DELAYED GRAFT FUNCTION DOES NOT REDUCE THE SURVIVAL OF RENAL TRANSPLANT ALLOGRAFTS

Abstract

Background. 

The aim of the present study was to investigate the effect of delayed graft function (DGF) in graft outcome when adjusted by the presence of acute rejection in the first month after transplantation.

Methods. 

A total of 437 cadaveric renal transplant patients on cyclosporine and steroids were included in the study. Variables related to donor, recipient, and graft were prospectively collected.

Results. 

The incidence of DGF was 44.4%. When patients dying with a functioning graft were censored, graft survival rates at 1 and 6 years were similar in patients with immediate function to those with DGF, when rejection was not present (96% and 81% vs. 95% and 83%, respectively). Rejection negatively influenced graft survival rates at 1 and 6 years, both in patients with immediate graft function (80% and 73%, P<0.05 vs. no DGF/no rejection) and more deeply in those with associated DGF (77% and 62%,P<0.001 vs. no DGF/no rejection). Rejection was more frequently diagnosed in patients with DGF than in those with immediate graft function(50% vs. 39.9%, P<0.05). Length of hospitalization was longer and the number of needle core biopsies was higher in patients with DGF or rejection. The presence of both complications had an additive effect.

Conclusions. 

This study showed that DGF did not adversely affect kidney graft survival in patients without rejection. However, it increased the length of hospitalization and the number of graft biopsies, thus increasing the cost of transplantation. Moreover, rejection was more frequent in patients with DGF, and it had a negative impact on graft outcome. Because the association of DGF and rejection gave the poorest outcome, an effort should be made to prevent both complications.

*Abbreviations: DGF, delayed grat function; PRA, panel-reactive antibodies.

Delayed graft function (DGF*) after cadaveric transplantation still remains an important clinical problem. Its incidence rate is about 25%^(1-10), but rates as low as 6% ⁽¹¹⁾ and as high as 50%^(12-14) have been reported. A nonfunctioning kidney after transplantation constitutes a challenge for the clinician because it makes it difficult to establish a diagnosis of technical and immunological complications, and consequently, more frequent use of invasive and noninvasive methods are required. Fluid and electrolyte management become more complicated in the postoperative period, and the complexity of the management of immunosuppressive drugs increases. The necessity of dialysis might cause morbidity associated to the procedure⁽⁷⁾. Moreover, the hospitalization is prolonged, and the cost of transplantation is higher ^(3,15). Several conditions such as donor hypotension, donor age, prolonged cold ischemia, method of preservation, harvesting procedure, prolonged anastomosis time, perioperative management of the recipient, and immunological variables^(2-13) have been reported to be associated with the occurrence of this complication. However, the exact causes have still to be clarified.

Whether primary dysfunction of the kidney graft has a negative influence on both patient and graft survival rates is a matter of debate. Decreased graft survival rates have been correlated with DGF in a number of studies^{(3-6,8,9,14,16-20)}, but other reports have found a limited impact or no impact on graft survival rates ^(1,7,21-26). Because an increased rate of graft rejection episodes has been found in recipients with DGF, the condition has been considered a predisposing factor, and the deleterious effect on graft survival rates could be mediated by graft rejection⁽¹⁸⁾. Other authors suggest that DGF can interact in synergy with rejection to result in decreased long-term graft survival rates^(7,9,10). More recently, Feldman et al.⁽¹⁴⁾ have shown that the deleterious effects on cadaveric allograft survival rates of DGF and graft rejection are independent from each other.

As a result of these considerations, the present study was performed with two main objectives: (1) determination of the incidence of DGF and identification of the possible risk factors involved in its development, and(2) analysis of the effect of posttransplant DGF in graft outcome and its relationship with graft rejection. Our study population was our cadaveric kidney allograft recipients with an unchanged immunosuppressive treatment consisting of cyclosporine and prednisone. Data from donor, recipient, and surgical procedure were prospectively collected throughout 10 years of surveillance.

