Quality- by-design in pharmaceutical development - Computer-Aided Applications in Pharmaceutical Technology - page 13

linked to CQAs of the product. Variability was reduced by product and
process understanding, which translated into quality improvement, risk
reduction, and productivity enhancement. The risk management approach
further led to a better understanding of the risks, ways to mitigate them,
and control strategy proposed commensurate with the level of the risk.
The production bioreactor step of an Fc-Fusion protein manufacturing
cell culture process was characterized following QbD principles (Rouiller
et al., 2012). Using scientifi c knowledge derived from the literature and
process knowledge gathered during development studies and
manufacturing to support clinical trials, potential critical and key process
parameters with a possible impact on product quality and process
performance, respectively, were determined during a risk assessment
exercise. The identifi ed process parameters were evaluated using a design
of experiment approach. The regression models generated from the data
characterized the impact of the identifi ed process parameters on quality
attributes. The models derived from characterization studies were used to
defi ne the cell culture process design space. The design space limits were
set in such a way as to ensure that the drug substance material would
consistently have the desired quality.
QbD principles were used to investigate the spray drying process of
insulin intended for pulmonary administration (Maltesen et al., 2008).
The effects of process and formulation parameters on particle
characteristics and insulin integrity were investigated. Design of
experiments and multivariate data analysis were used to identify
important process parameters and correlations between particle
characteristics. Principal component analysis was performed to fi nd
correlations between dependent and independent variables.
A multiparticulate system, designed for colon-specifi c delivery of
celecoxib for both systemic and local therapy, was developed using QbD
principles (Mennini et al., 2012). Statistical experimental design (Doehlert
design) was employed to investigate the combined effect of four
formulation variables on drug loading and release rate. Desirability
function was used to simultaneously optimize the two responses.
A QbD approach was also used to study the process of a nanosuspension
preparation (Verma et al., 2009), to establish appropriate specifi cations
for highly correlated active substance properties (Cui et al., 2011), to
develop analytical methods (Vogt and Kord, 2011), and its usage in lead
drug candidates optimization is proposed to address productivity in drug
discovery (Rossi and Braggio, 2011). The role of predictive
biopharmaceutical modeling and simulation in drug development, in the
context of QbD, was also presented (Jiang et al., 2011).

￿

￿ ￿