Pharmacokinetics and safety of roledumab, a novel human recombinant monoclonal anti-RhD antibody with an optimized Fc for improved engagement of FCγRIII, in healthy volunteers

Authors


Daniel Quagliaroli, Development Program Department, LFB Biotechnologies, 3, avenue des Tropiques, BP 50052, 91 942 Courtaboeuf Cedex, France
E-mail: quagliaroli@lfb.fr

Abstract

Background and Objectives  A human recombinant monoclonal anti-RhD IgG may be useful to prevent RhD allo-immunization. Roledumab is such an antibody with a glycosylation pattern optimized for biological activity. The objective of the study was to assess the safety and pharmacokinetics of roledumab in healthy RhD-negative volunteers.

Materials and Methods  A total of 46 subjects received doses of 30–3000 μg i.v. of roledumab or placebo using a double-blind escalating single-dose design; 12 of these subjects also received 300 μg i.m. of roledumab. Subjects were followed for 6 months after administration. Serum roledumab concentrations were determined using flow cytometry.

Results  Fourteen treatment-emergent adverse events related to treatment were reported in nine subjects, with no apparent difference in their frequency or nature after placebo or roledumab administration. No anti-roledumab antibodies were detected. AUClast increased from 4·4 ng/ml.day at 30 μg i.v. to 2257 ng/ml.day at 3000 g i.v. The t½ ranged from 18 to 22 days, and the absolute bioavailability after i.m. administration was between 73% and 80%.

Conclusion  Roledumab is safe and well tolerated in healthy RhD-negative volunteers and shows a pharmacokinetic profile similar to that of polyclonal anti-RhD immunoglobulin.

Introduction

Haemolytic disease of the foetus or newborn (HDFN) is a potential fatal immune consequence of the passage of anti-RhD antibodies from the mother to the foetus across the placenta [1, 2]. When the woman is RhD-negative and the foetus RhD-positive, the woman may develop an allo-immunization to the RhD antigen and produce anti-RhD antibodies. During a subsequent pregnancy with an RhD-positive foetus, higher quantities of anti-RhD antibodies will be produced, cross the placenta and can cause HDFN.

In the 1960s, prophylaxis against RhD allo-immunization was introduced, and this has reduced foetal and neonatal morbidity and mortality due to HDN dramatically [3–5]. This prophylaxis consisted of administrating human plasma-derived anti-RhD immunoglobulin to RhD-negative women after delivery of an RhD-positive infant, and this was later extended to the antenatal period by administration at the 28th week of pregnancy. This immunoglobulin recognizes any RhD-positive RBCs entering the maternal circulation and triggers their destruction through a cell-mediated immune response before they can stimulate proliferation of RhD-reactive B cells, the key process underlying allo-immunization. Post-natal and antenatal anti-RhD immunoprophylaxis practice reduced the incidence of allo-immunization in RhD-negative women from around 13% to around 0·35% [6].

Currently marketed human polyclonal anti-RhD immunoglobulins are produced by the fractionation of pooled plasma obtained from RhD-negative plasma donors. Due to the limited number of naturally immunized donors, active immunization of RhD-negative volunteers is required to produce anti-RhD immunoglobulins. Nonetheless, this practice presents concerns [7,8] and is no longer practiced in most West European countries. As a consequence, the great majority of polyclonal anti-RhD immunoglobulins used in Europe today are imported from the United States and obtained from paid donors.

In this context, the development of a human recombinant monoclonal anti-RhD IgG may present advantages, since it would ensure virtually unlimited supplies of a standardized and safe product without the practical and ethical barriers associated with preparation of immunoglobulins from human plasma.

Although numerous attempts have been made to develop recombinant monoclonal anti-RhD antibodies, [9] many of these have failed to show robust functional activity in antibody-dependent cellular cytotoxicity (ADCC) assays in vitro, and only a limited number have been evaluated in Phase II studies in human [10–12]. Several clinical trials designed to demonstrate that such antibodies can promote clearance of RhD-positive RBCs effectively in healthy RhD-negative volunteers have been inconclusive [9]. One possible explanation for this is that the low affinity of many of these antibodies for the FcγRIII receptor (CD16), whose activation is necessary for inducing ADCC [9, 13, 14], may compromise their efficacy [15]. One determinant of affinity for FcγRIII is the glycosylation pattern of the immunoglobulin heavy chains, which may be determined by the nature of the host cell in which the immunoglobulin is expressed. In this respect, a glycosylation pattern with a low fucose content confers on the immunoglobulin enhanced ADCC [9, 16].

