Clin Transplantation 2001: 15: 136–141Printed in Ireland. All rights reser6ed

Cyclosporin A withdrawal in live relatedrenal transplantation: long-term results

Dubey D, Kumar A, Srivastava A, Mandhani A, Sharma AP, GuptaA, Sharma RK. Cyclosporin A withdrawal in live related renal trans-plantation: long-term results.Clin Transplantation 2001: 15: 136–141. © Munksgaard, 2001

Abstract: Cyclosporin A (CsA) withdrawal after 1 yr of stable graftfunction has been shown to be beneficial in cadaveric renal transplanta-tion. This strategy could be even more suitable for ‘immunologicallyadvantaged’ grafts as in live related renal transplantation. We reportthe long-term outcome of patients in a live related transplantation pro-gramme undergoing early (between 1989 and 1992) and late (1993 on-wards) CsA withdrawal as compared with those on long-term low doseCsA (1993 onwards). Two-hundred and fifty-two patients were dividedinto three groups based on the following immunosuppressive protocol:group ECyW (n=99), early CsA withdrawal (9 months after trans-plantation); group LCyW (n=44), late CsA withdrawal (median 16months, range 13–22 months after transplantation); and group LDCy(n=109), long-term low dose CsA. The median period of follow-upwas 66 months after transplantation (range 43–84 months). There wasno difference in the actuarial 6-yr patient or graft survival among thethree groups. Acute rejection episodes were more frequent in ECyW(54.4%) than in LDCy (31.8%) and LCyW (23.8%) (p=0.001). Therisk of developing late (]9 months) acute rejection was highest inECyW 32/99 (32.3%) as compared with LCyW 8/44 (18.4%; p=0.08)and LDCy 8/109 (7.3%; p=0.0001). Of the 32 ECyW patients whodeveloped acute rejection episodes after CsA withdrawal, 13 (40.6%)lost their grafts either due to uncontrolled acute rejection or to chronicrejection. Chronic rejection was higher in ECyW (24%) than in LCyW(11%; p=0.04) and LDCy (17%; p=0.17). Antihypertensive require-ment was highest in patients maintained on low dose CsA. Graft func-tion, as measured by serum creatinine levels, was significantly better inLCyW (1.2490.4 mg%) as compared with ECyW (1.4990.5 mg%)and LDCy (1.4890.6 mg%). Early CsA withdrawal after live relatedrenal transplantation is associated with a significant risk of acute rejec-tion and subsequent chronic rejection. Slow withdrawal after 1 yr issafe and more economical than the long-term administration of lowdose CsA.

Deepak Dubeya, Anant Kumara,Aneesh Srivastavaa,Anil Mandhania, AjayPrakash Sharmab, Amit Guptab

and RK Sharmab

Departments of a Urology andb Nephrology, Sanjay Gandhi PostgraduateInstitute of Medical Sciences, Lucknow,India

Key words: cyclosporin A withdrawal –immunosuppression – renal transplantation

Corresponding author: Dr Anant Kumar,Additional Professor & Head, Department ofUrology, Sanjay Gandhi Postgraduate Insti-tute of Medical Sciences, Rae Bareli Road,Lucknow 226014, India. Tel.: +91 522440004-9 ext: 2133, 2110; fax: +91 522440017; e-mail: anant@sgpgi.ac.in

Accepted for publication 23 October 2000

Cyclosporin A (CsA) withdrawal in patients onimmunosuppressive therapy has been a disputedstep (1, 2). The reasons for discontinuing CsA stemfrom a desire to reduce costs (3) and potentialnephrotoxicity (4). Even though CsA has beenshown to decrease early acute rejection episodes,with a consequent increase in 1-yr graft survival(5), its long-term use has shown no significantadvantage over conventional immunosuppressionwith azathioprine (Aza) and prednisolone (Pred)

(6–8). Earlier studies that attempted CsA with-drawal less than 1 yr after transplantation demon-strated its long-term safety (9–12), whereas othersfailed to do so (13–15). The long-term safety andbenefits of CsA withdrawal in cadaveric renaltransplantation have recently been demonstrated inrandomized trials in which CsA was discontinuedafter 1 yr of stable graft function (16, 17). Othernon-randomized trials that discontinued CsA after1 yr have also shown good results (18). CsA with-

136

Cyclosporin A withdrawal in renal transplantation

drawal should be even safer in live related renaltransplantation as these recipients receive ‘im-munologically advantaged’ grafts (8). However,few such studies are available and these have pro-vided disparate results (19–21). Moreover, thesestudies included small numbers of patients (20, 21)and reported only short-term results (20).

