Drugs and teratogens in pregnancy

• Almost every pregnant woman needs some medicine, and the question is never "is this drug safe in pregnancy?" but "what happens if she takes it, and what happens if she does not?"^[10]
• The old FDA letter categories A to X were removed; modern labels now carry a narrative risk summary, clinical considerations and the underlying data instead.^[9]
• A handful of medicines are real human teratogens with predictable harm: valproate, isotretinoin, thalidomide, warfarin, and (in the second and third trimesters) angiotensin converting enzyme inhibitors.^{[1][2][3][4][5]}
• For common conditions there are well-established safer choices, such as labetalol for hypertension in pregnancy and inhaled corticosteroids for asthma, and expert services such as UKTIS exist to review individual cases.^[4][12]
• Alcohol is a human teratogen; no safe level in pregnancy has been defined, so UK guidance recommends complete abstinence, and smoking is the commonest reversible cause of growth restriction and stillbirth in pregnancy.^[6][7][8]
• Look up an exposure in the UKTIS monographs (or national equivalent) and, for an individual woman, arrange a maternal medicine or teratology referral when a high-risk drug has been taken.^[7]

By the end of this page you should be able to:

1. Explain why prescribing in pregnancy is a risk-benefit decision rather than a safety-versus-danger decision.
2. State what replaced the FDA pregnancy letter categories and what a modern label now contains.
3. Describe the dose, gestational timing and characteristic defects for valproate, isotretinoin, warfarin, ACE inhibitors and thalidomide.
4. Plan a medication review for a woman on long-term treatment who is planning a pregnancy or has just discovered she is pregnant.
5. Choose the commonly recommended first-line drug for hypertension and for asthma in pregnancy and explain why.
6. Counsel a woman about alcohol and tobacco in pregnancy, including nicotine replacement and electronic cigarettes, using current UK and US guidance.

A teratogen is any agent that crosses the placenta and can disturb normal development of the embryo or fetus, producing structural malformation, growth failure, neurodevelopmental impairment or other harm. Drugs, chemicals, infections and radiation can all be teratogens. Not every exposure causes harm: the outcome depends on the drug, the dose, the gestational age at exposure, the mother's handling of the drug and other risk factors in the pregnancy. Most prescribed drugs are not teratogens, and the aim of care is usually to keep the mother well rather than to stop every medicine. Stopping a needed treatment can be more dangerous than continuing it.

A medicine reaches the fetus in stages. Up to about week 10 of gestation the pregnancy is maintained by yolk-sac nutrition, and substances reach the embryo by active transport and diffusion. From about weeks 10 to 12 the placenta is functional and actively transports many substances to the fetal circulation. Once the fetal kidneys and liver begin to work, the fetus can metabolise some drugs, but usually more slowly than the mother. Alcohol levels in the fetal compartment can rise above maternal levels because fetal enzyme activity is lower and elimination is slower, and exposure is compounded by swallowing of excreted compounds in the amniotic fluid.^[6]

Passive diffusion across the placenta favours small, lipophilic, non-ionised molecules with low protein binding. Large molecules (for example insulin and heparin) cross poorly, which is why they are favoured when a medicine is needed in pregnancy. Even so, most small drug molecules do cross to some degree, so absence from the placenta cannot be assumed.

Timing matters as much as transfer. The first trimester is the period of organogenesis and is the window in which structural malformations are most likely after drug exposure. Later exposure more often affects growth, brain development, organ maturation and the neonate's adaptation after birth. ^[11]

Previously the United States Food and Drug Administration (FDA) used five pregnancy categories, A, B, C, D and X, which placed a drug into a single letter based on a mixture of animal and human data. The letter system was easy to misuse: a category C or D could be read as "dangerous", and a survey cited by the Teratology Society found that the system contributed to unnecessary terminations of wanted pregnancies. Under the Pregnancy and Lactation Labeling Rule (PLLR), newer prescription drug labels no longer use letters. Instead they contain a narrative risk summary, clinical considerations (including dosing in pregnancy and whether to adjust around delivery) and a data section describing the animal and human evidence. Labels may be updated when new information appears, and a drug may register a pregnancy exposure registry when one exists.^[9]

Two practical points follow. First, over-the-counter (OTC) medicines in the United States were not affected by the change and keep their older labelling, so a patient leaflet can still show an old letter even when the prescription version no longer does. Second, other countries use different systems. The United Kingdom assesses medicines case by case through UKTIS and the Medicines and Healthcare products Regulatory Agency (MHRA), and the Summary of Product Characteristics (SmPC) for a drug still contains its pregnancy advice. Labels differ between regulators because the underlying data and interpretation differ; this page uses the newest guidance it could open for each topic and states where sources disagree.

