OTHER DESIGN
CONSIDERATIONS
A. Study Participants
Study participants
should be representative of a typical patient population for the drug to be
studied including race, ethnicity, and trimester of pregnancy. Factors with
significant potential to affect the PK of a drug to be studied (e.g., age,
weight, diet, smoking, concomitant medications, ethnicity, renal function, other
medical conditions) can be considered depending on the pharmacologic properties
of the drug. The FDA recommends that uniform diagnostic measures be applied to
all pregnant women to ensure similarity of diagnosis for the treatment being
given and to reduce disease-specific variability in PK. The FDA recommends that
measures used for dating the pregnancy be stated clearly in the study protocol
and consistently applied throughout the study. Inclusion and exclusion criteria
can be tailored to the study.
For drugs that are
metabolized by enzymes known to exhibit genetic polymorphism (e.g., CYP2D6 or
CYP2C19), the FDA recommends that the investigator consider the metabolic status
of the enrolled subjects when analyzing the results of the study. Genotype has
been shown to have an effect on pregnancy-related changes in metabolism (Wadelius
1997).
B.
Postpartum Assessments
Physiology changes
rapidly at delivery but can take from weeks to months to return to the
pre-pregnancy state. The Agency recommends that drugs used only during the
peripartum period (e.g., labor and delivery) be studied only at that time. In
the peripartum period, PK and receptor sensitivity related to PD can change, so
PK/PD studies for drugs used in the peripartum period are important.
A woman’s own
postpartum PK/PD assessments can serve as a control or comparator for the
pregnant state. For women to whom drugs are administered chronically and for
whom a pregnancy on the medication of interest is planned, the pre-pregnancy PK/PD
assessment can serve as the comparison. For drugs used throughout pregnancy and
the postpartum period, PK studies can be performed during the postpartum period
to serve as the comparator or control group. Postpartum assessments can
potentially be done longitudinally (e.g., at 2, 4, 6, and 8 weeks postpartum) to
determine the time course for PK changes to return to the nonpregnant state.
Some pregnancy-related medical conditions rapidly improve after delivery such
that pharmacologic therapy is no longer needed in the postpartum period (e.g.,
some cases of pregnancy-induced hypertension or gestational diabetes). In this
scenario women can participate in a single-dose PK/PD study in the postpartum
period. If a drug possesses linear kinetics, the single-dose PK data can be
extrapolated to the multiple-dose steady state kinetics and then compared with
steady state kinetics obtained during pregnancy when the drug was administered
chronically.
If subjects are
breast-feeding during the postpartum portion of the study, the FDA recommends
that the study incorporate appropriate safety precautions concerning drug
excretion into breast milk and the effects of the drug on the breast-fed
infant. The study design should take into account data concerning the pediatric
pharmacology and adverse effects of the drug. A lactation study might be
performed in conjunction with postpartum sampling.
C. Sample Size
The objective and design of a study are determining factors
in deciding adequate sample size. The number of subjects enrolled in a study
should be sufficient to detect PK differences large enough to warrant dosage
adjustments. Sample size considerations include PK and PD variability for the
drug being studied, the study design (i.e., single-dose versus multiple-dose),
and the physiologic changes inherent in pregnancy. For a population PK
approach
sparse sampling with a larger number of subjects that span the gestational time
periods of interest is encouraged.
As a practical matter, it is prudent that the final number
of subjects enrolled be in excess of that originally determined by standard
sample size calculations to take into account withdrawal of subjects from the
study. Even if data for a subject are missing for one trimester, the Agency
suggests that the subject be retained in the study for the postpartum
assessments.
D. Drug Administration
In single-dose studies, the same dose can
usually be administered to all women in the study. Lower or less frequent doses
can be considered to minimize fetal risk in pregnant women who volunteer to take
the medication for study purposes, even if it is expected to pose minimal risk
at standard doses. The dosage regimen can be adjusted based on the best
available pre-study estimates of the PK of the drug and its active metabolites
and what is known about drug elimination. A concentration-controlled study
design or a dosage adjustment based on the patient’s response are alternative
methods to consider. For example, the study might be conducted to achieve a
specific target concentration using therapeutic drug monitoring procedures.
When studying pregnant patients who need the study drug, the dose can be
modified, either increased or decreased as pregnancy progresses, to achieve the
appropriate response (e.g., lowering of blood pressure, or to decrease adverse
events such as hypotensive episodes with antihypertensive therapy).
E. Sample Collection
and Analysis
The Agency recommends that plasma or whole
blood samples and urine samples
be analyzed for the parent drug and any metabolites with known or
suspected activity, therapeutic or adverse. It is recommended that the
frequency and duration of plasma sampling and urine collection be sufficient to
estimate accurately the relevant PK parameters for the parent drug and its
active metabolites (see Section VI, Data Analysis).
Plasma protein binding, like renal function,
is often altered in pregnancy.
For example, albumin and alpha-1-acid glycoprotein levels are reduced in
pregnancy, consequently the protein binding of drugs can be affected. With
systemically active drugs and metabolites, the unbound concentrations are
generally believed to determine the rate and extent of delivery to the sites of
action. For drugs and metabolites with a relatively low extent of plasma
protein binding (e.g., the extent of binding is less than 80 percent),
alterations in binding due to pregnancy are small in relative terms. In such
cases, description and analysis of the PK in terms of total concentrations would
be sufficient. For drugs where the extent of protein binding is greater than 80
percent, primarily to albumin, it is recommended that the PK be described and
analyzed with respect to the unbound concentrations of the drug and active
metabolites. Although unbound concentrations should be measured in each plasma
sample, if the binding is concentration-independent and unaffected by
metabolites or other time-varying factors, the fraction unbound can be
determined using a limited number of samples or even a single sample from each
patient during each trimester. The unbound concentration in each sample should
then be estimated by multiplying the total concentration by the fraction unbound
for the individual patient.
F. Studies with No Intended Therapeutic
Benefit
It is possible to study drugs that have no intended direct
therapeutic benefit to the pregnant woman provided that the risk to the fetus is
minimal (45 CFR 46). For example, probe substrates can be used to investigate
drug metabolism (e.g., cytochrome P-450 activity) or drug transporter status
(e.g., p-glycoprotein). Data from these studies offer generalizable information
to other pregnant women but do not offer direct therapeutic benefit to study
participants. The Agency encourages sponsors or investigators to explore
additional safeguards for human subject protection for this type of study. To
minimize exposure to a nontherapeutic drug, each pregnant woman can be exposed
to the drug once during pregnancy and in the postpartum period employing a
nonlongitudinal design (e.g., one cohort of women sampled in second trimester
and postpartum and another cohort of women sampled in third trimester and
postpartum). Examples of additional safeguards include administering only
products with a long or known record of safety in pregnancy, administering
products using only a single dose of the drug, using lower doses of the drug,
decreasing the number of drugs (probe substrates) used in any study subject, and
limiting study participants to pregnant women only in second or third
trimester.
G. Pharmacodynamic Assessments
PK studies are usually enhanced by including PD
assessments as part of the study. The Agency encourages sponsors to discuss the
selection of the PD endpoints with the appropriate FDA review staff. Endpoints
would be based on the pharmacological characteristics of the drug and
metabolites (e.g., the behavior of other drugs in the same pharmacological
class), and include consideration of relevant biomarkers.
|
