Manufacturing process used to produce long-acting recombinant factor VIII Fc fusion protein.

Recombinant factor VIII Fc fusion protein (rFVIIIFc) is a long-acting coagulation factor approved for the treatment of hemophilia A. Here, the rFVIIIFc manufacturing process and results of studies evaluating product quality and the capacity of the process to remove potential impurities and viruses are described. This manufacturing process utilized readily transferable and scalable unit operations and employed multi-step purification and viral clearance processing, including a novel affinity chromatography adsorbent and a 15 nm pore size virus removal nanofilter. A cell line derived from human embryonic kidney (HEK) 293H cells was used to produce rFVIIIFc. Validation studies evaluated identity, purity, activity, and safety. Process-related impurity clearance and viral clearance spiking studies demonstrate robust and reproducible removal of impurities and viruses, with total viral clearance >8-15 log10 for four model viruses (xenotropic murine leukemia virus, mice minute virus, reovirus type 3, and suid herpes virus 1). Terminal galactose-α-1,3-galactose and N-glycolylneuraminic acid, two non-human glycans, were undetectable in rFVIIIFc. Biochemical and in vitro biological analyses confirmed the purity, activity, and consistency of rFVIIIFc. In conclusion, this manufacturing process produces a highly pure product free of viruses, impurities, and non-human glycan structures, with scale capabilities to ensure a consistent and adequate supply of rFVIIIFc.

Use and application of hydrophobic interaction chromatography for protein purification.

The objective of this section is to provide the reader with guidelines and background on the use and experimental application of Hydrophobic Interaction chromatography (HIC) for the purification of proteins. The section will give step by step instructions on how to use HIC in the laboratory to purify proteins. General guidelines and relevant background information is also provided.

Enveloped virus inactivation using neutral arginine solutions and applications in therapeutic protein purification processes.

For the manufacturing of recombinant protein therapeutics produced from mammalian cell culture, demonstrating the capacity of the purification process to effectively clear infectious viruses is a regulatory requirement. At least two process steps, using different mechanisms of virus removal and/or inactivation, should be validated in support of the regulatory approval process. For example, exposure of the product stream to low pH, detergents or solvent/detergent combinations is commonly incorporated in protein purification processes for the inactivation of lipid-enveloped viruses. However, some proteins have limited stability at low pH or in the presence of the detergents, and alternative techniques for achieving the inactivation of enveloped viruses would be beneficial. We present here an alternative and novel approach for the rapid inactivation of enveloped viruses using pH-neutral buffer solutions containing arginine. The implementation of this approach in a monoclonal antibody or Fc-fusion protein purification process is described and illustrated with several different therapeutic proteins. The use of the neutral pH arginine solution was able to effectively inactivate two enveloped model viruses, with no measurable effect on the product quality of the investigated proteins. Thus, the use of pH-neutral arginine containing buffer solutions provides an alternative means of virus inactivation where other forms of virus inactivation, such as low pH and/or solvent/detergent treatments are not possible or undesirable due to protein stability limitations.

Mitigation of chromatography adsorbent lot performance variability through control of buffer solution design space.

The separation of undesired product-related impurities often poses a challenge in the purification of protein therapeutic species. Product-related impurity species, which may consist of undesirable isoforms, aggregated, or misfolded variants of the desired monomeric form of the product, can be challenging to remove using preparatory scale chromatographic techniques. When using anion exchange chromatography to remove undesirable product-related impurities, the separation can be highly sensitive to relatively small changes in the chromatography operating conditions, including changes to buffer solution pH, buffer solution conductivity protein loading, and operating temperature. When performing difficult separations, slight changes to the chemical and physical properties of the anion exchange adsorbent lot may also impact the separation profile. Such lot-to-lot variability may not be readily measurable by the adsorbent manufacturer, since variability can be highly dependent on a specific protein separation. Consequently, manufacturers of chromatographic adsorbents may not be able to control adsorbent lot to lot variability tightly enough to prevent differences from occurring when performing difficult product-related separations at the preparatory scale. In such cases, it is desirable to design a chromatography step with a control strategy which accounts for adsorbent lot to lot variability in the separation performance. In order to avoid the undesired changes to process consistency and product quality, a proper adjustment of the column operating conditions can be implemented, based on the performance of each adsorbent lot or lot mixture. In this work, we describe how the adjustment of the column buffer solution composition can be used as a design space based-control strategy used to ensure consistent process performance and product quality are achieved for an anion exchange chromatography step susceptible to adsorbent lot to lot performance variability. In addition, a "use test" is described that can be employed to determine the optimal buffer solution compositions for different anion exchange adsorbent lots based on the retention volume of the therapeutic protein during a gradient elution.

