IBC Life Sciences
The leading provider of scientific, technological and business information
Main menu
My IBC
Biological Assay Development & Validation
CSS
Event Information
Alternate Language Options:
May 12 - 14, 2010 · Sheraton Fisherman's Wharf · San Francisco, CA
Document Title
Agenda
Workshop: Wednesday | Main Conference: Thursday | Main Conference: Friday
Practical Statistical Tools for the Bioassay Scientist
These practical tutorials are designed to give you the statistical tools today to do a better and faster job at developing and maintaining your bioassays. Our hand-picked statisticians are well known for their ability to explain complex statistical concepts to a non-statistical audience. They promise to focus on introducing you to and explaining tools you can go home and start using.
8:30
Registration and Coffee
9:00
Tutorial #1 Begins: Outlier Management in Bioassay
This tutorial will cover the statistical and practical issues involved in outlier detection and management for biological assays. The methods presented will include graphical techniques and statistical calculations. There will be substantial emphasis on model-based outlier methods and why they are, in nearly all cases, more reliable than methods based on computing a mean and CV for a small number of replicates. For most of the methods discussed, there will be some theoretical justification, one or more real data examples, and the computational details of the calculations. Topics will include:
This tutorial will cover the statistical and practical issues involved in outlier detection and management for biological assays. The methods presented will include graphical techniques and statistical calculations. There will be substantial emphasis on model-based outlier methods and why they are, in nearly all cases, more reliable than methods based on computing a mean and CV for a small number of replicates. For most of the methods discussed, there will be some theoretical justification, one or more real data examples, and the computational details of the calculations. Topics will include:
- Key assumptions for common and recommended outlier methods
- Graphical and statistical tools for assessing assumptions
- How to choose a transformation so the rescaled data will satisfy assumptions
- Guidance on when it is reasonable to delete one of three observations
- For model-based methods for outlier detection:
- Why a model without an assumed functional form is desirable
- How modeling uses other doses and samples allowing fewer replicates
- Techniques to address masking and swamping
- When to use statistical methods for outlier detection vs. outlier deletion
- Using outlier detection before combining potency estimates
10:30
Networking Refreshment Break and Exhibit/Poster Viewing
11:00
Tutorial Part 1 Continues
12:00
Lunch and Exhibit/Poster Viewing
1:30
Tutorial #2 Begins: Practical Statistical Tools: LOD, MCDA, LOQ and SPC
The statistical (and other) literature is a confusing swamp when it comes to implementing measures for the fundamental concepts of "limit of detection" (LOD), "minimum consistently detectable amount" (MCDA), and "limit of quantification" (LOQ) -- it would seem that every organization takes it up itself to (re)define the meaning and implementation of these concepts. This tutorial will offer a clear, simple, statistically based framework for defining these concepts and obtaining numerical values for their implementation. Statistical process control (SPC) will be introduced as a means of monitoring these and other characteristics of bioassays, with particular emphasis on how SPC charts should not be used. Topics to be covered include:
The statistical (and other) literature is a confusing swamp when it comes to implementing measures for the fundamental concepts of "limit of detection" (LOD), "minimum consistently detectable amount" (MCDA), and "limit of quantification" (LOQ) -- it would seem that every organization takes it up itself to (re)define the meaning and implementation of these concepts. This tutorial will offer a clear, simple, statistically based framework for defining these concepts and obtaining numerical values for their implementation. Statistical process control (SPC) will be introduced as a means of monitoring these and other characteristics of bioassays, with particular emphasis on how SPC charts should not be used. Topics to be covered include:
- Sensitivity and its IUPAC definition
- Minimum consistently detectable amount and the false negative risk
- Selectivity and the selectivity ratio
- Limit of quantification and the assayist's ego
- Limit of detection and the false positive risk
- Statistical process control and continual validation
3:00
Networking Refreshment Break and Exhibit/Poster Viewing
3:30
Tutorial Part 2 Continues
5:00
Networking and Exhibit/Poster Viewing
Workshop: Wednesday | Main Conference: Thursday | Main Conference: Friday
7:30
Registration and Coffee
8:00
Chairperson's Opening Remarks
Laureen E. Little, Ph.D., Principal Consultant, Quality Services; Editor/Publisher, BioQuality
Laureen E. Little, Ph.D., Principal Consultant, Quality Services; Editor/Publisher, BioQuality
Keynote Address
8:10
Bioassays for Pharmaceutical Testing: Benefits and Limitations
This talk introduces the basics of different types of bioassays used for biopharmaceutical testing. Focusing primarily on cell-based assays, the impact /importance of choosing an adequate assay design, relevant validation parameters as well as suitable controls for the assay outcome are discussed. Practical approaches, combining scientific and statistical expertise, are presented to consider the requirements for valid analysis and interpretation of bioassay data under cGMP conditions.
