Multivariate analysis in the pharmaceutical industry / edited by Ana Patricia Ferreira, Jose C. Menezes, Mike Tobyn.

Format:

Book

Contributor:

Ferreira, Ana Patricia, editor.

Menezes, Jose C., editor.

Tobyn, Mike, editor.

Language:

English

Subjects (All):

Pharmaceutical industry.

Physical Description:

1 online resource (465 pages) : illustrations

Edition:

1st ed.

Place of Publication:

London, United Kingdom : Academic Press, an imprint of Elsevier, [2018]

Summary:

Multivariate Analysis in the Pharmaceutical Industry provides industry practitioners with guidance on multivariate data methods and their applications over the lifecycle of a pharmaceutical product, from process development, to routine manufacturing, focusing on the challenges specific to each step. It includes an overview of regulatory guidance specific to the use of these methods, along with perspectives on the applications of these methods that allow for testing, monitoring and controlling products and processes. The book seeks to put multivariate analysis into a pharmaceutical context for the benefit of pharmaceutical practitioners, potential practitioners, managers and regulators.Users will find a resources that addresses an unmet need on how pharmaceutical industry professionals can extract value from data that is routinely collected on products and processes, especially as these techniques become more widely used, and ultimately, expected by regulators.- Targets pharmaceutical industry practitioners and regulatory staff by addressing industry specific challenges- Includes case studies from different pharmaceutical companies and across product lifecycle of to introduce readers to the breadth of applications- Contains information on the current regulatory framework which will shape how multivariate analysis (MVA) is used in years to come

Contents:

Front Cover

Multivariate Analysis in the Pharmaceutical Industry

Copyright Page

Dedication

Contents

List of Contributors

About the Editors

Foreword

I. Background and Methodology

1 The Preeminence of Multivariate Data Analysis as a Statistical Data Analysis Technique in Pharmaceutical R&amp

D and Manufact...

1.1 Data Size Glossary (Table 1.1)

1.2 Big Data-Overall View

1.3 Big Data-Pharmaceutical Context

1.4 Statistical Data Analysis Methods in the Pharmaceutical Industry

1.5 Development of Multivariate Data Analysis as a Data Analysis Technique within the Pharmaceutical Industry

1.6 Current Status of the Use of Multivariate Data Analysis in the Pharmaceutical Space

1.7 What MVA Can be Used For/What it Cannot be Used For

1.8 Current Limitations and Future Developments

Acknowledgments

References

2 The Philosophy and Fundamentals of Handling, Modeling, and Interpreting Large Data Sets-the Multivariate Chemometrics App...

2.1 Introduction

2.1.1 The Nature of this Chapter

2.1.2 The History of Metrics

2.2 Univariate Data and How it is Handled

2.2.1 Data Vectors and Some Definitions

2.2.2 Some Statistics on Vectors

2.2.3 Some General Thoughts about Univariate Thinking

2.3 Multivariate Data With Definitions

2.3.1 Data Matrices, Two-Way Arrays

2.3.2 Three- and More-Way Arrays

2.3.3 Multiblock Data

2.3.4 General Thoughts About Multivariate Thinking

2.4 Modeling

2.4.1 General Factor Models

2.4.2 Principal Component Analysis

2.4.3 Multivariate Curve Resolution

2.4.4 Clustering-Classification

2.4.5 Regression Models

2.4.6 Model Diagnostics

2.4.7 Some General Thoughts About Modeling

2.5 Conclusions

3 Data Processing in Multivariate Analysis of Pharmaceutical Processes

3.1 Introduction.

3.1.1 Pharmaceutical Process Data

3.1.2 The Quality-by-Design Principle

3.2 Continuous Versus Batch Processes

3.3 Data Processing

3.3.1 Sampling

3.3.2 Noise Cancellation

3.3.3 Statistical Process Control

3.3.3.1 Implementation

3.3.3.2 Examples in the Pharma Industry

3.4 Conclusions and Trends

Acronyms

4 Theory of Sampling (TOS): A Necessary and Sufficient Guarantee for Reliable Multivariate Data Analysis in Pharmaceutical ...

