Eculizumab (anti-C5) has been sporadically reported as an efficient therapy for atypical hemolytic uremic syndrome (aHUS). However, the lack of series precludes any firm conclusion about the optimal use of anti-C5 for preventing or treating aHUS posttransplant aHUS recurrence. We thoroughly studied 22 renal transplant recipients with aHUS who received off-label therapy with anti-C5, including 12 cases, which have not been reported yet. Nine patients, all carrying a complement genetic abnormality associated with a high risk of aHUS recurrence, received prophylactic anti-C5 therapy to prevent posttransplant recurrence. Eight of them had a successful recurrence-free posttransplant course and achieved a satisfactory graft function, while the remaining patient experienced early arterial thrombosis of the graft. Thirteen renal transplant recipients were given anti-C5 for posttransplant aHUS recurrence. A complete reversal of aHUS activity was obtained in all of them. Importantly, the delay of anti-C5 initiation after the onset of the aHUS episode inversely correlated with the degree of renal function improvement. Three patients in whom anti-C5 was subsequently stopped experienced a relapse. Altogether these data suggest that long-term eculizumab is highly effective for preventing and treating posttransplant aHUS recurrence. Our study also indicates that anti-C5 should be promptly started if a recurrence occurs.
The interest in developing complement inhibitors has stemmed from the growing implication of complement activation in many pathological conditions, including atypical hemolytic uremic syndrome (aHUS) (1,2). Genetically determined or acquired uncontrolled activation of the complement alternative pathway has been found in up to 70% of patients with aHUS, in whom this constitutes a strong susceptibility factor for thrombotic microangiopathic processes triggered by endothelial cell damage (1,2). Eculizumab (anticomplement component 5 (C5)), a recombinant, fully humanized monoclonal antibody, is the first complement inhibitor licensed for clinical use (3). By specifically targeting the human C5, anti-C5 inhibits the generation of the proinflammatory mediator C5a and prevents the formation of the membrane attack complex (3). Anti-C5 has become a breakthrough therapy for paroxysmal nocturnal hemoglobinuria, a chronic and disabling disease, whose symptoms are dominated by complement-dependent hemolytic anemia and related vascular thrombosis (4).
Since 2009, anti-C5 has been successfully used in off-trial patients with aHUS, involving either native or transplanted kidneys (5–14). Four recent reports also indicated that anti-C5 could be efficiently used as prophylaxis to prevent posttransplant aHUS recurrence in patients in whom a high recurrence risk could be predicted from identified genetic abnormalities (15–18). In addition, anti-C5 was used with success in two cases of de novo posttransplant aHUS, occurring in simultaneous pancreas kidney transplant recipients (19,20). This is reminiscent of the finding that complement dysregulation may contribute to the pathogenesis of de novo posttransplant HUS as well (21). This preliminary experience based on anti-C5 off-label use was in line with the early results of two prospective phase II trials, reported in 2010 and 2011 at the congresses of the American Societies of Nephrology and Hematology, highlighting the safety profile and efficacy of anti-C5 in treating aHUS. These studies enrolled 37 patients with aHUS involving native or transplant kidneys. These patients had a plasma-resistant aHUS (n = 17) (22–24) or were on chronic plasma therapy (n = 20) (25–27). Anti-C5 succeeded in controlling the disease in most cases, as evidenced by normalization of hemolytic markers and thrombocytopenia and improvement of renal function (22–27).
However, accurate information regarding the use of anti-C5 in posttransplant settings, according to individualized recurrence risk assessment, is critically lacking. In an attempt to provide clues to optimize the chance for recovery of renal function with anti-C5 treatment and to draw the first lessons from the still limited experience in the settings of renal transplantation, we thoroughly studied 22 renal transplant recipients with aHUS who had received off-label anti-C5 therapy.
Patients and Methods
This was a retrospective multicenter study describing renal transplants recipients, who were given anti-C5 (eculizumab Soliris®, Alexion Pharmaceuticals, Cheshire, CT, USA) for preventing or treating posttransplant aHUS recurrence. French adult and pediatric renal transplant centers were interviewed to know whether any of the patients followed in their institutions had been provided anti-C5 for preventing or treating posttransplant aHUS recurrence. Eleven patients were identified including two cases, which have already been reported (6,18,28). In addition, all authors having reported the use of eculizumab in the prevention or treatment of aHUS recurrence in peer-reviewed journals (n = 8) (5,7,9,11,13,15–17), and congress abstracts (n = 3) (29–31) were contacted by C. Loirat and invited to share their experience. Altogether, 13 patients, identified in nine centers, including seven in France, one in Italy and one in Germany, were included in the study. Clinical and genetic data were obtained from medical records. Because the study population size was limited, we performed a pooled analysis of data from these 13 cases and from 9 other cases, already reported as case reports. When corresponding authors agreed to provide updated data, regarding late outcome of their patients, this was mentioned in Tables 1 and 2 as (personal communication, name of the corresponding author, date).
Table 1. Renal transplant recipients treated preemptively with anti-C5 to prevent aHUS recurrence
Age at onset of HUS previous renal transplantation
Type of donation
Previous prophylactic plasma therapy
Age at day 0* (yrs)
Therapeutic scheme of eculizumab
Response to aC5 s. creat (at last follow-up under aC5)
aC5 = anticomplement component 5 antibody; aHUS = atypical hemolytic uremic syndrome; mo = months; ND = not discontinued; PC = personal communication; PE = plasma exchange; RT = renal transplantation; TCC = terminal complement complex; wk = week; yr = year.
