Decreased Risk of Renal Allograft Thrombosis Associated With Interleukin-2 Receptor Antagonists: A Report of the NAPRTCS

Authors


Abstract

Graft thrombosis is the most common cause of first year graft failure in pediatric renal transplantation. The North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) database was analyzed for cases of graft failure due to thrombosis among patients transplanted from 1998 to 2004. The impact of interleukin-2 (IL-2) receptor antagonists as induction therapy was determined. There were a total of 51 graft failures due to thrombosis among the 2750 reported renal transplants (1.85%) (95% CI (1.39%, 2.41%)). This represents the most common cause of graft loss during the first year post-transplant accounting for 35% of first year losses and 18% of all graft losses. The incidence of thrombosis among patients who received IL-2 receptor antibodies was 1.07% (12/1126) compared to 2.40% (39/1624) among patients who did not (OR 0.44, 95% CI 0.23, 0.84, p = 0.014). Use of IL-2 receptor blockade was the only significant prognostic factor in a multivariate model with previously identified risk factors. Analysis of NAPRTCS data found that the use of IL-2 receptor antibodies as induction therapy is associated with a significantly decreased risk of graft failure due to thrombosis. This provocative finding requires further investigation to determine whether thrombotic failure can be decreased by this therapeutic strategy.

Introduction

Renal allograft thrombosis is a devastating complication of renal transplantation resulting in almost universal graft loss. In pediatric renal transplantation, it represents the most common cause of first year graft failure. Previous analyses of the North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) database have identified the following risk factors for graft thrombosis: cadaver donor source, peritoneal dialysis, >5 pre-transplant blood transfusions, cold ischemia time >24 h and prior transplant (1–3). From 1987 to 1995, the rate of thrombosis was constant at 3.1%. However, in the cohort of patients transplanted between 1996 and 2001, there was a reduction in the thrombosis rate to 2.1%. During the same time period, there were changes in the types of induction therapy used in pediatric renal transplantation with a decrease in the use of ATG/ALG and OKT3 and a dramatic increase in the use of interleukin-2 (IL-2) receptor antagonists. We performed a retrospective cohort study of the NAPRTCS database to assess the relationship between graft thrombosis and use of IL-2 receptor antagonists as induction therapy.

Methods

Study population

Since its inception in 1987, the scientific registry of NAPRTCS has enrolled over 8000 patients who have received one or more renal transplants. Data collection methods have been previously reported (4). Briefly, the NAPRTCS has over 150 participating medical centers in the United States, Canada, Mexico and Costa Rica. The data on pediatric transplantation were derived from centers that collect information at the time of transplant, 1 month post-transplant, 6 months post-transplant and every 6 months thereafter. The initial transplant information consists of recipient age, gender, race, primary disease, type and duration of dialysis, type of transplant, degree of HLA mismatch, ischemia time, peak panel reactive antibody, induction therapy, height and weight. At 30 days, data collected include graft outcome, initial immunosuppressive therapy, complications and re-hospitalizations. Thereafter, information is collected every 6 months regarding height, weight, serum creatinine, type and dose of immunosuppressive therapy and the use of other concomitant medications. In addition, extensive information is collected for each rejection episode. Cases were excluded from the antibody subgroups if the drug was first used on day 2 or beyond, and assigned to the IL receptor antagonist group if they received this induction treatment. Thrombosis refers to thrombosis as the cause of graft failure as we do not capture information on thrombotic events that do not lead to graft failure.

Statistical analysis

Descriptive analyses were used to report cohort characteristics. Univariate analysis of risk factors was done using the 2-tailed Fisher's exact test. Binomial confidence intervals used the Blyth-Still-Casella procedure. Logistic regression was used to determine the odds of developing graft thrombosis.

Results

Study population

Data from 2750 pediatric renal transplants performed between January 1, 1998 and October 2004 were analyzed. The patient characteristics are shown in Table 1.

Table 1. Patient characteristics
 Thrombosis rate
N/TotalRate (%)P value
Gender 0.32
 Male26/15951.63 
 Female25/11552.16 
Race 1.00
 Other42/22611.86 
 Black9/4891.84 
Age 0.50
 1–23/1462.05 
 3–510/3762.66 
 6–1218/9002.00 
 ≥1320/13281.51 
Prior transplant 1.00
 No42/22641.86 
 Yes9/4861.85 
Donor source 0.47
 Living27/16251.66 
 Deceased23/10962.10 
 Unknown1/293.45 
Main dialysis 0.61
 Unknown1/254.00 
 None11/7161.54 
 Hemodialysis14/8621.62 
 Peritoneal dialysis23/10082.28 
 Both2/1391.44 
Prior transfusion 0.49
 No44/24611.79 
 Yes7/2892.42 
Antibody induction 0.05
 ATG/ALG5/2751.82 
 Basiliximab8/5941.35 
 Daclizumab4/5320.75 
 OKT34/914.40 
 Other/None30/12582.38 
All51/27501.85 

