html xmlns=”http://www.w3.org/1999/xhtml” xmlns:mml=”http://www.w3.org/1998/Math/MathML” xmlns:epub=”http://www.idpf.org/2007/ops”>
Introduction
At least one in ten women is affected by mental ill health during pregnancy or the postpartum period (see Chapter 10). In most mild to moderate cases, nonpharmacological interventions will be appropriate. In women with more severe illness, however, psychotropic medication is often the mainstay of management. The focus of this chapter is the reproductive safety of the main psychotropic medications used in psychiatric disorders and how this can be applied to the management of women across childbearing.
A drug may harm a fetus early on in organogenesis, when major structural anomalies could occur, or later in pregnancy, when functional problems or suboptimal obstetric outcomes can result. Although advances have recently been made in this area, most evidence is very limited. Rigorous enquiry cannot include randomized controlled trials in pregnant or lactating women as they are not considered ethical. The next best approach is prospective cohort studies that include control groups and data collection before the studied outcome occurs. This reduces selection and recall bias, but cannot avoid “confounding by indication.” A confounder is a factor that is associated independently with both the exposure (to the medication) and the outcome (e.g., malformation). Confounding by indication implies that because women treated with antipsychotic medication are unwell, any adverse effects in their children may be associated with the illness for which the medication is prescribed rather than the medication itself.
Pregnancy registers with a prospective design have been established, particularly for antiepileptic drugs, and have begun to publish results. However, there are problems with the small size of these datasets and the bias in their sampling. If a study is set up to test whether an agent taken in the first trimester of pregnancy increases the risk of major congenital malformations two-fold above a similarly sized control group where the expected incidence is, say, 2%, then approximately 900 cases need to be included in both groups (Dellicour et al., 2008). More relevant for clinical practice is whether the risk of specific congenital malformations that have a major impact on the survival or functioning of the child is increased. If the background rate of such a condition is very small, 0.1%, for example, as it is for persistent pulmonary hypertension, and a study aims to test a clinically meaningful five-fold increase, then the required sample size is about 2,300 in each group (Dellicour et al., 2008). Existing pregnancy registers are beginning to approach such numbers, but usually only for groups of drugs, rather than individual agents, which may well differ in their potential for reproductive toxicity. Larger sample sizes can be achieved by epidemiological cohort and case-control studies, but the advantages of these may be offset in some population cohorts by uncertainty about timing and duration of exposure to the offending agent, as well as the accuracy of diagnosis and clinical significance of malformations. One example of a large pregnancy register prospectively collecting data about drug use in early pregnancy is the Swedish Medical Birth Register. The register has been covering around 98% of all pregnancies since 1995, including 1,764,458 deliveries to date. At the first antenatal care appointment the woman is questioned about any drug use since she became pregnant, hence recall bias is avoided.
Other challenges in this research are confounding factors for infant and pregnancy outcomes. Although suboptimal pregnancy and developmental outcomes in offspring of mothers with mental illness have been reported, it is not yet clear how much of this effect can be attributed to psychiatric disorders themselves, that is, confounding by indication. A host of other parameters can also act as confounders, including psychiatric comorbidities, physical illnesses, concomitant medication, smoking, maternal age, maternal obesity and low socioeconomic status. In women with mental illness, several of these characteristics are often present.
In this chapter, we discuss the challenges of the research in this complex area and consider classes of psychotropic drugs and individual agents to which mothers and their offspring are exposed. We not only review the literature on safety and potential adverse effects but also discuss implications for clinical practice.