Started at the time of renal transplantation not discontinued
Absence of recurrence. s. creat at 48 μmol/L (15 mo) Mixed rejection at 3 months posttransplant Biopsy-proven resolution at 15 months posttransplant CH50 was below detection level and subsequently rose to 12–19% after the doses had been spaced out to every 3 weeks
10-mo-old no previous RT
Started at the time of renal transplantation Not discontinued
Absence of recurrence. s. creat at 44 μmol/L (4.5 months) CH50 below detection level
Krid & Niaudet PC
1-yr-old no previous RT
Started at the time of renal transplantation not discontinued
Absence of recurrence. s. creat at 58 μmol/L CH50 below detection level upon every two weeks maintenance regimen (4 months)
3-yr-old 1 previous graft loss due to recurrent aHUS Relapsing aHUS during the dialysis period
1 PE session preceded the first aC5 dose Prophylactic aC5
Started at the time of renal transplantation
Early arterial thrombosis of the graft at day 1. Transplant nephrectomy at day 3.
Zuber & Legendre PC
33-yr-old 2 previous graft losses due to recurrent aHUS
Started at the time of renal transplantation not discontinued
Absence of recurrence s. creat at 176 μmol/l Mixed rejection at 6 weeks posttransplant. Biopsy-proven resolution 15 days later CH50 below detection level (2 months)
Table 2. Renal transplant recipients treated with anti-C5 to treat aHUS recurrence
Age at onset of aHUS previous renal transplantation
TI between RT and aHUS recurrence
Response to plasma of the current episode
TI between episode onset and aC5
Age [s. creat] at day 0* (μmol/L)
Therapeutic scheme of aC5
Response to aC5 s. creat (at last f/u under aC5)
aC5 = anti-complement component 5 antibody; aHUS = atypical hemolytic uremic syndrome; ARF = acute renal failure; AwRT = awaiting renal transplantation; BP-TMA = biopsy-proven TMA; CFH = complement factor H; CFI = complement factor I; C3 = complement component 3; CNS = central nervous system; CR = complete remission; Hb = hemoglobin (g/dL); HD = hemodialysis; HPN = hypertension; mo = month; ND = not discontinued; NI = not identified; NS = not specified; PC = personal communication; PE = plasma exchange; Prot = proteinuria; RT = renal transplantation; SG = sclerotic glomeruli; wk = week; THC = thrombocytopenia; TI = time interval; TMA = thrombotic microangiopathy; yr = year.
Chatelet & Hurault de Ligny (6) PC, Chatelet Feb 2012
30-yr-old 1 previous graft loss consecutive to recurrent aHUS
Complete adult protocol. Maintenance dose interval has been increased to 3 wks for the last 6 mo (ND)
CR of aHUS; 2 relapses occurred after aC5 doses had been delayed from 6 to 8 days s. creat at 230 μmol/L (43 mo) CH50 below the detection threshold
Started at 600 mg/week and then 600 mg every 3 weeks (ND)
CR of aHUS s. creat at 48 μmol/L (25 mo)
Zuber & Legendre PC
CFH (S1191L V1197A)
3-yr-old 1 previous graft loss consecutive to recurrent aHUS
Plasma resistant (prophylactic plasma therapy) FFP at day 0, 1 and 2, then 3 daily PE
Complete adult protocol (ND)
CR of aHUS s. creat at 65 μmol/L (17 mo) No TMA lesion on a 12-mo protocol biopsy CH50 below the detection threshold
Zuber & Legendre PC
CFH-CFHR1 Hybrid gene
1-yr-old 4 previous RT, including 2 losses by recurrence and 2 losses by early graft vessel thrombosis
3 days (persistent anemia, THC at 60000/mm3)
Incomplete response to prophylactic plasma therapy (20 PE and 8 FFP infusions during the first month)
Complete adult protocol (ND)
Rise in Hb and stabilization around 11 g/dL, s. creat at 204 μmol/L (14 mo) Low haptoglobin level between 6 and 12 wks and fresh TMA lesions on a renal biopsy 3 months after the onset of aC5 CH50 below the detection threshold
Zuber & Legendre PC
Anti-FH antibody Δ CFHR1/3
6-yr-old 4 previous RT, including 3 losses by recurrence
A 5-yr-post-RT graft biopsy (slight decrease in renal function) disclosed fresh TMA lesions
Incomplete response to prophylactic plasma therapy (1 PE every 6 wks as maintenance plasma therapy)
Complete adult protocol (ND)
Stabilization of s.creat at 80 μmol/L (9 mo) CH50 below the detection threshold
27-yr-old 1 previous graft loss consecutive to recurrent aHUS
15 mo Curative plasma exchanges were performed and progressively tapered over 8 months, and subsequently stopped.