MATERIALS AND METHODS

Between March 1, 1986 and December 31, 1996, a total of 461 cadaveric renal transplants (412 first grafts and 49 retransplants) were performed at the Hospital Ramon y Cajal. Twenty-four transplants lost due to early vascular complications were excluded from the study. There were 272 males and 165 females, the mean age was 40.1±13.5 years (range 18 to 70 years), and the mean time on dialysis was 35.6±33.2 months (range 0.5 to 180 months). A total of 192 patients (43.9%) had never been transfused. The follow-up period was 42.3±26.6 months (range 6 to 72 months). Data of donors (age, gender, multiple-organ or kidney-only donor), recipients (age at transplantation, gender, primary renal disease, pretransplant duration of dialysis, pretransplant blood transfusions, peak and current panel-reactive antibody [PRA] level, previous transplants), transplant variables (warm ischemia time, cold ischemia time, anastomosis time, HLA-DR mismatches, organ disposition local or shared, perfusion solution), and graft and patient outcome were prospectively collected. All patients were followed until death, resumption of chronic dialysis, loss to follow-up, or May 31, 1997, whichever occurred first.

The kidneys were procured from heart-beating donors, according to standard techniques, either locally or at other hospitals. The preservation solution for cold storage consisted of Euro-Collins solution for kidney-only donors or University of Wisconsin solution in the cases of multiple-organ donations. All kidneys were preserved by simple cold storage. The grafts were placed in the left or right iliac fossa through an extraperitoneal approach. Ureteral drainage was reestablished with an extravesical Leadbetter-Politano ureteroneocystostomy. Methylprednisolone (250 mg) and furosemide (80 mg) were routinely given by intravenous administration during the recipient operation before the renal artery was unclamped.

The immunosuppressive regimen included corticoids and cyclosporine. Methylprednisolone at 1 mg/kg/day was administered intravenously for 2 days, followed by oral prednisone at 0.5 mg/kg/day for a month, reducing to 20 mg on day 60 and to a maintenance dose of 10 mg/day at 1 year. Cyclosporine was begun at an initial dose of 5 mg/kg, given inravenously before surgery, followed by 3 mg/kg/day i.v. for 2 days and then 12 mg/kg/day in two doses, adjusted thereafter to maintain a trough blood level within a range of 400-800 ng/ml (polyclonal antibody with Tdx system; Abbott Laboratories, Diagnostic Division, North Chicago, IL) at the end of the induction treatment. Beginning in 1990, the initial cyclosporine dose was changed to 10 mg/kg/day, given orally, followed by 8 mg/kg/day in the following days, and adjusted to maintain total blood levels within a range of 150-250 ng/ml(monoclonal antibody with Tdx system; Abbott Laboratories Diagnostic Division). Azathioprine was not routinely administered after transplantation. Since 1989, azathioprine has only been added for patients in whom cyclosporine nephrotoxicity or chronic graft dysfunction develop⁽²⁷⁾. Episodes of rejection were treated with intravenous methylprednisolone, 250 mg daily for 3 to 7 consecutive days. No more than three courses of antirejection therapy were administered. Beginning in 1989, steroid-resistant rejection episodes were treated with OKT3, administered at a dose of 5 mg/day for 7 to 14 days.

Perioperative imaging studies (renal scintigram or graft ultrasound) were obtained every 2-4 days if the urine output was below 400 ml/day during the early posttransplant period, if the creatinine level failed to decrease appropriately, or in the case of graft function deterioration. DGF was defined as requiring dialysis in the immediate perioperative period, regardless of urinary volume. In the case of DGF, a core biopsy was obtained on the fifth to ninth day after transplant and every 7 days until renal function was adequate. Acute rejection was diagnosed by standard clinical criteria with histological confirmation when necessary. Graft biopsies were also performed whenever graft function deterioration occurred.

All results are reported as mean ± SD. The chi-square test was used to compare categorical data, and the Student's test was performed when indicated. Multiple logistic regression was used to determine the adjusted effect of the binary independent variables on DGF. The variables included in the model were: recipient age, recipient gender, pretransplant time on dialysis, pretransplant blood transfusions, peak and current panel-reactive antibody levels, donor age, donor gender, cold ischemia time, anastomosis time, organ disposition (local or shared), type of donor (multiple-organ or kidney-only donor), and HLA-DR mismatches. Graft and patient survival curves were calculated using the Kaplan-Meier test. Survival rates between the different groups were compared using the log-rank test. A Cox regression model was used to analyze the effect of DGF and rejection status on patient and graft survival rates.

