Renal Transplantation in Children —

A Report of the North American Pediatric Renal Transplant Cooperative Study

List of authors.
Paul T. McEnery, M.D., Donald M. Stablein, Ph.D., Gerald Arbus, M.D., and Amir Tejani, M.D.

Chronic Renal Disease

in children is frequently due to a congenital renal lesion, such as aplasia or dysplasia, which leads to
end-stage renal disease in early infancy. Such children have retarded growth and renal osteodystrophy.
Dialysis therapy tends to exacerbate renal bone disease and leads to early epiphyseal closure. To avoid this
additional injury, physicians frequently prefer to perform renal transplantation as soon as possible, even
before dialysis is initiated. Because of the small number of procedures performed annually at individual
centers, data on the outcome of renal transplantation in children are sparse. The North American Pediatric
Renal Transplant Cooperative Study was organized in 1987 to register and follow children up to 17 years of
age in the United States and Canada who receive renal allografts. This report is based on the 1550 patients
who underwent 1667 transplantation procedures that were reported to the registry during its first four years
of operation (1987 through 1990).

Methods
The North American Pediatric Renal Transplant Cooperative Study is made up of a clinical coordinating
center, a data-coordinating center, and 73 medical centers treating children with renal disease in the United
States and Canada. The data for this report, compiled in February 1991, include transplantations reported
during the four preceding years. Since January 1987 each allograft received by a child or adolescent ≤17 years
of age at a participating center has been reported to the study registry, along with information on graft
function and therapy one month after transplantation and every six months thereafter, as described
previously.1 Standard univariate and multivariate statistical methods, including product—limit estimates of
survival distributions, were used to analyze the data. Proportional-hazards survival models were constructed
that equated an individual patient's hazard to an underlying hazard multiplied by an estimated exponentiated
linear combination of risk factors. Multivariate models were scaled so that risk increased with larger values of
the covariates; the relative risk for a single dichotomous risk factor is the exponentiated parameter.

Results
CHARACTERISTICS OF THE PATIENTS
Table 1.
Characteristics of Pediatric Renal-Transplant Recipients, According to Year of Transplantation.
Table 2.

Characteristics of pediatric Renal-Transplant Recipients, According to Age at the Time of Transplantation.

Both the number of patients who received transplants during each of the four years and the number of
transplantation procedures have decreased slightly each year since 1987, although a lag in reporting probably
accounts for the small size of the 1990 group (Table 1). The age, race, and sex distribution of the patients did
not change significantly over time. The youngest patient who underwent transplantation was five months old.
Among the patients five months to five years of age, 70 percent were boys; the sex ratio was more nearly
equal in the two older age groups, which included 76 percent of the patients (Table 2).
The most common causes of renal failure were congenital lesions (renal dysplasia, obstructive
malformations, or both) in 42 percent of the patients and glomerulonephritis in 18 percent (including focal
segmental glomerulosclerosis in 12 percent). Lupus nephritis (5 percent) and hemolytic—uremic syndrome (3
percent) were rare, and only one patient had diabetic nephropathy. Of the patients 5 years of age or younger,
46 percent had congenital lesions, whereas various forms of glomerulonephritis — such as focal segmental
glomerulosclerosis, membranoproliferative glomerulonephritis, and lupus nephritis — were the most frequent
causes of renal failure among those 13 to 17 years of age. Although nonwhite patients made up 31 percent of
all transplant recipients, 44 percent of the patients with focal segmental glomerulosclerosis were nonwhite.
At the time of entry into the registry, 14 percent of the patients had just received their second or
subsequent transplants, with a median of 46 months (range, 5 to 178) since the previous transplantation.
Transplantation was used as initial therapy (without dialysis) in 22 percent of the patients; 34 percent of the
transplants from living related donors and 12 percent of those from cadavers were received by such patients.
The rate of transplantation as initial therapy was similar in all four age groups (≤1, 2 to 5, 6 to 12, and 13 to 17
years). Among the patients treated by dialysis, the median length of time from the initiation of dialysis to
transplantation was 12 months (mean, 21). All native renal tissue was removed in 29 percent of the patients,
and the existing grafts were removed in 62 percent of the patients who had undergone a previous
transplantation.

