Assessment of rVWF Pharmacokinetics in Overweight and Obese VWD Patients Using a Population Pharmacokinetic Model

Introduction: Adequate treatment and prevention of bleeding in patients with von Willebrand disease (VWD) requires replacement of von Willebrand factor (VWF) to thresholds often dictated by the clinical situation. While weight (WT)-based dosing is widely used for calculation of clotting factor dose, its validity in all populations has been debated. Further questions have arisen regarding the use of ideal vs actual body weight to calculate dose, likely driven by the large interindividual variability between factor concentrations and associated pharmacokinetics (PK). Understanding the relationship between body size and VWF PK in overweight and obese patients could contribute to optimization of dosing in these populations. Here, we aimed to understand the relationship between body size metrics (body mass index [BMI] and WT) and PK parameters of recombinant VWF (rVWF) in overweight and obese patients, using a population model that described von Willebrand Ristocetin Cofactor (VWF:RCo) PK following rVWF administration.

Methods: A previously developed 2-compartment population PK model, capturing the influence of body size on VWF:RCo PK by WT-based allometric scaling (Bauer A et al, Res Pract Thromb Haemost 2020;4[Suppl 1]), was utilized to investigate rVWF PK in overweight and obese VWD patients. VWF:RCo data used for model development originated from two phase 1 (study 070701 and 071104) and two phase 3 (study 071001 and 071101) VWD trials, for a total of 79 patients with a BMI and WT range of 16.6-47.8 kg/m² and 43.8-144.8 kg, respectively. The model performance in the target subgroup of patients was qualified by means of visual predictive checks (VPC). The relationship between BMI/WT and individual PK parameters was then investigated graphically. Finally, an exposure-matching exercise was performed to assess whether VWF:RCo exposure following rVWF administration in overweight subjects was comparable to that of the reference population (Food and Drug Administration, Guidance for Industry, Exposure-Response Relationships: Study Design, Data Analysis, and Regulatory Applications, 2003); to this end, VWF:RCo area under the curve from time zero to infinity (AUC_inf) for patients of 45-150 kg was simulated following a rVWF dose of 50 IU/kg. The AUC_inf was then compared to the target AUC_inf range, defined as the AUC_inf of a 74-kg patient receiving 50 IU/kg rVWF +/- 20% (Weiner D, Improving the Drug Company Information to Real World Patient Drug Dosing: a Rivaroxaban AF Case Study, American College of Pharmacometrics 2019 Meeting).

Results: The VPC stratified by BMI and WT showed satisfactory predictive performance of the model in overweight patients, with a slight underprediction of peak VWF:RCo activity in patients with BMI >25 kg/m². As expected, graphical analyses indicated an increase in VWF:RCo clearance and volume of distribution with rising BMI and WT, whereas half-life (t_1/2), incremental recovery (IR) and mean residence time appeared to be independent of body size. Trends in PK parameters vs. WT or BMI were similar because of the high correlation between the two body size indices (r=0.94 in the present data set). The simulations of VWF:RCo PK revealed that, for the typical patient, AUC_inf was within the target exposure across the weight range of 45-150 kg. Deviations from the target AUC_inf were observed when inter-individual variability was considered.

Conclusions: The available PK model adequately described VWF:RCo activity in overweight patients with VWD. Overall, t_1/2 and IR remained the same even as BMI or WT increased. Model-simulations indicated that, on average, WT-based dosing in patients of < 150 kg led to an AUCinf of VWF:RCo within the pre-defined target. These results support calculation of rVWF dosing based on actual WT in a patient population similar to the one included in the current analysis (BMI ≤ 47.8 kg/m²). However, PK assessments should remain regularly utilized by physicians to address any potential interindividual variability.