Cambridge Healthtech Institute’s Inaugural
Data Measurement, Monitoring and Modeling Allow Informed Control of Bioprocesses
January 12-13, 2017 | Hilton San Diego Bayfront | San Diego, CA
The biopharmaceutical industry is meeting increasing demands and costs for biotherapeutics through process optimization. Advanced instrumentation with sampling techniques, new sensor technologies and analyzers have emerged to monitor both upstream and
downstream processes. These analytical tools, however, result in large, complex datasets with multivariate interactions. The inherently complex nature of these datasets makes extraction of meaningful and relevant information a difficult task.
Cambridge Healthtech Institute’s Inaugural Bioprocess Analytics conference addresses statistical analysis strategies including multivariate data analysis (MVDA), quality by design (QbD), and process analytical technology (PAT), allowing for optimized
and informed control of bioprocessing.
THURSDAY, JANUARY 12
7:45 am Conference Registration and Morning Coffee
8:15 Chairperson’s Opening Remarks
Sohye Kang, Ph.D., Senior Scientist, Process Development, Amgen
8:20 Strategies to Enable High-Throughput Recombinant Protein Production in Mammalian Cells for Preclinical Studies
Athena Wong, Ph.D., Senior Scientist & Senior Group Leader, Early Stage Cell Culture, Genentech
Transient transfections in HEK293 and CHO cells are used to rapidly generate proteins for discovery research and early development studies. Here we present our approaches to express microgram to multigram amounts of protein in automated 96 deep
well plates and bioreactors. To increase transfection productivity, we performed host cell engineering followed by process optimization. Results showed that modifying media components provides significant benefits towards increasing yield
and/or modulating product quality.
9:00 Today a Caterpillar, Tomorrow a Butterfly: The Transformation of Bioprocess Analytics
Beth Junker, Ph.D., Principal Consultant, BioProcess Advantage LLC
The volume and complexity of bioprocess data are increasing dramatically, requiring a transformation of bioprocess analytics to effectively extract meaningful information. Whether in clinical development or commercial manufacturing, tomorrow’s
focus is moving away from Big Data towards Smart Data approaches. Available tools such as Multivariate data analysis as well as other statistical methods overlay the necessary structure. This permits purposeful analytics which ultimately
reveal valuable insights into process and analytical understanding.
9:30 Engineering Process Performance by Predictive Insilico Solutions
Dirk Müller, Ph.D., Head, R&D Services, Insilico Biotechnology AG
Genome-based network models have been reconstructed for a number of the biotech industry’s workhorses and find increasing application for the quantification of bioprocess performance and for the evaluation of strain engineering strategies
in silico. The biotech industry, however, is still a far cry from the situation in other areas like electronics or automotive where each new product is designed and optimized employing computer simulations, visualization, and
data management solutions. Since costs for both genomic data and computing resources are decreasing rapidly, we also expect the bioeconomy to transform into a knowledge-based industry within the next decade. This transformation will lead
to (i) significantly reduced development timelines, (ii) increased process performance, and (iii) improved quality of life-saving therapeutics. Nevertheless, there is a rather long way to go because a full mechanistic understanding of
cellular systems is presently lacking and the software tools and algorithms to swiftly do the job for every bioprocess are yet to be developed. In this contribution, we present first steps we have taken in this direction with a technology
platform comprising both content and software. The platform supports efficient setup and curation of genome based models as well as integration of metabolite data and transcript levels. Dedicated software tools enable automated quantification
of process performance and derivation of predictive dynamic models. We illustrate the application of such predictive models to bioprocess optimization including media design and strain engineering and discuss chances and current challenges
of this approach.
10:00 Coffee Break in the Exhibit Hall with Poster Viewing
11:00 Genome-Scale Big Data and Modeling Approaches to Optimizing Protein Production in CHO Cells
Bernhard Palsson, Ph.D., Galletti Professor, Bioengineering; Principal Investigator, Systems
Biology Research Group, Bioengineering; Professor, Pediatrics, University of California, San Diego
Three technological drivers that advanced the development of microbial production strains are now in place for CHO cells. These are: 1) whole-genome sequences, 2) genome editing tools, and 3) genome-scale models. The last item is reliant on
big data analysis against a structured network reconstruction for metabolism and protein secretion. Network reconstructions are knowledge bases that formalize our knowledge of biochemistry, genetics, and genomics in CHO. These are called
BiGG k-bases. The history, development and status of this field is reviewed in this talk, along with a view of what the future may hold.
11:30 Session Break
12:15 pm Luncheon Presentation (Sponsorship Opportunity Available) or Enjoy Lunch on Your Own
1:15 Ice Cream Break in the Exhibit Hall with Poster Viewing
2:30 Chairperson’s Remarks
Rainer Stahn, Ph.D., Director, Process Development, Glycotope GmbH
2:35 A Cost- and Time-Effective Approach for QbD and PAT in Upstream Bioprocess Development and Optimization: Combining Intensified Design of Experiments (iDoE) and Hybrid Modeling
Moritz von Stosch, Ph.D., Lecturer, School of Chemical Engineering and Advanced Materials,
iDoE compresses a classical DoE into a lower number of experiments by intra-experiment process condition changes. The process response is analyzed with hybrid models that combine fundamental knowledge with data-driven techniques, because their
development is cost effective. The results of different processes (microbial and mammalian) show that it is possible to reduce the number of experiments compared to classical DoE by a factor of 2-3, while increasing process understanding
beyond static endpoint correlations.