We therefore thought it worthwhile to reportour experience with early and late CsA withdrawalas compared with long-term low dose CsA admin-istration in a large number of patients undergoinglive related transplantation.

Materials and methods

The case records of 325 live related transplanta-tions performed at our Institute between 1989 and1995 were reviewed. The following group of pa-tients were excluded from the study: those inwhom CsA was never given (n=4); those whodied within the first 8 months of transplantation(n=24); those who developed CsA toxicity (n=10); and those with erratic follow-up (n=33)

Immunosuppression induction

For all of our patients, immunosuppression wasinduced with CsA 8 mg/kg, Pred 50 mg/d and Aza2.5 mg/kg/d on the day prior to transplantation.CsA was slowly tapered to 6 mg/kg/d by 1 monthand 3 mg/kg/d by 3 months, and was continued atthat dose until the ninth month. Pred was reducedto 30 mg/d by 1 month and 10 mg/d by 3 months,and Aza was reduced to 1.5 mg/kg/d by the thirdpostoperative day.

CsA withdrawal policy

1989–1992. All patients underwent elective CsAwithdrawal at 9 months after renal transplanta-tion. This was performed rapidly over a 4-wkperiod. Pred was increased to 30 mg/d and Aza to2.5 mg/kg/d 2 wk prior to withdrawal. Pred wassubsequently reduced to 10 mg/d after completionof CsA withdrawal and Aza was maintained at 2.5mg/kg/d.

Because this strategy led to a high incidence ofacute rejection we were compelled to change ourpolicy from 1993 onwards.

1993 onwards. Patients were continued on CsAbeyond the first year of transplantation and thiswas withdrawn only on patient request due tofinancial reasons. When requested, withdrawal wasperformed over an 8-wk period with a biweeklydose reduction of 0.5 mg/kg. Aza and Pred doses

were increased to 30 mg/d and 2.5 mg/kg/d, respec-tively, 2 wk prior to CsA taper and were main-tained at these doses until CsA was completelystopped. Subsequently, the Pred dose was reducedto 10 mg/d. The median time of CsA withdrawalafter renal transplantation was 16 months (range13–22 months). All patients who could afford CsAwere maintained on a ‘low dose’ of 2–4 mg/kg/d.This changing protocol led to the formation ofthree retrospective study groups:

group ECyW: patients with elective CsA with-drawal 9 months after transplantation (n=99);group LCyW: patients with CsA withdrawal af-ter 1 yr of transplantation (n=44); andgroup LDCy: patients maintained on low doseCsA (n=109).

Rejection

Rejection episodes were diagnosed using clinicalcriteria, ultrasound, 99Tc diethylenetriaminepen-taacetic acid (DTPA) nuclear scan and biopsy,which was performed in 61% of cases. Antirejec-tion treatment consisted of intravenous infusion of500 mg methylprednisolone or 80 mg dexam-ethasone for 3–5 d. Steroid-resistant rejections(SRRs) were managed with CsA rescue (CsA-R),which was achieved by adding CsA in a dose of 5mg/kg/d for 4–6 wk or anti-CD3 rnAb/antithrom-bocyte globulin (OKT3/ATG) therapy. OKT3 wasused in a dose of 5 mg/d for 6–10 d. ATG wasused in doses of 1.5–2.0 mg/kg body weight for7–10 d.

Clinical criteria, histology, or both were takeninto consideration to diagnose chronic rejection.The date of allograft loss was considered as thedate of nephrectomy, return to dialysis, or deathwith a functioning graft.

Statistical analysis

Group differences were calculated by analysis ofvariance (ANOVA), Fisher’s exact test, Student’st-test, or the chi-squared test as appropriate. Dif-ferences in graft survival were examined usingKaplan–Meier analysis and assessed for statisticalsignificance using the Wilcoxon log rank test. Bi-nary logistic regression was used to analyse theassociation between two variables.

Results

All patients had at least a single haplotype match.There were no significant differences in the mean

137

Dubey et al.

age, gender, pre-transplantation blood transfu-sions, diabetic status and post-transplantationearly graft dysfunction among the three groups(Table 1). Pre-withdrawal frequency of acute rejec-tions and mean serum creatinine at withdrawalwere not different between ECyW and LCyW(Table 1).