CHARACTERISTICS OF DONORS AND HLA MATCHING

 The patient's parent was the source of the allograft in 624 (37 percent) of the transplantations; in 94 (6
percent) the kidney came from a sibling or other living relative. Six of the sibling donors were under 18 years
of age, and three were identical twins of the patients; the youngest of these pairs of twins was 13 years old. In
949 (57 percent) of the transplantations, the kidney was obtained from a cadaver; 39 percent of the cadaver
donors were 10 years of age or younger. In the case of allografts from cadavers, 82 percent were maintained
with an iced electrolyte perfusion; the cold-storage times of the grafts were at least 24 hours in 56 percent of
the cases and more than 48 hours in 1 percent.
The percentage of recipients who received transfusions of blood from the donor of the allograft
decreased during the four years, from 43 percent in 1987 to 19 percent in 1990. Twenty-nine percent of the
patients who received kidneys from living related donors and 53 percent of those who received cadaver
kidneys received more than five transfusions of blood from unidentified donors. Among the recipients of
kidneys from living related donors, 87 percent had at least one haplotype match with the donor (HLA-B and
HLA-DR), whereas 12 percent had only one HLA-B or one HLA-DR antigen match. Matches of all A, B, and DR
alleles occurred in only 2 percent of the cases of transplantation of a kidney from a cadaver.

MEDICATION AFTER TRANSPLANTATION

The median doses of prednisone received by the patients were 0.31 mg per kilogram of body weight
per day 6 months after transplantation and 0.19 mg per kilogram per day after 30 months. The percentage of
transplant recipients who received prednisone therapy on alternate days increased from 12 percent to 36
percent during the same period. There was little change in the proportion of patients who received
prednisone, cyclosporine, and azathioprine at each follow-up evaluation. The dose of azathioprine was
constant throughout the study period (mean, 1.7 mg per kilogram per day); the median dose of cyclosporine
was 6.1 mg per kilogram per day at 6 months and 4.1 mg per kilogram per day at 30 months. The proportion
of recipients of kidneys from living related donors who received triple-drug therapy at six months increased
each year; 54 percent of those who underwent transplantation in 1987 and 87 percent of those who received
transplants in 1990 received all three drugs. For the recipients of kidneys from cadavers, the comparable
figures were 69 percent and 81 percent. The dose of each of the three maintenance immunosuppressive
medications correlated positively with age and weight, although the patients who weighed less received more
medication per kilogram of body weight.
One month after transplantation, 72 percent of the patients were receiving antihypertensive therapy,
but this percentage decreased to 53 percent at 30 months. Similarly, 59 percent initially received prophylactic
antibiotic therapy, and 31 percent continued to receive it at 30 months; such treatment was more frequent
among patients who had had obstructive lesions (40 percent).

