Effect of Recipient's Race on Pediatric Renal Allograft Survival: A Single-Center Study

(1)

Effect

of Recipient’s

Race

on Pediatric

Renal

Allograft

Survival:

A

Single-Center

Study

Seth L. Schulman, MD*; JoAnn Palmer, BSN*; Stephen Dunn, MDX,

Bruce A. Kaiser, MD*; Martin S. Polinsky, MD*; and

H. Jorge Baluarte, MD*

ABSTRACT. One hundred twenty-seven children (83

males, 44 females, 86 white, 41 nonwhite; mean age 12.1

years) who received 160 renal transplants between 1980

and 1989 were retrospectively studied. Variables such as

age, sex, primary diagnosis, type, HLA-DR mismatching,

and repeated transplants were compared between races

and found not to be significant. However, HLA-A and

-B cadaveric-graft mismatching, which was equivalent

between whites and nonwhites prior to 1985

(pre-cyclo-sporine A era), has significantly favored whites (49%

with 0 to 2 HLA-A and -B mismatch vs 16% in nonwhites)

since 1985 (P < .05), and a significantly higher proportion

of nonwhite patients (59%) were receiving medical

as-sistance (P < .0001). Graft survival was evaluated with

significantly poorer results in nonwhites as compared to

whites (P < .05). Although no difference was found

between white and nonwhite cadaveric-graft survival

before 1985, nonwhites had significantly worse graft

survival since 1985 (72% vs 59% for 1 year and 61% vs

24% for 3 years in whites and nonwhites, respectively; P

< .05). Subpopulations such as nonwhite adolescents,

nonwhite females, nonwhites with repeated transplants,

and all low socioeconomic patients were identified as

high-risk children with poor long-term survival. It is

concluded that secondary to poorer matching since 1985

there has been decreased graft survival in nonwhites

despite cyclosporine A. Attempts to improve matching

and attention to high-risk groups are needed for

equiv-alent survival. Pediatrics 1992;89:1055-1058; renal

trans-plantation, race, graft survival, HLA matching.

Although studied extensively in adults, the

so-called “racial effects or the impact of recipient race on

renal allograft survival remains unclear. In some

re-ports, apparent race-specific differences in allograft

survival among patient groups have been attributed

to the influence of HLA matching, medication

non-compliance, immunologic and “center> effects, and

socioeconomic status.’6 The presence of a racial effect

in the pediatric transplant population has not been

addressed in the literature to-date.

Consequently, we retrospectively evaluated

race-related differences in renal allograft survival for pe-diatric and adolescent recipients of renal transplants

at our center over a 10-year interval. The impacts of

From the Departments of >Pediatrics and Surgery, St Christopher’s Hospital for Children and Temple University School of Medicine, Philadelphia. PA. Received for publication Apr 19, 1991; accepted May 30, 1991.

Reprint requests to (H.J.B.) Section of Pediatric Nephrology, St Christopher’s Hospital for Children, Front St at Erie Ave. Philadelphia, PA 19134-1095. PEDIATRICS (ISSN 0031 4005). Copyright © 1992 by the American Acad-emy of Pediatrics.

allograft survival and race-specific differences after the initiation of cyclosporine A immunosuppression were also evaluated.

MATERIALS AND METHODS

Between January 1980 and December 1989, 127 children and adolescents received 160 renal allografts at St Christopher’s Hos-pital for Children. Of these, 1 1 1 allografts were transplanted into

white (69%), and 49 into nonwhite (31%) (39 African-American, 7

Latin-American, 1 Indian, 1 Cambodian, and 1 Jordanian) recipi-ents. Thirty-eight patients received living-related transplants, and

the remaining I 22 had cadaveric transplants. Other than for a medical contraindication, no patient was denied the opportunity to receive a transplant. Three patients with primary hyperoxaluria who had received a total of 5 kidney transplants were included;

those who had undergone combined liver-kidney transplantation

were not.

Data were analyzed for differences in graft survival by race, age, socioeconomic status, HLA-A, -B, and -DR mismatching, and donor

source. Graft loss was defined as follows: return to dialysis, whether

caused by rejection, suspected noncompliance, or recurrence of the original disease; patient death, with or without a functioning graft; or retransplantation with allograft function sufficient for avoidance of dialysis but not for growth (four children). Socioeconomic status

was simply defined by whether on not a patient’s family was

receiving public medical assistance.