The total number of acute rejection episodes inECyW, LCyW and LDCy were 54/99 (54.54%),14/44 (31.81%) and 26/109 (23.85%), respectively(p=0.001) (Table 2). The incidence of post-CsAwithdrawal rejection was significantly higher inECyW (32/99; 32.3%) than in LCyW (7/44; 15.9%)(p=0.04). In order to evaluate acute rejections inLDCy that occurred during the same time frame,we compared the incidence of late (]9 months)acute rejections in all three groups. ECyW patientshad a significantly higher proportion of late (]9months) acute rejection episodes (Fig. 1). The inci-dence of late acute rejections was highest in ECyW(32/99; 32.3%) as compared with LCyW (8/44;18.2%; p=0.08) and LDCy (8/109; 7.3%; p=0.001). Seven out of eight late acute rejectionepisodes in LCyW occurred after CsA withdrawal(months 13, 14, 15, 16, 16, 18 and 22). Of the 48late acute rejection episodes, 21 were SRR. Theincidence of SRR was significantly higher inECyW (59.4%) than in LCyW (14.3%; p=0.03)and LDCy (12.5%; p=0.04) (Table 3). Of thepatients with SRR, 12 ECyW and 1 LCyW pa-tients were treated with CsA-R. Response to CsA-R was seen in 8 patients (62%: 39% completeresponse with normalization of renal functionwhile 23% had partial response with at least 30%reduction in serum creatinine value). All 5 patientswho failed to respond died due to graft failure.Eight patients with SRR (7 ECyW and 1 LDCy)were treated with either OKT3 or ATG. Of these,5 had complete response, 2 had partial responseand 1 died due to graft failure. Thus, 6/39 (15%)patients with post-CsA withdrawal lost their graftsimmediately. All of these belonged to ECyW. On

long-term follow-up, the incidence of chronic rejec-tion was higher in ECyW (24.24%) than in LCyW(11.36%; p=0.04) and LDCy (17.43%; p=0.11)(Table 2). In ECyW and LCyW patients there wasno association between previous acute rejectionand acute rejection (p=0.7) or chronic rejection(p=0.72) after withdrawal. However, the inci-dence of chronic rejection was significantly higherin patients who presented acute rejection episodesafter withdrawal (p=0.001). The 6-yr graft sur-vival in ECyW, LCyW and LDCy was 82.83, 88.64and 91.74%, respectively (p=0.7) (Fig. 2). Themajor cause of mortality was graft failure (uncon-trolled acute rejection or chronic rejection) (n=24), followed by infections (n=5) and myocardialinfarction (n=2). Graft survival was similar topatient survival as none of the patients who hadgraft loss due to immunological causes could af-ford a second transplantation or find a suitabledonor for re-transplantation. The mean serum cre-atinine at final follow-up was significantly lower inpatients who underwent CsA withdrawal after 1 yr(Table 2). The mean number of antihypertensivemedications required at 3-yr post-transplantationwas significantly higher in LDCy (1.9590.81)than in ECyW (1.4391.03; p=0.001) and LCyW(1.4890.99; p=0.007).

Discussion

CsA has improved first year graft survival ratesafter kidney transplantation (5). However, its con-tinuation beyond the first year has not proven tobe of further benefit as the annual rate of graft lossthereafter has not changed significantly (6). On thecontrary, the prolonged administration of CsA hashad deleterious effects on graft function (4). Thishas urged clinicians to investigate the outcome ofCsA withdrawal in renal transplantation. Whilenumerous studies have dealt with CsA withdrawalin cadaveric transplantation, only a few concerninglive related transplantation have been conductedand have reported only short-term results (19, 20).

Table 1. Pre-transplantation recipient characteristics

pGroupRecipient characteristics

ECyW LDCyLCyW(n=99) (n =44) (n=109)

Age (mean9SD) 34.4910.3 33.398.7 36.2911.3 NSa

Sex (% male) 74.8 77 78 NSb

NSa4.191.113.892.3Pre-transplantation blood transfusions (mean9SD) 4.391.64Diabetes (%) 21.2 31.8 26.6 NSb

Acute tubular necrosis (n) 6 4 4 NSa

Serum creatinine at withdrawal (mg%; mean9SD) 1.3390.4 1.3990.3 – NSc

SD: standard deviation; NS: not significant (a ANOVA; b chi-square; c Student’s t-test).

