Joint Pharmaceutical Analysis Group
|
The science base underpinning pharmaceutical development
and manufacture has failed to keep pace with available technology.
The drivers, means and benefits of adopting a process analytical
technology strategy, its facilitation in the regulatory environment
and its impact on analytical science and the analyst were reviewed
in a workshop held by the Joint Pharmaceutical Analysis Group
at the Royal Pharmaceutical Society's London headquarters on
4 December. Dr Joseph Chamberlain reports
|
Process analytical technology: the challenge for analytical
science
Dr Chamberlain is a former editor of the Journal of Pharmacy and
Pharmacology
|
Ken Leiper, of Benson Associates, reviewed the present situation and
the drivers for the development of process analytical technology (PAT).
The largest cost to pharmaceutical companies was in manufacture so
that even a 1 per cent improvement in this area would mean a saving of
hundreds
of millions of dollars. Yet, the pharmaceutical industry compares unfavourably
with other major industries (food, automobiles) in its efficient use
of resources, characterised by large, inefficient batch equipment,
low product yields, long lead-times due to stage and final product testing,
high operating costs, high inventories and excessive warehouse space.
It is capital and labour intensive, and improvement is invariably limited
by regulatory constraints.
In the analytical area, too much emphasis is placed on aspects of the
analysis that are least critical, and a typical analytical procedure
involves consecutive, time-consuming steps. The analytical emphasis
needs to be on timely provision of information that contributes to
final quality.
Although the current approach is probably delivering as much as is
reasonably possible in respect of safety and efficacy, this is at the
cost of a
significant impact on regulation and manufacture.
Process analytical technology is not new: there has been industry interest
for at least 35 years as a problem-solving tool but there has been
little or no industry pressure to bring it into the mainstream regulatory
environment.
The United States Food and Drug Administration is now preparing to
discuss mechanisms to redress this situation. The aim now is to institute
systems
for analysis and control of manufacturing processes based on timely
measurements, during processing, of critical quality parameters and
performance attributes
of raw and in-process materials and processes to assure acceptable
end product quality at the completion of the process. The FDA will
encourage
companies to move away from current univariate prescriptive testing
to multivariate process-focused measurements.
These initiatives provide a dynamic regulatory framework that will
address the causes of poor performance rather than the symptoms, drive
the introduction
of innovative manufacturing measurement and control technologies, and
allow quality and performance to be designed into manufacturing scale
processes.
We must understand all our processes, said Mr Leiper, and promote a
regulatory and industry environment where relevant science is used
to drive quality
by design in development and manufacture.
The current toolbox
Because of the demands of a wide range of potential applications encompassing
chemical, physical and spatial characterisation, PAT requires a toolbox
of suitable techniques, said Don Clark, of Pfizer R&D, Sandwich,
Kent. Such a toolbox contains the full range of spectroscopic techniques
from X-rays to radio waves. Near infrared (NIR) spectroscopy is a well-established
technique for on- and off-line applications for both the solution and
solid state; it is non-invasive, non-destructive, and no sample preparation
is needed. NIR can be used for real-time process monitoring (qualitative
and quantitative analysis of excipients and active ingredients, blending
end-point determination, determination of polymorphism) and is easy
to use with simple, low-cost equipment. However, NIR data interpretation
is not intuitive or chemist-friendly and substantial data manipulation
is often needed to extract useful quantitative information. NIR cannot
always link results to simple properties of the system and is essentially
a black box approach.
Fourier transform infrared spectro-scopy (FT-IR) is another well-established
technique for on- and off-line use for solution and solid-state material;
it gives direct information on molecular vibrations and is relatively
amenable to direct interpretation. FT-IR is not so useful for polymorphism
(compared with other spectroscopic techniques) but it is readily available
at low to medium cost.
Raman spectroscopy is a powerful technique, becoming much more widely
used and provides chemical and physical information. On-line systems
are commercially available at reasonable cost. It is a particularly powerful
technique for polymorph identification.
Although ultraviolet spectroscopy is a central technique in analytical
chemistry (for example, as a high performance liquid chromatography detector),
it is of limited use in process monitoring; analytes must have a chromophore,
they must be in solution, and chemical discrimination is poor. Acoustic
emission spectroscopy is an emerging technique as described elsewhere
in this report.