Safety and prolonged activity of recombinant factor VIII Fc fusion protein in hemophilia A patients.

Current factor VIII (FVIII) products display a half-life (t(1/2)) of ∼ 8-12 hours, requiring frequent intravenous injections for prophylaxis and treatment of patients with hemophilia A. rFVIIIFc is a recombinant fusion protein composed of a single molecule of FVIII covalently linked to the Fc domain of human IgG(1) to extend circulating rFVIII t(1/2). This first-in-human study in previously treated subjects with severe hemophilia A investigated safety and pharmacokinetics of rFVIIIFc. Sixteen subjects received a single dose of rFVIII at 25 or 65 IU/kg followed by an equal dose of rFVIIIFc. Most adverse events were unrelated to study drug. None of the study subjects developed anti-rFVIIIFc antibodies or inhibitors. Across dose levels, compared with rFVIII, rFVIIIFc showed 1.54- to 1.70-fold longer elimination t(1/2), 1.49- to 1.56-fold lower clearance, and 1.48- to 1.56-fold higher total systemic exposure. rFVIII and rFVIIIFc had comparable dose-dependent peak plasma concentrations and recoveries. Time to 1% FVIII activity above baseline was ∼ 1.53- to 1.68-fold longer than rFVIII across dose levels. Each subject showed prolonged exposure to rFVIIIFc relative to rFVIII. Thus, rFVIIIFc may offer a viable therapeutic approach to achieve prolonged hemostatic protection and less frequent dosing in patients with hemophilia A. This trial was registered at www.clinicaltrials.gov as NCT01027377.

Prolonged activity of a recombinant factor VIII-Fc fusion protein in hemophilia A mice and dogs.

Despite proven benefits, prophylactic treatment for hemophilia A is hampered by the short half-life of factor VIII. A recombinant factor VIII-Fc fusion protein (rFVIIIFc) was constructed to determine the potential for reduced frequency of dosing. rFVIIIFc has an ∼ 2-fold longer half-life than rFVIII in hemophilia A (HemA) mice and dogs. The extension of rFVIIIFc half-life requires interaction of Fc with the neonatal Fc receptor (FcRn). In FcRn knockout mice, the extension of rFVIIIFc half-life is abrogated, and is restored in human FcRn transgenic mice. The Fc fusion has no impact on FVIII-specific activity. rFVIIIFc has comparable acute efficacy as rFVIII in treating tail clip injury in HemA mice, and fully corrects whole blood clotting time (WBCT) in HemA dogs immediately after dosing. Furthermore, consistent with prolonged half-life, rFVIIIFc shows 2-fold longer prophylactic efficacy in protecting HemA mice from tail vein transection bleeding induced 24-48 hours after dosing. In HemA dogs, rFVIIIFc also sustains partial correction of WBCT 1.5- to 2-fold longer than rFVIII. rFVIIIFc was well tolerated in both species. Thus, the rescue of FVIII by Fc fusion to provide prolonged protection presents a novel pathway for FVIII catabolism, and warrants further investigation.

Theory and use of hydrophobic interaction chromatography in protein purification applications.

Hydrophobic interaction chromatography (HIC) is a valuable tool used in protein purification applications. HIC is used in the purification of proteins over a broad range of scales-in both analytical and preparatory scale applications. HIC is used to remove various impurities that may be present in the solution, including undesirable product-related impurities. In particular, HIC is often employed to remove product aggregate species, which possess different hydrophobic properties than the target monomer species and can often be effectively removed using HIC. In this chapter, we provide a description of the basic theory of HIC and how it is used to purify proteins in aqueous-based solutions. Following the theoretical background, the latest in HIC adsorbent technology is described, including a list of commonly used and commercially available adsorbents. The basic procedures for using HIC adsorbents are described next, in order to provide the reader with useful starting points to apply HIC in protein purification applications.

