Key Points
- •
At baseline, each pregnancy has a 2% to 4% risk for a congenital anomaly diagnosed at birth.
- •
The adverse effects of exposures on embryo-fetal development depend on the agent, dose, and timing of exposure.
- •
Resources are available for up-to-date information on specific exposures during pregnancy.
Introduction
A teratogenic exposure is one that has the ability to interfere with normal development of the fetus. Some exposures increase the risk for structural anomalies, others may interfere with fetal organ development, and others may increase the risk for other adverse pregnancy outcomes, including intrauterine growth restriction, preterm birth and intrauterine fetal demise.
This chapter discusses some of the exposures that might increase the risk for abnormal development in human pregnancy. At baseline, each pregnancy has a 2% to 4% probability of a structural anomaly diagnosed at birth. Chemically induced anomalies, including those caused by medication exposure, are thought to occur in fewer than 1% of these cases.
Historical Perspective
In the 19th century, experimental teratology focused on frogs and birds in which, for example, hypoxia could cause developmental aberrations. However, the mammalian uterus was thought to be impervious to external hazards, and genetic abnormalities were blamed for the occurrence of malformations. The notion that human embryos might be harmed by environmental factors was a revolutionary concept until Norman Gregg, an Australian physician, noted in his practice an increase in children with congenital cataracts after a rubella epidemic. He identified what came to be called the ‘congenital rubella syndrome’ with a combination of additional findings, including heart defects, hearing loss, thrombocytopenia and poor growth.
In response to the increased interest in studying birth defects, the Teratology Society was formed in 1960. Shortly thereafter, the field was changed by the unexpected tragedy of thalidomide. Thalidomide, a sedative–hypnotic, given in the usual (medicinal) doses caused fetal malformations in the absence of toxicity to the mother. Since this discovery of what is now called selective embryotoxicity, drug testing regulations have changed. For example, in 1962, the Food, Drug and Cosmetic Act was expanded to include the Kefauver-Harris Amendment, which requires drug manufacturers to provide proof of effectiveness and safety before approval as well as provide information regarding side effects. By 1966, the US Food and Drug Administration (FDA) instituted standard test protocols for drug testing in pregnant laboratory animals.
Mechanisms of Teratogenicity
Experience in experimental teratology led to the development of potential mechanisms for abnormal development, articulated by James Wilson ( Table 4.1 ). Wilson’s mechanisms included the idea that impairment of survival and function in differentiated cells in key locations in an embryo could perturb development. This concept gave rise to the ‘all-or-nothing’ principle in which it was believed that before gastrulation (approximately postconception day 14), cells in the embryo were largely undifferentiated and could substitute for one another, decreasing the ability of an exposure to cause selective malformations without destroying the entire embryo. This all-or-nothing principle has not proved invariable in experimental teratology, but it remains true that it is more difficult to produce malformations with experimental damage to undifferentiated compared with differentiated embryonic tissues.
| Mutation Chromosomal aberrations Mitotic interference Altered nucleic acid synthesis and function Lack of precursors, substrates and coenzymes for biosynthesis Altered energy sources Enzyme inhibition Osmolar imbalance Changed membrane characteristics |
Mechanisms of congenital malformations more recently have been understood in terms of developmental pathways that might be inhibited by a specific exposure. For example, DiGeorge syndrome, which is often associated with a deletion in the long arm of chromosome 22, includes disruption of Tbx1 , a gene that plays a role in development of cells in the secondary heart field. As a consequence, conotruncal heart defects can be seen in affected children. It has not been identified, however, whether isotretinoin therapy, which also has been associated with conotruncal heart defects, works by way of interference with Tbx1 .
There are basic cell behaviours during embryo development that may serve as targets for exposures. Cell populations in the embryo migrate to targeted locations, and interference with migration may cause abnormal development. Neural crest cells migrate into the branchial arches, for example, and interference results in a group of disorders of the jaw, ears, and zygomatic arch, among other defects. Neurons migrate from their birthplace in the centre of the brain towards the periphery, and inhibition of migration can cause microcephaly. Cells also can induce behaviours in neighbouring cells. The tips of the ureteric buds induce the mesenchyme of the developing kidney to form functioning nephrons. Cells produce substances that act at a distance to modify cell differentiation. For example, cells in the medial (postaxial) limb bud secrete Sonic Hedgehog protein, which diffuses across the limb paddle to help determine the identity of the individual digits.
