- Top of page
- Materials and Methods
Thrombotic microangiopathy (TMA) and hemolytic uremic syndrome (HUS) represent serious threats to kidney allograft recipients.
During a 4-year period, among 850 kidney transplantations, seven recipients with primary HUS and seven recipients (eight transplants) with previous or de novo TMA/HUS were identified and given calcineurin inhibitor (CNI)-free immunosuppression by sirolimus (SRL), mycophenolate mofetil and steroids.
Thirteen out of 15 transplantations were successful in the long term; resulting in a mean creatinine of 101 μmol/L (16.4 months follow-up). In patients maintained on CNI-free regimen, no TMA/HUS recurrences were observed. A high rate of acute rejections (53%) may indicate insufficient immunosuppressive power and/or a causative relationship between TMA/HUS and rejection. Wound-related complications were abundant (60%), and call for surgical/immunosuppressive countermeasures.
Our experience supports the idea that CNI's are major offenders in TMA/HUS induction. Total CNI elimination seems essential, as the nephrotoxic combination CNI + SRL may promote TMA. Features of TMA/HUS should be carefully explored in recurrent 'high responders'.
- Top of page
- Materials and Methods
Thrombotic thrombocytopenic purpura (TTP) and hemolytic-uremic syndrome (HUS), are closely related disorders (1). When the dominant clinical feature is renal failure, the condition is most often designated as HUS. Furthermore, when the syndrome appears after renal transplantation and affects the graft, it is useful to distinguish between recurrent HUS, appearing in recipients with HUS as their primary renal disease, and de novo HUS. The histopathologic hallmark of HUS in graft biopsies is thrombotic microangiopathy (TMA) (2,3). In recent years, TMA has been introduced as a clinical entity, also encompassing recipients with minimal thrombocytopenia and minimal hemolysis/anemia (4). Localized, renal TMA, without the systemic manifestations of full-blown HUS probably represents a low-grade manifestation, sharing similar pathogenetic factors. On this background, the conception TMA/HUS is justified, and will be conveniently used in this paper.
Some progress in the basic understanding of TMA/HUS/TTP has been made, by characterizing the various entities at the molecular level. A genetic and/or acquired deficiency of the von Willebrand factor-cleaving protease (vWf-CP) has in recent years been found to distinguish TTP, while in HUS normal vWf-CP activity is found (5,6). The specific protease deficiency is responsible for the accumulation of large vWf multimers, which in turn causes intravascular thrombi formation through platelet aggregation (7). By investigating a genetic form of TTP, the responsible ADAMTS13 metalloprotease gene has recently been identified (8). Furthermore, IgG antibodies directed against vWf-CP have been detected in some TTP patients (6).
Primary HUS encompasses two distinct entities (1). In diarrhea-associated HUS (D-HUS), Shiga-toxin or verotoxin has been found to be directly responsible for endothelial injury. The etiology of atypical HUS (non-diarrheal; A-HUS) is still obscure. In a fraction of these patients, complement factor H deficiency, caused by germline mutations in the FH1 gene, has been demonstrated (9). Intriguingly, complement factor H deficiency has been suggested also to favor the development of mesangiocapillary glomerulonephritis type II (10) and acute allograft glomerulopathy (AAG) (11), a distinct form of allograft rejection, involving T-cell-mediated endothelial injury.
Calcineurin inhibitors (CNI) are considered to be major offenders in TMA/HUS induction post-transplant (2,3,12,13). There have been numerous reports of HUS induction by Cyclosporin A (CsA) and tacrolimus, and reversibility has in many cases been demonstrated with discontinuation of the drug. Furthermore, CNI-induced TMA/HUS is not restricted to renal allograft recipients. It also occurs in patients receiving CNI due to other organ transplantation or autoimmune diseases (14). The mechanisms of TMA/HUS induction are poorly understood; endothelial injury and increased platelet aggregation have been suggested (12). The therapeutic measures against HUS have traditionally been CNI reduction/withdrawal and plasma exchange. However, there have also been reports on successful switching from CsA to tacrolimus (3). In recent years, the introduction of new immunosuppressive agents has made CNI-free regimens possible. In one study, an immunosuppressive protocol consisting of mycophenylate mofetil (MMF), corticosteroids (CS) and antilymphocyte antibody induction therapy was administered to six patients, three of whom were undergoing retransplantation due to cyclosporine-associated recurrent HUS (15; abstract). At follow-up no patients had experienced recurrent disease, allograft loss or an episode of acute rejection (AR).
