Kidney transplantation is established as the treatment of choice for most patients with end-stage renal disease (1, 2). Advances in immunosuppression, surgical techniques, and medical management during the past three decades have significantly improved the early patient and graft outcomes of kidney transplantation (3). Unfortunately, this success in early outcomes has not been paralleled by an improvement in long-term outcomes. In fact, long-term graft survival has declined despite significant decrease in acute rejection rates since the year 1995 (4, 5). The cause of poor long-term graft outcome is probably multifactorial and not well understood. Currently, delineated major causes of death-censored graft failure (DCGF) include acute rejection, interstitial fibrosis/tubular atrophy (IFTA), transplant glomerulopathy, recurrent or de novo renal diseases, and a variety of other medical and surgical causes (6, 7). It is currently believed that the major contributors to DCGF are IFTA and transplant glomerulopathy (7, 8). The incidence and causes of graft failure often change when the demographics of the donor and recipient populations and immunosuppressive regimens change.
Infection remains a serious threat to successful renal transplantation especially with the recent advent of more potent immunosuppressive drugs. Currently, infection accounts for 17.2% to 29.7% of recipient deaths and is the second most common cause of death with functioning graft (9–12). Some recent studies have reported that infection-related death has become the major cause of renal transplant recipient mortality (13–15). Furthermore, the risk for all types of infections (nonfatal or fatal) has been increasing in renal transplant recipients since 1995, in particular bacterial infections in the elderly (16, 17). The introduction of newer immunosuppressive agents while significantly decreasing acute rejection rates is also responsible for the emergence of serious infections such as cytomegalovirus (CMV) and polyoma viral infections. In the face of these increasing infectious complications, the impact of infection as the cause of DCGF in renal transplant recipients is not well studied. The aim of our study is to assess the trend in infection-related DCGF (where infection is the primary or the major contributing factor to graft loss) by analyzing the United Network for Organ Sharing (UNOS) database.
Total Burden of Infection-Related DCGF
A total of 189,110 primary/nonconcomitant renal transplants are recorded in the UNOS database during our study between January 1, 1990 and December 31, 2006. Of this total number of transplants, 38,286 (20.2%) recipients experienced DCGF. Among these 38,286 patients with DCGF, information was available as to whether infection did or did not contribute to DCGF in 31,326 recipients. These 31,326 recipients were included in the final analysis. Infection-related DCGF occurred in 2397 of 31,326 (7.7%) recipients, and the remaining 28,929 (92.3%) DCGFs were unrelated to infection.
Trends in Infection-Related DCGF
Because the use of any induction agent significantly influences the risk for infectious complications and reduces the risk of rejection, trends in these three interrelated variables over time (by transplant year) was analyzed. In addition, because of significant decrease in rejection rates in association with increasing infection-related morbidity and mortality noted in recent years with the use of potent newer immunosuppressive agents, we wanted to compare and study the relationship between rejection rates and infection-related DCGF. During the study period, the use of any induction immunosuppressive agent increased from 27.4% in 1990 to 89.3% in 2006, and the infection-related graft loss concomitantly increased (in fact almost doubled) from 6.4% in 1990 to 10.1% in 2006. During the same period, rejection rate (based on “treated for rejection”) decreased from 27.7% to 7.1%. A marked increase in the use of induction immunosuppression occurred around 1993 to 1994 from 31% to 83%, and this was accompanied by a decrease in rejection rate from 23.2% to 18.5%. Subsequently, the pronounced increase in infection-related DCGF was noted around 1996. Therefore, we compared the infection-related DCGF rates for periods before and after December 31, 1996. There was a statistically significant higher rate of infection-related DCGF from 6.3% in the 1990 to 1996 period compared with 9.1% in the 1997 to 2006 period (P<0.001). In addition, there was an inverse relationship between the trends in infection-related DCGF and rejection rates during the study period with the former exceeding the latter starting in 2005 as shown in Figure 1.
