The North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) is a voluntary collaborative effort comprising over 150 pediatric renal disease treatment centers in the United States, Canada, Mexico and Costa Rica. It is supported by major, unrestricted educational grants from Novartis, AMGEN, and Genentech. Participating NAPRTCS centers are listed in the most recent NAPRTCS Annual Report: Seikaly M, Ho PL, Emmett L, Tejani A: The 12th Annual Report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS): Renal Transplantation from 1987 through 1998. Pediatr Transplant 2001; 5: 215–231.
Reduction in Acute Rejections Decreases Chronic Rejection Graft Failure in Children: A Report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS)1
Article first published online: 2 APR 2002
American Journal of Transplantation
Volume 2, Issue 2, pages 142–147, February 2002
How to Cite
Tejani, A., Ho, P. L., Emmett, L. and Stablein, D. M. (2002), Reduction in Acute Rejections Decreases Chronic Rejection Graft Failure in Children: A Report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS). American Journal of Transplantation, 2: 142–147. doi: 10.1034/j.1600-6143.2002.020205.x
- Issue published online: 2 APR 2002
- Article first published online: 2 APR 2002
- Received 16 March 2001, revised and accepted for publication 18 September 2001
- Chronic rejection;
- cyclosporine therapy;
- chronic rejection;
- pediatric patterns
Chronic rejection accounted for 32% of all graft losses in 7123 pediatric transplants. In a previous study acute, multiple acute and late acute rejections were risk factors for the development of chronic rejection. We postulated that the recent decrease in acute rejections would translate into a lower risk for chronic rejection among patients with recent transplants. We reviewed our data on patients transplanted from 1995 to 2000, and using multivariate analysis and a proportional hazards model developed risk factors for patients whose grafts had failed due to chronic rejection. A late initial rejection increased the risk of chronic rejection graft failure 3.6-fold (p < 0.001), while a second rejection resulted in further increase of 4.2-fold (p < 0.001). Recipients who received less than 5 mg/kg of cyclosporine at 30 days post-transplant had a relative risk (RR) of 1.9 (p = 0.02). Patients transplanted from 1995 to 2000 had a significantly lower risk (RR = 0.54, p < 0.001) of graft failure from chronic rejection than those who received their transplants earlier (1987–94). Since we were able to demonstrate that there is a decreased risk of chronic rejection graft failure in our study cohort, we would conclude that the goal of future transplants should be to minimize acute rejections.
The most common cause of graft loss in pediatric renal transplantation is chronic rejection. In a recent analysis of patients transplanted from 1987 through January 1999 (n = 5958), chronic rejection graft loss was 31% of all graft failures (1) and of graft losses reported in 1999, 38% were due to chronic rejection (1). However, the mechanisms of chronic rejection are not well elucidated (2–4). Whereas an immunological basis for chronic rejection has been well reviewed (5), several authors have suggested that nonimmunological factors also play an important role in the development of chronic rejection (6–10). Based on the theory of a decreased nephron mass and consequential hyperfiltration, nonimmunological risk factors, such as large recipient size, African–American recipient race, older or female donors, and longer preservation time, have also been identified as factors which increase the risk of chronic rejection (9, 10).
Numerous single and multicenter studies in adult transplantation have shown the correlation between acute rejection and the subsequent development of chronic rejection, and prolonged graft survival in a rejection-free milieu in adults has also been observed (11–13). In a single-center study of pediatric renal transplants from 1984 to 1994 (n = 217), one or more episodes of acute rejection were strong correlates of chronic rejection. Similarly, a small collaborative pediatric renal study in Europe (n = 314) (14) showed acute rejection to be a risk factor for the development of chronic rejection.
A study of NAPRTCS transplants between 1987 and 1994 showed that acute rejection, multiple acute rejections and late acute rejections are risk factors for the development of chronic rejection (15). In recent years the frequency of acute rejections has declined. The rejection rate and time to first rejection were also significantly better among patients transplanted more recently (16). We postulated that recently transplanted patients had a lower risk of chronic rejection, not only because of declining acute rejection rate but also because of improvements in transplant patient management in general. Therefore, we reviewed our data from 1995 to 2000, and present the results here.
Materials and Methods
The NAPRTCS is a voluntary organization with over 150 participating centers, of which 73 have an active transplant program. The data presented here are collected from the active transplant centers. Data regarding recipient and donor characteristics are collected at transplantation. Data regarding induction and immunosuppression dosing are collected at day 30 post-transplantation, thereafter follow-up data, including renal function, statural growth and hospitalizations, are collected every 6 months. Separate data are collected for each acute rejection episode and include outcome and therapy. Additionally, separate data are collected for graft loss, which requires identification of the cause. Chronic rejection graft loss is defined as ‘progressive loss of renal function that leads to graft failure which occurs at least 90 days post-transplant.’ Graft failure occurring in the first 90 days cannot be classified as chronic rejection.
