Emilie Hénin (1), Chafke Belmokhtar (2), Damien Barrois (2), Christian Laveille (1), Juliette Hassenboehler (3), Claude Négrier (3)
1) Calvagone, Liergues, France 2) Octapharma, Boulogne, France 3) Haematology division, Louis Pradel University Hospital, Bron, France
Objectives:
Prophylaxis for haemophilia B is effective but restrictive and expensive. Knowledge of a patient’s pharmacokinetic (PK) profile after infusion of factor IX is likely to be useful in clinical management. Management of haemophilia B has mostly been empirical so far to determine dosage, with a combination of outcome measures such as clinical bleeding phenotype and determination of trough level of factor IX (FIX) activity. However, inter-individual pharmacokinetic variability of FIX is large and PK-tailored personalized prophylaxis may help to adjust FIX activity at the correct level.
A prospective, open-label, multi-centre phase 2 study, PeKaFIX, has been conducted between July 2016 and August 2017, to evaluate, by rich sampling, the pharmacokinetics of a plasma-derived factor IX concentrate (OCTAFIX®, as a single IV injection at a dose of 75 IU/kg, after a wash-out period of at least 7 days) in patients with haemophilia B. Data from 3 other studies (YNE-201, YNE-202, YNE-203), performed between 1996 and 2005, in adult and paediatric patients with haemophilia B receiving OCTAFIX, were also available for model development.
The aim of this work was to analyse, using population modelling approaches, the factor IX activity profile after OCTAFIX administration, and to externally evaluate, on PeKaFIX data, the predictability of individual parameters and comparison criteria under several sparse sampling designs, in order to propose individual therapeutic monitoring strategy.
Patients & Methods:
Factor IX time-courses, measured in 45 adult and paediatric patients with moderate or severe haemophilia B, receiving single or multiple prophylactic OCTAFIX® administrations at doses ranging from 36 to 81 IU/kg, were issued from the previously conducted YNE-201, YNE-202 and YNE-203 studies, and were considered for model development.
Several scenarios, i.e. rich sampling (13 observations per individual, considered as the reference) and sparse sampling (3-4 observations at various time points), were considered to predict individual parameters in the 14 patients issued from PeKaFIX study, and compared in terms of Time-to-reach FIX xx% (5, 4, and 3%) and predicted FIX levels at time T (48, 72, 96 and 120 hours after administration).
Results & discussion:
The data from 45 patients, previously considered in YNE-201, YNE-202 and YNE-203 studies, were used to successfully build a population pharmacokinetic model, giving a good description of factor IX kinetics after OCTAFIX® single and multiple administrations, in adults and children with severe or moderate haemophilia B. The proposed model is a two-compartment model, with allometric scaling on parameters, and a baseline FIX level presenting endogenous production of factor IX. Inter-individual variability was found on clearance, central volume and basal FIX level.
Compared to the reference scenario (using all 13 samples per individual), the better sparse scenario accounted for 3 samples at 0, 0.5 and 48 hours after administration: absolute difference in FIX predicted levels was in median within the population 0.7% at 48 hours, 0.5% at 72 hours; 0.3% at 96 hours and 0.2% at 120 hours. Reference time-to-reach 5%, 4% and 3% were predicted in median at 91, 103 and 119 hours respectively; the above-mentioned scenario differed from 4 to 5 hours to reference values, allowing an adequate estimation of the dosing interval required to maintain FIX level above a given threshold.
All evaluated sampling schemes based on 3 or 4 observations gave acceptable individual predictions in terms of FIX profile, predicted FIX level at a given time and in terms of time-to-reach a given FIX level. However, better predictive performance was obtained when sampling schemes included samples up to 48 hours over schemes up to 24 hours after administration.
Using the proposed model, individual parameters for FIX kinetics can be predicted based on only few blood samples e.g. drawn at 0, 0.5 and 48 hours after OCTAFIX administration, while flexibility is still allowed to the physician in choosing sampling intervals adapted to routine constraints. The prophylactic dosing regimen to be administered can then be calculated (assuming dose-linearity) and adapted to each patient, depending on the threshold above which FIX level should be maintained (as a function of patient’s lifestyle), and its individual parameters. This approach will certainly help the management of an individualized prophylactic strategy in a routine manner.
Reference: PAGE 27 (2018) Abstr 8478 [www.page-meeting.org/?abstract=8478]
Poster: Drug/Disease Modelling - Other Topics