In support of this notion, LFB-R297, a first generation human recombinant monoclonal anti-RhD antibody [17], has demonstrated a high level of activity in FcγRIII-dependant ADCC in vitro assay and was subsequently shown to promote a similar clearance of antibody-coated RhD RBCs in healthy volunteers to the reference polyclonal anti-RhD immunoglobulin, Rhophylac®.

In this context, roledumab, a novel human recombinant monoclonal anti-RhD antibody derived from LFB-R297, has been developed. As a first-in-man study, in those subjects devoid of the molecular entity targeted by the antibody, a Phase I study in healthy RhD-negative volunteers has been initiated. The objective of the study were to assess the safety and pharmacokinetics of roledumab after intravenous and intramuscular administration.

Methods

The study was a Phase I, double-blind, randomized, placebo-controlled study in 46 healthy RhD-negative volunteers, performed in a single centre in France (Biotrial, Rennes) between October 2008 and December 2009. An escalating single-dose design was used to evaluate safety and pharmacokinetics in five groups of subjects receiving roledumab from 30 to 3000 μg or placebo by intravenous administration and a sixth group receiving roledumab 300 μg administered by the intramuscular route. In order to evaluate the intramuscular bioavailability, a two-way cross-over design was used in which, after 6 months, the group that had originally received roledumab by the intramuscular route received an identical dose by the intravenous route and vice versa.

Ethics

The study was performed according to European and French regulatory guidelines and current codes of Good Clinical Practice. Each subject was informed about the nature of the trial, its aim, its possible risks, its duration and the financial compensation they would receive. All subjects freely gave their written and informed consent prior to inclusion. The study protocol, the subject information sheet and the informed consent form were submitted to, and approved by, the relevant ethics committee (CPP Ouest V, Rennes) prior to the start of the study. Data handling for the study was authorized by the Commission Nationale d’Informatique et de Libertés (CNIL), a French-independent authority that ensure that all collected medical information is kept confidential and anonymous.

Subjects

Healthy men and women aged between 18 and 60 years with a body mass index between 18·5 and 30 kg/m2 were eligible. All subjects were tested for RhD status and only RhD-negative subjects included. Women of child-bearing potential were requested to use an effective method of contraception. Exclusion criteria included a history of immune disorders, a history of severe systemic reactions to immunoglobulins (e.g. anaphylaxis), hepatic, renal, endocrine or metabolic disorders, concomitant drug treatment, use of immunomodulatory drugs in the previous 3 months, alcohol abuse or a positive drug screen, and previous administration of antibodies in the previous year.

Preparation of roledumab and placebo

Roledumab is a fully human IgG1 targeted against the human RhD antigen on RhD-positive circulating RBCs. The genes coding for a human IgG1 have been isolated from the lymphocytes of a donor previously immunized with RhD-positive RBCs and transfected into a rat cell line YB2/0 selected for the optimized effector properties it confers on the antibody [18]. The original primary structure of this antibody (LFB-R297) was then modified by replacing a free cysteine by a phenylalanine at position 68 of the heavy chains in order to avoid changes in the 3D structure due to potential interactions between Cys68 and Cys22 or Cys96. This monoclonal anti-RhD antibody contains an optimized glycosylation profile characterized by a low fucose content which confers high ADCC activity. Roledumab demonstrates a similar high affinity for the RhD antigen to polyclonal anti-RhD antibodies [19]. This monoclonal anti-RhD antibody recognizes the epD 13·1 RhD epitope (Scott’s nomenclature) [20, 21]. The anti-RhD potency of roledumab was determined by the flow cytometry method described in the European Pharmacopoeia 5·0 section 2·7·13 using FACS Array equipment that measures antibody affinity from its binding to RhD-positive RBCs. In this method, OR1R1 RBCs were incubated with goat Fab anti-human IgG labelled with the fluorochrome FITC and median fluorescence intensity (MFI) was measured. The assay was calibrated using the international standard for anti-RhD immunoglobulins [preparation 01/572 from the National Institute for Biological Standards and Control (NIBSC)]. The specific anti-RhD activity of roledumab measured this way was 7·6 IU/μg with a concentration of antibodies determined using calculated molar epsilon. This ratio is slightly higher that generally accepted for polyclonal anti-RhD immunoglobulins (5 IU/μg). However, previous authors have shown that the potency of monoclonal anti-RhD antibodies measured by flow cytometry could be higher than that determined by other methods described in the European Pharmacopeia [22].

