Understanding Loss of Soluble High Molecular Weight Species during Filtration of Low Concentration Therapeutic Monoclonal Antibodies

Sterile filtration is an integral step in the manufacturing process of biological therapeutics. Protein adsorption to the surface of the filter is an unfortunate, common occurrence that can result in manufacturing difficulties, such as filter fouling or product loss. Although many filters have surface modifications to minimize adsorption, under certain conditions binding can still occur. We observed the loss of high molecular weight species (HMWS) during sterile filtration of eight different therapeutic monoclonal antibodies formulated at low protein concentrations across a commonly used hydrophilic polyvinylidene fluoride or polyvinylidene difluoride (PVDF) filter membrane. The protein absorption was specific to HMWS, and each antibody exhibited different degrees of filter adsorption. Debye screening length parameters of the solution (e.g. ionic strength) were adjusted, and influenced the amount of HMWS lost during filtration. Additionally, HMWS of a representative antibody (mAb1) were observed to be more positively charged than other size variants by ion-exchange chromatography. From these results, it is concluded that this HMWS loss is due to electrostatic interactions between HMWS and the filter surface. This adsorption can be reduced by increasing the ionic strength of the buffer matrix, demonstrating the influence of the Debye screening length in the filtration of low concentration proteins.     

Induction and analysis of aggregates in a liquid IgG1-antibody formulation.

The objective of this study was to compare different agitation stress methods (stirring in Reacti Vials versus horizontal shaking) in their effect on protein destabilization, to assess several analytical techniques (light obscuration, turbidimetric and light scattering analysis) for detection of aggregates of various sizes and to evaluate the protecting effect of polysorbate 80 on protein aggregation. A monoclonal IgG1 antibody was used as model protein. Both mechanical stress methods can provoke aggregate formation. The method of stirring induces particles in the range of 10-25 microm comparable to shaking stress. However, stirred samples show a much higher absorbance and reveal a second particle species in DLS analysis, suggesting that stirring stress induces a higher amount of smaller protein aggregates. Addition of polysorbate 80 protects the antibody against aggregation. Only in stirred samples a slight increase in sub-visible particles and turbidity was noted. However, a greater extent of aggregation products was detected by DLS as compared to surfactant-free formulations. Thus, polysorbate 80 appears to stabilise small aggregates and prevents further proceeding of the aggregation process. The induction of aggregates by stirring stress in Reacti Vials analysed by absorbance measurement seems to be a good combination for high-throughput formulation studies.     

Edetate Disodium as a Polysorbate Degradation and Monoclonal Antibody Oxidation Stabilizer

Polysorbates are frequently used in biotherapeutic formulations. Interest in assessing their stability, in particular the impact of their degradation products on the stability of therapeutic proteins, has been steadily growing in the past decade. The work presented summarizes a case study of a monoclonal antibody formulation that demonstrated a simultaneous loss of polysorbate and an increase in methionine oxidation. Spiking studies were conducted to determine both the cause and a potential mitigation for the monoclonal antibody (mAb) oxidation and polysorbate 80 (PS80) loss. The results indicated that a different source material exhibited different rates of mAb oxidation and PS80 loss and that in all evaluated materials, the addition of edetate disodium to the formulation mitigated both observed issues. The mAb was assessed for the presence of lipases and lipoprotein lipase was detected at low levels. It is proposed that edetate disodium was effective in mitigating the mAb oxidation and PS80 loss by chelating calcium in the formulation and therefore decreasing the activity of the lipases.     

Impact of extractables/leachables from filters on stability of protein formulations

Aqueous extractables/leachables from three sterilizing-grade filter membranes [polyvinylidene fluoride (PVDF), polyethersulfone (PES), and mixed cellulose ester (MCE)] were found to significantly reduce the surface tension of aqueous solutions. To evaluate the effect of these extractables/leachables from filter membranes on stability of protein formulations, model IgG2 formulations (with or without added surfactant) were spiked with different levels of filter extractables from stock solutions as a stress study. The stock solutions of extractables were created by processing the filter membranes through autoclaving and soaking steps. The IgG2 formulations were subsequently subject to agitation and temperature stress. Extractables/leachables from the filters were found to have a significant protective (PVDF, PES) and destabilizing (MCE) impact on both visible and subvisible particulates formation under agitation stress for formulations that did not contain any additional surfactant such as polysorbate 80. The impact of filter extractables/leachables on chemical stability of the antibody formulation displayed a more complicated pattern, but was generally destabilizing, causing increases in aggregation, oxidation, and acidic species. In conclusion, extractables/leachables from filter membranes may have impact on protein formulation stability and caution should be exercised during protein filtration, especially when filtering small volumes and in preformulation or high-throughput screening studies.     

