Several observational studies have associated acute rejection episodes (ARE) after renal transplantation with chronic allograft nephropathy and graft loss (1–5). However, not all ARE seem to have the same impact on long-term outcome. Number, type, severity, reversibility, and timing of rejection could be identified as determinants of renal graft survival (6–8). Recipients with one ARE have been found to show better survival rates than patients with two or more episodes of rejection (9–13). It could also be shown that the risk of graft loss increases with the severity of histological findings according to Banff criteria (5, 14). In addition, ARE followed by partial loss of graft function (as assessed by serum creatinine levels) exerted a more detrimental effect on long-term outcome than acute rejections with complete recovery (15–17). It has even been suggested that if serum creatinine levels return to baseline after ARE there is no increased risk of chronic allograft failure, and that an ARE predicts poor renal allograft survival only when associated with graft dysfunction (18, 19).
Several authors suggested an adverse effect of late ARE on long-term graft survival (1, 20–22). However, “late ARE” was inconsistently defined, ranging between more than 2 months and more than 1 year posttransplantation. The re-sponse to antirejection therapy in functional recovery depending on time of occurrence was usually not considered. We analyzed the association between timing of ARE and graft survival using retrospective data from the Collaborative Transplant Study (CTS) database (23). The long-term impact of reversibility of graft dysfunction in response to antirejection therapy at different time intervals after renal transplantation was also studied.
METHODS
Patients
Data from 28,867 renal transplant recipients receiving their graft between 1995 and 2005 from deceased donors formed the basis for this analysis. The data were reported to the CTS by 196 transplant centers from 39 countries (see Acknowledgements). Patients were included in the analysis if they had a functioning graft at the end of year 1, 2, or 3 after transplantation and if information was available on whether or not an ARE was treated during these years. These data were derived from centers completing the “Extended Follow Up” questionnaire, which is a voluntary part of the CTS. Patients were included if their serum creatinine concentration at the end of each interval considered in the analysis was known. Creatinine values were recorded in four categories: <130, 131–260, 261–400, and >400 μmol/L. Recipients of multiple organ transplants (e.g., kidney and liver or kidney and pancreas) were excluded. It was decided from the outset not to collect biopsy data because of the likelihood that such data would be controversial, based on reports showing that there is large variation among institutions in the histological grading of renal allograft biopsies and that the analysis of biopsy results combining data from different centers is problematic (24, 25).
Statistical Analysis
Graft, death-censored graft, and patient survival, as well as graft survival half-life were analyzed for the 3 years after the year during which ARE occurred. Survival curves were computed according to the Kaplan-Meier method, and statistical comparisons were made using the log-rank test or weighted regression analysis. The Mantel-Haenszel test was used for analysis of serum creatinine levels. Graft survival half-life and 95% confidence interval (CI) were calculated under the assumption of an exponential distribution of survival times (26). Multivariate analysis was performed using a Cox regression model to adjust for the potential confounders: year of transplantation, recipient geographical origin (continent), original disease leading to endstage renal disease, recipient and donor gender, race and age, number of HLA mismatches, preformed panel-reactive lymphocytotoxic antibodies, cold ischemia time, initial immunosuppressive medication, antibody induction, and time and number of rejections. Cox regression results are presented as hazard ratio (HR) and 95% CI. The software packages SPSS (version 15.0) and SAS (version 8.2) were used; P-values less than 0.05 were considered significant.
RESULTS
Study Population
A total of 28,867 patients qualified for inclusion in the analysis. Demographic characteristics of the patient population studied are listed in Table 1.
;)
TABLE 1: Characteristics of the study population
Incidence of Rejection and Time to Occurrence After Transplantation
Of the 4,449 patients whose graft functioned for at least 1 year and for whom the date of first rejection treatment during the first year was known, 83.2% experienced a first ARE within the first 3 months of posttransplantation, 9.4% at 4 to 6 months, and 7.4% at 7 to 12 months. More than one ARE was recorded in 21.6% of patients with rejection treatment during the first year.
Rejection in the First Year and Long-Term Outcomes
Patients who experienced an ARE during the first year had an inferior long-term graft survival when compared with patients without rejection, whereas patient survival was only modestly affected. After an additional follow-up of 3 years, 82.7%±0.5% (SE) of patients who were treated for rejection during the first year had a functioning graft when compared with 89.8%±0.2% of patients who were not treated for rejection (log rank P<0.001). Patient survival was affected to a lesser, albeit statistically significant degree (93.0%±0.3% vs. 94.6%±0.2%, P<0.001). When patient death was censored, the difference in graft survival between transplants in patients with or without ARE during the first year was 88.0%±0.4% vs. 94.5%±0.2%, respectively (P<0.001, data not shown).
