Quality- by-design in pharmaceutical development - Computer-Aided Applications in Pharmaceutical Technology - page 11

launch (Migliaccio, 2011), which indicates that QbD results in robust
processes and is able to rapidly improve process capability. Also, QbD
resulted in lower deviation rates in the fi rst year after launch than
achieved through traditional continuous improvement efforts (Migliaccio,
2011).
There are a variety of opportunities for the QbD concept to be applied
to existing products: processes can be redesigned, partial design spaces
can be defi ned, enhanced control strategies can be appointed (including
real-time release), or new technologies (i.e. continuous manufacturing)
can be developed (Migliaccio, 2011).

1.4 Scientifi cally based QbD -
examples of application

Some of the issues encountered by the regulatory agencies during the
assessment of a QbD based registration dossier are lack of relevant
explanations of the conclusions reached, insuffi cient graphical
presentations of the factor interactions, design space boundaries not
clearly described, no information on statistical validity of models, and
not enough structure in the presented data, etc. (Korakianiti, 2011).
Collaboration between scientists in industry, academia, and regulatory
bodies' experts is necessary to overcome the above-mentioned issues.
Many scientifi c projects are devoted to design space appointment, in-line
process monitoring, and modeling of products and processes. This
knowledge should serve to provide a foundation for the scientifi cally
based QbD concept application. Some of the peer-reviewed examples of
QbD elements development are presented below.
The QbD approach was used to establish a relationship between the
CPPs, CQAs, and clinical performance of the drug (Short et al., 2011).
Extended-release theophylline tablets were analyzed, showing that some
of the compendial tests are insuffi cient to communicate the therapeutic
consequences of product variability. Both critical and noncritical
attributes were used as inputs to the design space, which was conditioned
on quantitative estimates of ineffi cacy and toxicity risk.
A combined QbD and Discrete Element Model (DEM) simulation
approach was used to characterize a blending unit operation, by
evaluating the impact of formulation parameters and process variables
on the blending quality and blending end point (Adam et al., 2011). QbD
was used to establish content uniformity as CQA and link it to blend

￿

￿ ￿