REJECTION
A total of 1707 episodes of rejection (defined as the initiation of antirejection therapy or graft failure
due to rejection) were reported in 966 recipients, of whom 404 had 2 or more episodes of rejection
(maximum, 7). Overall, half of all transplant recipients had had an episode of rejection by 66 days after
transplantation (median, 187 days for recipients of grafts from living related donors and 39 days for recipients
of cadaver kidneys). At the end of the second year, 40 percent of the recipients of kidneys from living related
donors and 27 percent of the recipients of cadaver kidneys had not had an episode of rejection.
 The younger patients were at no disadvantage with respect to the length of time to the first episode of
rejection. The perioperative use of antithymocyte globulin—antilymphocyte globulin or OKT3 monoclonal
antibody was associated with a significantly longer time to the first episode of rejection among all kidney-
transplant recipients, regardless of the source of the graft. The recipients of kidneys from donors five years of
age or younger had a higher relative risk of rejection (1.5; 95 percent confidence interval, 1.3 to 1.6; P<0.001)
than those who received kidneys from older donors; 18 percent of the recipients of kidneys from cadaver
donors five years of age or younger had had no rejection episodes after one year, as compared with 35
percent of the recipients of kidneys from donors more than five years of age.
Overall, 54 percent of the episodes of rejection were completely reversed, 38 percent were partially
reversed, and 8 percent ended in graft failure or death. The rates of complete reversal declined with an
increasing number of rejection episodes, from 62 percent for the first episode to 30 percent when four or
more episodes had occurred. Treatment with antithymocyte globulin-antilymphocyte globulin or OKT3
antibody at the time of transplantation did not affect the probability that a later episode of rejection would be
completely reversed. Among the 28 patients who received OKT3 antibody at the time of transplantation and
again for the treatment of an episode of rejection, 18 had complete reversal and 8 partial reversal of the
rejection episode. OKT3 antibody was given during 469 (27 percent) of the episodes of rejection, and
intravenous methylprednisolone during 1187 (70 percent), whereas dialysis was used during 242 (14 percent)
of the episodes.

GRAFT SURVIVAL
Table 3.

Causes of Graft Failure in Pediatric Renal-Transplant Recipients.

Of the 1667 grafts, 378 (23 percent) failed at some time during follow-up (Table 3). Forty-nine percent of the
graft failures were caused by rejection, including 26 percent caused by acute rejection episodes. In 7 percent
of the patients, recurrence of the original disease resulted in graft failure; the most common recurrent disease
was focal segmental glomerulosclerosis. Vascular thrombosis was the cause of graft failure in 15 percent of the
cases.2
Figure 1.

Graft Survival According to Source.

Table 4.

Risk of Graft Failure in Recipients of Renal Allografts from Cadavers, Based on a Proportional-Hazards Model.*
Figure 1 shows the rates of graft survival in the recipients of kidneys from cadavers and from living related
donors. Among recipients of cadaver kidneys, 74 percent of grafts survived at one year and 62 percent at
three years; among recipients of kidneys from living relatives, 89 percent of the grafts survived at one year and
80 percent at three years (P<0.001). For patients who were two years old or younger at the time of
transplantation and received kidneys from cadavers, the one-year graft-survival rate was only 46 percent. For
all patients who received grafts from cadaver donors less than six years old, the two-year graft-survival rate
rose only to 57 percent. Proportional-hazards analysis identified several factors that were associated with an
increased risk of graft failure among recipients of cadaver kidneys (Table 4). Among these children, 74 percent
of those who did not undergo dialysis in the first week after transplantation had functional grafts at two years,
as compared with 52 percent among those in whom a delay in graft function caused by acute tubular necrosis
necessitated dialysis during the first week. Among children 2 to 17 years old who received kidneys from living
relatives, the graft-survival rate for various age groups (2 to 5, 6 to 12, and 13 to 17 years) ranged from 81 to
85 percent. In those less than one year old at the time of transplantation, the two-year graft-survival rate was
71 percent. The only factor associated with an increased risk of graft failure in the recipients of kidneys from
living related donors was acute tubular necrosis in recipients less than one year old (relative risk, 2.4; 95
percent confidence interval, 1.1 to 5.4; P<0.01).

GROWTH OF CHILDREN
Figure 2.

Mean (±SE) Change in Standardized Height Scores for Patients with Graft Function at Each Follow-up Interval,
According to Age at Transplantation.
At the time of transplantation, the mean height for all patients was 2.2 SD below the appropriate age-
and sex-adjusted mean for normal children and adolescents (z score). This deficit was comparable in both
sexes, although it was greater among younger patients ( — 2.8 SD for patients ≤5 years of age) and those with
previous transplantations ( — 3.2 SD). The mean height scores remained relatively constant after
transplantation in all but the two youngest groups of children (Fig. 2). The recipients who were 1 year old or
younger at the time of transplantation had an increase of 0.3 SD in height during the first 6 months after the
procedure; this increase was 0.8 SD by the 12th month. Accelerated growth did not continue, however, and
the z score for height changed little in subsequent months. An anticipated adolescent growth spurt or catchup
growth was not observed during the three years of observation after transplantation. In particular, patients
who were 13 years old or older at the time of transplantation had decreases in the mean z score. All but the
youngest age group had increases of 1.0 SD in weight scores by the 12th month after transplantation. After
two and three years, the mean weight values were comparable to those in normal children and adolescents.