Write down the principle you are using, because it is reused for every drug on this page.

• Risk-benefit, not "safe versus unsafe". A decision compares the harm from the drug with the harm from untreated illness, including maternal risk, fetal risk and neonatal risk. Inadequately treated asthma is a useful example: the UKTIS monograph states that consensus expert opinion holds that the risk of inadequately treated asthma in pregnancy outweighs any theoretical risk of asthma medication use.^[12]
• Use the available evidence, and say when it is thin. Information about risks in pregnancy is often limited because clinical trials rarely enrol pregnant women. Much of what is known comes from observational studies, registries, pharmacovigilance and case series. That is why guidelines are cautious in some areas and clear in others.
• Choose the established drug over the new one. When a cheaper, well-studied drug works, use it rather than a newer agent with little pregnancy data. This is a reason to continue a medicine a woman is already on and tolerating, rather than switching for no clinical reason.
• Preconception review is the best time. For a chronic condition, review medication before conception, document the plan, start folate where indicated, and counsel on contraception while a teratogen is still needed. If she discovers she is pregnant while on a teratogen, stop or switch at once after discussion with the specialist, and arrange the appropriate fetal surveillance rather than panicking or stopping everything abruptly.
• Avoid polypharmacy where possible. Teratogenicity may be enhanced by co-administration of several drugs. Check every medicine, including herbal and over-the-counter products. Some patients believe "natural" products are safe in pregnancy; they are not always safe.
• Folate matters for women on antiepileptic drugs. UKTIS advises women who take valproate to take high-dose folic acid (5 mg daily) prior to conception and throughout the first trimester, because valproate impairs the active metabolite of folate. An independent 2026 Australian review notes that folic acid has not been shown to reduce malformations associated with valproate and, if it is used, recommends a dose of at least 0.4 mg daily before conception and during pregnancy; UK practice retains the UKTIS recommendation of 5 mg daily. Other antiepileptics have similar concerns. Women on antiepileptic drugs should also be enrolled in the relevant pregnancy register where one exists.^[1][13]
• Dose from guidance, not memory. Drug doses, intervals and monitoring are those given in current guidelines and formularies. Check your local formulary and specialist advice before prescribing, and remember renal and hepatic dosing, drug interactions and therapeutic drug monitoring.
• Regional practice differs. The care pathways described here are mainly UK-based (UKTIS and NICE), with US (ACOG, FDA) and international input where those sources were available. East African and other low-resource practice is noted in the sections on alcohol, tobacco and low-resource settings.

The table gives the signature, and the sections below give the detail. Every entry is a drug that should generally be avoided or used only under specialist care during pregnancy, because well-documented fetal harm has been established.

Sodium valproate is an antiepileptic also used in mania and, off licence, in migraine prevention. UKTIS calls it an established human teratogen that should be used in women of childbearing potential only when there is no safer effective alternative and after a discussion of the risks and benefits, and states that it should not be used to treat bipolar disorder in pregnancy.^[1]

• Malformation. Valproate monotherapy in the first trimester is associated with an absolute risk of congenital malformation of up to 11%, more than three times above background rates. The malformations include neural tube defects (particularly spina bifida), cardiac defects, orofacial clefts and hypospadias, and some malformation reports extend to skull, limb and digit abnormalities.^[1]
• Dose-response. There is no "safe" threshold dose, but a dose-response relationship is observed. UKTIS, the source used here, identifies 1,000 mg/day as the dose above which malformation risk rises significantly and records malformation rates up to 62.1% at 3,000 mg/day. An independent 2026 Australian review instead reports 3.3% malformations below 1,100 mg/day and 30.2% at 1,100 mg/day or more, and identifies 1,100 mg/day as the dose at which risk increases significantly; the estimates come from different registers and cohorts, so the two figures are not directly comparable. No benefit to divided or modified-release dosing has been demonstrated in the one study that examined it, although slower or divided dosing is advised on theoretical grounds.^[1][13]
• Neurodevelopment. Valproate is a neurobehavioural teratogen affecting offspring IQ and motor skills. UKTIS reports that autism spectrum disorder may be increased up to three-fold against a roughly 1% background incidence, and an independent 2026 Australian review instead reports odds ratios of 2.4 to 4.5 for intellectual disability, 1.4 to 1.8 for attention-deficit hyperactivity disorder and 1.7 to 4.4 for autism spectrum disorder; 19% to 62% of valproate-exposed children in UK data need additional educational support. A 2014 Cochrane review identified poorer cognitive outcome with daily doses in excess of 800 to 1,000 mg. No valproate dose has been shown to be free of neurodevelopmental risk, and estimates vary between sources.^[1][13]
• Other outcomes. High doses above 750 mg/day have been associated with possible increased miscarriage risk, and weak evidence suggests folate may be protective below 1,000 mg/day. Growth and stillbirth data are limited.^[1]
• What to do. Give high-dose folic acid (5 mg daily) before conception and through the first trimester, enrol the woman in a pregnancy register, plan delivery in a unit with neonatal facilities, use the lowest effective dose if valproate is essential, and offer detailed anomaly scans from 16 weeks. Valproate should not be stopped abruptly without specialist advice; for women who do become pregnant the treatment is switched where possible and a management plan is made.^[1]