Behaviour of polysorbate 20 during dialysis, concentration and filtration using membrane separation techniques

During formulation development of a therapeutic protein, combinations of buffers, pH and excipients need to be tested. As the protein bulk solution used for formulation development usually contains a buffer component at a defined pH and potentially one or more excipients already, this bulk requires to be processed. In case low concentrations of non-ionic surfactants, for example polysorbate 20, are already present in the bulk, the surfactant needs to be removed in lab-scale for further development use. The scope of the work was to study the behaviour of low concentrations of polysorbate 20 during membrane separation processes. The first part focuses on evaluating the behaviour of polysorbate 20 during a dialysis process, whereas the second part analyses concentration changes of polysorbate during a membrane concentration process using a stirred cell. The third part analyses potential membrane absorption of polysorbate at sterilizing-grade filters. In conclusion, it was found that polysorbate could not be significantly reduced during a dialysis process and accumulated during a membrane concentration process in unreproducable manner. During sterile filtration, no significant influence on the concentration of polysorbate was measurable. In any case, it is recommendable to quantify the concentration of polysorbate during critical membrane process steps in pharmaceutical industry.     

Surface Activity of a Monoclonal Antibody

The development of high concentration antibody formulations presents a major challenge for the formulation scientist, as physical characteristics and stability behavior change compared to low concentration protein formulations. The aim of this study was to investigate the potential correlation between surface activity and shaking stress stability of a model antibody-polysorbate 20 formulation. The surface activities of pure antibody and polysorbate 20 were compared, followed by a study on the influence of a model antibody on the apparent critical micelle concentration (CMC) of polysorbate 20 over a protein concentration range from 10 to 150 mg/mL. In a shaking stress experiment, the stability of 10, 75, and 150 mg/mL antibody formulations was investigated containing different concentrations of polysorbate 20, both below and above the CMC. The antibody increased significantly the apparent CMC of antibody-polysorbate 20 mixtures in comparison to the protein-free buffer. However, the concentration of polysorbate required for stabilization of the model antibody in a shaking stress experiment did not show dependence on the CMC. A polysorbate 20 level of 0.005% was found sufficient to stabilize both at low and high antibody concentration against antibody aggregation and precipitation. © 2009 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 98:4525–4533, 2009     

Particles shed from syringe filters and their effects on agitation-induced protein aggregation

We tested the hypothesis that foreign particles shed from filters can accelerate the rate of protein aggregation and particle formation during agitation stress. Various types and brands of syringe filters were tested. Particle counts and size distribution (≥1 μm) in buffer alone or in solutions of keratinocyte growth factor 2 (KGF-2) were determined with a micro-flow imaging. Submicron particle populations were characterized by dynamic light scattering. Loss of soluble protein during filtration or postfiltration incubation was determined by ultraviolet spectroscopy and bicinchoninic acid protein assay. There was a wide range (from essentially none to >100,000/mL) in the counts for at least 1 μm particles shed into buffer or KGF-2 solution from the different syringe filters (with or without borosilicate glass microfibers). Filtration of KGF-2 with units containing glass microfibers above the membrane resulted in 20%-80% loss of protein due to adsorption to filter components. Filtration with systems containing a membrane alone resulted in 0%-20% loss of KGF-2. Effects of 24-h postfiltration incubation were tested on KGF-2 solution filtered with polyether sulfone membrane filters. Loss of soluble protein and formation of particles during agitation were much greater than that in control, unfiltered KGF-2 solutions. Similar acceleration of protein aggregation and particle formation was observed when unfiltered KGF-2 solution was mixed with filtered buffer and agitated. Particle shedding from syringe filters--and the resulting acceleration of protein aggregation during agitation--varied greatly among the different syringe filters and individual units of a given filter type. Our results demonstrate that nanoparticles and microparticles shed from the filters can accelerate protein aggregation and particle formation, especially during agitation.     