Pharmacokinetics in pregnancy
Pharmacokinetics is the science describing what “the body does with the drug,” that is, how the drug is absorbed, distributed and eliminated from the body. Pharmacokinetic data from studies in men or nonpregnant women cannot necessarily be extrapolated to pregnant women. The continuous physiological changes throughout pregnancy, particularly during the second and third trimester, alter pharmacokinetic properties for many drugs. Some examples of pregnancy-induced changes are slowed gastric emptying, increased volume of distribution, changes in hepatic clearance and decreases in α1-glycoprotein, which results in a higher proportion of the unbound, biologically active fraction of a drug. Furthermore, the activity of hepatic drug metabolizing enzymes changes during pregnancy. For example, the activity of cytochrome P4502D6 is increased and of P4502C19 decreased (see review by Hodge and Tracy, 2007). The renal elimination of drugs also changes across childbearing with rises from early pregnancy to a peak by the end of the second trimester up to 150% above baseline. In some women, decreases begin to occur in the third trimester and others only in the first few weeks postpartum. Little is as yet known about the net effect of these changes on the concentration of free and total parent compounds and their biologically active metabolites in human pregnancy.
Transfer of drugs to the fetus occurs mainly via diffusion across the placenta, favoring movement of smaller, lipophilic agents, such as psychotropic drugs. The rate-limiting step of this process is placental blood flow. Both the fetal liver and the placenta can metabolize drugs: Immature phase I and II metabolism can occur in the fetus at 8 weeks post conception. Metabolic enzyme activity, however, is low and 50% of the fetal circulation from the umbilical vein bypasses the fetal liver to the cardiac and cerebral circulation. Elimination from the fetus is by diffusion back to the maternal compartment. As most drug metabolites are polar, this favors accumulation of metabolites in the fetus.
Psychotropic drugs and breastfeeding
The amount of maternal drug present in the blood circulation of a breastfed infant is the result of complex interactions that could be regarded among the most challenging in pharmacology. Factors involved include maternal pharmacokinetics, the molecular size, lipophilicity and protein binding of the drug; the fat content of the milk; volume and frequency of feeding; the fate of the drug in the infant’s gut; the absorption rate; the maturity of the infant’s liver; and the infants volume of distribution. The concentration of psychotropic drugs and their metabolites in infants’ serum has seldom been measured, but where it is has, it has often fallen below the limit of detection of the estimation method used.
Overall, available studies of drug transfer into milk suggest that, compared with pregnancy, much smaller amounts of most psychotropic drugs are ingested by the offspring during breastfeeding although there are significant differences between drugs. One standardized measure of exposure that is frequently used in pediatric practice is the relative infant dose. This is defined as the daily amount of drug ingested by the infant during exclusive breastfeeding per kg bodyweight divided by the maternal daily dose per kg body weight. A value of more than 10% is regarded as indicating a higher probability of side effects in the infant (Bennett, 1996). Research into the potential behavioral and physical effects of psychotropic drugs to which infants are exposed to during breastfeeding is scarce and suffers from perhaps the most methodological problems in the field of reproductive psychopharmacology. Observations of infants involve isolated cases or case series and have not usually used standardized assessment tools or timing relative to drug exposure or control groups. Long-term effects have not been investigated.
Antidepressant drugs
Antidepressant drugs (ADs) are prescribed not only for unipolar depression, but also for a number of other psychiatric conditions, such as anxiety disorders, post-traumatic stress disorders, obsessive-compulsive disorder and bipolar depression. It has been estimated that up to 3% of women in Europe and 13% of women in the United States use antidepressant drugs during pregnancy. The choice of antidepressant is similar to that in other psychiatric patients; selective serotonin re-uptake inhibitors (SSRIs) are the most commonly prescribed. In a recent UK study, prescription rates decreased from 4.7% in the 3 months to conception to 2.8%, 1.3% and 1.3% in the first, second and third trimester and increased again to 5.5% in the first 3 months postpartum (Margulis et al., 2014). A large proportion of women (79.6%) discontinued ADs at some stage in pregnancy whereas only 0.4% of those who had not taken them before pregnancy commenced them in the third trimester.
Pharmacokinetics of antidepressants in pregnancy
The net result of pregnancy-induced pharmacokinetic changes on the pharmacokinetics of AD varies greatly among pregnant women, with some but not all experiencing faster AD metabolism in late pregnancy (for review see Deligiannidis et al., 2014). The lack of systematic studies of the relationship between drug levels and clinical response in pregnancy does not allow for any firm conclusions to be drawn for clinical practice.