Oral isotretinoin is a retinoid used for severe acne that has not responded to standard therapy. UKTIS calls it a confirmed teratogen causing major structural and neurodevelopmental abnormalities (isotretinoin embryopathy), states that use during pregnancy is contraindicated, and requires a negative pregnancy test plus confirmation that effective contraception, preferably two forms, is used before, during and for at least one month after the last dose.^[2]

• Defects. These include central nervous system abnormalities (hydrocephalus, cerebellar abnormalities, microcephaly), facial dysmorphism, cleft palate, external ear abnormalities (absence of the external ear, small or absent auditory canals), eye abnormalities (microphthalmia), cardiovascular abnormalities (conotruncal defects such as tetralogy of Fallot, transposition of the great vessels and septal defects), renal, thymus and parathyroid abnormalities.^[2]
• Risk. The absolute risk is unclear. Malformation rates in live-born infants from prospectively reported pregnancies range from 5% to 28%, and risk estimation is hampered by small numbers and by high rates of elective termination. One small study found no statistically significant difference in malformation risk between exposure before two weeks and exposure from two to nine weeks, and data suggest that even exposure of less than one week's duration may carry a high risk. No minimum dose or duration without risk has been specified, and because of the long half-life women should avoid conception for at least one month after stopping treatment. Miscarriage estimates reach up to 30%, with an increased incidence of ectopic pregnancy reported. Neurodevelopmental delay can occur with or without visible structural abnormalities, and the risk period may hypothetically extend throughout pregnancy.^[2]
• Topical. There are inadequate data to assess fetal risk from topical isotretinoin in pregnancy. Its use cannot be recommended, although women inadvertently exposed (including within a month before pregnancy) can be reassured that systemic absorption is generally low and existing studies show no teratogenic risk for topical retinoids as a group.^[2]
• What to do. Stop immediately if pregnancy is suspected, arrange a detailed fetal anomaly scan in every first-trimester exposure, and counsel the woman about the possibility of impaired neurodevelopment without structural abnormalities. Anyone starting oral isotretinoin needs enrolment in a pregnancy prevention programme.^[2]

Warfarin is a vitamin K antagonist used for thromboprophylaxis (mechanical heart valves, atrial fibrillation with rheumatic heart disease, venous thrombosis and pulmonary embolism). UKTIS confirms that warfarin should generally be avoided in pregnancy and replaced with a non-teratogenic anticoagulant, usually low-molecular-weight heparin (LMWH).^[3]

• First trimester. First-trimester exposure can cause fetal warfarin syndrome (FWS): nasal hypoplasia and skeletal abnormalities including short limbs and digits and stippled epiphyses. The critical period is thought to be weeks 6 to 12. Older literature suggested rates up to 30%, while more recent studies place the risk at 6 to 10%, with some evidence of dose-dependence: doses above 5 mg appear to carry higher risk, although lower doses are not free of risk.^[3]
• Later exposure. Exposure extending beyond or starting after the first trimester has been associated with central nervous system and eye anomalies. Second- and third-trimester exposure carries a risk of fetal, placental or neonatal haemorrhage.^[3]
• Other outcomes. Mixed reports of developmental delay are concentrated in children with FWS, and miscarriage may be increased; similar increases have been reported with other anticoagulants such as heparin.^[3]
• What to do. Switch to LMWH before conception when possible, or as soon as pregnancy is suspected. Patients with mechanical heart valves need individual specialist review because LMWH may be less effective for those valves; refer them for pre-conception counselling and contraceptive advice, and if pregnancy occurs refer immediately to a tertiary maternal cardiology centre. Offer detailed ultrasound for FWS anomalies where first-trimester exposure has occurred, plus fetal growth and wellbeing monitoring where warfarin continues.^[3]