Thirteen months after cessation of plasma therapy, a rise in s. creat. Associated with THC led to graft biopsy which disclosed fresh lesions of TMA. PE sessions were resumed but failed to improve renal function (n = 7)
Complete adult protocol (ND)
Increase in platelet count up to 134,000/mm3. A delay of 7 days in the fifth infusion of aC5 led to a mild relapse (rise in s. creat and LDH levels). Patient subsequently recovered baseline renal function (around 156 μmol/L) following re-initiation of treatment (4.5 mo)
Duran 2012 (13) PC, Campistol Feb. 2012
28-yr-old No previous RT
2.5 mo (triggered by cocaine consumption)
Partial response to 8 every other day PE sessions, frank relapse 1 mo later
Complete adult protocol (ND)
Normalization of platelet count and haptoglobin level Recovery from hemodialysis therapy. Renal function progressively improved with a last f/u creatinine level at 168 μmol/L (16 mo)
Heyne 2011 (30) PC
Post-partum 1 previous graft loss consecutive to recurrent aHUS
aC5 was used as first-line therapy
Complete adult protocol until 8 month post-RT when administration of aC5 was tapered off and stopped
Normalization of hematological parameters Renal function progressively improved with a creatinine nadir at 114 μmol/L Influenza vaccination triggered a relapse 3 months after aC5 had been stopped aC5 was resumed and subsequently spaced to every other month dose regimen The last f/u creatinine level was 123 μmol/L (14 mo)
All patients enrolled in the study from the French aHUS cohort underwent a complete complement work-up in the reference laboratory for complement assessment at the Georges Pompidou European Hospital in Paris, France, after having provided informed written consent for gene screening. Screening for anticomplement factor H (CFH) antibodies and complete exon sequencing of the genes encoding for CFH, membrane cofactor protein (MCP), complement factor I (CFI), complement component 3 (C3), complement factor B (CFB) and thrombomodulin (THBD) was performed as previously described. Patients were screened for a CFH/CFHR1 hybrid gene and CFHR1 deletion using multiplex ligation-dependent probe amplification (MLPA).
In the reference laboratory for complement assessment, the extent of C5 blockade is measured with the CH50 assay (complement hemolytic assay). This lytic assay uses antibody-sensitized sheep erythrocytes as the activator of the classical complement pathway. The proteins of the terminal pathway, from C5 to C9, are involved in the formation of the C5b-9 lytic complex. Various dilutions of the test plasma are used to determine the amount required to induce 50% of lysis. The CH50 test is therefore an indirect measure of residual functional C5 activity. Patients with hereditary homozygous C5 deficiency exhibit a CH50 lower than 10%. Titration has shown that a CH50 activity below the detection level (<10%), indicates a C5 activity lower than 1% of the normal value.
Immunostaining on graft biopsies
C5b-9 staining was performed in biopsies of four patients, for which frozen samples were available. C5b-9 staining was studied using a mouse antihuman C5b-9 monoclonal antibody (clone αE11, Dako) by indirect immunofluorescence on slides obtained from frozen specimen. C4d staining was studied using a rabbit antihuman C4d monoclonal antibody (DB107, clone A24-T, DB Biotech) by immunohistochemistry on slides obtained from AFA-fixed specimen.
Characteristics of 9 aHUS patients given prophylactic anti-C5
Six patients from our cohort had received prophylactic anti-C5 to prevent posttransplant aHUS recurrence (18), in addition to three other cases already reported in the literature (Table 1) (15–17). All but one cases had a child-onset aHUS history with a median age of 4 years old [range, 0.5–33] at diagnosis. They all harbored a genetic abnormality associated with a high risk of recurrence. Five patients carried a heterozygous mutation in C-terminus of CFH, located in the SCR20 (n = 4) or SCR19 (n = 1), and three others had a large nonhomologous genomic rearrangement between the CFH and CFHR1 genes. The remaining patient displayed a gain-of-function mutation in the C3 gene. Three patients had previously received four renal transplants, which had been lost from early aHUS recurrence (cases 4, 8 and 9). The median age at the time of the present renal transplantation was 9 years (range, 6.4–41). Two cases (5 and 9) exhibited a preformed donor-specific antibody (DSA), with a low level titer (MFI <1000) at the time of renal transplantation.
In all but one, the renal transplant was of deceased-donor origin. Basically three kinds of strategies were undertaken (Figure 1A). One of them consisted in sequential prophylaxis by switching plasma exchanges started just before transplantation to eculizumab therapy (cases 1 and 2). The first anti-C5 administration was thus postponed to days 10 and 5 postsurgery in cases 1 and 2, respectively. Another strategy relied on eculizumab-based prophylactic therapy, which was initiated 1 week or more before the transplantation, while the recipient was listed with urgent status for deceased-donor renal transplantation (case 3) or awaiting a scheduled living nonrelated donor renal transplantation (case 4). Finally, the last and simplest strategy was based on eculizumab therapy alone, started immediately before transplantation (cases 5–9). They all received the first dose of eculizumab treatment within the 24 h preceding the surgery.
Notably, five patients (cases 5–9) were given two consecutive daily anti-C5 infusions at the time of the transplantation to circumvent the concern of high complement activation triggered by ischemia reperfusion (17,32). Four patients received the starting dose a few hours before surgery and the second dose within the next 24 h (cases 5, 6, 8 and 9). The remaining patient (case 7) had two doses before surgery at days −1 and 0.
