2021 EUR Bioassay Conference
14th Annual | Virtual | September 20-23, 2021
2021 EUB Speaker Abstracts
Emma Bowen, Labcorp
Title: Case Study on the Challenges of Developing a Cell-Based Potency Assay with a Drug Product Containing Adjuvant
Abstract: Measuring the potency of a drug product is dependent on the ability to measure the mechanism of action in vitro. When an antigen-specific immunotherapy is formulated with the Alhydrogel® adjuvant, an aluminium hydroxide wet gel suspension, measuring cellular function is a challenge. Using DOE approaches, we have developed a two-cell line based, ELISA endpoint assay which overcomes the presence of Alhydrogel. Here we will outline the steps taken during development, robustness and validation of a method to measure the potency of this drug product in a very challenging formulation.
Anke Breckner, AbbVie
Title: Automation of Cell-Based Potency Assays – Not Only Useful in Times of Pandemic
Abstract: Alongside rising amounts of samples, the imposed remote work during the COVID-19 pandemic, significantly increased the demand for automation. As a milestone towards the full automation of our bioassays, we established a highly flexible, efficient and reliable semi-automated reporter gene assay method to determine the biological activity of drug substance and drug product samples under GMP conditions. Although initial automated method development and GMP validation required time investment and innumerous challenges had to be mastered, our efforts paid off. Using a Hamilton liquid handling system, we were able to increase sample throughput for potency assays by not only reducing the analysts hands-on time significantly in comparison to manual protocols but also minimizing assay failure rates.
Contributing Author: Anke Breckner
Bioassay Sciences and Development, Analytical Research and Development, AbbVie Deutschland GmbH & Co. KG, Ludwigshafen, Germany
Stanley N. Deming, Statistical Designs
Title: Bioassays from the Point of View of an Analytical Chemist
Abstract: The four-parameter logistic (4PL) model, commonly used to describe analytic dilution bioassay behavior, is often justified because it is a close approximation to the intractable sigmoid integral of a Gaussian distribution describing the tolerance of an organism to a stimulus. The 4PL can also be derived from simple chemical equilibrium related to the Hill equation. Because of proportionally heteroscedastic noise, it is difficult to differentiate between these two models. As an analytical chemist, I take the simpler view that a well-developed analytic dilution bioassay is a titration in which the assayed substance is the limiting reagent before the equivalence point and the association constant is large. It is for this reason that slope-ratio analysis, an alternative to 4PL analysis, works well when assessing relative potencies. An advantage of the 4PL model is that it does not require a subjective decisions as to where the equivalence point has occurred. Whichever model is used, the statistical analysis becomes easier if (as stated above) the assay is sufficiently well developed that it behaves as a titration in which the assayed substance is the limiting reagent and the association constant is large.
Denny Dimitrov, GlaxoSmithKline
Title: 384-Well Plate Automation of Bioassays
Abstract: Cell-based potency assays, also called bioassays, are required for a successful Biologic License Application (BLA) approval. They should reflect the mode of action of the tested compound, should be reproducible, accurate, robust, stability indicating and practical. However, developing and performing bioassays is technically challenging. Bioassays are very low throughput and are difficult to transfer. To overcome these challenges, we developed an automated 384-well plate bioassay method, using an antibody as a test compound. The assay was developed on Hamilton liquid handling robot and a modified version of the assay was performed on HP D300e digital dispenser instrument. Automation in a 384-well plate format increased the assay’s throughput by six times, decreased the hands-on time performing the assay and decreased the assay’s failure rate as compared to the manually performed 96-well plate assay. Three two-sample TOST tests were performed to confirm equivalence of the bioassays ran on the two instruments to the manually performed 96-well plate bioassay. The criteria were met, the bioassay results for all three methods were deemed equivalent and the automated bioassay could be used for GMP sample testing.
Brooke Franklin, Quantics Biostatistics
Title: How Can We Use Simulation to Reduce Lab Work?
Abstract: During development, scientists often repeat or rerun assay work, especially with cell-based potency assays with inherent variability. These repeats take time, resources and the cost can add up.Quantics encourage our clients to consider simulation to drastically reduce repetition. Simulating assay conditions can be done in either development or routine work and can help to define optimal doses, predict reference standard degradation, and more.Simulation is invaluable in drug development. Join us to discuss what it is, how it can be used and how it can save you time, money and a lot of extra, unnecessary laboratory work.