ACE inhibitors (captopril, enalapril, fosinopril, imidapril, lisinopril, perindopril, quinapril, ramipril, trandolapril) are used for hypertension, heart failure, nephropathy and cardiovascular prophylaxis. UKTIS describes a well-defined ACE-inhibitor fetopathy after exposure in the second and third trimesters: oligohydramnios, renal tubular dysgenesis, neonatal anuria, hypocalvaria, pulmonary hypoplasia, persistent patent ductus arteriosus, mild to severe intrauterine growth restriction, and fetal or neonatal death. These effects are thought to arise from a direct action on the fetal renin-angiotensin system, which begins to function at about 26 weeks. A small prospective case series suggests risk with exposure beyond 20 weeks, so use in the second and third trimesters is generally contraindicated.^[4]

First-trimester exposure has been linked inconsistently to malformation risk, mostly small increases in cardiovascular and urinary tract anomalies, and much of this may reflect the maternal disease rather than the drug: women needing an ACE inhibitor are more likely to have cardiac or renal disease or diabetes, and similar outcomes have been seen with other antihypertensives. For miscarriage, preterm delivery and growth restriction, UKTIS finds no reliable evidence that ACE inhibitors directly increase risk, because the findings appear confounded by the underlying maternal condition.^[4]

The practical rule is simple. Women planning a pregnancy on an ACE inhibitor should have a medication review, and women who become pregnant on one should be offered a different antihypertensive if that is clinically appropriate. NICE, as cited by UKTIS, recommends labetalol as first-line antihypertensive in pregnancy, with nifedipine or methyldopa as alternatives depending on pre-existing treatment, side effects, risks and the woman's preference. Where an ACE inhibitor is continued for compelling reasons beyond the first trimester, close surveillance by maternal and fetal medicine specialists (often in a tertiary centre) with serial growth and amniotic fluid scans is needed.^[4]

Thalidomide was marketed in 1957 as a sedative and for morning sickness and was withdrawn by 1962 after reports of limb reduction defects. Over 10,000 children were born worldwide with severe malformations including absent, reduced or hypoplastic limbs, dysplasia or absence of internal organs, ear and eye defects, facial palsy and congenital heart defects. The greatest risk is exposure between 20 and 36 days post-fertilisation (36 to 50 days after the last menstrual period), reported after a single 100 mg dose. The overall malformation risk in live-born infants exposed during that period is 20 to 50%, and around 40% of affected children die soon after birth. Later exposure effects are unknown, but thalidomide and its analogues are strictly contraindicated at all stages of pregnancy.^[5]

Lenalidomide and pomalidomide have no human pregnancy safety data, but are predicted to be teratogenic by similarity to thalidomide and from animal studies. All three drugs are licensed for myeloma and similar conditions under strict criteria. To avoid fetal exposure, patients, prescribers and pharmacies must register with the manufacturer, and both women and men must comply with the manufacturer's pregnancy prevention programme. If an early exposure occurs, the extremely high malformation rate means termination of the pregnancy should be discussed, and ongoing pregnancies should have a detailed fetal anomaly scan including echocardiography.^[5]

No drug is "absolutely safe" in pregnancy, and the correct framing is "safest option with best evidence". UKTIS monographs exist for many individual drugs; the entries in this table are the classes generally preferred, and the individual drug entry in UKTIS or your national formulary should be checked before prescribing.

The asthma entry deserves two sentences. Asthma is one of the commonest chronic conditions in pregnant women, and poorly controlled disease is associated with pre-eclampsia, preterm birth, low birth weight and perinatal mortality. UKTIS guidance therefore advises that pregnant women with asthma should generally be managed as for the non-pregnant patient, using the standard pharmacological stepped therapy (inhaled corticosteroid plus salbutamol, and escalation to long-acting beta-2 agonist combinations, montelukast, tiotropium, theophylline, oral prednisolone or biologics), and severe exacerbations in pregnancy require careful management that may include additional monitoring of fetal growth and wellbeing.^[12]

For hypertension, UKTIS reiterates NICE's approach: women taking an ACE inhibitor who are planning a pregnancy should have their medicines reviewed, those hoping to conceive should be switched if clinically appropriate, and labetalol is first line. For thrombosis, UKTIS advises switching warfarin to low-molecular-weight heparin before pregnancy wherever possible, or as soon as pregnancy is suspected. For mood and mental health, stopping a needed psychiatric drug suddenly can be more harmful to mother and baby than careful adjustment; UKTIS has individual monographs for antidepressants (citalopram, escitalopram, fluoxetine, sertraline, mirtazapine, venlafaxine, amitriptyline and others) and antipsychotics which should be consulted rather than applying a blanket rule.

