Preliminary Results on Pregnancy Outcomes in Women Using Lamotrigine
The International Lamotrigine Pregnancy Registry Scientific Advisory Committee consists of the following members: Dr. Janet Cragan, National Center on Birth Defects and Developmental Disabilities, Centers for Disease Control and Prevention (CDC), Atlanta, Georgia; Dr. Lewis Holmes, Massachusetts General Hospital, and Dr. Ellice Lieberman, Brigham and Women's Hospital, Boston, Massachusetts; Dr. John Messenheimer, CNS Clinical Research, GlaxoSmithKline, Research Triangle Park, North Carolina, U.S.A.; Dr. James I. Morrow, The Royal Victoria Hospital, Belfast, Northern Ireland; and Dr. Mark Yerby, 2455 Northwest Marshall, Suite 14, Portland, Oregon, U.S.A.
Address correspondence and reprint requests to Dr. P. Tennis at GlaxoSmithKline, Worldwide Epidemiology, Five Moore Drive, Research Triangle Park, NC 27709, U.S.A. E-mail: PST49347@gsk.com
Summary: Purpose: In 1992, the International Lamotrigine Pregnancy Registry was initiated to enroll prospectively and to monitor pregnancies exposed to lamotrigine (LTG) for the occurrence of major birth defects. This study presents results as of September 2001 on 168 outcomes exposed to LTG monotherapy and 166 outcomes after pregnancies exposed to LTG polytherapy during the first trimester.
Methods: LTG pregnancy exposures are voluntarily reported to the registry by health care providers before they are aware of each pregnancy outcome. Pregnancy-outcome ascertainment is obtained through subsequent follow-up with the reporting health care provider, and each reported birth defect is reviewed by an expert pediatrician. The percentage with major birth defects in pregnancies with known birth defect status was calculated for LTG monotherapy and for polytherapy stratified by trimester of exposure.
Results: The registry identified 334 first-trimester LTG pregnancy outcomes exposed to LTG monotherapy or polytherapy during the first trimester and involving either a live birth with or without a major birth defect or an abortion with a major birth defect. After exposure to LTG monotherapy, the percentage with major birth defects exposed to LTG monotherapy was three (1.8%) of 168 [95% confidence interval (CI), 0.5–5.5%]. There were five (10%) major birth defects observed in 50 outcomes after LTG polytherapy involving valproic acid (VPA; 95% CI, 3.7–22.6%) during the first trimester. The observed proportion of major defects after LTG polytherapy without VPA during the first trimester was five (4.3%) of 116 (95% CI, 1.6–10.3%). No specific patterns of major birth defects in any subgroup or within the registry as a whole were observed.
Conclusions: The sample sizes for individual regimens are too small to rule out small increases in frequency of all major birth defects or even large increases in frequency of rare major birth defects. However, the percentage of outcomes with major birth defects after LTG monotherapy in this study and in another similar pregnancy registry in the United Kingdom did not differ from that reported in the recent literature for women with epilepsy receiving antiepileptic drug monotherapy (4%). The frequency of major malformations after exposures of LTG–VPA is higher than that after the LTG monotherapy or LTG polytherapy regimens without VPA. Although there are published data on frequency of major malformations after VPA exposures in pregnancy, between-study differences in methods and source populations and the wide confidence intervals around the estimate for LTG and VPA limit the utility of comparison with such data, and no conclusions are made at this time about this combination. The continued registration of exposed pregnancies to an exposure registry as early as possible in the pregnancy before any knowledge of the outcome, and before any prenatal testing, will enhance the power of such data.
A number of new antiepileptic drugs (AEDs) have been introduced over the past decade. Although no AED is indicated as safe for use in pregnancy, in general, it is not feasible to discontinue AEDs for the duration of gestation (1). Even when this is the intention, the pregnancy may not be recognized for several weeks until the first missed menses, leaving the fetus exposed to AEDs during the crucial first weeks of embryonic development (2). Although most women with epilepsy have a normal pregnancy with healthy children, studies indicate an association between certain types of AEDs and an increased frequency of certain birth defects (BDs; congenital malformations). Moreover, the risk of BDs appears to increase with the use of multiple AEDs (1,3–5). For these reasons, and in view of the fact that clinical trials frequently exclude pregnant women, an observational exposure registry was initiated so that pregnancy outcomes could be monitored in patients exposed to lamotrigine (LTG) (6), an alternative to older AEDs.
LTG is a novel AED that acts by inhibiting the release of excitatory neurotransmitters, particularly glutamate, during seizures. LTG is indicated as adjunctive therapy for adults with partial seizures and as adjunctive therapy for generalized seizures of the Lennox–Gastaut syndrome. LTG also is indicated for conversion to monotherapy in adults with partial seizures who are also receiving treatment with a single enzyme-inducing AED.
The teratogenic potential of LTG has been assessed in preclinical studies using mice, rats, and rabbits. Oral doses of LTG ranging from 3 to 10 times the maximal recommended human dosage (500 mg/day or 7 mg/kg/day) produced no evidence of teratogenicity in these species, although some maternal toxicity and secondary fetal toxicity occurred at these large doses (6). LTG decreased fetal folate concentrations in the rat, and folate supplementation is known to decrease risk of BDs in humans. Clinical studies in adults have demonstrated no reduction in blood folate concentrations with LTG (7). No well-controlled clinical data are available concerning folate levels in the human fetus or in pregnant women after LTG exposure. Animal studies have found no evidence of mutagenicity or impairment of fertility with LTG. Several case reports of LTG pregnancy exposures (8–10) and a case series (11) reported no BDs in the pregnancies exposed to LTG; however, these cases were collected for other purposes and may not be representative of all exposed pregnancies.