We will hereby report our initial (single center) experience during 4 years, using CNI-free immunosuppression based on sirolimus (SRL) combined with MMF and steroids, in patients with primary/recurrent TMA/HUS and de novo TMA/HUS.
Materials and Methods
- Top of page
- Materials and Methods
At our department in Oslo, the single national transplant center of Norway, 200–250 kidney transplantations are performed each year. Our standard immunosuppressive regimen has during recent years consisted of CsA, MMF and CS. CsA has been monitored by C2 concentrations for the first 3 months and thereafter by C0 (initial dose 10 mg/kg/day; C2-target concentration 1500–2000 μg/L for the first month, 1400–1600 μg/L for month 2, 1000–1200 μg/L for month 3; thereafter C0-target 75–200 μg/L). The initial MMF dose has been 2 g/day, with individual dose reductions according to adverse effects, but without regard to C0 concentrations of mycophenolic acid (MPA), even though these have been routinely measured. CS have been tapered from prednisolon 80 mg/day initially, to 20 mg/day at 1 week post-Tx, 10 mg/day after 3 months and then further tapering toward 5 mg/day.
During 4 years from September 2000 to September 2004, all kidney recipients with primary HUS and all recipients with present or previous de novo TMA/HUS have been registered and followed in a prospective manner. In recipients with primary HUS or previous post-transplant TMA/HUS, the intention has been to give CNI-free immunosuppression from the time of transplantation. However, in a few cases during the first part of the series, the previous history was first acknowledged after some days and the immunosuppression converted then. Recipients with de novo HUS, recognized in the post-transplant period, have been treated by immediate CNI elimination and replacement by SRL.
The CNI-free regimen has in all cases consisted of SRL (loading dose 12 mg for 1–2 days, thereafter 6 mg/day; aiming at C0-target concentrations of 10–15 μg/L for the first 3 months; thereafter 5–10 μg/L) combined with MMF (2 g/day; according to standard protocol; not guided by C0-MPA) and CS (according to standard protocol). The conversion protocol has involved immediate withdrawal of CsA, full loading dose (12 mg) of SRL the next day, thereafter 6 mg/day and continuation of MMF/CS according to standard protocol.
All kidney recipients have had close follow-up (three times a week) at our single national center for 3 months post-transplant. Thereafter, follow-up has been transferred to nephrologists at the local hospitals. All surgical complications have exclusively been taken care of at our department.
Our main result parameters (‘end points’) have been ‘TMA/HUS-recurrence rate’, ‘kidney function’ evaluated by ‘glomerular filtration rate’ (GFR; by Cr-EDTA clearance) at 10–12 weeks post-Tx and ‘creatinine’ at 10–12 weeks/final follow-up and ‘graft survival’. The ‘AR rate’ and ‘rate of surgical complications’ have been considered secondary‘end points’.
Scheduled biopsies have not been performed.
- Top of page
- Materials and Methods
During these 4 years (September 2000–September 2004), we have performed 850 kidney transplantations, 38% from living donor and 62% from cadaveric donor.
Seven recipients with primary/recurrent HUS (Table 1; patients 1–7) and seven recipients with de novo TMA/HUS (Table 2; patients 8–14) have been included.
Table 1. Individual history and post-transplant events of recipients with HUS as primary renal disease
|Patient-Age at present Tx||Primary disease||Previous history||Present Tx/ Immunosuppression||Risk factors||Post-Tx events/Functional outcome|
|1. female 19 years||Primary HUS||HUS from 1983||Tx #1 Febrauary 2001|| ||Creatinine 87|
|(D-HUS (?))||Dialysis 3 months pre-Tx||CsA SRL day 1|| |
|2. male 25 years||Primary HUS||Tx #1 1984–Graft loss 1998;||Tx #2 May 2001||Chr-CS||Creatinine 94|
|(A-HUS)||Recurrent HUS/Chronic allograft nephropathy||CsA SRL day 1|| |
|3. female 35 years||CGN/Primary HUS|| ||Tx #1 Febrauary 2002||AR||AR; CS|
|(A-HUS)||Dialysis 11months pre-Tx||SRL from day 0|| ||Lymphocele; drainage Creatinine 140|
|4. female 59 years||Primary HUS|| ||Tx #1 April 2002||AR||AR; CS|
|(A-HUS)||Dialysis 10 months pre-Tx||SRL from day 0|| ||Lymphocele; reoperation Creatinine 116|
|5. female 54 years||Primary HUS||Tx #1 1998–Graft loss 2000;||Tx #2 May 2002||Chr-CS||AR; CS + PE|
|(A-HUS)||Chronic allograft nephropathy/Recurrent HUS||SRL from day 0||AR||Wound dehis./seroma; reoperation|
|(CyA due to psoriasis)|| ||Wound sinus; reoperation Creatinine 71|
|6. female 20 years||Primary HUS||Dialysis 11 months pre-Tx||Tx #1 September 2002||BMI 25.6||AR; ATG|
|(A-HUS)|| ||SRL from day 0||AR||Creatinine 77|
|7. male 53 years||Primary HUS||Dialysis 11 months pre-Tx||Tx #1 February. 2003 SRL from day 0||BMI 26.5||Creatinine 96|
All previous transplantations of these recipients were performed with CsA-based immunosuppression.