Alcohol is a human teratogen. UK guidance recommends complete abstinence from alcohol in pregnancy because no "safe" drinking level has been established. The UKTIS monograph states that UK government guidelines recommend complete abstinence, and that topical application of alcohol would not be expected to produce systemic absorption. Alcohol reaches the fetus at every stage of pregnancy, and fetal levels can exceed maternal levels because fetal enzyme activity and elimination are slower.^[6]

• Fetal alcohol spectrum disorders (FASD) are a continuum of physical and neurodevelopmental outcomes after prenatal alcohol exposure. The most severe form, fetal alcohol syndrome (FAS), is most often associated with chronic high consumption above 5 units a day and comprises growth retardation, characteristic facial features and central nervous system abnormalities. Other forms (partial FAS, alcohol-related neurodevelopmental disorder, alcohol-related birth defects) are subgroups; alcohol-related neurodevelopmental disorder has been estimated at up to 90% of all FASDs. Whether risk or severity is tied to a particular drinking pattern cannot be established from the current data. Gestational alcohol exposure has also been associated with miscarriage, stillbirth and preterm delivery.^[6]
• Interpersonal variation in FASD risk includes the duration and stage of pregnancy at exposure, maternal health and nutrition, maternal body size, other substance use and the genetically determined detoxification capacity of mother and fetus.^[6]
• Reassurance where appropriate. Occasional binge drinking before the pregnancy was recognised is common, and the limited available data on early pregnancy exposure to occasional binge drinking are not overall suggestive of harm. UKTIS advises that reassurance can be given, while also discussing the limitations of the data (small numbers, limited outcomes studied, possible insensitivity of neurodevelopmental tests).^[6]
• What to do. Advise complete abstinence; do not use abstinence as a way of avoiding the subject. After a confirmed exposure, the intensity of antenatal surveillance depends on the magnitude of risk. There are no definitive postnatal diagnostic tests for FASD; if it is possible, referral to a paediatrician or clinical geneticist is recommended, with consideration of differential diagnoses such as an underlying genetic disorder.

Tobacco smoking in pregnancy is associated with miscarriage, fetal congenital malformation, stillbirth, preterm birth, intrauterine growth restriction and placental abruption, as well as sudden infant death and later offspring outcomes such as neurodevelopmental disorders, asthma, obesity, diabetes and cancer. ACOG advises that pregnant women should be counselled about these significant perinatal risks, including orofacial clefts, fetal growth restriction, placenta previa, preterm prelabor rupture of membranes, low birth weight, increased perinatal mortality, ectopic pregnancy and decreased maternal thyroid function, together with increased respiratory infections, asthma, infantile colic, bone fractures and childhood obesity in the children. About 54% of women who smoke before pregnancy quit directly before or during pregnancy. Smoking cessation at any point in pregnancy benefits the woman and her fetus, with the greatest benefit observed before 15 weeks' gestation.^[7][8]

UKTIS translates this into practical care. Traditional nicotine replacement therapy products (gums, sublingual tablets, inhalators, lozenges, oromucosal or nasal sprays and transdermal patches) are recommended first line in the UK by the National Institute for Health and Care Excellence (NICE) for smoking cessation in pregnancy. Published human pregnancy exposure data for these traditional products do not show robust evidence of increased risk, unlike smoking itself, and exposure at any stage is not usually regarded as a reason for termination or additional fetal monitoring. Electronic cigarettes (vapes) are not recommended as nicotine replacement for pregnant women in the UK because fetal safety data are limited compared with traditional nicotine replacement, but some charities and the National Health Service advise pragmatically that, where traditional products are declined or have failed, vaping during pregnancy, although not risk-free, may be preferable to continued smoking. Women who have smoked heavily during later pregnancy may be offered additional fetal growth monitoring.^[7]

It is useful to sort medicines by what is known rather than by a letter.

• Documented human teratogens, where predictable structural or neurodevelopmental harm is established: valproate, oral isotretinoin, thalidomide and its analogues, warfarin (in the first trimester, and as a fetal/neonatal bleeding risk later) and ACE inhibitors in the second and third trimesters. These are avoided or used only under specialist care.
• Medicines where the risk of untreated disease exceeds the theoretical treatment risk: inhaled corticosteroids and beta-2 agonists for asthma, antiepileptic drugs chosen to minimise teratogenicity, insulin for diabetes, labetalol and similar agents for hypertension, and needed antibiotics or anticoagulation.
• Medicines with extensive experience and no convincing teratogenic signal: paracetamol for pain and fever, low-molecular-weight heparin, thyroid replacement and many short courses of common antibiotics. Absence of evidence is not proof of safety, but practice is guided by decades of data and registries.
• Drugs with little or no pregnancy data: the newest antidepressants, biologics and targeted therapies fall into this group; where possible they are avoided or used within a registry and with shared decision-making.
• Contraception and pregnancy prevention: oral isotretinoin and thalidomide cannot be prescribed without pregnancy prevention programmes; for valproate and some other drugs there are regulatory programmes or clinic-level rules. For all three, the patient, prescriber and (for thalidomide) dispensing pharmacy must be enrolled.