Outcome of renal transplantation performed in patients given prophylactic anti-C5
Of the nine patients treated preemptively with eculizumab, eight experienced a successful recurrence-free posttransplant course after a median follow-up of 14.5 months (range, 2–39) (Table 1). The mean ± SD creatinine level at last time point was 71.6 ± 44.8 μmol/L. Anti-C5 therapy was pursued in all patients except case 8, who lost her graft from immediate arterial thrombosis, despite undetectable CH50 activity. To confirm the full blockade of circulating C5, C5-dependent functional assay, terminal complement complex (TCC) dosing or CH50 were repeatedly monitored just before dosing in all cases (Table 1). They were constantly measured below or around the detection threshold, suggesting an optimal circulating C5 blockade. In cases 1 and 5, the C5-dependent functional assay rose significantly when maintenance doses were spaced out to every 3 weeks, instead of every 2 weeks. C5-dependent complement assay peaked two times in patient 3, concomitantly with BK virus and upper respiratory tract infections, respectively and once concurrently with Clostridium difficile infection in case 4. Notably, in this later, detection of functional C5 coincided with a slight and transient decrease in platelet count, which may indicate a mild reactivation of the aHUS process. In three patients (cases 5, 7 and 9), C5b-9 staining was performed in biopsies (Figure 1B). In cases 5 and 7, a negative C5b-9 staining revealed efficient and sustained in situ C5 blockade, while the two biospies performed in case 9, at 6 and 8 weeks posttransplant, disclosed a mild granular mesangial staining.
Cases 5 and 9 experienced a mixed rejection, including tubulitis, endothelitis, and glomerular and peritubular capillaritis as well as diffuse C4d positive staining, at 6 and 12 weeks posttransplant, respectively. Strikingly, none of these biopsies disclosed any lesions of TMA. Surprisingly, a de novo DSA, but not the preformed DSA, was readily detected in blood samples at the time of the rejection in both cases. The treatment consisted of three pulses of methylprednisolone steroid, followed by high-dose oral steroid therapy, with mycophenolate mofetil and tacrolimus, while eculizumab was pursued. Graft function improved and returned to baseline values in both of these cases. Although low-level DSA remained detectable throughout the follow-up, subsequent biopsies demonstrated a complete recovery from rejection in both patients. None of the other patients treated prophylactically had any DSA or C4d-positive staining on biopsy sections.
All of the patients were vaccinated against meningococcal infection and all but one received daily antibioprophylaxis during the treatment period. No significant infectious complications were reported.
Characteristics of 13 renal transplant recipients treated with anti-C5 for posttransplant aHUS recurrence
Seven patients from our cohort received anti-C5 to treat posttransplant aHUS recurrence (6,28–30), as did six other cases reported in the literature (Table 2) (5,7,9,11,13,31). Two were switched from plasma therapy to anti-C5 for severe intolerance to plasma or for personal convenience, while being considered sensitive to, yet dependent on, plasma (cases 10 and 11). Ten others were switched for plasma-resistant forms of aHUS (cases 12 to 21). The remaining received eculizumab as first-line therapy (case 22). The median age of aHUS onset was 6 years old (range, 1–28 years). All but one was screened for mutations in complement genes. Five, two and one had heterozygous mutations in the CFH, C3 and CFI genes, respectively. In addition, 10 of them were investigated by MLPA for genomic rearrangements in the CFH-CFHR1–5 regions. A harmful CFH/CFHR1 hybrid gene was identified in case 18. Anti-CFH antibodies, associated with the polymorphic CFHR1-CFHR3 deletion, were detected in case 19. No mutation was identified in cases 13 and 22, but not all genes have been studied. Nine of the 13 patients had previously received a total of 17 renal transplants, of which 14, 2 and 1 had been lost by aHUS recurrence, graft vessels thrombosis and chronic rejection, respectively. The median time interval between the current renal transplantation and aHUS recurrence was 2 months (range, 3 days to 5 years). Patients underwent the current transplantation at a median age of 32 years (range, 6–43 years) and received the first dose of anti-C5 after a median delay of 30 days (range, 1 day to 14 months) following the onset of posttransplant aHUS recurrence.
None of the patient exhibited preformed DSA. Atypical HUS recurrence was demonstrated by biopsy in 10 out of the 13 patients. Importantly, only one of these initial biopsies exhibited histological signs of antibody-mediated rejection. In case 19, the diagnostic biopsy displayed C4d negative glomerular and peritubular capillaritis, along with TMA lesions, associated with a low-level DSA. C4d staining was performed on 7 biopsies at the time of diagnosis and was always negative.
Outcome of renal transplant in patients treated by anti-C5 for posttransplant aHUS recurrence
All but three patients were given a complete anti-C5 protocol, still ongoing at last follow-up. Strikingly, cases 12 (11) and 13 (9) who received a single dose of anti-C5 experienced a relapse after 21 and 11 months of a HUS remission, respectively (personal communication from J. Nurnberger (Apr 2010) and M. Lozano (Jan 2011) to C. Loirat, with permission). Anti-C5 retreatment failed to prevent evolution toward ESRD in the former and did not allow complete recovery of renal function in the latter, who ultimately lost the graft from antibody-mediated rejection after anti-C5 had been withdrawn a second time owing to a severe pulmonary infection. In all 13 patients, hematological features of aHUS rapidly returned to or remained normal following the onset of anti-C5 therapy, while mean creatinine level (±SD) dropped from 295 ± 171 to 135 ± 69 μmol/L (p = 0.002) during the three subsequent months. Importantly, in those with incomplete or no response to plasma, the longer the interval between the onset of the HUS episode and anti-C5 treatment initiation the lower the recovery of renal function (Figure 2). Typically, patients treated with anti-C5 more than 28 days after the onset of the aHUS recurrence had a lower functional benefit from anti-C5 than those treated earlier (Figure 2). Patient 17 illustrates the case of a dramatic improvement of graft function following early onset anti-C5 therapy for a severe plasma-resistant form of aHUS recurrence (Figure 3A). Interestingly, on the biopsy performed 2 days after the first administration of anti-C5, while creatinine level was peaking, there were mild thrombotic microangiopathic lesions, including a fresh intraglomerular thrombosis, arterial wall edema and turgescent endothelial cells, while segmental C5 deposits were identified (Figure 3A). At 1-year posttransplant, these lesions as well as complement deposition had completely disappeared (Figure 3A). Cases 10 and 11 were given anti-C5 while being dependent on plasma (6,33). Interestingly, both of them experienced a mild but clinically significant improvement in renal function after switching to anti-C5, although aHUS was thought to be fully sensitive to plasma. The mean ± SD creatinine level at last follow-up was 140.6 ± 64.7 μmol/L in the 11 patients who were maintained on eculizumab and who were still off dialysis after a median period of treatment of 16 months [range, 4.5–49].
