© 2013 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd
Clin Transplant 2014: 28: 67–79 DOI: 10.1111/ctr.12280
Clinical Transplantation
Proteinuria and sirolimus after renal transplantation: a retrospective analysis from a large German multicenter database Naik MG, Heller KM, Arns W, Budde K, Diekmann F, Eitner F, Fischereder M, Goßmann J, Heyne N, Morath C, Riester U, Gwinner W and J€ urgensen JS for the German Sirolimus Study Group. Proteinuria and sirolimus after renal transplantation: a retrospective analysis from a large German multicenter database. Abstract: The German Sirolimus Study Group has established a database among 10 transplant centers throughout Germany to study the outcomes in 726 renal transplant patients being converted to a sirolimuscontaining therapy between 2000 and 2008 with a total of more than 1500 recorded patient years on therapy. In this study, we present a detailed description of the cohort, of characteristic changes over the observation period, proteinuria and graft survival, and new-onset proteinuria after conversion. Over the study period, age, graft function at the time of conversion, and the proportion of patients switched to sirolimus because of malignancy increased, whereas the proportion of patients with significant proteinuria at conversion decreased. Already modest proteinuria (151–268 mg/L) at conversion and new-onset proteinuria (>500 mg/L) after conversion were associated with inferior graft survival. Even mild proteinuria (>71 mg/L) at conversion was associated with new-onset proteinuria (>500 mg/L) post-conversion. Serum creatinine and urinary protein excretion at conversion together with age at transplantation had a significant impact on patient and graft survival. This large data set confirms and extends previous observations that proteinuria is an important indicator for graft outcome after conversion to sirolimus. We conclude that patients without any proteinuria have the greatest benefit from conversion to sirolimus.
Marcel G. Naika,*, Katharina M. Hellerb,*, Wolfgang Arnsc, Klemens Buddea, Fritz Diekmannd, Frank Eitnere, Michael Fischerederf, Jan Goßmanng, Nils Heyneh, Christian Morathi, Udo Riesterj, Wilfried Gwinnerk,* and € rgensenl for the Jan Steffen Ju German Sirolimus Study Group a
Department of Nephrology, Charité University – Mitte, Berlin, bDivision of Nephrology, Department of Medicine, University of Erlangen, Erlangen, cTransplant, Centre Cologne, Cologne General Hospital, Cologne Germany, dHospital Clinic, Barcelona Spain, e Division of Nephrology and Clinical Immunology, RWTH Aachen University, Aachen, fDepartment of Medicine IV, University Hospital LMU, Munich, gTransplant Care Centre Frankfurt, Frankfurt, hSection of Nephrology and Hypertension, Department of €bingen University Internal Medicine IV, Tu €bingen, iDepartment of Hospital, Tu Nephrology, University of Heidelberg, Heidelberg, jPfizer Pharma GmbH, Berlin, k Department of Nephrology, Hannover Medical School, Hannover and lDepartment of Nephrology, Charite´ University – Virchow, Berlin, Germany Key words: conversion – immunosuppression – mammalian target of rapamycin inhibitor – proteinuria – renal transplantation – sirolimus Corresponding author: Dr. Jan Steffen €rgensen, MPH, Department for Nephrology Ju and Medical Intensive Care, Universita¨tsmedizin Berlin, Campus Virchow Klinikum, Augustenburger Platz 1, Charite 13353 Berlin, Germany. Tel.: +49 30 450 570 606; fax: +49 30 450 7570606; e-mail: [email protected] *Both respective authors contributed equally. Conflict of interest: None. Accepted for publication 21 October 2013
67
Naik et al.
Patient and graft survival early after kidney transplantation increased dramatically in recent decades (1, 2). This successful development is largely based upon the introduction of new immunosuppressive agents with more distinct mechanisms of action that allow tailored regimens for individual patients (3). In contrast, long-term survival was not improved by all these strategies (1, 2). After its approval for use in organ transplantation in 1999, the mammalian target of rapamycin inhibitor (mTORi) sirolimus extended the therapeutic options. Early studies suggested that sirolimus was less nephrotoxic than calcineurin inhibitors (CNIs) and showed antifibrotic and anticancer properties (4). Sirolimus found its way into different immunosuppressive regimens both for de novo and maintenance immunosuppression (5–11). Early publications indicated favorable effects of sirolimus on long-term renal allograft function compared with CNI-based immunosuppressive regimens (9, 10, 12, 13). However, specific adverse effects limit the use of sirolimus in certain patients and require watchful use (14–16). Over the last decade, many randomized controlled trials (RCTs) have provided valuable information regarding the use and contraindications of sirolimus in distinguished patient groups that have been included in these studies and allowed a refined application of the drug (17). These robust RCTs aimed to exclude any systematic bias and supported clinical decision making with regard to the use of sirolimus. RCTs are designed as experiments with high internal validity, for example, the validity of the interferences drawn as they apply to the members of the source population. Clinical trials employ sophisticated methods to control for systematic errors by means of blinding, randomization, allocation concealment, and clear-cut inclusion and exclusion criteria. Usually, this leads to statistically sound and credible results. On the other hand, exactly those above-mentioned factors that determine the high internal validity of a clinical study often hamper its external validity. In other words, the validity of the interferences as they pertain to people outside the tightly defined source population or study cohort may be very limited. This lack of generalization with regard to patients with different characteristics, for example, elderly or those with specific comorbidities or comedication, is a major concern. On the other hand, observational studies and analysis based upon data gathered in clinical routine settings may be disposed to systematic errors, but offer additional valuable evidence and conclusions, which might be applicable for the full spectrum of patients under treatment (18, 19).
68
The German Sirolimus Study Group has established a database among 10 transplant centers in Germany to study sirolimus use under realistic conditions in more than 700 renal transplant patients who were switched to a sirolimus-containing regimen. This study provides a first analysis of the database and demographics of the population, with a focus of characteristic changes in key demographic variables over the eight-yr observation period in the conversion setting. Furthermore, in this study, we specifically wanted to thoroughly analyze proteinuria including new-onset proteinuria and clinical outcome in a large cohort of patients, who were converted to sirolimus.
Methods Patients
This multicenter, retrospective study comprises patients with a renal transplant or combined kidney transplantation with another solid organ who were converted to a sirolimus-containing immunosuppressive regimen three or more months after transplantation between January 1, 2000, and December 31, 2008. All patients were switched to sirolimus. Ten German transplant centers included their data to obtain a comprehensive database for the study of sirolimus in clinical practice. We aimed to include all converted patients in each of the participating centers regardless of success, length of sirolimus therapy, age, type of donation, number of previous transplants, and reason for conversion to sirolimus. There were no patients excluded from this database to reduce a selection bias and address the full spectrum of all therapeutic approaches. The study was approved by the local ethics committees, and all patients provided informed consent for the scientific use of their data. Data collection
Data collection included four distinct time points before initiation of sirolimus. The start of sirolimus therapy marked the study baseline to establish baseline values for allograft function, proteinuria, comorbidities, and medical therapies (immunosuppressive drugs and concomitant treatments). Data collection in the first year was intended at three, six, and 12 months after initiation of sirolimus and semiannually thereafter. Due to the retrospective nature of the study, actual time points for data collection in individual patients differed from this schedule. This was appreciated by building defined time periods and
German Sirolimus Study Group calculating the days from the study baseline for each patient. The time periods were calculated in days from baseline, with the pre-conversion visits 1Y (365 d before conversion), M6 (180 d before conversion), and M1 (30 d before conversion. Post-conversion visits were defined as M3 (days 31–135), M6 (days 136–270), Y1 (days 271– 547), Y2 (548–912), Y3 (913–1277), Y4 (1278– 1642), Y5 (1643–2007), and >Y5 (>2008). End points were defined as patient’s death and terminal allograft failure. Renal function was estimated using the abbreviated MDRD formula. Urinary protein excretion was recorded as a protein concentration in spot urine. Because each center used different units, all urinary protein concentrations were converted to the most frequently used unit (mg/L) assuming an urinary output of 2 L/d. A protein concentration of >500 mg/L was considered positive. Dipsticks were considered if no concentration was recorded. Positive dipsticks were imputed as the average protein concentration of patients with >500 mg/L proteinuria (750 mg/L), and negative dipsticks were imputed as 122 mg/L (the average protein concentration of patients 10% was detected before transplantation, and highly sensitized patients (PRA ≥ 85%) were present in only 0.5%. The mean creatinine at conversion was 2.16 0.97 mg/dL with an estimated glomerular filtration rate (eGFR) of 39.2 19.4 mL/min at
69
Naik et al. Table 1. Characteristics of the study population (N = 726)
Table 1. Continued
% General demographics Male Caucasian Living donation First renal transplantation Combined transplantation Thereof kidney pancreas BPAR before conversion BPAR after conversion RAAS before conversion RAAS after conversion DGF PRA > 10% Mismatches (MM broad) (mean SD) Age at TX (yr mean SD) Time on dialysis before TX (yr mean SD) Age at conversion (yr mean SD) Time from TX to conversion (yr mean SD) Time of follow-up (yr mean SD) eGFR at conversion (mL/min mean SD) Proteinuria at conversion (mg/L mean SD) First sirolimus dose (mg mean SD) Maintenance (sirolimus dose) (mg mean SD) Underlying renal disease Diabetes mellitus Hypertension Polycystic renal disease Glomerulonephritis Tubulointerstitial disease Other genetic disease Unknown Others Comorbidities Diabetes mellitus Arterial hypertension Neurologic disease Chronic liver disease Cardiovascular disease Coronary heart disease Peripheral arterial occlusive disease Cerebrovascular disease Immunosuppression Initial immunosuppression Induction with antibody Non-depleting Depleting Steroids CNI* Cyclosporine Tacrolimus Antiproliferative immunosuppressant* Mycophenolic acid Azathioprine At month 3 Sirolimus Steroids CNI* Cyclosporine Tacrolimus Antiproliferative immunosuppressant*
70
% 63.3 99.0 16.4 74.5 10.5 86.7 38.1 9.0 60.9 72.7 17.4 7.7 2.4 43.3 3.9 49.8 6.6
2.3 39.2 349 6.4 2.9 12.3 3.6 11.3 43.0 14.2 3.6 7.9 4.3 23.6 84.7 13.3 9.6 23.6 18.4 7.2 2.6
19.7 13.8 98.2 63.0 27.2 54.3 32.2 94.1 87.7 7.6 22.1
Mycophenolic acid Azathioprine Reason for conversion Study Tumor Creeping creatinine Side effect of other immunosuppressant Acute rejection Others Chronic allograft nephropathy CNI toxicity
1.64 13.57 3.13 13.35 6.11
1.82 19.4 754.56 4.78 1.64
39.2 4.5 11.1 23.0 22.6 10.1 12.0 24.9 17.7 26.5
TX, transplantation; BPAR, biopsy-proven acute rejection; CNI, calcineurin inhibitor; eGFR, estimated glomerular filtration rate; PRA, panelreactive antibodies. *Value may exceed 100% due to concomitant or overlapping use in some patients.
