http://www.jhltonline.org

INNOVATION

Use of height and a novel echocardiographicmeasurement to improve size-matching for pediatricheart transplantationWarren A. Zuckerman, MD,a Marc E. Richmond, MD, MS,a Rakesh K. Singh, MD, MS,a

Jonathan M. Chen, MD,b and Linda J. Addonizio, MDa

From the aDivision of Pediatric Cardiology, and the bDivision of Pediatric Cardiothoracic Surgery, Columbia University Medical

Center, New York, New York

BACKGROUND: A major limitation of pediatric heart transplantation is scarcity of pediatric donororgans, leading to longer waiting times and higher waiting list mortality. Current practice is to matchpotential pediatric recipients with donors by weight; however, we hypothesize height to be a betterpredictor of heart size as estimated by left ventricular end-diastolic diameter (LVEDd), as well as anovel measurement from the superior vena cava–right atrium junction to inferior vena cava–right atriumjunction (SVC-IVC distance). Our ultimate objective is to present a more effective means of size-matching for pediatric heart transplantation.METHODS: Measurements of LVEDd and SVC-IVC distance were taken from 254 normal echocar-diograms performed on individuals aged 7 days to 22 years, and correlated with demographic variables,including height, weight, and body surface area. Simulations were conducted using echocardiographicmeasurements and size parameters of past recipients with hypothetic donors to demonstrate practicality.RESULTS: There was a linear relationship between height and SVC-IVC distance (R2 � 0.904) andLVEDd (R2 � 0.889), whereas the relationships with weight were logarithmic (SVC-IVC distance,R2 � 0.855; LVEDd, R2 � 0.880), and the relationships with body surface area were polynomial(SVC-IVC distance, R2 � 0.880; LVEDd, R2 � 0.884). Three simulations demonstrate improvementsin efficiency of the size-matching process.CONCLUSIONS: The use of height and a novel SVC-IVC distance measurement to evaluate heart sizein potential pediatric heart transplant recipients and donors may allow for broadening of the donor pooland creation of a more efficient and accurate size-matching process. The prospective evaluation of thesenovel methods with respect to clinical outcomes is necessary.J Heart Lung Transplant 2012;31:896–902© 2012 International Society for Heart and Lung Transplantation. All rights reserved.

KEYWORDS:heart transplantation;pediatrics;size-matching;echocardiography;donor pool

One of the major limitations of pediatric heart transplantationas a treatment for end-stage pediatric heart disease is the scarcityof available pediatric donor organs, leading to longer waitingtimes and, ultimately, higher waiting list mortality.1 The UnitedNetwork for Organ Sharing (UNOS) reported recipient waitingtimes and waiting list mortality stratified by age, status, and bloodtype. In 2005 to 2009 data, depending on blood type, the median

Reprint requests: Warren A. Zuckerman, MD, Columbia UniversityMedical Center, Division of Pediatric Cardiology, 3959 Broadway, BH-2N, New York, NY 10032. Telephone: 212-305-4436. Fax: 212-342-1443.

E-mail address: wz2116@columbia.edu

1053-2498/$ -see front matter © 2012 International Society for Heart and Lunghttp://dx.doi.org/10.1016/j.healun.2012.03.014

waiting times for status 1A pediatric patients approached 6months. The waiting list mortality for babies aged younger than 1year with blood type O was more than 18% by 3 months afterlisting (UNOS data source 040710–18).

In addition to medical advances that improve the care ofpatients on the waiting list for organ donation, one of the majorfocuses of pediatric heart transplant teams has been to maxi-mize the organ donor pool. Several methods have been used toaccomplished this, including increasing public awareness oforgan donation,2 the introduction of ABO incompatible trans-

plantation in infants and toddlers,3–5 the acceptance of longer

Transplantation. All rights reserved.

897Zuckerman et al. Size-matching for Pediatric Heart Transplant

ischemic times,6–10 and increasing the range of size-matchingby allowing for under-sizing and over-sizing of donor heartscompared with prior convention.11,12

Currently, the most common listing practice for potentialpediatric heart transplant recipients is to match by weight. Atthe Morgan Stanley Children’s Hospital of New York (MS-CHONY), all patients are listed by weight with an acceptabledonor range of approximately 0.8 to 2.5 times the recipient’sweight, and this range is often broadened if clinically indicated.Conventionally, the measurement that is used most often toestimate heart size is left ventricular end-diastolic diameter(LVEDd). However, with recent studies demonstrating mor-bidity and mortality benefits of bicaval anastomosis for hearttransplantation,13–16 the “height” of the donor heart, or distancebetween the junction of the right atrium with superior venacava and the junction of the right atrium with inferior venacava (SVC-IVC distance), may be more relevant when tryingto determine if a donor heart will fit into a recipient.