All antidepressants readily cross the placenta, but to differing degrees. Existing studies of maternal serum and umbilical cord serum concentrations of SSRIs and their metabolites consist of descriptions of cases or case series with a tendency of fetal:maternal ratios at delivery being the lowest for sertraline and at the higher end for citalopram. Little is known about differences among tricyclic (TCAs) and other drugs.
The reproductive safety of antidepressant drugs in early pregnancy
An abundance of literature exists on the use of antidepressants during pregnancy, notably of SSRIs as the most commonly used drugs. A review of the literature on the association between antidepressant use and any infant congenital malformations and neonatal morbidity was recently published (Källén et al., 2013). The recently published analysis by Grigoriadis et al. (2013) found no association between exposure to SRRIs and other antidepressants and overall congenital anomalies or major congenital anomalies, but a small increase in the risk of cardiovascular malformations with an odds ratio of 1.40 (CI 1.10-1.77). A small effect was found specifically for paroxetine (RR = 1.43; CI 1.08-1.88). The meta-analysis of Myles et al. (2013) looked at four SSRIs individually. Fluoxetine and paroxetine were associated with an increased risk of major malformations (ORs of 1.14, CI 1.01-1.30 and 1.29, CI 1.11-1.49) and paroxetine was associated with a small excess of cardiovascular anomalies (1.44, CI1.12-1.86). In contrast, sertraline and citalopram were not linked with any teratogenic potential.
Although TCAs were introduced in the 1950s, comparatively little has been published on these agents and sample sizes have been too small to test reliably for an increase in congenital malformation rate above the general population rate. One notable signal, however, comes from two large Swedish epidemiological studies which have linked clomipramine use with a small increase in the risk for cardiovascular defects (Källén and Otterblad Olausson, 2006; Reis and Källén, 2010).
The use of antidepressants in late pregnancy
Many studies have shown an increased risk for infant morbidity after antidepressant exposure during the latter part of pregnancy. Delivery results in a neonatal antidepressant withdrawal syndrome that is fairly common, occurring in about 30% of exposed neonates compared to only about 10% of nonexposed children (Kieviet et al., 2013) which represents a three-fold increase over unexposed newborns (Moses-Kolko et al., 2005). The syndrome includes feeding difficulties, irritability, abnormal crying, jitteriness and cerebral excitation. Respiratory distress and tremors are also more common (Grigoriadis et al., 2013). In general, the neonatal effects are transient; most symptoms develop within 48 hours after birth and last for 2–6 days (Kieviet et al., 2013).
An increased risk for low Apgar score, hypoglycemia, prematurity and low birth weight (but not small for gestational age) has been reported (e.g., Källén et al., 2013). In a recent meta-analysis of published studies on the effects of antidepressant use in pregnancy on the preterm birth rate, Huybrechts et al. (2014) found an increased odds ratio (adjusted OR 1.53, CI 1.40 – 1.66), which was still significant when a diagnosis of depression was taken into account. The effect was significant if the exposure occurred in the second and third trimester but not in the first. Huang et al. (2014) confirmed in their meta-analysis an effect on preterm birth but also found evidence for lower birth weight.
Persistent hypertension of the newborn (PPHN) is a rare but potentially fatal complication, with a population incidence of 1–2 cases per 1000 live births (Walsh-Sukys et al., 2000). Several studies have suggested an increased risk for PPHN after maternal use of SSRIs late in pregnancy. Grigoriadis et al. (2014) conducted a meta-analysis and confirmed an effect of late pregnancy exposure, which was still significant after moderator variables were taken into account. Although the odds ratio was 2.50 (CI 1.32 to 4.73), the risk is still small for any given pregnancy considering the base rate. So far TCAs and other antidepressants have not been related to PPHN.
