Biomedical Translational Research Drug Design and Discovery (R.C. Sobti, Naranjan S. Dhalla)

table, called as design matrix. After screening the inﬂuential variables, experimental

studies deduce the factor levels. These screening designs invariably operate at two

levels, low (1) and high (+1). In the course of employing FbD, QRM can be

coupled up with factor screening to select high-risk CMAs and CPPs. This step is

essential, as improperly selected factor may bring about unjustiﬁable rise in ﬁnancial

expenditure and experimental effort.

18.4.3 Step III: DoE-Steered Experimentation and Search

for Optimum Nanoconstructs

Only after suitable prioritization of highly inﬂuential input variables, drug delivery

systems are subjected to optimization. DoE trials are performed, as per the chosen

experimental design taking the observed values of various CQAs, for establishing

deﬁnitive relationship(s) among factors and responses. Response surface plot is the

graphic presentation of this relationship to help in understanding the effect of each

input variable along with their plausible interaction(s) on the response variable

(Singh et al. 2005b, 2011a). This 3-D response surface plot is constructed between

two independent variables and a CQA, with their respective 2-D slices known as the

contour plots (Bhavsar et al. 2006; Weissman and Anderson 2015). The contour

plots are graphical representations of one independent factor varying versus another,

while the responses and other input factors are maintained as unaltered. For a deeper

insight, Fig. 18.7 is reproduced here as 3-D and the corresponding 2-D contour plot,

portraying the changes in response as the result of factor interactions.

An experimental design is imperative for response surface mapping based on the

desired goals. Several second-order experimental designs, like factorial design (FD),

central

composite

design

(CCD),

D-optimal

mixture

design

(D-OD)

and

Box-Behnken design (BBD) (Fig. 18.6), are the most often employed for optimizing

drug nanoconstructs. This is because of the fact that such designs can very well

analyse various plausible nonlinear responses, interactions and mixture effects

Fig. 18.6 Cubical representation of key experimental designs employed during QbD-enabled

product development

QbD-Steered Systematic Development of Drug Delivery Nanoconstructs:. . .

323