The manufacturing process of roledumab involved large-scale serum-free culture in bioreactors, three chromatographic purification steps, an ultra-filtration step and two specific viral inactivation steps. Extensive ‘in process’ control as well as quality control of the finished product were implemented to ensure identity, potency, sterility and purity of the antibody. Material for clinical use was formulated as a sterile-buffered solution and was supplied in 3·3 ml single use glass vials at a concentration of 300 μg of anti-RhD antibodies per millilitre.

The placebo was formulated using the formulation components of roledumab, namely sodium chloride, tri-sodium citrate dehydrate, polysorbate 80, mannitol and water for injection. The placebo was supplied in 3·3 ml single use glass vials.

Treatment and follow-up

After validation of the entry criteria, subjects were assigned consecutively to one of the six groups. In Groups 1–5, eight subjects were randomized to receive placebo or one of five increasing doses of roledumab administered by the intravenous route (Table 1). The first dose administered as well as the escalation ratio were in accordance with the NOAEL approach of the EMA guidelines [23]. In Group 6, six subjects received roledumab 300 μg administered by the intramuscular route. In order to evaluate the absolute bioavailability, subjects in Groups 3 and 6 received a second administration of roledumab at the end of this 6-month period. Subjects in Group 3 received their second treatment by the intramuscular route and those in Group 6 by the intravenous route. Subjects were hospitalized the day before treatment and for 48 h thereafter.

Table 1.   Description of treatment groups
 Dose escalation phaseCross-over phase
  1. aThe two subjects who had received placebo during the dose escalation phase did not participate in the cross-over phase.

Group 1Placebo: 2 subjects
Roledumab 30 μg i.v.: 6 subjects
Not applicable
Group 2Placebo: 2 subjects
Roledumab 100 μg i.v.: 6 subjects
Not applicable
Group 3Placebo: 2 subjects
Roledumab 300 μg i.v.: 6 subjects
Roledumab 300 μg i.m.: 6 subjectsa
Group 4Placebo: 2 subjects
Roledumab 1000 μg i.v.: 6 subjects
Not applicable
Group 5Placebo: 2 subjects
Roledumab 3000 μg i.v.: 6 subjects
Not applicable
Group 6Roledumab 300 μg i.m.: 6 subjectsRoledumab 300 μg i.v.: 6 subjects

In order to identify potential safety issues and thus minimize potential risks of treatment to the subject, subjects in Groups 1 and 2 were treated individually at 24-h intervals and subjects in Groups 3–5 were treated in groups of three or two at 48-h intervals. Dose escalation from one group to another was only authorized when all subjects had completed the lower dose group. At each dose, all safety data were reviewed at least 48 h after the last dosing by the study investigator and the sponsor. In addition, an independent Data Safety Monitoring Board had authority to stop the trial immediately should any safety issue arise.

Following treatment, all subjects were followed up for 6 months for safety evaluation including the immunogenicity of roledumab. Subjects in Groups 3 and 6, who received two administrations of treatment, were followed up for 6 months after each treatment.

