Alan Faraj1, Malte Selch Larsen1, Judi Møss2, Kristoffer Winther Balling1
1Pharmacometrics Novo Nordisk A/S, 2Clinical Pharmacology Novo Nordisk A/S
Objectives: Concizumab (CZM) is a humanized recombinant monoclonal antibody targeting tissue factor pathway inhibitor (TFPI) and is in development as a subcutaneous treatment for prevention of bleeds, including long-term prophylaxis in patients with haemophilia A/B, with and without inhibitors. Population pharmacokinetic (PK) and pharmacodynamic (PD) modelling has previously shown that CZM exhibits nonlinear PK with reduction in annualised bleeding rate with increasing CZM plasma concentrations. This work was conducted to support the dose selection of CZM based on modelling of accumulated data from phase 1-3 trials across healthy and haemophilia subjects. Methods: PK (measured with the clinical trial exposure assays), PD (free TFPI not in complex with CZM, and thrombin peak from ex vivo thrombin generation assay) and bleeding event data from eight clinical trials were pooled and used for model building. Direct Emax models were used to characterize the exposure-response based on free TFPI and thrombin peak levels whereas a repeated time-to-event approach was used to model the relationship between CZM concentrations and occurrence of bleeding events over time. Based on the developed models, clinical trial simulations were performed to study the performance of the proposed dosing regimen. Following a single loading dose of 1 mg/kg, the initial daily dose was 0.2 mg/kg. After at least 4 weeks on 0.20 mg/kg, the CZM concentration in plasma was measured and used to determine the individual daily maintenance doses for the patients at either 0.15, 0.20 or 0.25 mg/kg. The maintenance dose was set to 0.25 or 0.15 mg/kg if the CZM plasma concentrations were <200 or >4000 ng/ml on week 4, respectively. Any CZM levels in between led to continuation of the 0.2 mg/kg dose. The same individuals were also simulated without dose adjustment for comparison. In addition, simulations were performed to study resuming dosing (single or double maintenance dose or loading dose of 1 mg/kg) after up to seven missed doses. Furthermore, to the clinical trial simulations, predictions of exposure levels were generated with and without dose adjustment for the subjects included in the dataset used for model-building. The cutoff values of 200 and 4000 ng/mL for dosing adjustment were derived based on previous exposure-response analyses and preclinical toxicology data, showing desirable reduction in bleed rate and reducing risk of occurrence of adverse events. R version 4.2.0 was used for data handling and post-processing of results whereas NONMEM 7.5 [1] and Perl-Speaks-NONMEM 4.69 [2] was used for modelling and simulation. Results: A typical subject was found to reach above 95% of steady state concentrations already after four days of 0.2 mg/kg dosing following a loading dose of 1 mg/kg. The simulated dosing regimen reduced the variability in CZM exposure with a higher number of patients falling within the target levels using dosing adjustment as compared to no adjustment of the dose. The 90% prediction interval of average concentrations were 210-4000 and 170-8900 ng/mL with and without dosing adjustment, respectively. The decrease of CZM exposure in the subjects with exposures above 4000 ng/ml following dose reduction only led to 1%-point increase in bleeding risk compared to not reducing the dose. Increase of the dose in the subjects with lower exposure led to a reduction in bleeding risk by 5%-points compared to no dose adjustment. Simulations showed that a single missed dose could be mitigated with continuation of the maintenance dosing level whereas for subjects that missed two to six doses, the maintenance dose could be taken twice followed by continuation of daily dosing. For subjects that missed seven doses, a loading dose of 1 mg/kg should be considered followed by continuation of the daily subcutaneous maintenance dose. Conclusions: The clinical trial simulations supported the proposed dosing regimens including dose adjustment with and without missed doses. The local prescribing information should be followed for dosing decision.
[1] Beal SL, Sheiner LB, Boeckmann AJ, Bauer RJ 1989-2011. NONMEM Users Guides. Icon Development Solutions, Ellicott City, Maryland, USA [2] Keizer, R. J., Karlsson, M. O. & Hooker, A. Modeling and Simulation Workbench for NONMEM: Tutorial on Pirana, PsN, and Xpose. CPT Pharmacomet. Syst. Pharmacol. 2 (2013)
Reference: PAGE 33 (2025) Abstr 11502 [www.page-meeting.org/?abstract=11502]
Poster: Drug/Disease Modelling - Other Topics