MORBIDITY AND MORTALITY

The median length of hospitalization at the time of transplantation decreased from 19 to 16 days
during the study; the median stay was 2 days longer for the recipients of cadaver kidneys than for the
recipients of kidneys from living related donors. The number of hospital days after the procedure was
negatively correlated with the patient's age; the median hospital stay was 24 days for children 1 year old or
younger, 19 days for those 2 to 5 years old, 17 days for those 6 to 12 years old, and 15 days for those 13 to 17
years old. During the first one to five months after transplantation, 58 percent of the recipients were
rehospitalized (mean duration of hospital stay, nine days), 30 percent for symptoms of rejection. Bacterial
infection (12 percent), viral infection (12 percent), and hypertension requiring hospital treatment (8 percent)
were the other major causes of hospitalization. Both the frequency and the length of hospitalizations
decreased with increasing follow-up.
Cancer developed in 12 patients, of whom 6 had a lymphoproliferative disorder, 5 sarcoma, and 1 a
thyroid carcinoma. Eight of the patients with cancer died, five within a month of diagnosis. Four of these 12
patients had received more than one renal allograft.
Seventy-nine of the patients (including the 8 with cancer) died during the study period; infection was
the cause of death in 32 patients. In 38 patients the graft was reported to be functioning at the time of death.
Nineteen deaths occurred within 30 days after the initial transplantation during the study period, and 10
during the first postoperative week. At two years, 95 percent of the patients who received kidneys from living
related donors and 92 percent of those who received kidneys from cadavers were alive (P = 0.03). Sixteen (15
percent) of the patients one year old or younger died, nine of them with functioning grafts; they accounted for
20 percent of all deaths.

