ischemic time. The correlation coefficient (r2) between predicted 1-year mortality based on the IMPACT score and the actual, observed mortality for that IMPACT score was also calculated. Finally, we also evaluated the IMPACT score as a predictor of death at 30 days and 5 years ensure that its predictive ability was not limited to the 1-year period.

All frequency data are presented as number (percentage), and all continuous data as mean ± standard deviation. Statistical significance was defined for all analyses at the conventional two-tailed p value of less than 0.05. Odds ratios (OR) that were calculated in logistic regression analysis are presented with 95% confidence intervals (CI). All statistical analyses were performed with STATA 11 software (StataCorp LP, College Station, TX).

	Results

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
Study Population
We identified 2,518 eligible pediatric OHT recipients during the study period, of which 1,128 (44.8%) were girls. Mean age was 6.4 ± 6.1 years, with 714 patients (28.4%) aged younger than 1 year (Table 2). Indications for OHT included cardiomyopathy in 1,073 (42.6%) and congenital heart disease in 1,069 (42.5%). Although 513 recipients (20.4%) recipients were mechanically ventilated before OHT, only 14 (0.6%) had an intraaortic balloon pump. In addition, 202 (8.0%) required temporary circulatory support, and 144 (5.7%) were bridged with ventricular assist devices, most of which were older-generation pulsatile devices.

View larger version (4K): [in this window] [in a new window]   Fig 1. Distribution of Index for Mortality Prediction After Cardiac Transplantation (IMPACT) score in pediatric orthotopic heart transplant recipients.

View larger version (4K): [in this window] [in a new window]   Fig 2. Kaplan-Meier 1-year survival based on the Index for Mortality Prediction After Cardiac Transplantation (IMPACT) score.

In addition, to ensure that the significant effect of the composite score on death was not driven solely by donor factors, we examined the effect of the IMPACT score across various combinations of donor age and ischemic time. In all donor combinations using the respective median values of donor age and ischemic time as cut points, the IMPACT score continued to demonstrate predictive capability, with increasing scores consistently being associated with progressively worse 1-year survival (p < 0.001 for each strata; Table 5).

	Comment

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
The first year after OHT poses the highest risk of death in pediatric recipients, after which point long-term survival is favorable [6]. Indeed, infant OHT recipients who survive at least 1 year have a median survival of more than 20 years, and older children who survive to the 1-year mark have a median survival of 16 years [6]. In addition, survival rates have improved over time, principally due to a reduction in death in the early period after OHT [6]. This higher risk of death within the first year, coupled with the shortage of available donor organs, suggests that a risk index that could be used to stratify patients according to risk of short-term death would be a useful tool in organ allocation and prognostication. Accordingly, the aim of this study was to evaluate the utility of the previously adult-derived IMPACT score in predicting short-term death in pediatric recipients.

Individual Components of the IMPACT Score and Death
There were several important findings in this analysis. Foremost, several of the components of the IMPACT score had significant independent effects on death at 1 year. For instance, a high serum bilirubin concentration increased the odds of 1-year death, similar to a prior finding of the International Society for Heart and Lung Transplantation registry [7]. The indication for OHT also affected survival, such that recipients with congenital heart disease were at higher risk of death after OHT than those with cardiomyopathy. Congenital heart disease is a known risk factor for death, and especially early postoperative death, in both pediatric and adult populations [6, 8].

Female sex was also associated with increased death in our analysis. Although the sex effect on outcomes after cardiac operations is less clear in pediatric vs adult patients, prior studies have demonstrated worse survival in girls in this setting [9, 10]. African American race was also associated with increased death in our analysis, similar to findings from another multiinstitutional study [11]. Interestingly, reduced creatinine clearance values did not affect death after OHT in our study, although dialysis between listing and OHT did have a significant adverse effect. Dialysis between listing and transplant represents poor end-organ perfusion and has been linked to an increased risk of death, particularly in children on the OHT waiting list [12].

Although intraaortic balloon pump counterpulsation and newer generation ventricular assist devices predicted death perfectly (ie, all patients with those factors died after OHT), the number of patients with each of these factors was extremely low, and therefore, an adequate analysis would be difficult to conduct. Indeed, no significant effect on death was found with the older-generation devices that were used to bridge a significantly larger proportion of patients. Finally, mechanical ventilation and temporary circulatory support (including extracorporeal membrane oxygenation) before OHT significantly increased short-term death after OHT. The presence of these risk factors likely represent marginal cardiopulmonary reserve in the patient, and have indeed been shown in prior studies to adversely affect survival after transplant [13–15].