Shaken, not stirred: mechanical stress testing of an IgG1 antibody

Protein aggregation is known to occur under different stress conditions and displays a wide variety of morphologies. In this work, the aggregation behavior of a monoclonal antibody (IgG1) was investigated using two different mechanical stress methods namely stirring and shaking at two temperatures, various fill volumes and headspaces and different amounts of polysorbate present in the formulation. The detection of aggregates in terms of size and number was carried out using various analytical techniques including visible particle inspection, turbidity, sub-visible particle analysis, size exclusion chromatography and dynamic light scattering. The data showed that shaking and stirring resulted in different species of aggregates both qualitatively and quantitatively, where stirring was found more stressful than shaking on the IgG1 formulation. Mechanical stress testing performed at 5 and 25 degrees C only showed a difference on samples stressed by shaking and not by stirring. The headspace in the vials had great influence on the stability of the protein formulation when stressed by shaking. The presence of polysorbate had a protective effect on the antibody, however certain polysorbate concentrations even resulted in increased protein aggregation. An array of analytical methods was essential in order to cover the vast aggregate morphologies, which occurred during agitation.     

Modified Bundle of Capillaries Approximation of Sterilizing Filter Membranes and its use for Filter Characterization and Filtration Process Optimization

Sterilizing filtration is a common unit operation for the manufacture of parenteral drug products. However, filter performance can be impacted by properties of both the membrane material and the solution being filtered, requiring extensive multi-factor studies to optimize the filtration process for a given drug product. Here, we report the use of a modified bundle of capillaries approximation to predict filter performance. The model is directly applicable for both Newtonian and non-Newtonian solutions and does not require assumptions of steady state. Using a hydrophilic polyvinylidene difluoride (PVDF) filter as a test case, we show that the model fitting parameters align with expected values and both flux and shear are well predicted. Moreover, two case studies are presented to demonstrate the model's utility for filtration process optimization: 1) protein adsorption of an antibody formulation and 2) filter fouling of a 1% (w/v) carboxymethylcellulose (CMC) solution. In both cases, the model was able to accurately identify optimal filtration parameters to reduce the amount of adsorption or improve the filter capacity, respectively. This methodology can be easily extended to alternate filter types and provides an additional predictive tool to speed process development and minimize trial and error during filtration process design.     

Identification and Quantification of a Problematic Host Cell Protein to Support Therapeutic Protein Development

Monitoring of residual host cell proteins (HCPs) in therapeutic protein is essential to ensure product quality, safety and efficacy. Despite the development of advanced mass spectrometry techniques and optimized workflows, identifying and quantifying all problematic HCPs present at low levels remain challenging. Here, we developed a practical, effective strategy for the identification and quantification of low abundance HCPs, which facilitates the improvement of downstream purification process to eliminate potentially problematic HCPs. A case study of using this strategy to investigate a problematic HCP is presented. Initially, a commonly used native digestion approach coupled with UPLC-MS/MS was applied for HCP profiling, wherein several lipases and proteases were identified in a monoclonal antibody named mAb1 in early stages of purification process development. A highly active lipase, liver carboxylesterase (CES), was found to be responsible for polysorbate 80 degradation. To facilitate process improvement, after the identification of CES, we developed a highly sensitive LC-MS/MS-MRM assay with a lower limit of quantification of 0.05 ppm for routine monitoring of the CES in mAb1 produced through the different processes. This workflow was applied in low-level lipase identification and absolute quantification, which facilitated the investigation of polysorbate degradation and downstream purification improvement to further remove the problematic HCP. The current MRM method increased the sensitivity of HCP quantification by over 10-fold that in previously published studies, thus meeting the needs for quantification of problematic HCPs at sub-ppm to ppb levels during drug development. This workflow could be readily adapted to the detection and quantification of other problematic HCPs present at extremely low levels in therapeutic protein drug candidates.     