conversion. Mean proteinuria at conversion was 349 755 mg/L with a wide range (0–5480 mg/ L), with a median of 107.5 mg/L. There were 17% of patients with proteinuria exceeding 800 mg/d. The most frequent reasons for conversion were “CNI toxicity” (28.7%), “tumors” (25.4%), “creeping creatinine” (22.7%), and “chronic allograft nephropathy” (18.7%). It is important to note that multiple reasons could be ticked for individual patients. There were 80 patients (11.1%) who were included in various single and multicenter studies (e.g., skin cancer and creeping creatinine). The majority of patients converted due to tumors had skin tumors. Further analysis concerning tumors will be required. The immunosuppressive regimen before conversion to sirolimus consisted of corticosteroids (98.2%), CNIs such as cyclosporine or tacrolimus (90.2%), and antiproliferative immunosuppressants such as mycophenolic acid or azathioprine (86.5%). Prior antibody therapy was performed in 33.5% (13.8% depleting and 19.7% non-depleting antibodies). At conversion, the initial dose of sirolimus was 6.4 4.8 mg/d and the maintenance dose at the first visit post-conversion was 2.9 1.6 mg/d. The sirolimus trough level was 8.0 3.6 ng/mL and 8.2 1.2 ng/mL at six and 12 months, respectively. Three months after conversion, 70.3% were CNI free, 29.7% on low-dose CNI combination therapy (7.6% cyclosporine and 22.1% tacrolimus), 43.7% on antiproliferative medication (39.2% MPA and 4.5% azathioprine), and 87.7% on steroids. The vast majority of patients (72.9%) were treated with ACE inhibitors or ARB after conversion. In 65 patients (9.0%), BPAR was documented after conversion.
German Sirolimus Study Group dominated by conversion due to deteriorating renal function (e.g., “chronic allograft nephropathy” 50% in 2000 vs. 9.7% in 2008, p < 0.001), the potential antitumor effects of sirolimus accounted for a larger proportion of conversions at later time points (0% in 2000 vs. 25.8% in 2008, p < 0.001). As a consequence, a steady and statistically significant (p < 0.001) increase in the mean eGFR at conversion was observed from 2000 to 2008 (27.1 17.2 vs. 50.5 35.8 mL/min, Fig. 1C). Along with the increase in eGFR, the percentage of patients being converted with an eGFR 500 mg/L at conversion over the study period.
71
Naik et al.
in 2000 vs. 46.7% in 2008, p < 0.001). The use of CNIs showed a trend toward an increasing use of tacrolimus (37% in 2000 vs. 46.7% in 2008, p = 0.067) and rather stable use of cyclosporine (59.3% in 2000 vs. 53.3% in 2008, p = 0.79). Azathioprine (40.7% in 2000 vs. 20% in 2008, p < 0.001) was largely replaced by mycophenolic acid (40.7% in 2000 vs. 63.3% in 2008, p < 0.001) toward the final years of the study. The initial and maintenance dose of sirolimus fell from 12.9 4.1 mg/d (range 3–15 mg/d) and 4.7 1.8 mg/d in 2000 to 2.7 1.2 mg/d (range 1–5 mg/d) and 1.9 0.7 mg/d in 2008 (p < 0.001 each), respectively. The sirolimus trough levels at six months decreased significantly from 10.6 5.0 ng/mL in 2000 to 5.8 2.7 ng/mL in 2008 (p < 0.001). There were no significant changes during the study period regarding sex ratio, type of donation, number of transplant, combined transplantation, HLA mismatches, BPAR, delayed graft function, panel-reactive antibody level >10%, time from transplantation to conversion, time on dialysis, use of ACE/ARB agents before and after conversion, or underlying renal disease.
A
Conversion 417 87 18 6
Year 1 288 41 4 2
Year 2 207 25 0 1
Year 3 134 16 0 0
Year 4 77 11 0 0
Year 5 38 4 0 0
3500mg/l
Year 5 41 4 0 0
3500mg/l
Year 5 35 3 0 0
3500mg/l
Numbers at risk
B
Proteinuria at conversion and outcomes
Overall patient survival was 79.9%, and graft survival rates were 67.2% and 59.4% excluding and including death after five yr, respectively. There were 528 patients with data on proteinuria at the time of conversion (311 patients with urinary protein concentration and 217 dipsticks). In 110 patients (20.8%), proteinuria >500 mg/L was detected at conversion (65 patients with urinary protein concentration and 45 dipsticks). First we looked whether proteinuria >500 mg/L at conversion had a negative impact on patient, graft, and overall survival. The five-yr patient survival was 85.8% vs. 59.9% (p < 0.001). The five-yr death-censored graft survival was 76.6% vs. 31.7% (p < 0.001), and five-yr overall survival (including death) was 68.5% vs. 21.8% (p < 0.001). Next, we examined whether the extent of proteinuria at conversion is associated with patient, graft, and overall survival (Fig. 2A–C). In the group of proteinuria 3500 mg/L (N = 6), the five-yr patient survival was 85.8%, 62.6%, 41.7%, and 80%, respectively (p < 0.001), five-yr death-censored graft survival was 76.9%, 36.8%, 0%, and 50%, respectively (p < 0.001), and the five-yr overall survival (including death) was 68.7%, 24.7%, 10.3%, and 41.7%, respectively (p < 0.001).
72
Conversion 417 87 18 6
Year 1 293 42 5 2
Year 2 210 26 2 1
Year 3 136 15 0 0
Year 4 80 10 0 0
Numbers at risk
C
Conversion Year 1 417 295 87 46 18 5 6 2 Numbers at risk
Year 2 214 27 1 1
Year 3 137 15 1 0
Year 4 76 9 0 0
Fig. 2. (A) Patient survival and extent of proteinuria at conversion (p < 0.001). (B) Graft survival excluding death and extent of proteinuria at conversion (p < 0.001). (C) Graft survival including death and extent of proteinuria at conversion (p < 0.001).
German Sirolimus Study Group To systematically assess the relationship between proteinuria and survival, we performed a ROC curve analysis. The first ROC curve analysis for patient survival using quantitative value only showed an area under the curve (AUC) of 0.557, p = 0.39, N = 311. The second ROC curve analysis for patient survival using imputed values as well showed an AUC of 0.577, p = 0.13, N = 528. There was no significant impact on patient survival. We found a significant relationship between proteinuria at conversion and death-censored graft survival with an AUC of 0.728 (p < 0.001, N = 311, Fig. 3) and a cutoff value of 268 mg/L urinary protein (sensitivity 61.4%, specificity 83.9%, positive predictive value [PPV] 90.6%, and negative predictive value [NPV] 46.1%). Considering imputed values as well, the AUC was 0.71 (p < 0.001, N = 524) and the cutoff value was 151 mg/L (sensitivity 66.7%, specificity 72.9%, PPV 91.1%, NPV 34.4%, N = 528). As shown in Fig. 4A, already modest proteinuria >268 mg/L had a strong impact on graft failure. The five-yr death-censored graft survival was 79.7% in patients below and 34.8% in patients above the cutoff value of 268 mg/L. Using imputed values for the cutoff value of 151 mg/L, five-yr death-censored graft survival was 81.1% (151 mg/L), respectively (p < 0.001, Fig. 4B).
A
Conversion Year 1 235 162 76 31 Numbers at risk
Year 2 119 19
Year 3 77 10
Year 4 38 5
Year 5 18 3
268mg/l
B
Conversion 348 180
Year 1 248 93
Year 2 180 58
Year 3 116 36
Year 4 70 20
Year 5 34 11
151mg/l
Numbers at risk
Fig. 4. (A) Proteinuria cutoff value at conversion and graft failure. p < 0.001, N = 311. (B) Proteinuria cutoff value at conversion and graft failure. p < 0.001, N = 528.
Fig. 3. ROC curve of all patients with urinary protein concentration and death-censored graft failure after conversion. area under the curve 0.728. p < 0.0005, N = 311.