One patient (Table 2; patient 10) has been transplanted twice during the study period (Tx #2 and #3). Tx #2 ended in early graft loss due to AR and possible TMA/HUS. Patient 11 lost her graft by fulminant HUS 30 months post-Tx, and it is noteworthy that her immunosuppression had been supplemented by CNI/tacrolimus (TAC) from week 8, due to recurrent rejection episodes. Thereafter, she was maintained on quadruple regimen with SIR, MMF, CS and TAC.
All the other 13 transplantations have been considered successful, based on a mean follow-up period of 16.4 months, ranging from 6 to 47 months. Among these 13 kidneys there has not been observed any functional deterioration due to TMA/HUS, resulting in a mean (measured) GFR of 58 mL/min/1.73 m2 at 10–12 weeks post-Tx and a mean creatinine of 101 μmol/L at final follow-up (Table 3). About two anti-hypertensive drugs per patient were needed at 10–12 weeks post-Tx. In only one patient (Table 2; patient 9) a significant proteinuria (>0.5 g protein/24 h) was observed.
Table 3. Summarizing statistics regarding BMI, acute rejections (AR), SRL-concentrations, surgical complications, creatinine/measured GFR/antihypertensive drugs at 10–12 weeks post-Tx and creatinine at final follow-up
| ||All TMA/HUS 15 Tx/14 patients||Primary HUS 7 Tx/7 patients||De novo TMA/HUS 8 Tx/7 patients|
|BMI (kg/m2); mean [range)]||22.6 [17.7–27.2]||21.9 [17.7–26.5]||23.4 [18.9–27.2]|
|Acute rejections||53% (8 Tx)||57% (4 Tx)||50% (4 Tx)|
|SRL-concentrations day 0–30 post-Tx||9.8 [4.3–22.9]||11.3 [5.1–22.9]||7.9 [4.3–11.4]|
|(Mean [range] of median per patient; μg/L)||+ AR:10.0|| |
| In patients with/without AR||÷ AR:9.4|| |
|Surgical complications (% (number) of Tx)||60% (9)||43% (3)||75% (6)|
| Lymphocele||47% (7)||29% (2)||71% (5)|
| Wound dehiscence||20% (3)||14% (1)||25% (2)|
| Hernia||29% (2)||0||25% (2)|
| Others||20% (3)||14% (1)||25% (2)|
|Plasma creatinine 10–12 weeks post-Tx (μmol/L)||105 [79–153]||104 [83–147]||107 [79–153]|
|Measured GFR 10–12 weeks post-Tx||58 [29–86]||59 [36–81]||57 [29–86]|
|(Cr-EDTA-Clearance; mean [range]; mL/min/1.73 m2)|| |
|Antihypertensive drugs 10–12 weeks post-Tx||1.9 [0–5]||2.1 [1–4]||1.6 [0–5]|
|(Number; mean [range])|| |
|Plasma creatinine at final follow-up 16.4 [6–47] months (μmol/L)||101 [71–140]||97 [71–140]||105 [79–130]|
Patient 14 received a cadaveric kidney with distinctive TMA in baseline biopsies and inferior/delayed primary function. This obviously represents a very special case with TMA/HUS having evolved in the donor, possibly during the final disease.
There were eight AR episodes (53% of Tx), all verified by biopsy (Tables 1–3). No significant correlations were found versus SRL-concentrations during the first month post-transplant.
The surgical complications consisted of seven lymphoceles requiring drainage/reoperation (47%), three cases of wound dehiscence (20%), two patients with incisional hernia (three reoperations), as well as three reoperations due to hematoma, urinary leakage and wound sinus. Altogether, nine patients (60% of Tx) suffered surgical complications, requiring reoperation/intervention.