Some other pregnancy complications and maternal delivery diagnoses have been reported to occur more frequently after treatment with SSRIs, TCAs and other antidepressants. Examples are an increased risk for preeclampsia, hyperemesis and intrapartum bleeding, which was seen in a large Swedish sample (Reis and Källén, 2010). Palmsten et al. (2013) found a significant risk increase of postpartum hemorrhage. Moreover, an association between the use of SSRI and an increased risk of gestational hypertension and preeclampsia has been described (Toh et al., 2009). In women using antidepressants, however, factors that predispose to these conditions, such as higher maternal age, smoking, higher body mass index and the use of other drugs (Reis and Källén, 2010) are also more commonly present, suggesting that further research on these outcomes is required.
Neurodevelopment in children with intrauterine exposure to antidepressants
Nulman et al. (2012) conducted a small prospective study of children between 1.5 and 6 years old who were exposed to TCAs, SSRIs, venlafaxine or untreated maternal depression during fetal development. An unexposed control group of children with healthy mothers was included. An effect of antidepressant drug exposure on the children’s cognitive development or behavior was not found. Rather, the authors emphasized that untreated depression during pregnancy was associated with a high risk for postpartum depression. Further, fetal and childhood exposure to maternal depression was a predictor of child behavioral problems and long-term child psychopathology.
A concern was recently raised about the effects of intrauterine SSRI exposure on the social development of offspring. Of five epidemiological and case control studies, three found a significantly increased rate of autism spectrum disorder or autistic traits with odds ratios ranging between 1.9 and 3.3 (Croen et al., 2011; Rai et al., 2013; El Marroun et al., 2014). However, two studies did not find an association when the maternal mood disorder was controlled for (Sørensen et al., 2013; Clements et al., 2014). Interestingly, in one of these studies (Clements et al., 2014) exposed children had the same rate as their unexposed siblings, suggesting that there is a genetic link between maternal depression and autism spectrum disorders and that antidepressant therapy is not causally related. Further family genetic studies are needed to clarify this important question.
Antidepressants and breastfeeding
Most antidepressants are excreted in low concentrations in breast milk (Hale, 2012). Paroxetine and sertraline produce low relative infant doses in the 0.5 to 3% range, while fluoxetine, venlafaxine, and citalopram produce milk levels closer to, and sometimes even above, the 10% limit (Chad et al., 2013). In a pooled analysis of 57 studies, Weissman et al. (2004) found usually undetectable plasma levels for nortriptyline, paroxetine and sertraline during lactation in more than 200 infants tested, whereas fluoxetine, citalopram and the metabolite of venlafaxine, O-desmethylvenlafaxine, produced measurable but low levels in some infants.
Case reports and case series have reported some adverse events in infants who ingested antidepressants via breastmilk. These studies observed nonspecific symptoms such as irritability, decreased feeding and sleep problems, which were more often reported during exposure to citalopram and fluoxetine (see review by Weissman et al., 2004). However, the observations were uncontrolled and unstandardized and definitive conclusions about differences between drugs and how they affect breastfed children cannot be drawn.
Antipsychotic drugs
Consistent with pharmacoepidemiological studies in other psychiatric patient groups, two studies from the United States and UK have shown an increase in recent years in the prescribing of atypical antipsychotics to pregnant women, whilst the use of typical antipsychotics has remained the same or decreased (Toh et al., 2013; Petersen et al., 2014). The rate of prescribing was 8/1000 any time during pregnancy in the United States, and 3/1000 any time from 6 weeks post conception in the UK (Toh et al., 2013; Petersen et al., 2014). In the UK study, prescriptions for antipsychotic medications were often discontinued or no longer claimed by pregnant patients. Only 38% of atypical and 19% of typical antipsychotics were still prescribed in the third trimester (Petersen et al., 2014).