The greater understanding of cellular and molecular events during embryo development has given rise to an opportunity for understanding mechanisms of abnormal development in more detail. Although the mechanistic details of malformations in human beings associated with exposures are incompletely understood, we expect that developments in the genetic basis of malformations and in cell biology will lead to our improved understanding of exposure-associated malformations.
Underlying Principles
The mechanisms of abnormal development were the basis of one of James Wilson’s principles. The full set of principles is listed in Table 4.2 as developed in the 1950s and 1960s. Of note is Wilson’s fourth principle, which tells us that malformations are not the only kind of developmental toxicity of importance. An exposure that causes an organ, say the brain, not to function correctly can be devastating even if there are no recognisable structural malformations in the child. We commonly call these adverse effects developmental toxicity , which is a more helpful term than teratogenicity in not requiring a definition of what exactly counts as a malformation. An exposure producing nonmalforming developmental effects does not necessarily also produce malformations. Indeed, an exposure that produces one kind of malformation does not of necessity produce any other kind of malformation. Thalidomide, for example, produces only certain kinds of limb defects, not all kinds of limb defects. Effects of developmentally toxic exposures are specific, producing a finite grouping of adverse effects. Dr Wilson captured this idea of specificity in his third principle, and the Public Affairs Committee of the Teratology Society made it explicit in 2005.
|
Wilson’s sixth principle is among the most important for practitioners because it reminds us that for all medications, other chemicals, and physical agents, there are an exposure level that produces no harm, an exposure level that produces death and a range of exposures in between that produces a gradation of effects. It is not useful to talk about agents (drugs, chemicals, radiations) as teratogenic or nonteratogenic. It is preferable to talk about teratogenic exposures , an exposure including the identity of the agent and the dose level at which it is encountered. X-irradiation during pregnancy was associated with microcephaly and mental retardation when exposure levels were about 50 cGy from the atomic bombings of Japan. Flying across the United States is associated with estimated radiation exposures about 8000 times lower and would not be expected to have the same effects. We do not, therefore, characterise x-ray as teratogenic or nonteratogenic. It depends, among other things, on dose.
How much evidence is needed before it is worthwhile warning patients and health care providers about possible adverse effects of exposures in reproducing men and women? There is some controversy in this area because there are potential adverse effects of excessive warning, namely anxiety, the interruption of otherwise wanted pregnancies, and the discontinuation of important, even essential, medical therapy.
It has been fashionable from time to time to claim that most human teratogenic exposures were first identified by ‘astute clinicians.’ Thalidomide is often cited as evidence of this claim because the first publications on thalidomide birth defects came from one paediatrician in Germany and another in Australia who described clusters of children with phocomelia and wondered whether there was a pregnancy exposure as the cause. The astute clinician model has been described in mathematical terms, but the astute clinician model produces only a hypothesis that must be confirmed by other evidence. The association between rubella and congenital cataracts and between thalidomide and phocomelia were, in fact, doubted by some teratologists until additional evidence was forthcoming. Perhaps clinicians whose hypotheses are confirmed may be as lucky as they are astute.
For a determination of causation in teratology, methods have been advanced that are based on the Hill criteria. These criteria ( Table 4.3 ) were most famously applied in the consideration by the US Surgeon General of the evidence for a causal relationship between cigarette smoking and lung cancer. To entertain a conclusion of causation, you need not satisfy each of the Hill criteria, but the more you satisfy, the more confident you can be that an association is causal.
|
In this chapter, we discuss some exposures that have been associated with developmental toxicity in human pregnancy. In some instances, there is evidence that the association is causal, but in other cases, a causal association cannot be concluded. We recognise, however, that giving patients advice about exposures during pregnancy does not require conviction that causation criteria have been satisfied. For example, we recommend that women who have taken lithium during pregnancy consider fetal echocardiography, even though causation criteria are not satisfied that lithium causes Ebstein anomaly or any other heart defect. In the end, counselling about exposures relies on the same kinds of judgments about adverse outcomes that we use every day as clinicians considering possible risks and benefits of medication therapy.
Personalised Risk Assessment and Resources
Several excellent books are available as resources. However, soon after publication, books in this ever-changing field have the potential to be outdated. Online databases offer summaries of pregnancy exposures written and frequently updated by teratology experts include TERIS ( http://depts.washington.edu/terisdb/terisweb/index.html ), REPROTOX ( Reprotox.org ), and Briggs (available through http://wolterskluwer.com.easyaccess1.lib.cuhk.edu.hk ). Lactmed ( https://toxnet-nlm-nih-gov.easyaccess1.lib.cuhk.edu.hk/newtoxnet/lactmed.htm ) is a free online resource for medication exposure and lactation.