Incidence of graft thrombosis

There were a total of 51 graft failures due to thrombosis among 2750 transplants (1.85%) (95% CI (1.39%, 2.41%)). This era appears to have a significantly lower thrombosis rate that the 3.19% (p < 0.001) rate seen in 1987–1997. A quarter of the individuals with thrombotic graft failure occurred by day 1, half by day 3 and three-quarters by day 8. All but six were in the first month. Table 2 shows the causes of graft loss within the cohort. Thrombosis was the most common cause of graft failure in the first post-transplant year (35% of failures) and 18% of all grafts lost during the study period (1998–2004).

Table 2. Causes of graft failure
CausesYear 1 Failures N=142All Failures N=277
Thrombosis50(35%)51(18%)
Primary non-function13(9%)13(5%)
Chronic rejection7(5%)62(22%)
Acute, hyperacute, or accelerated acute rejection14(10%)30(11%)
Death with function11(8%)25(9%)
Recurrence16(17%)30(11%)
Other31(22%)66(24%)

Risk factors

The incidence of thrombosis was 1.07% (12/1126) (95% CI (0.60%, 1.80%)) among patients who received IL-2 receptor antagonist therapy compared to 2.40% (39(1624) (95% CI (1.71%, 3.27%)) among those who did not (p = 0.014 Fisher's). By univariate analysis, there was no significant difference in thrombosis rates when considering race, gender, donor source, history of prior transplant, mode of dialysis and transfusion history (Table 1). NAPRTCS does not collect information of the patient's sensitization (PRA) and thus, this was not able to be included in the analysis. IL-2 receptor use as reported in 23.5% of patients in 1998 and ranged from 42.1% to 49.6% in the subsequent years. The incidence of thrombosis was 1.82% (5/275) among patients who received polyclonal antibody induction. Comparing the thrombosis incidence among patients using IL-2 receptor antagonist 1.07% (12/1126) versus others, there was a significant difference (p = 0.018), while comparing IL-2 use versus polyclonal antibody use, the difference was not significant (p = 0.35). Regarding maintenance of immunosuppression, 26% of all recipients initiated cyclosporine and 26% initiated tacrolimus on days 0 or 1 post-transplant. The thrombosis rate appeared similar for the two types of calcineurin inhibitors. Table 3 shows the results of a logistic regression analysis for the univariate and a multivariate model including donor source, prior transplant, transfusion history and peritoneal dialysis. Patients who had received IL-2 receptor antagonists as induction therapy were estimated to have less than half the risk of thrombosis compared to that of non-recipients (OR = 0.44). Other previously identified risk factors (donor source, prior transplant, transfusion history and peritoneal dialysis) were not associated with an increased risk of thrombosis in this model. Donor age was not included in the original model due to the potential confounding by donor type. An additional model with donor age was performed and the reanalysis did not alter the results (p = 0.012). The incidence of thrombosis in patients receiving Basiliximab was 1.35% (8/594) compared to 0.75% (4/532) in the Daclizumab treated group (p = 0.39).

Table 3. Risk factors for graft thrombosis
 Odds ratio (95% CI)P value
Donor Source DD versus LD1.22 (0.69, 2.17)0.50
Pre-transplant peritoneal dialysis1.37 (0.78, 2.42)0.27
> 5 Random transfusions1.39 (0.60, 3.19)0.44
Prior transplant0.88 (0.41, 1.87)0.73
IL-2 receptor antibody use0.44 (0.23, 0.84)0.013
Adjusted for donor source, peritoneal dialysis, >5 pre- transplant blood transfusions, and prior transplant0.43 (0.23, 0.84)0.013

Discussion

Renal allograft thrombosis is an important cause of graft loss in pediatric renal transplantation. Analysis of the NAPRTCS database found that the use of IL-2 receptor antagonists was associated with a significantly decreased risk of thrombosis. The incidence of thrombosis was 1.07% among patients who received IL-2 receptor antagonist therapy compared to 2.40% among those who did not. There was no difference in risk with the two different formulations: basiliximab and daclizumab. Previously identified risk factors for thrombosis including donor source, peritoneal dialysis, >5 pre-transplant blood transfusions, cold ischemia time >24 h and prior transplant were not found to be associated with an increased risk of thrombosis (1–3).