Thomas Waerner, Ph.D., Head of Cell & Molecular Biology, Quality & Compliance, Boehringer Ingelheim RCV GmbH & Co KG, Germany
This talk introduces the basics of different types of bioassays used for biopharmaceutical testing. Focusing primarily on cell-based assays, the impact /importance of choosing an adequate assay design, relevant validation parameters as well as suitable controls for the assay outcome are discussed. Practical approaches, combining scientific and statistical expertise, are presented to consider the requirements for valid analysis and interpretation of bioassay data under cGMP conditions.
Thomas Waerner, Ph.D., Head of Cell & Molecular Biology, Quality & Compliance, Boehringer Ingelheim RCV GmbH & Co KG, Germany
Early Assay Development
8:50
Challenges in Using Bioassays for Product Characterization
Bioassays are widely used for product characterization of protein therapeutics. With relatively high assay variability and oftentimes lack of similarity/parallelism between the tested materials and reference standards, challenges in choosing appropriate assays and how to calculate and report a meaningful relative potency for product variants and stressed products will be discussed.
Mary Hu, Ph.D., Director, Seattle Genetics, Inc.
Bioassays are widely used for product characterization of protein therapeutics. With relatively high assay variability and oftentimes lack of similarity/parallelism between the tested materials and reference standards, challenges in choosing appropriate assays and how to calculate and report a meaningful relative potency for product variants and stressed products will be discussed.
Mary Hu, Ph.D., Director, Seattle Genetics, Inc.
9:30
How Miss-specified Models Affect Bioassay Performance
Analysis of bioassay data typically assumes normality, constant variance, independence, and a logistic response curve. A well-chosen transformation of the response will usually make bioassay data nearly normal, with nearly constant variance, and near logistic in shape; hence, it is the assumption of independence that is, in practice, the most challenging to address. This talk will illustrate with specific examples how common mismatches between the bioassay data and the model cause problems in the performance of bioassays. The mismatches will include misspecification of the shape of the concentration response curve, failure to address non-constant variance, failure to address non-normality, failure to detect outliers, failure to detect non-similarity and failure to address common departures from independence. Using representative simulated data from nonlinear, parallel straight line, and slope ratio assays the relative impact of all these will be quantified via their impact on assay acceptance, sample acceptance (similarity), and crucial properties of potency: bias and precision. Departures from independence (a failure to model design structure in the assay) will be studied in some detail illustrating how common practices such as use of multi-channel pipettes, serial dilution, blocking, and pseudoreplication affect assay performance. Finally, some relatively simple strategies for addressing the most important parts of assay design structure will be presented.
David Lansky, Ph.D., President, Precision Bioassay, Inc.
Analysis of bioassay data typically assumes normality, constant variance, independence, and a logistic response curve. A well-chosen transformation of the response will usually make bioassay data nearly normal, with nearly constant variance, and near logistic in shape; hence, it is the assumption of independence that is, in practice, the most challenging to address. This talk will illustrate with specific examples how common mismatches between the bioassay data and the model cause problems in the performance of bioassays. The mismatches will include misspecification of the shape of the concentration response curve, failure to address non-constant variance, failure to address non-normality, failure to detect outliers, failure to detect non-similarity and failure to address common departures from independence. Using representative simulated data from nonlinear, parallel straight line, and slope ratio assays the relative impact of all these will be quantified via their impact on assay acceptance, sample acceptance (similarity), and crucial properties of potency: bias and precision. Departures from independence (a failure to model design structure in the assay) will be studied in some detail illustrating how common practices such as use of multi-channel pipettes, serial dilution, blocking, and pseudoreplication affect assay performance. Finally, some relatively simple strategies for addressing the most important parts of assay design structure will be presented.