When performing a ROC curve analysis for overall survival (including death), we found a cutoff of 151 mg/L (p < 0.001, AUC 0.699, sensitivity 67.1%, specificity 72.3%, PPV 87.8%, NPV 42.6%, N = 311). The overall five-yr overall survival (including death) was 72.3% vs. 33.6% (p < 0.001, Fig. 5A). The same cutoff was calculated in the second ROC curve analysis using imputed values (AUC 0.692, p < 0.001, sensitivity 61.8%, specificity 74.3%, PPV 86.5%, NPV 42.2%, N = 528). The overall five-yr overall survival (including death) was 72.7% vs. 33.7% (p < 0.001, Fig. 5B) for this imputed cutoff. Lastly, we performed a Cox proportional hazard analysis on overall survival (including death) (Table 2). There was a 1.556-fold (CI: 1.252–1.934, p < 0.001) higher risk of graft loss or death per 1000 mg/L urinary protein concentration at conversion. Similarly, creatinine at conversion was
73
Naik et al. Table 2. Cox regression analysis for overall graft survival (including death)
A
95% CI Exp (B) Overall survival
Significance
Exp (B)
Lower
Upper
Creatinine at conversion (mg/dL) Proteinuria at conversion (g/L) Age at Tx (yr) Sex Time from Tx to conversion (yr) BPAR before conversion Time on dialysis (month) Glomerulonephritis
500 mg/L) was detected after conversion (Fig. 6), resulting in an estimated risk of newonset proteinuria of about 37.6% after five yr. Patients with tumors (N = 104) had similar rates of new-onset proteinuria compared with patients without (36.5% vs. 37.8%, p = 0.819). Patients having undergone a combined kidney/ pancreas transplantation (N = 38) showed a trend toward a less frequent incidence of newonset proteinuria compared with patients with kidney-only transplantation (23.7% vs. 39.7%, p = 0.060), although this analysis is limited by low numbers. In most cases, new-onset proteinuria (>500 mg/ L) started within the first six months after conversion (N = 57, 19.2%). Patients with a new-onset proteinuria had already significantly higher urinary protein concentrations at the time of conversion (mean 78 vs. 155 mg/L, p < 0.0005). The
74
Conversion Year 1 348 253 180 95 Numbers at risk
Fig. 5. (A) Proteinuria cutoff value (without imputation) at conversion and overall survival (including death), p < 0.001, N = 311. (B) Proteinuria cutoff value (with imputation) at conversion and overall survival (including death), p < 0.001, N = 528.
serum creatinine (2.33 1.14 vs. 1.98 1.56 mg/dL, p = 0.002) and consequently the eGFR (37.7 18.7 vs. 46.2 23.5 mL/min, p = 0.004) differed significantly between patients with and without new-onset proteinuria at the end of observation. To find a cutoff value for the development of new-onset proteinuria, we performed a ROC curve analysis with an AUC of 0.749 (Fig. 7, p < 0.001, sensitivity 76.4%, specificity 65.8%, PPV 81.1%, NPV 57.4%, N = 189). Already mild proteinuria above 71 mg/L showed a significant impact on the
German Sirolimus Study Group A
Conversion 332
Year 1 179
Year 2 119
Year 3 76
Year 4 49
Year 5 25
Numbers at risk
Fig. 6. Time to new-onset proteinuria (>500 mg/L) after conversion (N = 90/332 = 27.1%).
Conversion Year 1 89 61 77 31 Numbers at risk
Year 2 41 18
Year 3 27 12
Year 4 17 5
Year 5 8 3
71mg/l
Year 2 18 61
Year 3 12 38
Year 4 5 23
Year 5 3 13
71mg/l
B
Conversion Year 1 89 31 187 96 Numbers at risk
Fig. 8. (A) Cutoff value (without imputation) and development of new-onset proteinuria, N = 166, p < 0.0005. (B) Cutoff value (with imputation) and development of new-onset proteinuria, N = 276, p < 0.0005. Fig. 7. ROC curve of patients with new-onset proteinuria after conversion. area under the curve 0.749. p < 0.001, N = 189.
development of new-onset proteinuria (Fig. 8A, p < 0.0005, N = 166), being 74.6% vs. 32.5% after five yr, respectively. Using imputed values, the cutoff was the same with an AUC of 0.658 (Fig. 7B, p < 0.001, sensitivity 61.7%, specificity 38.3%, PPV 67.6%, NPV 63.4%, N = 316). As shown in Fig. 8B, development of new-onset proteinuria after five yr was 74.6 vs. 54.7% (N = 276, p < 0.0005).
New-onset proteinuria was also associated with a significantly lower graft survival (Fig. 9). After five yr, death-censored graft survival was 72.2% in patients with new-onset proteinuria and 84.1% in patients without new-onset proteinuria (p = 0.007). New-onset proteinuria had a strong tendency to deteriorate overall survival (including death), being 78.0% vs. 61.6% (p = 0.052) after five yr. Again, patient survival was not influenced significantly by new-onset proteinuria (85.2% vs. 86.5%).
75
Naik et al.
We did not observe a correlation between high sirolimus trough levels and the development of new-onset proteinuria at three, six, or 12 months after conversion. In a Cox proportional regression analysis for risk factors of new-onset proteinuria (Table 3), the urinary protein level at conversion increased the hazard ratio by 1.006-fold (CI: 1.004–1.009, p < 0.001) per mg/L. Serum creatinine at conversion increased the hazard ratio to 1.431 per mg/dL (CI: 1.041–1.966, p = 0.027). Age at transplantation, sex, time from transplantation to conversion,
Conversion Year 1 242 177 145 115 Numbers at risk
Year 2 130 83
Year 3 87 56
Year 4 49 37
Year 5 27 18
No Yes
Fig. 9. New-onset proteinuria and death-censored graft survival (p = 0.007).
Table 3. Cox regression analysis for new-onset proteinuria
95% CI Exp (B) New-onset proteinuria Creatinine at conversion (mg/dL) Proteinuria at conversion (g/L) Age at TX (yr) Sex Time from transplantation to conversion (yr) BPAR before conversion Time on dialysis (month) Glomerulonephritis
Significance
Exp (B)
Lower
Upper
0.027
1.431
1.041
1.966
three month after transplantation) was very broad to obtain a comprehensive pattern of the real-life experience with sirolimus in Germany. Overall, data assembly and database management were successful, and 726 patients and more than 1500 patient years were documented in a very detailed manner. The demographics truly represent the German transplant population and show a representative picture of the very broad range of converted patients in Germany and other European countries (1). Apart from the general characterization of the cohort, the database allows interesting time-dependent and longitudinal analytical approaches. The age at conversion increased gradually over the study time. In parallel, the amount of patients being converted to sirolimus because of malignancies increased in the late years of the study. This observation is consistent with global trends of an aging society and the more frequent transplantation in elderly recipients as reflected by the introduction of special allocation schemes for the elderly such as the Eurotransplant Senior Program (1, 2, 21). Furthermore, this increasing age may well reflect the higher burden of malignancies in elderly patients and the translation of growing evidence of sirolimus’ antitumor effects into clinical practice (17, 22–24). A steady and statistically significant increase in the mean eGFR at conversion was observed during the study period. Simultaneously, fewer patients with overt proteinuria were converted. These trends follow the published evidence that proteinuria and low GFR are negative predictors for a successful conversion to sirolimus (25, 26). As first
German Sirolimus Study Group reports about a proteinuric effect were published in 2004, the rate of converted patients with proteinuria decreased. The dose of sirolimus decreased during the study time as the transplant centers became more experienced with the sufficient use of sirolimus. Being afraid of increased rejection rates, therapy with sirolimus in the earlier years was aimed at higher drug levels. Consistently good results with low rejection rates after conversion to sirolimus and the observed toxicity with a high exposure to sirolimus led to a gradual decrease in target levels for sirolimus over time. An association of proteinuria with sirolimus was shown earlier (14, 15, 25, 27). The impact of mTORi on podocyte function and development of proteinuria remains controversial (27, 28). As shown in the CONVERT trial, proteinuria had a detrimental effect on graft survival (15). Hence, the observed trends over time nicely reflect the growing clinical knowledge on mTORi and lessons learned from published data on mTORi conversion. Our data clearly support this observation in a large multicenter cohort and provide further data on the magnitude, timing, consequences, and risk factors. The five-yr graft survival was lower than the five-yr patient survival in many analyses. This phenomenon was described by Meier-Kriesche and colleagues showing a higher death rate among patients after graft loss (29). The extent of proteinuria at conversion showed a detrimental impact on patient, graft, and overall survival. Due to small numbers, patient survival in the group with the highest extent of proteinuria was higher than in the second highest. It was shown earlier that overt proteinuria at conversion is a detrimental factor for graft survival (14, 30, 31). Both the emergence of new-onset proteinuria and the extent of proteinuria at conversion show an inverse association with graft survival. This is the first study aiming to identify a cutoff for inferior long-term outcomes with an immunosuppression containing mTOR inhibitors. Halimi et al. (32) have shown a detrimental effect of even low-grade proteinuria on graft survival at different time points after renal transplantation. However, in that retrospective analysis, no patient was treated with an mTORi (32). Similar to the CONVERT trial and Halimi et al., we found that rather low urinary protein concentrations at conversion already have a negative impact on outcome (15). Our cutoff values of 151–268 mg/L protein at conversion were clearly associated with inferior graft survival. Assuming an urinary output of around 2 l/d, these values correspond to a daily protein excretion of around 300–500 mg/d. Thus, our observations redefine initial recommendations (14)
that proteinuria >800 mg/d is associated with inferior outcomes and lowers the threshold further down. Extending the data from the CONVERT trial (15), a safe conversion to sirolimus with a successful outcome seems advisable predominantly for patients with very low urinary protein levels. However, overall risks and benefits have always to be balanced in the light of the individual patient. Most cases of new-onset proteinuria were detected within the first six month after conversion. This is earlier than that observed by Franco et al. (33), who found proteinuria occurring in about 50% after the first six months after conversion, and more consistent with Pinheiro et al. (34), who found 73.3% occurring within the first six months. However, in our study, we did not detect such high proteinuria levels as in those studies, possibly due to lower baseline values in most patients. We are not sure, whether the high incidence of proteinuria within six months after conversion is due to a direct toxic effect of sirolimus or a fading of described antiproteinuric effects of CNI use (35). As many patients in early years were converted to sirolimus with a low eGFR, they might have already been beyond a point of no return. As our study lacks of a control group, this question remains unsolved. Moreover, the chance of new-onset proteinuria increased already if proteinuria at conversion was >71 mg/L, and our data clearly demonstrate for the first time a negative association of new-onset proteinuria after conversion to sirolimus with outcome in a large database. Importantly, we could not find any association between sirolimus levels and new-onset proteinuria. There are several limitations to our study. One major issue is the lack of a control group in this retrospective data collection. Because patients were converted for an indication, or for several indications, it is difficult to estimate the true incidence of the natural course of proteinuria without a control group. However, this is one of the largest cohorts in which this issue was systematically assessed. The inclusion of a broad range of unselected patients in a real-life setting has some strengths but also some limitations. While this heterogeneity certainly is a potential confounder, we tried to eliminate this confounder as much as possible (e.g., by performing subgroup and multivariate analyses). Unfortunately, only about 60% of urine protein measurements were quantitative, as the imputation of dipstick values is not as accurate as quantitative measurements. Nevertheless, dipstick urine analysis provided important additional information, in particular whether dipstick remained negative or became highly positive. Urine protein measure-
77
Naik et al.