Notably, in cases 10 and 20, a 6- to 8-day delay in anti-C5 maintenance dose was three times associated with a mild aHUS relapse that was rapidly responsive to anti-C5 reinitiation (Table 2, cases 10 and 20). Yet, maintenance doses were eventually spaced again to every 3 weeks in case 10, without further relapse while CH50 remained below or around the detection level. In case 22, anti-C5 was successfully stopped for 3 months until a vaccine triggered a mild relapse, easily controlled by anti-C5 reinitiation. Progressive tapering of doses was undertaken again and detectable C5b-9 reflected incomplete C5 blockade under an every-other-month maintenance regimen. Notably, in the cases 18 and 19, a graft biopsy performed 3 and 15 months, respectively, after the onset of anti-C5 disclosed fresh lesions of thrombotic microangiopathy (TMA). Importantly, these features suggestive of active thrombotic microangiopathic processes coincided with tacrolimus overexposure (trough levels were between 18 ng/mL and 20 ng/mL) in the former and with persisting chronic active AMR in the latter, which might have triggered endothelial injury. In both of them, a CH50 activity below the detection threshold argued in favor of an optimal drug exposure. However, a positive C5b-9 staining, coexisting with TMA lesions, was documented in case 18 but not in case 19. Case 18 was switched back to plasma exchanges for three daily sessions while tacrolimus exposure was significantly reduced. Subsequently, anti-C5 was resumed. The biopsy performed 5 months later did not show any more active thrombotic microangiopathic lesions and barely detected granular C5b-9 deposits within the glomeruli (Figure 3B).
All patients were vaccinated against meningococcal infection and 5 of the 10 patients for whom information was available received daily antibioprophylaxis. Cases 13 and 19 each experienced a diffuse pulmonary infection by M. tuberculosis and H. influenza, respectively. Case 19 exhibited a profound hypogammaglobulinemia while she had undergone splenectomy during the early childhood and had received rituximab in an attempt to clear anti-CFH antibody. Her outcome was favorable with antibiotics. Weekly subcutaneous immunoglobulin supplementation successfully prevented recurrence.
We reported here the largest series of 22 renal transplant recipients treated with anti-C5 for prevention or treatment of aHUS recurrence.
Regarding prophylactic regimens, a first dose of anti-C5 was administered a few hours before surgery and an additional dose within the next 24 h (postoperative day 1) in three patients, as it has been proposed for preventing antibody-mediated rejections (ClinicalTrials.gov: NCT01399593) (34) and catastrophic antiphospholipid syndrome posttransplant recurrence (ClinicalTrials.gov: NCT01029587) (35). The day 1 infusion, incorporated in prophylactic protocols, stemmed from the finding that complement products are dramatically released following graft reperfusion (17,32). This led to the idea that conventional protocols might not sufficiently block circulating C5 during the early posttransplant course and that a greater anti-C5 exposure might be needed. Although no study has been designed to clarify whether this concern is clinically well grounded, this precaution seems to us quite reasonable. Among preemptively treated patients, anti-C5 treatment was started at least 1 week prior to transplantation in two cases because either a living-donor renal transplantation was scheduled or the patient was listed with urgent status for deceased-donor renal transplantation (15,16). However, our study shows that an anti-C5 schedule started at day 0 with an additional dose in the first 24 h after surgery was remarkably effective at preventing posttransplant recurrence and better accommodates with most deceased-donor transplantation situations.
Given the limited size of the study population and the absence of controls, one may nonetheless argue against our ability to firmly establish the efficacy of prophylactic anti-C5 therapy from these data. However, it is worth recalling that all of these patients harbored a genetic abnormality in complement genes associated with a very high recurrence risk. In the French adult aHUS cohort, nonhomologous recombinations of the 3′ region of the CFH gene, loss-of-function mutations in SCR20 of the CFH gene, and gain of function mutations in the C3 gene have been associated with a recurrence risk of 100, 100 and 80%, respectively (Le Quintrec, manuscript submitted). In addition, this is the subset of patients in whom prophylactic plasma therapy was found to be the least effective at preventing aHUS recurrence (Le Quintrec, manuscript submitted), as illustrated here by case 17. Our data suggest that prophylactic anti-C5 therapy may be the most suitable strategy in aHUS patients in whom an identified genetic abnormality in complement genes or a personal history of posttransplant recurrence predicts a very high risk of recurrence for the current transplantation (Table 3).