Data on graft loss due to an episode of chronic rejection were analyzed separately for cadaver donor (CD) and living donor (LD) transplant recipients. Factors analyzed consisted of donor age, recipient age, race, gender, type of graft, T-cell induction, HLA mismatch, prior transplantation, prior dialysis, prior transfusions, primary disease, number of acute rejection episodes, early acute rejection episodes (< 30 days post-transplantation), late acute rejection episodes (> 365 days post-transplant), and cyclosporine dose. To develop risk factors, patients whose grafts had failed due to chronic rejection were compared with patients whose grafts had failed due to other causes. Risk factors identified as correlating with chronic rejection were subjected to a multivariate analysis. A Cox proportional hazards model was constructed that equated the individual patient's hazard to an underlying hazard, multiplied by estimated exponentiated linear combination of risk factors. Number and timing of rejection episode and cyclosporine dose were included as time-varying covariates. Multivariate models were scaled so that with the exception of cyclosporine dose, risk increased with larger values of the covariates; the relative risk (RR) for single dichotomous risk factor is the exponentiated parameter.
The study cohort for this analysis comprised 1551 patients with LD and 1226 patients with CD transplants, entered between January 1995 and August 2000. Graft failure has been recorded for 134 (8.6%) LD and 189 (15.4%) CD transplants. Failure due to chronic rejection occurred in 16 (1.0%) LD and 40 (3.2%) CD transplants. Of all graft failures 12% of LD and 21% of CD were attributed to chronic rejection. Table 1 shows the overall percentage of failures attributed to chronic rejection relative to time post-transplant. Only one of the 25 grafts that failed during the first year was identified as due to chronic rejection, but it accounted for 15 of the 57 graft failures in the second year.
|Percent (#/Total failures)|
|Day 90–365||4% (1/25)|
|Year 2||26% (15/57)|
|Year 3||40% (19/48)|
|Year 4||54% (14/26)|
|Year 5||36% (5/14)|
We next categorized patients for various potential prognostic factors as to whether the graft failed due to chronic rejection, failed due to other causes or was still functioning. These data are shown separately for CD and LD grafts in Table 2. African–American patients and patients with a prior transplant were observed to have a significantly increased percentages of failure due to chronic rejection (p < 0.05).
|Living donor||Cadaver donor|
|Chronic rejection (n = 16)||Other failure (n = 118)||Functioning graft (n = 1095)||Chronic rejection (n = 40)||Other failure (n = 149)||Functioning graft (n = 802)|
|Age at Tx. (< 6 years)||18.8||17.8||23.3||15.0||12.1||12.0|
|> 5 Transfusions||6.3||16.1||10.9||25.0||27.5||17.5*|
|HLA-DR (2 mismatches)||18.8||23.7||29.9||32.5||55.7||50.1|
An analysis of cyclosporine dosing at day 30 revealed that among 1299 LD recipients with a functioning graft at day 90, 1.4% (4/278) of those who received less than 5 mg/kg of cyclosporine had subsequent chronic rejection graft failure, compared with 1.2% (12/1021) who received more than 5 mg/kg (p = 0.78). For 1028 CD graft recipients with a functioning graft for 90 days or longer, 3.3% (25/751) of patients who received less than 5 mg/kg had subsequent chronic rejection failure, compared with 5.4% (15/277) of those who received more than 5 mg/kg of cyclosporine (p = 0.044).
The following primary diseases were analyzed for comparative frequencies of graft loss due to chronic rejection: structural lesions (dysplasia, aplasia) glomerulonephritis, focal sclerosis, congenital nephrotic syndrome, hemolytic–uremic syndrome, renal infarct, cystinosis and familial nephritis. No significant differences were noted regarding the frequency of chronic rejection among various disease processes in either the LD or CD group.
In Table 3, we compare acute rejection episodes in patients with chronic rejection graft failure to all other graft recipients, separately for LD and CD grafts. In both donor groups, about 70% of failures from chronic rejection are preceded by an episode of acute rejection, while overall acute rejection occurred in 32% of the remaining transplants. Additionally, multiple episodes of acute rejection and initial late acute rejection occurred more frequently in the chronic rejection group.