ments increased in 2004 to a regular routine when first evidence of mTORi influencing proteinuria appeared. Finally, there has been recent evidence that antibody-mediated rejections occur with a higher frequency in patients being treated with another mTORi (16). We do not have data about antibody-mediated rejections as there were no protocol biopsies. Conclusion
In summary, we present a comprehensive database that reflects major medical trends and clinical changes over a long period. Despite the growing age of our transplant recipients reflecting the demographic change, we could see a further improvement in renal function as measured by the estimated GFR from the abbreviated MDRD formula. A new drug is used more efficiently after some lessons are learned. After the first reports about the proteinuric potential of sirolimus appeared, patients were selected more carefully before conversion. Our study suggests that patients with low proteinuria prior to conversion may benefit from a sirolimusbased regimen. As sirolimus has a proteinuric potential, intense follow-up is necessary to avoid deteriorating patient and graft survival. Acknowledgements
This work was supported by an unrestricted grant from Pfizer Pharma GmbH, Germany (formerly Wyeth Pharma GmbH). References 1. FREI U, SCHOBER-HALSTENBERG H-J. Nierenersatztherapie €ber Dialysebehandlung und Niin Deutschland – Bericht u erentransplantation in Deutschland 2006–2007. Berlin: Quasi Niere gGmbh, 2008: 1. 2. COLLINS AJ, FOLEY RN, HERZOG C et al. US Renal Data System 2010 Annual Data Report. Am J Kidney Dis 2011: 57(1 Suppl 1): A8, e1-A8, 526. 3. PONTICELLI C. Present and future of immunosuppressive therapy in kidney transplantation. Transplant Proc 2011: 43: 2439. 4. GEISSLER EK, SCHLITT HJ. The potential benefits of rapamycin on renal function, tolerance, fibrosis, and malignancy following transplantation. Kidney Int 2010: 78: 1075. 5. JOHNSON RW, KREIS H, OBERBAUER R, BRATTSTROM C, CLAESSON K, ERIS J. Sirolimus allows early cyclosporine withdrawal in renal transplantation resulting in improved renal function and lower blood pressure. Transplantation 2001: 72: 777. 6. OBERBAUER R, KREIS H, JOHNSON RW et al. Long-term improvement in renal function with sirolimus after early cyclosporine withdrawal in renal transplant recipients: 2year results of the Rapamune Maintenance Regimen Study. Transplantation 2003: 76: 364.
78
7. EKBERG H, TEDESCO-SILVA H, DEMIRBAS A et al. Reduced exposure to calcineurin inhibitors in renal transplantation. N Engl J Med 2007: 357: 2562. 8. FLECHNER SM. Sirolimus in kidney transplantation indications and practical guidelines: de novo sirolimus-based therapy without calcineurin inhibitors. Transplantation 2009: 87(8 Suppl): S1. 9. GUBA M, PRATSCHKE J, HUGO C et al. Renal function, efficacy, and safety of sirolimus and mycophenolate mofetil after short-term calcineurin inhibitor-based quadruple therapy in de novo renal transplant patients: one-year analysis of a randomized multicenter trial. Transplantation 2010: 90: 175. 10. LEBRANCHU Y, THIERRY A, THERVET E et al. Efficacy and safety of early cyclosporine conversion to sirolimus with continued MMF-four-year results of the Postconcept study. Am J Transplant 2011: 11: 1665. 11. WEIR MR, MULGAONKAR S, CHAN L et al. Mycophenolate mofetil-based immunosuppression with sirolimus in renal transplantation: a randomized, controlled Spare-theNephron trial. Kidney Int 2011: 79: 897. 12. KAHAN BD. Two-year results of multicenter phase III trials on the effect of the addition of sirolimus to cyclosporine-based immunosuppressive regimens in renal transplantation. Transplant Proc 2003: 35(3 Suppl): 37S. 13. LEBRANCHU Y, THIERRY A, TOUPANCE O et al. Efficacy on renal function of early conversion from cyclosporine to sirolimus 3 months after renal transplantation: concept study. Am J Transplant 2009: 9: 1115. 14. DIEKMANN F, ANDRES A, OPPENHEIMER F. mTOR inhibitor-associated proteinuria in kidney transplant recipients. Transplant Rev (Orlando) 2012: 26: 27. 15. SCHENA FP, PASCOE MD, ALBERU J et al. Conversion from calcineurin inhibitors to sirolimus maintenance therapy in renal allograft recipients: 24-month efficacy and safety results from the CONVERT trial. Transplantation 2009: 87: 233. 16. LIEFELDT L, BRAKEMEIER S, GLANDER P et al. Donor-specific HLA antibodies in a cohort comparing everolimus with cyclosporine after kidney transplantation. Am J Transplant 2012: 12: 1192. 17. HALLECK F, DUERR M, WAISER J et al. An evaluation of sirolimus in renal transplantation. Expert Opin Drug Metab Toxicol 2012: 8: 1337. 18. WEISS NS, KOEPSELL TD, PSATY BM. Generalizability of the results of randomized trials. Arch Intern Med 2008: 168: 133. 19. ZWARENSTEIN M, OXMAN A. Why are so few randomized trials useful, and what can we do about it? J Clin Epidemiol 2006: 59: 1125. 20. In: IAS Berlin, Germany [electronic mail system]. 2012. 21. COLLINS AJ, FOLEY RN, HERZOG C et al. Excerpts from the United States Renal Data System 2008 annual data report. Am J Kidney Dis 2009: 53(1 Suppl): 1. 22. PINTO JR, ARELLANO TORRES EM, FRANCO A et al. Sirolimus monotherapy as maintenance immunosuppression: a multicenter experience. Transpl Int 2010: 23: 307. 23. HALLECK F, BUDDE K. Transplantation: sirolimus for secondary SCC prevention in renal transplantation. Nat Rev Nephrol 2012: 8: 687. 24. EUVRARD S, MORELON E, ROSTAING L et al. Sirolimus and secondary skin-cancer prevention in kidney transplantation. N Engl J Med 2012: 367: 329. 25. DIEKMANN F, BUDDE K, OPPENHEIMER F, FRITSCHE L, NEUMAYER HH, CAMPISTOL JM. Predictors of success in conversion from calcineurin inhibitor to sirolimus in
German Sirolimus Study Group
26.
27.
28.
29.
30.
chronic allograft dysfunction. Am J Transplant 2004: 4: 1869. CAMPISTOL JM, COCKWELL P, DIEKMANN F et al. Practical recommendations for the early use of m-TOR inhibitors (sirolimus) in renal transplantation. Transpl Int 2009: 22: 681. SCHONENBERGER E, EHRICH JH, HALLER H, SCHIFFER M. The podocyte as a direct target of immunosuppressive agents. Nephrol Dial Transplant 2011: 26: 18. LETAVERNIER E, BRUNEVAL P, VANDERMEERSCH S et al. Sirolimus interacts with pathways essential for podocyte integrity. Nephrol Dial Transplant 2009: 24: 630. KAPLAN B, MEIER-KRIESCHE HU. Death after graft loss: an important late study endpoint in kidney transplantation. Am J Transplant 2002: 2: 970. ROODNAT JI, MULDER PG, RISCHEN-VOS J et al. Proteinuria and death risk in the renal transplant population. Transplant Proc 2001: 33: 1170.
31. ROODNAT JI, MULDER PG, RISCHEN-VOS J et al. Proteinuria after renal transplantation affects not only graft survival but also patient survival. Transplantation 2001: 72: 438. 32. HALIMI JM, LAOUAD I, BUCHLER M et al. Early low-grade proteinuria: causes, short-term evolution and long-term consequences in renal transplantation. Am J Transplant 2005: 5: 2281. 33. FRANCO AF, MARTINI D, ABENSUR H, NORONHA IL. Proteinuria in transplant patients associated with sirolimus. Transplant Proc 2007: 39: 449. 34. PINHEIRO HS, AMARO TA, BRAGA AM, BASTOS MG. Postrapamycin proteinuria: incidence, evolution, and therapeutic handling at a single center. Transplant Proc 2006: 38: 3476. 35. LETAVERNIER E, LEGENDRE C. mToR inhibitors-induced proteinuria: mechanisms, significance, and management. Transplant Rev (Orlando) 2008: 22: 125.