The characteristic defect often tells you the mechanism and the exposure window.

• Retinoids (isotretinoin). These are derivatives of vitamin A and act on retinoid signalling pathways that guide patterning of the craniofacial region, heart, brain and limbs during organogenesis. Disruption in this window produces the craniofacial, conotruncal cardiac, central nervous system, ear and thymic/parathyroid pattern of isotretinoin embryopathy. Because the pathways are active early and the drug has a long half-life, both early brief exposure and later exposure remain a concern. Topical absorption is small, which is why topical products carry a very different risk profile.
• Valproate. It impairs production of the active metabolite of folate and interferes with histone deacetylases and gene expression in the developing nervous system. This helps explain its combination of neural tube defects, structural anomalies and neurodevelopmental impairment (autism spectrum disorder, IQ, motor skills). High-dose folate (5 mg daily) is recommended before conception and in the first trimester even though the evidence that folate specifically prevents valproate-related defects is limited.
• Warfarin. It antagonises vitamin K, which is required for the carboxylation of clotting factors and of bone/cartilage proteins such as osteocalcin and matrix Gla protein. Defective carboxylation of these proteins in the embryonic skeleton, face and central nervous system underlies fetal warfarin syndrome and the later central nervous system and eye anomalies. Bleeding is an additional risk whenever warfarin continues in the second and third trimesters.
• ACE inhibitors. These block angiotensin converting enzyme and hence the renin-angiotensin system that becomes important to fetal renal and vascular development from roughly 20 to 26 weeks. The resulting fetal fetopathy is not primarily a structural malformation from the first trimester but a second- and third-trimester organ injury: oligohydramnios, renal tubular dysgenesis and possible hypocalvaria and pulmonary hypoplasia. First-trimester use has not produced a consistent malformation signal in the sources opened.^[4]
• Thalidomide. Its molecular target in the embryo (cereblon, leading to anti-angiogenic effects on limb bud development) was discovered decades after the epidemic. The extremely narrow critical window (20 to 36 days post-fertilisation, about 36 to 50 days after the last menstrual period) corresponds to early limb bud development, which is why a single dose in that window can produce limb reduction defects. Exposure later has no documented outcome data, but the drug remains contraindicated at all stages.
• Alcohol. Ethanol and its metabolite acetaldehyde are directly embryo- and fetotoxic, and teratogenicity is influenced by timing, dose, and maternal and fetal metabolism. Neuronal migration, facial midline development and growth are particularly affected, producing the FASD spectrum.

Teratogen management is as much about systems as about pharmacology.

• Pregnancy prevention programmes for oral isotretinoin and thalidomide require negative pregnancy tests, documented contraception, time-limited prescriptions, and (for thalidomide) registration of patients, prescribers and pharmacies. Valproate and some antiepileptics have comparable national rules or clinic protocols. The patient has to understand that conception while taking these drugs is a real emergency, not a theoretical risk.
• Folate. High-dose folic acid (5 mg daily) is recommended before conception and through the first trimester for women taking valproate, because valproate impairs folate metabolism. Women on other antiepileptics often receive the same. Local obstetric guidance and national fortification policy (for example, flour fortification in some countries) also influence neural tube defect risk independently of medication.
• Dose minimisation. When a teratogen cannot be substituted (some epilepsy, some psychiatric illness, some autoimmune disease), use the lowest effective dose, document the reasoning and arrange the right fetal surveillance.
• Stopping a needed drug suddenly can cause seizures, relapse of psychiatric illness, rebound hypertension or thrombosis, so the change must be planned with the specialist even after a positive pregnancy test.
• Peripartum planning. For continuing drugs in the third trimester, make a birth plan: neonatal adaptation after centrally acting drugs; vitamin K logistics when warfarin has been used; reversal agents and newborn care for drugs with neonatal toxicity; anaesthetic and analgesic plans; and breastfeeding compatibility reviewed before discharge.

A short list of what to look for after an exposure.