4.1 Introduction

4.2 Heterogeneity

4.2.1 Counteracting Heterogeneity: Composite Sampling

4.3 Heterogeneity: A Systematic Introduction for Multivariate Data Analysis

4.4 Sampling Is Always Involved in PAT and Multivariate Data Analysis

4.5 Measurement Uncertainty (MU)

4.6 The Role of Reliable Process Sampling in Multivariate Data Analysis

4.7 Sample Size, Purpose and Representativeness

4.8 Analytical Processes vs. Sampling Processes: A Monumental Difference

4.8.1 Case Illustration

4.9 TOS: The Necessary and Sufficient Framework for Practical Sampling

4.10 Process Sampling in the Pharma Industry

4.11 Variographics: A Breakthrough for Multivariate Process Monitoring

4.12 Conclusions and Further Resources

Glossary

Appendix A

A1 Pierre Gy (1924-2015): TOS's key concept of sampling errors

A2 TOS: Governing Principles (GPs) and Sampling Unit Operations (SUOs)

5 The "How" of Multivariate Analysis (MVA) in the Pharmaceutical Industry: A Holistic Approach

5.1 Background

5.2 Why Is a Holistic Approach Needed?

5.3 What Stands in the Way?

5.4 Key Enabling Tools

5.4.1 Voice of the Customer

5.4.2 Design of Experiments

5.4.3 Metadata

5.4.4 Data Handling

5.4.5 Model Diagnostics

5.4.6 FMEA and Risk Assessment

5.4.7 Process Automation

5.4.8 Visualization.

5.4.9 Calibration Strategy Space

5.4.10 Theory of Sampling

5.5 Case Study: Multivariate Calibrations for In-Process Control

5.5.1 Background

5.5.2 Before MVA

5.5.3 Calibration Strategy, v1

5.5.4 Model Deployment and Management

5.5.5 Outlier Diagnostics

5.5.6 Calibration Strategy, v.2

5.5.7 Building Sustainability

5.6 Summary

6 Quality by Design in Practice

6.1 Process Data and Its Analysis

6.1.1 Introduction

6.1.2 Traditional Approaches to Experimental Design and Data Analysis

6.1.3 Modern Approaches to Experimental Design and Data Analysis

6.2 The DoE Toolkit

6.2.1 The Right Tool for the Right Job

6.2.1.1 Factorial Designs

6.2.1.2 Optimization Designs

6.2.1.3 Mixture Designs

6.2.1.4 Other Design Types

6.3 Implementing DoE for QbD

6.3.1 Variability Starts With Raw Materials

6.3.2 Designed Experiments in Formulation

6.3.3 Designed Experiments for Calibration Model Development

6.3.4 Designed Experiments for Process Development and Understanding

6.3.5 A Practical Roadmap for Applying DoE

6.3.5.1 Take a Multivariate Mindset

6.3.5.2 Define the Objective of the Project

6.3.5.3 Use Risk Management Wisely

6.3.5.4 Design the Experimental Plan

6.3.5.5 Analyze the Data

6.3.5.6 Implementation

6.3.5.7 Improvement

6.4 Translating DoE Into Process Control: Maintaining the Design Space

6.4.1 The Relationship Between DoE and MVA Methods

6.4.2 A Short Note on Diametrically Opposed Systems

6.4.3 Implementing PAT to Maintain the Design Space

6.4.4 Bringing QbD and the Pharmaceutical Quality System Together

6.5 Modern Data Acquisition and PAT Management Systems

6.5.1 A Model of the Pharmaceutical Quality System

6.5.2 Architecture of a Modern Control System

6.5.2.1 The PAT and Automation LAN.

6.5.2.2 The Manufacturing LAN

6.5.2.3 The QbD Team Environment

6.5.3 The QbD Development and Deployment Environment

6.5.4 PQS for Continuous Manufacturing Systems

6.6 Summary and Future Perspectives

Terminology and Acronyms

II. Applications in Pharmaceutical Development and Manufacturing