Until 1985, all patients received only prednisone and

azathio-prime for maintenance immunosuppression. Rejection episodes were treated with pulse methylprednisolone and, when resistant, with antithymocyte globulin (Atgam; Upjohn, Kalamazoo, MI). By

1986, all patients were receiving cyclosporine A with prednisone

and azathioprine for maintenance immunosuppression.

Cyclospor-ine A levels were maintained between 100 and 200 ng/mL in the immediate postoperative period and between 50 and 100 ng/mL thereafter in all patients. The monoclonal antibody preparation Orthoclone OKT3 (Ortho Pharmaceuticals, Raritan, NJ) was added as therapy for acute rejection in 1986, and the use of equine antilymphocyte globulin (MALG; University of Minnesota) was initiated in 1988. All cadaveric-transplant recipients received at least three random donor transfusions. Between 1983 and 1988 all

living-related transplant recipients received three donor-specific

blood transfusions; after 1988 they received a single random donor

transfusion only.

Data were collected on a Medlog Databank (Information Analy-sis Corporation, Mountain View, CA). Statistical comparisons be-tween data groups were analyzed by either the 2 test or Student’s t test, where appropriate. Graft survival between groups was compared using the log rank test. A P value less than .05 was

considered statistically significant.

RESULTS

Patient demographic data are demonstrated in

Table 1. The majority of males were affected

second-ary to the increased risk of obstructive uropathy. No difference in sex, age, or diagnosis was detected when

data were analyzed by race. A significantly larger

proportion of nonwhites (59%) than whites were

(2)

Months Post Transplant

-0 White Liv.-Ret.

- N.wh. Liv..Rel.

-6- WhIte Cadaver

-‘. Nonwhite Cad.

20 30 40

Months Post Transplant TABLE 1. Patient Demographics TABLE 2. Transplant Data

1056 RENAL ALLOGRAFT SURVIVAL

White, No. (%) Nonwhite, No. (%)

(n=86) (n=41)

Male 57 (66) 26 (63)

Medical Assistance> 16 (19) 24 (59) Primary Disease

Congenital/obstruction 38 (44) 15 (37) Glomerulonephritis 24 (28) 17 (41)

Otherl 24 (28) 9 (22)

>P< .0001.

tIncludes patients with cystinosis (8), Alport’s syndrome (5), he-molytic-uremic syndrome (4), polycystic kidney disease (4), primary hyperoxaluria (3), Wilms’ tumor (3), sickle cell disease (2), medul-lary cystic disease (1), medullary sponge kidney (1), cortical necrosis (1), and interstitial nephritis (1).

An analysis of the individual transplants is detailed

in Table 2 and demonstrates a predominance of

liv-ing-related transplants toward white recipients just

missing statistical significance (.05 < P < .1). Analysis of the cadaveric donor’s race strongly favors whites

receiving kidneys from white donors (P < .0001) as

only 7 of 105 known donors were nonwhite. No

significant difference was noted between the propor-tion of first or subsequent allograft transplants when data were analyzed as a function of race over the 10-year study interval.

Evaluation of the differences in HLA-A and -B

mismatching by race (Table 3) reveals significantly

superior matching overall for whites compared with

nonwhites (P < .005). At our center, prior to the use of cyclosporine A, patients generally remained on the

cadaver waiting list until at least a 2 HLA-A,B

matched kidney was available; hence, no significant

difference in matching between races was present

prior to 1985. After 1985, poorer HLA matches were

considered acceptable for patients waiting more than

6 months for an allograft; consequently a significant

difference in mismatching was seen in both whites (P

< .005) and nonwhites (P < .005). The difference

between whites and nonwhites was accentuated since

1985 as only 4 (16%) of 25 nonwhites received a

cadaveric graft with a 0-2 HLA-A and -B mismatch

vs 19 (49%) of 39 whites (P < .05). A similar analysis

of HLA-DR mismatching showed no differences

be-tween whites and nonwhites between 1980 through

1984 and 1985 through 1989.

As shown in Fig 1, graft survival was significantly

greater for whites than nonwhites (P < .05). This

difference was lessened, however, when cadaveric

and living-related transplants were considered

sepa-rately (Fig 2). As expected, patients who received living-related transplants did significantly better than

those receiving grafts from cadaveric donors (P <

.005).

