Study design

A phase I, single-center, first-in-human, randomized, double-blind, placebo-controlled study of garadacimab in healthy male adult volunteers was carried out between October 28, 2016, and September 22, 2017, in Adelaide, Australia (ACTRN 12616001438448). The study design is outlined in Figure S1. Planned enrollment was six healthy male volunteers per each of the eight cohorts, with a total study population of 48 volunteers.

Garadacimab i.v. doses were 0.1, 0.3, 1, 3, and 10 mg/kg administered via infusion over 60 minutes. Garadacimab s.c. doses were 1, 3, and 10 mg/kg administered via s.c. injection into the abdomen. The doses in this study were selected based on data obtained from in vitro nonclinical pharmacologic and toxicology studies.

A total of six volunteers received garadacimab or placebo in a ratio of 2:1 in each cohort. Sentinel dosing was implemented for each i.v. cohort and the first (1 mg/kg) s.c. cohort, and enrollment was staggered between cohorts (Figure S1). The first two volunteers in each cohort undergoing sentinel dosing were randomized 1:1 to receive garadacimab or placebo (formulation buffer). If there were no safety concerns 48 h post-administration, the remaining four volunteers were randomized in a 3:1 ratio to receive garadacimab or placebo. Volunteers from each cohort were monitored for 14 days after drug administration; safety, PK, and select PD data were reviewed by the safety review committee before the next cohort was administered the next dose.

The study consisted of a screening period, a study drug administration period, and safety follow-up, which lasted 85 days after dosing. The protocol and amendment were approved by the independent Bellberry Human Research Ethics Committee, and the study was conducted in accordance with standards of Good Clinical Practice (as defined by the International Council for Harmonisation), ethical principles that have their origin in the Declaration of Helsinki, and all applicable national and local regulations. Written informed consent was obtained and documented for all volunteers before study participation.

Eligibility criteria

To be eligible for this study, volunteers had to be capable of providing written informed consent, willing and able to adhere to all protocol requirements, healthy, male, aged between 18 and 45 years inclusive, and with a body mass index of 18.0 to less than 30.0 kg/m2. Volunteers with a clinically significant medical condition, disorder, or disease, including clinically significant arterial or venous thrombosis, bleeding disorder, or any abnormal coagulation test result, were excluded from the study. Full inclusion and exclusion criteria can be found in the Supplementary Table S1.

Study population

The full analysis set (FAS) comprised all volunteers who provided informed consent and who were eligible for inclusion in the study after screening. The intention-to-treat (ITT) population comprised all volunteers in the FAS who were randomized. Overall analyses included i.v. and s.c. administration. The safety analysis set comprised all volunteers in the ITT population who received one dose of garadacimab. Safety analyses were overall and by route of administration and dose. The PK analysis set comprised all volunteers in the safety analysis set who received one dose of garadacimab and for whom at least one concentration value was reported. The PD analysis set comprised all volunteers in the safety analysis set for whom results were obtained for at least one of the exploratory biomarkers of aPTT or FXIIa-mediated kallikrein activity. PK/PD analyses were by route of administration.

Assessments

Safety was assessed by the frequency, severity, and relatedness of treatment-emergent adverse events (TEAEs) and the evaluation of changes in clinical laboratory tests, cardiac monitoring, vital signs, physical examination, and immunogenicity. Laboratory tests included hematology, biochemistry, coagulation, complement activity, and urinalysis. Cytokine activity was assessed, including interferon-γ, tumor necrosis factor-α, IL-6, IL-8, IL-13, IL­1β, IL-2, IL-4, IL-10, and IL-12p70. Immunogenicity was assessed as the proportion of volunteers with antidrug antibodies (ADAs) to garadacimab at the end-of-study visit.

A validated clinical assay for the detection, confirmation, and titration of ADAs was used. The method utilizes a bridging format where acid-dissociated samples are incubated in the presence of both biotinylated and sulfo-tagged garadacimab, which form complexes with any ADAs present. An anti-garadacimab antibody (4D3.11) isolated from a human antibody phage library at CSL Limited was used as a positive control. The complexes are captured onto a mesoscale discovery streptavidin plate, enabling an electrochemiluminescent signal to develop, which is proportional to the amount of ADA bound. The assay was validated according to guidance from the US Food and Drug Administration, European Medicines Agency, and Shankar et al. 2008.19 In the validation, sensitivity was determined to be 20 ng/ml with drug tolerance of 75 µg/ml for the high positive control (750 ng/ml), 50 µg/ml for medium (250 ng/ml), and low (100 ng/ml) positive controls. General acceptance criteria across performance parameters were set at less than or equal to 20% coefficient of variation. The validation exercise used a data set generated from the assay of 50 drug-naïve healthy donors to determine a screening assay cut point signal and likewise 50 donors were assayed with spiking of 100 µg/ml of garadacimab to determine the confirmatory assay cut point.

