Eculizumab prevents thrombotic microangiopathy in patients with atypical haemolytic uraemic syndrome in a long-term observational study

Abstract Background Eculizumab, a terminal complement inhibitor, is approved for atypical haemolytic uraemic syndrome (aHUS) to inhibit complement-mediated thrombotic microangiopathy (TMA). Methods In five parent studies, eculizumab effectively prevented TMA and improved renal and haematologic outcomes in patients with aHUS; therefore, these patients could enrol in this long-term, prospective, observational and multicentre study. The primary endpoint was the TMA manifestation rate off and on eculizumab post-parent study. Post hoc analyses evaluated rates during labelled versus non-labelled dosing regimens, and in those with versus without identified complement abnormalities. Serious targeted treatment-emergent adverse events (TEAEs) were evaluated. Results Of 87 patients in the current study, 39 and 76 had off- and on-treatment periods, respectively; 17 (44%) with off periods reinitiated eculizumab. TMA manifestation rate per 100 patient-years was 19.9 off and 7.3 on treatment [hazard ratio (HR), 4.7; P = 0.0008]; rates were highest off treatment and lowest during labelled regimens. TMA manifestations with hospitalizations/serious AEs occurred more frequently off versus on treatment. TMA rates were higher among patients with identified complement abnormalities (HR, 4.5; P = 0.0082). Serious targeted TEAEs occurred at similar rates off and on treatment. Conclusions As expected, patients with aHUS have increased risk of TMA manifestations after discontinuation of eculizumab or in the setting of non-labelled eculizumab dosing. Collectively, results show that maintaining eculizumab treatment minimizes risk of TMA, particularly in patients with identified complement abnormalities. Future studies are needed to further characterize TMA and longer term outcomes on labelled or non-labelled eculizumab regimens and after discontinuation of treatment.

Patients who remain untreated are at lifelong risk of renal impairment or failure, organ dysfunction and premature death [2,3]. Eculizumab (Soliris V R , Alexion Pharmaceuticals, Inc., New Haven, CT, USA), a humanized monoclonal antibody that inhibits C5a, C5b and C5b-9 formation by binding to C5, is the first and only approved treatment for patients with aHUS [5,6]. The efficacy and safety of eculizumab have been demonstrated in four prospective, multicentre clinical studies [7][8][9][10] and a retrospective study [11].

Study design and patients
This is a long-term, prospective, observational and multicentre study (NCT01522170) of patients with aHUS who were treated with eculizumab in any of five previous clinical studies (parent studies): four prospective studies (NCT00844545/NCT00844844 and NCT00838513/NCT00844428 [7,8], NCT01193348 [9], NCT01194973 [10]) and one retrospective study (NCT01770951 [11]). Patients who participated in a parent study were eligible for the current study, regardless of whether they completed or discontinued from the parent study or were on eculizumab at the time of enrolment. Patients could withdraw from the current study at any time. The protocol was approved by an institutional review board or independent ethics committee at each participating centre and the study was conducted in accordance with International Council for Harmonisation Guidelines and the Declaration of Helsinki. All patients and/or parents/guardians provided written informed consent before entry into the current study.
The current study consequently includes both prospective and retrospective data collection. Retrospective data were obtained from the date each patient ended participation in the parent study until the date of signed informed consent for the current study. All patients who received at least one infusion of eculizumab during the parent study and had signed consent forms for the current study were included in the analysis. Identification of complement abnormalities occurred during the parent studies and included analysis of complement factor I (CFI), complement factor B (CFB), complement factor H (CFH), membrane cofactor protein (MCP) and C3 mutations, complement factor H-related proteins 1-3 (CFHR1-3) deletions/polymorphisms and CFH autoantibodies [7,[9][10][11]. Patients received meningococcal vaccination in the parent studies [7,9,10] and were revaccinated according to country guidelines. The cut-off date for this interim analysis was 28 March 2015.
TMA manifestations were neither defined nor collected uniformly in the parent studies; thus, this analysis includes outcomes reported in this ongoing observational study only (i.e. beginning at the end of the parent study). Data were collected four times annually in both the retrospective and prospective portions of the current study.