Safety assessment

Safety was assessed by physical examination and ECG, vital signs and laboratory safety parameters. At each follow-up visit throughout the 6-month study period, any adverse events were documented from spontaneous and prompted report by the subject and clinical observation by the investigator. Laboratory safety assessment included standard haematology, blood biochemistry and urinalysis. Any clinically significant abnormalities in laboratory parameters, vital signs and physical examination were to be reported as adverse events. In addition, immune function was evaluated by measuring serum levels of certain cytokines, on the human Th1/Th2 11 plex FlowCytomix Multiplex platform (Bender Med Systems). The following cytokines were determined: interleukin-1b (IL-1b), interleukin-2 (IL-2), interleukin-4 (IL-4), interleukin-5 (IL-5), interleukin-6 (IL-6), interleukin-8 (IL-8), interleukin-10 (IL-10), interleukin-12 (IL-12), tumour necrosis factor-α (TNF-α), tumour necrosis factor-β (TNF-β) and interferon-γ (IFN-γ), complement components (C3 and C4) and elastase.

Particular attention was paid to the potential immunogenicity of roledumab. The presence of anti-roledumab antibodies was evaluated in serum samples taken before treatment and at 6 months thereafter. A specific quantitative enzyme-linked immunosorbent assay (ELISA) was developed for this purpose by the study sponsor, which conformed to FDA [24] and EMEA [25] guidelines on the assessment of the immunogenicity of biotechnology-derived therapeutic proteins. Roledumab was pre-coated onto a microplate. Clinical samples were pipetted into the wells and anti-roledumab antibodies bound to the immobilized roledumab. After washing away any unbound antibodies, biotinylated roledumab was added to the wells and bound to the free paratopes of the anti-roledumab antibodies. Following a wash to remove any unbound roledumab-biotin reagent, streptavidin-horseradish peroxidase (HRP) was added into the wells and bound to the immobilized biotin. After washing away any unbound streptavidin-HRP, the peroxidase bound to the complex was developed using 3,3′,5,5′-tetramethylbenzidine (TMB) as substrate and determined by spectrophotometry. The colour intensity on the spectrophotometer plate was proportional to the concentration of anti-roledumab antibodies. Calibration standards of mouse monoclonal anti-roledumab antibody were prepared in phosphate-buffered salt (PBS) containing 1% bovine serum albumin (BSA). The assay cut-off was 0·25 OD units.

PK assessment: determination of serum levels of roledumab

Blood samples were collected at baseline and at 10, 20 and 30 min and 1, 3, 6, 12, 24, 36, 48 h after intravenous administration and thereafter, 3, 4, 5, 7, 14, 21, 28 and 42 days after dosing and then once a month until 6 months. After intramuscular administration, the same schedule was used with the exception of the 10-, 20- and 30-min time-points. Serum samples were frozen at <−70°C at the investigational site before transfer to the central laboratory for analysis.

Serum roledumab concentrations were determined using flow cytometry. The assay fulfilled the methodological criteria recommended by current bioanalytical guidelines [26, 27]. RhD-positive RBCs were incubated in 96-well microtitre plates containing 50 μl of the calibration standards, quality controls or serum samples at a final density of 2 × 106 RBCs/well for 2 h at 37°C. Calibration standards and quality controls were prepared with the same clinical batch of roledumab that was administered to the healthy volunteers. At the end of the incubation period, 200 μl of PBS containing 1% bovine serum albumin (BSA) was added to the cells, which were then washed three times by centrifuging the microtitre plate at 770 g. After the final wash, the packed RBC pellet was resuspended with 50 μl/well of a 1:20 dilution of a human Fab anti-IgG antibody coupled to fluorochrome fluorescein isothiocyanate (FITC). The microtitre plate was mixed on a plate shaker for 20–30 min at room temperature, protected from light. After three further rounds of washing, erythrocytes tagged by the fluorochrome were then quantified using flow cytometry. The lower limit of quantification of the method was 3·87 ng/ml.

Sample analysis was performed on a Cytomics FC500 flow cytometer (Beckman Coulter Ltd, Brea, CA, USA). Primary data acquisition for the analysis was collected using a log-scaled forward and side scatter (Log FS/SS) histogram, to detect the RBC population. Antibody binding was determined by measuring the X-mean of the FITC-labelled RBCs on a single-parameter log-scaled histogram for detector FL1, set to collect 50 000 events. The amount of anti-RhD in samples was calculated by single-point determination reading from the assay calibration curves. Each serum sample was analysed in triplicate, the final reported result being the arithmetic mean of these triplicate values.

