Exposure to Lamotrigine monotherapy in pregnancy : Result from 18 years of international Lamotrigine pregnancy registry

Introduction

Scientifically valid data on the safety of drug use during pregnancy are a significant public health need. Data are rarely available on the fetal effects of in utero exposure in human pregnancies, particularly when a drug is first marketed. Data from animal reproductive toxicology studies, which function as a screen for potential human teratogenicity, are usually all that is available in a product's labelling. For practising clinicians, translating known animal risks into an accurate assessment of teratogenic risks in their patients is very difficult, if not impossible. Without human data on the effects of in utero drug exposure, it is difficult for physicians and other healthcare providers (e.g., genetic counsellors) to adequately counsel patients about fetal risks. Therefore, a pregnant woman may decide to unnecessarily terminate a wanted pregnancy or forego needed drug therapy. In spite of the lack of data on the safety of drug use during human pregnancies, pregnant women are exposed to drugs either as prescribed therapy or inadvertently before pregnancy is known (over one-half of pregnancies are unplanned). Because little is known about the teratogenic potential of a drug in humans before marketing, post-marketing surveillance of drug use in pregnancy is critical to the detection of drug-induced fetal effects. The existing passive mechanism of spontaneous reporting of adverse drug effects is inadequate to routinely detect drug-induced fetal risks or lack of such risks. Therefore, post-marketing pregnancy exposure registries are being increasingly used to proactively monitor for major fetal effects and to describe margins of safety associated with drug exposure during pregnancy. The 1990s saw the establishment of several pregnancy registries to monitor for the potential teratogenic effect of both the older and newer generation of antiepileptic drugs (AEDs). Among the first of these studies to be initiated was the International Lamotrigine Pregnancy Registry. While teratology studies in mice, rats, and rabbits, using oral doses up to 10, 3, and 4 times that of upper dose in humans, respectively, failed to provide evidence of teratogenicity, Lamotrigine was shown to decrease folate levels in rats. The latter has been associated with teratogenicity in both animals and humans.

Lamotrigine is an antiepileptic medication indicated for oral use as adjunctive therapy in the control of partial seizures with or without generalized tonic-clonic seizures. It is also used as a monotherapy in a number of countries. Lamotrigine is a drug of the phenyltriazine class and is chemically unrelated to existing antiepileptic medications. Teratology studies were conducted in mice, rats, and rabbits at oral doses up to 10, 3, and 4 times the upper human dose (500 mg/day or 7 mg/kg/day), respectively, and revealed no evidence of teratogenicity. Teratology studies were also conducted using bolus intravenous (i.v.) administration of the isethionate salt of Lamotrigine in multiples of the projected human dose. Intravenous Lamotrigine resulted in convulsions or impaired coordination in rat and rabbit dams at 30 mg/kg and 15 mg/kg, respectively. Clinical data indicate that Lamotrigine has no effect on blood folate concentrations in adults; however, the effects of Lamotrigine on fetal blood folate levels in utero are unknown. Animal reproduction studies are not always predictive of human response; therefore this drug should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus.

With the aim of monitoring the frequency of major congenital malformations (MCMs) among Lamotrigine-exposed pregnancies, in 1992, International Lamotrigine Pregnancy Registry was established. A surveillance approach was followed that depended upon voluntary enrollment and prospective follow-up and used reference data on the expected frequency of MCMs from US general population samples and cohorts of women exposed to AED monotherapy. The anticipated sample size lacked the power to determine with any precision the frequency of specific congenital malformation types, though malformation patterns were monitored to determine whether additional appropriately designed studies should be initiated.

Study Objective

Objective of present study was to monitor for a signal for major teratogenicity following in utero Lamotrigine exposure.

