Scalability of pre-packed preparative chromatography columns with different diameters and lengths taking into account extra column effects

Mechanistic modeling of minute virus of mice surrogate removal by anion exchange chromatography in micro scale

Biopharmaceutical products are often produced in Chinese hamster ovary (CHO) cell cultures that are vulnerable to virus infections. Therefore, it is a regulatory requirement that downstream purification steps for biopharmaceuticals can remove viruses from feedstocks. Anion exchange chromatography (AEX) is one of the downstream unit operations that is most frequently used for this purpose and claimed for its capability to remove viruses. However, the impact of various process parameters on virus removal by AEX is still not fully understood. Mechanistic modeling could be a promising way to approach this gap, as these models require comparatively few experiments for calibration. This makes them a valuable tool to improve understanding of viral clearance, especially since virus spiking studies are costly and time consuming. In this study, we present how the virus clearance of a MVM mock virus particle by Q Sepharose FF resin can be described by mechanistic modeling. A lumped kinetic model was combined with a steric mass action model and calibrated at micro scale using three linear gradient experiments and an incremental step elution gradient. The model was subsequently verified for its capability to predict the effect of different sodium chloride concentrations, as well as residence times, on virus clearance and was in good agreement with the LRVs of the verification runs. Overall, models like this could enhance the mechanistic understanding of viral clearance mechanisms and thereby contribute to the development of more efficient and safer biopharmaceutical downstream processes.

Status and future developments for downstream processing of biological products: Perspectives from the Recovery XIX yield roundtable discussions

Governments and biopharmaceutical organizations aggressively leveraged expeditious communication capabilities, decision models, and global strategies to make a COVID-19 vaccine happen within a period of 12 months. This was an unusual effort and cannot be transferred to normal times. However, this focus on a single vaccine has also led to other treatments and drug developments being sidelined. Society expects the pharmaceutical industry to provide an uninterrupted supply of medicines. However, it is often overlooked how complex the manufacture of these compounds is and what logistics are required, not to mention the time needed to develop new drugs. The overarching theme, therefore, is patient access and how we can help ensure access and extend it to low- and middle-income countries. Despite unceasing efforts to make medications available to all patient populations, this must never be done at the expense of patient safety. A major fraction of the costs in biopharmaceutical manufacturing are for drug discovery, process development, and clinical studies. Infrastructure costs are very difficult to quantify because they often depend on whether a greenfield facility or an existing, depreciated facility is used or adapted for a new product. To accelerate process development concepts of platform process and prior knowledge are increasingly leveraged. While more traditional protein therapeutics continue to dominate the field, we are also experiencing the exciting emergence and evolution of other therapeutic formats (bispecifics, tetravalent mAbs, antibody-drug conjugates, enzymes, peptides, etc.) that offer unique treatment options for patients. Protein modalities are still dominant, but new modalities are being developed that can be learned from including advanced therapeutics-like cell and gene therapies. The industry must develop a model-based strategy for process development and technologies such as continuous integrated biomanufacturing must be adopted. The overall conclusion is that the pandemic pace was unsustainable, focused on vaccine delivery at the expense of other modalities/disease targets, and had implications for professional and personal life (work-life balance). Routinely reducing development time from 10 years to 1 year is nearly impossible to achieve. Environmental aspects of sustainable downstream processing are also described.

Factors affecting therapeutic protein purity and yield during chromatographic purification

Therapeutic proteins are recombinant proteins generated through recombinant DNA technology and have attracted a great deal of interest in numerous applications, including pharmaceutical, cosmetic, human and animal health, agriculture, food, and bioremediation. Producing therapeutic proteins on a large scale, mainly in the pharmaceutical industry, necessitates a cost-effective, straightforward, and adequate manufacturing process. In industry, a protein separation technique based mainly on protein characteristics and modes of chromatography will be applied to optimize the purification process. Typically, the downstream process of biopharmaceutical operations may involve multiple chromatography phases that require the use of large columns pre-packed with resins that must be inspected before use. Approximately 20% of the proteins are assumed to be lost at each purification stage during the production of biotherapeutic products. Hence, to produce a high quality product, particularly in the pharmaceutical industry, the correct approach and understanding of the factors influencing purity and yield during purification are necessary.