Figure 3 depicts the survival of cadaveric donor

allografts over two periods of time. Between 1985

and 1989 statistically significant differences were

noted between whites and nonwhites (P < .05). This

corresponds to the time period when cyclosporine A was introduced and matching criteria were less strict. The graph also illustrates excellent early graft function

in nonwhite recipients that declines precipitously by

2 years. This difference was not seen prior to 1985,

when both whites and nonwhites had similar 1-, 3-,

White (n = 111) Nonwhite (n = 49) Age, y (mean ± SD) 11.8 ± 4.5 12.9 ± 4.9 Living-related donors, No. (%) 31 (28) 7 (14) Cadavenc donor, No. (%) 80 (72) 42 (86) Donor of same race,> No. (%) 64 (94) 3 (8) First transplant, No. (%) 73 (66) 39 (80) Second transplant, No. (%) 27 (24) 7 (14) Third or fourth transplant, 11 (10) 3 (6)

No.(%) *P< .0001.

TABLE 3. Grafts Comp

Number of HLA-A,B Mismatche aring 1980-1984 to 1985-1989>

d Cadaveric-Donor

Mis- 1980-1989 1980-1984 1985-1989

match

No. W NWt W NW W NWI

0-2 54 16 35 12 19 4

3-4 26 26 6 5 20 21

>W, white; NW, nonwhite.

tW vs NW, P < .005. WvsNW, P< .05.

U) 0

0 0.

Fig 1. Graft survival in all transplants comparing whites and

nonwhites (P < .05).

U)

0.

Fig 2. Graft survival in both living-related and cadaveric-donor

transplants comparing whites and nonwhites.

at Viet Nam:AAP Sponsored on September 1, 2020 www.aappublications.org/news

(3)

-0- 1980.4 WhIte

-a. 1985-9 WhIte

-.- 1980-4 N.Wh.

-a. 1985.9N.Wh.

Months Post Transplant

U) 0

0 0.

Fig 3. Graft survival in cadaveric-donor transplants in 1980 to 1984 (pre-cyclosporine A) and 1985 to 1989 (post-cyclosporine A)

com-paring whites and nonwhites. NS, not significant.

and 5-year graft survival rates. No change was noted in allograft survival in whites since 1985, but a

sig-nificant deterioration did occur among nonwhites

during this interval (1980 through 1984 vs 1985

through 1989, P < .05).

The 1-, 3-, and 5-year cadaveric-transplant survival rates for selected groups are shown in Table 4.

Non-white cadaveric-donor allograft recipients showed

poor 5-year survival for repeated transplants (0%),

age greater than or equal to 1 2 years at the time of

transplant (1 7%), females (10%), poor HLA-A and

-B mismatching (9%), and low socioeconomic status

(12%) compared with whites. Though these

differ-ences do not reach statistical significance, of note is

the increasing disparity between whites and

non-whites with increasing time posttransplantion in all

groups

except those receiving medical assistance,

where it was seen in both groups, indicating that a

relative increase in graft loss occurred in nonwhites after 1 year.

DISCUSSION

Several individual and multicenter studies have

been undertaken to determine the impact of race on

graft survival.’7 To date, all have looked at this effect

on the adult recipient. In those studies in which

statistical differences were noted, graft survival in

white recipients was superior to that in blacks. These

differences have remained true following the

intro-duction of cyclosporine A although they may not

become evident until long-term survival data are

ex-amined.1 The North American Pediatric Renal

Trans-plant Cooperative Study, begun in January 1987, has

not yet noted any racial differences,8 perhaps because reliable survival data greater than 2 years

posttrans-plant are not available. Although our analysis of

nonwhite recipients included both African-Americans and those of other ethnic origins, the observed

differ-ences in graft survival were the same whether the

comparisons were broad or were limited exclusively to blacks vs whites.

