Manufacturing is about “Lean” (i.e. efficient): How do we apply the same disciplines to bioassay? During development, many assays evolve in a complex way that is not necessarily “Lean”. It doesn’t have to stay that way! Once the biology is stable and well-characterized we can re-evaluate the design, before validation, to increase throughput (efficiency) whilst retaining precision (effectiveness). We’ll discuss case studies showing improved outcomes by embracing “Lean” principles.

A potency assay used for several years showed not to be sensitive to low-affinity insulin analogs as well as impacted by the Drug Product matrix. A story regarding implementation of two new potency assays as well as the learnings will be described.

This talk will address two common challenges faced by bioassay scientists. The first challenge is the selection of the proper dilution levels in relative potency assays, to observe the most precise and accurate estimates of both relative potency and similarity metrics. Common optimizers are challenging to use for nonlinear models and require the help of highly experienced statisticians. This talk will present a simple alternative that is easy to implement by non-statisticians. The second challenge addressed is the presence of outliers in bioassay data. This talk will present the different types of outliers in such data, and give pointers on how to detect, and deal with them.

Biomolecules are highly complex, physiochemical information alone is not sufficient to confirm high-order structure. Therefore, biological activity needs to be demonstrated for biomolecules. Biological activity is measured in bioassay to determine potency which is a critical attribute that measures the ability of drugs to elicit their function. Basic methodology of bioassay development and validation is the same for biologic, vaccine, and gene therapy modalities. However, the mechanism of action, variability, and challenges are drastically different. The talk will be focused on providing an overview of bioassay challenges and best practices for biologic, vaccine, and gene therapy products.

The gene therapy field is one of the fastest growing modalities in the pharmaceutical development realm. Potency assays are an essential regulatory aspect of the gene therapy field. The development and validation of these assays are paramount in licensing, but often pose time-consuming, multifaceted, and expensive challenges. This talk will explore effective optimization methodologies to overcome some of the more significant challenges that arise during the development process.

Combination therapy can significantly increase the impact of biologics as therapeutics. Patient-friendly delivery methods are essential for improved clinical outcomes. Factors like patient compliance, therapeutic efficacy, pharmacological specificity, and concordance could be addressed by co-formulation; making this approach a sought after outcome where possible.  Development of analytical methods including bioassay within CMC is crucial for delivery of co-formulation therapeutics. This presentation will explore the technical challenges and considerations for developing co-formulation biologics from a bioassay perspective. A case study will look at potential pinpoints to be examined and activities leveraged on for delivering bioassays to support a combination therapy.

Reference standards are essential to the development and control of biological products. USP recommends reporting potency of test articles relative to a reference standard, while some laboratories use these also or instead as a control. Considering their importance, however, there is no consensus on the source of a reference standard, the basis, and means of reference standard qualification and stability evaluation, or the use of a primary standard. This talk will discuss principles and practices related to reference standards used to report potency of biological products and propose strategies for their acquisition and evaluation. Those proposals will borrow from practices related to quality by design for analytical methods, highlighting fitness-for-use of a references standard as well as reduction of uncertainty and the decision risks associated with their uses during development and quality control.

• What does it mean to adopt a lifecycle approach to bioassay development and validation? 
• What matters most to regulators at Phase I, II and III when it comes to measuring potency?
• How to choose an Interim reference standard(s) and transition to a Primary reference standard and a Working standard?​

• When should bioassay scientists require help from a statistician?
• What can bioassay scientists learn to do themselves?
• When should bioassay scientists perform Design of Experiments?
• Should bioassay scientists use Statistical Process Control for their assays?

International standards have a long history supporting biological medicines. They are essential to the assignment of biological potency to many biological products and are responsible to maintain a global network of secondary standards. In the era of recombinant products, biosimilars, and other novel therapeutic modalities, the need for public standard preparations contributing to harmonization in the evaluation of biologicals is increasingly important. Recent data from international collaborative studies illustrate their role.

Small amounts of some types of non-similarity cause appreciable potency bias. Precision, sample size, and power considerations (or more simply the typical width of similarity confidence intervals) lead to a strategic approach: equivalence bounds based on bias limits with assay size based on precision and power needs for similarity. This approach has important implications for assay design and analysis. This talk will illustrate these issues with examples. 

Artificial Intelligence (AI) is increasing efficiencies in drug discovery and development by automating human tasks that are repetitive, demand accuracy, and/or require immense resources. Using some case studies, we will highlight the benefits of AI solutions, as well as discuss challenges that the industry faces in the adoption of AI solutions, such as:

• Regulatory: lack of guidance for low-risk AI tools; use of existing approval pathways for high-risk AI solutions without addressing some major issues
• Data: inaccessibility to large amounts of quality training data
• Awareness: limited understanding of AI
• Cognitive biases that shape erroneous and biased models
• Implementation challenges
  

USP <1033> discusses methods for the suitability of a bioassay for its intended use. This talk will cover the relationship between Cpm and the probability of being out of specification. Focus on the two terms related to assay in the denominator and a discussion of relative bias and intermediate precision.

• Discussion of relative bias and intermediate precision
• How does bioassay performance impact the probability of out of specification
• Is total error models better than Cpm?

Robustness variables are studied to optimize and evaluate bioassays. Experimental design is a technique to gain efficiency in this process. Response variables include system suitability parameters and relative potency (RP) measures. When assay development is complete, experimental design and RP variabilities are used to prove the assay is robust.

Development and characterization of product reference standards present a number of challenges that many fast become acquainted with and others perhaps years after commercialization. This talk will discuss briefly what a number of these challenges can be and how to first be aware and then possibly avoid them. A case study will be presented with actual data (masked/confidential) to demonstrate how to bridge to a new reference standard.

• Representativeness in the context of Reference Standard (connection to clinical material)?
• Reference Standard selection and bridging
  
• Reference Standard Lifecycle