Biomedical Engineering Reference
In-Depth Information
In addition to the ASTM and ISPE groups mentioned above, several other groups
offer opportunities for interaction with regulators and the exchange of best practices. The
Parenteral Drug Association (PDA) has a quality and regulatory department that interacts
with FDA, EMEA, WHO, ICH, USP, and numerous other regulatory bodies around the
world, and it also notifies members for updates in regulatory initiatives. The PDA also
develops technology reports and guidance documents where they perceive that gaps exist,
and it also arranges conferences with a focus on regulatory, manufacturing, and science
and technology topics around the world. The International Foundation for Process
Analytical Chemistry (IFPAC) runs annual meetings to share information on PAT across
a broad industry base. Initially dominated by large chemical and petrochemical manu-
facturers, IFPAC is increasing its focus on the pharmaceutical and biotechnology
industries, with the FDA functioning as one of the major sponsors and providing three
of the conference co-chairs over the past 4 years. It is unusual as one of the fewconferences
supports cross-industry learning in an increasingly specializedworld. The Pharmaceutical
Process Analytics Roundtable (PPAR) is an informal discussion group that meets annually
to discuss PAT-related topics in the pharmaceutical and biotechnology industries.
13.3.1 PAT Tools
As described in the PATGuidance, PAT tools fall into four main areas: process analyzers,
process control tools, multivariate tools for design, data acquisition, and analysis, and
continuous improvement and knowledge management tools [2].
There are a host of analyzers suitable for gathering information from biotech
processes, ranging from standard sensors such as pH, conductivity, and temperature,
through more sophisticated spectroscopic techniques, to sophisticated online analyzers.
UV is universally used to detect and track protein products and to make processing
decisions as to where to collect product or divert process streams to waste, but it is not
capable of discriminating between product and closely related contaminants. Near
infrared (NIR) is widely used in small-molecule processes and is useful for testing the
identity of incoming materials and with the potential to yield actionable information
about the condition of upstream processes [19]. Online high-performance liquid
chromatography (HPLC) has been used to determine peak quality in purification
processes [20, 21], but the problem downstream lies in shortening analysis times to
seconds rather than minutes. There is a significant need for developments in sensor
technology, specifically to cope with biological molecules and the relationships between
structural integrity/aggregation and biological properties. This gap was highlighted at a
recent workshop held at IFPAC 2008 and co-chaired by the FDA, NIH, NIST, and MIT.
Obvious candidates would be methods for rapid determination of glycosylation, the
presence of aggregates, and contaminants such as endotoxins [22, 23].
Multivariate data analysis (MVDA) is an essential tool. MVDA applies statistical
tools such as principal component analysis (PCA) and partial least squares (PLS) to
complex data sets to study their multidimensional nature and the interdependence of
multiple material attributes and process parameters [24]. These tools can be readily
applied to upstream applications where there is a multitude of input and process
parameters that can be studied [25] and whose interactions may need to be defined.