In addition to bridging with older-generation ventricular assist devices, recent infection and reduced creatinine clearance did not increase odds of short-term death in this pediatric cohort as it did initially in the adult cohort from which the IMPACT score was derived. Despite this, we wanted to assess the unmodified IMPACT score in pediatric OHT and thus still included these components.

Differences Between Pediatric and Adult Populations
The mean IMPACT score in our pediatric OHT cohort was more than 4 points higher than the adult population from which the score was derived (10.3 vs 6.1 points) [2]. This was due to several key baseline differences between the groups. For instance, there was approximately double the proportion of girls in the pediatric cohort, with significantly fewer white recipients, and a higher proportion recipients with congenital heart disease. The proportion of mechanical assistance was also significantly different: more pediatric patients required mechanical ventilation and temporary circulatory support, but fewer had ventricular assist device bridging and intraaortic balloon pump counterpulsation.

Clinical Utility of the IMPACT Score in Pediatric OHT Recipients
Several features should be evident in a clinically useful risk index. Foremost, the components of the risk index should be easily attainable by the clinician. In addition, there should be a broad range of scores within the population to be studied to ensure that low-risk and high-risk patients can be separated. Finally, the score should demonstrate significant clinical, in addition to statistical, differences in the outcome to be measured.

We felt the IMPACT score achieved these goals within pediatric OHT. Moreover, the components of the risk score represent data that are readily available to clinicians involved in pediatric OHT. Furthermore, as demonstrated in Figure 1, there was robust distribution of the IMPACT score in our study population, with a range of 0 to 38. In addition, as depicted in logistic regression and Kaplan-Meier analysis, there was a significant difference in survival according to risk score, such that patients with an IMPACT score of 0 to 4 had a 1-year survival of 96.7% vs 64.2% in those with a score of 20 or more, suggesting true clinical implications.

Limitations
This study has several key limitations. A principal limitation is that the IMPACT risk score was developed in the adult population and therefore may not represent the most optimal risk index for pediatric OHT recipients. For instance, several individual components of the risk index did not have a significant effect on death in the pediatric cohort in multivariable logistic regression analysis, such as recent infection or bridging with older-generation ventricular assist devices. Moreover, certain components of the risk score simply did not apply to the pediatric population, such as age older than 60 years.

In parallel to these limitations, certain factors that were not included in the IMPACT score might have significantly affected survival in pediatric OHT. Accordingly, their inclusion may have improved the predictive power of the risk index. This included age, pulmonary vascular resistance, and complexity of disease, including number of prior cardiac operations, all of which have been shown to influence survival after OHT in pediatric recipients [16, 17]. Despite these limitations, we applied the IMPACT score in an unaltered fashion to determine if it indeed had potential clinical utility in pediatric OHT.

Other limitations of this analysis included those inherent to a large, multiinstitutional data set such as the UNOS database. These include reporting biases and missing data, although the components of the IMPACT score were well reported in the pediatric cohort we examined. The UNOS database is also limited to transplantations in the United States and may therefore not be applicable to the international population.

Conclusions
We evaluated the predictive ability of the previously adult-derived and validated IMPACT score on short-term death after pediatric OHT. In this large-cohort analysis, the IMPACT score was indeed found to be highly predictive of death after OHT in pediatric recipients. Therefore, in addition to its role in the adult OHT population, the IMPACT score has potential clinical utility in organ allocation and prognostication after pediatric OHT.

	Discussion

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
DR CHARLES HUDDLESTON (St. Louis, MO): Thank you, President Coselli and Dr Jones. This is an excellent presentation, Dr Kilic. It gives us yet another tool to get a handle on, in a very objective way, what the risk of transplantation might be in any individual patient. First, a comment. I am fascinated by the fact that a simple blood test, that is, a serum bilirubin, has such wide application to risk across a number of transplant indices, however you want to look at it, including ventricular assist device implantation, adult heart transplantation, pediatric heart transplantation. This simple test probably costs 20 bucks.

Now, the question I have is what should we do with this information? It is a nice tool and how are we going to implement it? That is, are there patients with a certain Index for Mortality Prediction After Cardiac Transplantation (IMPACT) score that you would refuse transplantation for or that you would turn down on your transplant list?