Data analysis

A total of 46 subjects were planned to be included in the study in order to assess the safety and pharmacokinetics of roledumab. No statistical hypotheses were tested, and no formal sample size calculation was performed.

The safety analysis included a description of incidence and severity of adverse events, vital signs, ECG parameters, physical examination and laboratory data. The safety analysis was performed on all treated subjects and displayed by dose group and by route of administration.

The pharmacokinetic analysis was performed for all included subjects, with the exception of one subject in Group 4, for whom roledumab was not detected during the first 12 h after intravenous administration, presumably due to a dosing error. Two subjects who withdrew from the study prematurely before administration of the second dose of roledumab did not provide bioavailability information. The following PK parameters were determined: maximal serum concentration (Cmax) and time elapsed from dosing to Cmax (tmax), apparent terminal-phase rate constant (λz), t½, AUClast and AUC, mean residence time (MRT), clearance (CL) and volume of distribution at steady state (Vdss). Computation of the pharmacokinetic parameters was performed using WinNonlin® software (Pharsight, St Louis, MO, USA). Cmax and tmax were determined directly from the concentration–time data, λz was estimated by linear regression of the terminal phase of the logarithmically transformed concentration versus time data, and t½ was computed as = (ln 2) × inline image. This linear regression was only performed when three or more data points were available for the terminal part of the log-linear decline in serum levels of roledumab. AUClast was calculated by a combination of linear and logarithmic trapezoidal methods. The linear trapezoidal method was employed up to Cmax, and the logarithmic trapezoidal method was used thereafter. AUC was computed from AUClast as AUC = AUClast + (Clast × inline image). Clearance was computed as CL = dose × inline image, MRT as MRT = AUMC × (AUC)−1 − T/2, where T is the infusion duration and Vdss as Vdss = CL × MRT.

A compartmental PK analysis was performed using a mixed-effect model with NONMEM® software (ICON Development Solutions, Dublin, Ireland) to evaluate the population IM bioavailability and to investigate the dose proportionality of the IV exposure. The population PK model was qualified using visual inspection of goodness-of-fit plots and a visual predictive performance check using Monte Carlo simulations of the population PK data set under the final model.

Results

Subjects

A total of 46 healthy RhD-negative volunteers were enrolled in the study. Two subjects in Group 6 were prematurely withdrawn from the study, one due to the occurrence of a non-study drug-related Helicobacter pylori infection that required treatment and the other as a result of consent withdrawal due to a change in professional circumstances. The mean age of the subjects was 41 years (range: 22–59), and the majority were female (71·7%). The mean body mass index [23·6 kg/m2 (range: 19·2–30·1 kg/m2)] was in the normal range, as shown in Table 2.

Table 2.   Description of subjects included
 Placebo
N = 10
30 μg
N = 6
100 μg
N = 6
300 μg
N = 12
1000 μg
N = 6
3000 μg
N = 6
Total
N = 46
Age (years)
 Mean ± SD
43·0 ± 9·442·0 ± 15·739·0 ± 12·141·6 ± 11·542·3 ± 11·938·3 ± 15·141·3 ± 11·7
Sex
 Males:Females
1:92:43:34:83:30:613:33
Height(cm)
 Mean ± SD
163 ± 6169 ± 9166 ± 9174 ± 13160 ± 4163 ± 5166 ± 9
Weight (kg)
 Mean ± SD
63·6 ± 10·160·8 ± 6·467·0 ± 12·567·1 ± 15·374·7 ± 16·158·2 ± 6·165·3 ± 12·5
BMI (kg/m2)
 Mean ± SD
23·8 ± 3·122·9 ± 2·323·3 ± 2·624·1 ± 3·124·3 ± 2·422·8 ± 2·523·6 ± 2·7
Blood group
 A42232316
 B1013308
 AB0020002
 O54161320

Safety

A total of 117 treatment-emergent adverse events were documented over the course of the study in a total of 36 subjects (78·3%). There was no obvious difference in either the frequency or the nature of the adverse events between subjects receiving placebo and those receiving roledumab, between the different doses of roledumab administered or between the intramuscular and intravenous route of administration (Table 3). Most treatment-emergent adverse events (N = 103; 84·8%) started at least 48 h after administration of the study medication.