Timing of Rejection, Response to Treatment, and Graft Loss
The time at which the first ARE was observed greatly influenced subsequent graft survival. As shown in Figure 1, the longer the interval between transplantation and ARE, the worse the long-term survival rate (weighted regression, P<0.001). Within the 90-day interval there was no further separation of outcome depending on time of rejection; for instance, rejections that occurred within 14 days or between 14 and 90 days showed virtually the same 3-year graft survival rate (86.2% vs. 85.5%, P=0.64, data not shown). Serum creatinine levels examined at 1 year posttransplantion showed a worsening with increasing length of time between transplant surgery and ARE (Fig. 2; Mantel-Haenszel, P<0.001), indicating that late ARE were more difficult to reverse by antirejection treatment than early ARE.
FIGURE 1.:
Graft survival in relation to time of first treated acute rejection episode (ARE) during first year after renal transplantation. Patients whose graft functioned at least 1 year were analyzed.
FIGURE 2.:
Fraction of patients with serum creatinine <130, 130–260, or >260 μmol/L at 1 year after renal transplantation by time of occurrence of first treated acute rejection episode.
The impact of early versus late ARE on subsequent graft survival can also be expressed by graft survival half-lives calculated for the period after the ARE. These results are shown in Table 2 for all patients, and for patients whose ARE was treated successfully (i.e., serum creatinine returned to <130 μmol/L) in comparison with patients without complete functional recovery (≥130 μmol/L). In all comparisons, a striking decrease in graft survival half-life was noted when early and late occurring ARE were compared. ARE followed by partial loss of graft function exerted a more detrimental effect on long-term outcome than ARE with complete recovery. Patients whose ARE occurred within the first 90 days showed a relatively high rate of return to normal serum creatinine (38%) and their long-term half-life was a good 25.7 years (Table 2). Importantly, patients with late occurring rejections but complete functional recovery showed half-life times comparable with patients with no or early rejections associated with impaired serum creatinine (Table 2). Results of an analysis of the competing influences of time of rejection and functional recovery after rejection treatment are illustrated in Figure 3.
TABLE 2: Half-life of graft survival calculated for the 3 years following the year during which rejection occurred
FIGURE 3.:
Impact on graft survival of change in serum creatinine from time before rejection to time after rejection treatment. Graft survival curves for patients who were not treated for rejection are shown for comparison. Patients with a first rejection during posttransplant interval 91 to 180 days (left) or 181 to 365 days (right) were analyzed. Horizontal arrows depict no increase, vertical arrows increase in serum creatinine. + and − identify patients with or without rejection, respectively.
Time of Occurrence and Magnitude of Risk
In an attempt to quantify the impact of ARE on long-term outcomes by time of occurrence, and to account for the influence of potential confounders, multivariate Cox regression analysis was performed. As shown in Table 3, the HR was distinctly higher for patients with later occurring ARE. Compared with patients free of rejection, the risk of graft failure beyond the first posttransplant year associated with an ARE within 3 months of transplantation was 1.35 (CI 1.20–1.52, P<0.001), whereas the risk increased to HR 2.05 (CI 1.61–2.62, P<0.001) for ARE occurring during 4 to 6 months, and further to HR 2.74 (CI 2.12–3.53, P<0.001) for ARE occurring during 7 to 12 months. In general, the risk of death censored graft loss (approximating immunological rejection) increased much more than the risk of death. ARE occurring during the second and third year were associated with HR more than 4 in the analysis of death-censored graft survival (Table 3).
TABLE 3: Hazard ratios (HR) for survival during the 3 years following year of rejection
Number of Rejections
The number of treated ARE was identified as an additional risk factor. Data given in Table 4 show that patients who experienced two rejections during the first year had a 1.51 times (CI 1.23–1.85, P<0.001) higher risk of graft failure than patients who had only one ARE, and patients who were treated for more than two rejections showed a further increase in risk to an HR of 1.94 (CI 1.42–2.66, P<0.001).