Kate Getliffe, PsiOxus Therapeutics Ltd.
Title: Optimisation of Cell Handling Methods for the Development of Robust Bioassays – Part II: Cell Banking
Abstract: PsiOxus Therapeutics is an early phase tumour re-engineering company with several ongoing clinical trials involving our proprietary Tumour Specific Immune Gene (TSIGn) virus platform and its parent oncolytic virus, enadenotucirev. Our platform characterisation, release and stability tests include two main cell-based bioassays: a virus infectivity assay and an oncolytic potency assay. This presentation will describe an approach for improving the robustness of these bioassays by characterising the colorectal adenocarcinoma cell line (HT29 ATCC HTB-38) used to run them and optimising specific cell handling methods. The cell characterisation has been in terms of cell confluency (measured digitally), cell cycle status, metabolite concentrations in culture media and cell viability in monolayers. The handling methods have included cell counting, culture media supplementation, cell dissociation from flasks, cryopreservation and cell thaw, all investigated using DoE.In terms of cell handling methods, critical factors for cell counting, culture media supplementation, cell dissociation, cryopreservation and cell thaw were identified in DoE screening experiments. During the cryopreservation screening DoE, the most critical factors were identified to be DMSO concentration in freezing media and freezing cell density. These factors were then optimised using a response surface DoE design and the optimal conditions for cryopreservation of HT-29 cells were identified to be 13% DMSO and 1.4 x107 cells/mL. During the cell thaw screening DoE, the most critical factors were identified to be flask size and number of passages before use in assays, indicating the importance of recovery regimes on the performance of the cells. The optimal regime for recovering HT-29 cells after thaw, including flask size, passage timing and expansion seeding density, was determined using a non-DoE investigative approach.After completion of these studies, a master cell bank (MCB) and a working cell bank (WCB) were prepared using the optimal conditions determined. During a pre-qualification study of the oncolytic potency assay, performed using the WCB and before any optimisation studies had been performed on the assay itself, the %CV of the reportable result (relative potency) was determined to be 5.8%, indicating very good precision for a cell viability assay. The pre-qualification involved a total of 18 runs performed on 6 different days by 5 different analysts and using three different platform viruses.
Author Info: Dr Kate Getliffe
PsiOxus Therapeutics Limited
4-10, The Quadrant, Abingdon Science Park,
Barton Lane, Abingdon,
Oxfordshire, OX14 3YS
Email: Kate.Getliffe@PsiOxus.com
Sanne Gijsbers, Byondis B.V.
Title: Early Development of Potency Assays: How Do Concentration-Response Curve Parameters Relate to Method Performance Characteristics?
Abstract: For the development of a potency assay we strive for the most “optimal” concentration-response curve that, by inference, will lead to accurate and precise relative potency data. Conventionally, we target 4PL curve parameters to be “wide” (dose response range, while respecting equipment limitations), “not too steep or shallow” (slope or b-parameter), and to have “a robust location on the x-axis” (EC50 estimate). But to what extent do these targets really impact on accuracy and precision?To assess this impact, we generated different concentration-response curves (e.g., with lower-, intermediate, and higher response ranges), and conducted performance qualifications focusing on accuracy and precision.This allowed us to re-evaluate the importance of conventionally sought-after curve parameter targets in view of method performance. In this context and the anticipated release of ICH Q14, we want to readdress the objective for the use of system suitability limits, and the extent of interchangeability with assay acceptance criteria.
MaryRose Kammer, Organogenesis
Title: A Point Estimate of Relative Potency of a Complex Biologic from a 3-Point Dose-Response Curve
Abstract: Amniotic suspension allograft (ASA) is a biologic comprised of a mixture of cells derived from amniotic fluid and milled amniotic membrane, derived from the same donor, stored frozen at -80°C in a cryopreservation solution to support clinical development for the treatment of knee osteoarthritis.A cell-based assay has been developed to measure the potency of ASA relative to a positive control based on downregulation of inflammation at three doses. This talk will focus on the development and selection of this partial dose-response curve to calculate the point estimate of relative potency in the cell-based potency bioassay, as well as additional development work evaluating an expanded dose curve within the assay.
