18 Near-Infrared Spectroscopy in the Pharmaceutical Industry
393
18.2 ICH Guidance, Validation Principles, and Lifecycle
Management
The industry is highly regulated to ensure that patients are provided with safe and
effective active pharmaceutical ingredients (APIs). Health authorities publish guidelines that describe the clinical and manufacturing principles that companies should
consider to potentially gain market approval in their respective jurisdictions. In an
effort to harmonize expectations and requirements across the industry, the International Council for Harmonization of Technical Requirements for Pharmaceuticals for
Human Use (ICH) has published since 1990 several guidelines on manufacturing,
quality control and assurance and some specific countries have also published guidelines on the use of PAT [2], describing a framework designed to encourage companies to use process analytics tools to support development and innovation in the
manufacturing of medicines.
The development of PAT applications typically requires a multidisciplinary
approach, combining the use of simple process sensors and analyzers, chemometric
tools (multivariate tools for experimental design and analysis), process supervision
and control, and continuous improvement tools.
While the majority of spectroscopic PAT applications are for process understanding, an increasing number of submissions are specifically for assuring quality or
releasing product. Because spectroscopic PAT methods often rely on a multivariate
model, the impact of these models to the patient dictates the required level of model
verification and validation. Three levels were discussed by ICH [8]:
– Low-Impact Models: typically used to support product and/or process development
– Medium-Impact Models: assuring quality of the product but are not the sole
indicators of product quality
– High-Impact Models: prediction from the model is a significant indicator of
quality of the product.
ICH-Q2(R1) [9] is usually the framework followed by pharmaceutical scientists to validate analytical methods. It references the requirements for the various
figures of merit an analytical method must include (i.e., accuracy, precision, linearity,
range, robustness, …). A number of articles describing the validation of near-infrared
analytical methods have been published [10, 11]. In addition, frameworks of analytical development practices have been published such as the Analytical Quality by
Design [12]. Several standards discuss these in detail [13, 14] and regulatory agencies
have also published their expectations [15, 16].
Finally, it is necessary to state that the current challenge in the use of PAT methods,
including NIRS, is their lifecycle management. Method validation is now well understood but as these tools get used in commercial settings and need updates as the instruments, raw materials, and processes change, more experience needs to be developed
393
18.2 ICH Guidance, Validation Principles, and Lifecycle
Management
The industry is highly regulated to ensure that patients are provided with safe and
effective active pharmaceutical ingredients (APIs). Health authorities publish guidelines that describe the clinical and manufacturing principles that companies should
consider to potentially gain market approval in their respective jurisdictions. In an
effort to harmonize expectations and requirements across the industry, the International Council for Harmonization of Technical Requirements for Pharmaceuticals for
Human Use (ICH) has published since 1990 several guidelines on manufacturing,
quality control and assurance and some specific countries have also published guidelines on the use of PAT [2], describing a framework designed to encourage companies to use process analytics tools to support development and innovation in the
manufacturing of medicines.
The development of PAT applications typically requires a multidisciplinary
approach, combining the use of simple process sensors and analyzers, chemometric
tools (multivariate tools for experimental design and analysis), process supervision
and control, and continuous improvement tools.
While the majority of spectroscopic PAT applications are for process understanding, an increasing number of submissions are specifically for assuring quality or
releasing product. Because spectroscopic PAT methods often rely on a multivariate
model, the impact of these models to the patient dictates the required level of model
verification and validation. Three levels were discussed by ICH [8]:
– Low-Impact Models: typically used to support product and/or process development
– Medium-Impact Models: assuring quality of the product but are not the sole
indicators of product quality
– High-Impact Models: prediction from the model is a significant indicator of
quality of the product.
ICH-Q2(R1) [9] is usually the framework followed by pharmaceutical scientists to validate analytical methods. It references the requirements for the various
figures of merit an analytical method must include (i.e., accuracy, precision, linearity,
range, robustness, …). A number of articles describing the validation of near-infrared
analytical methods have been published [10, 11]. In addition, frameworks of analytical development practices have been published such as the Analytical Quality by
Design [12]. Several standards discuss these in detail [13, 14] and regulatory agencies
have also published their expectations [15, 16].
Finally, it is necessary to state that the current challenge in the use of PAT methods,
including NIRS, is their lifecycle management. Method validation is now well understood but as these tools get used in commercial settings and need updates as the instruments, raw materials, and processes change, more experience needs to be developed