April 01, 2020

Applications of modern high-throughput process development

By Cytiva

Highlights from the 5th International HTPD Conference, Part 2

The High-Throughput Process Development (HTPD) Conference Series is the key international forum for the presentation and discussion of topics relevant to high-throughput process development and “smart process development” for biopharmaceuticals.

The first of three articles in this series from the 5th conference provided takeaways from the preconference day which focused on giving an introduction to high throughput process development. This article focuses on “Application of Modern HTPD”. A third article covers “Smart process development, related analytics and a look into the future”.

Emerging high throughput techniques for developing processes of mAbs and antibody derivatives

High-throughput approaches are well established for the process development of mAbs and related molecules. Nevertheless, this area of HTPD is still open to innovation. New modalities, including mAb derivatives such as antibody fragments and bispecific antibodies, present new and more complex process development challenges. Some new technologies, such as continuous processing, might bring new challenges how HTPD approaches should be best applied. Other new technologies, like fiber-based chromatography, bring additional alternatives for speeding up process development.

Fibro chromatography does in minutes what traditional chromatography does in hours

Jennifer Spooner of AstraZeneca, Cambridge UK described an HTPD study related to fiber-based Protein A affinity chromatography [1]. The Biologics Expression Team she is part of expresses up to 1000 proteins a week and screens approximately 100 in regard to Protein A chromatography. They currently use thirteen ÄKTAxpress chromatography systems to purify samples using different modalities and scales.

At such early fermentation development stages, target protein expression is low and significant volumes of feed must be processed to obtain the protein mass needed for process development studies. This limitation can be overcome by utilizing the technology of a Protein A-derivatized cellulose fiber (Fibro PrismA), see Figure 1. The high flow rate capability of Fibro PrismA means that rapid cycling of the unit can be exploited so that a standard ÄKTA avant purification system can process hundreds of samples quickly and efficiently.

AstraZeneca´s approach is scalable, allowing volumes from just a few milliliter to larger sample volumes of 500 mL or more to be processed using different sized Fibro units. “A single Fibro PrismA unit allows large numbers of samples in the 10–25 mL range to be purified in quick succession. Using flow rates of 16 mL/min, each sample run takes approximately 9 min including a full 6 column volume (CV) cleaning-in-place (CIP) and re-equilibration in between each run.” As such, 100 samples can be processed in 16 h with one ÄKTA avant system, instead of over one week using thirteen ÄKTAxpress systems with 1 mL columns [1]. Jennifer pointed out that the method they are using does not require advanced robotic systems. Their experience is that Fibro PrismA results (HCP, LMW, HMW, yield) compare well with MabSelect PrismA resin column-based results even when the 0.4 mL Fibro-based units are flowing at 4 to 16 mL/min, which is 4-fold faster than column flow rates.

Read Case study: high throughput mAb purification with Fibro

Fig 1. Fibro chromatography is based on electrospun cellulose fibers that enables the combination of high flow rates with high binding capacity. This technology offers new opportunities for flexible, high-productivity purification solutions from process development to manufacturing scale.

Miniaturized continuous chromatography combines faster process development and enhanced chromatography performance

Razwan Hanif and Michael Rose of UCB presented on automated high-throughput and miniaturized semi-continuous chromatography [2] noting that continuous chromatography offers reduced buffer consumption, plus increased resin capacity and productivity of packed bed operations. UCB is using a novel, semi-continuous operation involving an unmodified commercial chromatography “skid” and Sequential Chromatography Recycling with Asynchronous Multiplexing (SCRAM). They have developed the ability to operate SCRAM using 600 µL microscale columns on an automated robotic platform to obtain information on optimal process parameters faster and earlier in the process development process.

HTPD is used beyond chromatography

HTPD continues to move beyond its initial chromatography-based applications with several presentations showcasing its use for filtration and capture membrane unit operation scouting. Lara Fernandez-Cerezo and coworkers, of Merck & Co, presented the use of HTPD to predict pilot-scale tangential flow filtration (TFF) performance during monoclonal antibody processing. The effects of loading (2 g/L to 25 g/L) on transmembrane pressure, as well as crossflow rate on the flux and product quality for two different monoclonal antibodies were presented [3].