|Acute rejection||Chronic rejection|
|Graft failure||Other recipients|
|Initial late rejection|
In both univariate and multivariate analyses of this study, we used a time-varying covariate approach to assess the association between acute rejection and chronic rejection failure, because the rejection history develops over time and the accompanying risk group also changes. For example, among patients with a 1-year follow-up, 24% of chronic rejection failures had late-onset initial rejections, compared to 32% of chronic rejection failure patients who had at least 2 years of follow-up. In univariate analysis, treating the rejection parameter as a time-varying covariate, a single acute rejection, a second acute rejection, and late acute rejection were significant (p < 0.001) predictors of chronic rejection failure. A multivariate analysis to simultaneously evaluate predictors of graft failure from chronic rejection is shown in Table 4.
|Relative risk increase||p-value|
|≥ 2 acute rejections||4.0||< 0.001|
|Late initial acute rejection||3.6||< 0.001|
|Prior transplant||3.1||< 0.001|
A late initial rejection increased the risk of chronic rejection graft failure 3.6-fold, while a second rejection resulted in a further increase of 4.0-fold. The risk of chronic rejection graft failure in patients with a late initial rejection and a second rejection was > 16 times the risk of patients who had no rejection. Table 4 shows that adjusting for acute rejection episodes, CD recipients, African–American patients, and patients with a repeat transplant are more likely to experience graft failure from chronic rejection. In addition, recipients who receive less than 5 mg/kg of cyclosporine at day 30 post-transplant are at a slightly higher risk (RR = 1.9, p = 0.02), as are patients hospitalized for infection in the previous 6 months at 2.7-fold (p = 0.001).
Since acute rejection is such an important factor in chronic rejection graft failure, subjects with acute rejection form a group at risk for the event. Figure 1 presents the distribution of time to graft failure from all causes and of time to failure from chronic rejection for both LD and CD source transplants. Time was measured from the first episode of acute rejection and all subjects with less than 90 days graft survival were excluded, as chronic rejection is defined as rejection occurring only after 90 days. At 3 years, the overall graft failure rate was 13.8% for LD and 24.1% for CD grafts, while the cumulative percentage of graft failure attributed to chronic rejection was 4.3% and 10.4%, respectively.
Lastly, to test our hypothesis that the recently transplanted patients had a lower risk for chronic rejection, we examined all transplants since 1987 (3655 LD and 3691 CD) and contrasted the recent transplants to earlier ones. These data are shown in Tables 5 and 6. Patients transplanted from 1995 to 2000 had a significantly lower risk (RR = 0.54, p < 0.001) of graft failure from chronic rejection than those who received their transplants earlier (1987–94), after adjusting for other factors (Table 5). When the effects of acute rejection episode factors were also adjusted, the relative risk of recently transplanted patients was still significantly lower (RR = 0.66, p = 0.005) than that of those transplanted earlier (Table 6). Furthermore, a single acute rejection was significantly predictive of chronic rejection graft failure (RR = 1.5, p = 0.006).
|Relative risk increase||p-value|
|Recent transplant (01/95–08/00)||0.54||< 0.001|
|Prior transplant||1.7||< 0.001|
|African–American race||1.9||< 0.001|
|Cadaver donor||1.7||< 0.001|
|Relative risk increase||p-value|
|Recent transplant (01/95–08/00) vs. earlier (01/87–12/94)||0.66||0.005|
|≥ 2 acute rejections||4.1||< 0.001|
|Late initial acute rejection||2.6||< 0.001|
|Prior transplant||2.4||< 0.001|
|African–American race||2.3||< 0.001|
|Cadaver donor||1.5||< 0.001|
For several years now our database has shown a steady improvement in short-term graft survival (16–19). In our latest published data, we note that CD graft survival has improved with each succeeding year so that for the 1997–98 cohort the 1-year CD graft survival is 92.6%, and 94% for LD transplants in the same cohort. However, the steady improvement in the short term is not reflected in the long-term graft survival since the 7-year graft survival rates are 73% (LD) and 59% (CD) (1).
In analyzing our data we continue to note the rising graft attrition due to chronic rejection, which has increased to 38% of all graft losses reported in 1999 (NAPRTCS 2000 Annual Report). In multicenter registry data like ours, biopsy confirmation of chronic rejection is impossible and our findings obviously under-represent the magnitude of the problem since we record only graft loss due to chronic rejection. As in adult transplantation, single-center pediatric studies (20) as well as our own previous data have shown that acute rejection is a strong correlate and risk factor for chronic rejection graft loss. With improving immunosuppression and changes in practice patterns, such as elimination of infantile cadaver kidneys and a reduction in the use of cadaver kidneys from 2 to 5–year-olds, there has been a substantial decrease in acute rejection rates in recent years (16), and the 12-month probability for a first rejection has decreased from 55% for LD and 71% for CD in 1987–88, to 38% and 45%, respectively, in 1995–96 (1). We therefore surveyed transplants performed from 1995 onwards to determine risk factors for chronic rejection graft loss.