Treatment
After completion of the parent study and entry into the current study, the labelled dosing regimen of eculizumab was defined as that specified in the prescribing information approved by regulatory authorities [5,6], and other dosing schedules (i.e. decreased or increased dosages, shortened or extended dosing intervals) were permitted and classified as non-labelled regimens. In this study, the first on-treatment period was defined as: from the date of the first infusion in the current study (i.e. beginning at the end of the parent study) through 3 weeks after the last infusion of eculizumab, or until the patient discontinued from the study, or data cut-off (whichever occurred first). The first off-treatment period was defined as: from 3 weeks after the last infusion of eculizumab within the current study until the patient restarted eculizumab therapy, discontinued from the study or data cut-off (whichever occurred first). Subsequent on-and off-treatment periods were defined similarly. Patient groups were not mutually exclusive; individual patients could be represented in both groups.

Endpoints
Primary endpoint was the rate of TMA manifestations (defined in Table 1) in the current study off and on treatment. Post hoc analyses evaluated TMA manifestation rates off and on treatment when TMA manifestations based on only a single laboratory value change were excluded; in patients receiving labelled compared with non-labelled eculizumab regimens; and in patients with and without identified complement abnormalities. Safety endpoints included assessment of serious targeted treatment-emergent adverse events (TEAEs; predefined as incidence of serious infection, meningococcal infection, sepsis, renal impairment or leucopaenia), as well as any serious AE (SAE).

Statistical methods
Time to first TMA manifestation was defined as time from the start of the current study (i.e. end of the parent study) to first TMA manifestation during the current study. Patients who did not have a TMA manifestation were censored at data cut-off or study discontinuation, whichever occurred first. Time to first TMA manifestation was analysed using Cox proportional hazards models with treatment status as a time-dependent explanatory variable and complement abnormality status as a covariate. Hazard ratios (HRs) and P values were obtained for comparisons off and on treatment and between identified and no identified complement abnormality subgroups.

Patients and exposure
Overall, 130 patients were enrolled in the parent studies. By the data cut-off for this analysis, 87 patients had enrolled in the current study. Of these, 39 (45%) had off-treatment periods whereas 76 (87%) had on-treatment periods. Seventeen patients (44%) with off-treatment periods reinitiated eculizumab; of these, 14 (82%) remained on therapy once they reinitiated ( Figure 1). Age, frequency of complement abnormalities and kidney transplant status were not different between patients with ongoing eculizumab therapy versus those who discontinued ( Table 2). Twenty-two patients (25%) had renal transplants, including eight patients (21%) who discontinued eculizumab and 14 (29%) on ongoing eculizumab. A median (range) of 11.0 (0.0-230.0) plasma exchanges or plasma infusions were used by the overall population before eculizumab initiation, including 7.0 (0.0-64.0) in patients who discontinued eculizumab and 13.3 (0.0-230.0) in patients who never discontinued eculizumab. Dialysis was required by 29/87 patients (33%) before parent study enrolment, and use was more frequent in those who discontinued eculizumab (39%) compared with those who never discontinued eculizumab (29%). Including parent studies, patients had a total median (range) of 45.9 (1.3-86.9) months of eculizumab exposure. In the current study, median (range) follow-up was 20.1 (0.7-79.5) months off and 26.1 (0.7-64.2) months on treatment.
Compared with patients who continued on eculizumab, those who discontinued had a shorter interval from initial aHUS diagnosis, as well as the most recent pretreatment TMA manifestation, to first-ever eculizumab dose in the parent study. Patients who discontinued also presented with lower estimated glomerular filtration rates (eGFRs) and were more frequently dialysis dependent at initiation of eculizumab (Table 2).
Non-labelled eculizumab regimens were received by 33/87 patients (38%) during the current study. Of these, 11 always received non-labelled doses and 22 had periods of labelled and non-labelled regimens. Median (range) duration of therapy was 25.7 (0.5-60.4) months during labelled and 14.3 (0.4-64.3) months during non-labelled regimens. Dosing higher than labelled accounted for 0.4% of the total patient-years for non-labelled regimens.

