A single-center cohort study of kidney and kidney–pancreas recipients was conducted to evaluate the association between new immunosuppressive regimens and risk of thrombotic microangiopathy (TMA). From January 1st,1996 to December 31, 2002, 368 patients received a kidney or kidney–pancreas transplant at our center. Four immunosuppressive regimens were evaluated as potential risk factors of TMA: cyclosporin + mycophenolate mofetil (CsA + MMF), cyclosporin + sirolimus (CsA + SRL), tacrolimus + myophenolate mofetil (FK + MMF), and tacrolimus + sirolimus (FK + SRL). Thirteen patients developed biopsy-proven TMA in the absence of vascular rejection. The incidence of TMA was significantly different in the four immunosuppressive regimens studied (p < 0.001). The incidence of TMA was highest in the CsA + SRL group (20.7%). The relative risk of TMA was 16.1 [95% confidence interval (CI): 4.3–60.8] for patients in the CsA + SRL group as compared with those in the FK + MMF group. We also investigated in vitro the pathophysiological basis of this association. The CsA–SRL combination was found to be the only regimen that concomitantly displayed pro-necrotic and anti-angiogenic activities on arterial endothelial cells. We propose that this combination concurs to development of TMA through dual activities on endothelial cell death and repair.
Post-transplant thrombotic microangiopathy (TMA) is a significant and growing cause of renal allograft dysfunction and graft loss (1). The incidence of TMA is remarkably higher in transplant patients than in the general population, reflecting a clustering of risk factors in the post-transplant period (1,2). Immunologic and nonimmunologic risk factors have been implicated in the development of TMA. Acute vascular rejection in renal transplant recipients may lead to development of a microangiopathic process triggered by alloimmune reactivity toward the allograft microvasculature (1,2). Certain immunosuppressive drugs such as calcineurin inhibitors and OKT3, have been implicated in the development of nonimmune TMA (1,2). Regardless of the underlying cause, acute damage to the microvascular renal endothelium plays a central role in the initiation of microangiopathy (1,3).
TMA is characterized by endothelial cell swelling and detachment from the basement membrane in association with formation of intra-luminal platelet thrombi and partial or complete obstruction of the microvessel lumina (3,4). Acute injury and death of renal microvascular endothelial cells are pivotal events leading to formation of platelet microthrombi (3–6). Endothelial cell death (necrosis or apoptosis) is associated with release of abnormal multimers of von Willebrand factor, increased tissue factor expression, exposure of the highly thrombogenic subendothelial matrix and in turn, platelet activation and aggregation (4,7–11). Recent data suggests that impaired angiogenic activity also contributes to development of microangiopathy in animal models and human TMA (12,13).
The number of immunosuppressive agents available in clinical practice for prevention of transplant rejection has considerably increased in the last decade; yet the impact of these new agents on drug-induced TMA has not been clearly delineated. In the present study, we evaluated in a single-center cohort of kidney and kidney–pancreas patients receiving various immunosuppressive regimens, whether new immunosuppressive combinations are associated with an increased risk of nonimmune TMA. Also, we studied in endothelial cell culture systems, the impact of these agents on mechanisms of endothelial cell death and repair.
Materials and Methods
Patients and study design
A cohort study involving all kidney and kidney–pancreas recipients transplanted at Notre-Dame Hospital from January 1st, 1996 to December 31st, 2002 was conducted to evaluate risk factors for drug-induced TMA. Clinical data and biopsy results were collected using a computerized database. All patients were studied from transplantation to occurrence of TMA, graft failure, death or until December 31, 2002. The following factors were considered: age of donor and recipient, primary renal disease, number of human leukocyte antigen (HLA) mismatches, cold ischemia time, panel reactive antibodies (PRA), and cytomegalovirus (CMV) status. Immunosuppressive regimens were also evaluated as potential risk factors for TMA.