David Lansky, Ph.D., President, Precision Bioassay, Inc.
10:10
Networking Refreshment Break and Exhibit/Poster Viewing
10:40
Case
Study
Objective Selection of Appropriate Non-Linear Curve, Weighting Scheme, and Curve Similarity Measure for BioassaysStudy
Modern day bioassay data are typically analyzed using a full dose-response non-linear curve with a possible mathematical weighting option. Reference standard and test sample curves are then compared for their similarity before calculating sample's relative potency value. These options are usually determined during the assay development phase. Historically, this process has been highly subjective and inconsistent due to the lack of industry best practice and regulatory guidance. In this work, we propose a unified approach that provides an objective and consistent way for selecting and defending the options for nonlinear curve, weighting, and curve similarity measure for bioassays. Application to two sets of assay development data will be illustrated.
Frank Ye, Ph.D., Director, Quality Engineering, Amgen Inc.
Fuat Doymaz, Ph.D., Senior Quality Engineer, Amgen Inc.
Pharmacoepial and Regulatory Guidance for Bioassays
11:20
Regulatory Update of Biological Assays for Vaccine Potency: A CBER Perspective
Juan L. Arciniega, D.Sc., LRSP, CBER, FDA
Juan L. Arciniega, D.Sc., LRSP, CBER, FDA
Technology Workshop
11:50
Speaker TBA
12:20
Luncheon and Exhibit/Poster Viewing
1:40
Update on USP Bioassay Chapters
This presentation will provide an update on chapter publication status and key ideas proposed in the new USP General Chapters pertaining to the bioassay (<1032> Design of Bioassays; <1033> Bioassay Validation; <1034> Analysis of Bioassays). The evolution of the chapters in the context of the regulatory environment will be discussed. Opportunities for submitting comment to the USP will be provided.
Robert Singer, M.S., Consultant, Biometry Associates LLC
This presentation will provide an update on chapter publication status and key ideas proposed in the new USP General Chapters pertaining to the bioassay (<1032> Design of Bioassays; <1033> Bioassay Validation; <1034> Analysis of Bioassays). The evolution of the chapters in the context of the regulatory environment will be discussed. Opportunities for submitting comment to the USP will be provided.
Robert Singer, M.S., Consultant, Biometry Associates LLC
Bioassays to Support Product Release - Monoclonal Products
2:15
Case
Study
Development of a MSD-based ADCC MethodStudy
One of the challenges in bioassay development is to produce data of consistently high quality in the face of increasingly more complex and time-consuming cell-based assays. To address these challenges, emerging analytical technologies such as MSD offer increased speed, sensitivity, dynamic range and ease of use, as well as adaptability to newer cell-based assay formats. This presentation will describe the development and optimization of a new MSD-based method for ADCC and its correlation with the traditional PBMC-based method.
Max L. Tejada, Ph.D., Scientist, Genentech, Inc.
2:50
Case
Study
Development and Validation of an ADCC-Reporter Gene AssayStudy
Antibody-dependent cell cytotoxicity (ADCC) is an important biological function attributed to the mechanism of action of several therapeutic antibodies. The classic ADCC assay is a not only complicated but also a highly variable assay to be implemented as a control assay in Quality Control labs. Hear a detailed description of the development and validation of an alternate approach, an ADCC-Reporter gene assay that is based on the key attributes of the classic ADCC. The ADCC-Reporter gene assay has performance characteristics (accuracy, precision as well as robustness) to be used not only as a potency assay for lot release and stability testing but also as a key assay for the characterization of therapeutic IgG1 molecules. Additionally, the general applicability of the approach for all IgG1s will be discussed.
Bhavin S. Parekh, Ph.D., Group Leader, Eli Lilly and Company
3:25
Networking Refreshment Break and Last Chance forExhibit/Poster Viewing
Bioassays to Support Product Release - Host Cell Protein Assays
3:55
Case
Study
Atypical Bioassays to Monitor Residual ImpuritiesStudy
Residual impurities, including Host Cell Proteins, can be difficult to detect in biotherapeutics with conventional analytical technology. This is especially the case when present at very low levels. Method development and validation is described for such a case, where specific cell lines and in vitro human tissue isolates were utilized to monitor a trace impurity. A combination of cell-based and immunoassay technologies were applied, as was a carefully designed control/release strategy, in order to develop an improved process and enable release of batches of drug.
