High Performance Size Exclusion Chromatography and High-Throughput Dynamic Light Scattering as Orthogonal Methods to Screen for Aggregation and Stability of Monoclonal Antibody Drug Products.

The presence of aggregates in monoclonal antibody (mAb) drug product (DP) formulations can present product quality challenges. Here we show that use of High Performance Size Exclusion Chromatography (HP-SEC), in conjunction with high-throughput dynamic light scattering (HT-DLS) analyses of mAb DPs can be a useful strategy to determine monomer content and the presence of aggregates under simulated stress conditions. This analytical approach was used to evaluate four commercially available mAb DPs under different conditions i.e.; original formulations, diluted, and thermo-mechanical stressed condition. Due to particle size limitations of HP-SEC columns, resulting in particles accumulating in the column frits prior to reaching the detector for analysis, there is a possibility that large mAb aggregates may not be detected. Both HP-SEC and HT-DLS were able to detect and resolve the mAb monomer (~10-12 nm) of the DPs in their recommended storage conditions. However, the ability to detect large aggregates (>40 nm) by both analytical methods differed, and HT-DLS was able to detect aggregates between 60 nm and 1400 nm under stress conditions. Our data indicates that HP-SEC, in conjunction with HT-DLS, may be beneficial to detect both mAb DP monomer content and multiple aggregate species (1-1000 nm) in the submicron size range.

The ELISA Detectability and Potency of Pegfilgrastim Decrease in Physiological Conditions: Key Roles for Aggregation and Individual Variability.

PEGylated recombinant human granulocyte colony stimulating factor (pegfilgrastim) is used clinically to accelerate immune reconstitution following chemotherapy and is being pursued for biosimilar development. One challenge to overcome in pegfilgrastim biosimilar development is establishing pharmacokinetic (PK) similarity, which is partly due to the degree of PK variability. We herein report that commercially available G-CSF and PEG ELISA detection kits have different capacities to detect pegfilgrastim aggregates that rapidly form in vitro in physiological conditions. These aggregates can be observed using SDS-PAGE, size-exclusion chromatography, dynamic light scattering, and real-time NMR analysis and are associated with decreased bioactivity as reflected by reduced drug-induced cellular proliferation and STAT3 phosphorylation. Furthermore, individual variability in the stability and detectability of pegfilgrastim in human sera is also observed. Pegfilgrastim levels display marked subject variability in sera from healthy donors incubated at 37 °C. The stability patterns of pegfilgrastim closely match the stability patterns of filgrastim, consistent with a key role for pegfilgrastim's G-CSF moiety in driving formation of inactive aggregates. Taken together, our results indicate that individual variability and ELISA specificity for inactive aggregates are key factors to consider when designing and interpreting studies involving the measurement of serum pegfilgrastim concentrations.

Multi-Attribute Method for Quality Control of Therapeutic Proteins.

Recent advances in high resolution mass spectrometry (MS) instrumentation and semi-automated software have led to a push toward the use of MS-based methods for quality control (QC) testing of therapeutic proteins in a cGMP environment. The approach that is most commonly being proposed for this purpose is known as the multi-attribute method (MAM). MAM is a promising approach that provides some distinct benefits compared to conventional methods currently used for QC testing of protein therapeutics, such as CEX, HILIC, and CE-SDS. Because MS-based methods have not been regularly used in this context in the past, new scientific and regulatory questions should be addressed prior to the final stages of implementation. We have categorized these questions into four major aspects for MAM implementation in a cGMP environment for both new and existing products: risk assessment, method validation, capabilities and specificities of the New Peak Detection (NPD) feature, and comparisons to conventional methods. This perspective outlines considerations for each of these main points and suggests approaches to help address potential issues.

Development of a modular virus clearance package for anion exchange chromatography operated in weak partitioning mode.