Four immunosuppressive regimens were studied: neoral + sirolimus (CsA + SRL), neoral + mycophenolate mofetil (CsA + MMF), tacrolimus + sirolimus (FK + SRL), and tacrolimus + mycophenolate mofetil (FK + MMF). Hence only patients who received one of these regimens at the time of transplantation were included in the analysis. All patients received steroids at the time of transplantation and for at least 6 months post-transplant. For patients who experienced a TMA episode, the immunosuppressive regimen received prior to and at the time of TMA onset was considered in the analysis. For patients who did not present a TMA episode over the course of the study, the immunosuppressive regimen received during the first 2 years post-transplantation was considered in the analysis. For these patients, when a change in immunosuppressive regimen occurred during the first 2 years post-transplant, the immunosuppressive regimen received for more than 12 months was considered in the analysis. Immunosuppressive regimens received after 2 years post-transplantation were not considered in the analysis (as late occurrence of TMA is exceptional and did not occur in our cohort).
Episodes of biopsy-proven TMA occurring in the absence of vascular rejection were considered in the analysis. A pathological diagnosis of TMA was made based on the presence of one or more of the following conditions: (i) presence of intraglomerular and/or arteriolar thrombi; (ii) occlusion of glomerular capillaries by amorphous material that corresponded to subendothelial accumulation of electron lucent deposits on electron microcopy; and (iii) accumulation of electron lucent subendothelial material with widening of the subendothelial space. Vascular rejection was defined according to the Banff 97 working classification of renal allograft pathology (14). TMA developing in association with acute rejection of grade IIA, IIB or III were excluded from the analysis. Hence, all cases showing evidence of TMA in association with either (a) mild-to-moderate intimal arteritis in at least one arterial cross section, (b) severe intimal arteritis comprising >25% of the luminal area or (c) transmural arteritis, were not considered in the analysis. C4D staining was not done systematically in all TMA cases and thus was not included as a diagnostic criterion.
The baseline characteristics of subjects with TMA and those without evidence of TMA were compared using Fisher's exact test or the chi-square or their nonparametric equivalents when required. The incidence of TMA for the four immunosuppressive regimens was determined and analyzed using a contingency table analysis. The association between TMA occurrence and immunosuppressive regimens was tested using the chi-square test. Relative risks of TMA were computed according to immunosuppressive regimens. The immunosuppressive regimen with the lowest incidence of TMA was considered as the reference category for relative risk computation.
Results are presented as means ± standard deviation. All p-values were two-tailed, and values of less than 0.05 were considered to indicate statistical significance. Confidence intervals (CI) were calculated at the 95% level. Analyses were performed using Statistica (StatSoft, Tulsa, OK, USA).
Culture of human endothelial cells
Human umbilical artery endothelial cells (HUAEC) were purchased from Clonetics (San Diego, CA, USA) and used at passages 2–4. The cells were seeded on gelatin-coated (1%) tissue culture plates (Costar, Cambridge, MA, USA) and cultured in endothelial growth medium (EGM-MV) (Clonetics, CA) and maintained in a humidified atmosphere containing 5% CO2/95% air at 37 °C as we described previously (15).
Screening for apoptosis and necrosis by fluorescence microscopy
For assessment of apoptosis and necrosis we used fluorescence microscopy of endothelial cells stained with Hoechst 33342 and propidium iodide as described in our previous work (15–18). We showed that this technique accurately differentiates all stages of apoptosis from primary necrosis and is more accurate for evaluation of apoptosis than TUNEL assays (17–19). In brief, HUAEC were grown to confluence in 24-well gelatinized polycarbonate culture plates (Becton-Dickinson, Franklin Lakes, NJ, USA) at 37 °C with 5% CO2/95% air. After exposure to the various experimental conditions, Hoechst 33342 [2′-(4-ethoxyphenyl)-5-(4-methyl-1-piperazinyl)-2.5′-bi-1H-benzimidazole] was added (1 μg/mL for 10 min at 37 °C). To prevent further uptake of Hoechst (HT), cells were washed with phopshate-buffered saline (PBS). Propidium iodide (PI) was added to each sample to a final concentration of 5 μg/mL, immediately before fluorescence microscopy analysis (excitation filter = 360–425 nm). The percentages of normal, apoptotic and necrotic cells adherent to the dish were estimated by an investigator blinded to the experimental conditions.
