Dosing considerations for medicines

The physiological changes of pregnancy can significantly affect plasma drug concentrations in pregnant women compared to non-pregnant women receiving the same dose of the drug. Determining the safe and effective dose for a medicine in pregnancy requires an understanding of the pharmacokinetics (PK) and pharmacodynamics (PD) of a drug in pregnancy and how they differ from non-pregnant individuals. PK studies aim to establish how a drug is absorbed, distributed and metabolized in, and excreted from, the body. PD studies measure the pharmacological effect that a medicine has on the body.

A PK/PD study aims to inform dosing regimen selection by establishing the quantitative relationship between different doses of a drug and their resulting pharmacological response and effect on the body. If strong data exists on the relationship between the PK and the PD of a particular dose, and there is scientific confidence that the PK/PD relationship holds true during pregnancy and the breastfeeding period, then scientifically there likely can be a heavy reliance on PK as a surrogate for efficacy data. This is noted when examining therapeutics for infectious diseases.

In reality, however, without clinical testing directly in pregnant and lactating populations, regulators are hesitant to include authorization for use of medications during pregnancy and the breastfeeding period within their labels. The information contained in regulatory labels differs from country to country. Ideally, therefore, all medications warrant further PK/PD studies to quantitatively characterize their relationship in pregnancy and during the breastfeeding period. If a PK/PD study is conducted, it must adequately represent the populations where the drug will be used, including in age, body weight, diet, and ethnicity.

If the drug examined is intended to treat chronic medical conditions, an intra-individual comparison is the ideal approach where the PK of the drug is assessed in the same women before, during, and after pregnancy and during the breastfeeding period. If intra-individual comparisons are not feasible, comparison with PK data from non-pregnant women is a reasonable alternative, although such information can be difficult to obtain because male and female data are often presented together in publications and regulatory reviews.

It is not possible to develop a standardized sampling strategy for all drugs. Any strategy should, however, be informed by prior knowledge of the disposition of the specific drug being studied. Measuring the concentration of a drug sample in pregnancy and postpartum women can be difficult to sample and the following issues need to be considered.

The tissue required, including placenta, fetal tissue, cord blood collection or breast milk, may be hard to extract for analysis, although there have been recent advances to support at-home sample collection of tissue, including micro-sampling (e.g., dried blood spots and other volumetric absorptive micro-sampling techniques). Emerging non-animal methods exist (e.g., placenta on a chip) and can be used to supplement traditional studies.
Bioanalytical methods are needed to accurately detect and quantify drug concentrations in plasma (and breast milk) within the required range. Methods used for non-pregnant participants therefore need to be reviewed and validated for quantification of analytes from pregnant and lactating women to ensure they remain fit for purpose. For example, oxytocin clearance is significantly increased during the third trimester and at the time of delivery. A recent study indicated that to effectively characterize the PK of clinical doses of oxytocin immediately post-delivery, a bioanalytical method with a 20-fold increase in sensitivity to that used in studies of the non-pregnant population would be required. This level of precision helps ensure safe and effective dosing for mothers during and after childbirth.
Sampling strategies should minimize discomfort and disruption to postnatal bonding between the mother and her baby and ensure that they respect the physical and emotional needs of pregnant and breastfeeding participants.

PK parameters of the parent drug and its active metabolites (where applicable) should be estimated from total and unbound plasma concentration data and based on a deep understanding of the determinants of a drug’s distribution and elimination processes. For highly protein‐bound drugs, changes in protein binding during pregnancy should also be factored into the determination of target exposures, as total drug concentrations may be affected in pregnancy, whereas unbound drug concentrations may remain unchanged. If pharmacogenomics (how genes affect the response of the body to a drug) play a substantial role in drug exposure, this should be taken into account when deciding the acceptability of decreased drug concentrations in pregnancy. This is important as pregnancy may modify this effect.

PK results must be presented in an accessible language that is clear for ethics committees, regulators, and participants, using lay summaries or visuals to enhance understanding.

Alternative approaches to traditional PK/PD studies exist whereby data can be modelled to inform dosing regimen selection. These approaches ensure that maximum value is derived from limited data sets. Methods include Population PK modelling, PK/PD modelling, and physiologically based PK modelling. Many of these techniques are accessible and applicable in low- and middle-income countries. Specialist expertise is recommended in the selection and use of the method.

SPECIAL CONSIDERATIONS: LACTATION

When a woman takes a medicine during the breastfeeding period, the child may also be exposed to the medicine through breastmilk. Some risks (such as teratogenicity) no longer exist, but there are other situations that pose increased risk to both the mother and child.

There is little reason to anticipate that the efficacy of a medication differs between a lactating mother and a woman who is not lactating. After the first 4–6 weeks postpartum, a breastfeeding mother’s pharmacokinetics are also not considered to differ substantially from a non-pregnant, non-postpartum woman. However, it is important to know whether, and how much of, a medicine passes into breast milk, what the exposure of the child is to this medicine and any metabolites, and what the effects of this exposure may be.

By assessing the milk/plasma ratio of a medicine, we can predict how much an infant will ingest. From this information, the relative infant dose (the dose the infant gets relative to the dose of the woman) can be calculated.

For a drug to reach the breast milk, it must reach a concentration in the mother’s blood and then transfer (by active transport or passive diffusion) into the breast milk. The infant receives a ‘dose’ through breastfeeding, and then absorption is necessary to reach an infant maximum concentration prior to metabolism and elimination by the infant. The maximum concentrations in breast milk are often reached later than in plasma, and the elimination from the breast milk is often slower.

When adapting a plasma assay for a new matrix, such as breast milk, it is necessary to perform a partial validation using the same principles. It is essential that standards and quality controls are diluted in the same matrix as the test samples. As for plasma bioanalysis, such standards and controls are injected into blank plasma. Thus, when establishing the breast milk assay, it is important to obtain breast milk from donors who are not taking any medication. This can be done through collaboration with a donor milk bank where donors have consented for any surplus sample to be used in future research. Alternatively, an ethically approved consent form for breast milk donors among students and staff in a research institute may be used, drawn from similar approaches for the collection of donor blood.

Key resources

FDA U.S. Department of Health and Human Services. Food and Drug Administration (FDA). Bioanalytical method validation guidance for industry