Which medication taken by a pregnant woman may show a delayed teratogenic effect in the offspring?

TERATOGENIC DRUGS

A teratogen is an agent that can disturb the development of the embryo or fetus, resulting in spontaneous abortion, congenital malformations, intrauterine growth retardation, mental retardation, carcinogenesis, or mutagenesis.1,2 Known teratogens include radiation, maternal infections, chemicals, and drugs. Among possible teratogens, drugs account for approximately 1% of all congenital malformations of known etiology.3,4 All teratogenic drugs generally produce a specific pattern or single malformation during a sensitive period of gestation with a dose-dependent effect.5

Teratogenic drugs generally produce a specific pattern of abnormalities or a single malformation during a sensitive period of gestation with a dose-dependent effect.

The American Food and Drug Administration (FDA) instituted a rating system for drugs marketed after 1980 based on their safety during pregnancy. Five pharmaceutical categories have been elaborated: A, B, C, D, and X. Drugs under Category A are the safest drugs in which no fetal risks have been demonstrated during controlled human studies, while those under Category X present proven teratogenicity that clearly outweighs their benefits. Category D drugs have demonstrated risks to the human fetus, but in serious diseases or life-threatening situations their benefits outweigh these risks.6

Teratology

Pregnancy

Leflunomide is a teratogenic drug and should not be used in pregnancy or in women who may become pregnant. Consequently, women of childbearing potential should use effective contraception during and up to 2 years after treatment. If women do become pregnant while taking leflunomide, the washout procedure may be used (see material following). In that animal studies show that leflunomide or its metabolites pass into breast milk, breast-feeding women should also not receive leflunomide. Paternal exposure is similarly advised against. The actual risk of malformations after leflunomide exposure is difficult to assess and the evidence is as follows.

One survey69 recorded experience from 175 North American rheumatologists who reported 10 pregnancies in women receiving leflunomide, none of which resulted in congenital malformations.

An Italian group70 reported experience in five women, four exposed in the first trimester and one conceived 6 months after stopping leflunomide (with one case of paternal exposure). Three women had voluntary abortions, and three had live births with healthy babies.

A single U.K. patient has been reported in which the mother had an apparently healthy male baby born 9 weeks prematurely who was subsequently found to be blind in the right eye and to have cerebral palsy with left-sided spasticity.71

Other antirheumatic drugs should be used in patients who are likely to become pregnant, and there are well-known options.72 However, when patients do conceive while taking leflunomide they need to have access to the available information to make an informed choice.

34.5.3 Other Teratogenic Agents: Thalidomide and Misoprostol

Thalidomide and misoprostol are two teratogenic drugs, known to induce a variety of systemic malformations. Thalidomide was commercialized as a sedative drug in the late 1950s before being withdrawn from the market in 1961. Teratogenicity is due to its angiogenesis inhibiting activity, which causes multiple systemic malformations, as well as abnormal cortical development and neuronal hyperexcitability (Hallene et al., 2006). Misoprostol is a methyl ester derivative of prostaglandin E1, used especially in Central and South America to treat gastric ulcers, but also a popular abortion inducer due to its powerful stimulatory effect on uterine contractions: the teratogenic effects of misoprostol have been studied in children born after unsuccessful abortion attempts (Bandim et al., 2003). These two drugs display several interesting parallels: both hamper fetal blood perfusion either directly (thalidomide) or indirectly (misoprostol); both produce systemic and especially ophthalmologic malformations, primarily coloboma and microphtalmos (Miller et al., 2004, 2005); both frequently cause prenatally exposed children to develop signs of Moebius sequence, including horizontal strabismus (Duane syndrome) and facial nerve palsy due to the involvement of the VI and VII cranial nerves (Bandim et al., 2003; Miller et al., 2005); both are associated with enhanced risk of autism and/or mental retardation, provided exposure occurs early in development (Miller et al., 2005; Strömland et al., 1994). The critical period for teratogenetic induction of autism has been defined in great detail for thalidomide, where it appears to be restricted to as early as 4–6 weeks into gestation (i.e., 6–8 weeks since the last menstrual cycle) (Miller et al., 2005; Strömland et al., 1994). The critical period for misoprostol has not been defined with the same precision, but it is known that maximum fetal vulnerability occurs during the first 2 months of pregnancy, and possible 5–6 weeks after fertilization (i.e., 7–8 weeks since the last menstrual cycle) (Bandim et al., 2003). Patients with idiopathic Moebius sequence and with no history of prenatal exposure to thalidomide or misoprostol have been found at enhanced risk of autism by some (Gillberg and Steffenburg, 1989), but not by others (Briegel et al., 2009). It will be important to determine conclusively whether there is a significant association between Moebius sequence and autism, because this would demonstrate that autism specificity is conferred more by a sensitive time window during development, rather than by the specific nature of prenatal insults or teratogenic mechanisms involved.