Flow behavior of protein solutions in a lab-scale chromatographic system

A fluid dynamics model has been developed to describe flow behavior in a lab-scale chromatographic system dedicated for protein processing. The case study included a detailed analysis of elution pattern of a protein, which was a monoclonal antibody, glycerol, and their mixtures in aqueous solutions. Glycerol solutions mimicked viscous environment of the concentrated protein solutions. The model accounted for concentration dependences of solution viscosity and density, and dispersion anisotropy in the packed bed. It was implemented into a commercial computational fluid dynamics software using user-defined functions. The prediction efficiency was successfully verified by comparing the model simulations in the form of the concentration profiles and their variances with the corresponding experimental data. The contribution of the individual elements of the chromatographic system to protein band broadening was evaluated for different configurations: for the extra-column volumes in the absence of the chromatographic column, for the zero-length column without the packed bed and for the column containing the packed bed. The influence of the operating variables, including: the mobile phase flowrate, the type of the injection system, i.e., the injection loop capillary or the superloop, the injection volume and the length of the packed bed, on band broadening of the protein was determined under nonadsorbing conditions. For protein solutions having viscosity comparable with the mobile phase, the flow behavior either in the column hardware or in the injection system made major contributions to band broadening, which depended on the type of the injection system. For highly viscous protein solution, the flow behavior in the packed bed exerted a dominant influence on band broadening.

Mechanistic modeling of cation exchange chromatography scale-up considering packing inhomogeneities

In process development and characterization, the scale-up of chromatographic steps is a crucial part and brings a number of challenges. Usually, scale-down models are used to represent the process step, and constant column properties are assumed. The scaling is then typically based on the concept of linear scale-up. In this work, a mechanistic model describing an anti-Langmuirian to Langmuirian elution behavior of a polypeptide, calibrated with a pre-packed 1 ml column, is applied to demonstrate the scalability to larger column volumes up to 28.2 ml. Using individual column parameters for each column size, scaling to similar eluting salt concentrations, peak heights, and shapes is experimentally demonstrated by considering the model's relationship between the normalized gradient slope and the eluting salt concentration. Further scale-up simulations show improved model predictions when radial inhomogeneities in packing quality are considered.

Prediction of the performance of pre-packed purification columns through machine learning

Pre-packed columns have been increasingly used in process development and biomanufacturing thanks to their ease of use and consistency. Traditionally, packing quality is predicted through rate models, which require extensive calibration efforts through independent experiments to determine relevant mass transfer and kinetic rate constants. Here we propose machine learning as a complementary predictive tool for column performance. A machine learning algorithm, extreme gradient boosting, was applied to a large data set of packing quality (plate height and asymmetry) for pre-packed columns as a function of quantitative parameters (column length, column diameter, and particle size) and qualitative attributes (backbone and functional mode). The machine learning model offered excellent predictive capabilities for the plate height and the asymmetry (90 and 93%, respectively), with packing quality strongly influenced by backbone (∼70% relative importance) and functional mode (∼15% relative importance), well above all other quantitative column parameters. The results highlight the ability of machine learning to provide reliable predictions of column performance from simple, generic parameters, including strategic qualitative parameters such as backbone and functionality, usually excluded from quantitative considerations. Our results will guide further efforts in column optimization, for example, by focusing on improvements of backbone and functional mode to obtain optimized packings.