Contributing Authors: MaryRose Kammer1, Miranda Burnette1, and Katie Mowry1
1 Organogenesis, Birmingham, AL, USA
David Lansky, Precision Bioassay Inc.
Title: When You Can’t Show Similarity, Make Lemonade?
Abstract: Potency is defined only between samples that can reasonably be assumed similar. We will examine and compare several require alternative measures of (relative) activity to understand why and how they can be used for samples that are not similar. Some of these may also be useful for samples when assay data do not support the assumption of similarity. For context, we will also discuss some of the biological and statistical risks of operating without clear evidence of similarity.
Laureen Little, BEBPA
Title: Introduction to the Problem: Should we be Requiring 4PL or Linear Dose Response Curves?
Abstract: (Pending)
Lorna McInroy, Public Health England
Title: Developing and Qualifying a SARS-CoV-2 Neutralisation Assay in a Pandemic, Challenges and Risks
Abstract: In December 2019, several pneumonia cases of unknown cause emerged in Wuhan, Hubei, China. The causative agent of a novel coronavirus disease (COVID-19) was discovered and named SARS-CoV-2. There was a need for a functional assay to assess specific viral neutralisation activity. Public Health England (PHE), Porton Down responded by developing and qualifying a viral microneutralisation assay. This assay can be used to test the functional neutralisation ability of sera from convalescent or vaccinated individuals against SARS-CoV-2. An overview of the challenges and risks the team faced whilst developing an assay in a pandemic will presented. Specifically, the approach taken to define assay acceptance criteria will be discussed.
Nancy Niemuth, Battelle
Title: Establishing Cross-species Parallelism in an Antibody Assay with a Partial Curve
Abstract: The Filovirus Animal Nonclinical Group (FANG) developed an anti-EBOV glycoprotein (GP) immunoglobulin G (IgG) ELISA to measure vaccine-induced ebolavirus antibodies in serum samples. The assay produces a partial curve for both reference and test samples. Despite this drawback, the 4PL model fits well in the range of the concentrations tested and can be used to measure antibody levels in test samples. The assay was qualified and validated for both human and nonhuman primate samples. For direct immunobridging comparability between humans and NHPs, cross-species parallelism was evaluated. Based on these assessments, the anti-EBOV GP IgG ELISA method is considered suitable for the intended use of testing with both human and NHP serum samples in the same assay for immunobridging purposes.
Ásdís Pétursdóttir, Novo Nordisk
Title: End-to-End Robot System for Cell-Based Potency Assays
Abstract: This presentation will discuss our end-to-end robot system for running cell-based potency assays, its benefits, challenges and results. The current system is built around a Spinnaker robot arm and consist of an incubator, liquid handler, plate reader and an ELISA washer. The robot system pushes the boundaries of what can be done manually in the lab, where more samples can be handled at the same time and analyses can be run over night. This provides unique possibilities for development as well as increasing the throughput of analyses giving more time for non-routine projects. It is a flexible system where parts of the analysis can be run manually, and parts automated. There are certain challenges one has to face when qualifying a complicated automated system for GMP use. Especially when it is built from equipment from more than one manufacturer as this can give challenges for communications within the system. Also, the automated cell-based assays have to be adapted to the automated system. An example of the adaptation of such a method will be given and results discussed.
Mike Sadick, Precision Biosciences
Title: Resetting Expectations for CAR-T Potency Assessment – Baby, You Can Drive my CAR
Abstract: (Pending)
Johannes Solzin, Boehringer Ingelheim Pharma GmbH & Co KG
Title: Infectivity Assays for Virus-Based Biologics
Abstract: A robust and precise infectivity assay iQs a prerequisite for the development and market supply of virus-based biologics such as AAV-based gene therapy or oncolytic viruses. Besides that, such cell-based assays suffer a relatively high variability common to all biological assay systems and in addition are prone to operator induced variability and they required extensive hands-on time. Therefore, a faster and more robust method to measure infectivity would be desirable. Here we present an overview about the currently available infectivity assays discussing their pros and cons. Furthermore, we present our recently developed TCID50-assay with an automated, algorithm-based readout that determines label-free the cell confluence to discriminate between cytopathic (CPE)-positive and -negative wells. To increase sample throughput the assay is performed with semi-automated bench-top pipetting robots. The automated, label-free readout as well as the robotization results in an infectivity assay that is faster, more robust, and more precise compared to classical TCID50-assays.