DR KILIC: In reference to your first comment about serum bilirubin, it is truly impressive to see how strong of a predictor it is of mortality, not only in the pediatric population but also in the adult heart transplant population. In terms of your question, I think that it depends on the institution. Different institutions will be willing to assume different levels of patient risk for the patients that they transplant. However, I can say with our data analysis that really an IMPACT score of 20 or more, that group seems to be a group that is at very high risk for short-term mortality. Their 1-year survival was 64%, but maybe even more alarming is their 30-day survival was only 80%. So, 20% of those patients in that group will die within 1 month.

DR HUDDLESTON: So are you saying you would turn them down?

DR KILIC: Well, that is a difficult question, again, to answer. It depends on what the other options are, but this does suggest that there is a group that can be identified that is associated with a very high risk and maybe we need to do some further analysis to figure out whether or not those patients could benefit from some other type of therapy.

DR HUDDLESTON: The next question along this line is how can we impact the IMPACT score? Specifically, is it reasonable to use this as an indication for implantation of a ventricular assist device in order to lower the bilirubin by improving liver function or getting patients off dialysis or whatever it is that is impacting the score? Now, some of these things, we can't change such as the gender, well, not chromosomally at least, and we can't change whether they are African American or Caucasian, et cetera, but do you think it is reasonable to use the IMPACT score as an indication for a ventricular assist device (VAD) implant in these kids?

DR KILIC: I think that is a great comment and I think there may be potential utility there. We did not analyze to see whether or not there was prognostic utility within the VAD population. However, it does identify a subset of patients who are bridged with VADs, that if those patients have other high-risk factors that drive their score up, those are a group of patients that will be very high risk for mortality after heart transplant. So, again, we may need to do some further analysis to figure out what is the best treatment algorithm for those patients.

DR HUDDLESTON: I didn't quite understand the difference between temporary support and ventricular assist devices. You implied that temporary support was extracorporeal membrane oxygenation (ECMO), but you said that ECMO was included in that. What is the difference?

DR KILIC: We had defined those two categories originally in our derivation in the adult population. Temporary circulatory support was defined as either ECMO or extracorporeal ventricular assist device and the ventricular assist device category was defined as either paracorporeal or intracorporeal VAD.

DR HUDDLESTON: And then my last question has to do with the so-called conditional survival. For instance, in the International Society for Heart and Lung Transplantation database, they will take all the patients that survived to 1 year and then analyze their survival out from that. And it has been particularly interesting for the adults with congenital heart disease who have very high mortality in a 30-day period of time, but those who survive to a year actually do much better over the long haul than the ischemic and the cardiomyopathic patients in that particular age group. I wondered if you take the IMPACT score then and see whether that still had an impact for those who had already survived a year.

DR KILIC: That is a great question as well. Although the data were not shown here, we actually did look at 30-day and 5-year mortality in addition to the 1-year mortality data that I presented, and the IMPACT score is highly predictive in each of those categories. In addition, it is highly predictive of conditional survival. So if we look at patients who survive to 30 days, what the impact of the IMPACT score would be on their 1-year mortality is highly significant. I believe the odds ratio was 1.11, which was a p-value of less than .001. So it was highly significant.

DR MUHAMMAD MUMTAZ (Norfolk, VA): Is the impact of the IMPACT score different in the various diagnostic groups, congenital vs cardiomyopathy, in the pediatric population?

DR KILIC: It does have predictive capability within each of the cohorts. I am not sure if you remember the slide, but the congenital group automatically gets 5 points assigned. So they are automatically ahead of the game. And that is why in the pediatric population the mean IMPACT score is actually about 4 points higher than what we see in the adult population, because such a high percentage are congenital heart disease. But if we do stratify based on the two most common indications in the study sample, which were congenital and cardiomyopathy, within those groups, increasing IMPACT score is highly predictive of mortality.

	Acknowledgments

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 
This study was supported by Departmental Funds from the Department of Surgery, Johns Hopkins Hospital.

