Statistical Issues in Estimating, Tracking, and Comparing Complex Nonlinear Parameters Determined When Developing and Manufacturing Biologics

• Activity Number: 293
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 8:30 AM to 10:20 AM
• Sponsor: Biopharmaceutical Section
• Author(s): Patrick J. Gaffney*+ and Anthony Lonardo
• Companies: ImClone Systems and ImClone Systems
• Keywords: biologics ; Bayesian ; cell growth

Abstract

Measurement is a key concern in development and manufacture of biologics. It is important in establishing the design space and in subsequent control in the manufacturing process. This paper looks at measurement in the final bioreactor where the antibody is primarily produced. Traditionally the nutrients (e.g. glucose and glutamine), by-products (e.g. lactate and ammonia) and the bioreactors conditions (e.g. pH, dissolved oxygen and carbon dioxide) are tracked. This paper examines measurement of the cell culture growth, which ultimately captures how the cell culture responds. While the cell growth rate in earlier bioreactors is, for the most part, exponential, the growth pattern of the cell culture in this bioreactor is more complex and exhibits distinct phases of growth and death. Two methodologies are examined - one based on the Bayesian paradigm and the other using a frequentist framework. This paper will explore the unique statistical issues associated with precise estimation and control.