Age-Associated Trend in Infection-Related DCGF
Because immunocompetence declines with age, we evaluated the age-related trends in infection-related DCGF and rejection rates during the entire study period. The contribution of infection to DCGF showed a positive association with increasing patient age (4.8% in age group 20 to 24 years compared with 14.1% for age group >65 years) as shown in Figure 2. Conversely, rejection rate showed a negative relationship to increasing age (16.6% for age group 20 to 24 years compared with 13.3% for age group >65 years). The crossover between these two variables occurred between 60 and 64 years of age above which the rate of infection-related DCGF exceeded the rejection rate (Fig. 2).
Variables That Contributed Significantly to Infection-Related DCGF
The variables that showed statistically significant association for infection-related DCGF in univariate analysis are shown in Table 1. All variables shown in Table 1 had a positive association with an increased risk for infection-related DCGF except race: African Americans were less likely to have infection-related DCGF when compared with other races. Although induction agents were associated with an increased risk for infection-related DCGF, we were unable to demonstrate the influence of individual induction agents because of incomplete availability of data. Similarly, treatment for rejection that will be expected to increase the risk of infectious complications did not reach statistical significance, probably again because of incomplete data availability. Interestingly, history of cardiovascular or cerebrovascular disease before transplantation was found to be a significant risk factor for infection-related DCGF.
Multivariable Logistic Regression Analysis of Variables Contributing to Infection-Related DCGF
Renal allograft loss secondary to chronic rejection was the most common cause for death-censored graft loss recognized in UNOS database. Using it as the reference category for the specific causes of graft failure, various potential contributors to infection-related DCGF were analyzed by logistic regression, and the result is shown in Table 2. Among all the factors, the two most significant risk factor for infection- related graft loss include urological complications and polyoma viral infection with an odds ratio for graft loss of 8.77 (confidence interval: 5.11–14.93, P=<0.0001) and 2.55 (confidence interval 1.42–4.61, P=0.0019), respectively.
Kaplan-Meier Infection-Free Renal Allograft Survival
Finally, the probability of survival free of infection- related graft failure at 1, 5, 10, and 15 years after renal transplantation was analyzed using all the first and nonconcomitant kidney transplants performed during the study period. The Kaplan-Meier plot in Figure 3 summarizes the survival probability with values at different time intervals as shown within the figure.
To summarize the findings of our study based on the UNOS database, the risk of infection-related DCGF after primary/nonconcomitant renal transplantation increased from 6.4% in 1990 to 10.1% in 2006. Overall, during this 17-year period, infection contributed to 7.7% of all DCGF. The contribution of infection-related DCGF was most pronounced (14.1%) in recipients older than 65 years. Conversely, the rejection rate showed a decrease from 27.7% in 1990 to 7.1% in 2006. Logistic regression analysis identified many variables associated with an increased risk of infection-related DCGF with the most important being urological complications and polyoma virus infection with odds ratios of 8.77 and 2.55, respectively. The trends in infection-related DCGF and rejection rates showed an inverse relationship during the study period. In the final 2-year period of the study, the rate of infection-related DCGF was higher than the rejection rate. Finally, the probabilities of survival free of infection-related graft failure at 1, 5, 10, and 15 years after transplantation is shown as Kaplan-Meier plot in Figure 3.
Infection is an important cause of morbidity and mortality among transplant recipients. A major goal in the management of transplant recipients is to achieve a balance between maintaining adequate immunosuppression for prevention of rejection and avoiding overimmunosuppression with its attendant risks of infection and malignancy. Our findings suggest that this balance seems to have tilted toward overimmunosuppression during the 17-year period of our study coinciding with the availability and use of more potent immunosuppressive agents.