Blood samples for PK evaluation were collected during in-house days (at predose and 0.5, 1, 6, 12, 24, and 48 h post-i.v. administration, and at predose and 1, 6, 12, 24, 32, and 48 h post-s.c. administration) and up to 85 ± 5 days after the end of garadacimab administration. Garadacimab plasma concentrations were measured using a validated, clinical enzyme-linked immunosorbent assay. Garadacimab present in the samples was captured by a microtiter plate coated with an anti-garadacimab antibody (4D3.7) isolated from a human antibody phage library at CSL Limited. Captured garadacimab was detected with a mouse monoclonal anti-human lambda chain antibody conjugated with horseradish peroxidase (Abcam, Cambridge, MA, USA), which, when exposed to an enzymatic substrate, results in a color change directly proportional to the amount of garadacimab in the sample. The assay uses an 11-point (plus blank) calibration curve of 200–0.56 ng/ml. During validation of this assay, a lower limit of detection of 100 ng/ml and a minimal required sample dilution of 1:100 were defined. Validation parameters, such as precision (intra-assay, interassay, and interoperator), accuracy, dilutional linearity, selectivity, and robustness were defined within acceptance criteria limits of 20% for coefficient of variation and relative error (30% relative error for dilutional linearity). These criteria are applied to the validity of clinical test data.

PK assessments included the maximum plasma concentration (Cmax), area under the plasma concentration–time curve (AUC) from the time of dosing up to collection time t (AUC0–t), AUC extrapolated to infinity (AUC0–inf), time to maximum concentration in plasma, terminal elimination half-life (t1/2), total systemic clearance (CL for i.v. infusion and CL/F for s.c. injection), volume of distribution during the elimination phase (Vz for i.v. infusion and Vz/F for s.c. injection), and drug bioavailability after s.c. dosing. The acceptable percentage AUC that was extrapolated for the AUC0-inf calculation was less than 20%, as per an internal standard operating procedure. Dose proportionality was assessed separately for i.v. and s.c. doses for the PK parameters Cmax, AUC0–inf, and AUC0–t. Dose proportionality was explored with the Hummel power model analysis.20

Exploratory biomarkers of the PD effects of garadacimab were assessed, including aPTT and FXIIa-mediated kallikrein activity. The aPTT was measured by standard methodology using the Siemens system with Dade Actin reagent as activator of intrinsic coagulation. To measure FXIIa-mediated kallikrein activity, 40 µl of sodium citrate plasma was mixed with 10 µl of a reaction buffer containing 10 mM HEPES, 137 mM NaCl, 4 mM KCl, 11 mM d-glucose, pH 7.4, and 40 µl of the same buffer containing dextran sulphate MW40,000 to a final assay concentration of 2.5 µg/ml in 96-well microtiter plates. The plate was placed in a 37°C warmed reader and read for 3 min before addition of 10 µl chromogenic substrate H-D-Pro-Phe-Arg-pNA•2HCl (S-2302) to 0.6 mM. Amidolytic activity, which reflects the kallikrein activity generated, was monitored in the sample by absorbance at 405 nm over 30 min and the maximal rate of cleavage derived from the kinetic curves.

Statistical analysis

Dose proportionality was assessed for Cmax, AUC0–inf, and AUC0–t for the garadacimab i.v. and s.c. doses separately. Exploratory dose proportionality was analyzed using a power model by the Hummel et al. method.20 After ln-transformation of the power model, the slope and its 90% confidence interval (CI) were estimated with a linear mixed-effect model in SAS (PROC MIXED). When the 90% CI for the slope is within the predefined critical interval, based on the ratio between the highest and lowest dose levels of garadacimab, dose proportionality can be declared. The critical interval to declare dose proportionality was 0.85–1.15 for the 90% CI of the slope estimate for i.v. infusion PK parameters and 0.7–1.3 for the 90% CI of the slope estimate for s.c. administration.

The bioavailability of s.c. garadacimab relative to that of i.v. garadacimab infusion was assessed. Comparisons were completed using an analysis of variance model and log transformed PK parameters AUC0–inf/total dose. The geometric mean ratio of dose normalized AUC0–inf of garadacimab administered s.c. to i.v. was estimated by exponentiating the difference in least-squares mean, resulting in absolute bioavailability with corresponding 90% CI. Eight subjects were included in the analysis of each dose and 24 subjects were included in the analysis of all doses. SAS version 9.3 was used for statistical analysis. All statistical tests were exploratory and nonconfirmatory. No formal hypothesis was tested in this study, and all results were descriptive. The sample size was considered appropriate for assessing safety.