We hypothesize that in the healthy population, this novelmeasurement is more closely related to height than to weight orother patient demographic data. We also hypothesize thatheight is a better predictor of heart size using the more con-ventional measurement of LVEDd. The objective of the cur-rent study is to present a more effective means of size-match-ing for pediatric heart transplantation.

Methods

Approval from the Columbia University Institutional ReviewBoard was obtained before research activities. All completepediatric transthoracic echocardiograms coded for innocentmurmur with normal cardiac structure were identified via aquery of the echocardiogram database of MS-CHONY betweenMarch 2009 and February 2010. A brief review of the finalreport was conducted to confirm diagnosis, and the echocardio-gram was screened to ensure that the appropriate images wereobtained for measurement acquisition. Demographic data ob-tained for each individual included sex, age, weight, height,body surface area (BSA) calculated by Haycock formula, andsquare root of the BSA (sqrtBSA). An individual with multiplestudies performed during the queried period was only includedonce, using the earliest acceptable echocardiographic study.

Echocardiographic measurements

Three separate measurements of the SVC-IVC distance (Figure1) were performed on 3 separate beats in the sub-xyphoidshort-axis view by a single reviewer (W.Z.). In addition, 3measurements of the LVEDd were performed on 3 separatebeats in the standard parasternal short-axis M-mode. A meanwas calculated from the recorded SVC-IVC distance andLVEDd measurements, and these means were used for allfurther calculations.

Statistical analysis

Correlation analysis was performed for each of the 2 measure-

ments with the demographic variables, and a best-fit curve and

Pearson correlation coefficient were calculated using ExcelStatistics Add-in (Microsoft Corp, Redmond, WA).

Size-match simulations

To demonstrate the practicality of the findings from the initialportion of the study, simulations were conducted using echo-cardiographic measurements and size parameters of actual pastlisted recipients from our institution with hypothetical donorpatients.

Results

Echocardiographic review

Echocardiograms were deemed acceptable for inclusionin the study if 3 separate measurements of the SVC-IVCdistance were obtainable from 3 separate beats in thesub-xyphoid short-axis view. This resulted in the exclu-sion of nearly 30% of individuals. There were 254 ac-ceptable transthoracic echocardiograms reviewed on in-dividuals aged 7 days to 22 years. Of the 254 individuals,152 (60%) were female and 102 (40%) were male; 54(21%) were aged younger then 1 year old, 87 (34%) wereaged between 1 and 6 years, 64 (25%) were aged between6 and 12 years, and 49 (19%) were aged between 12 and22 years.

SVC-IVC distance was plotted against weight, height,age, BSA, and sqrtBSA. The relationship between SVC-IVC distance and weight was determined to be logarith-mic (R2 � 0.855; Figure 2A); the relationships betweenSVC-IVC distance and height (R2 � 0.904; Figure 2B),and also sqrtBSA (R2 � 0.879; Figure 2C), were linear;and the relationships between SVC-IVC distance and age(R2 � 0.874; Figure 2D), and also BSA (R2 � 0.880;Figure 2E), were best described by polynomial curves.LVEDd was also plotted against weight, height, age,

Figure 1 A still echocardiographic clip in the sub-xyphoid,short-axis view demonstrates novel superior vena cava (SVC)-inferior vena cava (IVC) distance (arrow).

BSA, and sqrtBSA. The relationship between LVEDd

898 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

and weight was logarithmic (R2 � 0.880; Figure 3A); therelationships between LVEDd and height (R2 � 0.889;Figure 3B), and sqrtBSA (R2 � 0.887; Figure 3C) werelinear; and the relationships between LVEDd and age(R2 � 0.842; Figure 3D), and BSA (R2 � 0.884; Figure

Figure 2 Novel superior vena cava (SVC)-inferior vena cava (Iis shown plotted against (A) weight, (B) height, (C) square root opatients, aged 7 days to 22 years.

3E) were polynomial.