Viable cells display normal nuclear and cytoplasmic morphology and stain blue. Apoptotic cells are characterized by cell shrinkage, nuclear condensation and preservation of plasma membrane integrity. Chromatin condensation is associated with enhanced fluorescence for HT (bright blue) whereas preservation of cell membrane integrity precludes PI staining. Primary necrotic cells are characterized by increased cell size, absence of chromatin condensation and rapid disruption of cell membrane integrity associated with positive PI staining (16–18).
Assessment of proliferation using a BrdU incorporation assay
Measurement of BrdU incorporation during DNA synthesis was performed using a Cell Proliferation ELISA BrdU (colorimetric) kit (Roche Diagnostics, Mannheim, Germany) according to the protocol provided by the manufacturer and as described previously (16). BrdU incorporation was measured in subconfluent (50% confluence) HUAEC. Incorporation is expressed as: 100× (incorporation in cells exposed to experimental condition/incorporation in cells exposed control).
Evaluation of angiogenesis with branching formation assay
Angiogenesis was studied in vitro using a branching formation assay on matrigel as we described previously (16). In brief, HUAEC were seeded on a 96-well plate coated with growth factor-reduced matrigel (BD PharMingen, Mississauga, Ontario, Canada) and incubated for 16 h at 37 °C. Branching formation was evaluated at a magnification of 200×, by an investigator blinded to the experimental condition. A branch was defined as a straight cellular segment connecting two cell masses (nodes) (16).
All reagents were purchased from Sigma-Aldrich Canada Ltd (Coakville, Ontario, Canada) unless stated otherwise. Cyclosporine (CsA) was purchased from Novartis Pharma (Dorval, Quebec, Canada). Sirolimus was a kind gift of Dr Huifang Chen (University of Montreal). Tacrolimus was purchased from Fujsawa Canada Inc. (Markham, Ontario, Canada) and Mycophenolate mofetil was from Hoffman-LaRoche (Mississauga, Ontario, Canada).
Increased incidence of thrombotic microangiopathy in patients receiving a combination of cyclosporin and sirolimus
From January 1st, 1996 to December 31, 2002, 368 patients received a kidney (355 patients) or kidney–pancreas transplant (13 patients) at our centre. Three hundred and forty-nine (349) patients received one of the following immunosuppressive regimens: CsA + MMF, CsA + SRL, FK + MMF, FK + SRL. Four hundred and sixty-four biopsies were performed during this period (404 biopsies because of allograft dysfunction and 60 protocol biopsies). Thirteen patients developed biopsy-proven nonimmune TMA (11 kidney transplant recipients and 2 kidney–pancreas transplant recipients). Baseline characteristics are shown in Table 1. The age of donor, type of donor (cadaveric vs. living), primary renal disease, number of HLA mismatches, percentage of panel reactive antibodies (PRA) at the time of transplantation, duration of cold ischemia and CMV status were not significantly different in patients who developed TMA compared with those who did not (Table 1). The age of recipient was significantly lower in patients with TMA.
Table 1. Baseline characteristics
Patients with TMA (n = 13)
Patients without TMA (n = 355)
Values are mean ± SD, unless otherwise indicated. PRA, percentage of panel reactive antibodies; CMV D+/R-, proportion of cases with a CMV positive donor receiving a CMV negative allograft. *p = 0.03. For all other characteristics the p-value was >0.05.