Anion exchange chromatography (AEX) operated under weak partitioning mode has been proven to be a powerful polishing step as well as a robust viral clearance step in Pfizer's monoclonal antibody (mAb) platform purification process. A multivariate design of experiment (DoE) study was conducted to understand the impact of operating parameters and feedstream impurity levels on viral clearance by weak partitioning mode AEX. Bacteriophage was used initially as a surrogate for neutral and acidic isoelectric point mammalian viruses (e.g., retrovirus and parvovirus). Five different mAbs were used in the evaluation of process parameters such as load challenge (both product and impurities), load pH, load conductivity, and contact time (bed height and flow-rate). The operating ranges obtained from phage clearance studies and Pfizer's historical data were used to define an appropriate operating range for a subsequent clearance study with model retrovirus and parvovirus. Both phage and virus clearance evaluations included feedstreams containing different levels of impurities such as high molecular mass species (HMMS), host cell proteins (HCPs), and host cell DNA. For all the conditions tested, over 5 log10 of clearance for both retrovirus and parvovirus was achieved. The results demonstrated that weak partitioning mode AEX chromatography is a robust step for viral clearance and has the potential to be included as part of the modular viral clearance approach.

Quality by design approach for viral clearance by protein a chromatography.

Protein A chromatography is widely used as a capture step in monoclonal antibody (mAb) purification processes. Antibodies and Fc fusion proteins can be efficiently purified from the majority of other complex components in harvested cell culture fluid (HCCF). Protein A chromatography is also capable of removing modest levels of viruses and is often validated for viral clearance. Historical data mining of Genentech and FDA/CDER databases systematically evaluated the removal of model viruses by Protein A chromatography. First, we found that for each model virus, removal by Protein A chromatography varies significantly across mAbs, while remains consistent within a specific mAb product, even across the acceptable ranges of the process parameters. In addition, our analysis revealed a correlation between retrovirus and parvovirus removal, with retrovirus data generally possessing a greater clearance factor. Finally, we describe a multivariate approach used to evaluate process parameter impacts on viral clearance, based on the levels of retrovirus-like particles (RVLP) present among process characterization study samples. It was shown that RVLP removal by Protein A is robust, that is, parameter effects were not observed across the ranges tested. Robustness of RVLP removal by Protein A also correlates with that for other model viruses such as X-MuLV, MMV, and SV40. The data supports that evaluating RVLP removal using process characterization study samples can establish multivariate acceptable ranges for virus removal by the protein A step for QbD. By measuring RVLP instead of a model retrovirus, it may alleviate some of the technical and economic challenges associated with performing large, design-of-experiment (DoE)-type virus spiking studies. This approach could also serve to provide useful insight when designing strategies to ensure viral safety in the manufacturing of a biopharmaceutical product.

Fermentanomics informed amino acid supplementation of an antibody producing mammalian cell culture.

Fermentanomics, or a global understanding of a culture state on the molecular level empowered by advanced techniques like NMR, was employed to show that a model hybridoma culture supplied with glutamine and glucose depletes aspartate, cysteine, methionine, tryptophan, and tyrosine during antibody production. Supplementation with these amino acids prevents depletion and improves culture performance. Furthermore, no significant changes were observed in the distribution of glycans attached to the IgG3 in cultures supplemented with specific amino acids, arguing that this strategy can be implemented without fear of impact on important product quality attributes. In summary, a targeted strategy of quantifying media components and designing a supplementation strategy can improve bioprocess cell cultures when enpowered by fermentanomics tools.

Meeting report--workshop on virus removal by filtration: trends and new developments.