Discussion
The North American Pediatric Renal Transplant Cooperative Study was established to study
systematically the effects of renal transplantation in a large population of children and adolescents and to
generate data that could ultimately improve the probability of successful transplantation in such patients. The
problems of the adverse effects of immunosuppression, growth deficits, and questions about the optimal
timing of transplantation surgery are apparent in our results. These observations also highlight several factors
affecting prognosis that are different from those in adults.
The geometric mean hospital stay authorized for adults undergoing renal transplantation under the
Medicare system of prospective payment according to diagnosis-related groups was 15.8 days in 1987—1988
and 15.4 days in 19881989.3 The median length of the hospital stay after transplantation surgery in children
was 17.6 days, but the number of hospital days was negatively correlated with the patient's age; thus, the
median hospital stay for transplant recipients one year old or younger was 24.5 days.
In 1989 a total of 8882 renal transplantations were performed in adults, of which 1900 (21 percent)
used kidneys from living related donors.4 We found that 47 percent of the kidneys transplanted into children
in 1989 came from living related donors; the comparable figure was 43 percent during the four years of this
study. In 1983 through 1986, the European Dialysis and Transplant Association recorded 226 transplantations
of kidneys from living related donors (20 percent) among 1105 transplantations in children.5 Whether the
proportion of renal allografts from living related donors in children in North America is high because a smaller
number of kidneys from cadavers are allocated to children or because of the presence of a more altruistic
attitude than that in Europe is a matter for conjecture, but the high rate could be due to the knowledge that
the graft-survival rate for cadaver kidneys in children has been inferior to that for kidneys from living related
donors. In this study, the 89 percent graft-survival rate at one year for kidneys from living related donors
decreased to 80 percent by the third year, whereas the 74 percent one-year graft-survival rate for kidneys
from cadavers decreased to 62 percent at three years.
An age of two years or less in the recipient, more than 24 hours of cold storage, an age of five years or
less in the donor, black race, and delayed graft function due to acute tubular necrosis were associated with a
lower rate of graft survival for cadaver kidneys. Almost 40 percent of all cadaver kidneys were obtained from
donors under 10 years of age, with donors from 1 to 5 years old accounting for 25 percent of all such
transplants. The graft-survival rate for kidneys from cadaver donors five years of age or younger was lower
and the rejection rates were higher than when the donors were older. In addition, the incidence of graft
thrombosis was inversely related to the age of the donor.2 , 5 These results suggest that the current practice
of transplanting kidneys from very young donors into young recipients does not lead to optimal graft survival.
At the time of transplantation, the mean height deficit for all the recipients exceeded 2.0 SD, with the
youngest children having the greatest deficit. Deceleration of growth is more frequent than stable growth in
patients undergoing peritoneal dialysis or hemodialysis.6 , 7 The children who were one year old or younger
had an acceleration of growth soon after transplantation, but this group also had the worst long-term
outcome. In contrast, adolescent recipients (13 to 17 years of age) had a deceleration in growth after
transplantation. Our results suggest that delaying transplantation beyond the 12th year of age for children
who need a transplant is disadvantageous in terms of improvement in height. Thus, vigorous attempts should
be made to allocate more kidneys to children from 2 to 12 years of age. The process of allocating kidneys from
cadavers to children has been addressed by the United Network for Organ Sharing,8 which now adds points
for children 10 years old and younger on the waiting list. The results of this policy are not yet available.
An additional important result that emerges from this study relates to the role of immunosuppression
after transplantation. As in adults, sequential therapy after the transplantation of a cadaver kidney is
increasingly popular in children. The use of multiple drugs in children increases the risk of cancer as well as
that of infection, however. Of the deaths in this study group, 45 percent were attributable to infection. The
increase in triple-drug immunosuppression therapy —from 52 percent of the recipients of kidneys from living
related donors in 1987 to 80 percent in 1990 —indicates the need for a study of the effects of various
treatment regimens (with three drugs, two drugs, or lower doses of therapeutic agents) on graft survival and
patient growth.
 Abstract
BACKGROUND.
Previous studies of renal transplantation in children have focused on the survival of grafts and patients.
Little information is available about the cause of renal disease, the sources of donated organs, or children's
growth after transplantation. The North American Pediatric Renal Transplant Cooperative Study was organized
to identify the diseases that require transplantation and to analyze factors that affect the success of
transplantation in children.

METHODS.
We collected data from 73 pediatric transplantation centers from 1987 through 1990. These data
included information about demographic characteristics of patients, graft function, and therapy one month
after transplantation and every six months thereafter for each patient 17 years of age or younger.

RESULTS.
Altogether, 1550 children received 1667 renal allografts during this period; 31 percent of the children
were five years of age or younger. Forty-three percent of the transplanted kidneys came from a living related
donor, and 57 percent from a cadaver. The two most common causes of renal disease leading to
transplantation were congenital malformations of the kidneys and urinary tract (42 percent of the patients)
and focal segmental glomerulosclerosis (12 percent). One year after transplantation, the rate of graft survival
in recipients of a kidney from a living related donor was 89 percent; it was 80 percent after three years. For
recipients of cadaver kidneys, the comparable rates were 74 percent and 62 percent, respectively (P<0.001).
The best growth was observed in patients who were no more than five years old at the time of
transplantation. During follow-up, 79 patients died, and cancer developed in 12 patients.

CONCLUSIONS.
The most common causes of end-stage renal disease in children and adolescents are congenital
malformations of the kidneys and urinary tract and focal segmental glomerulosclerosis. The rates of graft
survival at one and three years are better in children and adolescents who receive a kidney from a living
related donor than in those who receive a kidney from a cadaver. (N Engl J Med 1992;326:1727–32.)