Applications of Statistical Regression and Modeling in Fill–Finish Process Development of Structurally Related Proteins

In order to increase efficiency and reduce cost, many biotechnology and pharmaceutical companies utilize platform approaches for discovery and development of structurally related therapeutic proteins. In the case of the monoclonal antibody modality, retention and reuse of prior development knowledge is especially useful to gauge risks, improve speed and reduce cost for developing similar molecules in the future. In this paper, we present two applications of statistical regression and modeling to help decision making during antibody drug product fill-finish process development. The applications are for estimating viscosity and filter capacity (Vmax) values. Experiments were performed to obtain relevant data sets of viscosity, protein concentration, density, and Vmax values for various candidate antibodies. Then, statistical models were developed and optimized to estimate viscosity and filtration Vmax values for new antibodies. Viscosity of protein formulations is an important physical property that impacts almost all manufacturing operations, as well as delivery or administration of drug products. Vmax is a critical parameter for filter size selection in manufacturing processes. Development and optimization of both models followed similar steps: identifying multicollinearity and interactions, removing unnecessary explanatory variables, performing appropriate data transformation, and evaluating different model options. We obtained 95% prediction limits for the mean and individual values from the models, and further verified the models by comparing predicted values with additional experimental data. These applications of statistical tools enabled leveraging prior knowledge for process development of new molecules belonging to the same class of structurally related proteins. Although the two specific models presented here may not be directly applicable for all proteins, the approach and methodology presented can be broadly useful for structurally related protein products during their development.     

The Role of Polysorbate 80 and HP��CD at the Air-Water Interface of IgG Solutions

Purpose

To test the hypothesis of surface displacement as the underlying mechanism for IgG stabilization by polysorbates and HP??CD against surface-induced aggregation.

Methods

Adsorption/desorption-kinetics of IgG-polysorbate 80/-HP??CD were monitored. Maximum bubble pressure method was used for processes within seconds from surface formation. Profile analysis tensiometry was applied over long periods and to assess surface rheologic properties. Additionally, the kinetics of adsorption, desorption and surface displacement was followed by a double-capillary setup of the profile analysis tensiometer, allowing drop bulk exchange.

Results

Weak surface activity for HP??CD vs. much higher surface activity for polysorbate 80 was shown. Protein-displacement when exceeding a polysorbate 80 concentration close to the CMC and a lack of protein displacement for HP??CD was observed. The drop bulk exchange experiments show IgG displacement by polysorbate 80 independent of the adsorption order. In contrast, HP??CD coexists with IgG at the air-water interface when the surface layer is built from a mixed IgG-HP??CD-solution. Incorporation of HP??CD in a preformed IgG-surface-layer does not occur.

Conclusions

The results confirm surface displacement as the stabilization mechanism of polysorbate 80, but refute the frequently held opinion, that HP??CD stabilizes proteins against aggregation at the air-water interface in a manner comparable to non-ionic surfactants.     

The behavior of different carrageenans in pelletization by extrusion/spheronization

kappa-Carrageenan is known as a novel pelletization aid in manufacturing of pellets by extrusion/spheronization. The implentation of kappa-carrageenan as a pelletization aid can overcome several disadvantages of commonly used microcrystalline cellulose (MCC) such as lacking disintegration and adsorption of several actives. The aim of this study was to compare different types of carrageenans from different suppliers: one iota-, five kappa-, and one lambda-carrageenan. The kappa-carrageenans had the best pelletization behavior. Four of the five tested kappa-carrageenans resulted in pellets with acceptable shapes, sizes, and size distributions using a high drug load of 80% hydrochlorothiazide. These pellets have similar properties over a wide range of water contents ranging from 90 to 105%. The filler, dicalcium phosphate, affected the pelletization process and the pellet properties of all investigated kappa-carrageenans.     