One possible explanation for variable graft survival

is the increased immunogenicity seen in blacks,2

al-though this would be difficult to explain in light of

reports from several centers documenting equivalent

results among blacks and whites.4’6’7 In addition,

Rodey and colleagues9 found no difference in the

frequency of alloimmunization between blacks and

whites

among the 323 patients they studied. Increased

immunologic responsiveness may explain why

re-ports have indicated that younger recipients do worse than older ones.2

The effect of HLA matching has been reported as

a possible etiology for improved allograft

The significance of a possible difference in HLA

an-tigens among blacks and whites may be magnified at

a center where the allocation of scarce allografts is

heavily dependent on matching. After cyclosporine

A was introduced, we decreased our dependence on

well-matched kidneys selecting zero HLA antigen

matched ABO compatible kidneys to shorten waiting

time on dialysis. This may explain the poorer graft

survival noted in our nonwhite patients after 1985.

These results and a national emphasis on matching

have caused us to modify our policy. After studying

the Pennsylvania state donor pool in 1987, Nathan

et al’2 found 9% of actual donors were nonwhite.

This may be secondary to the fact that fewer

non-whites gave consent when asked and several

non-whites were never recognized as potential donors.

Perez et al’3 looked at differences in organ donation

and have speculated several reasons for less organ

donation in blacks and Latinos. Nevertheless, one

solution would be to increase the number of nonwhite donors, which may result in better matching for

non-whites. However, some investigators have noted

poorer graft function in any patient receiving kidneys from black donors14.’5

In our study a strong association was observed

between nonwhites and low socioeconomic status,

making

it difficult to isolate the factors responsible for poor graft survival in this population of limited size. Moreover, the use of a descriptor of socioeco-nomic status as imprecise as whether or not a patient

TABLE 4. Aflograft Su rvival for Selec ted Groups of Ca daveric-Transpla nt Recipients>

Category No. W/NW 1 Year 3 Years 5 Years P

W NW W NW W NW

All

Repeatedtransplant

12 y

Females

Since 1985 3-4 HLA mismatch

Medical assistance 80/42 37/10 47/31 30/17 39/25 26/26 15/29 0.65 0.64 0.60 0.60 0.68 0.70 0.70 0.53 0.72 0.59 0.73 0.62 0.80 0.59 0.53 0.44 0.53 0.13 0.53 0.46 0.54 0.40 0.61 0.24 0.59 0.42 0.55 0.39 0.44 0.20 0.48 0.00 0.42 0.17 0.44 0.10 0.44 0.09 0.22 0.12 .20 .09 .19 .08 .02 .07 .25

(4)

1058 RENAL ALLOGRAFT SURVIVAL

was receiving public medical assistance prior to

transplantation’6 is suboptimal; while other variables such as family income, family size, and highest level

of education achieved should be considered, these

criteria were impossible to develop from our

retro-spective analysis.

Noncompliance is a documented problem in

pedi-atric renal transplantation,’7 especially with regard to females, adolescents, and children with unstable fam-ilies, and is considered a major cause of late graft failure.’8 In a large retrospective study, Schweizer et

al’6 found a higher incidence of noncompliance in

lower socioeconomic status patients as well as young adults. Although our data did not reach statistical significance, the particularly poor long-term graft sur-vival in our nonwhite adolescents and females as well as children from a low socioeconomic status confirms these results and implies that graft loss may be

sec-ondary to noncompliance. This cause of graft failure

merits a great deal of attention as it is potentially preventable.

We suspect that noncompliance has become more

clinically apparent since the introduction of

cyclo-sporine A, because episodes of acute rejection may be

readily associated with inadequate drug levels. None-theless, because compliance is difficult to quantify independently of other criteria, no attempt was made to compare groups by labeling patients as “compliant” or “noncompliant.”

Finally, the so-called center effect has been shown

to magnify differences in outcome with regards to

race.” We cannot address this in a single-center study

but hope that multicenter studies like the North

American Pediatric Renal Transplant Cooperative

Study will be able to detect potential center dispari-ties.

SUMMARY

Our comparison of white and nonwhite children

over the past 1 0 years reveals superior graft survival overall in white recipients. This is, in part, due to an increased rate of living-related transplants in whites. Thus efforts should be made to increase living-related transplantation, when possible, in nonwhite patients.

Another concern remains when we evaluate the data

from the past 5 years which show a significant

dif-ference in white vs nonwhite cadaveric-graft survival.