Table 3. Promises, uncertainties and challenges for eculizumab in aHUS recurrence settings
Possible design of future studies
Key end points of future studies
This table summarizes the findings of the present study and our proposals for the prevention or treatment of aHUS recurrence according to an individualized risk assessment. Aims, design and endpoints of future studies are briefly discussed to address the pending questions related to the management of aHUS recurrence. aHUS = atypical hemolytic uremic syndrome; DGF = delayed graft function; eGFR = estimated glomerular filtration rate. *Please see references 46–48, **Please, see references 43 and 45, ***Complete remission is defined as normalization of platelet count and LDH level and recovery of baseline renal function.
Prophylactic anti-C5 therapy
Posttransplant aHUS recurrence is a strong predictor of early graft failure. Anti-C5 efficiently prevents aHUS posttransplant recurrences. Prophylactic plasma therapy may fail to prevent aHUS recurrences or may mask subclinical, yet progressive, aHUS recurrences in the subgroup of patients with high- risk mutations.
We propose to stratify renal transplantation candidates into risk levels of recurrence by their medical history and genetic background. Patients at high risk of recurrence include those who have already experienced a recurrence in a previous graft and those whose mutations are associated with a risk of recurrence greater than 80% (mutation in CFH, CFH-CFHR1 hybrid gene, gain of function mutations in C3 and CFB) Patients at moderate risk of recurrence include those with detected anti-CFH antibody, isolated mutations in CFI, mutations of unknown functional significance and no mutation identified. Patients at low risk of recurrence include those with isolated mutations in MCP and those in whom anti-CFH antibodies have been cleared from circulation. Prophylactic therapy should be undertaken in patients with high and moderate risks while those at low risk should only be monitored. We propose the use of anti-C5 instead of plasma therapy to prevent posttransplant recurrences and to attempt discontinuation of the treatment in patients at moderate risk.
Inclusion criteria: Adult aHUS patients Candidates for renal transplantation Completion of thorough complement investigations Moderate to high risk of recurrence Local availability of daily plasma exchanges Exclusion criteria: Severe allergy to one of the components of both therapeutic arms Ongoing meningococcal infection Contraindication to vaccination against N. meningitidis or to daily antibioprophylaxis Design: A randomized study could compare two arms over a 3-year study period: – Prophylactic plasma therapy* – Anti-C5 complete protocol**, including an additional dose 1 day post-surgery. Patients in whom the initial treatment would fail to prevent aHUS recurrence would be switched to the other arm. Prophylactic therapy will be discontinued at 1 year post-transplant only in the subset of patients with a moderate risk of recurrence. Anti-C5 would be (re)-initiated if aHUS recurrence occurred.
Primary Incidence of aHUS recurrence at 1 year postrenal transplant Secondary Incidence of aHUS recurrence at the end of the study (3 yrs) Period of DGF Graft survival Occurrence of aHUS recurrence after discontinuation of prophylaxis 3- and 12-month protocol biopsies 12-, 24- and 36-month eGFR Cost-effectiveness analysis
Curative anti-C5 therapy
Plasma therapy has failed to improve graft survival in renal transplant recipients with aHUS recurrence The earlier the initiation of anti-C5 after aHUS recurrence the greater the recovery of graft function.
We propose the prompt initiation of anti-C5 for the treatment of overt posttransplant aHUS recurrence. Full-blown as well as incomplete presentation of aHUS should be treated in a similar manner as they both lead to rapid irreversible damages of the graft. Given the multiple endothelium-insulting factors in renal transplant recipients, we would not recommend discontinuation of anti-C5 treatment in aHUS patients who have experienced a posttransplant recurrence in the current graft.
Inclusion criteria: Adult renal transplant recipients with aHUS recurrence Local availability of daily plasma exchanges Exclusion criteria Same as for the prophylactic protocol Design: The study could last 3 years. Two randomized arms: – Curative plasma therapy* – Anti-C5 complete protocol** Patients, in whom the initial treatment would fail to obtain partial and complete remission*** of aHUS by day 5 and 10, respectively, would be switched to the other arm.
Primary Achievement of complete remission (%) Secondary Recovery of renal function Number of dialysis Graft survival Changes from one arm to the other (%) 12-month protocol biopsy 12, 24 and 36-month eGFR Cost-efficacy analysis
With respect to curative protocols, several lessons could be drawn from this pioneer experience. First, a single-dose regimen was consistently associated with subsequent relapses, with uncertainty about sequela-free recovery. Indeed, aside from the two renal transplant recipients reported here (9,11), two patients with aHUS involving native kidneys experienced a relapse after a single dose of anti-C5 (10,36). Strikingly, anti-C5 reinitiation failed to prevent evolution toward ESRD in three of them and achieved an incomplete functional recovery in the fourth one, who ultimately also progressed to ESRD (2,9–11,36) (personal communication from M. Lozano [Jan 2011] and J. Nurnberger [Apr 2010] to C. Loirat, with permission). Thus, a single-dose regimen should not be recommended. Second, among the 11 patients who received anti-C5 for overt and uncontrolled aHUS recurrence, the interval between the onset of the HUS episode and anti-C5 initiation critically determined the extent of recovery of renal function. The earlier anti-C5 treatment was started following the onset of HUS recurrence, the better the chances were for recovering baseline renal function. It is likely that renal allografts, exposed to ischemia-prone conditions, including calcineurin inhibitor-related toxicity (2,37) and immune-mediated allograft vasculopathy (38), should be more prone than native kidneys to rapidly develop ischemia-induced graft scarring in the settings of incomplete aHUS control. Moreover, we recently provided evidence that curative plasma therapy failed to improve graft survival in patients with posttransplant aHUS recurrence (Le Quintrec, submitted). Altogether, these data suggest that prompt introduction of anti-C5 should be recommended as a first-line therapy for the treatment of posttransplant aHUS recurrence. Third, aHUS renal transplant recipients under long-term plasma therapy, considered sensitive to plasma, may experience an improvement in either renal function or hemoglobin level after being switched to anti-C5 (cases 9 and 10, Table 2). This finding is in line with other cases of aHUS involving native kidneys (39) and is reminiscent of the fact that mild and clinically truncated thrombotic microangiopathic processes may continue to progress under long-term plasma therapy, as evidenced by case 19. Therefore, we recommend a liberal renal biopsy policy, especially in the cases of slightly decreased renal function, appearance of de novo proteinuria, poorly explained anemia, increasing need of antihypertensive drugs. In this context, biopsy-proven TMA lesions would require a switch to anti-C5.