There are two networks of teratology information services, one in North America called the Organization of Teratology Information Specialists (OTIS; http://www.mothertobaby.org ) and one serving Europe, the European Network of Teratology Information Services (ENTIS; http://www.entis-org.eu ). Both networks are staffed with physicians, genetic counsellors and teratology experts who offer individualised risk assessments and up-to-date information for any drug or environmental exposure during pregnancy and lactation. These complimentary services are available to both health care providers and patients. These networks also conduct prospective studies and patient follow-up after pregnancy exposures.
Pregnancy registries open to enrolment may be located from the FDA’s Office of Women’s Health ( www.fda.gov/ScienceResearch/SpecialTopics/WomensHealthResearch/ucm251314.htm ).
Selected Human Exposures
Medication
A survey of 1000 pregnancies in southwest France found that 99% of the women received a prescription for at least one drug during pregnancy with a mean of 13.6 medications per woman. In other studies, approximately two thirds of pregnant women took at least one medication during pregnancy, and about 60% of patients used a prescription medication. In the past few decades, the overall use of medication and the number of women using 4+ prescription medications anytime in pregnancy more than doubled, and for the first trimester, it more than tripled.
Thalidomide
No other therapeutic agent has had a greater impact on how we think about teratogenic potential as thalidomide. In 1957, thalidomide was marketed as a sedative and antiemetic. Thalidomide did not cause acute toxicity in adults, making it one of the few agents that are selectively embryotoxic. In the late 1950s and 1960, there were few case reports of phocomelia, an unusual limb reduction defect in which the hand and foot arise from the shoulder or hip. By 1961, Dr Lenz in Germany and Dr McBride in Australia independently noted an association between phocomelia and exposure to thalidomide. Thalidomide exposure was associated with specific limb reduction defects, oesophageal and duodenal atresia, congenital heart defects (tetralogy of Fallot), external ear and cranial nerve abnormalities, and renal agenesis. The sensitive time period for the limb defects was 21 to 36 days postconception.
Although thalidomide was removed from the market, it was found later to be effective for erythema nodosum leprosum. In 1998, the US FDA approved Thalomid for this use. In 2006, the medication was approved for multiple myeloma. Prescription and dispensing of thalidomide is strictly controlled in the US through a Risk Evaluation and Mitigation Strategy (REMS) to prevent exposure during pregnancy.
Isotretinoin
Isotretinoin (13-cis-retinoic acid), a derivative of vitamin A (retinol), is an effective treatment of cystic acne vulgaris. Use of the medication increases the risk for spontaneous abortion and the development of a specific set of anomalies, including neural crest-related facial and palate defects, micrognathia, external and internal ear anomalies (microtia or anotia), conotruncal heart defects, thymic abnormalities and deficits in intelligence. Effects on cognition can occur in the absence of structural anomalies.
There is no associated risk for poor fetal outcome in cases in which isotretinoin was discontinued before pregnancy. The current recommendation is to discontinue isotretinoin 1 month before conception, but considering the half-life of 29 hours, after 1 week off therapy, maternal blood concentrations should be negligible.
Use of topical retinoids does not increase the risk for structural malformations or developmental delay due to decreased availability of the medication and its metabolites in maternal plasma.
Warfarin
In 1948, warfarin was marketed as a potent rodenticide because of its capability of inducing internal haemorrhage. Warfarin prevents vitamin K from acting as a cofactor in hepatic synthesis of factors II, VII, IX and X and was adapted for human clinical use because of its oral bioavailability and reversibility by vitamin K. Warfarin was associated with embryo-fetal growth restriction, nasal hypoplasia, fibula hypoplasia and stippled epiphysis. Embryotoxicity is associated with exposure between 6 and 9 weeks of gestation, although one report did not note a warfarin embryopathy in those exposed before 8 weeks of gestation.
Other poor perinatal outcomes associated with warfarin exposure include an increased risk for stillbirth, spontaneous abortion, preterm birth and low birth weight. The underlying maternal disease may contribute to the increased risk for poor pregnancy outcomes.