Chemical imaging combines microscopy with spectroscopic information to
give spatial characterisation of solid samples. A particularly impressive
example was the demonstration of the distribution of individual particles
of active drug material in a tablet using X-ray microtomography. Imaging
technology is evolving rapidly, the first imaging system being demonstrated
in 1998. However, at the moment, it is still an expensive, low-throughput
technique.
Additions to the toolbox are likely to include nuclear magnetic resonance
(NMR) spectroscopy and mass spectrometry. Although NMR is a solution-based
technique, it provides much structural information and newer technologies
may make it viable for on-line use in manufacturing areas. Mass spectrometry
is sensitive and can track single components in a mixture using molecular
weights. Formulation and process design
Robust formulation and process design starts in R&D, said Dr James
Krausnoe, AstraZeneca, Charnwood, Leicestershire. The impact of process
analytical technology is to emphasise that R&D must employ a holistic
approach and understand the nature of all parts of the manufacturing
process; in practice this will mean carrying out a thorough review of
raw material and the viability of intermediary processes. Available information
must be maximised to build process understanding.
NIR may be used both for screening of raw materials and for characterisation
of the final product (hardness, content uniformity of tablets). Its importance
in improving the manufacturing process can be illustrated by considering
the effect of including NIR early in the analysis cycle, providing prediction
of final properties to correlate with measured properties at a later
stage. This continual evaluation of the manufacturing process provides
information on such parameters as dry mixing, granulation, drying, milling,
compression and subsequent formulation of a control strategy.
The full implementation of PAT will involve the integration of analysts,
formulators, process engineers, and support functions including quality
assurance, engineering and regulatory affairs, requiring management commitment
across the whole of R&D. Any new product introduction will require
a strategy that includes PAT technology. R&D should ensure the analytical
function stays off the critical path of the manufacturing process, Dr
Krausnoe said. Variability in manufacture
Martin Warman, of Pfizer Global Manufacturing, Sandwich, Kent, said
that the role of PAT was to provide process knowledge during, rather
than
at the end, of the process. This would be achieved by monitoring
multiple parameters to enhance process knowledge. Attributes of critical
process
parameters need to be established and used for direct feedback control
to reduce variability where possible and ultimately put an end to
the monitoring of non-critical parameters. For example, NIR data were
used
to monitor a reaction to ensure completion, but more importantly
confirmed the reproducibility of the time course of the reaction. In
another
example, crystal size was monitored during a crystallisation step
to allow the control of fines or agglomerates as well as control of particle
size and the elimination of the need for milling. NIR was used to
monitor
blending, with subsequent benefits for safety (no operator contact),
reduction in sampling errors, real-time information, multi-ingredient
uniformity, process understanding, and reduced cycle times (fast
release of the blend). NIR was also used for estimation of blend segregation
in the bin and for automated analysis of tablets.
There has been a paradigm shift, said Mr Warman, from a laboratory-based
approach to a process-based approach. Acoustics
In process analytical technology, an ideal method would be non-invasive,
non-destructive, relatively inexpensive, have a short measurement time,
and be intrinsic-ally safe. The emerging technique of acoustics had
these attributes, said Professor David Littlejohn (University of Strathclyde).
In this simple technique a transducer containing piezoelectric material
is glued to the outside wall of a reaction vessel and the acoustic
information collected and processed. Acoustic measurement could be
passive (where the process is the source of the acoustic wave) or active
(where the acoustic wave is put into the process and the change in
velocity or attenuation is monitored). In an example to demonstrate
monitoring of a mixing process, acoustic measurement was shown to give
similar results to the established near infrared methods. Additional
information could be obtained from the acoustic spectra; for example,
at lower frequencies (less than 50kHz), the signals obtained were dependent
on the particle size of citric acid added to Avicel, whereas at higher
frequencies (50–150kHz) the signal depended only on the amount
of citric acid added.