The workshop was held on 27 June 2011 in Barcelona, in conjunction with the PDA Virus & TSE (transmissible spongiform encephalopathy) Safety Forum 2011. Virus-retentive filters are important tools to assure a high virus safety level of biological medicinal products. Important parameters such as properties of virus spike preparations, mechanism of virus retention by different filter brands, use of prefilters to improve the filtration performance, and, finally, strategies to select the most appropriate filter for a specific product were discussed on the workshop. The panel discussion at the end of the workshop that involved speakers and regulators from different global areas came to following conclusions: The major mechanism of virus retention is size exclusion; filtration, however, is complex and protein and virus can interact with the membrane in multiple ways. Pressure interruption during filtration resulted in enhanced virus passage. It has never been reported that murine leukemia virus (MuLV) passes a parvovirus filter. It makes sense that a small virus can be used to provide a claim for a large virus like MuLV. This relies on the assumption that there is no aggregation or interaction of the model parvovirus with proteins leading to aggregates larger than retroviruses. Several prefilters are under investigation to improve flow rate and throughput of filtration in large-scale manufacture. It was discussed whether the prefilter and the virus-retentive filter can be viewed as one unit operation so that virus retention by both can be claimed as the viral clearance capacity of this manufacturing step. This question engendered some controversy: whereas some saw the combination as a correct reflection of manufacturing conditions, others discussed the different mechanisms of virus retention, which need to be studied separately. All together, the workshop was seen as a valuable forum for the discussion between regulators and industry; it was proposed that such forum should be provided again if possible in connection with one of the next PDA Virus & TSE Safety Conferences. LAY ABSTRACT: The workshop was held on 27 June 2011 in Barcelona, in conjunction with the PDA Virus & TSE (transmissible spongiform encephalopathy) Safety Forum 2011. Virus-retentive filters are important tools to assure a high virus safety level of biological medicinal products. Important parameters such as properties of virus spike preparations, mechanism of virus retention by different filter brands, use of prefilters to improve the filtration performance and, finally, strategies to select the most appropriate filter for a specific product were discussed on the workshop. At the end of the workshop, aspects of the discussion were summarized by the following: The major mechanism of virus retention is size exclusion, but interactions are complex. Pressure interruption during filtration resulted in enhanced virus passage. It has never been reported that murine leukemia virus (MuLV) passes a parvovirus filter, and thus the parvovirus may provide a claim for a large virus like MuLV. Combination of prefilter and the virus-retentive filter are seen by some panelists as a correct reflection of manufacturing conditions; others discussed the different mechanisms of virus retention, which need to be studied separately. All together, the workshop was seen as a valuable forum for the discussion between regulators and industry.

Bioreactor environment-sensitive sentinel genes as novel metrics for cell culture scale-down comparability.

Scale-down of bioreactors is currently done based on matching one or more measurable parameters such as k(L) a and P/V, which could result in insufficient process comparability. Currently, there is a lack of genomic translational studies in cell culture scale-down, which could help delineate measurable cellular attributes for improved scale-down. In this study, we scaled-down from a typical bench-scale 5-L bioreactor to a novel high-throughput 35-mL minibioreactor based on matching oxygen transfer rate, which resulted in cell growth and product-related discrepancies using Sp2/0 cells. Performing DNA microarrays on time-course samples from both systems, we identified ∼200 differentially expressed transcripts, presumably because of bioreactor aeration and mixing differences with scale-down. Evaluating these transcripts for bioreactor-relevant cellular functions such as oxidative stress response and DNA damage response, we chose 18 sentinel genes based on their degree of difference and functionality, which we further verified by quantitative real-time polymerase chain reaction (qRT-PCR). Tracking the differential expression of Sod1, Apex1, and Odc1 genes, we were able to correlate sparging-related damage and poor mixing, as possible causes for physiological changes such as prolonged culture in minibioreactors. Additionally, to verify our sentinel gene findings, we performed follow-up improved scale-down studies based on gene analysis and measured transcriptomic changes. As a result, qRT-PCR-based genomic profiles and cell growth profiles showed better convergence between the improved minibioreactor conditions and the model 5-L bioreactor. Our results broadly show that based on the knowledge from transcriptomic changes of sentinel gene profiles, it is possible to improve bioreactor scale-down for more comparable processes.

A case study in converting disposable process scouting devices into disposable bioreactors as a future bioprocessing tool.

In this study, we perform mass transfer characterization (k(L) a) on a novel mechanically driven/stirred Process Scouting Device, PSD, (SuperSpinner D 1000®, SSD) and demonstrate that this novel device can be viewed as disposable bioreactor. Using patch-based optical sensors, we were able to monitor critical cell culture environmental conditions such as dissolved oxygen (DO) and pH in SSD for comparison to a 1 L standard spinner (SS) flask. We also coupled these mass transfer studies with mixing time studies where we observed relative high mixing times (5.2 min) that are typically observed in production scale bioreactors. Decreasing the mixing time 3.5-fold resulted in 30% increase in k(L) a (from 2.3 to 3.0 h(-1) ) and minimum DO level increased from 0% to 20% for our model hybridoma cell line. Finally, maximum viable cell density and protein titer stayed within ±20% of historical data, from our standard 5 L stirred bioreactor (Biostat®) operated under active DO control.