TMA manifestations
When using the per-protocol definition (Table 1), 28 TMA manifestations occurred. This included 14 TMA manifestations in 11/39 patients (28%) off treatment and 14 TMA manifestations in 10/76 patients (13%) on treatment (Tables 3 and 4). TMA manifestation rate per 100 patient-years was 19.9 off and 7.3 on treatment (63% lower; HR, 4.7; P ¼ 0.0008; Table 5). Measurements were required to be confirmed by a second measurement !28 days apart with no interruption. c Baseline was defined for each on period as the last laboratory value during the preceding off period, and for each off period as the last value during the preceding on period.
d As determined at the discretion of the investigator.
Patients treated with non-labelled and labelled regimens had TMA manifestation rates of 12.1 and 5.2 per 100 patient-years, respectively (39% and 74% lower, respectively, versus off treatment).

Characteristics of TMA manifestations
TMA manifestations off treatment were more frequently associated with multiple laboratory criteria for TMA, clinical signs and symptoms of TMA, interventions, SAEs and/or hospitalizations (Tables 3 and 4

TMA manifestations by complement abnormality status
The majority of patients who reported TMA manifestations as defined per protocol had complement abnormalities (Tables 3  and 4

Safety
Overall, treatment with eculizumab was well tolerated. The occurrence of serious targeted TEAEs during the current study was similar off and on treatment ( Table 6). Two patients from the parent retrospective study reported meningococcal infections during the current study; both were determined to be probably related to eculizumab. Both patients were treated and recovered while continuing eculizumab on schedule. Diagnoses/underlying conditions associated with reported serious targeted TEAEs of renal impairment that did not meet criteria for TMA included new kidney transplant, renal graft rejection, multiorgan failure, dehydration, infection and interstitial tubulopathy. One adult patient, who received non-labelled dosing during the current study, died due to severe intensive care complications and severe multiorgan dysfunction after gastrointestinal haemorrhage, lithiasic cholecystitis and severe sepsis, which were determined to be unrelated to eculizumab.