Application of a novel affinity adsorbent for the capture and purification of recombinant factor VIII compounds.

Recombinant Factor VIII (FVIII) therapies have been created to provide treatment for Hemophilia A, an inherited bleeding disorder caused by mutation in the FVIII gene. A major challenge in the purification of recombinant FVIII molecules is the development of an affinity chromatography step. Such a step must be highly specific and selective for the FVIII molecule, but also must be designed appropriately to ensure the FVIII molecule can be effectively recovered without resorting to harsh elution conditions which may be harmful to the product. Additionally, it is desirable to have affinity adsorbents designed to be reusable over a large number of column cycles while maintaining consistent purification performance. In this work, we describe the use of VIIISelect, a commercially available affinity adsorbent designed specifically for the purification of FVIII compounds. The VIIISelect adsorbent consists of a 13kDa recombinant protein ligand attached to a cross-linked agarose base matrix. The structure of the recombinant ligand is based upon Camelid-derived single domain antibody fragments. The VIIISelect adsorbent is produced using a process free of animal-derived raw materials, which is a highly desirable attribute for adsorbents used in the purification processes of recombinant protein therapeutics. The VIIISelect adsorbent was used as the initial capture column to purify a FVIII compound directly from clarified cell culture fluid prior to further downstream purification. The purification performance of the VIIISelect was evaluated, which included measurement of the static binding capacity, dynamic binding capacity, product recovery, impurity clearance, and adsorbent reuse. Following laboratory-scale process development, the VIIISelect adsorbent was scaled up and used in the large scale manufacturing of a FVIII compound.

Modeling the effects of column packing quality and residence time changes on protein monomer/aggregate separation.

The packing quality of chromatography columns used for the purification of protein therapeutics is routinely monitored to ensure consistent and reproducible performance. In this work, we used established chromatography models to determine the effect of column packing quality and fluid residence time on the separation of protein therapeutic monomer and aggregate species using a hydrophobic interaction chromatography adsorbent (Phenyl Sepharose Fast Flow). The relationship between the number of theoretical plates, fluid residence time, and column separation performance was quantified using modeling simulations. The simulations showed the separation depended on both the fluid residence time and the number of theoretical plates. However, when the number of theoretical plates was increased to >/=150, the simulations predicted that the separation performance of the column was not significantly improved. The approach described here could be used as a method to quantify acceptable height equivalent of a theoretical plate values for columns, and serve as a tool to understand how column packing quality impacts a given chromatographic separation prior to column scale-up, as well as during the monitoring of column lifetime in the manufacturing of large scale protein therapeutics.

Effect of phenyl sepharose ligand density on protein monomer/aggregate purification and separation using hydrophobic interaction chromatography.

In the large-scale manufacturing and purification of protein therapeutics, multiple chromatography adsorbent lots are often required due to limited absorbent batch sizes or during replacement at the end of the useful column lifetime. Variability in the adsorbent performance from lot to lot should be minimal in order to ensure that consistent product purity and product quality attributes are achieved when a different lot or lot mixture is implemented in the process. Vendors of chromatographic adsorbents will often provide release specifications, which may possess a narrow range of acceptable values. Despite relatively narrow release specifications, the performance of the adsorbent in a given purification process could still vary from lot to lot. In this case, an alternative use test (one that properly captures the lot to lot variability) may be required to determine an acceptable range of variability for a specific process. In this work, we describe the separation of therapeutic protein monomer and aggregate species using hydrophobic interaction chromatography, which is potentially sensitive to adsorbent lot variability. An alternative use test is formulated, which can be used to rapidly screen different adsorbent lots prior to implementation in a large-scale manufacturing process. In addition, the underlying mechanism responsible for the adsorbent lot variability, which was based upon differences in protein adsorption characteristics, was also investigated using both experimental and modeling approaches.

Modeling of protein monomer/aggregate purification and separation using hydrophobic interaction chromatography.