Antirheumatic therapy using MTX

Ten publications listing more than 110 pregnancies with exposure during the first trimester refer to the so-called “low-dose” therapy for rheumatic diseases. However, with the exception of a recently published small prospective study from France (Lewden 2004) and our own unpublished data, all listed cases represent retrospective reports or, at best, small prospective case studies describing a maximum of four pregnancies (Østensen 2000, Donnenfeld 1994).

There were a total of four children with characteristic malformations. Two of the mothers (Del Campo 1999, Powell 1971) had taken more MTX (one took 3 ÷ 2.5mg per week until week 10, and the other took 5mg per day until week 8) than is usual for a “low-dose” therapy (i.e. maximally 25mg per week). Another mother received 7.5mg per day for 2 days in week 6 (Nguyen 2002), and the fourth woman received 12.5mg per week until week 10 in combination with a daily dose of 1mg folic acid (Buckley 1997).

In contrast to these publications, there are case reports of 14 healthy children whose mothers had a dosage of between 7.5 and 15mg per week (Østensen 2000, Donnenfeld 1994, Feldkamp 1993, Kozlowski 1990), 4 spontaneous miscarriages (Østensen 2000, Kozlowski 1990), and 2 induced abortions without embryopathic background. Chakravarty (2003) describes 38 retrospective ascertained pregnancies with “low-dose” MTX therapy without giving details regarding the period of administration and the dose; 21 children were born healthy, 3 had malformations (no details given), there were 7 spontaneous miscarriages, and 8 pregnancies were electively terminated. A prospective French study (Lewden 2004) with 28 cases and a weekly median dose of 10.5mg MTX reports 4 spontaneous miscarriages, 5 induced abortions, and 19 live births, none of which was a MTX embryopathy.

These findings are in agreement with those of the TIS Berlin: of the 22 prospectively ascertained pregnancies with exposure during the first trimester (with a weekly dose of 10–25mg MTX), 3 pregnancies were terminated despite inconspicuous ultrasound findings, 5 resulted in spontaneous miscarriages, and 13 healthy children were born (1 premature infant at 36 weeks). One child whose mother additionally took phenprocoumon and other drugs weighed 1600g at birth and was growth restricted, had an inguinal hernia, and was highly irritable for 14 days.

The dose ranges of MTX administered in combination chemotherapy, attempted abortions, and for rheumatic indications (“low-dose” treatment) overlap. Therefore, the conclusion is inadmissible that there are safe and risky indications for MTX. However, since there has been only one case with suspicious symptoms following 10mg MTX per week, the hypothesis proposed by Feldkamp (1993) seems plausible: that MTX is teratogenic only at a weekly dose in excess of 10mg. In addition, the author postulates a sensitive phase between weeks 8 and 10. However, the data available do not permit a definitive conclusion in this respect.

Recommendation.

Developmental anomalies resulting from treatment with the teratogenic drug methotrexate have been observed in a number of pregnancies, essentially comprising growth restriction beginning at a prenatal stage, severe ossification defects of the calvaria, facial dysmorphisms, CNS anomalies with or without diminishing of intelligence, and defects of the extremities. A safe dose cannot be defined; however, at present there are no indications for teratogenic effects to occur below a weekly dose of 10mg. Antirheumatic “low-dose” therapy which had been continued (inadvertently) during the first trimester seems to be associated at most with a slightly increased risk of malformations. In general, exposure during the first trimester does not necessarily lead to malformations, even when used to treat malignant diseases.

Although antirheumatic MTX treatment should be stopped before planning a pregnancy, the data available at present do not justify the advice to postpone pregnancy for at least 3 months after stopping methotrexate therapy. Pregnant women who have been (inadvertently) exposed to MTX during the first trimester should be offered a detailed ultrasound scan to obtain confirmation of the normal development of the fetus.

1.21.11.2 Policy Shifts and Public Concern

Although not approved in the United States, the teratogenic drug thalidomide caused birth defects in other countries and increased public interest in drug regulations in the early 1960s. In 1962, the Congress amended the FDCA to require the FDA to approve new drugs. Consequently, FDA approval for NDAs required evidence of safety and effectiveness for all drugs activated between 1938 and 1962.62 Moreover, the FDA concluded that generic drugs would require their own approved NDAs.63 This greatly increased the time and money required for a company to introduce a generic drug to market, and was the first real barrier to the entry of generic substitutes.