DISCUSSION
Results from this interim analysis of a non-randomized, prospective observational study demonstrate that rates of TMA manifestations in patients with aHUS were 2.7-fold higher off compared with on eculizumab (63% lower), despite longer follow-up on treatment. TMA manifestation rates were lowest during labelled dosing regimens (74% lower than off treatment), higher during non-labelled regimens (39% lower than off treatment) and highest off treatment. The per-protocol definition of TMA manifestations was broad, including changes in laboratory values, clinical signs and symptoms of TMA related to aHUS, and/or interventions related to TMA. Thus, reported TMA per this definition represented varying degrees of clinical deterioration. Importantly, there is no single, agreed-upon definition of TMA based upon a single laboratory value in clinical practice. TMA manifestations off treatment were associated with multiple laboratory criteria, clinical sequelae (e.g. renal impairment and acute renal failure), SAEs, hospitalizations and/or required interventions (e.g. transfusion) in 13/14 cases (93%). In contrast, TMA manifestations on treatment typically comprised changes in single laboratory values with no clinical signs/symptoms. Therefore, post hoc analyses were conducted to provide insights using a more stringent TMA definition that we believe more closely defines TMA in the setting of aHUS. When TMA manifestations based only on changes in single laboratory values were excluded, the rate off treatment was 15.6-fold higher than on treatment. These results taken together could suggest worse outcomes for patients who discontinue eculizumab, although it is possible that changes in single laboratory values may signal subclinical disease processes.
Patients with identified complement abnormalities had statistically significantly higher TMA rates than patients with no identified abnormalities. CFH and CFI mutations were predominant among patients who experienced TMA, regardless of treatment status. This finding is consistent with previous  c Dialysis at baseline was defined as any dialysis that occurred within 7 days prior to or 14 days following the first eculizumab dose in the parent study. d eGFR was defined as 10 mL/min/1.73 m 2 when a patient was on dialysis.
PE/PI, plasma exchange/plasma infusion.  observational studies of the natural history of aHUS [2,3]. In a small observational study, Ardissino et al. [28,29] also noted particular risk for TMA in patients who chose to discontinue eculizumab with CFH mutations, and CFI mutations to a lesser extent. In a retrospective study of eculizumab discontinuation in a French cohort (n ¼ 38) [27], all 12 patients (32%) with TMA post-eculizumab discontinuation had rare or novel CFH or MCP variants; both were independent risk factors for TMA following discontinuation. However, patients with transplant, on chronic dialysis or 'secondary' aHUS were excluded from the French cohort. Results of the current analysis, which are from the single largest prospective study of TMA risk following eculizumab discontinuation, reinforce previous findings that patients who discontinued eculizumab were at greater risk for TMA compared with patients on treatment, and particularly those with identified complement abnormalities. In this study, patients without identified abnormalities were at a significantly lower yet distinct risk for TMA after eculizumab discontinuation. However, genetic analyses were performed during the parent studies several years ago; thus, it is possible that novel mutations not known at that time have been left unrecognized.
Overall, characteristics of patients with TMA manifestations were highly heterogeneous with respect to age, treatment The patient's first on-treatment period was 19 months, followed by an off-treatment period of 31 months and another on-treatment period of 10 months.
LDH, lactate dehydrogenase; SCr, serum creatinine. duration and duration of eculizumab before onset of TMA. Age, frequency of complement abnormalities and kidney transplant status did not differ between patients who discontinued versus remained on eculizumab. However, patients who discontinued appeared to have poorer renal function at baseline in the parent study and initiated eculizumab more rapidly. Such patients may have had clinically significant renal improvement with eculizumab, since they initiated treatment in a rapid manner [31], followed by the clinical decision to discontinue treatment after recovery. Discontinuation of therapy was not randomized, potentially allowing selection bias for continuing versus discontinuing eculizumab. In the previously cited studies [26][27][28], not all patients with aHUS were included. In the current study, all patients who met inclusion criteria for the parent studies (including those with dialysis and renal transplants) were allowed to enrol. Further studies are needed to identify patient characteristics potentially associated with increased risk for TMA after eculizumab discontinuation.
Collectively, the current results reinforce the need for ongoing treatment with eculizumab to minimize risk of TMA in patients with aHUS, particularly those with an identified complement abnormality. Although thorough genetic testing informs prognosis, additional considerations when optimizing treatment strategy include the patient's unique clinical situation, age, TMA and family histories, as well as recognition of the complex and unpredictable nature of aHUS. For individual patients in whom discontinuation of eculizumab is being considered, clinicians would be well advised to consult an expert centre in the field while ensuring that the patient: (i) has been treated for a sufficiently long period to ensure maximal organ function recovery; (ii) can be monitored closely for signs and/or symptoms of TMA; and (iii) has immediate access to eculizumab so treatment can be restarted at the first signs and/or symptoms of TMA [27,32].
As was observed in the parent studies [7][8][9][10], eculizumab was generally well tolerated in the current study. Rates of serious targeted TEAEs, including infections, reported off and on treatment were similar. In particular, rates of renal impairment were relatively high both off and on treatment, but commonly associated with new kidney transplantation and existing graft failure. Two patients reported meningococcal infections during the current study. Both recovered and there were no changes in eculizumab dosing. Frequency of meningococcal infections [2/87 patients (2%)] is similar to that from the overall parent trial programme of eculizumab in aHUS [two cases/100 total patients (2%)] [7][8][9][10]. Overall, the reported meningococcal infection rate in patients treated with eculizumab is 0.3 events/100 patientyears [33]. Regulatory guidance for eculizumab includes increased susceptibility to meningococcal infection [5,6]. Patients should be counselled in order to fully understand potential benefits and risks of treatment, early signs of meningococcal disease and processes for seeking immediate medical care. Risks of potentially severe complications, including meningococcal infection, should be considered during the decision-making process regarding initiating treatment or discontinuing eculizumab. Long-term evaluations of the eculizumab safety profile will be included in future analyses from the Global aHUS Registry [34].
An important study limitation was its open-label and observational nature. Voluntary patient enrolment into this prospective study may have introduced selection bias because data are not available for patients who completed a parent study but did not consent to enrolment in the current study. In this analysis, 43/130 patients (33%) who enrolled in one of the parent studies had not continued into the current study. During the parent and current studies, which together included a median exposure of 45.9 months, withdrawal of eculizumab due to an AE was uncommon. One adult patient died due to multiorgan failure following a reduced dosing regimen. One paediatric patient discontinued eculizumab in the parent study due to agitation [9]. Three patients with previous renal transplants and poor renal function (eGFR <30 mL/min/1.73 m 2 ) at the start of treatment discontinued eculizumab following reports of renal impairment in the current study; of these, one later restarted eculizumab and the other two received additional renal transplants. Additional studies are needed to further understand patient and physician rationale for discontinuing and reinitiating treatment.
Taken together, findings from this interim analysis suggest that patients with aHUS have an increased risk for TMA manifestations after discontinuation of eculizumab or during non-labelled regimens compared with labelled eculizumab dosing. These results support current regulatory guidance [5,6] in noting potential risk for TMA following discontinuation of eculizumab. Evidence demonstrates that patients had a 63-94% lower risk of TMA on eculizumab therapy, depending on the definition used. Future analyses will allow for further characterization of TMA and evaluation of longer term outcomes on labelled or non-labelled regimens of eculizumab and after therapy discontinuation.