Table 3.   Adverse events documented during the study
 Placebo i.v.
N = 10
30 μg i.v.
N = 6
100 μg i.v.
N = 6
300 μg i.v.
N = 10
1000 μg i.v.
N = 6
3000 μg i.v.
N = 6
300 μg i.m.
N = 12
EventsSubjects (%)EventsSubjects (%)EventsSubjects (%)EventsSubjects (%)EventsSubjects (%)EventsSubjects (%)EventsSubjects (%)
  1. AE, adverse event; TEAE, treatment-emergent adverse event; SAE, serious adverse event.

Any AEs279 (90)194 (67)84 (67)217 (70)124 (67)156 (100)187 (58)
Any TEAEs279 (90)184 (67)84 (67)217 (70)124 (67)135 (83)187 (58)
Any SAEsNone11 (17)NoneNoneNoneNoneNone
Any drug-related AEs74 (40)21 (17)11 (17)21 (10)None22 (33)None
Any TEAEs within 48 h53 (30)21 (17)11 (17)32 (20)None22 (33)11 (8)

One serious adverse event, which was not considered to be related to roledumab, was documented. One subject administered roledumab 30 μg i.v. experienced a thoraco-abdominal trauma caused by a kick from a horse more than 3 months after the study drug administration. With the exception of this serious adverse event, all adverse events reported were considered to be of mild or moderate severity. Fourteen treatment-emergent adverse events potentially related to treatment were reported in nine subjects. These events are listed in Table 4.

Table 4.   Treatment-related adverse events documented during the study
GroupSubjectEventSeverity
PlaceboN° 20HeadacheModerate
HeadacheModerate
Muscle spasmMild
N° 51Orthostatic hypotensionModerate
Orthostatic hypotensionModerate
N° 18PruritusMild
N° 48Pityriasis roseaMild
30 μg i.v.N° 5HeadacheMild
Pain in extremityModerate
100 μg i.v.N° 17Maculopapular rashMild
300 μg i.v.N° 38Orthostatic hypotensionModerate
Vasovagal syncopeModerate
1000 μg i.v.N° 64Orthostatic hypotensionModerate
N° 70HeadacheMild

No clinically significant change in vital signs was detected, with the exception of mild hypertension (148/89 mm Hg) in one subject receiving roledumab 100 μg whose blood pressure was borderline at the inclusion visit. No clinically relevant changes potentially related to treatment were observed in the evolution of any haematological or biochemical variable during the study. The investigated immunology parameters (C3, C4, TNF-α, TNF-β and interleukins) did not indicate any immune response to roledumab.

No human anti-roledumab antibodies were detected 6 months after administration of study medication in any subject.

Pharmacokinetics

The pharmacokinetic parameters determined in this study are presented in Table 5. After intravenous administration, the mean maximal serum concentrations (Cmax; geometric mean) of roledumab increased from 6 ng/ml at the dose of 30 μg to 1904 ng/ml at the dose of 3000 μg. Over this dose range, the mean serum exposure to roledumab, determined as AUClast, increased from 4·4 to 22 557 ng/ml.day. The time–course of serum concentrations of roledumab after intravenous administration of different doses is shown in Fig. 1.

