DELETERIOUS EFFECTS OF DELAYED GRAFT FUNCTION IN CADAVERIC RENAL TRANSPLANT RECIPIENTS INDEPENDENT OF ACUTE REJECTION^1,2,3

Abstract

Background. 

In cadaveric renal transplantation, delayed graft function (DGF) correlates with poor long-term graft survival; however, whether its effects are independent of acute rejection is controversial. We wished to study the effect of DGF on graft survival, controlling for acute rejection, discharge creatinine, and human leukocyte antigen match.

Methods. 

We analyzed 27,096 first cadaveric donor renal transplant report to the UNOS Scientific Renal Transplant Registry between January 1994 and November 1997. DGF was defined as dialysis need in the first week. Acute rejection was recorded for initial hospitalization and within 6 months. Kaplan Meier survival curves were analyzed with the log rank test.

Results. 

DGF increased the incidence of acute rejection before discharge (8% without DGF; 25% with DGF, P<0.01) and any acute rejections by 6 months (25% without DGF, 42% with DGF, P<0.01). Without early rejection, DGF reduced 1-year graft survival from 91 to 75% (P<0.0001) and graft half-time from 12.9 to 8.0 years. In kidneys with acute rejection within 6 months, DGF decreased 3-year graft survival from 77 to 60% and graft half-life from 9.4 to 6.2 years (P<0.001). With a discharge creatinine of less than 2.5 mg/dl, the difference in graft half-life between no DGF and no rejection (13.4 years) and DGF with rejection (9.8 years) was significant (P<0.001). Increased donor age and cold ischemia time additionally decreased graft survival, whereas a good human leukocyte antigen match could not overcome the deleterious effects of DGF or acute rejection.

Conclusions. 

DGF is an important independent predictor of poor graft survival. Never immunosuppressive strategies must minimize nonimmune and immune renal injury if long-term graft survival is to improve.

*Abbreviations: CIT, cold ischemia time; DGF, delayed graft function; HLA, human leukocyte antigen.

Delayed graft function (DGF*) has long been identified as one of the principal correlates of poor graft survival in cadaveric renal transplantation ^(1,2); however, its strength as an independent variable has been questioned. Previous single-center studies had conflicting results over the importance of DGF in the absence of acute rejection, with some groups finding no effect ⁽³⁾, and other groups finding a DGF effect independent of early acute rejection ^(4,5). Indeed, DGF may very well increase the likelihood of acute rejection, either by the release of cytokines and up-regulation of major histocompatibility complex class II antigens ⁽⁶⁾ or by masking the diagnosis of rejection. It is difficult to understand how DGF alone could lead directly to chronic allograft nephropathy, particularly if there is complete early recovery of function. One possible mechanism might be that ischemia-reperfusion injury leads to an ongoing injury response involving inflammation, fibrosis and, ultimately, hyperfiltration injury ^(7,8).

We report a preliminary study ⁽⁹⁾ of the short- and long-term effects of DGF and rejection based on data from the UNOS Scientific Renal Transplant Registry, paying particular attention to the interaction between DGF, early rejection, and serum creatinine (as a surrogate of nephron mass) with graft survival and half-life. Four this study, we analyzed updated Registry data and focused on the most recent 3-year period to insure the results were applicable for current organ procurement and immunosuppressive protocols.

MATERIALS AND METHODS

We analyzed 27,096 first cadaveric donor renal transplants reported to the UNOS Scientific Renal Transplant Registry between January 1994 and December 1997. Recipients of multiorgan transplants were excluded. DGF was defined as the need for dialysis during the first week after transplant. Early rejection episodes were those reported during hospitalization for transplantation. The dialysis requirement and rejection before discharge were reported for 99% of the cases, and those cases missing either of these data elements were not analyzed. Early rejection episodes were not biopsy-confirmed in the majority of cases but were episodes of treated rejection confirmed by the report of the specific drugs administered for rejection. Late acute rejection was reported as occurring after the initial hospitalization but within 6 months of transplant. The serum creatinine level was as reported at the initial hospital discharge. Graft survival rates were calculated as Kaplan-Meier product limit estimates, including patient deaths as graft failures; graft half-lives were computed as the rate of graft loss after the first posttransplant year, assuming a constant rate. Statistical significance was estimated by log rank comparisons of survival curves, by z-test comparisons of graft half-lives, or by chi-square analysis of groupings.

RESULTS

The annual incidence of DGF and early and late acute rejection from 1988 to 1997 is shown in Figure 1. The incidence of DGF has made a slight but steady decline, from 27 to 22%, in 1997, and acute rejection has significantly decreased, both for incidence at discharge (29 to 8%) and at 6 months (55 to 22%). Focusing only on the first cadaveric transplants between 1994 and 1997 (Fig. 2), there was a significant relationship between DGF and incidence of acute rejection before discharge (8% without DGF, 25% with DGF, P<0.01) and any acute rejections by 6 months (25% without DGF, 42% with DGF, P<0.01). There was no increase, however, in the incidence of acute rejection between discharge and 6 months and DGF (15 vs. 13%), suggesting that the difference seen in acute rejection by 6 months was primarily because of early rejections and the drop in initial hospital length of stay shown in the bar graph in Figure 1 did not simply push early rejections associated with DGF into the postdischarge period.