Warfarin is still used by some practitioners in pregnancy, especially in women with mechanical heart valves. Heparin and low-molecular-weight heparin (LMWH) are the mainstays of anticoagulant therapy during pregnancy, but they may not be effective enough, especially in women with mechanical heart valves. More favourable maternal and fetal outcomes may be associated with the use of low-dose warfarin (<5 mg/day) followed by LMWH close to the time of anticipated delivery, but low-dose warfarin therapy has also been associated with warfarin embryopathy. Second and third trimester exposure may increase the risk for central nervous system (CNS) defects, possibly associated with microhaemorrhages in neuronal tissues.
Methotrexate and aminopterin
Folic acid is a cofactor in the synthesis of thymidylate, a rate limiting step in DNA synthesis. Methotrexate (MTX), a folic acid analog, has been used for the treatment of malignancy, rheumatic disorders, psoriasis and ectopic pregnancy. Aminopterin is also a folic acid antagonist that is a close structural analog of MTX. Aminopterin interferes with early human fetal development and was used as an abortifacient in the 1940s and 1950s. Successful and failed abortion attempts using this agent were associated in case reports with fetal malformations, including hydrocephalus, meningomyelocele, anencephaly, limb malformations, cleft lip with cleft palate and developmental delay.
Similar human malformations were noted after pregnancy exposure to MTX. Feldkamp and Cary (1993) presented a review of case reports. Based on six malformed infants, they concluded that 6 to 8 weeks after conception is the sensitive period for malformations induced by MTX exposure. They suggested that a MTX dose of more than 10 mg/week was necessary to produce anomalies such as clover-leaf skull with a large head, swept back hair, low-set ears, prominent eyes and a wide nasal bridge. A disproportionality analysis study of MTX and aminopterin case reports and case series provided support for pulmonary atresia, craniosynostosis and limb deficiencies, which were reported more often than expected in MTX-exposed children.
Because MTX-associated birth defects have only been described in case reports, the incidence of malformations after MTX exposure is not known. Of clinical pertinence is the incidence of malformations after MTX exposure for a misdiagnosed ectopic pregnancy when an intrauterine pregnancy exists. The dose of MTX used for this purpose (75–100 mg) is higher than the antirheumatic doses seen in relatively large series of MTX exposure, and the medication is given earlier in gestation than the proposed critical window, usually before 6 weeks postconception. However, case reports have described malformations in these cases and raised the question of a distinct syndrome caused by such early exposures.
Evidence is still not clear as to how long a woman should wait to conceive after she has been successfully treated for a previous ectopic pregnancy. MTX can persist in the maternal liver for months. In 2009, a study showed no difference in outcome in pregnancies conceived less than 6 months compared with more than 6 months after MTX therapy.
Misoprostol
Misoprostol (Cytotec) is a synthetic prostaglandin E1 analog indicated for the prevention of gastric ulcers but also used to empty the uterus after an incomplete abortion, to ripen the cervix in preparation for labour and in the treatment of a postpartum haemorrhage. In combination with other agents, the use of oral misoprostol was approved by FDA for the termination of pregnancy of less than 49 days’ duration. With the use of 200 μg of misoprostol, the uterine artery resistance indices increase, supporting the theory of reduced uteroplacental perfusion as one mechanism for abnormal development. Other mechanisms propose an increase in intrauterine pressure or interruption of normal embryonic vascular development. Most misoprostol exposure in pregnancy occurs between 5 and 8 weeks of gestation. Misoprostol as an abortifacient fails in 10% of cases, and the risk for failure is higher when it is used as a single agent instead of in combination with mifepristone or MTX. Birth defects after misoprostol use are reported most often in countries where abortion is illegal and not widely available.
Case reports and case series report a possible misoprostol association with terminal transverse limb reduction defects, abdominal wall defects, palsies of the sixth and other cranial nerves (Möbius syndrome), bladder exstrophy and autism spectrum disorder with Möbius syndrome. A review of four studies that included 4899 cases of congenital anomalies and 5742 normal control participants reported an association with Möbius syndrome (odds ratio (OR), 25.31; 95% confidence interval (CI), 11.11–57.66) and terminal transverse limb defects (OR, 11.86; 95% CI, 4.86–28.90).