Infectious complications can contribute to DCGF in several ways: (1) infections can activate immune system and trigger cytokine release that can result in acute or chronic rejection that eventually results in IFTA, thereby resulting in graft loss (18–20). However, the extent of contribution of this mechanism to graft loss is not well defined. (2) Infections that directly involve the allograft such as bacterial pyelonephritis, BK virus nephropathy, CMV-related interstitial nephritis, and Epstein-Barr virus-related posttransplant lymphoproliferative disorder are well-documented causes of graft loss (7, 21, 22). A perfect example of infection-related DCGF is BK virus nephropathy that occurs in 5% to 10% of all renal transplant recipients and results in graft loss in 40% to 60% of affected patients (23–25). In a recent study by El-Zoghby et al. (7), BK virus nephropathy-mediated IFTA accounted for 23.4% of all IFTA, which in turn was responsible for 30.7% of all DCGF in that study. (3) CMV infection is an important contributor to graft loss by causing a variety of clinical problems in renal transplant recipients: CMV syndrome, tissue-invasive CMV disease, predisposition to superinfections, acute allograft rejection, chronic allograft dysfunction, new-onset posttransplant diabetes mellitus, and increased patient mortality. Opelz et al. and others showed that early CMV infection and disease are independent risk factors for allograft rejection and new-onset diabetes within 100 days of posttransplantation, and it is also significantly associated with recipient mortality beyond 100 days posttransplantation (21, 26, 27). (4) Severe systemic infections and sepsis frequently results in acute tubular necrosis that may not recover, ultimately result in graft failure or chronic allograft dysfunction (28, 29). (5) Clinically, severe infectious complications may mandate reduction or discontinuation of immunosuppression posttransplantation consequently resulting in acute or chronic rejection, and also, unavoidable nephrotoxicity of antimicrobial agents used to treat posttransplant infections are other mechanisms that may result in infection-related graft loss (30–33).
Understanding the individual causes of graft loss is important in achieving improved long-term outcomes after renal transplantation. A recent single-center study that aimed to identify the specific causes of renal allograft loss among 1317 transplant recipients during 50.3±32 months of follow-up showed a total of 330 renal allograft failure of which 153 were secondary to DCGF. In this study, infectious complications (sepsis, polyoma nephropathy, and acute and recurrent pyelonephritis) contributed to 17.4% and acute rejection episodes (excluding immunological causes of IFTA) to 11.7% of all DCGF (7). Similar to our own findings, this important study clearly demonstrates that in recent years, antirejection therapy has swung more toward overimmunosuppression with an increased risk for infectious complications contributing to DCGF more than acute rejection.
The weaknesses of our study include those inherent to any retrospective analysis of registry data, especially incomplete availability of information in the database. Further, information regarding specific microbiologic agents causing infections, site(s) of infection, and the specific mechanism(s) by which and the degree to which infection contributed to graft loss was unavailable. Other possible caveats in this study include advances in microbiologic diagnostic techniques as in case of BK virus nephropathy may have contributed to higher rate of infection-related DCGF in recent years, and finally uneven duration of posttransplant follow-up between recent and remote years of transplantation could have influenced the event rates. However, contrary to the expectation of remote years of transplantation with longer follow-up to have higher event rates, our study showed the opposite of lower infection-related DCGF during earlier years supporting our assumption of overimmunosuppression in recent years. Despite these drawbacks, the large number of renal transplants that we studied and the consistent trends that we observed during a 17-year period strengthen our conclusion that the use of potent induction agents (overall net state of immunosuppression) during the period of our study has significantly increased the risk of infection-related DCGF and may be an additional risk factor for poor long-term outcome. Our findings highlight the importance of minimization and individualization of immunosuppression, particularly in the elderly, especially given the increasing trend in the acceptance of older patients for renal transplantation in recent years. We suggest that in addition to rejection rates, all infectious complications should be prospectively monitored and detailed microbiologic data/site(s) of infection/exact mechanism(s) and contribution of infections to graft loss should be reported to national databases. This information should become an important benchmark of the quality of posttransplant outcomes and guide decisions about immunosuppressive management of transplant recipients.