Hydrophobic interaction chromatography (HIC) is commonly used to separate protein monomer and aggregate species in the purification of protein therapeutics. Despite being used frequently, the HIC separation mechanism is quite complex and not well understood. In this paper, we examined the separation of a monomer and aggregate protein mixture using Phenyl Sepharose FF. The mechanisms of protein adsorption, desorption, and diffusion of the two species were evaluated using several experimental approaches to determine which processes controlled the separation. A chromatography model, which used homogeneous diffusion (to describe mass transfer) and a competitive Langmuir binary isotherm (to describe protein adsorption and desorption), was formulated and used to predict the separation of the monomer and aggregate species. The experimental studies showed a fraction of the aggregate species bound irreversibly to the adsorbent, which was a major factor governing the separation of the species. The model predictions showed inclusion of irreversible binding in the adsorption mechanism greatly improved the model predictions over a range of operating conditions. The model successfully predicted the separation performance of the adsorbent with the examined feed.

Use of an alternative scale-down approach to predict and extend hydroxyapatite column lifetimes.

Ceramic hydroxyapatite (CHT) chromatography offers unique selectivity for protein purification. However, columns composed of CHT, a crystalline form of calcium phosphate, often suffer from short column lifetimes, particularly under acidic operating conditions. In this paper, CHT was used under slightly acidic conditions (pH 6) for the production scale purification of a recombinant protein. Under these conditions, the packing quality of production scale CHT columns (45 cm diameter) degraded after 5-10 cycles of operation. This was not reproduced using a conventional scale-down chromatography model, in which a constant column bed height was maintained across scales. Thus, an alternative scale-down model was developed to better predict the lifetime of large scale CHT columns. The alternative approach, which utilized a constant column diameter-to-height aspect ratio, was able to predict column failure that approximated that of the manufacturing scale column. The alternative scale-down approach was then used to test alternate buffer formulations that significantly improved the CHT column lifetime. Screening studies, which assessed the effects of mobile phase pH and composition on the dissolution (weight loss) of CHT, were used to identify the alternative mobile phase formulations. Results from the study showed that slight changes to the existing mobile phase compositions significantly increased the column lifetime, from approximately 10 cycles to approximately 65 cycles of use, without altering the purification of the recombinant protein. The alternative scale-down model, together with relatively rapid mobile phase screening studies, provides a practical approach for predicting and optimizing the useful lifetime of CHT columns for large scale applications.

Application of a two-dimensional model for predicting the pressure-flow and compression properties during column packing scale-up.

A two-dimensional model was formulated to describe the pressure-flow behavior of compressible stationary phases for protein chromatography at different temperatures and column scales. The model was based on the assumption of elastic deformation of the solid phase and steady-state Darcy flow. Using a single fitted value for the empirical modulus parameters, the model was applied to describe the pressure-flow behavior of several adsorbents packed using both fluid flow and mechanical compression. Simulations were in agreement with experimental data and accurately predicted the pressure-flow and compression behavior of three adsorbents over a range of column scales and operating temperatures. Use of the described theoretical model potentially improves the accuracy of the column scale-up process, allowing the use of limited laboratory scale data to predict column performance in large scale applications.

Evaluation of protein-A chromatography media.

In process-scale antibody purification, protein-A affinity chromatography is commonly used as the initial purification step. In this paper, two different protein-A media were evaluated. These adsorbents have a porous glass backbone with different pore sizes: 700 A and 1000 A. Adsorption equilibrium data of human immunoglobulins on these media were measured via a batch technique and correlated using the Langmuir isotherm model. A larger static capacity was found for the smaller pore size material, which is probably a result of the larger specific surface area and associated higher ligand concentration. The protein uptake kinetics were also obtained via a stirred tank experiment using different initial protein concentrations. A surface layer model was used to represent the protein uptake by the media and to estimate values of a concentration-independent effective diffusivity within the particle. Experimental breakthrough curves were also obtained from packed beds operated under different conditions. Calculated breakthrough profiles were found to be in good agreement with the experimental results. Experimental breakthrough data were used to determine the dependence of the dynamic capacity of the media as a function of the fluid residence time. A larger dynamic capacity was also found for the smaller pore size media. The permeability of large scale packed beds was also reported and used in conjunction with the dynamic capacity to calculate the process production rate.