Section II: Benefit-Risk Management Principles and Practices
Chapter 6: The Impact of Preclinical Planning and Study Outcome on the Risk Management of Biologicals
88
information for risk mitigation purposes through
additional target organ function analyses or
pathology characterization to address anticipated
toxicities. The dose range, as well as the number of
dose groups, should anticipate a pharmacological
or no observable adverse effect level (NOAEL),
as well as a maximum feasible dose for a biother-
apeutic drug candidate. In some cases, it may be
strategically advantageous to include four or five
dose groups rather than the standard three dose
levels, especially when it is important to have a
NOAEL or no observed effect level (NOEL),
and the in vivo pharmacology data in the toxi-
cology species are lacking or weak. This must be
balanced by consideration of the principles of
3Rs—replacement, reduction, and refinement—
which may limit the number of animals that can
be used. Another case may be one in which a
drug has biphasic or dual activity depending on
the dose (e.g., target potentiation at high doses
and suppression at low doses). At the very least,
the dosing paradigm should be based on the can-
didate’s absorption, distribution, metabolism and
excretion (ADME) properties and the intended
therapeutic application in the clinic.
The design of the preclinical GLP toxicol-
ogy studies ideally would take advantage of the
ETLA information and the early pharmacology
profiling. If, for example, the literature or early
evidence indicates particular organ toxicity,
parameters that assess effects on this system
certainly should be included in the study design,
even if it means adding so-called nonstandard
endpoints, such as humoral immunity parameters
in the case of a suspected immunomodulator.
In general, the GLP preclinical testing phase
also should be viewed as a risk management tool.
Being able to write a convincing interpretation
of the safety implications around a liability for
an investigational new drug (IND) applica-
tion requires having pertinent preclinical data
addressing an anticipated liability. Reviewers
and clinicians generally are more open to safety
arguments supported by data than those with
speculative declarations about a preclinical
finding’s relevance to clinical safety. Similarly,
demonstrating the utility of a potential phar-
macodynamic or toxicity biomarker in the GLP
setting sometimes can be a deciding factor for
taking the drug into human studies.
Several aspects of preclinical study design
stand out above others because they are often the
pivot points or deciding factors in a successful
IND submission. These include proving relevant
pharmacology (animal models and study design),
using characterized test material at maximum
dosing levels, and employing studies of suffi-
cient duration (with adequate exposure).13 For
instance, clinical studies in the area of obesity
treatments often need longer treatment time-
frames (e.g., six weeks) to provide convincing
evidence of efficacy. Incorrectly addressing these
study elements can lead to project timeline delays
and to the need to repeat studies.
Study Design and Species Selection
For biopharmaceuticals, the toxicology plan
must be tailored to match the drug candidate’s
pharmacology and species specificity more than
is the case with small molecule standard tox-
icology plans. A key point of emphasis in the
original ICH S6 guidance, Preclinical safety
evaluation of biotechnology-derived pharma-
ceuticals, is the design of appropriate toxicology
studies and the use of a relevant animal model
a species that is pharmacologically responsive
to the intended human drug. However, because
biopharmaceuticals have unique attributes
and mechanisms of action, choosing the most
appropriate animal model for toxicology testing
requires a case-by-case approach. The ICH S6
guidance section on species selection reinforces
the notion that two species are needed, but in the
case where there is only one relevant pharmaco-
logically active model, there is no need to create a
transgenic species or use a homologous molecule
just to achieve that goal.14 However, because the
decision about whether one animal species is
sufficient for the preclinical GLP studies is based
mostly on experience and scientific judgment
of the candidate biotherapeutic’s mechanism of
action, there is usually some degree of sponsor
trepidation about whether regulatory author-
ities will concur with this judgment prior to
submission of the IND. This is especially true
when there are potentially viable alternatives for
preclinical testing, such as the use of surrogate
Chapter 6: The Impact of Preclinical Planning and Study Outcome on the Risk Management of Biologicals
88
information for risk mitigation purposes through
additional target organ function analyses or
pathology characterization to address anticipated
toxicities. The dose range, as well as the number of
dose groups, should anticipate a pharmacological
or no observable adverse effect level (NOAEL),
as well as a maximum feasible dose for a biother-
apeutic drug candidate. In some cases, it may be
strategically advantageous to include four or five
dose groups rather than the standard three dose
levels, especially when it is important to have a
NOAEL or no observed effect level (NOEL),
and the in vivo pharmacology data in the toxi-
cology species are lacking or weak. This must be
balanced by consideration of the principles of
3Rs—replacement, reduction, and refinement—
which may limit the number of animals that can
be used. Another case may be one in which a
drug has biphasic or dual activity depending on
the dose (e.g., target potentiation at high doses
and suppression at low doses). At the very least,
the dosing paradigm should be based on the can-
didate’s absorption, distribution, metabolism and
excretion (ADME) properties and the intended
therapeutic application in the clinic.
The design of the preclinical GLP toxicol-
ogy studies ideally would take advantage of the
ETLA information and the early pharmacology
profiling. If, for example, the literature or early
evidence indicates particular organ toxicity,
parameters that assess effects on this system
certainly should be included in the study design,
even if it means adding so-called nonstandard
endpoints, such as humoral immunity parameters
in the case of a suspected immunomodulator.
In general, the GLP preclinical testing phase
also should be viewed as a risk management tool.
Being able to write a convincing interpretation
of the safety implications around a liability for
an investigational new drug (IND) applica-
tion requires having pertinent preclinical data
addressing an anticipated liability. Reviewers
and clinicians generally are more open to safety
arguments supported by data than those with
speculative declarations about a preclinical
finding’s relevance to clinical safety. Similarly,
demonstrating the utility of a potential phar-
macodynamic or toxicity biomarker in the GLP
setting sometimes can be a deciding factor for
taking the drug into human studies.
Several aspects of preclinical study design
stand out above others because they are often the
pivot points or deciding factors in a successful
IND submission. These include proving relevant
pharmacology (animal models and study design),
using characterized test material at maximum
dosing levels, and employing studies of suffi-
cient duration (with adequate exposure).13 For
instance, clinical studies in the area of obesity
treatments often need longer treatment time-
frames (e.g., six weeks) to provide convincing
evidence of efficacy. Incorrectly addressing these
study elements can lead to project timeline delays
and to the need to repeat studies.
Study Design and Species Selection
For biopharmaceuticals, the toxicology plan
must be tailored to match the drug candidate’s
pharmacology and species specificity more than
is the case with small molecule standard tox-
icology plans. A key point of emphasis in the
original ICH S6 guidance, Preclinical safety
evaluation of biotechnology-derived pharma-
ceuticals, is the design of appropriate toxicology
studies and the use of a relevant animal model
a species that is pharmacologically responsive
to the intended human drug. However, because
biopharmaceuticals have unique attributes
and mechanisms of action, choosing the most
appropriate animal model for toxicology testing
requires a case-by-case approach. The ICH S6
guidance section on species selection reinforces
the notion that two species are needed, but in the
case where there is only one relevant pharmaco-
logically active model, there is no need to create a
transgenic species or use a homologous molecule
just to achieve that goal.14 However, because the
decision about whether one animal species is
sufficient for the preclinical GLP studies is based
mostly on experience and scientific judgment
of the candidate biotherapeutic’s mechanism of
action, there is usually some degree of sponsor
trepidation about whether regulatory author-
ities will concur with this judgment prior to
submission of the IND. This is especially true
when there are potentially viable alternatives for
preclinical testing, such as the use of surrogate