Vaccine production, exosome production, and other emerging applications

Cell, gene, nucleic acid therapies, and other novel therapeutic modalities have unique production, analytical, and drug delivery challenges. Therefore, in contrast to mAbs, the possibility to use a platform approach for these novel molecules is limited, and process developers for these emerging therapies face new challenges. This topic was an important focus area during HTPD conference; the session gathered the highest number of abstracts, and the session was also extended to fit as many talks as possible.

Exosomes

Michael F. Doherty of Codiak Biosciences presented on HTPD for exosomes, which are bionanoparticles produced by cells and which can be taken up by other cells [4]. HTPD is particularly attractive for such bionanoparticles, as current purification methods which include gradient ultracentrifugation, are not readily scalable for manufacturing.

The development of robust high throughput screening (HTS) tools for exosomes reduces process development timelines, improves process performance, and serves as a model for integrating bio-nanoparticles with HTPD. The method uses the framework developed for recombinant protein and monoclonal antibodies, however minimal changes to standard robotic screening workflows were required.

Initial screening approaches involved light scattering, however the 50–200 nm size of the exosomes made this challenging and error prone. Codiak Biosciences found that rapid size exclusion chromatography (SEC) HPLC allowed for rapid identification of promising capture resins to be evaluated further in their process development.

Viral vectors

With the rapid development of cell and gene therapy programs, virus and viral vector process development were also in focus. For example, Vimal H. Vaidya of bluebird bio [5] noted their success at using HTPD for development with mini-columns of an ion exchange step for primary capture of a lentiviral vector.

Sheng-Ching Wang of MSD [6] discussed an automated HTPD chromatography screening platform for virus purification. He noted limitations imposed on standard porous chromatography resins by the relatively large size of the viral particles. These limitations cannot be entirely overcome by use of convective media due to the reduced choice of ligands available for the latter.

Merck employs an entirely automated system to screen a wide range of bead and membrane stationary phases. The screening covers the entire process development workflow, from buffer preparation to chromatography data analysis. Generic and flexible robotic methods are deployed for analytical testing, and their results are compiled and processed in an automated fashion through custom developed tools.

Cell therapy

Guy Caspary of Juno Therapeutics presented on use of high-throughput approach for chimeric antigen receptor T cell (CAR-T) immunotherapies [7]. Such cell-based therapies present several challenges including those related to scaling for a single patient-specific lot of drug product. Guy Caspary noted that “the unique challenge in cell therapy manufacturing is reducing the cost per dose, given that there are currently few economies of scale to exploit.”

Juno Therapeutics have developed and qualified an agitated scale-down model system which allows them to model and characterize a CAR-T manufacturing process in a manner that reduces process development costs. They are currently working to add in-process monitoring of pH, dissolved oxygen (DO), metabolic markers, and cytokines. Personalized therapy HTPD was also addressed by Chris Ladd of Moderna Therapeutics regarding personalized mRNA cancer vaccine [8]. He presented how “mRNA process unit operations from both upstream and downstream steps can be automated, designed and intensified. Yield, purity, and productivity are critical attributes for mRNA processes.” Standard mRNA chromatography approaches usually work well but come with target related challenges, such as reduction in dynamic binding capacity. HTPD provides a good approach to identify optimal chromatography conditions.

Vaccines

More classic vaccine process development was also discussed at the meeting. Jessica Olson and coworkers of Vaccines Process Development, MSD, [9] presented the use of HTPD to scout microcarrier production system for vaccine candidates. Anchorage-dependent cell lines grown on microcarrier beads present unique challenges to cultivation in scale-down systems. These challenges can be addressed through vessel modifications, improvement in motor speed control, and careful liquid handling practices. The microbial vaccine candidate HTPD data demonstrated good comparability to standard lab-scale systems following evaluation of several scale-up methods and sampling method improvement.