Heterogeneous reaction monitoring is sometimes difficult to monitor
with optical spectroscopic techniques and acoustics could assist in
this area
by monitoring the consumption or production of particles, appropriate
kinetic studies, and end-point detection. For example, in the reaction
of itaconic acid with 1-butanol in toluene to give mono- and di-esters,
only the itaconic acid is a solid under the reaction conditions. Acoustic
emission spectra successfully monitored the disappearance of the acid
in real time, comparing favourably with the standard HPLC method for
determining the formation of the esters. In work carried out so far — and
Professor Littlejohn emphasised this work is continually expanding — it
is apparent that acoustics can be used to monitor a wide range of processes,
providing complementary information to molecular spectroscopic techniques
to give real-time information from product discovery through to full-scale
manufacture. The use of multiple sensors appropriately placed around
the reaction or mixing vessels would also provide spatial information.
The wider use of acoustics for process monitoring will require advances
in transducer technology, signal processing, signal interpretation (including
mathematical modelling), and most of all a multidisciplinary approach,
Professor Littlejohn said. The reality
PATs provide additional tools for manufacturing process development,
said Bob Chisholm, of AstraZeneca Engineering. The use of these tools
has led to greatly improved process understanding in the pharmaceutical
development phase which translates through technology transfer into
effective and efficient manufacturing processes. Identified critical
formulation and processing factors are monitored and controlled to
prevent or mitigate the risk
to quality (real-time quality control and
quality assurance).
Regulatory policies and procedures will be tailored to recognise such
approaches and regulatory scrutiny or inspections appropriately applied.
This risk-based approach by the pharmaceutical industry will be mirrored
by a risk-based approach to regulation.
The AstraZeneca solid dosage facility in Germany incorporates networked
in-line NIR analysers at each unit operation providing the capability
to monitor key process variables identified in the risk assessment
process. Each incoming raw material in the active and excipient dispensaries
is
identified and characterised; moisture in the fluid bed drier is monitored
and there is continuous on-board monitoring of powder blending. Statistically
based in-line monitoring of tablet quality parameters is carried out
throughout the batch.
The system includes four NIR analysers with seven spectral inputs and
appropriate computer stations. The resulting real-time quality control
and quality assurance provides the platform for real-time release of
product.
Thus PAT increases process understanding and when used in a manufacturing
execution system is a key component in the toolbox to deliver manufacturing
excellence. Through this understanding, PAT is an effective tool to
improve robustness of existing as well as newly developed processes,
Mr Chisholm
said. Regulatory facilitation
For the final paper Dr Ajaz Hussain, FDA, Washington, joined the meeting
by satellite link and presented the agency’s draft guidance for
industry, “PAT — a framework for innovative pharmaceutical
manufacturing and quality assurance”. Working with existing regulations
the draft guidance describes a regulatory framework to encourage the
voluntary development and implementation of innovative pharmaceutical
manufacturing and quality assurance, said Dr Hussain. There were two
components to this framework: a set of scientific principles and tools
supporting innovation, and a strategy for regulatory implementation
that will accommodate innovation.
Atypically for FDA guidelines, this guidance is written for a broad
industry audience in different organisational units and scientific
disciplines
and it discusses principles with the goal of highlighting technological
opportunities and developing regulatory processes that encourage innovation.
Companies ready with innovative ideas for implementation should propose
to the agency a scientific, risk-based implementation plan and the preferred
regulatory path for implementation. The agency is ready to provide a
scientific assessment of the proposal before a submission to define the
type of data needed to evaluate the proposal and provide a mutually acceptable
regulatory path.
PAT is a system for designing, analysing, and controlling manufacturing
through timely measurements of critical quality and performance attributes
of raw and in-process materials and processes, with the goal of ensuring
final product quality. It works on the principle of quality by design.
The design of manufacturing processes using sound principles of engineering,
material science, and quality assurance ensures acceptable and reproducible
product quality and performance throughout a product’s shelf life.
Gains in quality, safety and efficiency will vary depending on the product
and are likely to come from: reducing production cycle times by using
on-, in-, and at-line measurements and controls; preventing rejects,
scrap and reprocessing; allowing the possibility of real-time release;
giving increased automation; facilitation of continuous processing; use
of small-scale equipment (to eliminate certain scale-up issues) and dedicated
manufacturing facilities; improved energy and material use; and increased
capacity.
A desired goal of the PAT framework is to design and develop processes
that can consistently ensure a predefined quality at the end of the manufacturing
process.
The next steps will be to collect public comments, issue the final guidance
early next year, and mount workshops on this guidance, said Dr Hussain. |