Donor age (years)
40.9 ± 15.4
39.0 ± 25
Recipient age (years)
37.9 ± 8.3*
45.7 ± 12.6
Primary renal disease
Mean HLA mismatch
2.9 ± 1.2
2.9 ± 1.2
Mean PRA (%)
7.4 ± 17.9
12.9 ± 23.8
Cold ischemia time (h)
14.8 ± 5.9
13.3 ± 6.1
CMV D+/R- (%)
Cadaveric donor (%)
TMA occurred within the first 6 months after transplantation in 9 out of 13 patients and within the first year after transplantation in all cases but one. Median time to diagnosis was 27 days. At TMA onset, hematological abnormalities were uncommon as the triad of anemia, increased lactate dehydrogenase (LDH) serum levels and thrombocytopenia was found in only 23% (3/13) of cases (Table 2). Schizocytes were absent in all patients with biopsy-proven TMA except one. In all cases of TMA, the biopsy was performed because of increased serum creatinine whereas no protocol biopsy disclosed evidence of TMA. Proteinuria was found in a majority of TMA cases at the time of biopsy (Table 2).
Table 2. Hematological and renal abnormalities at the time of diagnosis in patients with thrombotic microangiopathy
Table 3 summarizes the immunosuppressive regimens received at the time of transplantation and at TMA onset for all patients who developed TMA and the dose of immunosuppressive agents at TMA onset. For all cases except one, the immunosuppressive regimens at transplantation and at onset of TMA were the same. One patient was transferred from FK + MMF to FK + SRL 49 days after kidney transplantation (or 156 days before onset of TMA). Therapeutic interventions after diagnosis of TMA varied: in most patients calcineurin inhibitors were decreased or changed (transfer from CsA to FK or the opposite), while four patients were transferred to a combination of SRL and MMF.
Four patients with TMA lost their graft. Patient no. 1 presented with primary nonfunction of the renal allograft and was on dialysis at the time of TMA onset (day 7 post-transplant). After diagnosis of TMA, CsA was stopped. A week later, the patient was still dialysis-dependent and a second biopsy was performed which disclosed acute rejection grade III. Patient no. 2 developed severe renal dysfunction at the time of TMA diagnosis [baseline creatinine: 1.3 mg/dL (115 μmol/L), creatinine at TMA onset 4.6 mg/dL (412 μmol/L)]. After TMA diagnosis, CsA was replaced with half-dose FK and SRL was changed for MMF. Hemodialysis was initiated in the following days and the patient was never weaned off dialysis. Patient no. 12 also presented with severe renal dysfunction at the time of TMA onset [baseline creatinine: 0.8 mg/dL (71 μmol/L), creatinine at TMA onset: 3.2 mg/dL (291 μmol/L)]. FK was stopped and replaced with SRL but nonetheless the creatinine progressively increased in the following weeks and hemodialysis was initiated. Finally, patient no. 13 also presented with significant renal dysfunction at the time of TMA diagnosis [baseline creatinine: 1.2 mg/dL (105 μmol/L), creatinine at TMA onset: 3.7 mg/dL(329 μmol/L)]. FK was stopped and SRL initiated. Renal function improved with a creatinine at 1.4 mg/dL (128 μmol/L) 1 month post-TMA which remained stable for a year. Finally, as serum creatinine increased again progressively, a renal biopsy was repeated which disclosed focal segmental glomerulosclerosis. Renal function deteriorated during the following year and the patient went back to hemodialysis.
We evaluated the association between immunosuppressive regimens received during the first year after transplantation and risk of TMA. Nineteen out of 368 patients received an immunosuppressive regimen at the time of transplantation that consisted of steroids and a calcineurin inhibitor in the absence of an antiproliferative agent (MMF or SRL) and were excluded from the analysis. None of these patients developed TMA. Table 4 shows the incidence of TMA according to immunosuppressive regimens in the remaining 349 patients. The incidence of TMA was significantly different in the four immunosuppressive regimens studied (p < 0.001). The incidence of TMA was highest in the CsA + SRL group (20.7%) whereas it was significantly lower for the other immunosuppressive regimens (3.7% for CsA + MMF, 6.1% for FK + SRL and 1.3% for FK + MMF). The relative risk of TMA was 16.1 (95% CI: 4.3–60.8) for patients in the CsA + SRL group as compared with those in the FK + MMF group. The immunosuppressive regimen with the lowest incidence of TMA (FK + MMF) was selected as the reference category. There was no other statistically significant difference in TMA occurrence between the other immunosuppressive regimens. We also performed an analysis excluding patients who experienced a change in immunosuppressive regimen and the difference in incidence of TMA in patients receiving the CsA + SRL combination compared with other immunosuppressive regimens remained statistically significant (data not shown).