RESULTS

The overall incidence of DGF was 44.4% (194 grafts). It was 47.5% for the periods from 1986 to 1990 and from 1991 to 93 and declined to 37.5% for the period from 1994 to 1996 (P=0.055). The characteristics of the recipients with immediate graft function and with DGF are summarized inTable 1. Time on dialysis was longer in recipients with DGF: 39.2±34.8 months vs. 32.9±31.7 months(P<0.05). Interestingly, acute rejection during the first month occurred in 39.9% of the patients with immediate graft function and in 50% of those with DGF (P<0.05). Logistic regression analysis was used to identify variables influencing the development of DGF. The risk factors and associated odds ratios for DGF are shown in Table 2. Among these predictive factors, the strongest associated was observed for male recipients and kidney-only donors. Donor age >50 years, cold ischemia time >24 hr, and time on dialysis were also associated with DGF. Because the number of recipients with a cold ischemia time lower than 12 hr and higher than 36 hr was very small (seven recipients and three recipients, respectively), we were unable to classify recipients according to the cold ischemia time in more than two groups. Organ retrieval has changed in the last years: multiple-organ donors accounted for 9.2% of renal transplants for the period from 1986 to 1990 and increased thereafter to 85% in 1996. The incidence of DGF was lower with kidneys obtained from multiple-organ donors than that observed with kidney-only donors (37.2% vs. 62.7%,P<0.01) in the period from January 1990 to December 1996. Moreover, DGF had a less severe course: length of DGF, number of hemodialysis sessions, and days of hospitalization were lower in recipients with kidney transplants from multiple-organ donors (Table 3). Kidney graft function measured by the levels of serum creatinine at 12 months after transplantation was slightly better in recipients without DGF than in those with DGF (1.6±0.7 vs. 1.9±1.2 mg/dl, respectively,P<0.05), but it was similar after 2 years.

Table 1: Characteristics of cadaveric renal transplant recipients grouped according to early graft function(n=437)

Table 2: Risk factors for DGF obtained by logistic regression

Table 3: Characteristics of DGF according to the type of donor

One-hundred and nineteen grafts (27.2%) were lost during follow-up. The most frequent cause was acute and chronic rejection (n=55, 46.2%), followed by death with function (n=33, 27.7%), recurrence of primary renal disease(n=8, 6.7%), and other causes (n=23, 19.3%). Graft survival rates for the whole study group were 86% at 1 year and 66% at 6 years. Univariate survival analysis demonstrated that DGF did not compromise long-term graft survival rates (Fig. 1): graft survival rates at 1 and 6 years were 87% and 68% for the no-DGF group and 83% and 63% for the DGF group(P=0.261). Even in recipients older than 55 years, DGF did not decrease graft survival rates: 83% and 62% at 1 and 6 years for recipients with immediate graft function; and 86% and 68% for recipients with DGF. However, acute rejection during the first month after transplantation had a deleterious effect on graft survival rates: 94% and 74% at 1 and 6 years for recipients without rejection; and 75% and 55% at 1 and 6 years for recipients with rejection (P<0.0001) (Fig. 2).

Figure 1: Actuarial graft survival. Effects of immediate function and DGF (P=0.261).

Figure 2: Actuarial graft survival. Effects of rejection in the first month and no rejection (P<0.0001).

When the patients were grouped according to DGF and rejection, 147 recipients (33.6%) had neither DGF nor acute transplant rejection (group 1), 96 recipients (22%) had rejection only (group 2), 96 recipients (22%) had DGF but no rejection (group 3), and 98 (22.4%) had both DGF and rejection (group 4). As can be observed in Figure 3, kidney graft survival rates were similar at 1 and 6 years in groups 1 and 3 (95% and 75% vs. 93% and 72%, respectively). Furthermore, the presence of DGF did not have an additive deleterious effect to rejection on graft survival rates at 1 and 6 years, which were 76% and 57% for group 2 and 74% and 54% for group 4 (group 1 vs. 2, P<0.01; group 1 vs. 4,P<0.001; group 2 vs. 3, P<0.05; group 3 vs. 4, P<0.01). When patients dying with a functioning graft were censored (Fig. 4), graft survival rates at 1 and 6 years were similar in recipients without DGF to recipients with DGF if rejection was not present (96% and 81% vs. 95% and 83%, respectively). In addition, graft survival rates decreased in patients with acute rejection to 80% at 1 year and 73% at 6 years (P<0.05 vs. no DGF/no rejection, and P0.05 vs. DGF/no rejection), and it was even lower when the two complications were present (77% at 1 year and 62% at 6 year in patients with DGF/rejection) (P<0.001 vs. no DGF/no rejection, and P<0.01 vs. DGF/no rejection). Using the Cox proportional hazard model to explore the factors influencing graft survival rates, rejection during the first month after transplantation was related to decreased graft survival rates at 6 years (relative risk=1.79,P<0.01) as was type of donor (kidney only) (relative risk=1.85,P<0.01). Other variables (cold ischemia >24 hr, DGF, age of donor, age of recipient at transplantation, HLA-DR match, and PRA at transplantation) were not associated with a significant risk of graft loss.