3:05 New Solutions for Production of Difficult-to-Express Proteins
Sebastian Schuck, Ph.D., Head, Business Development, Wacker Biotech GmbH
Wacker Biotech will present highly competitive solutions for production of difficult-to-express proteins based on its proprietary E. coli expression systems ESETEC® and FOLDTEC®. Recent case studies will include secretion of functional
antibody fragments and enzymes to the fermentation broth with up to 14 g/L. Together with its E. coli refolding platform FOLDTEC®, Wacker Biotech offers a novel and comprehensive approach to rapidly assess manufacturability of therapeutic
3:35 Refreshment Break in the Exhibit Hall with Poster Viewing
4:15 Selected Poster Presentation: Automated Peptide Mapping for Quantitative Comparison of Critical Quality Attributes of Biotherapeutics
Aude Tartiere, MS, Scientific Account Manager, Expressionist Business Unit, Genedata
4:45 Leveraging the MAM for Process Control with Real-Time Monitoring and Feedback to Bioreactors
Richard Rogers, Ph.D., Scientist 4, Just Biotherapeutics
We have developed and implemented a mass spectrometry-based multi-attribute method (MAM) that monitors known CQAs but also has the ability to identify new CQAs on the biotherapeutics. New peak detection is an instrumental element of the MAM
that ensures novel modifications on the biotherapeutics are not overlooked. We are able to detect unexpected events during manufacturing by comparing a test sample to a reference standard.
5:15 Scalability of Growth Characteristics and Product Quality: Efficient Downscale Perfusion Bioprocess Development Using DOE Studies
Rainer Stahn, Ph.D., Director, Process Development, Glycotope GmbH
The sedimentation-based down-scale perfusion system SAM (10mL reactor volume) has been developed to characterize the upstream process parameters and their influence on product quality. Using DoE studies we gain a highly efficient method for
media development as well as process optimization to achieve higher cell densities and higher productivities. Scalability and reproducibility of perfusions bioreactors (10mL-1000L) will be highlighted with data of the fully human, high-yield
production and glycol-optimization platform GlycoExpress (GEX).
5:45 Close of Day
FRIDAY, JANUARY 13
8:00 am Conference Registration and Morning Coffee
8:30 Chairperson’s Remarks
Barthélemy Demeule, Ph.D., Senior Scientist & Senior Group Leader, Late Stage Pharmaceutical Development, Genentech
8:35 QbD Approaches to Design Leaner Formulation Robustness Studies
Barthélemy Demeule, Ph.D., Senior Scientist & Senior Group Leader, Late Stage
Pharmaceutical Development, Genentech
QbD offers a formal framework for process and formulation design. Through case studies, we show how historical data and multivariate studies can be used to design lean, worst-case formulation robustness studies supporting formulation specifications.
9:05 Understanding Origins of Sequence Variants during Drug Development
Thomas Slaney, Ph.D., Scientist I, Molecular and Analytical Development, Global Manufacturing & Supply,
Biologics Development & Operations, Bristol-Meyers Squibb Co.
Liquid chromatography-mass spectrometry analysis is an information-rich technique for the characterization of therapeutic proteins. This method can be employed to examine the primary sequence of expressed biotherapeutics to determine its integrity.
The distribution of sequence variants observed can be utilized to discern the mechanism of their occurrence during process development.
9:35 QbD Applied to Analytical Method Optimization for Biotherapeutic Products
Alessandra Tieri, Ph.D., Researcher Scientist, Protein Chemistry, Analytical Development, Merck
The quality of a product should be measurable, thus accurate measure of CQAs is a must. Robust and capable analytical methods are a strict requirement for the delivery of a quality product to patients. A QbD approach is the key for long-lasting
analytical methods, successful troubleshooting and effective investigation closures. In the frame of the analytical development and validation, tools such as RA and DoE guarantee faster flows, robustness and reliability of the results.
In the frame of the methods LCM, while the RA and RCA bring to light the method criticalities and weaknesses, the DoE drive to their right adjustments.
10:05 Coffee Break with a Poster Pavilion
11:00 Sequential Injection Capillary Electrophoresis for Bioprocess Monitoring
Rosanne Guijt, Ph.D., Alexander von Humboldt Fellow and Senior Lecturer, Australian Centre for Research
on Separation Science (ACROSS), University of Tasmania
Biological processes are naturally susceptible to variability because living cells consume substrates and produce metabolites and products in a dynamic way with variations in metabolic rate across short time intervals. This presentation explores
the potential of capillary electrophoresis (CE) for bioprocess monitoring. Using a novel injection strategy, this fully automated system offers high sample throughput, good temporal resolution and low sample consumption combined with robustness,
sensitivity and flexibility which provides a promising new platform for pharmacological and biotechnological studies.
View Speaker Interview
11:30 High-Throughput Protein Analysis and Engineering Using Microcapillary Arrays
Spencer Alford, Ph.D., Protein Engineer, xCella Biosciences, Inc.
We developed a high-throughput screening platform that allows researchers to assay the functional activity of millions of protein variants, displayed on or secreted from cells. This talk describes several protein analysis and engineering applications
performed with this new technology platform.
12:00 pm IT’S A WRAP: PEPTALK 2017 CLOSING PLENARY PANEL DISCUSSION
1:15 Close of Conference