Drugs During Pregnancy: Methodological Aspects 1st ed.

11. Timing of Drug Use and Effects on the Embryo or Fetus

Bengt Källén¹

(1)

Tornblad Institute, Lund University, Lund, Sweden

The timing of the drug use is often crucial for a harmful effect on the embryo or fetus. This is perhaps most evident for the origin of congenital malformations. For these, the first trimester of pregnancy is of greatest importance. For other outcomes the period of sensitivity is less strict even though it is often thought that the strongest effects on, for instance, preterm birth and neonatal morbidity will be obtained by exposures during the second and third trimester.

The actual time of drug exposure is often uncertain, less so when it is based on interview data than when it is based on prescription data. In a few situations, drugs taken long before pregnancy may remain in the body and cause damage during pregnancy. An example of a drug which disappears very slowly from the patient is the teratogenic retinoid etretinate. Patients who had used that drug are recommended not to get pregnant for a period of 2 years after stopping the drug. This is unusual; in most instances drugs are rapidly metabolized or excreted and their half-life relatively short.

11.1 Exposure Before Conception

The use of a drug before conception could affect sperm or egg cell formation resulting in an abnormal development of the embryo after conception.

Direct damage of eggs or sperm should most likely cause gene mutations or chromosome anomalies. Candidates for such effects are mainly drugs with known mutagenic effects, notably drugs used for treatment of cancer or as immunosuppressant drugs. The typical result of a mutagenic effect would be a condition which is due to a dominant mutation, e.g., achondroplasia. The spontaneous rate of this condition is about 1/10,000 births, and a very strong increase would be needed in order to detect it in a material consisting even of a few thousand exposed pregnancies.

Note that in order to get a mutagenic effect which causes a malformation, the exposure most likely must occur before or around conception. Mutagenesis which takes place during the fetal development may increase the risk of childhood cancer but is unlikely to cause a malformation.

Sperm damage leading to infertility occurs with some chemicals, including some drugs, for instance, methotrexate and finasteride. Such effects are usually temporary and disappear a time after stopping the drug. Another possibility is that the drug causes a mutation in the sperm which will participate in the conception. A third possibility, suggested in animal experiments, is that the drug can cause epigenetic modifications (Cordier 2008).

The most critical period for mutations in sperms is about 3 months before conception, while mutations in eggs can occur any time since the woman’s birth which makes studies of the phenomenon very difficult. So far studies of women who had been treated with potentially mutagenic drugs or radiation because of childhood malignancy and later had pregnancies have not demonstrated an increased risk for malformations in the offspring.

Paternal exposure to drugs like azathioprine/6-mercaptopurine has not been linked to an increased risk for congenital malformations (e.g., Hoeltzenbein et al. 2012, but the studied number of exposures was low, n = 115). In a paper from Motherisk (Lee et al. 2010), 301 TIS questions on paternal exposures were tabulated. They referred both to drugs with a potential to mutagenesis (methotrexate, azathioprine) and drugs with a known or suspected teratogenic activity when used by the woman (e.g., isotretinoin, valproic acid). Among 43 live births, one infant had a congenital malformation, preaxial polydactyly.

11.2 Period of Organogenesis: First Trimester Exposure

Most congenital malformations are a result of a disturbed organogenesis, that is, the formation of the organs during embryonic development. Generally speaking this occurs during the first trimester, but a few congenital malformations can be formed later during pregnancy. Examples are microcephaly and hydrocephaly which may origin late in pregnancy and even postnatally. Some other malformations like pyloric stenosis are also thought to develop late. It is also possible that the growth of the organ (which occurs throughout the whole pregnancy) can be affected with hypoplasia as the result.

In some animal experiments, one has seen teratogenic effects to occur before embryo implantation. Already during the passage of the egg through the fallopian tube, an exchange of chemicals between the maternal organism and the embryo may occur. Such very early damage will probably result in embryonic death and will go unnoticed clinically.

The main pregnancy period when a malformation can arise is, however, the first trimester. Within that period specific time windows exist for the origin of specific malformations, that is, periods when the relevant structures develop. Such “sensitive periods” may be the most relevant ones for teratogenesis and were very typically demonstrated for thalidomide. Theoretically, a drug treatment before the formative period of an organ can cause a malformation because every structure comes from an earlier rudiment. Even though the organogenesis of the heart starts during week 5, the first cardiogenic areas exist already toward the end of the third week and damage at that time may result in later disturbances of heart morphogenesis.

A crude tabulation of the weeks of interest for some congenital malformations is given in Table 11.1. The weeks given are only approximate but can give an idea of the time pattern. What can be seen is that many malformations may be formed already so early that the woman hardly suspects her pregnancy.

Table 11.1

Approximate timetable of organogenesis for some common malformations

Modified after Czeizel (2008)

The stated weeks refer to time after conception, not time after last menstrual period

?The exact timing of gastroschisis is unclear as its mode of embryogenesis is debated

If the drug under study is used during the whole first trimester – a common situation at chronic use of the drug – the period of sensitivity for a specific malformation is relatively uninteresting, but if the drug has been used only for a short time like a week (e.g., an antibiotic), a risk estimate based on exposure during the whole first trimester will be strongly biased toward null, even if we concentrate the study to exposures during the second and third month after LMP as has been suggested (Czeizel 2008).