Size-match simulations

Simulation 1The recipient is an 8.3-year-old child with dilated cardiomyopa-thy, with a weight of 25 kg (25th–50th percentile for age) and a

istance in centimeters, as measured on echocardiography images,surface area, (D) age, and (E) body surface area, in 254 control

VC) df body

height of 124 cm (10th–25th percentile for age). Conventionally,

ontrol

899Zuckerman et al. Size-matching for Pediatric Heart Transplant

our program would list with a weight range of 20 to 63 kg. Thesimulated potential donor is 17 years old with weight and heightat the 50th percentile for age (73 kg and 177 cm, respectively).Because the weight of the potential donor is outside the listingrange of the recipient, this organ would not be offered to thisrecipient. By plotting SVC-IVC distance and LVEDd againstheight (Figure 4), the donor will have an estimated SVC-IVC

Figure 3 Left ventricular end-diastolic diameter (LVEDd) in cenof body surface area, (D) age, and (E) body surface area in 254 c

distance of 6.5 cm and LVEDd of 5.0 cm. Measurements of these

dimensions on the recipient were SVC-IVC distance of 6.8 cmand LVEDd of 6.1 cm. By these measures, there is no significantsize mismatch, and if offered, this organ should be acceptable fortransplant into this recipient.

Simulation 2The recipient is a 2.7-year-old toddler with dilated cardio-

rs is shown plotted against (A) weight, (B) height, (C) square rootpatients aged 7 days to 22 years.

timete

myopathy, with a weight of 10.2 kg (� 5th percentile for

900 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

age) and a height of 88 cm (10th–25th percentile for age).Conventionally, our program would list with a weight rangeof 8 to 26 kg. In this case, the simulated potential donor isa 6.5-year-old obese (35 kg, 95th–99th percentile for age)child of average height (116 cm, 50th percentile for age). Asin Simulation 1, with the weight of the potential donorfalling outside the listing range of the recipient, this organwould not be offered to this recipient. By plotting SVC-IVCdistance and LVEDd against height (Figure 5), the donorwill have an estimated SVC-IVC distance of 4.5 cm andLVEDd of 3.7 cm. Measurements of these dimensions onthe recipient were SVC-IVC distance of 4.3 cm and LVEDdof 5.2 cm. Again, as in the first case, these measures dem-onstrate no significant size mismatch, and if offered, thisorgan should be acceptable for transplant into this recipient.

Simulation 3The recipient is a 1.1-year-old baby with noncompactioncardiomyopathy, with a weight of 7.3 kg (5th percentile forage) and a height of 68 cm (10th percentile for age). Con-ventionally, our program would list with a weight range of6 to 18 kg. The simulated potential donor in this case is a5.5-year-old tall (116 cm, 75th percentile for age), thin (17kg, 25th percentile for age) child. Because the weight in thissimulation falls inside the listing range of the recipient, this

Figure 4 (A) Superior vena cava-inferior vena cava (SVC-IVCshown plotted against height in a potential donor. For a height of 1to be 5.0 cm (Simulation 1).

organ would be offered to this recipient. By plotting SVC-

IVC distance and LVEDd against height (Figure 5), thedonor will have an estimated SVC-IVC distance of 4.5 cmand LVEDd of 3.7 cm. Measurements of these dimensionson the recipient were SVC-IVC distance of 3.5 cm andLVEDd of 4.4 cm. In this case, there is a significant sizemismatch, and by these measures, this organ would likelybe unacceptable for transplant into this recipient.

Discussion

UNOS data as of April 2010 report an unacceptably highmortality rate for pediatric patients on the waiting list forheart transplantation. For children ages 1 to 5 years andbabies aged younger than 1 year with blood group O, themortality rate by 12 months after listing was approximately15% and 20%, respectively. Therefore, recent focuswithin the pediatric heart transplant community has beentoward broadening the donor pool. Although changessuch as the introduction of ABO incompatible transplan-tation,3–5 the allowance of under-sized and over-sizeddonors by weight,11,12 and increasing organ donor aware-ness have helped considerably,2 there have been no ex-aminations of the process by which pediatric heart trans-plant recipients are listed for transplantation, which is

nce and (B) left ventricular end-diastolic diameter (LVEDd) are, SVC-IVC was estimated to be 6.5 cm and LVEDd was estimated

) dista77 cm

typically by the creation of a donor weight range.

901Zuckerman et al. Size-matching for Pediatric Heart Transplant

Our results demonstrate that perhaps height is a betterand more direct method of evaluating heart size than weight,because it correlates linearly to the LVEDd as well as to thenovel SVC-IVC distance. In addition, we suggest that byusing height to list pediatric patients for heart transplanta-tion and SVC-IVC distance to examine heart size of poten-tial recipients as well as donors, we may be able to broadenthe organ donor pool and create a more efficient and accu-rate size-matching process.