79
Clin Transplant 2014: 28: 67–79 DOI: 10.1111/ctr.12280
Clinical Transplantation
Proteinuria and sirolimus after renal transplantation: a retrospective analysis from a large German multicenter database Naik MG, Heller KM, Arns W, Budde K, Diekmann F, Eitner F, Fischereder M, Goßmann J, Heyne N, Morath C, Riester U, Gwinner W and J€ urgensen JS for the German Sirolimus Study Group. Proteinuria and sirolimus after renal transplantation: a retrospective analysis from a large German multicenter database. Abstract: The German Sirolimus Study Group has established a database among 10 transplant centers throughout Germany to study the outcomes in 726 renal transplant patients being converted to a sirolimuscontaining therapy between 2000 and 2008 with a total of more than 1500 recorded patient years on therapy. In this study, we present a detailed description of the cohort, of characteristic changes over the observation period, proteinuria and graft survival, and new-onset proteinuria after conversion. Over the study period, age, graft function at the time of conversion, and the proportion of patients switched to sirolimus because of malignancy increased, whereas the proportion of patients with significant proteinuria at conversion decreased. Already modest proteinuria (151–268 mg/L) at conversion and new-onset proteinuria (>500 mg/L) after conversion were associated with inferior graft survival. Even mild proteinuria (>71 mg/L) at conversion was associated with new-onset proteinuria (>500 mg/L) post-conversion. Serum creatinine and urinary protein excretion at conversion together with age at transplantation had a significant impact on patient and graft survival. This large data set confirms and extends previous observations that proteinuria is an important indicator for graft outcome after conversion to sirolimus. We conclude that patients without any proteinuria have the greatest benefit from conversion to sirolimus.
Marcel G. Naika,*, Katharina M. Hellerb,*, Wolfgang Arnsc, Klemens Buddea, Fritz Diekmannd, Frank Eitnere, Michael Fischerederf, Jan Goßmanng, Nils Heyneh, Christian Morathi, Udo Riesterj, Wilfried Gwinnerk,* and € rgensenl for the Jan Steffen Ju German Sirolimus Study Group a
Department of Nephrology, Charité University – Mitte, Berlin, bDivision of Nephrology, Department of Medicine, University of Erlangen, Erlangen, cTransplant, Centre Cologne, Cologne General Hospital, Cologne Germany, dHospital Clinic, Barcelona Spain, e Division of Nephrology and Clinical Immunology, RWTH Aachen University, Aachen, fDepartment of Medicine IV, University Hospital LMU, Munich, gTransplant Care Centre Frankfurt, Frankfurt, hSection of Nephrology and Hypertension, Department of €bingen University Internal Medicine IV, Tu €bingen, iDepartment of Hospital, Tu Nephrology, University of Heidelberg, Heidelberg, jPfizer Pharma GmbH, Berlin, k Department of Nephrology, Hannover Medical School, Hannover and lDepartment of Nephrology, Charite´ University – Virchow, Berlin, Germany Key words: conversion – immunosuppression – mammalian target of rapamycin inhibitor – proteinuria – renal transplantation – sirolimus Corresponding author: Dr. Jan Steffen €rgensen, MPH, Department for Nephrology Ju and Medical Intensive Care, Universita¨tsmedizin Berlin, Campus Virchow Klinikum, Augustenburger Platz 1, Charite 13353 Berlin, Germany. Tel.: +49 30 450 570 606; fax: +49 30 450 7570606; e-mail: [email protected] *Both respective authors contributed equally. Conflict of interest: None. Accepted for publication 21 October 2013
67
Naik et al.
Patient and graft survival early after kidney transplantation increased dramatically in recent decades (1, 2). This successful development is largely based upon the introduction of new immunosuppressive agents with more distinct mechanisms of action that allow tailored regimens for individual patients (3). In contrast, long-term survival was not improved by all these strategies (1, 2). After its approval for use in organ transplantation in 1999, the mammalian target of rapamycin inhibitor (mTORi) sirolimus extended the therapeutic options. Early studies suggested that sirolimus was less nephrotoxic than calcineurin inhibitors (CNIs) and showed antifibrotic and anticancer properties (4). Sirolimus found its way into different immunosuppressive regimens both for de novo and maintenance immunosuppression (5–11). Early publications indicated favorable effects of sirolimus on long-term renal allograft function compared with CNI-based immunosuppressive regimens (9, 10, 12, 13). However, specific adverse effects limit the use of sirolimus in certain patients and require watchful use (14–16). Over the last decade, many randomized controlled trials (RCTs) have provided valuable information regarding the use and contraindications of sirolimus in distinguished patient groups that have been included in these studies and allowed a refined application of the drug (17). These robust RCTs aimed to exclude any systematic bias and supported clinical decision making with regard to the use of sirolimus. RCTs are designed as experiments with high internal validity, for example, the validity of the interferences drawn as they apply to the members of the source population. Clinical trials employ sophisticated methods to control for systematic errors by means of blinding, randomization, allocation concealment, and clear-cut inclusion and exclusion criteria. Usually, this leads to statistically sound and credible results. On the other hand, exactly those above-mentioned factors that determine the high internal validity of a clinical study often hamper its external validity. In other words, the validity of the interferences as they pertain to people outside the tightly defined source population or study cohort may be very limited. This lack of generalization with regard to patients with different characteristics, for example, elderly or those with specific comorbidities or comedication, is a major concern. On the other hand, observational studies and analysis based upon data gathered in clinical routine settings may be disposed to systematic errors, but offer additional valuable evidence and conclusions, which might be applicable for the full spectrum of patients under treatment (18, 19).
68
The German Sirolimus Study Group has established a database among 10 transplant centers in Germany to study sirolimus use under realistic conditions in more than 700 renal transplant patients who were switched to a sirolimus-containing regimen. This study provides a first analysis of the database and demographics of the population, with a focus of characteristic changes in key demographic variables over the eight-yr observation period in the conversion setting. Furthermore, in this study, we specifically wanted to thoroughly analyze proteinuria including new-onset proteinuria and clinical outcome in a large cohort of patients, who were converted to sirolimus.
Methods Patients
This multicenter, retrospective study comprises patients with a renal transplant or combined kidney transplantation with another solid organ who were converted to a sirolimus-containing immunosuppressive regimen three or more months after transplantation between January 1, 2000, and December 31, 2008. All patients were switched to sirolimus. Ten German transplant centers included their data to obtain a comprehensive database for the study of sirolimus in clinical practice. We aimed to include all converted patients in each of the participating centers regardless of success, length of sirolimus therapy, age, type of donation, number of previous transplants, and reason for conversion to sirolimus. There were no patients excluded from this database to reduce a selection bias and address the full spectrum of all therapeutic approaches. The study was approved by the local ethics committees, and all patients provided informed consent for the scientific use of their data. Data collection
Data collection included four distinct time points before initiation of sirolimus. The start of sirolimus therapy marked the study baseline to establish baseline values for allograft function, proteinuria, comorbidities, and medical therapies (immunosuppressive drugs and concomitant treatments). Data collection in the first year was intended at three, six, and 12 months after initiation of sirolimus and semiannually thereafter. Due to the retrospective nature of the study, actual time points for data collection in individual patients differed from this schedule. This was appreciated by building defined time periods and
German Sirolimus Study Group calculating the days from the study baseline for each patient. The time periods were calculated in days from baseline, with the pre-conversion visits 1Y (365 d before conversion), M6 (180 d before conversion), and M1 (30 d before conversion. Post-conversion visits were defined as M3 (days 31–135), M6 (days 136–270), Y1 (days 271– 547), Y2 (548–912), Y3 (913–1277), Y4 (1278– 1642), Y5 (1643–2007), and >Y5 (>2008). End points were defined as patient’s death and terminal allograft failure. Renal function was estimated using the abbreviated MDRD formula. Urinary protein excretion was recorded as a protein concentration in spot urine. Because each center used different units, all urinary protein concentrations were converted to the most frequently used unit (mg/L) assuming an urinary output of 2 L/d. A protein concentration of >500 mg/L was considered positive. Dipsticks were considered if no concentration was recorded. Positive dipsticks were imputed as the average protein concentration of patients with >500 mg/L proteinuria (750 mg/L), and negative dipsticks were imputed as 122 mg/L (the average protein concentration of patients 10% was detected before transplantation, and highly sensitized patients (PRA ≥ 85%) were present in only 0.5%. The mean creatinine at conversion was 2.16 0.97 mg/dL with an estimated glomerular filtration rate (eGFR) of 39.2 19.4 mL/min at
69
Naik et al. Table 1. Characteristics of the study population (N = 726)
Table 1. Continued
% General demographics Male Caucasian Living donation First renal transplantation Combined transplantation Thereof kidney pancreas BPAR before conversion BPAR after conversion RAAS before conversion RAAS after conversion DGF PRA > 10% Mismatches (MM broad) (mean SD) Age at TX (yr mean SD) Time on dialysis before TX (yr mean SD) Age at conversion (yr mean SD) Time from TX to conversion (yr mean SD) Time of follow-up (yr mean SD) eGFR at conversion (mL/min mean SD) Proteinuria at conversion (mg/L mean SD) First sirolimus dose (mg mean SD) Maintenance (sirolimus dose) (mg mean SD) Underlying renal disease Diabetes mellitus Hypertension Polycystic renal disease Glomerulonephritis Tubulointerstitial disease Other genetic disease Unknown Others Comorbidities Diabetes mellitus Arterial hypertension Neurologic disease Chronic liver disease Cardiovascular disease Coronary heart disease Peripheral arterial occlusive disease Cerebrovascular disease Immunosuppression Initial immunosuppression Induction with antibody Non-depleting Depleting Steroids CNI* Cyclosporine Tacrolimus Antiproliferative immunosuppressant* Mycophenolic acid Azathioprine At month 3 Sirolimus Steroids CNI* Cyclosporine Tacrolimus Antiproliferative immunosuppressant*
70
% 63.3 99.0 16.4 74.5 10.5 86.7 38.1 9.0 60.9 72.7 17.4 7.7 2.4 43.3 3.9 49.8 6.6
2.3 39.2 349 6.4 2.9 12.3 3.6 11.3 43.0 14.2 3.6 7.9 4.3 23.6 84.7 13.3 9.6 23.6 18.4 7.2 2.6
19.7 13.8 98.2 63.0 27.2 54.3 32.2 94.1 87.7 7.6 22.1
Mycophenolic acid Azathioprine Reason for conversion Study Tumor Creeping creatinine Side effect of other immunosuppressant Acute rejection Others Chronic allograft nephropathy CNI toxicity
1.64 13.57 3.13 13.35 6.11
1.82 19.4 754.56 4.78 1.64
39.2 4.5 11.1 23.0 22.6 10.1 12.0 24.9 17.7 26.5
TX, transplantation; BPAR, biopsy-proven acute rejection; CNI, calcineurin inhibitor; eGFR, estimated glomerular filtration rate; PRA, panelreactive antibodies. *Value may exceed 100% due to concomitant or overlapping use in some patients.