In 1970, the FDA instituted Abbreviated NDAs (ANDAs) for the approval of drugs instituted prior to 1962. The ANDA required all information for an NDA, but no data regarding safety and efficacy testing.63 The ANDA approval process relied on information from the National Academy of Sciences’ Drug Efficacy Study Implementation (DESI) program for the approval process.50 While generics of pre-1962 pioneer drugs benefited from the ANDA process, the FDA had no such process for post-1962 drugs. Beyond this limitation, all information for IND and NDA applications were regarded as confidential proprietary business information, and could not be disclosed by the FDA to the public or any competitor. As a result, post-1962 generics required independent animal and human tests to demonstrate safety and effectiveness.62 This created significant time and cost investments for new generic drugs entering the market.

In 1981, the FDA began to allow some generic versions of post-1962 drugs to be marketed under a ‘paper NDA’ policy whereby FDA approval could be sought based on evidence derived primarily from published reports of well-controlled studies.63 However, the FDA requirements to meet this definition were strict, as they estimated this information to be unavailable for 85% of all post-1962 drugs. Furthermore, the lack of clear specific guidelines, and excessive FDA discretionary power prevented any significant number of approvals from paper NDAs.63–65 The legislation and litigation of the period from 1962 to 1984 made generic market entry difficult, and as a result only one in ten drugs had a generic equivalent.66

Premedication and Patient Positioning

Premedication should be appropriate for the specific cardiac lesion and physical status of the patient. Teratogenic drugs should be avoided, especially in the first trimester of pregnancy. After the 34th week of gestation, stomach emptying is delayed and patients are at increased risk for pulmonary aspiration. Although it is not possible to ensure gastric emptying before anesthesia induction, sodium citrate and an H2-receptor antagonist may provide some protection against aspiration pneumonia. The gravid uterus obstructs aortic flow and IVC blood return to the heart. Gravid patients should never be supine; they must be positioned with left uterine displacement throughout the perioperative period.

Abnormal Chromosome Structure

Various abnormalities of chromosome structure can give rise to malformations in development. Some chromosomal abnormalities result from chromosome breakage induced by environmental factors such as radiation and certain chemical teratogens. This type of structural error is usually unique to a given individual and is not transmitted to succeeding generations.

Other types of structural abnormalities of chromosomes are generated during meiosis and, if present in the germ cells, can be inherited. Common types of errors in chromosome structure are reciprocal translocations, isochromosome formation, and deletions and duplications (Fig. 8.11). One well-defined congenital malformation resulting from a deletion in the short arm of chromosome 5 is the cri du chat syndrome. Infants with this syndrome are severely mentally retarded, have microcephaly, and make a cry that sounds like the mewing of a cat.

3 Congenital Lesions

Spontaneous congenital lesions of the ear in F344 rats are infrequently reported. This may be due in part to the lack of routine examination of this structure in toxicologic studies. Microtia is seen occasionally.

Chemical teratogens produce a variety of congenital lesions of the ear. These include persistence of the meatal plug, branching of the primordium of the external acoustic meatus, narrowing of the tympanic cavity, presence of only one to two primordia of the middle ear ossicles, hypoplasia of the stapedial artery, a maximum of two turns of the cochlea, and incomplete differentiation of the organ of Corti and the semicircular ducts.

Teratogenesis

Women who called two Italian Teratology Information Services after being exposed to itraconazole during the first trimester and a contemporary group of pregnant women who contacted the Services because they had been exposed to a non-teratogenic drug during the first trimester have been compared in a prospective cohort study [63c]. Information was obtained by a trained operator via a structured questionnaire no earlier than 1 month after delivery. A conducted the interview. Information about major congenital anomalies, type of delivery, birth weight, and any pregnancy or neonatal complications was collected on 206 women who had been exposed to itraconazole and 207 controls. There were no significant differences in major congenital anomalies (3/163 versus 4/190 respectively). There were no statistical differences in the rates of vaginal delivery, premature birth, low birth weight, or high birth weight. However, the rates of live births (163/206 versus 190/207), spontaneous abortions (23/206 versus 10/207), and termination of pregnancy (19/206 versus 7/207) were significantly different. Thus, itraconazole exposure in the first trimester did not increase the risk of major congenital anomalies, but did increase the rates of spontaneous and induced abortion. Given the relatively small sample size, larger studies are warranted.