189 Application of Multivariate Analysis in Supporting QbD

Tuesday, October 30, 2012: 8:30 AM
Allegheny II (Westin )
The application of multivariate latent variable methods (PCA, PLS, LPLS, JYPLS, OPLS ..) has dramatically increased in the pharmaceutical sector after the introduction of concepts like Process Analytical Technology and Quality by Design . Early applications of these techniques were heavily concentrated on pure Chemometrics for the development of soft sensors or monitoring. Applications have since evolved from pure Chemometrics isolated from process engineering, to an integral usage of the data from analytical instruments combined with data and models from the process engineering perspective to derive fundamental process understanding from a system. This session seeks submissions where multivariate methods where used to understand analytical information (NIR, UV, HPLC, NMR, FBRM,…etc) in combination with process engineering principles to support or derive process understanding, ultimately seeking for a valid design space. Examples of this work include (but not limited to): Reaction kinetic studies supported by spectra; drug product and process performance and using intermediate complex analytical data; and scale-up.

Topical I: Comprehensive Quality by Design in Pharmaceutical Development and Manufacture
Applied Mathematics and Numerical Analysis (10D)

Koji Muteki
Email: koji.muteki@pfizer.com

Huiquan Wu
Email: huiquan.wu@fda.hhs.gov

8:50 AM

9:30 AM
(189d) Implementing a Recipe Based Framework to Support Product Lifecycle Management
Adam Fermier, John Cunningham, Thomas Difeo, John Dingerdissen, James Kenyon, Steve Mehrman, Terry Murphy, Eugene Schaefer, Tom Schultz, Brian Sherry, James Weber and Paul McKenzie

9:50 AM
(189e) Variable Selection in Multivariate Modeling of Drug Product Formula and Process
Yong Cui, Xiling Song, King Chuang, Cadapakam Venkatramani, Sueanne Lee, Gregory Gallegos, Thirunellai Venkateshwaran and Minli Xie

10:10 AM
(189f) A Novel QbD Tool for Statistical Data Analysis
Linas Mockus, José Miguel Laínez and G. V. Reklaitis

10:30 AM
(189g) Mixture Component Prediction Using Iterative Optimization Technology
Koji Muteki, Daniel O. Blackwood, Kyle R. Leeman, George L. Reid, Yong Zhou, Yang A. Liu, Brent Maranzano, Paul R. Gerst, Howard W. Ward, Andreas AM. Mühlenfeld and Matthias Danner