conversion. Mean proteinuria at conversion was 349 755 mg/L with a wide range (0–5480 mg/ L), with a median of 107.5 mg/L. There were 17% of patients with proteinuria exceeding 800 mg/d. The most frequent reasons for conversion were “CNI toxicity” (28.7%), “tumors” (25.4%), “creeping creatinine” (22.7%), and “chronic allograft nephropathy” (18.7%). It is important to note that multiple reasons could be ticked for individual patients. There were 80 patients (11.1%) who were included in various single and multicenter studies (e.g., skin cancer and creeping creatinine). The majority of patients converted due to tumors had skin tumors. Further analysis concerning tumors will be required. The immunosuppressive regimen before conversion to sirolimus consisted of corticosteroids (98.2%), CNIs such as cyclosporine or tacrolimus (90.2%), and antiproliferative immunosuppressants such as mycophenolic acid or azathioprine (86.5%). Prior antibody therapy was performed in 33.5% (13.8% depleting and 19.7% non-depleting antibodies). At conversion, the initial dose of sirolimus was 6.4 4.8 mg/d and the maintenance dose at the first visit post-conversion was 2.9 1.6 mg/d. The sirolimus trough level was 8.0 3.6 ng/mL and 8.2 1.2 ng/mL at six and 12 months, respectively. Three months after conversion, 70.3% were CNI free, 29.7% on low-dose CNI combination therapy (7.6% cyclosporine and 22.1% tacrolimus), 43.7% on antiproliferative medication (39.2% MPA and 4.5% azathioprine), and 87.7% on steroids. The vast majority of patients (72.9%) were treated with ACE inhibitors or ARB after conversion. In 65 patients (9.0%), BPAR was documented after conversion.
German Sirolimus Study Group dominated by conversion due to deteriorating renal function (e.g., “chronic allograft nephropathy” 50% in 2000 vs. 9.7% in 2008, p < 0.001), the potential antitumor effects of sirolimus accounted for a larger proportion of conversions at later time points (0% in 2000 vs. 25.8% in 2008, p < 0.001). As a consequence, a steady and statistically significant (p < 0.001) increase in the mean eGFR at conversion was observed from 2000 to 2008 (27.1 17.2 vs. 50.5 35.8 mL/min, Fig. 1C). Along with the increase in eGFR, the percentage of patients being converted with an eGFR 500 mg/L at conversion over the study period.
71
Naik et al.
in 2000 vs. 46.7% in 2008, p < 0.001). The use of CNIs showed a trend toward an increasing use of tacrolimus (37% in 2000 vs. 46.7% in 2008, p = 0.067) and rather stable use of cyclosporine (59.3% in 2000 vs. 53.3% in 2008, p = 0.79). Azathioprine (40.7% in 2000 vs. 20% in 2008, p < 0.001) was largely replaced by mycophenolic acid (40.7% in 2000 vs. 63.3% in 2008, p < 0.001) toward the final years of the study. The initial and maintenance dose of sirolimus fell from 12.9 4.1 mg/d (range 3–15 mg/d) and 4.7 1.8 mg/d in 2000 to 2.7 1.2 mg/d (range 1–5 mg/d) and 1.9 0.7 mg/d in 2008 (p < 0.001 each), respectively. The sirolimus trough levels at six months decreased significantly from 10.6 5.0 ng/mL in 2000 to 5.8 2.7 ng/mL in 2008 (p < 0.001). There were no significant changes during the study period regarding sex ratio, type of donation, number of transplant, combined transplantation, HLA mismatches, BPAR, delayed graft function, panel-reactive antibody level >10%, time from transplantation to conversion, time on dialysis, use of ACE/ARB agents before and after conversion, or underlying renal disease.
A
Conversion 417 87 18 6
Year 1 288 41 4 2
Year 2 207 25 0 1
Year 3 134 16 0 0
Year 4 77 11 0 0
Year 5 38 4 0 0
3500mg/l
Year 5 41 4 0 0
3500mg/l
Year 5 35 3 0 0
3500mg/l
Numbers at risk
B
Proteinuria at conversion and outcomes
Overall patient survival was 79.9%, and graft survival rates were 67.2% and 59.4% excluding and including death after five yr, respectively. There were 528 patients with data on proteinuria at the time of conversion (311 patients with urinary protein concentration and 217 dipsticks). In 110 patients (20.8%), proteinuria >500 mg/L was detected at conversion (65 patients with urinary protein concentration and 45 dipsticks). First we looked whether proteinuria >500 mg/L at conversion had a negative impact on patient, graft, and overall survival. The five-yr patient survival was 85.8% vs. 59.9% (p < 0.001). The five-yr death-censored graft survival was 76.6% vs. 31.7% (p < 0.001), and five-yr overall survival (including death) was 68.5% vs. 21.8% (p < 0.001). Next, we examined whether the extent of proteinuria at conversion is associated with patient, graft, and overall survival (Fig. 2A–C). In the group of proteinuria 3500 mg/L (N = 6), the five-yr patient survival was 85.8%, 62.6%, 41.7%, and 80%, respectively (p < 0.001), five-yr death-censored graft survival was 76.9%, 36.8%, 0%, and 50%, respectively (p < 0.001), and the five-yr overall survival (including death) was 68.7%, 24.7%, 10.3%, and 41.7%, respectively (p < 0.001).
72
Conversion 417 87 18 6
Year 1 293 42 5 2
Year 2 210 26 2 1
Year 3 136 15 0 0
Year 4 80 10 0 0
Numbers at risk
C
Conversion Year 1 417 295 87 46 18 5 6 2 Numbers at risk
Year 2 214 27 1 1
Year 3 137 15 1 0
Year 4 76 9 0 0
Fig. 2. (A) Patient survival and extent of proteinuria at conversion (p < 0.001). (B) Graft survival excluding death and extent of proteinuria at conversion (p < 0.001). (C) Graft survival including death and extent of proteinuria at conversion (p < 0.001).
German Sirolimus Study Group To systematically assess the relationship between proteinuria and survival, we performed a ROC curve analysis. The first ROC curve analysis for patient survival using quantitative value only showed an area under the curve (AUC) of 0.557, p = 0.39, N = 311. The second ROC curve analysis for patient survival using imputed values as well showed an AUC of 0.577, p = 0.13, N = 528. There was no significant impact on patient survival. We found a significant relationship between proteinuria at conversion and death-censored graft survival with an AUC of 0.728 (p < 0.001, N = 311, Fig. 3) and a cutoff value of 268 mg/L urinary protein (sensitivity 61.4%, specificity 83.9%, positive predictive value [PPV] 90.6%, and negative predictive value [NPV] 46.1%). Considering imputed values as well, the AUC was 0.71 (p < 0.001, N = 524) and the cutoff value was 151 mg/L (sensitivity 66.7%, specificity 72.9%, PPV 91.1%, NPV 34.4%, N = 528). As shown in Fig. 4A, already modest proteinuria >268 mg/L had a strong impact on graft failure. The five-yr death-censored graft survival was 79.7% in patients below and 34.8% in patients above the cutoff value of 268 mg/L. Using imputed values for the cutoff value of 151 mg/L, five-yr death-censored graft survival was 81.1% (151 mg/L), respectively (p < 0.001, Fig. 4B).
A
Conversion Year 1 235 162 76 31 Numbers at risk
Year 2 119 19
Year 3 77 10
Year 4 38 5
Year 5 18 3
268mg/l
B
Conversion 348 180
Year 1 248 93
Year 2 180 58
Year 3 116 36
Year 4 70 20
Year 5 34 11
151mg/l
Numbers at risk
Fig. 4. (A) Proteinuria cutoff value at conversion and graft failure. p < 0.001, N = 311. (B) Proteinuria cutoff value at conversion and graft failure. p < 0.001, N = 528.
Fig. 3. ROC curve of all patients with urinary protein concentration and death-censored graft failure after conversion. area under the curve 0.728. p < 0.0005, N = 311.
