Improving the pharmaceutical properties of Factor VIII with protein modification.

摘要

Factor VIII (FVIII) is a large glycoprotein of ∼280 kDa that is dysfunctional or deficient in Hemophilia A (HA), a bleeding disorder. Patients with HA suffer from an inability to effectively form blood clots following vascular injury. The overall goal of this thesis was to explore how different modification strategies might ultimately impact the pharmaceutical properties of FVIII.;In Chapter 1, we characterized the allometric scaling of FVIII with a literature meta-analysis of over 100 PK profiles across five species. The Wajima method was successfully used to relate FVIII concentration-time profiles between species. We developed a method of 'informed scaling' to predict PK parameters for next generation products in a way that incorporates known information for the unmodified reference product and demonstrated the applicability of the method using a case study with an Fc-fusion FVIII.;In Chapter 2, a PK study was conducted in the dose range of 20-200 IU/kg comparing free FVIII and FVIII-PI. Nonlinearity was observed and characterized by a two compartment model with saturable Michaelis-Menten (MM) elimination. Association with PI was identified to reduce affinity for the clearance mechanism (km), distributional clearance (CLd), and peripheral volume (V2). These changes are consistent with PI mediated shielding from the low density lipoprotein receptor-related protein 1 (LRP1) clearance receptor and reduced uptake of high molecular weight FVIII species. Interpretation of changes in PK for modified FVIII products may be complicated by pronounced nonlinearity at higher doses and preclinical product testing in HA mice is recommended in the clinically relevant 25-50 IU/kg dose range.;In Chapter 3, PK studies were conducted to determine whether the previously observed reduction of clearance from associating full length FVIII with PI particles was extendable to wild-type (WT) B-domain deleted (BDD)-FVIII. WT-PI exhibited an increase in half-life compared to the free WT protein similar to what was observed with full length FVIII, indicating specific interactions with the B-domain did not play a prominent role in the benefits of association with PI.;In chapter 4 we investigated whether the CAT could be used as a PD endpoint in mice, primates, and dogs. Substantial optimization was performed for the assay in mouse plasma, but the assay was unable to provide acceptable reproducibility. The CAT proved equally ineffective in primates, but represents a viable strategy in dogs where it was used to compare thrombin generation in plasma collected from a HA dog receiving free FVIII and FVIII-PI. Despite prolonging the half-life of FVIII in mice, association with PI did not appear to impact thrombin generation in this limited canine sample.;In Chapter 5 we examined the structural changes resulting from mutations in D519V/E665V with multiple biophysical techniques and investigated whether in vitro stability improvements would translate to an in vivo animal model. The mutations did not alter the native secondary or tertiary structure of the WT protein, but the transition temperature for D519V/E665V was increased in thermal unfolding studies indicating improved stability. PK studies were conducted in HA mice to evaluate the impact of this enhanced stability on plasma survival of the protein.;N-linked glycans are the predominant oligosaccharides found on FVIII. 21 of 25 consensus N-glycosylation sites have been identified to display active glycosylation, mostly concentrated in the heavily glycosylated B-domain. In Chapter 6 we investigated how enzymatic digestion of these N-linked sugars with a series of enzymes would ultimately impact the stability, activity, intermolecular interactions, and immunogenicity of FVIII. (Abstract shortened by UMI.).