7 Multivariate Analysis Supporting Pharmaceutical Research

7.1 Overview of Multivariate Analysis as a Part of Pharmaceutical Product Design

7.2 Classification and Experimental High-Throughput Screening

7.3 Exploring Complex Analytical Data

7.3.1 Imaging: Raman Spectra

7.4 Product and Process Understanding

7.5 Summary

Abbreviations

8 Multivariate Data Analysis for Enhancing Process Understanding, Monitoring, and Control-Active Pharmaceutical Ingredient ...

8.1 Introduction

8.2 Process Understanding

8.2.1 Univariate Trending

8.2.2 Multivariate Trending

8.2.2.1 Unsupervised Methods

8.2.2.2 Supervised Methods

8.2.2.2.1 Direct Approaches

8.2.2.2.2 Inverse Approaches

8.2.3 Post-Hoc Analyses for Process Improvements and Optimization

8.3 Process Control

8.3.1 Crystallization Control

8.3.2 Reaction Control

8.4 Multivariate Statistical Process Control

8.4.1 Batch-Wise Unfolding (or Batch Level)

8.4.2 Observation-Wise Unfolding (or Observation Level)

8.5 Conclusion

9 Applications of MVDA and PAT for Drug Product Development and Manufacturing

9.1 Introduction

9.2 Method Design and Development

9.2.1 Method Requirements and Performance Criteria

9.2.2 Risk Assessment

9.2.3 Calibration Design

9.3 Method Validation

9.4 Outlier Detection and System Suitability Test

9.5 Method Maintenance and Life Cycle Management

9.6 Example Data During Commercial Implementation

9.6.1 Blending Homogeneity.

9.6.2 Near-Infrared-Based Tablet Potency and Content Uniformity Measurements

9.7 Conclusions

10 Applications of Multivariate Analysis to Monitor and Predict Pharmaceutical Materials Properties

10.1 Introduction

10.2 Spray-Dried Dispersions

10.3 Case Study 1: Investigate the Impact of Spray-Dried Dispersion Particle Properties on Formulation Performance

10.3.1 Material and Methods

10.3.1.1 Particle Size Characterization by Imaging With Malvern Morphologi G3

10.3.1.2 Particle Size Characterization by Laser-Light Scattering

10.3.1.3 Density

10.3.1.4 Surface Area

10.3.1.5 Mercury Intrusion Porosimetry

10.3.1.6 Flow

10.3.1.7 Compactibility

10.3.1.8 Dissolution

10.3.1.9 Data Analysis

10.3.2 Results and Discussion

10.3.2.1 Exploring the Relationships Between Spray-Dried Dispersion Particle Properties

10.3.2.2 Impact of Particle Properties on Formulation Performance

10.3.2.2.1 Flow

10.3.2.2.2 Compactibility

10.3.2.2.3 Dissolution Rate

10.3.3 Summary

10.4 Case Study 2: Development of a Surrogate Measurement for Particle Morphology

10.4.1 Material and Methods

10.4.1.1 Light Transmission Data

10.4.1.2 Scanning Electron Microscopy

10.4.1.3 Mercury Intrusion Porosimetry

10.4.1.4 Data Processing and Analysis

10.4.2 Results and Discussion

10.4.2.1 High-Pressure Intrusion Volume

10.4.2.2 Light Transmission Data

10.4.2.3 Regression Analysis and Morphology Factor

10.4.3 Summary

10.5 Conclusions

11 Mining Information From Developmental Data: Process Understanding, Design Space Identification, and Product Transfer

11.1 Introduction

11.2 Latent-Variable Modeling Techniques

11.2.1 Principal Component Analysis

11.2.1.1 Process Monitoring Using PCA.

11.2.2 Projection to Latent Structures.

Notes:

Description based on print version record.

ISBN:

9780128110669

012811066X

9780128110652

0128110651