Although current convention is to list potential pediatricheart transplantation recipients by weight, with a suitabledonor weight range of approximately 0.8 to 2.5 times therecipient’s weight, this system leaves gaps in which poten-tially acceptable donor hearts are not offered to appropriaterecipients for size. These gaps may relate to both recipientand donor issues.

In Simulations 1 and 2, we present examples of donor-recipient matches that effectively expand the donor pool forthose recipients. The recipient in Simulation 1 has a bigheart, which is often the case in patients with dilated car-diomyopathy, and might be able to accommodate a heartfrom a larger donor. Patients with dilated cardiomyopathyoften have failure to thrive and poor weight gain, furtherwidening the gap between the recipient’s expected heart

Figure 5 (A) Superior vena cava-inferior vena cava (SVC-IVCshown plotted against height in a potential donor. For a height oestimated to be 3.7 cm (Simulations 2 and 3).

size for weight and the actual heart size.17

In Simulation 2, the potential donor is obese (95th–99thpercentile for age) but of average height. With currentconvention of listing recipient’s using a weight range, thisdonor would fall above that range. Once again, however, thepotential recipient has dilated cardiomyopathy and thereforeincreased heart size, and when using height as a means ofevaluating the potential donor, this would result in an ade-quate size match.

In addition, we demonstrate in our review of echocar-diograms a linear relationship between height and LVEDdand SVC-IVC distance, whereas the relationship betweenthose measurements and weight is logarithmic in nature.Therefore, the use of height as an estimate of heart sizewould be a more direct and potentially more accuratemethod.

We acknowledge that although our goal is to broaden thedonor pool, this novel measurement and method of evalu-ation for donor hearts will also on occasion result in the lossof a potential donor offer. Simulation 3 is an example ofsuch a situation: the potential recipient is a toddler, and thepotential donor is a tall, thin child. Using the current con-ventional listing method, the potential donor would fall intothe listing weight range. However, this potential donorwould more than likely be rejected given the significant

nce and (B) left ventricular end-diastolic diameter (LVEDd) arecm, the SVC-IVC was estimated to be 4.5 cm and LVEDd was

) distaf 116

differences in age and height. Therefore, although our pro-

902 The Journal of Heart and Lung Transplantation, Vol 31, No 8, August 2012

posed method of evaluation would result in this offer beinglost, this would most likely not result in the loss of an actualdonor, thus creating a more efficient process.

One of the limitations of the current study is that theechocardiograms were reviewed retrospectively, and there-fore, special attention was not paid to acquiring the idealimages for the non-standard SVC-IVC measurement. Inaddition, evaluation of the reliability of the SVC-IVC mea-surement was not conducted for the study. This study alsodid not evaluate other measurements of fit, such as cavaldiameter. In our center’s experience, diameter discrepancyis less of a concern than overall heart size when performingthe bicaval anastomosis. Although caval anastomotic steno-sis is a risk, it rarely causes immediate postoperative mor-bidity and has been easily addressed in the catheterizationlaboratory at a later time.

Future evaluations will include testing of the reliabilityof the SVC-IVC measurement. In addition, it is not entirelyclear that this novel measurement and evaluation processwould apply similarly to potential recipients with cardio-myopathy and large hearts, as well as to potential recipientswith congenital heart disease and small hearts, or to poten-tial recipients with and without failure to thrive. The appli-cability of the measurement and evaluation process will beexamined in recipients with different classes of heart dis-ease.

Finally, a retrospective outcomes review of prior pediat-ric heart transplants should be conducted in an attempt todetermine the ideal maximum amount of recipient–donormismatch using the proposed listing mechanism. We be-lieve that this study provides an excellent starting pointfor future evaluations of the applicability of this novelheart measurement and proposed pediatric transplant list-ing mechanism.

In conclusion, we demonstrate height is a better andmore direct method of evaluating heart size than weightbecause it correlates linearly to an established heart mea-surement (LVEDd) as well as to a novel heart measurement(SVC-IVC distance). We suggest that by using patientheight as well as the novel SVC-IVC distance to evaluateheart size in both potential recipient and donor patients, wemay be able to broaden the organ donor pool and create amore efficient and accurate size-matching process. Prospec-tive analysis of these novel methods and measurements withrespect to clinical outcomes are required and are a next steptoward the overall improvement of the size-matching pro-cess for pediatric heart transplantation.

