Simulating transplacental pharmacokinetics — how maternal medications reach the fetal compartment, and how molecular properties, gestational timing, and risk classification shape clinical decisions
For decades the placenta was popularly described as a "barrier" implying protection from everything external. Physiologically this is misleading. The placenta is a metabolically active, selectively permeable interface between maternal and fetal blood — and the overwhelming majority of drugs administered to a pregnant patient reach the fetal circulation in some measurable concentration. Complete exclusion is the exception, not the rule.
The placental barrier separating maternal and fetal blood is composed of a small number of cellular layers whose thickness and composition change across gestation:
• Syncytiotrophoblast — a single continuous, multinucleated layer bathed directly in maternal blood; the primary site of exchange and active transport • Cytotrophoblast — present as a discontinuous layer beneath the syncytium early in pregnancy, largely disappearing by term • Villous stroma and fetal capillary endothelium — the final layers before fetal blood
Early in pregnancy this composite barrier is four cell layers thick; by term it has thinned to essentially two (syncytiotrophoblast plus fetal capillary endothelium), and the exchange surface area has expanded enormously. Both changes increase the rate of passive diffusion as pregnancy progresses — meaning, all else equal, a drug taken near term often reaches the fetal compartment more readily than the same drug taken in the first trimester.
Most drugs cross the placenta by simple (passive) diffusion, driven by the concentration gradient between maternal and fetal blood — no energy or carrier protein required. Transfer rate is described conceptually by Fick's law of diffusion: proportional to the concentration gradient and the exchange surface area, and inversely proportional to membrane thickness.
Beyond passive diffusion, several other mechanisms contribute for specific substances:
• Facilitated diffusion — carrier-mediated, down a concentration gradient, for molecules like glucose • Active transport — energy-dependent, can move substances against a gradient (relevant for some nutrients and, notably, efflux transporters that pump certain drugs back toward maternal blood) • Pinocytosis — bulk engulfment, relevant mainly for larger molecules such as immunoglobulins
Because diffusion never truly reaches zero for small molecules, the practical clinical assumption for most systemically absorbed maternal medications is that some fetal exposure occurs — the open question is usually one of degree and timing, not presence or absence.
Clinically, "does this drug cross the placenta?" is rarely the most useful question — nearly all small-molecule drugs cross to some extent. The more actionable questions are how much reaches the fetus, at what point in gestation, and what that exposure plausibly means for development.
Not all drugs cross the placenta at the same rate. Three physicochemical properties dominate the prediction of transfer: molecular weight, lipid solubility, and the fraction of drug bound to plasma proteins. Smaller, more lipophilic, less protein-bound molecules move across the syncytiotrophoblast membrane more readily — a pattern that echoes how drugs cross other lipid membrane barriers throughout the body, including the blood-brain barrier.
Molecular weight — the syncytiotrophoblast behaves like a lipid membrane with limited aqueous pores. Small molecules (roughly under 500–600 Da) diffuse readily; very large molecules (heparins, insulin, most monoclonal antibodies outside of active IgG transport, and other biologics) cross poorly or not at all by passive diffusion.
Lipid solubility — because the barrier is fundamentally a lipid bilayer, lipophilic (fat-soluble) drugs partition into and across the membrane more efficiently than hydrophilic (water-soluble) drugs of similar size. This is the same principle that governs central nervous system penetration.
Plasma protein binding — only the free (unbound) fraction of a drug is available to diffuse across the membrane. Highly protein-bound drugs (bound to albumin or alpha-1-acid glycoprotein) have a smaller free fraction and, all else equal, transfer more slowly — though pregnancy itself alters maternal protein binding, complicating simple predictions.
The syncytiotrophoblast expresses efflux transporter proteins — including P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP) — embedded in the membrane facing maternal blood. These transporters actively pump certain substrates back toward the maternal circulation, reducing net fetal exposure below what passive diffusion alone would predict.
This matters clinically because it means molecular size and lipophilicity are necessary but not sufficient predictors: a small, lipophilic molecule that also happens to be a strong efflux transporter substrate may still show lower-than-expected fetal transfer. Conversely, transporter inhibition (by other co-administered drugs) can increase fetal exposure to a substrate drug beyond its baseline diffusion-predicted level.
A simple mental model: think of transfer likelihood as size and lipophilicity setting the "speed limit," protein binding setting how much drug is "on the road" to begin with, and active transporters acting as selective checkpoints that can pull certain molecules back before they cross.
Identical drug exposure can carry very different implications depending on when in pregnancy it occurs. Early gestation, during organogenesis, is the period of peak sensitivity to structural teratogenic effects, because organ systems are actively forming. Later in pregnancy, when organs are largely formed, the same or related exposures more typically influence fetal growth, organ maturation, or function rather than gross structural anomalies.
Organogenesis — the period during which the major organ systems form — spans roughly the third through eighth week post-conception. Within this window, different organ systems have staggered periods of peak sensitivity: the neural tube closes very early (around week four), while the heart, limbs, and palate each have their own critical windows later in the first trimester.
A key clinical complication: this highest-risk window frequently occurs before a pregnancy is clinically recognized, since a missed period is typically noticed around week four to five. This is part of why preconception counseling and cautious prescribing in women of reproductive potential are emphasized — by the time pregnancy is confirmed, some of the most sensitive developmental windows may have already passed or be actively underway.