Dissolved oxygen and pH profile evolution after cryovial thaw and repeated cell passaging in a T-75 flask.

Routine cell culture is done in small-scale disposable vessels (typically 0.1-100 mL volumes) in academia and industry. Despite their wide use in bioprocess development (i.e., process optimization and process validation), miniature process scouting devices (PSDs) are considered "black boxes" because they are generally not equipped with sensors. In this study, we show that on-line monitoring of dissolved oxygen (DO) and pH in a T-75 flask-based PSD can be achieved during cell passaging and that this information can be linked to different cellular metabolic states. In this case, on-line monitoring of DO and pH show three distinctive metabolic regions in passages 1-18, 19-28, 29-54 and in particular, the shift in the pH curve, the specific oxygen uptake rate (q(O2)), and the lactate production rate to the oxygen consumption rate yield (Y(Lac/ox)) confirm the existence of these distinctive metabolic regions. These findings are particularly useful because they show that sensor equipped PSDs can help to monitor cell culture behavior after thaw, in pre- and seed culture prior to scale-up and in development/optimization studies. Such routine monitoring will help to develop more consistent cell culture techniques.

Multiplex RT Q-PCR assay for simultaneous quantification of three viruses used for validation of virus clearance by biopharmaceutical production.

Virus removal studies are used to insure the safety of biopharmaceutical products by quantitatively estimating the viral clearance capacity by the manufacturing process. Virus quantification assays are used to measure the log(10) clearance factor of individual purification unit operations in spike recovery studies. We have developed a multiplex RT Q-PCR assay that detects and quantifies three commonly used model viruses X-MuLV, SV40, and MMV simultaneously. This RT Q-PCR multiplex assay has a 6log(10) dynamic range with a limit of detection (LOD) of approximately 1 genome copy/microL. Amplification profiles are similar to existing singleplex assays. Overall, this RT Q-PCR multiplex assay is highly quantitative, accurately identifies multiple viruses simultaneously, and may prove useful to validate viral clearance of biological products in small scale studies.

A novel, Q-PCR based approach to measuring endogenous retroviral clearance by capture protein A chromatography.

Quantification of virus removal by the purification process during production is required for clinical use of biopharmaceuticals. The current validation approach for virus removal by chromatography steps typically involves time-consuming spiking experiments with expensive model viruses at bench scale. Here we propose a novel, alternative approach that can be applied in at least one instance: evaluating retroviral clearance by protein A chromatography. Our strategy uses a quantitative PCR (Q-PCR) assay that quantifies the endogenous type C retrovirus-like particle genomes directly in production Chinese Hamster Ovary (CHO) cell culture harvests and protein A pools. This eliminates the need to perform spiking with model viruses, and measures the real virus from the process. Using this new approach, clearance values were obtained that was comparable to those from the old model-virus spike/removal approach. We tested the concept of design space for CHO retrovirus removal using samples from a protein A characterization study, where a wide range of chromatographic operating conditions were challenged, including load density, flow rate, wash, pooling, temperature, and resin life cycles. Little impact of these variables on CHO retrovirus clearance was found, arguing for implementation of the design space approach for viral clearance to support operational ranges and manufacturing excursions. The viral clearance results from Q-PCR were confirmed by an orthogonal quantitative product-enhanced reverse transcriptase (Q-PERT) assay that quantifies CHO retrovirus by their reverse transcriptase (RT) enzyme activity. Overall, our results demonstrate that protein A chromatography is a robust retrovirus removal step and CHO retrovirus removal can be directly measured at large scale using Q-PCR assays.

Robustness of virus removal by protein A chromatography is independent of media lifetime.

The robustness of virus clearance with respect to protein A media reuse was demonstrated using media with four matrix chemistries: Protein A immobilized ProSep A, Poros A50, Protein A ceramic Hyper DF and MabSelect SuRe, an alkali resistant protein A ligand. Endogenous retrovirus clearance, step yield, impurity clearance and other performance parameters were evaluated periodically in media cycled up to 300 times. Media lifetime was generally limited by either declining step yield or media fouling. However, clearance of endogenous retrovirus remained in an acceptable range, either increasing or remaining constant. Multiply cycled media were tested for clearance of three viruses (SV40, X-MuLV, and MMV); clearance was comparable to naïve media. Overall, virus clearance by protein A chromatography appears to be extremely robust with respect to media age.