TABLE 4: Hazard ratios (HR) for number of rejections during first year
DISCUSSION
This retrospective analysis of a large cohort of renal transplant recipients addressed the effect of acute rejection at different times posttransplantation on reversibility of graft function and long-term graft failure. The principal findings can be summarized as follows: (i) long-term graft survival is significantly better in patients without ARE than that in patients with ARE during the first posttransplant year; (ii) ARE occurring during the first posttransplant year are increasingly deleterious for long-term outcome the later they occur; (iii) ARE occurring during the first 3 months and associated with a return to normal serum creatinine (<130 μmol/L) are associated with good long-term outcome; (iv) late ARE are usually more severe in nature than early ARE as reflected by incomplete functional recovery; (v) ARE followed by partial loss of graft function exert a more detrimental effect on long-term outcome than ARE with complete recovery; and (vi) patients experiencing a first rejection during the second or third year show particularly high HR (>4) for death censored graft survival.
An association between late ARE and reduced long-term graft survival has been reported in the literature; however, “late ARE” was inconsistently defined (1, 2, 20–22). In our analysis, the time from renal transplantation to first-treated ARE was divided into defined intervals. This enabled us to quantify the increase in long-term risk for different posttransplant time intervals during which ARE occurred. The results of the analysis of death censored graft survival, which approximates the rate of immunological rejection and shows the effect particularly strongly (Table 3), suggest that the differential impact of late versus early ARE most likely has an immunological cause. Recipient T cells recognizing donor class I molecules presented by recipient antigen presenting cells (reflecting the indirect pathway of allorecognition) were suggested to be important mediators of late ARE and chronic rejection (22, 27). Late ARE might signal ongoing immunological activity associated with a higher likelihood of chronic rejection and graft loss, as suggested by others (28, 29). One should also consider the possibility that late ARE may differ from early rejections because triggering by infections or other unknown triggers may play a role, and that nonadherence to immunosuppressive medication is more likely to affect late than early rejections.
Previous reports encouraged the use of elevated serum creatinine at 1 year as a surrogate marker for predicting long-term transplant survival. Posttransplant renal function within the first year has been correlated with ultimate graft failure in single center studies as well as large database analyses, and a serum creatinine of more than or equal to 130 μmol/L was shown to be associated with poor long-term survival (15, 30, 31). Moreover, it has been proposed that the degree of renal damage conferred by an ARE, as estimated by the difference in serum creatinine before and after the ARE (DeltaCr), is directly proportional to the severity of ARE (10). It was furthermore shown that the time from transplantation to the first ARE correlated well with DeltaCr (10, 32). Other authors reported that ARE severity (severe being defined as greater than 50% reduction in estimated glomerular filtration rate) is an independent risk factor of chronic rejection (33). Our analysis confirmed that late ARE are usually more difficult to reverse by antirejection treatment than early ARE. Thirty-eight percent of patients with an ARE during the first 3 months showed a return to serum creatinine less than 130 μmol/L in 1 year, when compared with only 26% of patients who experienced an ARE at a later time point. Normalization of serum creatinine after ARE was associated with superior long-term outcome as indicated by graft survival half-life and HR. This is in agreement with previous findings demonstrating that the risk for graft loss is different between ARE that lead to functional deterioration as opposed to those that do not (3). It has been suggested that ARE resulting in graft dysfunction are associated with a reduction in functional nephron mass leading to hyperfiltration injury of the remaining nephrons with subsequent chronic graft failure (10). In agreement with previous reports (2, 10, 12, 13), we identified the number of ARE as an additional predictor of decreased long-term graft survival.
A limitation of the present study is the absence of biopsy data. However, collecting biopsy data from 200 centers seemed impractical and would in all likelihood have introduced a bias. It has been shown that there is large variation between institutions in the histological grading of renal allograft biopsies (24, 25). Based on recent studies it seems that the incidence of treated ARE is approximately 4% to 8% higher than that confirmed by biopsy (34, 35). The incidence rates of ARE observed in our analysis may therefore have been slightly overestimated, although the transplants analyzed in this study were performed at experienced university hospitals and most ARE were probably biopsy-confirmed. A further limitation is inherent in the method of data collection from multiple centers which, for example, with respect to the exact assessment of graft function or the exact method of rejection treatment is less detailed than single-center data. However, it must also be considered that the advantage of gaining access to large numbers of cases for analysis resulting from multicenter collaboration enabled us to perform this analysis of quantitative effects that otherwise would not have been possible.
In conclusion, the results of our analysis show that the time period elapsed between renal transplantation and occurrence of first ARE plays a critical role in predicting functional recovery after rejection treatment and long-term graft survival.
ACKNOWLEDGMENTS
The authors are indebted to the following 196 transplant centers for generously providing the data on which this analysis was based.