When performing a ROC curve analysis for overall survival (including death), we found a cutoff of 151 mg/L (p < 0.001, AUC 0.699, sensitivity 67.1%, specificity 72.3%, PPV 87.8%, NPV 42.6%, N = 311). The overall five-yr overall survival (including death) was 72.3% vs. 33.6% (p < 0.001, Fig. 5A). The same cutoff was calculated in the second ROC curve analysis using imputed values (AUC 0.692, p < 0.001, sensitivity 61.8%, specificity 74.3%, PPV 86.5%, NPV 42.2%, N = 528). The overall five-yr overall survival (including death) was 72.7% vs. 33.7% (p < 0.001, Fig. 5B) for this imputed cutoff. Lastly, we performed a Cox proportional hazard analysis on overall survival (including death) (Table 2). There was a 1.556-fold (CI: 1.252–1.934, p < 0.001) higher risk of graft loss or death per 1000 mg/L urinary protein concentration at conversion. Similarly, creatinine at conversion was
73
Naik et al. Table 2. Cox regression analysis for overall graft survival (including death)
A
95% CI Exp (B) Overall survival
Significance
Exp (B)
Lower
Upper
Creatinine at conversion (mg/dL) Proteinuria at conversion (g/L) Age at Tx (yr) Sex Time from Tx to conversion (yr) BPAR before conversion Time on dialysis (month) Glomerulonephritis
500 mg/L) was detected after conversion (Fig. 6), resulting in an estimated risk of newonset proteinuria of about 37.6% after five yr. Patients with tumors (N = 104) had similar rates of new-onset proteinuria compared with patients without (36.5% vs. 37.8%, p = 0.819). Patients having undergone a combined kidney/ pancreas transplantation (N = 38) showed a trend toward a less frequent incidence of newonset proteinuria compared with patients with kidney-only transplantation (23.7% vs. 39.7%, p = 0.060), although this analysis is limited by low numbers. In most cases, new-onset proteinuria (>500 mg/ L) started within the first six months after conversion (N = 57, 19.2%). Patients with a new-onset proteinuria had already significantly higher urinary protein concentrations at the time of conversion (mean 78 vs. 155 mg/L, p < 0.0005). The
74
Conversion Year 1 348 253 180 95 Numbers at risk
Fig. 5. (A) Proteinuria cutoff value (without imputation) at conversion and overall survival (including death), p < 0.001, N = 311. (B) Proteinuria cutoff value (with imputation) at conversion and overall survival (including death), p < 0.001, N = 528.
serum creatinine (2.33 1.14 vs. 1.98 1.56 mg/dL, p = 0.002) and consequently the eGFR (37.7 18.7 vs. 46.2 23.5 mL/min, p = 0.004) differed significantly between patients with and without new-onset proteinuria at the end of observation. To find a cutoff value for the development of new-onset proteinuria, we performed a ROC curve analysis with an AUC of 0.749 (Fig. 7, p < 0.001, sensitivity 76.4%, specificity 65.8%, PPV 81.1%, NPV 57.4%, N = 189). Already mild proteinuria above 71 mg/L showed a significant impact on the
German Sirolimus Study Group A
Conversion 332
Year 1 179
Year 2 119
Year 3 76
Year 4 49
Year 5 25
Numbers at risk
Fig. 6. Time to new-onset proteinuria (>500 mg/L) after conversion (N = 90/332 = 27.1%).
Conversion Year 1 89 61 77 31 Numbers at risk
Year 2 41 18
Year 3 27 12
Year 4 17 5
Year 5 8 3
71mg/l
Year 2 18 61
Year 3 12 38
Year 4 5 23
Year 5 3 13
71mg/l
B
Conversion Year 1 89 31 187 96 Numbers at risk
Fig. 8. (A) Cutoff value (without imputation) and development of new-onset proteinuria, N = 166, p < 0.0005. (B) Cutoff value (with imputation) and development of new-onset proteinuria, N = 276, p < 0.0005. Fig. 7. ROC curve of patients with new-onset proteinuria after conversion. area under the curve 0.749. p < 0.001, N = 189.
development of new-onset proteinuria (Fig. 8A, p < 0.0005, N = 166), being 74.6% vs. 32.5% after five yr, respectively. Using imputed values, the cutoff was the same with an AUC of 0.658 (Fig. 7B, p < 0.001, sensitivity 61.7%, specificity 38.3%, PPV 67.6%, NPV 63.4%, N = 316). As shown in Fig. 8B, development of new-onset proteinuria after five yr was 74.6 vs. 54.7% (N = 276, p < 0.0005).
New-onset proteinuria was also associated with a significantly lower graft survival (Fig. 9). After five yr, death-censored graft survival was 72.2% in patients with new-onset proteinuria and 84.1% in patients without new-onset proteinuria (p = 0.007). New-onset proteinuria had a strong tendency to deteriorate overall survival (including death), being 78.0% vs. 61.6% (p = 0.052) after five yr. Again, patient survival was not influenced significantly by new-onset proteinuria (85.2% vs. 86.5%).
75
Naik et al.
We did not observe a correlation between high sirolimus trough levels and the development of new-onset proteinuria at three, six, or 12 months after conversion. In a Cox proportional regression analysis for risk factors of new-onset proteinuria (Table 3), the urinary protein level at conversion increased the hazard ratio by 1.006-fold (CI: 1.004–1.009, p < 0.001) per mg/L. Serum creatinine at conversion increased the hazard ratio to 1.431 per mg/dL (CI: 1.041–1.966, p = 0.027). Age at transplantation, sex, time from transplantation to conversion,
Conversion Year 1 242 177 145 115 Numbers at risk
Year 2 130 83
Year 3 87 56
Year 4 49 37
Year 5 27 18
No Yes
Fig. 9. New-onset proteinuria and death-censored graft survival (p = 0.007).
Table 3. Cox regression analysis for new-onset proteinuria
95% CI Exp (B) New-onset proteinuria Creatinine at conversion (mg/dL) Proteinuria at conversion (g/L) Age at TX (yr) Sex Time from transplantation to conversion (yr) BPAR before conversion Time on dialysis (month) Glomerulonephritis
Significance
Exp (B)
Lower
Upper
0.027
1.431
1.041
1.966
three month after transplantation) was very broad to obtain a comprehensive pattern of the real-life experience with sirolimus in Germany. Overall, data assembly and database management were successful, and 726 patients and more than 1500 patient years were documented in a very detailed manner. The demographics truly represent the German transplant population and show a representative picture of the very broad range of converted patients in Germany and other European countries (1). Apart from the general characterization of the cohort, the database allows interesting time-dependent and longitudinal analytical approaches. The age at conversion increased gradually over the study time. In parallel, the amount of patients being converted to sirolimus because of malignancies increased in the late years of the study. This observation is consistent with global trends of an aging society and the more frequent transplantation in elderly recipients as reflected by the introduction of special allocation schemes for the elderly such as the Eurotransplant Senior Program (1, 2, 21). Furthermore, this increasing age may well reflect the higher burden of malignancies in elderly patients and the translation of growing evidence of sirolimus’ antitumor effects into clinical practice (17, 22–24). A steady and statistically significant increase in the mean eGFR at conversion was observed during the study period. Simultaneously, fewer patients with overt proteinuria were converted. These trends follow the published evidence that proteinuria and low GFR are negative predictors for a successful conversion to sirolimus (25, 26). As first
German Sirolimus Study Group reports about a proteinuric effect were published in 2004, the rate of converted patients with proteinuria decreased. The dose of sirolimus decreased during the study time as the transplant centers became more experienced with the sufficient use of sirolimus. Being afraid of increased rejection rates, therapy with sirolimus in the earlier years was aimed at higher drug levels. Consistently good results with low rejection rates after conversion to sirolimus and the observed toxicity with a high exposure to sirolimus led to a gradual decrease in target levels for sirolimus over time. An association of proteinuria with sirolimus was shown earlier (14, 15, 25, 27). The impact of mTORi on podocyte function and development of proteinuria remains controversial (27, 28). As shown in the CONVERT trial, proteinuria had a detrimental effect on graft survival (15). Hence, the observed trends over time nicely reflect the growing clinical knowledge on mTORi and lessons learned from published data on mTORi conversion. Our data clearly support this observation in a large multicenter cohort and provide further data on the magnitude, timing, consequences, and risk factors. The five-yr graft survival was lower than the five-yr patient survival in many analyses. This phenomenon was described by Meier-Kriesche and colleagues showing a higher death rate among patients after graft loss (29). The extent of proteinuria at conversion showed a detrimental impact on patient, graft, and overall survival. Due to small numbers, patient survival in the group with the highest extent of proteinuria was higher than in the second highest. It was shown earlier that overt proteinuria at conversion is a detrimental factor for graft survival (14, 30, 31). Both the emergence of new-onset proteinuria and the extent of proteinuria at conversion show an inverse association with graft survival. This is the first study aiming to identify a cutoff for inferior long-term outcomes with an immunosuppression containing mTOR inhibitors. Halimi et al. (32) have shown a detrimental effect of even low-grade proteinuria on graft survival at different time points after renal transplantation. However, in that retrospective analysis, no patient was treated with an mTORi (32). Similar to the CONVERT trial and Halimi et al., we found that rather low urinary protein concentrations at conversion already have a negative impact on outcome (15). Our cutoff values of 151–268 mg/L protein at conversion were clearly associated with inferior graft survival. Assuming an urinary output of around 2 l/d, these values correspond to a daily protein excretion of around 300–500 mg/d. Thus, our observations redefine initial recommendations (14)
that proteinuria >800 mg/d is associated with inferior outcomes and lowers the threshold further down. Extending the data from the CONVERT trial (15), a safe conversion to sirolimus with a successful outcome seems advisable predominantly for patients with very low urinary protein levels. However, overall risks and benefits have always to be balanced in the light of the individual patient. Most cases of new-onset proteinuria were detected within the first six month after conversion. This is earlier than that observed by Franco et al. (33), who found proteinuria occurring in about 50% after the first six months after conversion, and more consistent with Pinheiro et al. (34), who found 73.3% occurring within the first six months. However, in our study, we did not detect such high proteinuria levels as in those studies, possibly due to lower baseline values in most patients. We are not sure, whether the high incidence of proteinuria within six months after conversion is due to a direct toxic effect of sirolimus or a fading of described antiproteinuric effects of CNI use (35). As many patients in early years were converted to sirolimus with a low eGFR, they might have already been beyond a point of no return. As our study lacks of a control group, this question remains unsolved. Moreover, the chance of new-onset proteinuria increased already if proteinuria at conversion was >71 mg/L, and our data clearly demonstrate for the first time a negative association of new-onset proteinuria after conversion to sirolimus with outcome in a large database. Importantly, we could not find any association between sirolimus levels and new-onset proteinuria. There are several limitations to our study. One major issue is the lack of a control group in this retrospective data collection. Because patients were converted for an indication, or for several indications, it is difficult to estimate the true incidence of the natural course of proteinuria without a control group. However, this is one of the largest cohorts in which this issue was systematically assessed. The inclusion of a broad range of unselected patients in a real-life setting has some strengths but also some limitations. While this heterogeneity certainly is a potential confounder, we tried to eliminate this confounder as much as possible (e.g., by performing subgroup and multivariate analyses). Unfortunately, only about 60% of urine protein measurements were quantitative, as the imputation of dipstick values is not as accurate as quantitative measurements. Nevertheless, dipstick urine analysis provided important additional information, in particular whether dipstick remained negative or became highly positive. Urine protein measure-
77
Naik et al.
