Risk Management Principles for Devices and Pharmaceuticals
89
animal models or when a drug homolog could
be substituted. Fortunately, a good resource is
available that reviews recommendations for
species selection and the rationale for adopting
different alternative approaches in situations
where a conventional toxicology-screening model
will not suffice for a candidate biotherapeutic’s
safety assessment.15 If there are persistent uncer-
tainties, a meeting with regulators to request
guidance on the proposed submission package
may be warranted. Designing studies for repro-
ductive/developmental risk can be a challenging
task, as the guidelines are flexible around species
choice and design alternatives. The original ICH
S6 guidance did not provide specific direction
on which test species was acceptable, on study
design or on the number of animals per dose
group. The ICH S6 addendum incorporated
recently into the main guidance14 recommends
developmental toxicity studies should not be
conducted in non-human primates unless they
are the only relevant species. Regulators prefer
testing the clinical candidate (in nonhuman
primates if that is the only relevant model), but
alternatives to the drug molecule can be consid-
ered if scientifically justified.
The addendum further says if the mech-
anism of action, class effects, or information
on genetic mutants indicates an adverse effect
on fertility or pregnancy outcome, it may be
sufficient to communicate the potential devel-
opmental or reproductive risk without further
preclinical studies.
For monoclonal antibody drug candidates
active only in nonhuman primates, it is recom-
mended that combined embryo-fetal postpartum
developmental (EFPPD) studies be considered
rather than separate embryo-fetal and perinatal/
postnatal toxicity studies. It is further recom-
mended the dosing interval cover gestation days
(GD) 20–100 (rather than GD 20–50) because
of the prolonged circulating half-lives of most
intact human monoclonal antibodies in primates.
For fertility testing, a standalone nonhuman pri-
mate study is not necessary, but there should be
an assessment of reproductive organs in standard
toxicity studies with the candidate biological.
If there are special concerns, parameters should
be added to provide more detail such as gonad
weights, sperm viability/motility, and reproduc-
tive hormones. Sponsors should realize regulators
recognize these studies are for hazard identifica-
tion rather than true risk assessment due to the
relatively low number of study animals assigned
to each treatment or dose group.
Dose Levels and Duration
One of the most important aspects of preclinical
study design is the choice of dose levels. Much
consideration is given to selecting treatment
doses that will yield the maximum pharmacolog-
ical effect or provide a tenfold exposure multiple
over the maximum estimated exposure in the
clinic, as well as a pharmacologically active dose.
Again, the more extensive the early pharmaco-
logical profiling studies have been in terms of
pharmacokinetic or ADME characterization,
the greater the confidence level for projecting
initial GLP toxicity study dose levels. Factors
that strongly influence the dose range selected
for a preclinical study include the study duration,
gender differences and inherent ADME proper-
ties, including those that are species-specific. For
example, the drug’s in vivo Cmax and its halflife
are properties determining the dose frequency
and the potential toxicity profile in the study.
For many biologicals, the clearance of the drug
is dependent on receptor-binding mechanisms
rather than kidney filtration mechanisms. If a
justification is provided, the maximum dose can
be predicated on data calculations showing that
if all the receptors are occupied at a given dose,
increasing the dose beyond this maximum level
is not informative, according to the ICH S6
guideline addendum.14 For chronic preclinical
toxicology studies, six months is considered a
sufficient length of time to evaluate the haz-
ards associated with chronic, repeat dosing. The
addendum to the ICH S6 guidance recom-
mends an evaluation of recovery (from toxicity)
be included with a statement that recovery is
not intended to detect delayed toxicity. This
topic presumably refers to the characteristically
long circulating halflives of many monoclonal
antibodies, which may take six to nine months
to clear completely from an animal given the
very high doses administered. Operationally,
this referral to delayed toxicity in the guidance
89
animal models or when a drug homolog could
be substituted. Fortunately, a good resource is
available that reviews recommendations for
species selection and the rationale for adopting
different alternative approaches in situations
where a conventional toxicology-screening model
will not suffice for a candidate biotherapeutic’s
safety assessment.15 If there are persistent uncer-
tainties, a meeting with regulators to request
guidance on the proposed submission package
may be warranted. Designing studies for repro-
ductive/developmental risk can be a challenging
task, as the guidelines are flexible around species
choice and design alternatives. The original ICH
S6 guidance did not provide specific direction
on which test species was acceptable, on study
design or on the number of animals per dose
group. The ICH S6 addendum incorporated
recently into the main guidance14 recommends
developmental toxicity studies should not be
conducted in non-human primates unless they
are the only relevant species. Regulators prefer
testing the clinical candidate (in nonhuman
primates if that is the only relevant model), but
alternatives to the drug molecule can be consid-
ered if scientifically justified.
The addendum further says if the mech-
anism of action, class effects, or information
on genetic mutants indicates an adverse effect
on fertility or pregnancy outcome, it may be
sufficient to communicate the potential devel-
opmental or reproductive risk without further
preclinical studies.
For monoclonal antibody drug candidates
active only in nonhuman primates, it is recom-
mended that combined embryo-fetal postpartum
developmental (EFPPD) studies be considered
rather than separate embryo-fetal and perinatal/
postnatal toxicity studies. It is further recom-
mended the dosing interval cover gestation days
(GD) 20–100 (rather than GD 20–50) because
of the prolonged circulating half-lives of most
intact human monoclonal antibodies in primates.
For fertility testing, a standalone nonhuman pri-
mate study is not necessary, but there should be
an assessment of reproductive organs in standard
toxicity studies with the candidate biological.
If there are special concerns, parameters should
be added to provide more detail such as gonad
weights, sperm viability/motility, and reproduc-
tive hormones. Sponsors should realize regulators
recognize these studies are for hazard identifica-
tion rather than true risk assessment due to the
relatively low number of study animals assigned
to each treatment or dose group.
Dose Levels and Duration
One of the most important aspects of preclinical
study design is the choice of dose levels. Much
consideration is given to selecting treatment
doses that will yield the maximum pharmacolog-
ical effect or provide a tenfold exposure multiple
over the maximum estimated exposure in the
clinic, as well as a pharmacologically active dose.
Again, the more extensive the early pharmaco-
logical profiling studies have been in terms of
pharmacokinetic or ADME characterization,
the greater the confidence level for projecting
initial GLP toxicity study dose levels. Factors
that strongly influence the dose range selected
for a preclinical study include the study duration,
gender differences and inherent ADME proper-
ties, including those that are species-specific. For
example, the drug’s in vivo Cmax and its halflife
are properties determining the dose frequency
and the potential toxicity profile in the study.
For many biologicals, the clearance of the drug
is dependent on receptor-binding mechanisms
rather than kidney filtration mechanisms. If a
justification is provided, the maximum dose can
be predicated on data calculations showing that
if all the receptors are occupied at a given dose,
increasing the dose beyond this maximum level
is not informative, according to the ICH S6
guideline addendum.14 For chronic preclinical
toxicology studies, six months is considered a
sufficient length of time to evaluate the haz-
ards associated with chronic, repeat dosing. The
addendum to the ICH S6 guidance recom-
mends an evaluation of recovery (from toxicity)
be included with a statement that recovery is
not intended to detect delayed toxicity. This
topic presumably refers to the characteristically
long circulating halflives of many monoclonal
antibodies, which may take six to nine months
to clear completely from an animal given the
very high doses administered. Operationally,
this referral to delayed toxicity in the guidance