Figure 1: Changes in incidence of delayed graft function (DGF, ▴), acute rejection before discharge (disch, ▵), acute rejection within 6 months (6 mo, □) and length of initial hospital stay (bars) in cadaveric renal transplant recipients from 1988 to 1997.

Figure 2: Relationship between delayed graft function (DGF) and incidence of acute rejection, whether diagnosed before hospital discharge (Disch), between discharge and 6 months (Disch-6 mo) or any time within 6 months (6 mo).

Both DGF and early acute rejection had independent negative impacts of graft survival at 1 and 3 years and on graft half-life (Fig. 3). Even in the absence of early rejection, DGF reduced 1-year graft survival from 91 to 75% (P<0.0001) and graft half-life from 12.9 to 8.0 years. When acute rejection occurred before discharge, DGF reduced 1-year graft survival from 81 to 66% (P<0.0001) and graft half-life from 9.5 to 7.7 years. When any rejections within the first 6 months were considered, DGF had less of an effect on early graft survival, but it still had a significant effect on graft half-life (Fig. 4). Comparing kidneys without any acute rejection, DGF had a minimal effect on 1-year graft survival, and it decreased 3-year graft survival from 89 to 84% and graft half-life from 14.2 to 9.7 years (P<0.001). Among kidneys with acute rejection within 6 months, initial DGF decreased 3-year graft survival from 77 to 60% and decreased graft half-life from 9.4 to 6.2 years (P<0.001).

Figure 3: Effect of DGF and acute rejection (Rej) before discharge on graft survival and half-life. n=total number of cases; t ½, half-life of the graft.

Figure 4: Effect of DGF and acute rejection (Rej) within 6 months of transplantation on graft survival and half-life. n=total number of cases; t ½, half-life of the graft.

Some studies have questioned whether DGF has a deleterious effect if there is full recovery of function. In Figure 5, the effect of DGF and early acute rejection were compared in patients with a discharge serum creatinine of less than 2.5 mg/dl. In the absence of early rejection, DGF reduced 1-year graft survival from 93 to 88% and 3-year graft survival from 83 to 75%. Nevertheless, even with a discharge creatinine of less than 2.5 mg/dl, the difference in graft half-life between no DGF and no rejection (13.4 years) and DGF with rejection (9.8 years) was significant (P<0.001).

Figure 5: Effect of DGF and acute rejection (Rej) in patients with a discharge serum creatinine of <2.5 mg/dl. n=total number of cases; t ½, half-life of the graft.

Advanced donor age is associated with reduced total nephron mass, so older kidneys would seem particularly vulnerable to added insult. When the effects of DGF and early acute rejection were stratified according to young (19-30 years) and older (46-60 years) donor age, DGF and rejection reduced graft survival from both age groups; however, the combination with kidneys from older donors was particularly harmful (Fig. 6). By comparison, no rejection or DGF with a young donor gave 1 and 3-year graft survivals of 93% and 86%, which fell to 61% and 51% with early acute rejection and DGF in an older donor.

Figure 6: Effect of DGF and acute rejection according to donor age on graft survival and half-life.

A positive relationship between increasing cold ischemia time (CIT) and incidence of DGF has long been known. As seen in Figure 7, there was a stepwise increase in DGF as CIT increased above 24 hr, coupled with an increase in risk of acute rejection before discharge. Nevertheless, CIT had a smaller influence on graft survival if considered independently of DGF or early acute rejection (Fig. 8).

Figure 7: Effect of cold ischemia time (CIT) on incidence of DGF and acute rejection before discharge.

Figure 8: Effect of cold ischemia time (CIT) either with no rejection (Rej) or no DGF on graft survival.

If the deleterious effect of DGF was only through the increased cellular expression of human leukocyte antigens (HLAs), then well matched kidneys would be expected not to show an effect. When we analyzed graft survival with or without DGF and early acute rejection, we saw an identical pattern of differences in 0-1 mismatched transplants and 5-6 mismatched transplants (Fig. 9). What is striking, however, is the loss of survival advantage in 0-1 mismatched kidneys in the presence of DGF with or without rejection. Although the well matched kidneys have a significantly longer half-life than the poorly matched kidneys in the absence of DGF or rejection (17.7 vs. 10.4 years), the advantage was lost with DGF alone (6.2 vs. 10.9 years) or DGF with early acute rejection (5.6 vs. 7.2 years).

Figure 9: Effect of DGF and acute rejection according to HLA mismatch (HLA MM).