138

Cyclosporin A withdrawal in renal transplantation

Table 2. Outcome in different groups

Group

ECyW(I) LCyW(II) LDCy(III)(n=99) (n=44) (n=109)

54 (55)Acute rejections n (%)a 14 (31.81) 26 (23.7)24 (24) 5 (11)Chronic rejection n (%)b 19 (17)

Serum creatinine (mg%)c (mean9SD) 1.4990.5 1.2490.4 1.4890.667.5913Time to death/final follow-up (months) (mean9SD) 55.9913 56.9914

a p (Fisher’s exact test): 0.001 (I vs II), 0.001 (I vs III), 0.15 (II vs III).b p (Fisher’s exact test): 0.04 (I vs II), 0.11 (I vs III), 0.17 (II vs III).c p (Student’s t-test): 0.04 (I vs II), 0.9 (I vs III), 0.01 (II vs III).

In our study, patients undergoing CsA with-drawal (ECyW and LCyW) had a higher incidenceof acute rejection than did those who continuedCsA (LDCy). The overall incidence of post-CsAwithdrawal acute rejection was 27.3%. The early(ECyW) CsA withdrawal group had a greater riskof acute rejections than those with late (LCyW)withdrawal (32.3 vs 16%). In a meta-analysis of 22studies in which CsA was withdrawn between 3and 12 months post-renal transplantation, Helder-man et al. (22) reported a 33.2% risk of acuterejection. Other investigators have also demon-strated that time under CsA therapy is inverselyproportional to the number of acute rejectionepisodes during the first year after CsA withdrawal(21, 23). Studies in which CsA was withdrawn lessthan 1-yr post-transplantation have reportedhigher acute rejection rates (11, 15) compared withthose withdrawing CsA after 1 yr (16, 17). Thereason for this is unclear, but it could be related tothe development of some kind of immune toler-ance after 1 yr of stable graft function. The higherrate of acute rejections in ECyW might also havebeen influenced by the duration over which CsAwas withdrawn as these patients had rapid taper (4wk) compared with LCyW (8 wk). The rapid taper-ing of CsA can lead to higher acute rejection ratesas compared with more controlled withdrawal (18).Of interest was the finding of late rejections inECyW and LCyW (Fig. 1). Post-CsA withdrawalacute rejection episodes were more severe in ECyWas seen by the high proportion of SRRs (Table 3).This led to increased admissions, cost and compli-cations due to high dose steroid treatment. Thir-teen out of 32 (40.6%) patients, who developedacute rejection post-withdrawal (ECyW), lost theirgrafts (6 due to uncontrolled acute rejection and 7due to chronic rejection on long-term follow-up).Chronic rejection was highest in ECyW (24%) ascompared with LCyW (11%; p=0.04) and LDCy(17%; p=0.11). In another study (11) of CsAwithdrawal in live related renal transplantation, 9

out of 34 (26.5%) patients developed acute rejec-tions within 3 months of CsA withdrawal. Of these9 patients, 5 (55%) developed chronic rejection.Others have also demonstrated a high rate ofchronic rejection and graft loss after acute rejec-tions after early CsA withdrawal (14, 23). Al-though most of these rejections were rescued, graftlosses were clearly described. Furthermore, it hasbeen documented that late acute rejections (7–12months post-transplantation) are more severe anddetrimental to long-term allograft survival as com-pared with early (0–6 months) rejections (24, 25).Thus, any factor contributing to a high incidenceof acute rejections during this period should beavoided. This would be even be more pertinent ina country where the socio-economic scenario issuch that the graft is equivalent to the patient interms of survival (similar graft and patient survivalrates). Given these results and the fact that thenumber of acute rejection episodes in the firsttransplantation year is a major antigen-dependentrisk factor related to long-term graft loss (26), itwould seem logical to extend the duration of CsAtreatment beyond 1 yr. As CsA withdrawal after 1

Fig. 1. Proportion of early (B9 months) and late (]9months) acute rejection episodes in each group.

139

Dubey et al.