Angiotensin-converting enzyme inhibitors and angiotensin II receptor antagonists
Angiotensin-converting enzyme (ACE) inhibitors reduce the activity of ACE and lower blood pressure. These medications have been used in the treatment of hypertension, cardiac failure and diabetic nephropathy. ACE inhibitor exposure in the second and third trimesters can reduce fetal blood pressure and renal function and has been associated with oligohydramnios, growth restriction, hypocalvaria, and renal failure.
In one study, first trimester exposure to ACE inhibitors showed an increased risk for cardiovascular defects (relative risk (RR), 3.72; 95% CI, 1.89–7.30) and CNS defects (RR, 4.39, 95% CI, 1.37–14.02) in offspring of women who filled a prescription for an ACE inhibitor. These findings might have been due to maternal diabetes, for which adequate adjustments did not appear in the analysis. These findings were not replicated by subsequent studies that might have been more successful in adjusting for maternal diabetes.
Angiotensin II receptor antagonist exposure in the first trimester was not associated with an increase in congenital anomlies. After the first trimester, these drugs have been associated with oligohydramnios, limb contractures, pulmonary hypoplasia and fetal renal dysfunction.
Lithium
Lithium has been used for the treatment of bipolar disorder. Based on reports from the 1970s, lithium was thought to increase the risk for Ebstein anomaly in which the tricuspid valve is severely displaced toward the right ventricular apex, causing significant tricuspid regurgitation. Although very few cases of Ebstein anomaly were noted in a lithium exposure registry, the prevalence was higher than the expected rate of 1 in 20,000 62 ; however, this registry included cases reported after the diagnosis of a birth defect had been made. According to one author, lithium might increase the risk for Ebstein anomaly to 1 in 1000 exposed children. There has been disagreement with the conclusion that lithium exposure causes Ebstein anomaly. If the structural malformation risk exists, it is likely to be on the order of 0.1%. Fetal echocardiography can be considered.
Mycophenolate
Mycophenolate mofetil and mycophenolate sodium are used as immunosuppressants after organ transplantation or in treatment of autoimmune disease. The US National Transplant Pregnancy Registry (NTPR) reported 26 mycophenolate exposed pregnancies born to 18 women. There were 15 liveborn children, 4 of whom had malformations. Three children had microtia, and of those, two also had cleft lip and palate. The fourth child had hypoplastic nails and shortened fifth fingers. Many case reports emerged showing different malformations; however, it was unclear if all of these malformations represented a mycophenolate embryopathy. The most characteristic abnormalities appeared to be ear abnormalities, facial clefts, and perhaps conotruncal heart defects.
The existence of a mycophenolate embryopathy was verified by epidemiology studies. The National Transplantation Pregnancy Registry reported an increase in malformations and other adverse outcomes among pregnancies of women exposed to mycophenolate compared with other transplant patients on different regimens. ENTIS reported 57 prospectively ascertained pregnancies after maternal therapy with mycophenolate. There were 29 liveborn children, 16 spontaneous abortions and 12 elective terminations. There were two late terminations because of multiple malformations consistent with mycophenolate exposure. Of the liveborn infants, six had major congenital defects: two with external auditory canal atresia, one with tracheoesophageal atresia, one with severe hydronephrosis, one with an atrial septal defect and one with a myelomeningocele. The malformation rate is thought to be greater than 20%. In this study, there were also increases in preterm birth (62%) and low birth weight (31%). A 2014 British study followed nine pregnant women exposed to mycophenolate. There were no malformations noted, but there was an increase in composite poor pregnancy outcome (OR, 5.31, 95% CI, 1.05–26.96).
Anticonvulsants
Most anticonvulsants have been associated with an increase in structural malformations, although it has not always been easy to separate medication use from a possible genetic or other contribution of maternal epilepsy. Pregnancy registries currently are collecting and evaluating outcome data with which to address the possible risks of anticonvulsant use with more precision. One such registry estimated malformation risks of common anticonvulsants compared with that of lamotrigine as shown in Table 4.4 . In this study, valproic acid was associated with increases in spina bifida, hypospadias, oral clefts and some heart defects; phenobarbital was associated with heart defects and oral clefts; and topiramate was associated with oral clefts. Valproic acid has also been associated with cognitive impairment and autism.
| Drug | Malformation risk estimate | 95% confidence interval |
|---|---|---|
| Valproic acid | 5.1 | 3.0–8.5 |
| Phenobarbital | 2.9 | 1.4–5.8 |
| Topiramate | 2.2 | 1.2–4.0 |
| Lamotrigine | 1.0 | Reference |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