Table 4. Risk of TMA associated with the various immunosuppressive combinations
Incidence no./total no.
Relative risk (95% CI)
CI denotes confidence interval.
The relative risk is for each immunosuppressive regimen as compared with the FK + MMF group.
CsA + SRL
CsA + MMF
FK + SRL
FK + MMF
Mechanisms of endothelial cell injury after exposure to a combination of cyclosporin and sirolimus
To delineate the potential mechanisms of endothelial damage associated with the combination of CsA and SRL, we characterized the impact of these agents on endothelial cell death, proliferation and angiogenesis in vitro. We first evaluated whether SRL induces endothelial cell death or potentiated pro-death activity of other immunosuppressive agents. Development of apoptosis and necrosis in HUAEC exposed to CsA, FK, MMF, SRL and combinations of SRL with either CsA or FK was evaluated by fluorescence microscopy with Hoechst 33342 and PI staining. HUAEC exposed to clinically relevant concentrations of SRL for 24 h did not show evidence of apoptosis or necrosis (Figure 1A). Also, exposure to FK or MMF for 24 h did not increase development of apoptosis or necrosis in HUAEC (Figure 1A). CsA was the only immunosuppressive agent associated with a significant pro-necrotic activity on arterial endothelial cells at clinically relevant concentrations. Concomitant exposure to CsA and SRL did not significantly increase development of endothelial cell death in HUAEC as compared with CsA alone (Figure 1A), suggesting that SRL alone or in combination with CsA does not enhance endothelial cell death.
We then evaluated whether the CsA + SRL combination may lead to enhanced endothelial injury by concomitant activities on endothelial cell death and repair. Endothelial proliferation was evaluated in HUAEC exposed to CsA (1 μg/mL), SRL (10 ng/mL), MMF (10 ng/mL) or a combination of CsA and SRL for 24 h. SRL, either alone or in combination with CsA, significantly inhibited endothelial proliferation whereas exposure to CsA or MMF did not (Figure 1B). Using a Matrigel branching formation assay we also evaluated the impact of these immunosuppressive molecules on angiogenesis. CsA and MMF did not significantly inhibit branching formation. However, SRL significantly inhibited angiogenesis and a combination of SRL and CsA led to a very robust inhibition of angiogenic activity. (Figure 1C). Hence these results suggest that the combination of CsA and SRL is the only immunosuppressive combination that targets concomitantly the molecular control of endothelial cell death and repair.
The number of immunosuppressive agents available in clinical practice increased dramatically in the last decade. This was associated with an important decrease in the incidence of acute rejection in renal transplant recipients (20). Yet other causes of renal allograft dysfunction in the early post-transplant period are being increasingly recognized. Post-transplant TMA affects a growing number of solid organ and bone marrow transplant recipients worldwide (1). We conducted a cohort study to evaluate whether new immunosuppressive regimens may be associated with an increased risk of TMA. Also we used an in vitro system to delineate the mechanisms of endothelial injury that underlie the association between new immunosuppressive regimens and TMA.
We found that the incidence of TMA is significantly higher in patients receiving a combination of CsA and SRL (CsA + SRL group: 20.7%) compared with the other combinations of immunosuppressive agents (3.7% for CsA + MMF, 6.1% for FK + SRL and 1.3% for FK + MMF). Hence, the relative risk of TMA was 16.1 (95% CI: 4.3–60.8) for patients in the CsA + SRL group as compared with those in the FK + MMF group. Thus, among all immunosuppressive combinations, the risk of TMA was highest in patients receiving a combination of CsA and SRL. Hence, although SRL in itself was not associated with significant nephrotoxicity, our results suggest that the combination of SRL and CsA exerts synergistic effects on renal endothelial injury.