Antibody-mediated rejection (AMR) is the most common cause of de novo posttransplant TMA (40). Interestingly, a recent case report described a renal transplant recipient with severe AMR associated with TMA in whom the complete polymorphic deletion of CFHR3/1, but not the anti-CFH antibody, was identified (41). This finding raised the question of how genetic susceptibility factors may predispose transplant recipients to develop de novo TMA in the setting of AMR. This finding is further supported by our previous study, which identified mutations in the CFH and CFI genes in roughly one-third of renal transplant recipients with de novo posttransplant aHUS (21). We therefore wondered whether antibody-mediated rejection might have triggered posttransplant aHUS recurrence, through initiation of endothelial insult. However, evidence of coexisting AMR and TMA lesions at the time of the diagnosis of aHUS recurrence was observed only in one patient under long-term plasma therapy. Moreover, two patients treated prophylactically with eculizumab remained free of any aHUS recurrence, although they experienced a severe, yet steroid responsive, mixed rejection.
A 6- to 8-day delay in anti-C5 infusion has been associated three times, in two patients, with a mild aHUS relapse that was rapidly responsive to anti-C5 re-initiation. This is consistent with the incomplete TCC blockade before doses observed under every 3 week instead of every 2-week regimens (17). Taken together, our data suggest that the 14-day delay between two maintenance doses should not be spaced out in individuals who weigh more than 10 kg or should be spaced out with extreme caution in rare situations and under close monitoring of the complement assay. We also report evidence that active aHUS disease may occur in patients treated with anti-C5 according to the recommended schedule. The transient increase in circulating C5 and TCC in patients with concurrent infections, including BK virus replication, upper respiratory tract infection and infectious colitis (15,16), suggests that an infection-related complement activation might induce temporary underdosing in anti-C5, when treating with a dose that is typically sufficient to block the complement assay at steady state. Notably, a significant decrease in platelet count was noticed in the patient with detectable blood C5, concurrently with Clostridium difficile infection (15,16). Active TMA lesions were also observed in two cases, in which undetectable CH50 indicated a complete depletion of circulating C5 (below 1% of normal values). This finding suggests that TMA lesions may continue to progress through complement-independent mechanisms in the setting of overwhelming and continuous endothelial damages. However, in one of these cases, biopsies performed 3 and 9 months after the onset of anti-C5 revealed C5b deposition coexisting with active TMA lesions, although the deposition was noticeably less in the latest biopsy. As suggested by others, this may reflect a prolonged half-life of pretreatment C5b deposits, rather that insufficient in situ C5 blockade (42). Taken together, these findings suggest that aHUS patients, regardless of the treatment used, should be closely monitored when exposed to endothelium-insulting factors and/or to intercurrent events associated with heightened activation or production of complement factors, including infections, vaccines, CNI overexposure and antibody-mediated rejection.
The optimal duration of anti-C5 therapy in aHUS remains an elusive question, which has not been properly addressed so far. Life-long treatment has been approved in off-dialysis aHUS patients by the European agency (43). In the present series, anti-C5 was withdrawn in three cases, either after a single dose in two of them or after 8 months of treatment in another. They all subsequently experienced a relapse. However, in our opinion the duration of the treatment might be individualized based on genetic findings (Table 3). If complement genetic abnormalities are associated with a very high risk of posttransplant recurrence, i.e. mutations in CFH and gain of function mutations in C3 and CFB, this suggests an overwhelming contribution of genetics over environmental triggers to aHUS pathogenesis. In these patients, as well as in those patients treated for a recurrence in the current or a previous graft, long-term anti-C5 therapy should not be discontinued throughout the graft lifespan unless a combined liver–kidney transplantation is undertaken to definitively cure the disease. However, one should acknowledge that nobody can predict the new posttransplant history of aHUS, now that a prophylactic treatment efficiently prevents the recurrence during the first months, associated with the highest risk of ‘endothelial trigger’. Therefore, attempts to taper off anti-C5 therapy after a 12-month aHUS recurrence-free period of full complement blockade might be a reasonable strategy, at least in those with a moderate risk of recurrence (Table 3). In this respect, prospective trials are critically warranted to address how long the treatment should be continued according to an individual's genetic background and history of recurrence. These future studies will also be mandatory to definitively confirm the superiority of anti-C5-based therapy over plasma therapy (Table 3), to demonstrate the benefit of anti-C5-based prophylaxis in candidates for renal transplantation with a moderate to high risk of recurrence, and to establish that anti-C5 should be used as first-line rather than as rescue therapy for the treatment of posttransplant recurrence (Table 3). However, for obvious ethical reasons, patients in whom the initial treatment would fail to prevent or fully resolve aHUS recurrence should be rapidly switched to the other arm.