Aachen, Adelaide (2), Akron, Ancona, Ankara, Antalya, Auckland, Augsburg, Australia, Banska-Bystrica, Baracaldo, Barcelona (2), Bari, Barquisimeto, Basel, Belfast, Belo Horizonte (2), Bergamo, Berlin, Bern, Bochum, Botucatu, Bremen, Brescia, Brisbane, Bristol, Budapest, Buenos-Aires (4), Cagliari, Cali, Cape Town (4), Caracas, Carshalton, Christchurch, Cincinnati, Cleveland, Cologne, Cordoba, Dallas (4), Debrecen, Detroit, Duesseldorf, Durban, Edmonton, Erlangen, Essen, Exeter, Florence, Frankfurt, Freiburg, Fulda, Gdansk, Geneva, Giessen, Glasgow, Goettingen, Grand Rapids, Guadalajara, Haifa, Halle, Hamilton (CDN), Hamilton (NZ), Hann-Muenden, Heidelberg, Helsinki, Homburg, Hong Kong, Innsbruck, Izmir, Jena, Jerusalem, Kaiserslautern, Kansas City, Kashihara, Katowice, Kaunas, Kiel, L'Aquila, Lausanne, Leicester, Leuven, Lexington, Liege, Lille, Lima, Limoges, Linz, Ljubljana, Louisville, Luebeck, Lyon, Maastricht, Maceio, Madrid, Mainz, Manila, Mannheim, Mar del Plata, Maracaibo, Marburg, Martin, Medellin, Melbourne (5), Mexico City, Milan (2), Modena, Muenster, Nancy, Nantes, Nashville, New Delhi, Newcastle (AUS), Newcastle u Tyne, Nottingham, Novara, Oklahoma, Omaha, Orlando, Oviedo, Pamplona, Panama, Parma, Pato Branco, Pavia, Pecs, Perth, Plzen, Poitiers, Porto Alegre, Prague, Quebec, Regensburg, Reims, Rennes, Rio de Janeiro, Rome, Rosario, Rostock, Santa Fe, Santander, Santiago, Sao Paulo, Seattle, Seoul, Shreveport, St. Etienne, St. Gallen, Stanford, Stony Brook, Strasbourg, Stuttgart, Sydney (5), Szeged, Tehran, Tel Aviv, Temuco, Toledo, Toulouse, Tours, Treviso, Tuebingen, Turin, Udine, Ulm, Uppsala, Valdivia, Valencia (2), Valhalla, Vilnius, Wellington, Wichita, Winnipeg, Wuerzburg, Zagreb, Zurich.
REFERENCES
1. Matas AJ, Gillingham KJ, Payne WD, Najarian JS. The impact of an acute rejection episode on long-term renal allograft survival (t1/2).
Transplantation 1994; 57: 857.
2. Humar A, Kerr S, Gillingham KJ, Matas AJ. Features of acute rejection that increase risk for chronic rejection.
Transplantation 1999; 68: 1200.
3. Meier-Kriesche HU, Ojo AO, Hanson JA, et al. Increased impact of acute rejection on chronic allograft failure in recent era.
Transplantation 2000; 70: 1098.
4. Hariharan S, Johnson CP, Bresnahan BA, Taranto SE, McIntosh MJ, Stablein D. Improved graft survival after renal transplantation in the United States, 1988 to 1996.
N Engl J Med 2000; 342: 605.
5. Pallardo Mateu LM, Sancho Calabuig A, Capdevila Plaza L, Franco Esteve A. Acute rejection and late renal transplant failure: Risk factors and prognosis.
Nephrol Dial Transplant 2004; 19(suppl 3): iii38.
6. Ponticelli C. Renal transplantation 2004: Where do we stand today?
Nephrol Dial Transplant 2004; 19: 2937.
7. Joosten SA, Sijpkens YW, van Kooten C, Paul LC. Chronic renal allograft rejection: Pathophysiologic considerations.
Kidney Int 2005; 68: 1.
8. Tantravahi J, Womer KL, Kaplan B. Why hasn't eliminating acute rejection improved graft survival?
Annu Rev Med 2007; 58: 369.
9. Vanrenterghem YF. Acute rejection and renal allograft outcome.
Nephrol Dial Transplant 1995; 10(suppl 1): 29.
10. Pelletier RP, Cosio F, Henry ML, et al. Acute rejection following renal transplantation. Evidence that severity is the best predictor of subsequent graft survival time.
Clin Transplant 1998; 12: 543.
11. Humar A, Payne WD, Sutherland DE, Matas AJ. Clinical determinants of multiple acute rejection episodes in kidney transplant recipients.
Transplantation 2000; 69: 2357.