Pharmacokinetics of antipsychotics in pregnancy
In a first study of serial plasma level estimations of antipsychotics, Windhager et al. (2014) found large decreases in three women taking aripiprazole during pregnancy, followed by an increase in levels in the early postpartum period. Clear deteriorations in mental state were not seen. No other systematic data are available on how antipsychotic drug levels may change during pregnancy and whether this would affect the course of illness.
When measuring the transplacental passage of several antipsychotic drugs at birth, Newport et al. (2007) found the highest fetal-to-maternal serum concentration ratio for olanzapine, followed by haloperidol, risperidone and quetiapine. Whether differences in placental transfer translate into differences in infant outcome is currently unknown.
Congenital anomalies associated with first trimester exposure to antipsychotics
When considering potential teratogenicity, the challenges in existing datasets outlined earlier mean that most antipsychotic drugs have not been studied as individual agents but within groups. In a meta-analysis of (predominantly) prospective cohort studies, first-trimester use of phenothiazines was associated with a small but significant excess of congenital anomalies in infants exposed to phenothiazines as a group (odds ratio: 1.21, 95% CI 1.01–1.45), but there was no association with a specific anomaly or a specific agent (Altshuler et al., 1996). In the population study by Reis and Källen (2008), there was a nonsignificant trend for an association of antipsychotic drugs (mostly first generation, but second generation agents were included as well) with an excess of anomalies (odds ratio 1.45, CI 0.99-1.41). The authors noted that the excess seemed to be due to cardiovascular anomalies, and in particular ventricular and atrial septum defects. In a prospective cohort study of women contacting a teratology information service (Habermann et al., 2013) only second generation antipsychotics were associated with an increased risk of congenital malformations (OR 3.21, CI 1.34-7.67) and again there seemed to be an excess of septal defects which mostly occurred as isolated anomalies. In these three studies, limited adjustments were made for confounding factors, which are known to be associated independently with cardiovascular anomalies and which are more prevalent in women who suffer from psychosis. No conclusions can be drawn in respect of individual antipsychotic agents as data are too limited (Barnes et al., 2011).
The literature indicates therefore that antipsychotic therapy is associated with a small increase in the risk of cardiovascular (mainly septal) defects. However, it is at present uncertain whether this is a causal relationship and how clinically significant these findings are, or whether there are differences among antipsychotics.
Other pregnancy outcomes
Obesity and diabetes mellitus are two side effects of antipsychotic medication that are particularly important in the context of childbearing since they are independently associated with adverse pregnancy and infant outcomes (Aberg et al., 2001; Sebire et al., 2001; Owens et al., 2010). There are differences among antipsychotic drugs in their propensity to cause weight gain and increases in blood glucose levels, with olanzapine and clozapine having the greatest impact (Rummel-Kluge et al., 2010).
The rate of diabetes increases from the second trimester onwards and antipsychotic medication may amplify this effect. In two population studies from Sweden (Reis and Källén, 2008; Bodén et al., 2012a), antipsychotic therapy was indeed associated with a significant, almost two-fold, increase of gestational diabetes. Bodén et al. (2012a) also controlled for confounding factors, and when they took early pregnancy BMI into account, the odds ratios for gestational diabetes were slightly attenuated and no longer significant. Gestational exposure was based on pharmacy claims data in this study and the duration and precise timing in pregnancy was not reported. Despite this, it is noteworthy that the two epidemiological studies found similar odds ratios.
There is evidence that antipsychotic therapy is associated with an increased rate of preterm birth (Reis and Källén, 2008) with more support for an effect of first-generation agents than second-generation agents (Diav-Citrin et al., 2005; Bodén et al., 2012a; Lin et al., 2010; Newham et al., 2008). Whether severe mental illness itself contributes to early delivery is unclear (Bodén et al. 2012a; Lin et al., 2010). Studies of the effect of antipsychotic drugs on infant birth weight, length and head circumference have resulted in conflicting findings. Although most of these studies controlled for other factors that influence birth weight, the inconsistent results may be due to differences in their relative balance, the diagnostic profile and other population characteristics. In two epidemiological studies, there was no increased risk of stillbirth following the use of antipsychotics in pregnancy (Reis and Källén, 2008; Bodén et al., 2012b).