The high cost of the drug may prompt clinicians to prematurely discontinue the drug or to prefer initiating treatment with plasma therapy. To address this financial concern, we compared the cost of both plasma and anti-C5-based strategies for a 3-month treatment of an acute episode (Figure 4). This cost model was based on the assumption that all procedures would be undertaken during day care hospitalizations. We postulated that most cases would have arteriovenous fistulas and did not take into account the cost of central venous catheter insertion and related complications. Our cost analysis indicates that the difference in cost between anti-C5 and plasma therapies varies significantly depending on the weight of the individual and the frequency of maintenance plasma therapy (Figure 4). In addition, the over-cost related to a 3-month anti-C5 treatment, particularly significant in patients heavier than 40 kg, is on the same order of magnitude as the money saved by rescuing a renal allograft from early failure (Figure 4). A medico-economic study conducted by the French Health Authority (Hautes Autorités de Santé) and reported in 2010 estimated the cost of 1 year spent on hemodialysis to be approximately 88 608 € ($ 112 199), while the cost of a first and subsequent years following renal transplantations were around 86 471 € ($ 109 493) and 20 147 € ($ 25 511), respectively (44). Therefore we believe that short-term anti-C5 therapy is the most cost-effective strategy to treat aHUS recurrence, although we agree that this issue should be more thoroughly addressed for long-term treatment. The duration of hospitalization stays, period of dialysis need, graft survival, vascular access-related complications, gain in health-related quality of life and the preservation of satisfactory working ability would have to be taken into considerations in a medico-econonomic study of long-term anti-C5 treatment in the subset of patient with high risk of recurrence.
Finally, two of the 22 patients experienced a pulmonary infection, which could be explained in one case by the patient's profound hypogammaglobulinemia. However, no meningococcal infection was reported. Although the incidence of meningococcal infection has been reported to be approximately 4.2/1000 patient year in patients with paroxysmal nocturnal hemoglobinuria treated with long-lasting eculizumab therapy (45), we are not aware of any infection in patients receiving daily antibioprophylaxis. Importantly, to date serogroup B has not been covered by any vaccine, despite its status as the most frequent cause of invasive meningococcal disease in both America and Europe. In addition, the uncertainty surrounding the vaccine's efficacy in immunocompromised patients further supports the use of daily antibioprohylaxis during the eculizumab treatment period in renal transplant recipients.
In conclusion, anti-C5 therapy is highly effective for preventing and treating posttransplant aHUS recurrence. Our study shows that a single-dose anti-C5 regimen should not be used and that maintenance anti-C5 doses should not be spaced out more than 14 days for patients heavier than 10 kg. Despite the limitations of a small sample size, this series also emphasized that early anti-C5 initiation seems to ensure better recovery of renal function. We also provide the first evidence that in situ C5 deposition may be indicative of partial escape of C5 blockade in the setting of excessive endothelial damage triggered by environmental factors.
We wish to thank Ms. Laura Leboucher (Financial Administrator, Necker Hospital, Paris France) and Pr. Jean-Bernard Zuber (Laboratoire de Physique Théorique et Hautes Energies, Université Pierre et Marie Curie, Paris, France) for their invaluable help in performing the cost analysis study. The authors are also grateful to Dr. S. Al-Akash (Driscoll Children's Hospital, Texas, USA), Dr. J.M. Campistol (Hospital Clinic, ICNU, University of Barcelona, Barcelona, Spain), Dr. J.C Davin (Emma Children's Hospital, Amsterdam, The Netherlands), Dr. M. Lozano (Hospital Clinic, IDIBAPS, University of Barcelona, Barcelona, Spain), Dr. C. Nester (University of Iowa Hospitals and Clinics, Iowa City, Iowa, USA), Dr. J. Nürnberger (University Duisburg-Essen, Essen, Germany), Dr. M. Riedl (Medical University Innsbruck, Innsbruck, Austria), and Dr. Weitz (University Children's Hospital Tuebingen, Tuebingen, Germany) for providing updated data on the late outcomes of their patients transplanted under eculizumab therapy. We would also like to thank all the members of the French Study Group for Atypical HUS who participated in fruitful and constructive discussions about this study, including Pr. M. Buchler (Tours, France), Pr. S. Burtey (Marseille, France), Dr. Y. Delmas (Bordeaux, France), Pr. G. Deschênes (Paris, France), Pr. F. Fakhouri (Nantes, France), Dr. A. Karras (Paris, France), Pr. B. Knebelmann (Paris, France), Pr. B. Moulin (Strasbourg, France), Pr. C. Pouteil-Noble (Lyon, France), Dr. F. Provost (Lille, France).
The authors have conflicts of interest to disclose as described by the American Journal of Transplantation. J.Z., V.F.-B., C.L. and C.L. have received fees from Alexion Pharmaceuticals for invited lectures.