Recent case reports and case series of renal and bone marrow transplant recipients also suggest an increased risk of TMA in patients receiving SRL (21–24). However, an earlier paper by Langer et al. alluded to a possible protective effect of SRL on development of hemolytic–uremic syndrome (HUS) in renal transplant recipients treated with CsA-based immunosuppression (25). Various factors may account for the apparent discrepancy between this report and our results. First, we focused on drug-induced TMA and thus excluded all cases of TMA that developed in association with acute vascular rejection. In the report of Langer et al. diagnosis of TMA was based on clinical features (25) and thus likely included immune-mediated TMA. Hence, it is possible that SRL, through its potent anti-proliferative activity on lymphocytes, may prevent immune-mediated TMA. In cases of drug-induced TMA, however, inhibition of endothelial repair by SRL could accentuate the severity of renal endothelial injury. Second, in the series by Langer et al. all patients received a CsA/SRL/steroid regimen. Hence, the conclusion that SRL may afford a decreased risk of HUS largely relied on comparison with historical controls (25) whereas we studied head-to-head the risk of TMA associated with the various combinations of immunosuppressive agents in a single cohort of patients.
We evaluated in an endothelial cell culture system potential mechanisms that may underlie the increased incidence of TMA observed with the CsA + SRL combination. Damage to renal microvascular endothelial cells is a central pathophysiological factor implicated in formation of platelet microthrombi within the renal microvasculature and development of TMA (3–6). Necrosis and more recently apoptosis of endothelial cells have been implicated in the mechanisms of endothelial damage associated with TMA (4,7–11,26,27). We showed recently that CsA favors development of a necrotic type of programmed cell death in arterial endothelial cells regulated by increased production of reactive oxygen species and secondary lysosomal damage (15). This pro-necrotic activity is not dependent on calcineurin inhibition, as FK at equipotent concentrations, did not induce necrosis of endothelial cells (15). We found that SRL, at clinically relevant concentrations, does not induce apoptotic or necrotic forms of endothelial cell death in vitro and does not potentiate the pro-necrotic activity of CsA on arterial endothelial cells. Hence, increased endothelial injury associated with the CsA–SRL combination cannot be attributed to a synergistic effect of SRL and CsA on endothelial cell death.
Inhibition of endothelial repair by SRL was evaluated as another possible mechanism for potentiation of endothelial injury by the combination of CsA and SRL. Endothelial proliferation and angiogenic activity are important factors for vascular repair and containment of the microangiopathic response. Pro-angiogenic molecules, such as VEGF, have been shown recently to hasten renal recovery in an animal models of TMA (12,13). We found that proliferative rates and angiogenic activity were maintained in endothelial cells exposed to CsA alone, but almost completely abolished by concomitant exposure to clinically relevant concentrations of SRL and CsA. In keeping with our clinical data showing a lower risk of TMA in patients receiving the CsA + MMF combination compared with the CsA + SRL combination, we found that angiogenic activity was not significantly attenuated by MMF in vitro at equipotent concentrations. Hence these results suggest that the CsA + SRL combination is the only immunosuppressive regimen that concomitantly targets the molecular control of cell death and repair at the endothelial level.
In summary, we found that risk of TMA is significantly increased in patients receiving a combination of CsA and SRL. As this combination is the only immunosuppressive regimen that affects concomitantly endothelial viability and angiogenesis, we propose that the increased risk of TMA in this setting may be explained, at least in part, by these dual activities, favoring development of endothelial cell death while inhibiting endothelial repair.
This work was supported by an operating grant to MJH from the Canadian Institutes of Health Research (grant no. mt-15447). MJH is a scholar of the Canadian Institutes of Health Research. We thank the Fondation J.-L. Lévesques and the Fondation CHUM for their support.