Table 3. Outcome of late (]9 months) acute rejections in the three groups

Group p-value (chi-square)Outcome

ECyW(I) LCyW(II) LDCy(III)(n=32) (n=7) (n=8)n (%) n (%) n (%)

19 (59.4) 1 (14.3) 1 (12.5)Steroid resistant I vs II (0.03), I vs III (0.02)17 (53.1) 1 (14.3) 2 (25) I vs II (0.06) I vs III (0.15)Chronic rejection13 (40.6)a 1 (14.3)b 1 (12.5)b I vs II (0.19), I vs III (0.13)Death

a Six uncontrolled acute rejection, 7 chronic rejection.b Chronic rejection.

yr in LCyW occurred at variable times after trans-plantation (median 16 months), we cannot com-ment on the optimum time of CsA withdrawal.However, there is convincing evidence of the safetyof CsA withdrawal after completing 1 yr of stablegraft function (12, 16). The actuarial 6-yr graftsurvival in patients with early and late withdrawaland those who continued on CsA was 82.83, 88.64and 91.74%, respectively (p=0.7). Although thisdid not attain statistical significance at a medianfollow-up of 66 months, it could become signifi-cant on even longer follow-up. This seems likely asECyW had the highest incidence of chronic rejec-tion, which has been the main cause of graft loss inour study. None of our patients with immunologi-cal causes (uncontrolled acute rejection/chronic re-jection) of graft loss could afford maintenancedialysis or re-transplantation, leading to similarpatient and graft survival rates. This is in con-tradistinction to most Western figures, where pa-tient survival figures are higher than graft survival(5).

Graft function, as measured by serum creatininelevels, was best preserved in LCyW (1.2490.4mg%) as compared with LDCy (1.4890.6 mg%)and ECyW (1.4990.5 mg%). The higher serumcreatinine levels in ECyW were probably due tothe higher incidence of chronic rejection in thisgroup. Two randomized trials have documentedinferior graft function (as measured by serum crea-tinine levels and glomerular filtration rates) in pa-tients maintained on chronic low dose CsA therapyas compared with patients undergoing CsA with-drawal at 1 yr. The antihypertensive requirementin our study was maximal in patients continued onlow dose CsA. This corroborates with the resultsof others (16, 17), who found increasing antihyper-tensive medication requirement in patients main-tained on CsA doses between 2 and 4 mg/kg/d ascompared with those who were withdrawn fromCsA. This clearly demonstrates that long-term lowdose CsA has no beneficial effects in terms of graftfunction and provides inferior blood pressurecontrol.

Conclusions

We conclude that early CsA withdrawal in liverelated renal transplantation involves a consider-able risk of acute rejection with subsequent chronicrejection. Late withdrawal (\1 yr) performedgradually is safer and provides better long-termblood pressure control compared with long-termlow dose CsA. Patients wishing to withdraw fromCsA should be encouraged to discontinue the drugonly after the first year of transplantation. CsA canbe withdrawn after 1 yr of stable graft function asit is safe and economical.

References

1. NEWSTEAD CG, JOHNSTON PA, WILL EJ, DAVISON EA.The case for withdrawal of cyclosporin after renal trans-plantation. Nephrol Dial Transplant 1998: 13: 28.

2. HELDERMAN HJ, GORAL S. Cyclosporin withdrawal: to beor not to be!. Nephrol Dial Transplant 1998: 13: 31.

Fig. 2. Actuarial graft survival of patients continued on lowdose CsA, and early and late CsA withdrawal. Differenceswere not statistically different by the Wilcoxon log rank test(p=0.7).

140

Cyclosporin A withdrawal in renal transplantation

3. SANDERS CE, CURTIS JJ, JULIAN BA et al. Tapering ordiscontinuing cyclosporin for financial reasons – a singlecenter experience. Am J Kid Dis 1993: 21 (1): .

4. BENNET WM, DEMATTOS A, MEYER MM, ANDOH T,BARRY JM. Chronic cyclosporin nephropathy: the Achillesheel of immunosuppressive therapy. Kidney Int 1996: 50:1089.

5. CECKA M, TERASAKI P. Clinical Transplants 1995. LosAngeles, CA: UCLA Tissue Typing Laboratory, 1996.

6. NAIMARK DMJ, COLE E. Transplant Rev 1994: 8: 93.7. FISCHEL RJ, PAYNE WD, GILLINGHAM KJ et al. Long-

term outlook for renal transplant recipients with 1-yearfunction: ‘Doctor what are my chances?’. Transplantation1991: 51: 118.

8. GHONEIM MA, SOBH MA, SHOKEIR AA, BAKR MA,EL-SHERIEF AK, FOUDA MA. Cyclosporin in primaryhaploidentical live-donor kidney transplantation: is itworthwhile. Nephrol Dial Transplant 1993: 8: 551.