Thrombosis is the most common cause of first year graft failure but its incidence is decreasing. According to the NAPRTCS database, the thrombosis rate was 3.19% in patients transplanted between 1987 and 1997 compared to 1.85% among patients transplanted since 1998 (5). Change in transplant practice could account for the decreased incidence of thrombosis. Specifically, changes in practice related to the modifiable risk factors such as cold ischemia time, transfusions and young donor age may contribute to the decreased rates of thrombosis. NAPRTCS data demonstrate that the number of transplant recipients with a history of >5 pre-transplant blood transfusions has decreased from 44% (1987) to 8% since 2003 (5). Among cadaver transplants, the cold ischemia times have decreased from a median of 23.5 h in 1987 to 17.3 h in 2003 (5). Similarly, the number of donors less than 6 years has dramatically decreased from 25% in the first 3 years of the registry to 5% in the last 3 years (5). However, looking at the study period from 1998 to 2004, we find that there has been little change in these factors (5).

Immunosuppression use has changed significantly over the past decade. During the study period, induction therapy has shifted away from the polyclonal and monoclonal antibody preparations toward the IL-2 receptor antagonists (5). In 1998, 14.7% of patients received OKT3, 27.9% ATG(ALG, 5.0% Daclizumab and 0.7% Basiliximab. In contrast, the 2004 NAPRTCS Annual Report found that among patients transplanted in 2003, no patients received OKT3, 12.3% received ATG/ALG, 18.2% received Daclizumab and 29.9% received Basiliximab (5).

The relationship between medications and graft thrombosis has been studied and remains a controversial area. Several studies have implicated immunosuppression medication in the development of thrombosis. There are numerous possible mechanisms whereby cyclosporine can lead to thrombosis including hemostatic alterations (6), enhancement of platelet aggregation (7,8), induction of glomerular thrombosis (9) and reduction in prostacyclin synthesis (10). However, large clinical studies report conflicting results with some identifying cyclosporine as a risk factor (11–14) and others not (15,16). Similarly, OKT3 has been associated with graft thrombosis in adults and pediatric transplant recipients (17,18). There is evidence of possible endothelial activation in response to OKT3, which may contribute to thrombosis (19). In addition, some have implicated HLA antibodies and subsequent activation of the complement system in the pathogenesis of thrombosis (20). Other authors have found OKT3 use not to be an independent risk factor for graft thrombosis (2).

The current study found that the use of IL-2 receptor antagonists was associated with a significantly decreased risk of graft thrombosis. OKT3 use cannot account for the risk difference due to the small number of subjects who received it as induction therapy. To our knowledge, this benefit of IL-2 receptor antagonists has not previously been described. Possible mechanistic explanations for this observation are based on the studies of IL-2 immunotherapy. IL-2 administration has been associated with microvascular thrombosis (21) and it has been shown to increase platelet activation (22). In addition, IL-2 therapy is known to cause the activation of vascular endothelial cells through the induction of cytokines such as TNF and IL-1 (23). Blockage of the IL-2 receptor may prevent activation of this cascade that otherwise may lead to thrombosis.

We acknowledge the limitations of this study including selection biases and reporting accuracy associated with any registry report. Specifically, NAPRTCS does not collect information on patient sensitization data (PRA). The lack of this information may have impacted our results. We observe that graft thrombosis is an early event, typically occurring within the first 48 h of post-transplant. As such, use (or not) of induction antibody could be affected by the development of thromboses related graft dysfunction in the peri-transplant period, an event that could impact the observations from this report. In addition, there could be a ‘center effect’ whereby centers that adopt novel immunsuppression may be more advanced in their peri-operative management, which would also impact thrombosis rates. There were insufficient numbers of polyclonal antibody recipients to obtain precise estimates of differences between the polyclonal and anti-IL2 receptor groups. Finally, we cannot exclude that the relationship observed is merely associative and not causal.

In conclusion, this analysis found that IL-2 receptor antagonists were associated with a significantly decreased risk of graft thrombosis in pediatric renal transplantation. In addition, previously identified risk factors (donor source, prior transplant, transfusion history and peritoneal dialysis) were not associated with an increased risk of thrombosis in this cohort. Further studies to confirm these findings are warranted.

Acknowledgments

The North American Pediatric Renal Transplant Cooperative Study (NAPRTCS) is supported by unrestricted educational grants from Novartis, AMGEN and Genentech. The NAPRTCS is a voluntary collaborative effort comprising 150 pediatric renal disease treatment centers in the United States, Canada, Mexico and Costa Rica. Participating centers are listed in the NAPRTCS annual report.

Ancillary