12. Pascual M, Theruvath T, Kawai T, Tolkoff-Rubin N, Cosimi AB. Strategies to improve long-term outcomes after renal transplantation.
N Engl J Med 2002; 346: 580.
13. Dickenmann MJ, Nickeleit V, Tsinalis D, Gürke L, Mihatsch MJ, Thiel G. Why do kidney grafts fail? A long-term single-center experience.
Transpl Int 2002; 15: 508.
14. Bates WD, Davies DR, Welsh K, Gray DW, Fuggle SV, Morris PJ. An evaluation of the Banff classification of early renal allograft biopsies and correlation with outcome.
Nephrol Dial Transplant 1999; 14: 2364.
15. Opelz G, for the Collaborative Transplant Study. Critical evaluation of the association of acute with chronic graft rejection in kidney and heart transplant recipients.
Transplant Proc 1997; 29: 73.
16. Madden RL, Mulhern JG, Benedetto BJ, et al. Completely reversed acute rejection is not a significant risk factor for the development of chronic rejection in renal allograft recipients.
Transpl Int 2000; 13: 344.
17. Meier-Kriesche HU, Schold JD, Srinivas TR, Kaplan B. 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.
18. Cosio FG, Pelletier RP, Falkenhain ME, et al. Impact of acute rejection and early allograft function on renal allograft survival.
Transplantation 1997; 63: 1611.
19. Vereerstraeten P, Abramowicz D, de Pauw L, Kinnaert P. Absence of deleterious effect on long-term kidney graft survival of rejection episodes with complete functional recovery.
Transplantation 1997; 63: 1739.
20. Heaf JG, Ladefoged J. The effect of acute rejection on long-term renal graft survival is mainly related to initial renal damage.
Transpl Int 1998; 11(suppl 1): S26.
21. Joseph JT, Kingsmore DB, Junor BJ, et al. The impact of late acute rejection after cadaveric kidney transplantation.
Clin Transplant 2001; 15: 221.
22. Sijpkens YW, Doxiadis II, Mallat MJ, et al. Early versus late acute rejection episodes in renal transplantation.
Transplantation 2003; 75: 204.
23. Collaborative Transplant Study. Available at:
http://www.ctstransplant.org. Accessed August 2007.
24. Furness PN, Taub N. International variation in the interpretation of renal transplant biopsies: Report of the CERTPAP Project.
Kidney Int 2001; 60: 1998.
25. Furness PN, Taub N, Assmann KJ, et al. International variation in histologic grading is large, and persistent feedback does not improve reproducibility.
Am J Surg Pathol 2003; 27: 805.
26. Collet D. Modelling survival data in medical research. London: Chapman & Hall 1994; 116.
27. Sayegh MH. Why do we reject a graft? Role of indirect allorecognition in graft rejection.
Kidney Int 1999; 56: 1967.
28. Chapman JR, O'Connell PJ, Nankivell BJ. Chronic renal allograft dysfunction.
J Am Soc Nephrol 2005; 16: 3015.
29. McDonald S, Russ G, Campbell S, Chadban S. Kidney transplant rejection in Australia and New Zealand: Relationships between rejection and graft outcome.
Am J Transplant 2007; 7: 1201.
30. Hariharan S, McBride MA, Cherikh WS, Tolleris CB, Bresnahan BA, Johnson CP. Post-transplant renal function in the first year predicts long-term kidney transplant survival.
Kidney Int 2002; 62: 311.
31. Siddiqi N, McBride MA, Hariharan S. Similar risk profiles for post-transplant renal dysfunction and long-term graft failure: UNOS/OPTN database analysis.
Kidney Int 2004; 65: 1906.
32. Yates PJ, Nicholson ML. The aetiology and pathogenesis of chronic allograft nephropathy.
Transpl Immunol 2006; 16: 148.
33. Massy ZA, Guijarro C, Wiederkehr MR, Ma JZ, Kasiske BL. Chronic renal allograft rejection: Immunologic and nonimmunologic risk factors.
Kidney Int 1996; 49: 518.
34. Vincenti F, Friman S, Scheuermann E, et al. Results of an international, randomized trial comparing glucose metabolism disorders and outcome with cyclosporine versus tacrolimus.
Am J Transplant 2007; 7: 1506.
35. Silva HT Jr, Yang HC, Abouljoud M, et al. One-year results with extended-release tacrolimus/MMF, tacrolimus/MMF and cyclosporine/MMF in de novo kidney transplant recipients.
Am J Transplant 2007; 7: 595.