Neonatal effects
Motor or withdrawal symptoms have been described in neonates following late pregnancy exposure to antipsychotic medication in case reports, retrospective case series and small prospective observational studies (American College of Obstetricians and Gynecologists, 2008; Gentile, 2010; Gilad et al., 2011). In a search of their spontaneous Adverse Event Reporting System Database the US Food and Drug Administration (2011) reported that, up to 2008, 69 cases of neonatal extrapyramidal or withdrawal symptoms after late pregnancy antipsychotic exposure had been identified. Some neonates recovered within hours or days without specific treatment, while others required intensive care unit support and prolonged hospitalization. In the majority of cases, there were confounding factors, including concomitant use of other drugs known to be associated with withdrawal symptoms, prematurity, congenital malformations and obstetric and perinatal complications. However, there were some cases where these symptoms were associated with antipsychotic medication alone. Habermann et al. (2013) found an increased risk of discontinuation symptoms in their prospectively followed cohort of exposed and control children, but whether the result remained significant after accounting for concomitant medication was not reported.
Because of these concerns, the Federal Drug Administration in the United States added information about neonatal effects to the pregnancy risk labeling for all antipsychotic drugs in 2011. In the UK, the Medicines and Health Care Products Regulatory Agency (2011) also issued advice to healthcare professionals to examine neonates for discontinuation symptoms.
Neurodevelopmental effects
Data on neurodevelopmental effects of antipsychotics are sparse. In a prospective study of 2,141 children exposed to phenothiazines in the first 4 lunar months of pregnancy and 26,217 unexposed children, there were no differences in IQ in children with heavy, intermediate and no intrauterine exposure (Slone et al., 1977). Many of these mothers were mentally well and took phenothiazines for nausea. A different result was found in a recent prospective study. Johnson et al. (2012) examined the effects of psychotropic exposure in pregnancy on neuromotor performance in 6-month-old infants. All mothers had a psychiatric history with predominantly affective disorders but they differed in medication status in pregnancy. Infants prenatally exposed to first or second generation antipsychotic drugs demonstrated significantly lower scores compared with both antidepressant-exposed infants and infants with no psychotropic exposure. Results were also significantly associated with variables relating to maternal psychiatric history, including depression, psychosis and overall severity or chronicity. However, the study sample was very small and the majority of women taking antipsychotics during pregnancy were also co-prescribed antidepressants, anxiolytics or hypnotics. In a well-designed, prospective study of 76 infants whose mothers had taken first or second generation antipsychotics in pregnancy and 76 unexposed children of mothers without severe mental illness, the Bayley’s neurodevelopment scores indicated delay in several domains in the early postnatal months in the exposed children (Peng et al., 2013). However, these differences resolved by the end of the first year, indicating that any potential adverse effects were only temporary.
These studies suggest that there are no long-term neurodevelopmental disadvantages arising from antipsychotic exposure in pregnancy. However, the database remains small and further prospective investigations of monotherapy exposure with larger samples, longer follow up and measures of potential confounders (e.g., maternal physical and mental health, socioeconomic status, social environment) are needed.
Antipsychotic drugs and breastfeeding
Based on the best available published evidence, Hale (2012) summarized relative infant doses for several frequently used antipsychotics. The small case numbers to date suggest relative infant doses of much less than 10% except for amisulpride, haloperidol, sulpiride and risperidone, where doses can approach or exceed 10% (Hale, 2012; Teoh et al., 2011). In published clinical observations of breastfed infants, relatively few adverse effects have been reported (Gentile, 2008; Dayan et al., 2011) although definite conclusions cannot be drawn because of the small numbers of infants observed and a lack of standardized assessments. Dev and Krupp (1995) reported agranulocytosis in one breastfed infant whose mother took clozapine.
