Biomedical Engineering Reference
In-Depth Information
or to be mediated, linearly or nonlinearly, by the presence of the target or of other
nonspecific binding elements. The model could also be further extended to incor-
porate specific signaling events in the tumor cell population. The caveat to this,
much like developing biochemical or molecular systems-biology model, is that
with increasing granularity comes an increase in the assumptions and supportive
data necessary for model building and validation.
Until recently, systems biology and pharmacology have existed separately and
very little overlap or integration has occurred. QSP is an integrative discipline,
incorporating computational data analysis and modeling with molecular and cellu-
lar biology. Although QSP modeling is still in its infancy, there are a couple of
elegant examples of multi-scale modeling and simulation incorporating biochemi-
cal and molecular reaction networks and whole-body models [
21
,
97
]. Wu et al. [
97
]
modeled the VEGF system from the molecular level to the cellular level to tissue
level to organism and Eissing et al. present a software platform to conduct multi-
scale systems modeling. In the work published by Eissing et al., the authors created
a virtual patient with a tumor in the pancreas. The tumor was subdivided into
proliferating, resting and dead cells and nested a molecular model of the EGFR-
MAPK pathway into the cells. They then simulated the treatment effects on the
tumor. The drug used in their virtual patient is a prodrug converted to the active
metabolite in the liver by CYP2D6, which has known a polymorphism that affects
the metabolism rate. They were able to model the effect different phenotypes of
CYP2D6 would have on the effect of treatment. They were also able to vary the
concentrations of the proteins in the EGFR signal transduction to model the effect
that would have on drug treatment. These two examples illustrate the power and
utility of multi-scale, mechanistic QSP models.
12.5 Translation
There are significant challenges in the development of novel targeted oncology
drugs [
31
,
47
,
100
] and the predictable translation of impressive preclinical activity
of agents to successful clinical compounds remains elusive. Recently, greater
emphasis has been placed in developing knowledge of the concentration of drugs
at the site of action and understanding how this relates to modulation of the target
pathway and the measurement of therapeutic effect [
12
,
53
,
95
]. Preclinical QSP
studies are of critical importance because they allow us to establish how plasma
concentrations relate to target concentrations, target modulation, surrogate
biomarkers (a clinical or translational biomarker), and growth effect so we can
extrapolate the relationship to patients, where often the only measures obtained are
plasma PK and/or a surrogate biomarker (Figs.
12.1
and
12.4
).
In clinical studies, tremendous variability is often observed both in the pharma-
cokinetics and in response to therapy. Recently, efforts have been made to develop
“biomarkers” to predict which patients may respond to a given therapy versus those
who will not. This has been successfully employed in non-small cell lung cancer,
Search WWH ::
Custom Search