Comparisons of optical pH and dissolved oxygen sensors with traditional electrochemical probes during mammalian cell culture.

Small-scale upstream bioprocess development often occurs in flasks and multi-well plates. These culturing platforms are often not equipped to accurately monitor and control critical process parameters; thus they may not yield conditions representative of manufacturing. In response, we and others have developed optical sensors that enable small-scale process monitoring. Here we have compared two parameters critical to control in industrial cell culture, pH and dissolved oxygen (DO), measured with our optical sensors versus industrially accepted electrochemical probes. For both optical sensors, agreement with the corresponding electrochemical probe was excellent. The Pearson Correlations between the optical sensors and electrochemical probes were 98.7% and 99.7%, for DO and pH, respectively. Also, we have compared optical pH sensor performance in regular (320 mOsm/kg) and high-osmolality (450 mOsm/kg) cell culture media to simulate the increase in osmolality in pH-controlled cultures. Over a pH range of 6.38-7.98 the average difference in pH readings in the two media was 0.04 pH units. In summary, we have demonstrated that these optical sensors agree well with standard electrochemical probes. The accuracy of the optical probes demonstrates their ability to detect potential parameter drift that could have significant impact on growth, production kinetics, and protein product quality. We have also shown that an increase in osmolality that could result from controlling pH or operating the reactor in fed-batch mode has an insignificant impact on the functionality of the pH patches.

Validation of an optical sensor-based high-throughput bioreactor system for mammalian cell culture.

Cell culture optimization is a labor-intensive process requiring a large number of experiments to be conducted under varying conditions. Here we describe a high-throughput bioreactor system that allows 12 mini stirred-tank bioreactors to be operated simultaneously. All bioreactors are monitored by low-cost minimally invasive optical sensors for pH and dissolved oxygen. The sensors consist of single-use patches affixed inside the bioreactors and monitored optically from the outside. Experimental results show that different sensing patches with the same composition respond consistently. The discrepancy between different pH sensors is less than 0.1 pH units over most of their responsive range. The discrepancy between different dissolved oxygen sensors is less than 10% over the whole range from 0% to 100% dissolved oxygen. The consistency of the sensing system ensures that only an initial one-time calibration is required for the sensing patches. After that, a calibration code is generated and sensing patches of the same composition can be used directly. This greatly reduces the time and cost required for monitored multi-bioreactor operations. We used SP2/0 myeloma/mouse hybridoma cell cultures to demonstrate reactor performance consistency. Transcriptional profiling, HPLC analysis, viable cell count, and viability inspection show that the presence of sensing patches and the use of optical monitoring have no apparent effect on the metabolism of the cells.

Generic/matrix evaluation of SV40 clearance by anion exchange chromatography in flow-through mode.

The potential of viral contamination is a regulatory concern for continuous cell line-derived pharmaceutical proteins. Complementary and redundant safety steps, including an evaluation of the viral clearance capacity of unit operations in the purification process, are performed prior to registration and marketing of biotechnology pharmaceuticals. Because process refinement is frequently beneficial, CBER/FDA has published guidance facilitating process improvement by delineating specific instances where the bracketing and generic approaches are appropriate for virus removal validation. In this study, a generic/matrix study was performed using Q-Sepharose Fast Flow (QSFF) chromatography to determine if bracketing and generic validation can be applied to anion exchange chromatography. Key operational parameters were varied to upper and lower extreme values and the impact on viral clearance was assessed using simian virus 40 (SV40) as the model virus. Operational ranges for key chromatography parameters were identified where an SV40 log(10) reduction value (LRV) of >or=4.7 log(10) is consistently achieved. On the basis of the apparent robustness of SV40 removal by Q-anion exchange chromatography, we propose that the concept of "bracketed generic" validation can be applied to this and potentially other chromatography unit operations.

Identification of protein A media performance attributes that can be monitored as surrogates for retrovirus clearance during extended re-use.