Study Design

Data Collection

Pregnant women exposed to Lamotrigine, from anywhere in the world, were enrolled on a voluntary basis by their health care provider via phone, facsimile, or mail. The timing of enrollment during pregnancy varied, but was encouraged as early in pregnancy as possible. The reasons for this were several fold: to maximize prospective follow-up and to minimize the potential for bias due to knowledge of the pregnancy outcome from prenatal screening; to increase the accuracy of exposure and pregnancy details with reporting close to the time of exposure; and to maximize information on exposure during the first trimester of pregnancy representing the period of organogenesis. The information sought at enrollment included maternal demographics, exposure details (timing and dose), pregnancy details (last menstrual period or expected delivery date), details of prenatal testing, and family history of epilepsy.

Outcome Classification

Based upon the pregnancy outcome information obtained during follow-up, the Scientific Advisory Committee classified each pregnancy as with a MCM, without a MCM, or a spontaneous pregnancy loss. Pregnancy outcomes with a MCM were further categorized as live births, fetal deaths, and induced abortions. Retrospectively reported pregnancy outcomes, where the pregnancy outcome was known at the time of enrollment, were also reviewed by the Scientific Advisory Committee for patterns of reported defect types, but were not included in analyses. Retrospective reports included malformations identified by prenatal screening prior to enrollment in the Registry. Full details of the retrospective reports can be found in the final Registry report. To provide consistency in the definition of MCMs, the Registry adopted the US Centers for Disease Control and Prevention's (CDC) Metropolitan Atlanta Congenital Defects Program (MACDP) case definition for congenital malformations, though all reported malformations were reviewed by the clinicians on the Scientific Advisory Committee, with whom the final classification rested. All malformations were listed in the interim and final Registry reports, though only MCMs were included in analyses. As access to pediatric evaluations and records to obtain longer term follow-up information was beyond the scope of the Registry, the findings were restricted to those malformations that were identifiable shortly after birth and before hospital discharge. The Registry was not able to systematically include minor abnormalities and dysmorphisms due to inconsistent recognition.

The analyses included all MCMs identified after enrollment (i.e., prospectively identified) and meeting MACDP inclusion criteria regardless of whether the infant was live born. Pregnancies ending in a stillbirth or elective termination without a malformation reported, and all spontaneous pregnancy losses, were excluded from analyses given the potential for inconsistent identification of malformations in those circumstances. This was considered a conservative approach, likely to overestimate, rather than underestimate, the proportion of infants with a MCM following in utero Lamotrigine exposure. Chromosomal defects were also excluded as these were unlikely to be associated with drug exposure. The proportion of infants with MCMs was calculated as follows:

Number of pregnancy outcomes with MCMs / [(number of pregnancy outcomes with MCMs) + (number of live births without MCMs)].
The accompanying 95% confidence intervals (CIs) were calculated using the Fleiss method (normal approximation of the binomial distribution with continuity correction). The proportion of infants with a MCM was stratified by earliest trimester of exposure (second trimester beginning at 14 weeks and third trimester at 28 weeks) and by Lamotrigine therapy type reported at enrollment (monotherapy and polytherapy with and without valproate).
The potential effect of dose on the proportion of infants with a MCM was investigated among first- trimester monotherapy exposures using logistic regression with maximal first-trimester dose level (100 mg per day increments) as the independent variable and MCM status as the dependent variable. The odds ratio (OR) and 90% confidence intervals were based on Wald statistics. Dosage information was based on prescribed AED dose, as reported by the health care provider at enrolment and follow- up, rather than serum levels.

Study Results

As of Registry closure on March 31, 2010, 3,416 Lamotrigine-exposed pregnancies had been enrolled prospectively from across 43 countries. Of those, 2,444 had known outcomes and 972 (28.5%) had been lost to follow-up. The majority of pregnancies with known outcomes were enrolled from the United States (65.1%). A further 5 countries (Poland, United Kingdom, Germany, Sweden, and Denmark) recruited more than 50 pregnancies each, constituting a further 20.0% of the pregnancies with known outcomes.