Figure 3: Actuarial graft survival. Effects of DGF and rejection. No DGF/no rejection versus no DGF/rejection(P<0.01) and versus DGF/rejection (P<0.001). DGF/no rejection versus no DGF/rejection (P<0.05) and versus DGF/rejection (P<0.01).

Figure 4: Actuarial graft survival. Effects of DGF and rejection when patients dying with a functioning graft were censored. No DGF/no rejection versus no DGF/rejection (P<0.05) and versus DGF/rejection (P<0.001). DGF/no rejection versus no DGF/rejection(P<0.05) and versus DGF/rejection (P<0.01).

Recipients with DGF had a longer hospital length of stay when compared with recipients without DGF, but it was similar to that caused by rejection. Recipients with both complications had the longest hospital stay: 16.5±11.4 days for recipients without either DGF or rejection; 29.4±21.2 days for recipients without DGF and with rejection; 31.2±25.8 days for recipients with DGF and no rejection; and 39.9±21.9 days for recipients with both DGF and rejection(P<0.001 vs. the other three groups). The number of needle core biopsies was studied in 123 consecutive transplants. A total of 86 biopsies were performed, and 40% of patients had at least one core biopsy during the first admission. The percentage of patients biopsied was 10% for those without either DGF or rejection (4 of 38), 38.2% for those without DGF and with rejection (13 of 34), 42.4% for those with DGF and no rejection (11 of 26), and 100% for those with DGF and rejection (25 of 25).

During follow-up, 44 patients (11.2%) died, 12 of them (27.3%) after returning to dialysis. Cardiovascular disease was the leading cause of death(n=16, 36.4%); infection (n=11, 25%), liver failure (n=5, 11.4%), malignancy (n=5, 11.4%), and other causes (n=7, 15.9%) accounted for the remaining deaths. Overall patient survival rates were 96% at 1 year and 85.7% at 6 years. By univariate analysis, patient survival rate was influenced by the age of the recipient (P<0.0001), rejection(P<0.01), and type of donor (P=0.03). Patient survival rate was not affected by DGF: 96.2% at 1 year and 85.1% at 6 years for recipients without DGF and 95.8% at 1 year and 85.8% at 6 years for recipients with DGF. By Cox regression analysis, recipient survival rate was only significantly affected by age >55 years at transplantation (relative risk=3.20, P<0.001) and was nearly significantly affected by rejection (relative risk=1.80, P<0.1).

DISCUSSION

As in other studies ^(7,21-26,28), DGF did not constitute an independent risk factor for short-term and long-term graft survival rates. Following some previous studies ^(7,9,10), we explored the relationship of DGF to acute rejection and allograft survival rates. When patients were classified according to the presence or absence of DGF and acute rejection, graft survival rates were similar in patients with or without DGF when the patients were free from rejection during the first month after transplant. By multivariate analysis, DGF was not a significant risk factor for decreased graft survival rates for patients without rejection, but rejection significantly decreased graft survival rates, particularly in combination with DGF. Similar results have been reported by Troppman et al.⁽⁷⁾. These findings are not in agreement with those published in other recent papers. Ojo et al. ⁽⁹⁾ observed that DGF and rejection yielded an identical poor graft survival rate when compared with recipients without either DGF or rejection, and their simultaneous occurrence had an additive negative effect. Identical results have been reported by Nicholson et al. ⁽¹⁰⁾, who also found that DGF exerts the single most important deleterious effect on graft outcome. For other authors ⁽¹⁴⁾, DGF and acute graft rejection have important independent and deleterious effects on cadaveric allograft survival rates, but the effects of the two conditions were separately evaluated. The discrepancies with the first two studies are not easy to explain. The definition of DGF and the inclusion criteria were similar to ours. The differences from our findings could be due to the very low graft survival rate in recipients with DGF observed in these series, particularly when associated with rejection. This poor graft survival rate could be attributed to a more severe graft damage caused by the conditions contributing to the development of DGF. Two examples could be a longer ischemia time (we have a very small number of patients with cold ischemia time above 36 hr) or immunologic factors, such as higher PRA (high PRA was rare in our series). However, data related to the severity of DGF are not available.