Antiepileptic mood stabilizers
The antiepileptic mood stabilizers valproate, carbamazepine and lamotrigine are used for the treatment of bipolar disorder in acute phases of illness and in maintenance therapy. Over the last two decades, there has been a large decline in the UK in the use of carbamazepine and sodium valproate in pregnant women; lamotrigine is now the most commonly prescribed antiepileptic in this patient group with 0.25% of pregnancies affected (Man et al., 2012). Similar trends have been reported for other countries (Man et al., 2012). Little data exist about prescribing of antiepileptic mood stabilizers specifically to pregnant women with psychiatric disorders, but it is of concern that a nationally representative survey in the United States recently reported no decline in the recent prescribing of valproate to women of childbearing age with nonepileptic conditions (Adedinsewo et al., 2013).
Pharmacokinetics of antiepileptic medication in pregnancy
As pregnancy progress, the serum concentrations of total and free lamotrigine decline markedly. This may be because of pregnancy-induced alterations in the activity of glucuronosyl transferases or renal function. The extent is highly variable with values of 248% being reported (Pennell et al., 2008; Tomson et al., 2013). The return to the nonpregnant baseline occurs within the first month postpartum. In a prospective study of pregnant women with epilepsy on lamotrigine monotherapy, seizure control significantly decreased when serum lamotrigine concentrations had fallen by more than 35% from their optimal serum concentration before pregnancy (Pennell et al., 2008). In a case series of bipolar patients, Clark et al. (2013) found that several women required dose increases of lamotrigine in pregnancy. However, the threshold at which deterioration in mental state becomes likely has not been explored.
The changes in total and unbound carbamazepine levels seem to be small in pregnancy (Tomson et al., 2013). Little reliable data are as yet available in respect of the pharmacokinetics of valproate that could inform clinical practice (Tomson et al., 2013).
The extent to which total valproate is transferred from the maternal to the fetal circulation is high, whereas the concentration of biologically active free fraction is lower in the fetal than the maternal circulation with a ratios between 0.5 and 0.8 (see review by Wieck, 2011). A reverse relationship exists for carbamazepine where the ratio for the free fraction is higher and ranges between 1.0 and 1.4 (for review see Wieck, 2011). About 90% of total lamotrigine is transferred to the fetus (for review see Wieck, 2011).
Major congenital anomalies after pregnancy exposure to antiepileptic drugs
It has been known for several decades that maternal use of valproate is associated with congenital anomalies. In their systematic review and meta-analysis of cohort and pregnancy register studies, Meador et al. (2008) found an anomaly rate for monotherapy with valproate in the first trimester of 10.7% which was much higher than among the offspring of healthy control mothers (3.3%). Among specific anomalies, significant differences were found for spina bifida with a particularly high odds ratio (12.7), atrial septal defects, cleft palate, hypospadias, polydactyly and craniosynostosis (Jentink et al., 2010a). A relationship of the teratogenic effect with daily dose has been widely described although the critical threshold has varied between daily doses of 600–1500 mg (Tomson and Battino, 2012).
Carbamazepine and lamotrigine have consistently been shown to be less teratogenic than valproate, and in the study by Meador et al. (2008) they were associated with congenital malformation rates similar to those of children of healthy mothers. However, carbamazepine has been implicated in an increased risk of spina bifida (Matalon et al., 2002; Jentink et al., 2010b) with an odds ratio of 2.6 compared to no antiepileptic exposure (Jentink et al., 2010b). An earlier finding of an increased rate of oral clefts following pregnancy exposure to lamotrigine has not been substantiated by subsequent studies (Tomson and Battino, 2012).
The general population incidence of neural tube defects and several other congenital anomalies can be markedly reduced by folic acid use from preconception to early pregnancy. However, clinical guidelines are misleading if they recommend high-dose folic acid (4–5 mg daily) to women taking antiepileptic drugs since there is little evidence that this will prevent their teratogenic effect (Wlodarczyk et al., 2012).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