In Tables 1 and 2 risk factors for surgical complications have been indicated for each patient: body mass index (BMI), diabetes mellitus (DM), previous chronic use of CS (Chr-CS) and AR after the present Tx. These TMA/HUS patients were nonobese (Table 3). However, a 'cutoff' at BMI above 25 has been applied as a risk factor. Patients suffering surgical complications had an average of 1.6 risk factors (14 in 9 patients) versus 1.0 (5 in 5 patients) without such complications.
- Top of page
- Materials and Methods
HUS is considered a serious post-transplant complication. In patients with HUS as primary renal disease, the rate of recurrent TMA/HUS after transplantation is reported as high as 25–50% (16,17). TMA has been reported to occur as frequent as up to 14% of recipients on CNI-based immunosuppression (3). These figures may, however, represent an overestimate, since the vascular histopathological changes are hard to distinguish from vascular rejection and CNI toxicity. Among our primary HUS patients, only one (Table 1; patient 1) seems to have had D-HUS, which is associated with a low rate of recurrence. The other six patients must be classified as A-HUS, which according to the literature (17) confers a distinctly unfavorable prognosis with regard to renal recurrence. On this background, our results with primary HUS recipients (Table 1; patients 1–7) on SRL-based immunosuppression must be considered interesting, as no recurrences were detected, and the functional outcome was excellent (creatinine 71–140 μmol/L).
Some patients with an underlying renal disease different from HUS, may develop de novo TMA/HUS after transplantation. The incidence of the syndrome is considered to be low (2), and our series is consistent with this view. Only seven patients (<1%) with de novo TMA/HUS (Table 2; patients 8–14) in previous or present grafts were identified among 850 recipients during the study period. Three of these de novo recipients (patients 8–10) were considered particularly difficult cases with previous early graft losses; so-called 'high responders'. Hereditary factors, that could be attributed to polymorphisms in the ADAMTS13 gene, FH1 gene or other related genes, may make certain patients predisposed to the development of TMA/HUS. The risk of HUS induction may also be influenced by the serum level of inhibitors, which in turn could be defined by environmental (CNI ?) and hereditary factors. Accordingly, deficiency of vWf-CP and the presence of ADAMTS13 inhibitors were demonstrated in a renal allograft recipient with TMA on CsA-based immunosuppression (18). Discontinuation of CsA and plasma exchange restored the ADAMTS13 activity, followed by resolution of TMA and improved graft function. In future studies on TMA/HUS, molecular analysis on ADAMTS13 and factor H activity should be included.
In patients 9, 10 and 12, the diagnosis of TMA/HUS was partly based on a retrospective evaluation of the previous transplantations with graft loss. Clear indications of HUS were found, including thrombocytopenia, elevated LDH and recurring need for transfusions. Pathologic reevaluation of biopsy specimens showed evident TMA (Figure 1). The primary missing of TMA/HUS diagnosis in these three cases may indicate that the de novo syndrome often escapes diagnostic attention, perhaps in part because the histological findings are hard to distinguish from CNI toxicity and rejection.
Figure 1. Light micrographs of graft biopsies, showing TMA. (All stained with hematoxylin, eosin and safran.) (A) Patient 5; Recurrence of primary TMA/HUS (Tx #1). Fibrin and fragmented red blood cells in arteriolar lumina and vessel walls (arrows) with thickened intima. (B) Patient 9; Recurrence of de novo TMA/HUS (Tx #3). Glomeruli with capillary thrombi (arrows). (C) Patient 13; De novo TMA, combined with (C4d positive) vascular rejection (Tx #1). Glomeruli with capillary thrombi (arrow). (D) Patient 14; TMA in baseline biopsy (Tx #1). Glomeruli with capillary thrombi (arrows). Acute tubular necrosis (left hand side).
Download figure to PowerPoint
A combination of TMA/HUS and AR in the same transplant at the same time has been recognized in four patients (8, 10, 12 and 13). In patient 10, the clinical and pathological picture during transplantations #2 and #3 was dominated by a severe vascular rejection, which after Tx #2 was highly developed at the time of conversion to SRL, and led to graft loss 2 weeks post-transplant. At Tx #1 he had (by retrospective evaluation) developed a severe, full-blown picture of HUS (+ rejection ?); with severe haemolytic anemia, need for transfusions and early graft loss. Systemic TMA/full-blown HUS has been shown to predict a worse graft prognosis than localized TMA (4). The second graft loss of our series (patient 11) is consistent with this notion, as the graft deterioration 30 months post-transplant coincided with fulminant HUS. Intriguingly, in this particular patient, the recurrence of HUS may have been induced by CNI, as her immunosuppression had been supplemented by TAC from week 8.