MATERIALS AND METHODS
We analyzed the data on all kidney transplants performed between January 1, 1990 and December 31, 2006 available in the UNOS database. Only adult (older than 17 years) recipients of a first and nonconcomitant (combined renal and extrarenal transplants were excluded) renal transplants performed during this 17-year period in whom graft failure occurred and information was available as to whether infection did or did not contribute to DCGF were included in our analysis. Graft failure was considered infection related when infection was documented as a primary factor or a contributing factor to DCGF in the UNOS database. The year of transplantation was the main variable of interest to determine the trend in infection-related DCGF. We also analyzed the trends in the relationship between rejection rates (documented as treated for rejection), use of any induction immunosuppressive agent, and infection-related DCGF. The Student's two-sample t test for normally distributed numeric variables, the wilcoxon rank sum test for nonnormally distributed numeric variables, and the Chi-square test for categorical variables were used to perform univariable comparisons. Multivariable logistic regression analysis was performed to assess the independent contribution of individual variables to infection-related DCGF. Because the response that was evaluated was whether graft failure was caused by infection with in the set of all patients who experienced graft failure, logistic regression analytic method was applied in this study. The specific causes of graft failure used in the modeling (acute/hyperacute rejection, primary nonfunction, graft thrombosis, urological complications, recurrent disease, and polyoma virus) were evaluated in comparison with chronic rejection, which is the most commonly reported cause of DCGF in the UNOS database. Finally, Kaplan-Meier plots summarizing the freedom from infection-caused graft failure were analyzed among all the first and nonconcomitant kidney transplants performed during the study period.
The authors thank Mr. Emmanuel Kumar for his technical support in preparation of the final version of tables and figures.
1.Wolfe RA, Ashby VB, Milford EL, et al. Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med
1999; 341: 1725.
2.McDonald SP, Russ GR. Survival of recipients of cadaveric kidney transplants compared with those receiving dialysis treatment in Australia and New Zealand, 1991–2001. Nephrol Dial Transplant
2002; 17: 2212.
3.Hariharan S, Johnson CP, Bresnahan BA, et al. Improved graft survival after renal transplantation in the United States, 1988 to 1996. N Engl J Med
2000; 342: 605.
4.Meier-Kriesche HU, Schold JD, Srinivas TR, et al. Lack of improvement in renal allograft survival despite a marked decrease in acute rejection rates over the most recent era. Am J Transplant
2004; 4: 378.
5.Annual Data Report. Atlas of chronic kidney disease and end stage renal disease in the United States. Am J Kidney Dis
2008; 51: S1.
6.Womer KL, Kaplan B. Recent developments in kidney transplantation —A critical assessment. Am J Transplant
2009; 9: 1265.
7.El-Zoghby ZM, Stegall MD, Lager DJ, et al. Identifying specific causes of kidney allograft loss. Am J Transplant
2009; 9: 527.
8.Jevnikar AM, Mannon RB. Late kidney allograft loss: What we know about it, and what we can do about it. Clin J Am Soc Nephrol
2008; 3(suppl 2): S56.
9.Ojo AO, Hanson JA, Wolfe RA, et al. Long-term survival in renal transplant recipients with graft function. Kidney Int
2000; 57: 307.
10.Chang SH, Russ GR, Chadban SJ, et al. Trends in kidney transplantation in Australia and New Zealand, 1993–2004. Transplantation
2007; 84: 611.
11.Diethelm AG, Deierhoi MH, Hudson SL, et al. Progress in renal transplantation: A single center study of 3359 patients over 25 years. Ann Surg
1995; 221: 446.
12.Excerpts from the United States Renal Data System 2008 annual data report, transplantation. Am J Kidney Dis
2009; 53(suppl 1): S228.