FUNDING
This interim analysis was funded by Alexion Pharmaceuticals, Inc. Medical writing/editorial support was provided by Kristen W. Quinn, PhD, of Peloton

CONFLICT OF INTEREST STATEMENT
This interim analysis was sponsored by Alexion Pharmaceuticals, Inc. J.M. receives lecture and/or advisory fees from Alexion Pharmaceuticals, Inc., AstraZeneca, Berlin-Chemie, Daiichi Sankyo, Boehringer Ingelheim and Novartis. F.F. received fees for participation in advisory boards, experts' meetings and/or teaching courses from Alexion Pharmaceuticals, Inc. J.F.K. was a stockholder and employee of Alexion Pharmaceuticals, Inc. when the study was conducted. C.L. has received grant/research support and/or consultancy fees from Achillion Pharmaceuticals, Inc., Alexion Pharmaceuticals, Inc. and CSL Behring; has received honoraria from Alexion Pharmaceuticals, Inc. and CSL Behring; has submitted patents for CSL Behring and Finnegan, Henderson, Farabow, Garrett & Dunner; is a member of the Editorial Boards for Kidney International, Nephrology Dialysis Transplantation and Pediatric Nephrology; is a Steering Committee Member of the Alport Syndrome Treatments and Outcomes Registry (ASTOR) and an Appointed Member of the Safety Board of the European Treatment Trial for Alport Syndrome (EARLY PRO-TECT); is the aHUS International Registry Scientific Advisory Board Chair and HUS International Chair for Alexion Pharmaceuticals, Inc.; and has participated in the Eculizumab in Adolescent Patients With Plasma Therapy-Resistant aHUS (C08-002; NCT00844844), Eculizumab in Adolescent Patients With Plasma Therapy-Sensitive aHUS (C08-003; NCT00844428), and Eculizumab in Pediatric Patients With Atypical Hemolytic-Uremic Syndrome (C10-003; NCT01193348) clinical studies for Alexion Pharmaceuticals, Inc. E.E.M. has participated in the C10-004 adult interventional study (NCT01194973) and in the C11-003 observational, follow-up study (NCT01522170) of atypical haemolytic uraemic syndrome patients for Alexion Pharmaceuticals, Inc. C.M. was a stockholder and employee of Alexion Pharmaceuticals, Inc. when the study was conducted. F.P. has received honoraria from Alexion Pharmaceuticals, Inc. E.R. has received fees for participation in advisory boards, experts' meetings and/or teaching courses from Alexion Pharmaceuticals, Inc. N.S.S. has received research funding from GlaxoSmithKline plc. J.W. is a stockholder and employee of Alexion Pharmaceuticals, Inc. L.E.W. has received fees for participation in advisory boards from Alexion Pharmaceuticals, Inc. L.A.G. has received research funding for Emory University from Alexion Pharmaceuticals, Inc., for his participation in the Eculizumab in Pediatric Patients and aHUS International Registry clinical studies; has received grant/research support and/or consultancy fees from AbbVie Inc., Alexion Pharmaceuticals, Inc., Bristol Myers Squibb, Advicenne Pharmaceuticals, Mallinckrodt Pharmaceuticals, Otsuka America Pharmaceutical, Inc. and Vifor Pharma; has served as a member of a scientific advisory board for Alexion Pharmaceuticals, Inc., and as a member of data safety monitoring boards for Retrophin, Inc. and Relypsa Pharmaceuticals. Y.D. has no relevant financial relationships to disclose.