Mechanistic modeling and simulation of a complex low and high loading elution behavior of a polypeptide in cation exchange chromatography

The mechanistic modeling of preparative liquid chromatography is still a challenging task. Non-ideal thermodynamic conditions may require activity coefficients for the mechanistic description of preparative chromatography. In this work, a chromatographic cation exchange step with a polypeptide having a complex elution behavior in low and high loading situations is modeled. Model calibration in the linear range of the isotherm is done by applying counterion-induced linear gradient elution experiments between pH 3.3 and pH 4.3. Inverse fitting with column loads up to 25 mg/mL_CV is performed for parameter estimation in the non-linear range. The polypeptide elution peak shows an anti-Langmuirian behavior with fronting under low loading conditions and a switch to a Langmuirian behavior with increasing load. This unusual elution behavior could be described with an extended version of the sigmoidal Self-Association isotherm, including two activity coefficients for the polypeptide and counterion in solution. The activity coefficient of the solute polypeptide shows a strong influence on the model parameters and is crucial in the linear and non-linear range of the isotherm. The modeling procedure results in a unique and robust model parameter set that is sufficient to describe the complex elution behavior and allows modeling over the full isotherm range. This article is protected by copyright. All rights reserved.

Facile production of quercetin nanoparticles using 3D printed centrifugal flow reactors

A 3D printed reactor-in-a-centrifuge (RIAC) was developed to produce drug nanocrystals. Quercetin nanocrystals were manufactured at varying operational and formulation conditions, and had a small size (190–302 nm) and low size dispersity (PDI < 0.1).

Influence of the geometry of extra column volumes on band broadening in a chromatographic system. Predictions by computational fluid dynamics

A computational fluid dynamics method was used for prediction of flow behavior and band profiles of small- and macro-molecule compounds eluting in extra-column volumes (ECV) of an Äkta chromatographic system. The model compounds were: acetone, bovine serum albumin and an antibody. The construction of ECV was approximated by different types of geometries, starting from the simplest two-dimensional (2D) arrangement consisting of a straight capillary tube, and ending with a three-dimensional system (3D), which accounted for the flow path curvature of individual elements of ECV, including: injection loop capillary, multi-way valve, connecting capillary and detector cell. The accuracy of the model predictions depended on the flow path length and the eluent flowrate. 2D-geometry models reproduced pretty well the shapes of band profiles recorded at the lowest eluent flowrate used, but they failed for increased flowrates. The 3D-geometry model was found to be sufficiently accurate for all conditions investigated. It was exploited to analyze band broadening in the individual ECV elements. The simulation results revealed that the flow behavior in the injection loop capillaries strongly influenced the shape of band profiles, particularly at higher eluent velocities. This was attributed to the formation of Dean vertices triggered by centrifugal forces in curved parts of the eluent flow path.

A revised definition of dynamic adsorption coefficient for characterizing activated carbon instead of retention bed

Activated carbon (AC) retention beds are widely used in nuclear facilities, removing radioactive contaminants from exhaust air. Dynamic adsorption coefficient (DAC) is the core parameter to quantify the performance. Its definition has not been unified and it is affected by the geometry of the retention bed, the presence, the flow rate, and the concentration of adsorbate. So, DAC is currently a parameter characterizing the adsorption performance of the retention bed instead of the AC. In this regard, the definition of DAC should be revised, stripping away the influence of other factors. In this study, a 1D model for the AC column, a 2D model for blank piping, and a mathematical model for retention factor is developed. All are validated with simulations and experiments based on the "pulse dynamic method". They are used to analyze the factors affecting DAC quantitatively in detail, including the direct effect of blank piping, the indirect effect of blank piping by affecting the pulse height into the column, and the effect of krypton concentration distribution in the column. Finally, an improved definition of DAC characterizing AC instead of retention bed is given. This definition can be used as a reference for scholars who formulate relevant standards.