Our results confirmed the increased incidence of rejection during the first months after transplantation in patients with DGF ^{(7,9,14,16,18,25,29,30)}. The reasons are still speculative. Routine graft biopsies at 7 days after transplantation in patients with DGF might have overdiagnosed mild rejection episodes, the diagnosis of which was based solely on histological evidence of an interstitial inflammatory cell infiltration, which may occur even in the absence of other clinical signs of acute rejection ^(31,32). Alternatively, histological examination can diagnose mild rejection episodes that would have otherwise passed unnoticed. Furthermore, ischemia might be a predisposition to rejection by up-regulating DR expression, and DGF would be a manifestation of early accelerated rejection ^(3,33). On the other hand, Yokoyama et al. ⁽³⁴⁾ concluded that prolonged DGF was associated with a higher incidence of graft failure, mostly due to chronic rejection, and they suggested that chronic rejection might be related to undiagnosed acute rejection episodes occurring during the period of DGF.

The 44.4% prevalence of DGF in the present study was higher than the prevalence found in our recipients on azathioprine ⁽³⁵⁾ and the prevalence recently reported in some series of patients on cyclosporine ^(1-11,31,34). However, the prevalence was similar to that in some other studies of patients on cyclosporine ^{(12,14,22,25,28,30)}. Conditions influencing the incidence of acute renal failure were: male recipient, kidney-only donor, donor age >50 years, cold ischemia time >24 hr, and time on dialysis. All of these had previously been identified as risk factors for DGF^(5,9,17,36). Other factors, such as HLA matching, number of pre-transplant blood transfusions, retransplantation, antibody status, anastomosis time, diabetes, and cytomegalovirus status did not influence the development of DGF in our recipients ^{(5,9,17,24,34,36,37)}. In general, total ischemia time was the variable most frequently associated with DGF^{(2,3,6,7,9,13,17,23-26,28,36-39)}. Male gender has been found to be associated with DGF by other investigators⁽³⁷⁾; the explanation for this association is not obvious at present, but in nontransplant patients, male gender constitutes a risk factor for the development of acute tubular necrosis⁽⁴⁰⁾. Unexpectedly, having a kidney-only donor was a strong predictor of DGF in our study. It has been published previously that having a multiple-organ donor increased the rate of DGF with respect to having a kidney-only donor ⁽²⁵⁾ or had no effect⁽¹⁷⁾, but there are no previous reports on the benefits of multiple-organ donation in the early evolution of kidney allografts. Having a multiple-organ donor reduced the rate of DGF and ameliorated its severity in those recipients, because the duration (defined as the days before a spontaneous fall in serum creatinine level and the number of dialysis sessions) was lower than in recipients with DGF and organs from kidney-only donors. Among the possible explanations: the utilization of University of Wisconsin solution in multiple-organ donors could have been responsible for this positive effect, because this preservation solution decreased the rate of initial nonfunction in a prospective randomized European trial ⁽⁴⁾. A further explanation could be the quality of donors: kidney-only donors were, in the last years, those who were not appropriate to be multiple-organ donors.

DGF led to a prolonged hospital length of stay and a higher number of graft biopsies when compared with recipients with immediate function^(15,16,24). This accounted for an increased cost of transplantation ^(3,15). Hospitalization and number of biopsies have been analyzed comparing recipients with or without immediate renal function, and there are few data comparing DGF and rejection ⁽⁴¹⁾. We have observed that DGF and acute rejection prolonged the hospital length of stay to the same extent and increased the number of diagnostic procedures, such as graft biopsy, when compared with recipients without either of these complications. The presence of both DGF and rejection increased the length of hospitalization and the number of biopsies even more.

In conclusion, our study demonstrates that DGF per se does not decrease short- and long-term graft survival rates, but it prolongs the length of hospital stay and the number of diagnostic procedures and consequently increases the cost of transplantation. Moreover, recipients with DGF have a higher risk of acute rejection, which has a detrimental effect on graft survival. As the association of DGF and acute rejection yield the poorest graft survival rates and the highest cost, efforts have to be made to prevent both complications, and therefore to improve transplant outcome. These include avoiding long cold ischemia time, careful monitoring and maintenance of the donor, and using immunosuppression protocols according to the immunological risk.