Disclosure statementUNOS data source 040710-18 was supported in part by HealthResources and Services Administration contract 234-2005-370011C. The content is the responsibility of the authors alone anddoes not necessarily reflect the views or policies of the U.S.Department of Health and Human Services, nor does mention oftrade names, commercial products, or organizations imply en-

dorsement by the U.S. Government.

None of the authors has a financial relationship with a com-mercial entity that has an interest in the subject of the presentedmanuscript or other conflicts of interest to disclose.

References

1. Sweet SC, Wong HH, Webber SA, et al. Pediatric transplantation inthe United States, 1995-2004. Am J Transplant 2006;6:1132-52.

2. Policy statement—pediatric organ donation and transplantation. Pedi-atrics 2010;125:822-8.

3. West LJ, Karamlou T, Dipchand AI, Pollock-BarZiv SM, Coles JG,McCrindle BW. Impact on outcomes after listing and transplantation,of a strategy to accept ABO blood group-incompatible donor hearts forneonates and infants. J Thorac Cardiovasc Surg 2006;131:455-61.

4. West LJ, Pollock-Barziv SM, Dipchand AI, et al. ABO-incompatibleheart transplantation in infants. N Engl J Med 2001;344:793-800.

5. West LJ, Pollock-Barziv SM, Lee KJ, Dipchand AI, Coles JG, Ruiz P.Graft accommodation in infant recipients of ABO-incompatible hearttransplants: donor ABH antigen expression in graft biopsies. J HeartLung Transplant 2001;20:222.

6. Razzouk AJ, Chinnock RE, Gundry SR, et al. Transplantation as aprimary treatment for hypoplastic left heart syndrome: intermediate-term results. Ann Thorac Surg 1996;62:1-7; discussion 8.

7. Kawauchi M, Gundry SR, de Begona JA, et al. Prolonged preservationof human pediatric hearts for transplantation: correlation of ischemictime and subsequent function. J Heart Lung Transplant 1993;12:55-8.

8. Mitchell MB, Campbell DN, Clarke DR, et al. Infant heart transplan-tation: improved intermediate results. J Thorac Cardiovasc Surg 1998;116:242-52.

9. Canter C, Naftel D, Caldwell R, et al. Survival and risk factors fordeath after cardiac transplantation in infants. A multi-institutionalstudy. The Pediatric Heart Transplant Study. Circulation 1997;96:227-31.

10. Morgan JA, John R, Park Y, et al. Successful outcome with extendedallograft ischemic time in pediatric heart transplantation. J Heart LungTransplant 2005;24:58-62.

11. Fullerton DA, Gundry SR, Alonso de Begona J, Kawauchi M, Raz-zouk AJ, Bailey LL. The effects of donor-recipient size disparity ininfant and pediatric heart transplantation. J Thorac Cardiovasc Surg1992;104:1314-9.

12. Razzouk AJ, Johnston JK, Larsen RL, Chinnock RE, Fitts JA, BaileyLL. Effect of oversizing cardiac allografts on survival in pediatricpatients with congenital heart disease. J Heart Lung Transplant 2005;24:195-9.

13. Sun JP, Niu J, Banbury MK, et al. Influence of different implan-tation techniques on long-term survival after orthotopic heart trans-plantation: an echocardiographic study. J Heart Lung Transplant2007;26:1243-8.

14. Weiss ES, Nwakanma LU, Russell SB, Conte JV, Shah AS. Out-comes in bicaval versus biatrial techniques in heart transplantation:an analysis of the UNOS database. J Heart Lung Transplant 2008;27:178-83.

15. Davies RR, Russo MJ, Morgan JA, Sorabella RA, Naka Y, Chen JM.Standard versus bicaval techniques for orthotopic heart transplanta-tion: an analysis of the United Network for Organ Sharing database.J Thorac Cardiovasc Surg 2010;140:700-8, 708e 701-2.

16. Schnoor M, Schafer T, Luhmann D, Sievers HH. Bicaval versusstandard technique in orthotopic heart transplantation: a systematicreview and meta-analysis. J Thorac Cardiovasc Surg 2007;134:1322-31.

17. Alvarez JA, Orav EJ, Wilkinson JD, et al. Competing risks for deathand cardiac transplantation in children with dilated cardiomyopathy:results from the pediatric cardiomyopathy registry. Circulation 2011;

124:814-23.