A cuboid chromatography device having short bed-height gives better protein separation at a significantly lower pressure drop than a taller column having the same bed-volume

Simultaneously reducing the bed-height and increasing the area of cross-section, while keeping the bed-volume the same, would substantially reduce the pressure drop across a process chromatography column. This would minimize problems such as resin compaction and non-uniformity in column packing, which are commonly faced when using soft chromatographic media. However, the increase in macroscale convective dispersion due to the increase in column diameter, and the resultant loss in resolution would far outweigh any potential benefit. Cuboid-packed bed devices have lower macroscale convective dispersion compared to their equivalent cylindrical columns. In this paper, we discuss how and why a flat cuboid chromatography device having a short bed-height gives better protein separation, at a significantly lower pressure drop, than a taller column having the same bed-volume. First, we explored this option based on computational fluid dynamic (CFD) simulations. Depending on the flow rate, the pressure drop across the flat cuboid device was lower than that in the tall column by a factor of 6.35 to 6.4 (i.e. less than 1/6^th the pressure). The CFD results also confirmed that the macroscale convective dispersion within the flat cuboid device was significantly lower. Head-to-head separation experiments using a 1 mL flat cuboid device having a bed-height of 10 mm, and a 1 mL tall column having a bed-height of 25.8 mm, both packed with the same chromatographic media, were carried out. The number of theoretical plates per unit bed-height was on an average, around 2.5 time times greater with the flat cuboid device, while the total number of theoretical plates in the two devices were comparable. At any given superficial velocity, the height equivalent of a theoretical plate in the tall column was on an average, higher by a factor 2.5. Binary protein separation experiments showed that at any given flow rate, the resolution obtained using the flat cuboid device was significantly higher than that obtained with the tall column. This work opens up the possibility of designing and developing short bed-height chromatography devices for carrying out high-resolution biopharmaceutical purifications, at very low pressures.

Intensified Downstream Processing of Monoclonal Antibodies Using Membrane Technology

The need to intensify downstream processing of monoclonal antibodies to complement the advances in upstream productivity has led to increased attention toward implementing membrane technologies. With the industry moving toward continuous operations and single use processes, membrane technologies show promise in fulfilling the industry needs due to their operational flexibility and ease of implementation. Recently, the applicability of membrane-based unit operations in integrating the downstream process has been explored. In this article, the major developments in the application of membrane-based technologies in the bioprocessing of monoclonal antibodies are reviewed. The recent progress toward developing intensified end-to-end bioprocesses and the critical role membrane technology will play in achieving this goal are focused upon.

Relationship between HETP measurements and breakthrough curves in short chromatography columns

An analysis of the relationship between the number of plates measured with a small molecule tracer and the breakthrough curve of a strongly bound protein in short laboratory chromatography columns (1-5 cm) considering flow nonuniformity is presented. For practical conditions, while axial dispersion has only a small impact on the breakthrough curve, radial flow nonuniformity has a profound effect. Radial parabolic velocity profiles lead to tailing tracer peaks and broader breakthrough curves. Profiles where the velocity varies radially only in a thin region near the column wall lead to fronting tracer peaks and early breakthrough when the velocity at the wall is higher than the average and to tailing peaks and tailing breakthrough curves when the velocity at the wall is lower than the average. Experiments conducted in laboratory minicolumns (0.5-1 cm diameter, 0.5-1 ml volume) show tracer peaks and protein breakthrough curves that are consistent with higher velocities at the wall. The model presented in this work provides a tool to model experimental breakthrough data and to assess the degree of flow uniformity required to obtain meaningful dynamic binding capacity measurements using minicolumns in a high-throughput lab setting.

Compartment Model of Mixing in a Bubble Trap and Its Impact on Chromatographic Separations

Chromatography equipment includes hold-up volumes that are external to the packed bed and usually not considered in the development of chromatography models. These volumes can substantially contribute to band-broadening in the system and deteriorate the predicted performance. We selected a bubble trap of a pilot scale chromatography system as an example for a hold-up volume with a non-standard mixing behavior. In a worst-case scenario, the bubble trap is not properly flushed before elution, thus causing the significant band-broadening of the elution peak. We showed that the mixing of buffers with different densities in the bubble trap device can be accurately modeled using a simple compartment model. The model was calibrated at a wide range of flow rates and salt concentrations. The simulations were performed using the open-source software CADET, and all scripts and data are published with this manuscript. The results illustrate the importance of including external holdup volumes in chromatography modeling. The band-broadening effect of tubing, pumps, valves, detectors, frits, or any other zones with non-standard mixing behavior can be considered in very similar ways.