PAT implementation in pharmaceutical manufacturing and its economical impact

The overall aim of this thesis was to evaluate the feasibility of implementing Process
analytical Technology (PAT) into pharmaceutical manufacturing. First of all the
implementation of PAT was looked at from a bird perspective, taking all unit
operations into account. This task was divided into two parts, the first one developing
a PAT strategy that could result in a real time release in tablet manufacturing for one
product, identifying the critical variables in the manufacturing process and selecting
the analytical techniques to measure these critical variables. The second part of the
bird perspective determined the economical feasibility of the PAT implementation
based on the strategy formed in part one for three products. After determining the
overall feasibility of the PAT implementation a closer look was given to special
technologies which contribute to the PAT philosophy in ways of product efficiency,
manufacturing flexibility, process understanding and building quality into the
products instead of testing quality into products.
Critical process variables were determined by analysis of validation data, discussion
with plant operators and literature analysis of the fundamental operation of each
process step. The critical process variables were determined to be moisture content,
identity, and particle size distribution of the raw materials, blend homogeneity of the
mixing steps, particle size distribution after granulation, content uniformity of the
drug substance and hardness of tablets, coating thickness, final tablet moisture
content, droplet size distribution of coating solution, identity of material packaged and
package integrity.
Analytical techniques to measure each of the critical variables were chosen based on a
literature search to evaluate what techniques are available and further narrowed based
on availability of equipment. Diffuse reflectance near infrared spectroscopy (DRNIR)
is recommended for determining component identity, moisture content, blend
homogeneity, hardness, coating thickness, and material identity. Transmittance NIR
(T-NIR) is recommended for determining content uniformity. Laser light diffraction
(LLD) is recommended for determining particle size distribution. The PAT strategy
that combines process monitoring and product characterization will hopefully
eliminate or at least reduce the need for Quality Assurance / Quality Control (QA /
QC) laboratory testing.
The first technologies to be implemented should be the ones that are easiest to
implement and least expensive. Full adoption of the PAT philosophy will require an
alteration of the method used to develop pharmaceuticals. A better scientific
understanding of the process and materials will be required so as to result in a process
that is fully understood, has a process model and produces materials with built-in
quality assurance for real time release.
The two areas where PAT can provide the greatest savings were found to be: (1)
reduction in Quality Control (QC) testing, leading to real time release and (2)
improvement of product yield from better process control. The financial analysis is
sensitive to the time when these benefits begin to take effect. Since there is some
uncertainty as to when real time release will be approved, two scenarios were created
for implementation of PAT. The first scenario (base case) assumes that QC savings
start accruing after the installation and qualification of analytical instruments, at the
end of the first year. The second scenario is a conservative case in which it is
assumed that the QC savings begin in the third year, after correlations have been
developed for tablet friability, disintegration time and dissolution time.
PAT implementation is envisioned to occur in three stages. The first phase involves
the installation of on-line analytical equipment and results in benefits related to
savings in QC costs, reduced time for investigations, and reduced throughput time.
During the first phase of PAT implementation, correlations will be developed for
process understanding and to replace current analytical tests such as disintegration,
dissolution and friability. The second phase is the implementation of process control
systems which results in savings due to fewer Food and Drug Administration (FDA)
audits, fewer investigations, fewer batch rejections, shorter throughput times, and
increased product yield. The third phase is the integration of PAT into the
development process of a new product, resulting in quality by design. The third phase
was not addressed within this project as PAT implementation was only evaluated for
products already on the market and so concentrating on the production side. In order
to investigate the impact of PAT on the development process of a new drug product a
separate dissertation work would be necessary.
For the first scenario (QC savings after the first year), the results of the economic
analysis indicate that the implementation of PAT for solid tablet manufacturing is
very cost-effective. The Return on Investment (ROI) for implementation of PAT after
phase I is 69% for Product X, 49% for Product Z and 46% for Product Y. The Return
on Investment for implementation of PAT in phase II is 147% for Product X, 92% for
Product Z and 60% for Product Y.
For the second scenario (no QC savings in phase I), the Return on Investment drops to
130% for Product X, 81% for Product Z and 49% for Product Y. Thus, PAT
implementation is cost-effective even in the conservative case.
Companies tend to move ahead with a project when the Return on Investment is at
least 25% and these results demonstrate that PAT should be implemented. The
Return on Investment results indicate that the installation of on-line analytical
equipment creates great savings for companies while implementation of control
systems is even more cost effective.
In matters of new manufacturing technology, the Glatt Multicell®was evaluated as an
innovative approach for granulation. The Multicell®, being a semi-continuous
granulator producing small sub-batches of approximately 7kg, is able to circumvent
the unpleasant practice of scale-up as it is capable of producing one sub-batch or just
as many as required. The capacity of the Multicell® was compared to the capacity of
conventional granulation equipments. It did not only show that it could hold up with
the conventional granulation but also increased capacity dramatically and at the same
time reducing the needed personnel or Full Time Equivalents (FTE) from 18 down too
7, bringing in yearly savings of approximately 1100kCHF.
Another technology which contributes to process understanding and building quality
into the product is the MCC Presster®. The Presster® is a rotary press simulator which
is capable of replicating different rotary presses by utilizing the same compression
tooling and using the same dwell time as the original rotary press with only one pair
of punches. This allows the comparison of different rotary presses with only one
machine and very little powder formulation. In this study the Presster® was used to
replicate the results of three different rotary presses. Two commercial product
formulations were run on all three rotary presses with different dwell times. The two
product formulations were chosen considering their flowing properties, one product
having bad flowing properties and the other very good flowing properties. It was
shown that the Presster® was capable of simulating the effects of a full scale tableting
machine if the Presster® is equipped with the corresponding compression tooling.
Further it was shown that prolonged pre-compression dwell time due to pneumatic
compensation leads to tablets with higher crushing strength than regular precompression
rolls.
Last but not least a Near Infrared (NIR) Spectrometer was evaluated for
implementation on a packaging line. This technology also contributes to the proposed
PAT philosophy by enhancing process understanding and enabling 100% real time
analysis of the product being packed. For this study a packaging line simulator was
used with an integrated NIR spectrometer system from Visiotec called VisioNIR®.
The products used for this study were all strengths of Sandimmun Neoral® which are
in soft gelatin capsules and different tablet products for mix-up studies. The aim of
this study was to estimate which product flaws the VisioNIR could detect within the
soft gelatin capsules, such as capsules without content, capsules with wrong content
and product mix-ups. It was shown that the VisioNIR was capable of distinguishing
empty capsules, excessively smudged capsules and capsule mix-ups from the
Sandimmun Neoral® capsules for all strengths. It became apparent that Sandimmun
Neoral® strengths which were grey were harder to analyze as the lighter soft gelatin
capsules as the halogen light did not penetrate deep enough into the capsule shell.