9. ADU D, MICHAEL J, MCMASTER P. Conversion fromcyclosporin to azathioprine prednisolone. Lancet 1987: 1:392.

10. HOITSMA AJ, WETZELS JFM, VAN LIER HJJ, BERDEN

HJM, KOENE RAP. Cyclosporin treatment with conver-sion after 3 months versus conventional immunosuppres-sion in renal allograft recipients. Lancet 1989: 1: 584.

11. VAN WEZEL HBM, HILBRANDS LB, KOENE RAP,HOITSMA AJ. Conversion from cyclosporin to azathio-prine at 3 months after renal transplantation: long termresults. Transplant Proc 1994: 26 (4): 2550.

12. HOLLANDER AAMJ, VAN SAASE JLCM, KOOTTE AMMet al. Beneficial effects of conversion from cyclosporin toazathioprine after kidney transplantation. Lancet 1995:345: 610.

13. MORRIS PJ, FRENCH ME, DUNNILL MS et al. A con-trolled trial of cyclosporin conversion in renal transplanta-tion with conversion to azathioprine and prednisone after3 months. Transplantation 1983: 36: 273.

14. HIESSE C, NEYRAT N, DEGLISE FAVRE A et al. Ran-domised prospective trial of elective cyclosporin with-drawal from triple therapy at 6 months after cadavericrenal transplantation. Transplant Proc 1991: 23 (1): 987.

15. DELMONICO FL, CONTI D, AUCHINCLOSS H et al. Long-term, low dose cyclosporin treatment of renal allograftrecipients. Transplantation 1990: 49 (5): 899.

16. MACPHEE IAM, BRADLEY JA, BRIGGS JD et al. Long-

term outcome of a prospective randomised trial of conver-sion from cyclosporin to azathioprine treatment one yearafter renal transplantation. Transplantation 1998: 66 (9):1186.

17. PEDERSEN PB, HANSEN HE, KORNERUP HJ, MADSEN S,SORENSEN AWS. Long term graft survival after conver-sion from cyclosporin to azathioprine 1 year after renaltransplantation. A prospective, randomised study from 1to 6 years after transplantation. Nephrol Dial Transplant1993: 8: 250.

18. HEIM DUTHOY KL, CHITWOOD KK, TORTORICE KL,MASSY ZA, KASISKE BL. Elective cyclosporin withdrawal1 year after renal transplantation. Am J Kid Dis 1994: 24(5): 845.

19. KAHN D, OVNAT A, PONTIN AR, SWANEPOEL CR, CAS-

SIDY MJD. Long term results with elective cyclosporinwithdrawal at three months after renal transplantation-ap-propriate for living-related transplants. Transplantation1994: 58: 1410.

20. AGARWAL SK, DASH SC, TIWARI SC et al. Results ofconversion from triple drug to double drug therapy inliving related renal transplantation. Transplantation 1995:59 (1): 27.

21. CHEW-WONG A, ALBERU J, ABASTA-JIMINEZ E, AL-

VAREZ-SANDOVAL E, GABILONDO-NAVARO F, CORREA-ROTTER R. Withdrawal versus continous cyclosporin inkidney transplant recipients of one haplotype matcheddonors. Transplant Proc 1999: 31: 1106.

22. HELDERMAN JH, VAN BUREN DH, AMEND JC JR, PIRSCH

JD. Chronic immunosuppression of the renal transplantrecipient. J Am Soc Nephrol 1994: 4 (1): S2–S9.

23. KEITEL E, MICHELON T, DOMINIGUEZ V et al. Long-termevaluation of two protocols of elective cyclosporin with-drawal in renal transplant recipients. Transplant Proc1999: 31: 3013.

24. LEGGAT JE JR, OJO AO, LEICHTMAN AB et al. Long-termrenal allograft survival: prognostic implication of the tim-ing of acute rejection episodes. Transplantation 1997: 63(9): 1268.

25. HUMAR A, KERR S, GILLINGHAM AJ, MATAS AJ. Fea-tures of acute rejection that increase risk of chronic rejec-tion. Transplantation 1999: 68 (8): 1200.

26. MATAS AJ, GILLINGHAM KJ, PAYNE WD et al. Theimpact of an acute rejection episode on long term allograftsurvival (t1/2). Transplantation 1994: 57 (6): 837.

141

.