Let us take an example: use of erythromycin has been associated with an increased risk of cardiovascular defects (Källén et al. 2005). This finding was based on 31 infants with such malformations among 1844 infants exposed to erythromycin in the first trimester which gave an OR =1.84 (95% CI 1.29–2.62). The first trimester is 10 weeks long (counted from conception) and we suppose that the cardiac rudiment is sensitive during 5 weeks. If erythromycin as an average was used for 1 week, crucial exposure might have occurred during a total of 35 days of the sensitive period. Thus the actual number of infants exposed during the sensitive period would be 1844 × 35/70 = 922, and the actual risk at exposure during the sensitive period could be twice as high as that estimated. These are hypothetical calculations but show in what direction exposure outside the sensitive period will affect risk estimates.

If a drug has been used after the formative period of a malformation, it is unlikely to cause it. Some years after the thalidomide tragedy, a rumor was spread in Sweden of an association between the use of meclizine and spina bifida (as far as I know never published outside local media). Meclizine is an antihistamine mainly used for NVP which seldom starts before week 6. Extensive information exists on the harmlessness of meclizine in early pregnancy (Källén and Mottet 2003) – but nearly all data represent treatments of NVP and that would be too late to cause spina bifida. Only if meclizine had been used for other purposes, e.g., moving sickness, before the woman knew about her pregnancy, could an effect on the spina bifida rate occur.

In efforts to use exact timings of exposures during the first trimester in order to increase the sensitivity of the study, one often has the added problem of uncertainty in dating. Most interview or questionnaire information asking which week the drug was used relies on the woman’s statement. If the information is obtained in early pregnancy, an uncertainty about pregnancy week may exist to which may be added an uncertainty if the woman counts from LMP, from conception, or from the first missed period. If the interview is made after delivery when the woman knows about the presence or absence of a congenital malformation, exact dating will add to the uncertainty of exposure data due to the possibility of recall bias.

Studies based on prescription registers can identify the earliest possible exposure but not the actual exposure time. The possibility that the woman had access to and used the drug before the data of filling the prescription adds to the uncertainty of time of exposure.

The possibility that teratogenic drugs could be transferred from the man to the woman at intercourse during early pregnancy and reach and damage the embryo is regarded as unlikely. This was discussed, for example, for isotretinoin which is a drug with a strong teratogenicity and recommendations existed to avoid conception during male therapy with this drug. Estimates have shown, however, that the transferred amounts are so small that it is unlikely that it could harm the embryo (Millsop et al. 2013). No such effect was known of paternal use of thalidomide.

11.3 Exposure After the First Trimester

With few exceptions such exposures will be irrelevant for the origin of congenital malformations but may cause other adverse outcomes like preterm birth, low birth weight, neonatal morbidity, and long-term effects. It is, however, possible that also early exposures may affect outcomes around delivery. Placentation and placenta development may be affected which could, for instance, increase the risk of placental abruption but also of preterm birth. A further consideration is that women who used a drug in early pregnancy may be more likely than other women to use it also later in pregnancy even if she got no new prescription, so early use could be a proxy for later use.

In a recent study of the effect of air pollution of term infant birth weight (Rich et al. 2015), it was suggested that an effect was found only during the eighth month of pregnancy.

References

Czeizel A (2008) Specified critical period of different congenital abnormalities: a new approach for human teratological studies. Congenit Anom (Kyoto) 48:103–109CrossRef

Cordier S (2008) Evidence for at role of paternal exposures in developmental toxicity. Basic Clin Phamracol Toxicol 102:176–181CrossRef

Hoeltzenbein M, Weber-Schoendorfer C, Borisch C, Allignol A, Meister R, Schaefer C (2012) Pregnancy outcome after paternal exposure to azathioprine/6-mercaptopurine. Reprod Toxicol 34:364–369CrossRefPubMed

Källén B, Mottet I (2003) Delivery outcome after the use of meclizine in early pregnancy. Eur J Epidemiol 18:665–669CrossRefPubMed

Källén BAJ, Otterblad Olausson P, Danielsson BR (2005) Is erythromycin therapy teratogenic in humans? Reprod Toxicol 20:209–214CrossRefPubMed

Lee CYW, Jin C, Mata AM, Tanska AM, Einarsson A, Koren G (2010) A pilot study of paternal drug exposure: the Motherisk experience. Reprod Toxicol 29:353–360CrossRefPubMed

Millsop JW, Heller MM, Eliason MJ, Murase JE (2013) Dermatological medication effects on male fertility. Dermatol Ther 26:337–346CrossRefPubMed

Rich DQ, Liu K, Zhang J, Thurston SW, Stevens TP, Pan Y et al (2015) Differences in birth weight associated with the 2008 Beijing Olympics air pollution reduction: results from a natural experiment. Environ Health Perspect 123:880–887PubMedPubMedCentral