The 2,444 Lamotrigine-exposed pregnancies yielded 2,492 infants including 43 sets of twins and one set each of triplets and quadruplets. Of those infants, 1,817 had been exposed to Lamotrigine monotherapy, 173 to Lamotrigine polytherapy including valproate, and 502 to Lamotrigine polytherapy without valproate. For the majority, the earliest exposure was during the first trimester of pregnancy: 93.5% of monotherapy, 93.1% of polytherapy with valproate, and 94.4% of those without valproate exposures. The classification of outcomes for the 2,492 infants is described in Table 1.

Lamotrigine Monotherapy

Of the 1,817 Lamotrigine monotherapy exposed infants, 1,699 were exposed during the first trimester of pregnancy. Of those, 141 were excluded from analyses due to the pregnancies ending in a fetal death, elective termination, or spontaneous loss without a reported malformation. Those outcomes were excluded due to the potential for inconsistent identification of malformations in those situations and hence uncertainty that a malformation was truly absent. Among the remaining 1,558 infants, there were 35 infants with major congenital malformations: 2.2% (95% CI 1.6%±3.1%). The observed MCMs, which ranged in

Table 1: Lamotrigine exposure in pregnancy by earliest trimester of exposure and pregnancy outcome

21956-1

type and severity, are listed in Table 2 according to organ system. The Scientific Advisory Committee noted the presence of 3 cases of anencephaly, all ending in elective terminations, and several severe obstructive cardiac defects. Efforts to obtain additional information to aid case interpretation, for example a family history of similar defects, were unsuccessful.

Lamotrigine Polytherapy

Among the 150 analysable pregnancy outcomes with earliest exposure to Lamotrigine polytherapy with valproate in the first trimester, there were 16 outcomes with MCMs: 10.7% (95% CI 6.4%±17.0%). Four of the infants with MCMs had an orofacial cleft defect. Among the 430 analyzable pregnancy outcomes with earliest exposure to Lamotrigine polytherapy without valproate in the first trimester there were 12 outcomes with MCMs: 2.8% (95% CI 1.5%±5.0%).

Proportion of Infants with MCM by Dose

A potential trend in the proportion of infants with a MCM by dose was explored within the 1,523 first-trimester monotherapy exposures for which dose information was available. The distribution of maximal first-trimester dose exposures (i.e., highest dose recorded during first trimester), and the corresponding MCM frequencies, is given in Table 3. There was no evidence of an increase in the proportion of infants with a MCM with increasing dose (OR per 100 mg increment: 1.0, 90% CI 1.0±1.0). There were, however, relatively few exposures above 600 mg per day.

Table 2: Description of specific malformation types following first-trimester lamotrigine exposure in pregnancy by therapy type

image-1 (1)

Table 3: Proportion of infants exposed to lamotrigine monotherapy with major congenital malformations by first trimester maximal dose exposure

Study Analysis

The purpose of the Registry is two-fold: a) to assess whether there is any large risk of major malformations following exposure to Lamotrigine in pregnancy and b) to provide information on outcomes following pregnancy exposure to Lamotrigine so that patients and physicians can best determine how to manage pregnancies. The combination of the large number of women with epilepsy who are of reproductive capacity and the lack of data concerning Lamotrigine use during pregnancy makes such a Registry an essential component of the ongoing program of epidemiologic studies of the safety of Lamotrigine.

This Registry was considered essential because of the potential for exposure in the first trimester of pregnancy, the unknown risks in pregnancy for any new chemical entity, the known teratogenicity of specific existing anticonvulsants, and the increased risk of teratogenicity with polytherapy. Because Lamotrigine is used to treat women with epilepsy, results from this Registry were compared with published data on women with epilepsy who did not take Lamotrigine. This study is an observational, exposure-registration, and follow-up study. Patient confidentiality is strictly upheld. Furthermore, the Registry initiated a registration process which will protect patient anonymity at the Registry Office. Over an 18-year period, the International Lamotrigine Pregnancy Registry collected pregnancy outcome data on over 1,500 infants exposed to Lamotrigine monotherapy in the first trimester of pregnancy. There were 35 infants with MCMs observed following 1,558 first-trimester monotherapy exposures, a frequency of 2.2% (95% CI 1.6%±3.1%).