• Valproate: high-dose folic acid; enrolment in a pregnancy register where available; detailed anomaly scan from 16 weeks; delivery in a unit with neonatal care; postnatal assessment of the child's development.
• Isotretinoin: pregnancy test before starting; at least two effective contraceptives before, during and for one month after; detailed anomaly scan in the first trimester for any exposure; counselling about neurodevelopment. No antidote exists; stopping the drug and supportive care are the interventions.
• Warfarin: pre-conception switch to low-molecular-weight heparin wherever possible; for first-trimester exposure, detailed ultrasound for fetal warfarin syndrome (nasal hypoplasia, skeletal abnormalities); for women who must continue warfarin, fetal growth and wellbeing monitoring and a tertiary maternal cardiology plan.
• ACE inhibitors: medication review before conception or at diagnosis of pregnancy; if continued for compelling reasons, maternal and fetal medicine surveillance with serial growth and amniotic fluid scans, often in a tertiary centre.
• Thalidomide/analogues: registration and pregnancy prevention for men and women; if exposure occurs in early pregnancy, discussion of termination; ongoing pregnancies require a detailed anomaly scan including echocardiography.
• Alcohol: complete abstinence counselling; after confirmed exposure, surveillance guided by clinical assessment of the magnitude of risk; postnatal referral to paediatrics or clinical genetics if fetal alcohol spectrum disorder is possible.
• Smoking and nicotine: smoking status and cessation support (nicotine replacement first line in the UK); additional fetal growth monitoring for women who have smoked heavily in late pregnancy.

Good counselling is as important as the prescription.

• "Risk comparison" language. Patients understand proportions better than category letters. Contrast the drug's risk with the background risk of the same outcome, and with the risk of untreated maternal illness.
• Avoid blame. Many exposures occur before pregnancy is recognised. UKTIS explicitly supports reassurance for common early exposures such as occasional binge drinking, while being honest about data limits.
• Consistency across professionals. Contraindications, monitoring plans and contraception advice should match between general practice, pharmacy, obstetrics and the specialist. A shared written plan helps.
• Documentation. Record the exposure, the gestational age, the decision, the alternatives discussed and the monitoring arranged. Pharmacovigilance and pregnancy registers improve knowledge for future patients.
• Contraception. For women needing isotretinoin, thalidomide or valproate, contraception advice must be specific, and often needs two methods; follow-up at each prescription.
• Mental health. Do not present a stark choice between maternal treatment and fetal safety when switching a drug is possible; involve perinatal mental health and obstetric teams.

Prescribing in lactation is a separate risk-benefit question from pregnancy, and many drugs that are avoided in pregnancy are compatible with breastfeeding.

• Most small-molecule medicines transfer into breast milk in small amounts; the clinically important questions are the infant dose received, the infant's ability to metabolise it, and whether the effect (for example drowsiness or kidney handling) is clinically relevant.
• Heparin and insulin do not enter milk in significant amounts and are considered compatible; warfarin is also compatible with breastfeeding, although close neonatal monitoring is required if the mother was anticoagulated around delivery.
• Antiepileptic, antidepressant and antipsychotic use in lactation is individualised: lamotrigine, sertraline and several others are commonly continued, and monitoring of the infant is prudent. UKTIS monographs often include breastfeeding sections, and specialist pharmacy/pharmacology services can advise.
• Alcohol and smoking are also relevant to lactation: alcohol passes into breast milk and its metabolism in the infant is slow; nicotine passes into milk and smoking reduces milk and affects infant health. Practical advice (delayed feeds, limits) should follow national guidance.
• Teratogens do not automatically mean "no breastfeeding". Isotretinoin is usually avoided in breastfeeding; valproate is compatible with appropriate infant monitoring. A medicine is usually scored separately for pregnancy and for lactation, so never conflate the two when counselling.
• When uncertain, use the lactation pharmacology resources available locally (specialist pharmacy, or regional medicines information) rather than stopping breastfeeding unnecessarily.

The guidance written in this page comes from the newest editions of the references opened directly (listed in the front matter and cited throughout). Three services are worth knowing by name, because a medical student will call them repeatedly.

• UKTIS (the United Kingdom Teratology Information Service, commissioned by UKHSA and hosted by the Newcastle upon Tyne Hospitals NHS Foundation Trust) provides evidence-based monographs on medicines, vaccines, chemicals and radiological exposures in pregnancy, a public-facing leaflet service at medicinesinpregnancy.org, surveillance, and a telephone advisory service for healthcare professionals. It is the source of the monographs for all the drugs and exposures discussed above. Regional equivalents exist (for example MotherToBaby in the United States).^[7]
• ACOG (the American College of Obstetricians and Gynecologists) is the source used here for tobacco cessation risks in pregnancy; its Committee Opinion No. 807 (2020, reaffirmed 2026) was opened for this page.^[8]
• StatPearls open-access study guides (2026 for pregnancy medications, 2024 for teratogenic medications) were used to confirm the removal of the letter categories by the PLLR and the first-trimester predominance of malformation risk. ^[11]
• FDA pregnancy and lactation labelling pages explain the PLLR process that replaced the A-X categories.