13.Pirsh JD, Miller J, Dierhoi MH, et al. A comparison of tacrolimus (FK506) and cyclosporine for immunosuppression after cadaveric renal transplantation. FK506 Kidney Transplant Study Group. Transplantation
1997; 63: 977.
14.Kahan BD. Efficacy of sirolimus compared with azathioprine for reduction of acute allograft rejection: A randomised multicentre study. The Rapamune US Study Group. Lancet
2000; 356: 194.
15.MacDonald AS, for the Rapamune Global Study Group. A worldwide, phase III, randomized, controlled, safety and efficacy study of a sirolimus/cyclosporine regimen for prevention of acute rejection in recipients of primary mismatched renal allografts. Transplantation
2001; 71: 271.
16.Dharnidharka VR, Caillard S, Agodoa LY, et al. Infection frequency and profile in different age groups of kidney transplant recipients. Transplantation
2006; 81: 1662.
17.Dharnidharka VR, Agodoa LY, Abbott KC. Risk factors for hospitalization for bacterial or viral infection in renal transplant recipients: An analysis of USRDS data. Am J Transplant
2007; 7: 653.
18.Byrd LH, Tapia L, Cheigh JS, et al. Association between Streptococcus faecalis
urinary tract infections and graft rejection in kidney transplantation. Lancet
1978; 2: 1167.
19.Simmons RL, Weil R, Tallent MB, et al. Do mild infections trigger the rejection of renal allografts? Transplant Proc
1970; 2: 419.
20.Heemann UW, Tullius SG, Schmid C, et al. Infection associated cellular activation accelerates chronic renal allograft rejection in rats. Transpl Int
1996; 9: 137.
21.Opelz G, Dohler B, Ruhenstroth A. Cytomegalovirus prophylaxis and graft outcome in solid organ transplantation: A collaborative transplant study report. Am J Transplant
2004; 4: 928.
22.Randhawa PS, Magnone M, Jordan M, et al. Renal allograft involvement by Epstein-Barr virus associated post-transplant lymphoproliferative disease. Am J Surg Pathol
1996; 20: 563.
23.Hirsch HH, Knowles W, Dickenmann M, et al. Prospective study of polyomavirus type BK replication and nephropathy in renal-transplant recipients. N Engl J Med
2002; 347: 488.
24.Hariharan S. BK virus nephritis after renal transplantation. Kidney Int
2006; 69: 655.
25.Dall A, Hariharan S. BK virus nephritis after renal transplantation. Clin J Am Soc Nephrol
2008; 3(suppl 2): S68.
26.Sagedal SS, Hartmann A. Cytomegalovirus infection in renal transplant recipients is associated with impaired survival irrespective of expected mortality risk. Clin Transplant
2007; 21: 309.
27.Hartmann A, Sagedal S, Hjetmesaeth J. The natural course of cytomegalovirus infection and disease in renal transplant recipients. Transplantation
2006; 82: S15.
28.Mattoso R, Khouri N, de Jesus L, et al. Risk factors for graft dysfunction in the late period of renal transplantation. Transplant Proc
2009; 41: 1594.
29.Witzke O, Schmidt C, Kohnle M, et al. Impact of febrile infections on the long-term function of kidney allografts. J Urol
2001; 166: 2048.
30.Häyry P, Mennander A, Yilmaz S, et al. Towards understanding the pathophysiology of chronic rejection. Clin Investig
1992; 70: 780.
31.Kutinova A, Woodward RS, Ricci JF, et al. The incidence and costs of sepsis and pneumonia before and after renal transplantation in the United States. Am J Transplant
2006; 6: 129.
32.Pourmand G, Pourmand M, Salem S, et al. Posttransplant infectious complications: A prospective study on 142 kidney allograft recipients. Urol J
2006; 3: 23.
33.Manitpisitkul W, McCann E, Lee S, et al. Drug interactions in transplant patients: What everyone should know. Curr Opin Nephrol Hypertens
2009; 18: 379.