Table 5.   Pharmacokinetic parameters of roledumab in healthy subjects
 DoseNon-compartmental analysisDerived from population PK model
N Median [Range]Geometric mean N Estimate
Intravenous administration
 Cmax (ng/ml)3036·5 [5·7–6·7]6·3  
100628·4 [18·2–37·0]26·0
30010144·1 [101·2–192·9]144·5
10005302·6 [250·5–464·7]335·3
300062009·0 [1490·7–2250·5]1904·2
 AUClast (ng/ml.day)3034·5 [2·3–8·3]4·4  
1006193 [104–285]186
30010949 [529–1777]995
100053715 [3295–6781]4299
3000623 441 [18 530–26 294]22 557
 t½ (days)30 620·5
100 618·5
300 1218·0
1000418·5 [14·8–21·9]18·3619·9
3000622·8 [17·6–24·6]21·4621·5
 AUC (ng/ml.day)30 6101
100 6345
300 121106
100044764 [3458–6977]479764535
3000623 713 [18 705–26 553]22 777620 868
Intramuscular administration
 tmax (h)30012228 [36 – 504]  
 Cmax (ng/ml)3001222·28 [14·70–48·90]24·52  
 AUClast (ng/ml.day)30012673 [350–1637]727  
Figure 1.

 Time–course of the serum concentrations of roledumab after intravenous administration of different doses. Open circles: 3000 μg; open squares: 1000 μg; filled circles: 300 μg; filled squares: 100 μg.

A number of PK parameters could not be estimated when serum concentrations were found to be below the lower limit of quantification of the assay (<3·87 ng/ml). No Cmax could be estimated for three subjects administered 30 μg roledumab. No elimination half-life could be estimated for subjects administered 30, 100 or 300 μg roledumab. At the two highest doses, terminal elimination half-lives were comparable with median values of approximately 19 days at the 1000 μg dose and 23 days at the 3000 μg dose.

After administration by the intramuscular route, absorption of rodelumab 300 μg was quite slow, as expected for a human immunoglobulin, with a median tmax of 228 h, corresponding to around 10 days. Compared with the intravenous route, the maximal serum concentrations and serum exposure achieved were lower after intramuscular administration (Table 4). However, beyond 18 days, serum concentrations of roledumab were comparable or even slightly higher after intramuscular administration than after intravenous injection (Fig. 2).

Figure 2.

 Time–course of serum concentrations of roledumab after intravenous (□) or intramuscular (○) administration at the dose of 300 μg.

Under the assumption of a common PK model for all doses and for both intravenous and intramuscular routes, the population PK model provided individual Bayesian estimates from which individual parameters for exposure and elimination half-life at all doses were derived (Table 4). The terminal elimination half-life was estimated to be between 18 and 22 days at doses ranging from 30 to 3000 μg, and this is consistent with the values estimated for the higher doses from the individual subject non-compartmental approach. After intramuscular administration, the estimated terminal elimination half-life was 18 days. The mean serum exposure (AUC) predicted by the population pharmacokinetic model ranged from 101 ng/ml.day at the 30 μg dose to 20 868 ng/ml.day at the highest dose of 3000 μg.

An evaluation of the relationship between ln(Cmax) and ln(AUClast) and ln(dose) revealed that exposure increased more than proportionally to the dose over the dose range of 30–3000 μg roledumab (Fig. 3). For Cmax, the slope was 1·21 (90% CI: 1·14; 1·28) and for AUClast, the slope of the linear regression line between ln(AUClast) and ln(Dose) was 1·65 (90% CI: 1·51; 1·79).

Figure 3.

 Relationship between intravenous dose of roledumab administered and serum exposure in terms of Cmax (left) and AUClast (right). The open circles represent the parameter obtained for each individual and the solid line the best fit to a log–log regression model.

The absolute IM bioavailability determined by the ratio of the geometric mean of AUClast over 42 days after intramuscular and intravenous administration of 300 μg roledumab could be estimated at around 73%. The population PK model estimate of the absolute bioavailability was comparable at 80%.

Discussion

The primary objective of this study was to evaluate the safety of roledumab in healthy RhD-negative volunteers. No safety issue was identified after intravenous administration of doses ranging from 30 to 3000 μg. Following intramuscular administration of the 300 μg dose, no specific local injection site reactions were reported. The nature and frequency of the adverse events documented were similar in subjects receiving roledumab, whatever the dose or route of administration, and in subjects receiving placebo. No dose relationship was observed in the overall or individual frequency of adverse events. The only serious adverse event that was documented (a thoraco-abdominal trauma caused by a kick from a horse) was clearly unrelated to administration of roledumab. Adverse events considered potentially related to treatment were mild or moderate in severity also evenly distributed between subjects receiving placebo and roledumab, the most frequently reported of these being headache and orthostatic hypotension.