Polysorbate 20 prevents the precipitation of a monoclonal antibody during shear

Monoclonal antibodies (MAbs) are widely used as therapeutic proteins and they are frequently exposed to a high degree of stress during manufacturing or delivery. MAbs shear thin upon increasing shear rates. After undergoing multiple shear cycles, with a cone-and-plate rheometer, the solution viscosity of high concentration antibodies increases due to the formation of insoluble aggregates. These shear-induced insoluble aggregates do not form when polysorbate 20 is present in solution. We hypothesize that monoclonal antibodies form a thin protein layer at the air-water interface. MAbs at the interface expose their hydrophobic core to air leading to unfolding, multiple non-specific intermolecular interactions and, upon continuous high shear, precipitation. Surface tension analysis confirms that monoclonal antibodies are surface active and that polysorbate 20 can prevent their interaction with the air-water interface. In addition, we complement these findings with a viscometer that measures bulk viscosity without the influence of an air-liquid interfacial viscosity and find that the bulk viscosity increases slightly when Mab solutions contained polysorbate 20. These methods of analysis could be used when designing manufacturing systems in which a protein solution is subject to shear forces.     

Stabilization of a human recombinant factor VIII by poloxamer 188 in relation to polysorbate 80

Detection of enhanced surface tension depression by surfactant in the presence of protein was recently suggested as a basis for determining whether protein stabilization by that surfactant is owing to surfactant forming a steric barrier at interfaces or surfactant association with the protein. In particular, protein interaction with surfactant aggregates may lead to an increased concentration of monomers thus enhancing surfactant adsorption, or to formation of surfactant–protein complexes having little or no effect on adsorption. We compared the initial rates of surface tension depression by poloxamer 188 and polysorbate 80 (PS 80) in the presence and absence of a human recombinant factor VIII (rFVIII). Indirect evidence had suggested poloxamer 188 enters into stable associations with rFVIII in solution but does not form a steric barrier at the interface, while PS 80 behaves in contrary fashion. In this study, we show the presence of rFVIII caused an increase in the rate (reduction in the activation energy) of PS 80 adsorption, while no such change was recorded in the case of poloxamer 188. Thus, we provide substantiation for detection of protein-mediated acceleration of surfactant adsorption as a means to compare different surfactants in relation to their favored mechanism for protein stabilization.     

The Behavior of Different Carrageenans in Pelletization by Extrusion/Spheronization

κ-Carrageenan is known as a novel pelletization aid in manufacturing of pellets by extrusion/spheronization. The implentation of κ-carrageenan as a pelletization aid can overcome several disadvantages of commonly used microcrystalline cellulose (MCC) such as lacking disintegration and adsorption of several actives. The aim of this study was to compare different types of carrageenans from different suppliers: one ι-, five κ-, and one λ-carrageenan. The κ-carrageenans had the best pelletization behavior. Four of the five tested κ-carrageenans resulted in pellets with acceptable shapes, sizes, and size distributions using a high drug load of 80% hydrochlorothiazide. These pellets have similar properties over a wide range of water contents ranging from 90 to 105%. The filler, dicalcium phosphate, affected the pelletization process and the pellet properties of all investigated κ-carrageenans.     

The Degradation of Polysorbates 20 and 80 and its Potential Impact on the Stability of Biotherapeutics

Purpose

To study the potential impact of the degradation of Polysorbates (PS) 20 and 80 on the stability of therapeutic proteins in parenteral formulations.

Method

First, degradation products of PS20 and 80 were identified. Subsequently, the effect of degraded polysorbate on physical characteristics and long-term stability of protein formulations was assessed. Further, the impact of polysorbate degradation on protein stability was evaluated via shaking stress studies on formulations spiked with artificially degraded polysorbate or degradants like fatty acids. Additionally, aged formulations with reduced polysorbate content were shaken.

Results

The degradation of polysorbate leads to a buildup of various molecules, some of which are poorly soluble, including fatty acids and polyoxyethylene (POE) esters of fatty acids. Spiking studies showed that the insoluble degradants could potentially impact protein stability and that the presence of sufficient intact polysorbate was crucial to prevent this. End-of-shelf-life shaking of protein formulations showed that the stability of various monoclonal antibodies was, however, not affected.

Conclusions

Although some degradants can potentially influence the stability of the protein (as discerned from spiking studies), degradation of polysorbates did not impact the stability of the different proteins tested in pharmaceutically relevant temperature and storage conditions.     