The high rate of AR (Table 3) may indicate insufficient immunosuppressive power and/or suggests a causative relationship between TMA/HUS and rejection. It should be pointed out that the 'background' AR rate on our standard immunosuppressive regimen (CsA (C2-monitored; high dose) + MMF (2 g/day) + CS) is about 40%, which is high compared with other centers using exactly the same protocol. We believe this is partly due to a very close follow-up during the first 2–3 months, and a low threshold for performing biopsy (average of two biopsies per patient during the first 10 weeks); thereby detecting borderline/'self-limiting' rejection activity. The histological cutoff point for defining rejection may be rather exact; however, the threshold for performing biopsy is not.
In Europe, there has been an attitude toward accepting somewhat higher AR rates, to avoid induction theraphy by Thymoglobulin-/Anti-thymocyte globulin on a routine basis. However, many centers have introduced interleukin-2-receptor-inhibitors (IL-2-RI), and based on the present AR rate, basiliximab/daclizumab may be warranted in addition to the present SIR + MMF + CS regimen. We are presently running a study on recipients of marginal donor kidneys (Wyeth 0468-101466) using this quadruple protocol in the trial arm.
The possible pathogenetic relation between TMA/HUS and rejection remains to be explored. Interestingly, factor H deficiency may turn out to be a common pathogenetic factor, as also involved in the AAG type of rejection (11). An exciting future project would be to analyze 'high responders' for ADAMTS13 and factor H activity (including inhibitor levels). The idea that some of the 'high responders' may hide a combination of TMA/HUS and rejection, represents an exciting hypothesis.
The very high rate of lymphoceles (7/15), wound dehiscence (3/15) and hernia (2/15) is clearly related to the potent antiproliferative action of SRL, and has previously been reported in both SRL (19) and everolimus (EVE) studies. The specific combination with MMF (also exerting antiproliferative actions) and CS may be particularly unfavorable with regard to wound healing.
Measures to counteract these frequent and undesirable side effects are being considered. We are presently planning a study on performing prophylactic fenestration during the Tx procedure. Furthermore, by using nonabsorbable suture for musculofascial closure the incidence of wound dehiscence and hernia may be lowered. However, many of these cases are due to tissue friability, not suture disruption. Our data do underscore the importance of the established risk factors for wound healing problems, which were predominant in those of our patients suffering surgical complications (Tables 1 and 2). Furthermore, wound dehiscence and hernia did only occur in retransplanted patients, subjected to long-lasting CS theraphy. CS rejection therapy may be reduced by intensifying the basic immunosuppression, preferably by adding an IL-2-RI. And by using a more potent combination, a lowering of SRL dose/target concentration may be attempted, in order to reduce dose-dependent adverse effects.
SRL (and EVE) has so far predominantly been used in addition to CNI, even though this combination has been proven to be particularly nephrotoxic (19). It seems like SRL/EVE potentiates the nephrotoxic effects of CNI at the pharmacodynamics level. The mechanism may very well involve increased CNI-induced microangiopathic/endothelial injury, enhancing platelet aggregation (12) and thereby possibly promoting TMA. This may explain why recent studies indeed have identified SRL as a risk factor for TMA (20). In our series, the best results were obtained with the primary HUS patients (Table 1; patients 1–7), in whom CNI mostly were avoided from the start. We think these considerations advocate the use of SRL/EVE without CNI, and support our strategy of CNI elimination in the present series.
This study was open, uncontrolled and exploratory by design. Due to the low incidence and heterogenous nature of TMA/HUS, a controlled/randomized study was not attempted here, and has not been launched elsewhere. However, based on the promising results regarding TMA/HUS recurrence, we have decided to employ such a 'tailored' SRL-based/CNI-free immunosuppressive regimen, in all renal allograft recipients with primary, previous and de novo TMA/HUS. However, in the future we will add an IL-2-RI to the present regimen.
When investigating graft failures, the features of HUS should be kept in mind, and the pathologist should be alerted to look for TMA. And, when dealing with 'high responders', requiring multiple transplantations, the option of using a CNI-free regimen should be considered even though exact evidence toward TMA/HUS is missing.