Medicinal products of biological origin present specific potential safety risks. In particular, idiosyncratic species-specific immune reactions may occur, which may be serious, such as the unanticipated severe anaphylactic reaction observed in a Phase I study of an anti-CD28 monoclonal antibody [28]. In this study, care was taken to space the dosing between individual subjects to limit the number of subjects exposed to roledumab in case of unexpected reactions. In the study, no adverse event or clinical sign suggestive of an immune reaction was observed. We also measured serum levels of a panel of 14 cytokines and other biochemical markers of immune function, and no abnormality was detected. Another potential problem with biological agents is that they may stimulate the formation of host antibodies that may neutralize their anticipated biological activity, potentially trigger a severe hypersensitivity reaction or lead to the development of autoimmunity through the activity of antibodies that recognize the endogenous form of the target protein [25]. In this study, no anti-roledumab antibody was detected in any of the 36 treated subjects 6 months after treatment.

The secondary objective of the study was to determine the pharmacokinetic characteristics of roledumab in human. We found that serum exposure to roledumab after intravenous administration was acceptable, the AUC reaching >20 000 ng/ml.day at the highest dose of 3000 μg with a terminal elimination half-life of 18–22 days. Interindividual variability was relatively low (<25%).

The terminal elimination half-life obtained can be compared to that observed in a previous study of the pharmacokinetics of purified polyclonal anti-RhD immunoglobulin after intramuscular administration in healthy volunteers [29], which was 23 days. The estimates of AUClast and t½ were also close to those reported for polyclonal IgG in pregnant RhD-negative women [30], namely 1015 ng/ml.day and 16 days at the dose of 300 μg i.v. This suggests that the specific glycosylation profile of rodelumab does not modify the overall PK profile of the antibody compared to a natural IgG1. In contrast, for another recombinant monoclonal anti-RhD antibody, MonoRho (CSL Behring) [31], serum exposure was less extensive (AUC of 216 ng/ml.day at the dose of 300 μg i.v.) and bioavailability lower (46%), which may account for the poorer performance of MonoRho in clearing RhD-positive RBCs compared to polyclonal IgG [14].

One of the difficulties encountered in the study was the quantification of serum roledumab levels at low doses. To overcome this issue, a population pharmacokinetic analysis was performed to determine AUC and t½ in subjects administered the lower doses. To our knowledge, the dose linearity of exposure to anti-RhD immunoglobulins has not been investigated previously. We found that serum exposure to rodelumab was dependent on the administered dose, but that the dose–response relationship departed somewhat, although slightly, from linearity. This may in part be due to the difficulties in estimating serum rodelumab concentrations at low doses.

In conclusion, roledumab was safe and well tolerated in doses up to 3000 μg i.v. and at 300 μg i.m. in healthy RhD-negative volunteers, which is 10 times the anticipated therapeutic dose, and shows a pharmacokinetic profile similar to that of purified human polyclonal anti-RhD immunoglobulin. Taken together with its robust efficacy in in vitro assays of cytotoxicity towards RhD-positive RBCs, the current data suggest that roledumab can be seen as a promising alternative to polyclonal anti-RhD immunoglobulin for preventing pregnancy-related allo-immunization of RhD-negative women and potentially the first monoclonal antibody to match polyclonal anti-RhD antibody. A Phase II study of roledumab is currently in progress to assess whether roledumab can promote clearance of RhD-positive RBCs in healthy RhD-negative volunteers.

Acknowledgements

The authors would like to thank all the volunteers who participated in this study as well as the staff of Biotrial for their professionalism in the implementation of the study. They are also grateful to the members of the Data and Safety Monitoring Board, Pr Bruno Carbonne, Pr Hervé Watier and Dr Pierre Guéret, for their time and their commitment to the smooth conduct of the study. The authors would like to express their thanks to Dr Yong-Un Kim, Dr Beatrice Perron and Dr Tony Waegemans for their contribution to the design and implementation of this study.

Ancillary