Once major organ structures are established, continued drug exposure shifts the risk profile rather than eliminating it. Later-gestation concerns more typically include:
• Fetal growth restriction from drugs affecting placental blood flow or fetal metabolism • Effects on organ maturation still in progress — for example the fetal central nervous system continues developing throughout gestation and postnatally • Direct pharmacologic effects on the fetus near term (for example, drugs affecting fetal cardiovascular tone or causing neonatal withdrawal/adaptation syndromes after birth) • Effects on labor physiology or uterine tone from drugs administered late in pregnancy
Compounding this, placental transfer efficiency itself generally increases across gestation as the syncytiotrophoblast thins and the exchange surface area expands — so a drug that transferred modestly in the first trimester may reach the fetal compartment more efficiently by the third.
Gestational age is not a simple "safer later" gradient — it reframes what the relevant risk even is. Early exposure risk is often framed in terms of structural malformation; later exposure risk is more often framed in terms of growth, functional maturation, and peripartum effects.
Because individually assessing every drug from first principles is impractical, structured risk classification frameworks have been developed to summarize known or theoretical fetal risk. The most widely recognized historical system in the United States was the FDA's lettered pregnancy category system (A, B, C, D, X), which was retired in 2015 in favor of a narrative labeling format intended to convey nuance the letters could not.
The original FDA pregnancy categories assigned each drug a single letter — A, B, C, D, or X — intended to summarize fetal risk from a mix of animal and human data. In practice the system drew persistent criticism:
• The letters implied a false sense of graded, linear risk ("C is worse than B") when the underlying evidence quality and risk type varied enormously within each category • A single letter could not convey how the risk changed across trimesters, or distinguish theoretical/animal-only risk from confirmed human risk • The letters said little about lactation risk or reproductive-potential considerations, which are clinically distinct questions • Category X sometimes conflated genuine teratogenic risk with simply lacking any indication that would justify use in pregnancy
Recognizing these limitations, the FDA phased out the letter system for prescription drugs and biologics beginning in 2015 under the Pregnancy and Lactation Labeling Rule (PLLR).
Under PLLR, prescription drug labeling now includes three narrative subsections instead of a single letter:
• Pregnancy — a summary of risk (including any pregnancy exposure registry), plus discussion of clinical considerations and available data, organized to distinguish human data from animal data • Lactation — risk to a breastfeeding infant and effects on milk production • Females and Males of Reproductive Potential — information relevant to pregnancy testing, contraception, and fertility, when applicable
This narrative approach conveys considerably more nuance than a single letter, but it also demands more of the reader — there is no longer a quick-glance shorthand. Other complementary resources (teratogen information services, clinical pharmacology references, and pregnancy exposure registries) remain important supplements, since even narrative labeling reflects a snapshot of evidence that may be incomplete, especially for newer medications.
No classification system — lettered or narrative — substitutes for individualized clinical assessment. Classification frameworks are a starting point for gathering evidence, not a final answer about whether a specific medication is appropriate for a specific patient at a specific point in pregnancy.
A common but incomplete framing of medication use in pregnancy treats "avoid the drug" as the automatically safer choice. In reality, prescribing decisions during pregnancy require weighing the risk of fetal drug exposure against the risk of leaving the maternal condition untreated or undertreated — and the latter is not risk-free for the pregnancy.
It is tempting to treat "no medication" as the risk-free default option and any drug exposure as an added risk on top of it. This framing breaks down for many maternal conditions, where poor disease control itself carries documented risk to the pregnancy:
• Poorly controlled epilepsy — seizures themselves can cause fetal hypoxia and trauma, independent of any medication • Untreated severe depression or anxiety — associated with poor prenatal care engagement, substance use, preterm birth, and impaired maternal-infant bonding • Uncontrolled hypertension or diabetes — associated with preeclampsia, growth restriction, stillbirth, and other serious complications • Untreated maternal infection — can itself cross to the fetus or trigger harmful inflammatory responses
In each case, the clinically relevant comparison is not "drug exposure vs. zero risk" but "drug exposure vs. the risk of the untreated or undertreated condition" — and for many conditions, that untreated risk is substantial.
A practical approach to weighing maternal treatment need against fetal exposure risk typically considers several dimensions together rather than any single factor in isolation:
• Severity and trajectory of the maternal condition if left untreated or undertreated • Strength and quality of available evidence on the specific drug's fetal effects, and whether that evidence is trimester-specific • Availability of alternative therapies with a more established safety profile, and whether they are equally effective for this patient • Gestational timing of the anticipated exposure relative to organogenesis and later developmental windows • Dose, duration, and route — a short course or lowest effective dose may carry materially different risk than prolonged high-dose therapy • The patient's own values and preferences, incorporated through shared decision-making
This is rarely a one-time decision — it is revisited as pregnancy progresses, as new data emerges, and as the maternal condition evolves, ideally with input from obstetrics, the prescribing specialist, and clinical pharmacy.
The goal of pregnancy pharmacotherapy is not zero fetal exposure — it is the treatment strategy that minimizes overall risk to both mother and fetus, which sometimes means treating and sometimes means withholding or switching therapy. Neither default is correct in every case.