Although the lack of an internal control group complicates interpretation of this estimate, and precludes the calculation of relative risk estimates, the proportion of Lamotrigine-exposed infants with a MCM, and the associated confidence limits, is consistent with estimates from general population cohorts: the proportion of infants with MCMs detected by the end of the first week of life in the Metropolitan Atlanta population between 1968 and 2003 was 2.1% and the Brigham and Women's Hospital Surveillance Program reported a MCM frequency of 1.6%±2.2% at birth depending on whether chromosomal and genetic abnormalities were included. Although direct comparisons may result in biased risk differences given methodologic differences in data collection and population differences, cohorts of women exposed to AED monotherapy in the literature have yielded slightly higher point estimates of the proportion of infants with MCMs: 3.3%±4.5%. These descriptive comparisons yield no signal of a substantial increase in MCM frequency in association with in utero Lamotrigine exposure. Three cases of anencephaly, all ending in elective terminations, were observed among the 1,558 first-trimester monotherapy exposures. This is higher than the prevalence estimates of 0.19±0.68 per 1,000 live births and stillbirths and elective terminations reported through population-based studies. However, a review of 5 other ongoing AED pregnancy registries, with denominators ranging from 51 to over 1,000 first-trimester Lamotrigine monotherapy exposures, failed to support the findings of the International Lamotrigine Pregnancy Registry: a single anencephaly case was reported to the North American Anti-Epileptic Drug (NAAED) Registry and proved a duplicate case matched on maternal age, Lamotrigine dose, and the date of the termination (L.B. Holmes, personal communication, 2009). Several cases of severe cardiac defects were also noted by the Scientific Advisory Committee. While attempts to obtain additional information to aid case interpretation were unsuccessful, a review of data from other ongoing AED registries again failed to support an increased frequency of cardiac defects following in utero Lamotrigine exposure.

Earlier, through the NAAED and the Swedish Medical Birth Register signal for an increased rate of orofacial clefts among Lamotrigine-exposed pregnancies was reported. This apparent increase was not observed in the International Lamotrigine Pregnancy Registry. A case-control study, with increased power to detect associations of exposures with rare outcomes, was mounted in the European Congenital Anomalies and Twins Registers (EUROCAT) system to specifically test the hypothesis of a potential association between Lamotrigine and orofacial clefts. An OR for all isolated clefts of 0.8 (95% CI 0.1±2.9) compared to other non- cleft defects was reported following first-trimester Lamotrigine monotherapy exposure. The EUROCAT study is continuing recruitment to increase its power to assess the possible association between Lamotrigine and orofacial clefts. The EUROCAT framework will also be used to assess any potential associations between Lamotrigine and neural tube defects and cardiac defects. Although there is no evidence of teratogenicity from preclinical studies of Lamotrigine, this Registry was managed as a part of an ongoing program in epidemiologic safety monitoring.

The Lamotrigine Pregnancy Registry Advisory Committee noted the higher frequency of major malformations within the group exposed to the combinations including valproate compared with other polytherapies or compared with Lamotrigine monotherapy.

The observed frequency of major defects (2.8% in women exposed to Lamotrigine and at least one other antiepileptic drug, excluding valproate, and 10.7% in women exposed to Lamotrigine and at least one other antiepileptic drug including valproate) is consistent with published studies which report that women using valproate have experienced elevated rates of birth defects. UK Epilepsy and Pregnancy Register have reported a positive dose-response effect for MCMs with Lamotrigine use was previously. The International Lamotrigine Pregnancy Registry's final results failed to confirm a positive dose-response relationship.

However, there were a limited number of first-trimester exposures above 600 mg per day. Interpretation is further complicated by the fact that serum level data were not available, data pertinent to Lamotrigine given the increase in Lamotrigine clearance during pregnancy and consequent decreased serum levels, though this caveat is common to all AED registries completing such analyses. The proportion of infants with a MCM following first-trimester exposure to Lamotrigine polytherapy including valproate (10.7%, 95% CI 6.4%±17.0%) was elevated compared with the proportions associated with Lamotrigine polytherapy without valproate and monotherapy exposure.