Contributing Author: Johannes Solzin, Analytical Development Biologics, Boehringer Ingelheim Pharma GmbH & Co KG
Jaana Vesterinen, Finnish Medicines Agency
Title: Regulator’s View on the New Types of Bioassays and Assay Controls
Abstract: An increasing number of publicly available international and pharmacopoeial reference standards for bioassays have been established, including the standards for multiple therapeutic antibodies. From a regulatory viewpoint this is an important development in the field, and the value of using them is being recognized in an increasingly evolving multi-product market.
At the same time, there is an increasing number of advanced therapy medicinal products (ATMPs) in clinical trials and a handful of gene and cell therapy products have reached authorization. Although each ATMP product is unique, subgroups of these new product modalities share certain characteristics relevant for their biological activity. Thus, the methods used for measuring their biological activity and purity may be very similar.
From a regulator’s viewpoint it is of interest to consider the commonalities between the new product modalities and to explore the possibility to standardize certain assays and assay standards. For example, the copy number and infectivity rate are important indicators of the biological activity of gene therapy products.
The bioassays for novel product modalities and common factors of the bioassays will be discussed in the light of examples. Further, the quality surveillance testing by regulatory laboratories and the need for public reference standards will be discussed.
Steven Walfish, USP
Title: Statistical Model Selection Using USP <1034>
Abstract: USP General Chapter <1034> includes statistics concepts and methods of analysis for the calculation of potency and confidence intervals for a variety of relative potency bioassays. Relative potency is a measure obtained from the comparison of a test sample to a standard based of the capacity to produce the expected biological activity. USP <1034> presents several different models and suitability criteria to determine the reliability of the estimate. This talk will cover traditional linear and non-linear bioassay models with an emphasis on model suitability including parallelism. This talk will focus on the quantitative analysis, thought qualitative models are presented in the chapter. A case study comparing a parallel line assay to a slope ratio assay is presented to highlight the differences.
Sébastien Wieckowski, FHNW - University of Applied Sciences and Arts Northwestern Switzerland
Title: A Better Understanding of Bioassays – Comparison of Binding Assays, Reporter Gene Assays and Primary Cell-Based Assays of Glycoengineered Monoclonal Antibodies
Abstract: Robust bioassays are very often limited to reporter gene assays and the use of immortalized cell lines that are supposed to mimic immune cells such as natural killer (NK) cells to the detriment of primary materials, which are appreciated for their biological validity but are also difficult to exploit due to the great diversity between individuals. However, data from the literature demonstrating a biological benefit in addition to the QC-suitability to use in vitro systems such as surface plasmon resonance or bioassays based on cell lines engineered to express FcγRs rather than ex vivo systems based on primary cells are lacking. The influence of IgG1 Fc galactosylation and sialylation on ADCC activity in vitro and binding to FcγRs has been investigated using different glyco-variants produced from anti-HER2 trastuzumab monoclonal antibodies by in vitro glycoengineering, as published previously. In the current study, we measured the ADCC activity induced by variants of trastuzumab, namely samples generated to obtain deglycosylated (deglyc), degalactosylated (G0), galactosylated (G2), as well as galactosylated followed by α2,3- (ST3) or α2,6-sialylation (ST6), in the CD56+ population of cytotoxic cells isolated ex vivo from the blood of 22 healthy donors, in the presence of HER2-positive BT474 target cells. Although the effector functions of primary CD56+ cells reflected the receptor binding properties and cytolytic activity in NK92 cells in vitro, for the least affine G0 and ST3 glycoengineered variants of trastuzumab, the functional avidity plateau reached using unmodified trastuzumab could not be overcome by the most affine G2 and ST6 variants. These new data suggest that the use of primary cells could compromise the detection of subtle affinity differences between antibody variants, such as glycoengineered variants of trastuzumab, due to inherent biological limits that cannot not be exceeded in such ex vivo systems and do not allow to have an optimal amplitude margin in the measurement of improved functional avidities unlike with NK92 cells. As a result, this observed biological saturation might drastically limit the use in a quality control environment to the benefit of other systems such as cell-line based and binding assays.