This is despite the use of cyclosporine A and may be

related to poorer matching. Attempts at waiting for

better matched kidneys might improve graft survival, but this might be at the expense of fewer transplants for nonwhite children. Increased cadaveric-allograft

donation from nonwhites may ensure better matched

kidneys, but this needs to be evaluated as a source of potential benefit. Although not statistically

signifi-cant, the poorest graft function was seen in our

ado-lescent and female nonwhites as well as patients

receiving medical assistance. Attempts toward

im-proving compliance by simplifying medication

sched-ules and access, assessing patients’ concerns regarding their self-image, and individualized attention directed toward the high-risk patient might improve their fate.

Clearly multicenter studies with a greater number of

patients and long-term results tested with multivar-iant analysis may allow for better definition of high-risk patients to target for special consideration.

REFERENCES

1. Kasiske BL, Neylan JF, Riggio RR, et al. The effect of race on access and

outcome in transplantation. N Engl IMed. 1991;324:302-307

2. Takemoto 5, Terasaki P1. A comparison of kidney transplant survival in white and black recipients. Transplant Proc. 1989;21:3865-3868

3. Dawidson IJ, Coorpender L, Fisher D, et al. Impact of race on renal transplant outcome. Transplantation. 1990;49:63-67

4. Garvin PJ, Castaneda M, Codd JE, Mauller K. Recipient race as a risk factor in renal transplantation. Arch Surg. 1983;118:1441-1444

5. Dunn 1’Vasthala A, Golden D, et al. Impact of race on the outcome of renal transplantation under cyclosporine-prednisone. Transplant Proc.

1989;21 :3946-3948

6. Shapiro R, Tzakis AG, Hakala TR. Lopatin WB, Steiber AC, Starzl TE. Renal transplantation in black recipients at the University of Pittsburgh.

Transplant Proc. 1989;21:3921-3925

7. Sommer BG, Sing DE, Henry ML, Ferguson RM. The influence of recipient race on renal allograft survival: a single institution analysis.

Transplant Proc. 1989;21:3929-3930

8. Tejani A, Stablein D, Alexander 5, Fine R. Renal allograft outcome in North American children: a report of the North American Pediatric Renal Transplant Cooperative Study. JASN. 1990;1:772. Abstract 9. Rodey G, Parker M, Neylan J. Lowance D, O’Brien D, Whelchel J.

Antibodies to public class I epitopes in the American black population.

Transplant Proc. 1989;21:3878-3879

10. Opeltz G, Engelmann A. Effect of HLA matching in cyclosporine-treated black kidney transplant recipients. Transplant Proc. 1989;21:3881-3883 1 1. Sanfilippo F. Vaugn WK. Speis EK, Light JA, LeFor WM. Benefits of

HLA-A and HLA-B matching on graft and patient outcome after cadav-eric donor renal transplantation. N Engl I Med. 1984;311:358-364 12. Nathan HM, Jarrell BE, Broznik B, et al. Estimation and characterization

of the potential renal organ donor pool in Pennsylvania. Transplantation.

199 1;51:142-149

13. Perez LM, Schulman B, Davis F, Olson L, Tellis VA, Matas AJ. Organ donation in three major American cities with large Latino and black populations. Transplantation. 1988;46:553-557

14. Opeltz G, Pfarr E, Englemann A, Keppel E. Kidney graft survival in black cyclosporine-treated recipients. Transplant Proc. 1989;21:39 I 8-3920

15. Galton J. Racial effects in renal transplantation. In: Terasaki P1, ed.

Clinical Kidney Transplants I 985. Los Angeles, CA: UCLA Tissue Typing

Laboratory, 1985:153-166

16. Schweizer RT, Rovelli M, Palmeri D, Vossler E, Hull D, Bartus S. Noncompliance in organ transplant recipients. Transplantation.

1990;49:374-377

17. Beck DE, Fennell RS, Yost RL, Robinson JD, Geary D, Richards GA. Evaluation of an educational program on compliance with medication regimens in pediatric patients with renal transplants. I Pediatr.

1980;96:1094-1097

18. Didlake RH, Dreyfus K, Kerman RH, Van Buren CT, Kahan BD. Patient noncompliance: a major cause of late graft failure in cyclosporine-treated renal transplants. Transplant Proc. 1988;20(suppl 3):63-69

(5)

1992;89;1055

Pediatrics

H. Jorge Baluarte

Seth L. Schulman, JoAnn Palmer, Stephen Dunn, Bruce A. Kaiser, Martin S. Polinsky and

Study