Bruno Andre of GlaxoSmithKline Biologicals, Rixensart, Belgium presented a poster on “Development of a High Throughput Chromatography Screening Platform to expediate vaccine process development” [10]. GSK has set up a high-throughput chromatography screening platform for the development of purification processes for vaccines. “The platform uses small-scale parallel chromatography in a column format, robotic worktable, automatic buffer preparation, standardized screening methodologies, and new analytical tools.” The best resins and buffer conditions are identified regarding target yield and contaminant reduction using powerful analytical technologies. The process development time for purification of an antigen produced in CHO cells could be reduced by a factor of four when compared to a traditional screening approach.

The challenge of analytics

The challenge for high-throughput analytics has been a repeated topic on the HTPD conferences. For emerging therapies, this challenge becomes even more critical and it is important to consider best approach.

Brenna Kelley-Clarke of Juno Therapeutics [11] noted that traditional cell-based infectivity assays for viral vectors are resource intensive, requiring highly specialized operators. They can take upwards of a week or more from start to finish, and the resulting data can be highly variable. To implement analytical support for lentivirus viral vector high throughput process development, they analyzed which analytical approaches were needed, and then developed higher throughput methods for such assays. One example is using physical particle titer as an alternative to infectious titer for a variety of test cases spanning both upstream and downstream unit operations.

Measuring physical particle titer can offer faster turnaround time and better precision than cell-based infectious titer approaches. However, it is poorly predictive of biologically active vector. As such physical particle titer is only good for measuring unit operation yield and establishing peak collection criteria for column chromatography steps. For situations where process variables may affect vector quality (e.g., plasmid transfection optimization), they also designed high(er) throughput versions of infectious titer methods.

Incorporating high throughput process development approach into commercial organizations

Several presenters, including those from Merck & Co, Sanofi, Takeda and AstraZeneca, shared experiences with incorporating HTPD into their organization, including supporting fast-tracked timelines, summarizing benefits, revealing implementation challenges and solutions to such challenges.

The implementation of a multisite purification platform for downstream processing requires harmonization of equipment, buffers, resins, and analytical assays. Once implemented, advantages include lower material demands, shorter duration of experiments, time and cost savings, and easy transfer of projects between sites. To ensure successful implementation, analytical work is essential to compare results between the various sites [12].

Increasing numbers of parallel programs and limited resources call for the co-localization of process development and analytics to reduce sample turnaround times and increase flexibility. Traditional approaches often feature automated tasks linked by manual manipulations. A recommendation is to link the complete workflow from downstream to analytics in a fully automated and integrated platform. A presentation from Brian Murray of Sanofi and related companies noted that a 3 d-process can generate 100 pH and conductivity data points, 1500 concentration measurements, and 200 samples for impurity analysis. Such a data challenge supports establishment of standardized, rapid, and efficient process development methods for mAbs and mAb derivatives as well as other complex modalities requiring multistep orthogonal separation chemistries [13].

Kelcy Newell of AstraZeneca [14] noted that process development often occurs across multiple development “silos” which can lead to parallel efforts to produce manufacturing processes, analytical testing protocols and formulations. However, this approach can lead to data from one silo not being readily available to, or understood by, another silo. In the example given, for reasons of cost and time, 80% of process bioreactors used to design and optimize the manufacturing process are often never harvested and purified to understand the effect of bioreactor conditions on product quality attributes. The integration of upstream and downstream HTPD scale-down models in conjunction with high-throughput process analytic technologies (HTPAT) enables reconsideration of earlier decisions and challenge of the parallel process development paradigm. Process development and analytic silos can be broken down by identifying best practices, as well as common problems that can be jointly addressed, and via standardization of hardware and related software including data acquisition and analysis. Some solutions are relatively “low tech” such as common test tube bar coding and cap color designations. One recommendation is that focusing process analytical technologies (PAT) on quality promotes interdepartmental collaboration.

Meeting conclusions

Overall the conference attendants were appreciative of the 5th international HTPD conference and are looking forward to “HTPD VI” in two years. They are optimistic about the future for high throughput process development—especially regarding expansion of applications, as well as development of better models and analytical methods.
There is a need for better programming tools for some of the equipment used in HTPD.
Use of HTPD in the biopharmaceutical industry is expanding regarding both applications areas as well as focus.
Analytical methodologies remain key to the value and accuracy of HTPD.
There is a common desire to continue to discuss and share best practices at the next HTPD meeting.