We have not addressed other presumptive nonimmunological factors, such as nephron mass, that may accelerate chronic rejection graft failure since we do not collect data on the size and weight of the transplanted kidney. Other factors, such as hyperlipidemia and hypercholesterolemia, which have been implicated in adult transplant patients (21) have not been demonstrated as risk factors for children (22). We have been unable to look at the role of the other major calcineurin inhibitor, tacrolimus, since 89–91% of our patients are receiving cyclosporine at 5 years post-transplant (19) and only 11% of the patients transplanted from 1996 to 1998 were initiated on tacrolimus (1). Our database shows that, during the study period, only 118 LD and 86 CD patients were receiving tacrolimus at day 30 post-transplant. A Kaplan–Meier analysis of chronic rejection contrasting tacrolimus and cyclosporine showed no significant difference (p = 0.43 LD and 0.10 CD), probably due to smaller numbers in the tacrolimus arm We have previously observed that the median cyclosporine dose at 1 month post-transplant is 9 mg/kg (23) and we have also observed that the mean cyclosporine dose at 6 months was higher for those who kept their grafts for the subsequent 6-month period compared to those who lost their grafts in the same period (24). For this study, however, we focused on the day-30 dose of cyclosporine to exclude the possibility that the dose lowering may be dictated by chronic cyclosporine nephrotoxicity or may be due to noncompliance, both of which are rare in the first month. Our observation that low cyclosporine dose is a risk factor for the development of chronic rejection graft failure would suggest that there is a window of immunosuppression with cyclosporine and that both high and low doses should be guarded against.
The mechanism by which acute rejection induces chronic rejection has not been elucidated. The histological picture of acute rejection has been well studied (25) and intrarenal expression of cytotoxic attack molecules during acute rejection has been documented in both adult (26) and pediatric renal biopsies (27). It is interesting to note that all patients who develop an acute rejection do not progress to chronic rejection. It could be postulated that in patients who progress to chronic rejection following an episode of acute rejection the cytokine profile is robust, and that, whereas antirejection therapy may provide clinical quiescence, a smoldering of cytotoxic attack molecules continues leading to the development of chronic rejection. However, when renal biopsies of patients with chronic rejection are analyzed, intrarenal transforming growth factor (TGF)-β1 mRNA expression has been shown to be a molecular correlate of chronic rejection, whereas it is neither a negative nor a positive correlate of acute rejection (28). These findings would suggest that other nonimmunological mechanisms, such as ischemic tissue injury, are also involved since TGF-β1 is a prominent member of the cytokine cascade involved in tissue repair. That ischemic injury may play a role is further attested to by our observation that CD transplants are at a higher risk for the development of chronic rejection.
This study was prompted by our observation of a decrease in the incidence of acute rejection, and is an attempt to determine if that translates into a decreased risk for the development of chronic rejection. We have been able to confirm our hypothesis by showing that there is a decrease in the risk for chronic rejection graft failure in our present study cohort compared with the total transplant population (Table 6), and by showing that the long-term (5-year) graft survival for the present study cohort is superior to that of the total transplant population. A recent review of the impact of acute rejection on the development of chronic rejection in pediatric renal transplantation (21) has pointed out that the goal must be to minimize acute rejection. A rejection-free milieu can truly provide improved long-term graft survival. In a recent study of ours we noted that of 5075 children with 1-year graft survival, 52% were rejection free at 1 year, and for 87% of those patients 5 additional years of graft survival are observed (29).
- 2T cell mediated immune responses in chronic rejection. Role of indirect allorecognition and costimulation. Graft 1998; 1 (Suppl. II): 11–17., , ,
- 4Allorecognition pathways. In: Tejani AH, Harmon WE, Fine RN, eds. Pediatric Solid Organ Transplantation. Copenhagen: Munksgaard; 2000. pp 17–25.,
- 6Renal mass: an important determinant of late allograft outcome. Transplant Rev., 1998; 12: 74–84., , ,
- 10Antigen-independent determination of graft survival in living-related kidney transplantation. Kidney Int 1997; 63 (Suppl.): S84–S86., , ,
- 12Lessons learned and future hopes. Three thousand renal transplants. In: Terasaki PI, ed. Clinical Transplants 1990. Los Angeles, CA: UCLA Tissue Typing Laboratory; 1991. p. 217., , et al.
- 17Renal transplantation in children from 1987 to 1996. The Annual Report of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS). Pediatr Transplant 1997; 1: 146–162., , , ,
- 23Cyclosporine dosing and its relationship to outcome in pediatric renal transplantation. Kidney Int 1993; 44 (2): S50–S55.,
- 27Intragraft expression of T-cell genes in human renal allograft rejection. Kidney Int 1996; 49 (2): S7–S12., ,