Ned Mozier, Ph.D., Director, Pfizer Inc.
4:30
Case
Study
Multi-Product Host Cell Protein Impurity Assays Provide Superior Means Of Monitoring Manufacturing ConsistencyStudy
Multiproduct-ELISAs for host cell proteins (HCPs) are superior to product-specific ELISAs for quantifying HCP impurities and are standard for process validation studies and in-process tests to demonstrate manufacturing consistency. Two dimensional gel electrophoresis (2DE) of HCPs showed minimal variation in protein pattern between cultures grown under different conditions and at different scales, providing a strong scientific rationale for the multiproduct application of a standard assay. Immunoaffinity fractionation of HCPs showed 25% of the total proteins form an immunodominant sub-population, which is representative of the total population of HCPs.
Martin Vanderlaan, Ph.D., M.B.A., Director of Analytical Operations, Genentech, Inc.
5:05
Case
Study
Host Cell Protein Analysis: Measuring the Forest or Counting the TreesStudy
The challenges of testing for Host Cell Proteins are well known. The accuracy and linearity, as well as antibody coverage and standardization, can complicate assay development. A scaled down HCP study was conducted and a mathematical model is presented which illustrates this complexity and sheds light on some of the underlying causes. When the HCP assay is considered as the sum of many individual interactions, the relevance of the assay issues is better understood. This allows for the development of more robust and dependable assays to support process development, characterization and product release.
Ted Kocot, B.Sc., Scientist, Pfizer Inc.
5:40
Close of Day 2
Workshop: Wednesday | Main Conference: Thursday | Main Conference: Friday
7:30
Coffee
Quality by Design and Use of DOE for Bioassays
8:00
Chairperson's Remarks
Sally Seaver, Ph.D., President, Seaver Associates, LLC
Sally Seaver, Ph.D., President, Seaver Associates, LLC
8:10
A QbD Concept for Test Method Development and Validation and its Application on Bioassays
The pharmaceutical industry is currently embracing Quality by Design concepts to help improve the robustness of manufacturing processes and to facilitate continuous improvement strategies to shape and enhance product quality and manufacturing productivity. Transferring these concepts to the analytical methods environment could offer opportunities e.g. for enhanced method robustness and ruggedness, facilitation of continuous improvement and technological innovation and more robust method knowledge transfer due to enhanced analytical method understanding. The talk will provide a general overview of the current discussion on QbD for analytical methods and will address how these concepts can be applied on bioassay development.
Alexander Lenhart, Ph.D., Section Head, Abbott GmbH & Co. KG, Germany
The pharmaceutical industry is currently embracing Quality by Design concepts to help improve the robustness of manufacturing processes and to facilitate continuous improvement strategies to shape and enhance product quality and manufacturing productivity. Transferring these concepts to the analytical methods environment could offer opportunities e.g. for enhanced method robustness and ruggedness, facilitation of continuous improvement and technological innovation and more robust method knowledge transfer due to enhanced analytical method understanding. The talk will provide a general overview of the current discussion on QbD for analytical methods and will address how these concepts can be applied on bioassay development.
Alexander Lenhart, Ph.D., Section Head, Abbott GmbH & Co. KG, Germany
8:50
Case
Study
DOE Application during Cell-based Potency Assay DevelopmentStudy
Response Surface Method (RSM) is a statistical technique for modeling responses via polynomial equations. The model becomes the basis for 2-D contour maps and 3-D surface plots for the purpose of optimization. Computerized optimization needs to be verified by running the actual complex experiment: cell-based assay. As a matter of fact, RSM should be applied only when right questions are asked. This cell-based potency assay development case study will demonstrate how to establish the relationship among the factors, identify the interactions between the factors, and finalize where the so-called best process settings are located. Last but not the least, the pre-qualification of this statistically modified assay eventually confirmed the required performance characteristics and showed acceptable assay accuracy, precision, recovery, linearity and ruggedness.