	References

Top Abstract Introduction Patients and Methods Results Comment Discussion Acknowledgments References

 

1. Irving CA, Webber SA. Immunosuppresion therapy for pediatric heart transplantation Curr Treat Options Cardiovasc Med 2010;12:489-502.[Medline]
2. Weiss ES, Allen JG, Arnaoutakis GJ, et al. Creation of a quantitative recipient risk index for mortality prediction after cardiac transplantation (IMPACT) Ann Thorac Surg 2011;92:914-921.[Abstract/Free Full Text]
3. Schwartz GJ, Munoz A, Schneider MF, et al. New equations to estimate GFR in children with CKD J Am Soc Nephrol 2009;20:629-637.[Abstract/Free Full Text]
4. Staples A, LeBlond R, Watkins S, Wong C, Brandt J. Validation of the revised Schwartz estimating equation in a predominantly non-CKD population Pediatr Nephrol 2010;25:2321-2326.[Medline]
5. Mosteller RD. Simplified calculation of body-surface area N Engl J Med 1987;317:1098.[Medline]
6. Kirk R, Edwards LB, Kucheryavaya AY, et al. The registry of the International Society for Heart and Lung Transplantation: fourteenth pediatric heart transplantation report–2011 J Heart Lung Transplant 2011;30:1095-1103.[Medline]
7. Boucek MM, Aurora P, Edwards LB, et al. Registry of the International Society for Heart and Lung Transplantation: tenth official pediatric heart transplantation report–2007 J Heart Lung Transplant 2007;26:796-807.[Medline]
8. Stehlik J, Edwards LB, Kucheryavaya AY, et al. The registry of the International Society for Heart and Lung Transplantation: twenty-eighth adult heart transplant report–2011 J Heart Lung Transplant 2011;30:1078-1094.[Medline]
9. Chang RK, Chen AY, Klitzner TS. Female sex as a risk factor for in-hospital mortality among children undergoing cardiac surgery Circulation 2002;106:1514-1522.[Abstract/Free Full Text]
10. Seifert HA, Howard DL, Silber JH, Jobes DR. Female gender increases the risk of death during hospitalization for pediatric cardiac surgery J Thorac Cardiovasc Surg 2007;133:668-675.[Abstract/Free Full Text]
11. Singh TP, Almond CS, Gauvreau K. Improved survival in pediatric heart transplant recipients: have white, black and Hispanic children benefitted equally? Am J Transplant 2011;11:120-128.[Medline]
12. Almond CS, Thiagarajan RR, Piercey GE, et al. Waiting list mortality among children listed for heart transplantation in the United States Circulation 2009;119:717-727.[Abstract/Free Full Text]
13. Patel ND, Weiss ES, Nwakanma LU, Scheel J, Cameron DE, Vricella LA. Effect of pre-transplant extracorporeal membrane oxygenation on pediatric heart transplantation outcomes: an analysis of the United Network for Organ Sharing database Circulation 2007;116:II-564.
14. Auerbach SR, Richmond ME, Chen JM, et al. Multiple risk factors before pediatric cardiac transplantation are associated with increased graft loss Pediatr Cardiol 2012;33:49-54.[Medline]
15. Gandhi R, Almond C, Singh TP, Gauvreau K, Piercey G, Thiagarajan RR. Factors associated with in-hospital mortality in infants undergoing heart transplantation in the United States J Thorac Cardiovasc Surg 2011;141:531-536.[Abstract/Free Full Text]
16. Zales VR, Pahl E, Backer CL, Crawford S, Mavroudis C, Benson Jr DW. Pharmacologic reduction of pretransplantation pulmonary vascular resistance predicts outcome after pediatric heart transplantation J Heart Lung Transplant 1993;12:965-972.[Medline]
17. Lamour JM, Kanter KR, Naftel DC, et al. The effect of age, diagnosis, and previous surgery in children and adults undergoing heart transplantation for congenital heart disease J Am Coll Cardiol 2009;54:160-165.[Medline]

This article has been cited by other articles:

				K. Howard-Quijano, J. C. Schwarzenberger, J. C. Scovotti, A. Alejos, J. Ngo, J. Gornbein, and A. Mahajan Increased Red Blood Cell Transfusions Are Associated with Worsening Outcomes in Pediatric Heart Transplant Patients Anesth. Analg., June 1, 2013; 116(6): 1295 - 1308. [Abstract] [Full Text] [PDF]

				A. Kilic, J. G. Allen, and E. S. Weiss Reply Ann. Thorac. Surg., July 1, 2012; 94(1): 334 - 334. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Add to Personal Folders Download to citation manager Author home page(s): Arman Kilic George J. Arnaoutakis Duke E. Cameron Luca A. Vricella Eric S. Weiss Permission Requests Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Kilic, A. Articles by Weiss, E. S. Search for Related Content PubMed PubMed Citation Articles by Kilic, A. Articles by Weiss, E. S. Related Collections Transplantation - heart