While this increased frequency is consistent with that reported for valproate monotherapy in the literature, it is beyond the scope of this Registry to determine the relative contributions of the constituent parts of the combination therapy. This is further complicated by the potential for confounding by the number and severity of seizures inextricably linked to therapy choice. Several factors may have introduced bias into the calculation of the proportion of infants with a MCM in the International Lamotrigine Pregnancy Registry.

The voluntary nature of recruitment may have introduced an inherent sampling bias, favouring high- or low- risk pregnancies, and is likely to underlie the high loss to follow-up rate (28.5% of enrolled pregnancies). An incentive scheme for complete reporting was introduced during the later years of the Registry, but did little to reduce loss to follow-up, leaving the possibility of bias if the pregnancies lost to follow-up systematically differed in terms of outcome from pregnancies with outcome information.

Due to prior knowledge of pregnancy outcome, the Registry concentrated on prospectively enrolled pregnancies to minimize bias. However, 40% of pregnancies exposed to monotherapy during the first trimester were enrolled after 16 weeks' gestation, introducing the potential for bias associated with knowledge of fetal outcome due to prenatal testing. This bias could only be removed by restricting recruitment of pregnancies to the period preceding prenatal testing, as is the case in some registries. The limited follow-up of infants after birth may also have introduced differential reporting by malformation type with the under-representation of malformations likely to be identified after several months, for example some heart malformations.

Despite these study limitations, confidence in data from the International Lamotrigine Pregnancy Registry is increased given consistency in data emerging from several large AED registries, each capturing at least 500 first-trimester monotherapy exposures. The proportion of infants with a MCM following first-trimester monotherapy exposure was 2.2% (95% CI 1.6%±3.1%) in the International Lamotrigine Pregnancy Registry, which is close to estimates from the NAAED Registry (2.3%, 95% CI 1.3%±3.8%), the UK Epilepsy and Pregnancy Register (3.2%, 95% CI 2.1%±4.9%), and the International Registry for Anti-epileptics in Pregnancy (EURAP) (2.9%, 95% CI 2.1%±4.1%). Due to a lack of personal identifiers across Registry datasets, to protect patient confidentiality, researchers cannot accurately estimate duplicate enrollment of pregnancies across these registries. Consistency of results is encouraging across geographically independent regions.

A number of ongoing pregnancy registries were designed to quantify pregnancy outcomes in women with epilepsy using AEDs during pregnancy. The International Lamotrigine Pregnancy Registry collected prospective data on LTG exposures from physicians throughout the world; the U.K. AED Pregnancy Register prospectively collected pregnancy-outcome data on pregnancies exposed to any AED from U.K. neurologists; and the North American AED Pregnancy Registry collected data on any North American AED pregnancy exposures in women with epilepsy from the subjects themselves. In addition, the International Antiepileptic Drug and Pregnancy Registry coordinated the collection of data on pregnancies exposed to AEDs throughout the world. To facilitate comparison of results across registries, it was preferred that individual pregnancies not be intentionally enrolled into more than one registry.

Conclusion

In conclusion, the Registry did not detect an appreciable increase in MCM frequency following first trimester Lamotrigine monotherapy exposure. With over 1,500 first-trimester monotherapy exposures, the Registry was powered to detect major teratogenicity. The proportion of infants with MCMs following Lamotrigine / valproate polytherapy exposure was high, but similar to that previously reported with valproate monotherapy. The Registry failed to observe an increased MCM frequency with increasing Lamotrigine dose. Monitoring of specific malformations among Lamotrigine-exposed pregnancies will continue through case-control surveillance in the European Congenital Anomalies and Twins Registers network.

Neurology. 2011; 76: 1817-23

Epilepsy Times - Issue 2: Exposure to Lamotrigine Monotherapy during Pregnancy is Safe: Results from 18 Years of International Lamotrigine Pregnancy Registry