Liming Shi, Ph.D., Senior Scientist, Amylin Pharmaceuticals, Inc.
9:30
Networking Refreshment Break
10:00
Poster Presentations
Meeting posters will be selected for a 15 minute presentation. This session is designed to generate discussion about cutting edge and evolving projects.
Meeting posters will be selected for a 15 minute presentation. This session is designed to generate discussion about cutting edge and evolving projects.
Assay Comparability
10:40
Mini-Tutorial: Introduction to Bland-Altman Analysis of Comparative Data
D. G. Altman and J. M. Bland stated the problem well in the introduction to their 1983 paper in The Statistician: "Methods of analysis used in the comparison of two methods of measurement are reviewed. The use of correlation, regression and the difference between means is criticized." They went on to propose a simple parametric approach based on analysis of variance and simple graphical methods. This introduction will describe the original Bland-Altman method and its wide acceptance over the years since it was first published.
Stanley N. Deming, Ph.D., President, Statistical Designs
D. G. Altman and J. M. Bland stated the problem well in the introduction to their 1983 paper in The Statistician: "Methods of analysis used in the comparison of two methods of measurement are reviewed. The use of correlation, regression and the difference between means is criticized." They went on to propose a simple parametric approach based on analysis of variance and simple graphical methods. This introduction will describe the original Bland-Altman method and its wide acceptance over the years since it was first published.
Stanley N. Deming, Ph.D., President, Statistical Designs
11:20
Lunch on your own
12:50
Case
Study
Strategies for Assessing Potency Method ComparabilityStudy
The need for evaluation of comparability between substituted potency methods may occur at different times along a product development life cycle. The strategy is different depending on the development stage in the commercialization of the product. The different scenarios may include: 1) replacing a "like for like" method, such as a receptor-ligand binding assay with another type of receptor-ligand binding assay, 2) justifying the use of a non-cell based receptor-ligand binding assay over a cell-based assay at product licensure, and 3) replacing a cell-based assay with another cell-based assay following commercialization of the product. Examples of each scenario will be presented.
Jill Crouse-Zeineddini, Ph.D., Principal Scientist, Amgen Inc.
1:30
Case
Study
Characterization and Implementation of an In Vitro Potency Assay for Gardasil®Study
As an alternative to the in vivo potency assays typically used to release vaccine products, an in vitro relative potency (IVRP) assay has been developed for release of Merck's human papillomavirus (HPV) vaccine, Gardasil®. The advantages and limitations of using an in vitro assay and the strategy used to characterize and implement the assay will be presented.
Mary Retzlaff, Ph.D., Research Fellow, Merck Research Laboratory
2:10
2:40
Networking Refreshment Break
Maintaining and Transferring Assays
3:10
Case
Study
Post Approval Management of Potency BioassaysStudy
Post approval management of potency bioassay provides many challenges. Many of these challenges are similar to those in the clinical development phase. However, the emphasis becomes maintaining control of the assay over the many years of the product lifespan in a regulated commercial QC environment rather than assay development and improvement. Maintenance of the assay will require monitoring systems that are predictive and will proactively identify changes in assay performance. Managing change in cell line and culture critical reagents and in the assay personnel remain as key activities in the maintenance of bioassay post approval..
Marianne Hayes, Ph.D., Fellow, Merck Research Laboratories
3:50
Case
Study
Demonstration of Robustness for a High Throughput Toxin Neutralization Assay for Mouse SerumStudy
Jamie Bird, BSc., Supervisor-Manufacturing Assay Development, Emergent BioSolutions
4:30
Case
Study
Transfer of Potency Methods: from Challenges to Lessons LearnedStudy
The transfer of potency methods to production sites, partners, or Contract Manufacturing Organizations can be particularly challenging due to various factors such as the reliance of these methods on analyst experience, availability of critical equipment and/or reagents as well as the handling of cells. This presentation summarizes several case studies, which illustrate the challenges encountered and lessons learned from method transfer.
Carl Co, Ph.D., Associate Scientist, Genentech, Inc.
5:10
Close of Conference
Bottom menu
Copyright IBC Life Sciences, a part of Informa Life Sciences Group. | Phone: 800.390.4078 | Email: reg@ibcusa.com