ments increased in 2004 to a regular routine when first evidence of mTORi influencing proteinuria appeared. Finally, there has been recent evidence that antibody-mediated rejections occur with a higher frequency in patients being treated with another mTORi (16). We do not have data about antibody-mediated rejections as there were no protocol biopsies. Conclusion
In summary, we present a comprehensive database that reflects major medical trends and clinical changes over a long period. Despite the growing age of our transplant recipients reflecting the demographic change, we could see a further improvement in renal function as measured by the estimated GFR from the abbreviated MDRD formula. A new drug is used more efficiently after some lessons are learned. After the first reports about the proteinuric potential of sirolimus appeared, patients were selected more carefully before conversion. Our study suggests that patients with low proteinuria prior to conversion may benefit from a sirolimusbased regimen. As sirolimus has a proteinuric potential, intense follow-up is necessary to avoid deteriorating patient and graft survival. Acknowledgements
This work was supported by an unrestricted grant from Pfizer Pharma GmbH, Germany (formerly Wyeth Pharma GmbH). References 1. FREI U, SCHOBER-HALSTENBERG H-J. Nierenersatztherapie €ber Dialysebehandlung und Niin Deutschland – Bericht u erentransplantation in Deutschland 2006–2007. Berlin: Quasi Niere gGmbh, 2008: 1. 2. COLLINS AJ, FOLEY RN, HERZOG C et al. US Renal Data System 2010 Annual Data Report. Am J Kidney Dis 2011: 57(1 Suppl 1): A8, e1-A8, 526. 3. PONTICELLI C. Present and future of immunosuppressive therapy in kidney transplantation. Transplant Proc 2011: 43: 2439. 4. GEISSLER EK, SCHLITT HJ. The potential benefits of rapamycin on renal function, tolerance, fibrosis, and malignancy following transplantation. Kidney Int 2010: 78: 1075. 5. JOHNSON RW, KREIS H, OBERBAUER R, BRATTSTROM C, CLAESSON K, ERIS J. Sirolimus allows early cyclosporine withdrawal in renal transplantation resulting in improved renal function and lower blood pressure. Transplantation 2001: 72: 777. 6. OBERBAUER R, KREIS H, JOHNSON RW et al. Long-term improvement in renal function with sirolimus after early cyclosporine withdrawal in renal transplant recipients: 2year results of the Rapamune Maintenance Regimen Study. Transplantation 2003: 76: 364.
78
7. EKBERG H, TEDESCO-SILVA H, DEMIRBAS A et al. Reduced exposure to calcineurin inhibitors in renal transplantation. N Engl J Med 2007: 357: 2562. 8. FLECHNER SM. Sirolimus in kidney transplantation indications and practical guidelines: de novo sirolimus-based therapy without calcineurin inhibitors. Transplantation 2009: 87(8 Suppl): S1. 9. GUBA M, PRATSCHKE J, HUGO C et al. Renal function, efficacy, and safety of sirolimus and mycophenolate mofetil after short-term calcineurin inhibitor-based quadruple therapy in de novo renal transplant patients: one-year analysis of a randomized multicenter trial. Transplantation 2010: 90: 175. 10. LEBRANCHU Y, THIERRY A, THERVET E et al. Efficacy and safety of early cyclosporine conversion to sirolimus with continued MMF-four-year results of the Postconcept study. Am J Transplant 2011: 11: 1665. 11. WEIR MR, MULGAONKAR S, CHAN L et al. Mycophenolate mofetil-based immunosuppression with sirolimus in renal transplantation: a randomized, controlled Spare-theNephron trial. Kidney Int 2011: 79: 897. 12. KAHAN BD. Two-year results of multicenter phase III trials on the effect of the addition of sirolimus to cyclosporine-based immunosuppressive regimens in renal transplantation. Transplant Proc 2003: 35(3 Suppl): 37S. 13. LEBRANCHU Y, THIERRY A, TOUPANCE O et al. Efficacy on renal function of early conversion from cyclosporine to sirolimus 3 months after renal transplantation: concept study. Am J Transplant 2009: 9: 1115. 14. DIEKMANN F, ANDRES A, OPPENHEIMER F. mTOR inhibitor-associated proteinuria in kidney transplant recipients. Transplant Rev (Orlando) 2012: 26: 27. 15. SCHENA FP, PASCOE MD, ALBERU J et al. Conversion from calcineurin inhibitors to sirolimus maintenance therapy in renal allograft recipients: 24-month efficacy and safety results from the CONVERT trial. Transplantation 2009: 87: 233. 16. LIEFELDT L, BRAKEMEIER S, GLANDER P et al. Donor-specific HLA antibodies in a cohort comparing everolimus with cyclosporine after kidney transplantation. Am J Transplant 2012: 12: 1192. 17. HALLECK F, DUERR M, WAISER J et al. An evaluation of sirolimus in renal transplantation. Expert Opin Drug Metab Toxicol 2012: 8: 1337. 18. WEISS NS, KOEPSELL TD, PSATY BM. Generalizability of the results of randomized trials. Arch Intern Med 2008: 168: 133. 19. ZWARENSTEIN M, OXMAN A. Why are so few randomized trials useful, and what can we do about it? J Clin Epidemiol 2006: 59: 1125. 20. In: IAS Berlin, Germany [electronic mail system]. 2012. 21. COLLINS AJ, FOLEY RN, HERZOG C et al. Excerpts from the United States Renal Data System 2008 annual data report. Am J Kidney Dis 2009: 53(1 Suppl): 1. 22. PINTO JR, ARELLANO TORRES EM, FRANCO A et al. Sirolimus monotherapy as maintenance immunosuppression: a multicenter experience. Transpl Int 2010: 23: 307. 23. HALLECK F, BUDDE K. Transplantation: sirolimus for secondary SCC prevention in renal transplantation. Nat Rev Nephrol 2012: 8: 687. 24. EUVRARD S, MORELON E, ROSTAING L et al. Sirolimus and secondary skin-cancer prevention in kidney transplantation. N Engl J Med 2012: 367: 329. 25. DIEKMANN F, BUDDE K, OPPENHEIMER F, FRITSCHE L, NEUMAYER HH, CAMPISTOL JM. Predictors of success in conversion from calcineurin inhibitor to sirolimus in
German Sirolimus Study Group
26.
27.
28.
29.
30.
chronic allograft dysfunction. Am J Transplant 2004: 4: 1869. CAMPISTOL JM, COCKWELL P, DIEKMANN F et al. Practical recommendations for the early use of m-TOR inhibitors (sirolimus) in renal transplantation. Transpl Int 2009: 22: 681. SCHONENBERGER E, EHRICH JH, HALLER H, SCHIFFER M. The podocyte as a direct target of immunosuppressive agents. Nephrol Dial Transplant 2011: 26: 18. LETAVERNIER E, BRUNEVAL P, VANDERMEERSCH S et al. Sirolimus interacts with pathways essential for podocyte integrity. Nephrol Dial Transplant 2009: 24: 630. KAPLAN B, MEIER-KRIESCHE HU. Death after graft loss: an important late study endpoint in kidney transplantation. Am J Transplant 2002: 2: 970. ROODNAT JI, MULDER PG, RISCHEN-VOS J et al. Proteinuria and death risk in the renal transplant population. Transplant Proc 2001: 33: 1170.
31. ROODNAT JI, MULDER PG, RISCHEN-VOS J et al. Proteinuria after renal transplantation affects not only graft survival but also patient survival. Transplantation 2001: 72: 438. 32. HALIMI JM, LAOUAD I, BUCHLER M et al. Early low-grade proteinuria: causes, short-term evolution and long-term consequences in renal transplantation. Am J Transplant 2005: 5: 2281. 33. FRANCO AF, MARTINI D, ABENSUR H, NORONHA IL. Proteinuria in transplant patients associated with sirolimus. Transplant Proc 2007: 39: 449. 34. PINHEIRO HS, AMARO TA, BRAGA AM, BASTOS MG. Postrapamycin proteinuria: incidence, evolution, and therapeutic handling at a single center. Transplant Proc 2006: 38: 3476. 35. LETAVERNIER E, LEGENDRE C. mToR inhibitors-induced proteinuria: mechanisms, significance, and management. Transplant Rev (Orlando) 2008: 22: 125.
79