DISCUSSION

It is well established that DGF is deleterious to cadaveric renal allograft survival when analyzed as a solitary variable ⁽²⁾. Risk factors for DGF include immune and ischemic components ⁽¹⁰⁾. Furthermore, ischemic injury can produce an injury response that may potentiate the risk of acute rejection ^(6,11) and interact with other forms of renal injury, ultimately leading to chronic allograft nephropathy ⁽⁷⁾. Notwithstanding these theories, several studies have called into question the true impact of DGF in the absence of other factors. For instance, some centers report a deleterious effect of DGF only when combined with early acute rejection episodes ⁽³⁾. Others have felt that DGF is deleterious only insofar as it reduces renal mass, leading directly to hyperfiltration injury ⁽¹²⁾.

Our data confirmed that early acute rejection was more prevalent in patients with DGF, but not acute rejections after discharge. This supports the hypothesis that DGF may increase the immunogenicity of the graft. Known effects of recovery from ischemic injury that may increase immunogenicity include up-regulation of HLA expression ⁽⁶⁾, release of inflammatory cytokines ⁽¹³⁾, stimulation of nitric oxide synthase ⁽¹⁴⁾, and release of chemokines ⁽¹⁵⁾. Nevertheless, DGF lowered graft survival and half-life whether or not acute rejection occurred during the initial hospitalization or in the first 6 months. This implies either a subclinical degree of acute rejection (which has been detected in studies of protocol biopsies ^[16,17]) or a direct nonimmune injury to the kidney. There is considerable evidence that nonimmune and immune factors interact synergistically to produce chronic allograft nephropathy ⁽¹⁸⁾. Regardless of the mechanism of renal damage, a stereotypical injury response can lead to a cycle of inflammation, further injury, cellular stress and, eventually, fibrosis and sclerosis ⁽¹⁹⁾.

That DGF was harmful in 0-1 HLA-mismatched kidneys suggests that induction of HLA expression alone may not be a sufficient explanation for poor graft survival in DGF, contrary to a previous single-center study ⁽²⁰⁾. There is no disputing that the long-term survival of well matched cadaveric kidneys is superior ⁽²¹⁾. However, our study demonstrated that in the face of DGF, the survival advantage of well matched kidneys over very poorly matched kidneys disappears. In fact, the graft half-life of 5-6 HLA-mismatched kidneys without DGF or early acute rejection was superior to the half-life of any 0-1 HLA-mismatched kidney that had DGF and/or early acute rejection. HLA-matched kidneys from donors more than 60 years of age reportedly provided no short- or long-term graft survival benefit ⁽²²⁾. The incidence of DGF among kidneys from donors more than 60 years of age was very high in that study, and diminishing nephron mass in the aging kidney was suggested as the underlying cause of failures that were attributed to chronic rejection.

Others have suggested that loss of nephron mass with resultant hyperfiltration injury may be the "final common pathway" leading to chronic allograft nephropathy ⁽¹²⁾. In animal models of chronic rejection, changes of nephron mass alone can modulate immune and inflammatory responses. In a prior study examining the effect of DGF in primary cadaveric transplant recipients from 1987 to 1997, we found that DGF significantly reduced 1-year allograft survival and half-life whether the discharge creatinine was above or below 2.0 mg/dl ⁽⁹⁾. In the present study, however, the effect of DGF was much less pronounced in patients with discharge creatinines below 2.5 mg/dl. Although this lends more support to the importance of reduced nephron mass in reducing graft survival, it should be noted that the present study lost significant power because of the reduced number of patients who fell in categories of DGF and/or acute rejection whose discharge creatinines were low.

Factors that lead to nephron loss are additive. Stepwise reduction in graft survival and half-life is seen when combining DGF and early acute rejection with elevated discharge creatinine, prolonged CIT, or older donor age (46-60 years). In particular, the graft survival seen in DGF kidneys from older donors (but still under 61 years of age) is typical of renal allograft survival in the precyclosporine era. Even in the absence of acute rejection, the half-life of kidneys in younger donors without DGF was more than 2.5 times that of kidneys from older donors with DGF (16.7 vs. 6.2 years). Unfortunately, older donors with preexisting medical problems are becoming more the rule than the exception ⁽²²⁾. One solution has been the use of both donor kidneys from older marginal donors, an approach that has resulted in excellent short-term graft survival ⁽²³⁾. It should be noted, however, that the CIT and DGF rates in studies using double adult kidneys have been very low, which may also contribute to improved survival independent of nephron mass.

An inherent limitation of the analysis of large multicenter registries such as ours is the variation among centers in their practices and policies that affect the rates of DGF and treatment of suspected rejection episodes. Nevertheless, the very large numbers of cases from the more than 250 U.S. transplant centers available for these analyses provide a valuable overview of the effects of delayed function and rejection during a relatively recent period of time. What is lost in uniformity of approach is made up for in statistical power.

In conclusion, the findings in this study support the theory that DGF can influence long-term graft survival both through acute rejection and by nonimmune mechanisms. Current immunosuppressive protocols are "lymphocentric"; that is, they focus exclusively on the prevention of T-cell mediated acute rejection, often at the expense of nephrotoxicity, chronic ischemia, and infection. To reduce chronic graft loss, we therefore require a paradigm shift that takes a "nephrocentric" approach: one that minimizes all forms of renal injury, not just acute rejection ⁽⁷⁾.