A potential safety concern in biotechnology purification schemes that employ re-use of column media, often for large numbers of chromatography runs, is loss of the virus removal capacity of the chromatographic purification operation over time. To define chromatography performance attributes that best predict retrovirus clearance during extended re-use of protein A media, small-scale protein A columns were cycled 150 to 460 times using concentrates of murine hybridoma cell culture supernatants, standard low pH elution buffers and different cleaning solutions (6 M urea, 6 M guanidine, 100 mM NaOH or 500 mM NaOH). Load, flow-through and eluate samples were taken periodically and assayed for reverse transcriptase (RT, an enzyme component of retroviruses) activity, bovine IgG (a component of the culture media), genomic DNA, leached protein A, and mouse IgG. Under all cleaning conditions tested, the log,10 reduction value (LRV) of RT activity did not decrease and impurity co-elution did not increase during the 150 to 460 purification/cleaning cycles. In the two studies in which the columns were cleaned with NaOH, the chromatography performance attribute that best predicted the column media lifespan was column capacity, as measured by antibody (Ab) step yield and breakthrough. In both studies, Ab capture decayed in a biphasic manner starting at cycle 200 (100 mM NaOH) or cycle 50 (500 mM NaOH). For media cycled 300+ times using 6 M urea or 6 M guanidine cleaning buffers, column performance, including RT activity LRV, was more stable, although small upward trends in Ab breakthrough were evident. In summary, our studies identify Ab step yield and breakthrough as performance attributes that decay prior to retrovirus LRV when protein A media is multiply-cycled. Thus, we propose that virus removal validation studies should be performed on new media only and these attributes can be monitored during protein A unit operations in lieu of performing virus removal validation studies with cycled protein A media.

Bracketed generic inactivation of rodent retroviruses by low pH treatment for monoclonal antibodies and recombinant proteins.

Viral safety is a predominant concern for monoclonal antibodies (mAbs) and other recombinant proteins (RPs) with pharmaceutical applications. Certain commercial purification modules, such as nanofiltration and low-pH inactivation, have been observed to reliably clear greater than 4 log(10) of large enveloped viruses, including endogenous retrovirus. The concept of "bracketed generic clearance" has been proposed for these steps if it could be prospectively demonstrated that viral log(10) reduction value (LRV) is not impacted by operating parameters that can vary, within a reasonable range, between commercial processes. In the case of low-pH inactivation, a common step in mAb purification processes employed after protein A affinity chromatography, these parameters would include pH, time and temperature of incubation, the content of salts, protein concentration, aggregates, impurities, model protein pI, and buffer composition. In this report, we define bracketed generic clearance conditions, using a prospectively defined bracket/matrix approach, where low-pH inactivation consistently achieves >or=4.6 log(10) clearance of xenotropic murine leukemia virus (X-MLV), a model for rodent endogenous retrovirus. The mechanism of retrovirus inactivation by low-pH treatment was also investigated.

Evaluation of a quantitative product-enhanced reverse transcriptase assay to monitor retrovirus in mAb cell-culture.

Murine hybridoma cells used in the production of monoclonal antibodies (mAbs) produce endogenous type C retrovirus particles. Regulatory agencies require a demonstration that mAbs intended for human use are free of retrovirus with an adequate margin of safety. This is usually achieved by evaluation studies, performed at small scale, to demonstrate that the manufacturing process is capable of removing or inactivating several different model viruses, including a murine retrovirus. In a previous report, we demonstrated the utility of TaqMan fluorogenic 5'-nuclease product-enhanced reverse transcriptase (TM-PERT) assays for measuring reverse transcriptase (RT) activity in laboratory-scale cell-culture samples and RT removal by laboratory-scale models of processing steps. In this report, we evaluate the specificity, accuracy, range, precision and robustness of TM-PERT for this purpose. We find that this assay detects RT activity contained in xenotropic murine leukemia virus (X-MuLV) and CHO cell type C particles and quantifies particle numbers comparably to other assays (e.g. transmission electron microscopy, viral sequence specific TaqMan). Cell derived DNA polymerases appear to contribute only modestly to the assay background and RT activity in clarified cell culture harvests is contained largely in Type C particles. TM-PERT is linear and precise between 10(7)and 10(13) pU/ml, establishing the assay range. The assay is robust in that test article storage condition and DNA/protein content had little impact on assay performance. Thus, TM-PERT appears to be an acceptable assay to measure type C particles in rodent cell culture samples.