Biomedical Engineering Reference
In-Depth Information
humans so may not elicit an anti-protein antibody in human
trials, while in preclinical studies the species are heterol-
ogous. Therefore, the relevance of an immunogenic
response in nonclinical studies generally does not translate
to humans, thus the impact is on interpretation of
nonclinical study and the design and conduct of subsequent
studies. Immunogenicity may limit the conduct of chronic,
DART, and carcinogenesis studies.
dictate different approaches for safety evaluation (
Demp-
ster, 2000; Tsang and Cortez, 2008
).
Due to their small size, small molecules generally bind
to multiple active sites (e.g. pharmacophores, toxico-
phores). Also, small molecules are often metabolized to
compounds with different activity than the original mole-
cule. As a result, small molecule toxicity is broad and
difficult to categorize given that toxicity often has an off-
target component or is due to a metabolite and can, there-
fore, occur via a wide variety of mechanisms.
On the other hand, large molecules are usually so large
that they can only fit into one receptor sufficiently to activate
it and, therefore, usually bind only to the pharmacological
target. Further, metabolismof largemolecules usually results
in their inactivation without production of any other active
molecules. As a result, largemolecule effects are typically an
extension of exaggerated pharmacological effects
(e.g. erythropoietin causing erythrocytosis, hypoglycemia
with insulin). Additional types of toxicity include conditions
such as immune-complex formation (e.g. antigen
Toxicities with Small and Large Molecules
All compound-related effects are evaluated in nonclinical
toxicity/safety studies, not just effects that are considered
adverse (toxic). Compound-related effects may be phar-
macological, physiological or toxicological, adverse or
nonadverse, primary or secondary, or adaptive, reactive, or
reparative. Not all drug-induced changes in animal models
are relevant for humans, whether healthy or diseased.
However, without scientific justification for determining
that a finding is not relevant, by default it is assumed
relevant until proven otherwise.
Effects seen in toxicity studies may be on-target, off-
target, or both. On-target effects are those that occur as
a result of the candidate molecule binding to the intended
pharmacological target receptor. On-target effects may be
the expected pharmacological effect or may be unexpected
effects associated with the targeted receptor itself or an
extension of the biochemical pathway associated with the
targeted receptor. Off-target effects are those due to the
candidate molecule or its metabolite binding to receptors or
sites other than the targeted receptor or other effects not
related to the pharmacological
antibody
complex formation leading to glomerulonephritis with
recombinant human gamma interferon (
Terrel and Green,
1993
)), and activation of pleiotropic receptors (e.g. vascular
leak syndrome of recombinant human interleukin-2 stimu-
lated lymphocytes) (
Anderson et al., 1993
).
Special attention for assessing potential toxicity arises
when the targeted receptor is normally locally activated but
is widely distributed throughout the body. In these cases,
biopharmaceuticals administered systemically may cause
massive activation of a receptor that is biologically regu-
lated for local activation. A recent catastrophic clinical trial
occurred with a CD28 agonist antibody, TGN1412, being
developed for disease related to low numbers of activated
T cells, such as B-cell lymphoma, or for treatment of
autoimmune diseases, such as rheumatoid arthritis (
Atter-
wala, 2010
). This antibody caused systemic activation and
proliferation of T cells, bypassing cellular control and
regulation. High doses of antibody were given in multidose
nonclinical studies in both rhesus and cynomolgus
monkeys; no adverse findings occurred. The expected
pharmacodynamic effect of increased levels of CD4
þ
and
CD8
þ
and moderately increased cytokine levels (i.e. IL-2,
IL-6, and IL-5) occurred at ~2 weeks, but these were
interpreted as nonadverse due to low magnitude of change,
lack of clinical signs, and lack of systemic and organ
toxicity. When given parenterally in phase 1 clinical trials,
human volunteers developed multi-organ failure within 8
hours. As a result of this tragic event, multiple changes in
expectations for clinical trials resulted, including: conduct
of nonclinical studies and assessment of parameters for
assessing theoretical toxicities based on pathways and
pharmacological effects anticipated to be caused by the
drug; evaluation of the pathogenesis and comparative
medicine aspects of even minimal findings in nonclinical
e
target or its associated
pathways.
Preclinical toxicity studies have the same objectives for
both small (chemical substances) and large (biological
substances; e.g. proteins) molecules, and these are to
identify target organ toxicity, starting doses for clinical
trials, establish dose
response relationships, and provide
information for biomarker selection (
Dempster, 2000;
Baumann, 2009
). Small molecules, considered conven-
tional pharmaceuticals, are characterized by low molecular
weight, well-defined physical and chemical characteristics,
chemical synthesis, stability, heat insensitivity, and
metabolism to potentially toxic or pharmacologically
active components. Biotechnologically derived pharma-
ceuticals are characterized by large molecular weight,
complex physicochemical characteristics inherent in three-
dimensional protein structures, are produced from geneti-
cally manipulated cell lines or isolated from biological
fluids, are heat and physically sensitive, and are catabolized
to inactive naturally occurring components (
Dempster,
2000; Cavagnaro, 2008
). While testing objectives are the
same, these differences between small and large molecules
e