• Preconception counselling is the single most effective intervention. For any woman on a teratogen, plan the pregnancy, review the drug, discuss contraception and document the agreed plan.
• If a woman discovers she is pregnant on a teratogen, stop or switch the drug where there is a safer alternative, and arrange fetal surveillance; do not delay but do not panic. For valproate, UKTIS specifies anomaly scans from 16 weeks; for thalidomide exposure, a detailed anomaly scan including echocardiography; for warfarin first-trimester exposure, scans for fetal warfarin syndrome; and for ACE inhibitors continued beyond the first trimester, serial growth and amniotic fluid monitoring.
• Men and teratogens. Thalidomide requires registration of men as well as women. Valproate has MHRA advice that men should avoid fathering a pregnancy during treatment and for three months afterwards, although UKTIS notes that the signal of offspring neurodevelopmental impairment in children of exposed fathers has not been replicated and advises a pragmatic case-by-case review.
• Multiple exposures and synergistic effects. Isotretinoin toxicity may be modulated by dose, long or multiple courses, prolonged treatment and individual metabolism. Individual case assessment always remains necessary.
• Paternal exposures contribute to risk assessment and planning in some cases; see the UKTIS monographs for specific notes.
• What not to do. Do not stop all medication in pregnancy out of fear; do not use a category letter rather than the evidence; do not rely on a search snippet for a dose.

• Use the best available national guidance. In the UK National Health Service that means UKTIS monographs and NICE; in other settings the national formulary, the WHO Model Formulary and specialist services all have roles, and where possible an exposure should be discussed with a teratology service.
• Where imaging for anomaly detection is limited, the history of exposure and the gestational age at exposure become even more important; counselling should be honest about what scans can and cannot see.
• For hypertension, labetalol and nifedipine are usually more accessible than specialist agents; for smoking, behavioural support plus whatever nicotine replacement is available is the appropriate offer.
• When a high-risk drug is essential (for example valproate where no alternative controls seizures), document the decision, use the lowest effective dose, give folate where indicated and arrange whatever monitoring the system can provide.
• Alcohol use and tobacco use are often under-asked about. A simple, non-judgemental routine question at every antenatal visit is the cheapest intervention available.

These points are areas where the sources differ or the evidence is limited, and they should be read as such rather than as facts to be memorised.

• The absolute risk of isotretinoin embryopathy remains unclear, with malformation estimates of 5 to 28% in live-born infants from prospectively reported pregnancies, and small and skewed samples.
• Warfarin embryopathy risk has been quoted as up to 30% in older literature and 6 to 10% in more recent studies, possibly dose-dependent above 5 mg. ^[3]
• ACE inhibitors in the first trimester have inconsistent associations with malformations, and the available data suggest the signal may be driven by maternal disease rather than the drug.
• Valproate dose-response and formulation data remain incomplete: the dose-response for malformations is observed, but no study demonstrates benefit from divided or modified-release dosing.
• FASD risk by drinking pattern cannot be robustly analysed from the current data, and occasional early binge drinking does not appear to carry measurable harm in the small available studies.
• Electronic cigarettes in pregnancy lack robust fetal safety data in the UK sources, hence the pragmatic but unresolved position.
• Sources disagree on method and timing of stopping smoking: ACOG emphasises that cessation at any point benefits the pregnancy, with the greatest benefit before 15 weeks, while UKTIS frames nicotine replacement as the first-line cessation aid; both are consistent in their recommendation that smoking should stop.

Related MedGrasp topics, listed as plain text (topics may be linked once published and verified): preeclampsia, HELLP and eclampsia; gestational and chronic hypertension in pregnancy; pregnancy diagnosis and physiology; diabetes in pregnancy; asthma in pregnancy; urinary tract infection and pyelonephritis in pregnancy; substance use and dependency in pregnancy; oral contraception and preconception care; epilepsy in pregnancy; thyroid disease in pregnancy; anaemia in pregnancy.

This topic is a curriculum gap with no heading in the source syllabus, so no book pages are mapped to it; its scope is the standard medical student curriculum for drugs and teratogens in pregnancy.

The newest open-access editions opened were used: UKTIS monographs dated July 2026 (asthma), April 2026 (tobacco and NRT), October 2026 (valproate), July 2025 (ACE inhibitors), October 2024 (isotretinoin), January 2024 (alcohol), April 2023 (warfarin), September 2022 (thalidomide); ACOG Committee Opinion No. 807 May 2020 (reaffirmed 2026); the FDA pregnancy and lactation labelling page (content current as of March 2025); and StatPearls open-access study guides dated 2026 and 2024. Where sources disagree (for example valproate embryopathy risk estimates, and the unresolved role of electronic cigarettes in pregnancy) both values or both positions are given rather than presenting one as settled.