A rapid filtration method for receptor binding: characterization with mu and delta opiate receptors

A commercially available (Skatron) cell harvester was adapted for use in mu (3H-naloxone-labeled) and delta (3H-DADLE-labeled) opiate receptor assays and compared with a widely used conventional manifold for a number of binding characteristics. Whatman GF/B glass fiber filters and the less expensive filters available with the harvester were also compared and produced similar binding characteristics on the harvester and manifold if the harvester filters were used double-ply, and if the rinse time was less than 12.5 sec. Longer rinse times produced lower binding with 2-ply Skatron filters. Kd values, Hill coefficients, and Scatchard plot regression coefficients were very similar for the two filtration devices and filter types. A significantly reduced maximum number of sites (Bmax) was observed after filtration on the harvester, reflecting the smaller filter surface area relative to that of the manifold. The filter surface area on the harvester, nevertheless, is considerably larger than that of other manifolds with microplate spacing. This provides the advantages of rapid filtration with less restriction on tissue concentrations. Specific binding was linear with protein concentration up to at least 800 micrograms protein, which is well within the range of most neurotransmitter and peptide receptor binding studies. At about 1 mg protein the rinse buffer flow was slower due to the high tissue concentration. Although the results of filtration with the harvester and the conventional manifold were similar, the time requirements differed considerably. With the harvester, one experimenter could conduct the filtration process 2-3 times faster than 2 experimenters using the manifold.(ABSTRACT TRUNCATED AT 250 WORDS)     

Polysorbates 20 and 80 used in the formulation of protein biotherapeutics: structure and degradation pathways

Polysorbates 20 and 80 (Tween 20 and Tween 80) are used in the formulation of biotherapeutic products for both preventing surface adsorption and as stabilizers against protein aggregation. The polysorbates are amphipathic, nonionic surfactants composed of fatty acid esters of polyoxyethylene sorbitan being polyoxyethylene sorbitan monolaurate for polysorbate 20 and polyoxyethylene sorbitan monooleate for polysorbate 80. The polysorbates used in the formulation of biopharmaceuticals are mixtures of different fatty acid esters with the monolaurate fraction of polysorbate 20 making up only 40-60% of the mixture and the monooleate fraction of polysorbate 80 making up >58% of the mixture. The polysorbates undergo autooxidation, cleavage at the ethylene oxide subunits and hydrolysis of the fatty acid ester bond. Autooxidation results in hydroperoxide formation, side-chain cleavage and eventually formation of short chain acids such as formic acid all of which could influence the stability of a biopharmaceutical product. Oxidation of the fatty acid moiety while well described in the literature has not been specifically investigated for polysorbate. This review focuses on the chemical structure of the polysorbates, factors influencing micelle formation and factors and excipients influencing stability and degradation of the polyoxyethylene and fatty acid ester linkages.     

Understanding Protein-Interface Interactions of a Fusion Protein at Silicone Oil-Water Interface Probed by Sum Frequency Generation Vibrational Spectroscopy

Protein adsorbed at the silicone oil-water interface can undergo a conformational change that has the potential to induce protein aggregation on storage. Characterization of the protein structures at interface is therefore critical for understanding the protein-interface interactions. In this article, we have applied sum frequency generation (SFG) spectroscopy for studying the secondary structures of a fusion protein at interface and the surfactant effect on protein adsorption to silicone oil-water interface. SFG and chiral SFG spectra from adsorbed protein in the amide I region were analyzed. The presence of beta-sheet vibrational band at 1635 cm^?1 implies the protein secondary structure was likely perturbed when protein adsorbed at silicone oil interface. The time-dependent SFG study showed a significant reduction in the SFG signal of preadsorbed protein when polysorbate 20 was introduced, suggesting surfactant has stronger interaction with the interface leading to desorption of protein from the interface. In the preadsorbed surfactant and a mixture of protein/polysorbate 20, SFG analysis confirmed that surfactant can dramatically prevent the protein adsorption to silicone oil surface. This study has demonstrated the potential of SFG for providing the detailed molecular level understanding of protein conformation at interface and assessing the influence of surfactant on protein adsorption behavior.