The primary aim of the International Lamotrigine Pregnancy Registry is to monitor LTG exposures in pregnancy to detect major structural BDs that may be drug related. Because of the high potential for fetal exposure to LTG during gestation and the background BD rate associated with epilepsy and other AEDs (1,3,5,12–15), the systematic monitoring of pregnancy outcomes after exposure will yield important information for the medical community, women with epilepsy, and the sponsor. In 1992, the International Lamotrigine Pregnancy Registry was created by GlaxoSmithKline (GSK) as part of an ongoing postmarketing epidemiologic program for LTG. Managed by GlaxoSmithKline, the registry is an observational study that follows up pregnant women exposed to LTG. This publication covers all prospective data available through September 30, 2001. The registry data are intended to supplement the available preclinical, clinical, and epidemiologic data on LTG and to provide prescribing physicians and pregnant women with information to assist them in weighing the risks and benefits of taking LTG while pregnant.
Prospectively, while the pregnancy is ongoing and before the pregnancy outcome occurs, health care professionals throughout the world report pregnancies exposed to LTG to the registry on a voluntary basis. Registration may be made by telephone contact, fax, or by completing a mailed questionnaire. At the time of registration, the enrolling physician identifies the week of gestation and states whether a BD has already been detected. If the pregnancy is no longer ongoing or if a BD has already been identified, the pregnancy is not included in the registry but is maintained in another database of retrospectively reported pregnancy outcomes. In anticipation of collecting follow-up information on pregnancy outcome, the physician is given a unique patient-tracking number and provides de-identified information to safeguard patient privacy. Shortly after the expected date of birth, the registry contacts reporting professionals to ascertain the pregnancy outcome. Reported pregnancies from outside North America are funneled through the safety surveillance departments of local GSK companies into the coordinating center in the United States.
We include each prospectively reported pregnancy in the registry when information on the pregnancy outcome and the extent of LTG exposure (dose, duration of exposure, concomitant use of other AEDs) have been obtained. We define cases as lost to follow-up when three or more unsuccessful attempts to contact the reporting professional to obtain outcome information have been made or when the professional has lost contact with the patient and does not know the outcome.
For defining major BDs, the Advisory Committee use the code list of Centers for Disease Control and Prevention's Metropolitan Atlanta Congenital Defects Program (MACDP) (16) as a guide and are not blinded to the available exposure information. Final decisions about BD status and inclusion in analysis as a major BD are made by the Scientific Advisory Committee on a case-by-case basis. Any BDs included by the MACDP classification but excluded from analysis by the committee are described here. The Scientific Advisory Committee consists of independent experts in neurology, teratology, pediatrics, epidemiology, and a medical advisor and epidemiologist from the manufacturer.
This Registry adopts the following definition from BDs surveillance programs, which define a child with a BD as any live or stillborn infant, or electively terminated fetus, of any gestational age with a major structural abnormality diagnosed before age 6 years. However, outcomes are generally reported during the first year of life. Because accessing pediatric evaluations and records for follow-up information about the presence of defects is beyond the scope of its methods, the Registry primarily monitors the frequency of major defects that are external, recognizable in the delivery room, and/or symptomatic shortly after birth. Minor defects and dysmorphisms are not included in the analysis because they are not consistently ascertained, and they may be differentially reported based on exposure information. The registry disqualifies as defects those findings that are present in infants of 36 weeks gestation or less or weighing ±2,500 g and are attributable to prematurity itself. Anatomic findings observed on prenatal sonography and that would not be routinely detected by physician examination at birth (e.g., mild hydronephrosis or choroid plexus cysts) are not included as defects in the analysis but are reported if observed. Chromosomal anomalies and genetic disorders also are excluded from the analysis. Infants having transient or infectious conditions or biochemical abnormalities are classified as being without defects unless the condition is subsequently found to be a consequence of a previously unrecognized BD.
Analysis of data
We stratified pregnancy outcomes by the earliest trimester of exposure to LTG and calculated gestational weeks from the first day of the last menstrual period. For each trimester, we defined polytherapy exposure as having exposure to more than one AED during that trimester. These exposures may or may not occur concurrently during the trimester. The second trimester begins at day 1 of week 14, and the third trimester, at day 1 of week 28.
Percentage of outcomes with major BDs are stratified by trimester of exposure and by AED monotherapy versus polytherapy status and are calculated as the total number of outcomes with major BDs divided by the sum of the number of outcomes with major BDs + number of live births without major BDs. Pregnancy outcomes are categorized as those with or without major structural BDs. These are further classified as live births, spontaneous pregnancy losses, or induced abortions. All major BDs detected after registration into the study and meeting inclusion criteria, regardless of whether the infant is born alive, are included in the calculation of percentage with BDs. Because spontaneous and elective abortions are not routinely examined for defects after delivery, spontaneous abortions or terminations that do not have a BD reported to the registry are considered to have unknown BD status and are not included in analysis. The exclusion of these outcomes facilitates comparison with rates in the literature, most of which are expressed as percentage of live births, and their small number is not expected to have an impact on the rates.
For each percentage we calculate 95% confidence intervals based on Fleiss (17). The proportion of outcomes with major BDs in LTG-exposed pregnancies is compared with the percentage reported in the literature for women with epilepsy. Citations of percentages from other studies of AED exposure during pregnancy include only those that ascertain and include major BDs observed in elective or spontaneous abortions. Percentages of outcomes with both major and minor BDs are not calculated because ascertainment of minor defects is highly variable across observers (18).
The Scientific Advisory Committee review the methodology, the raw data, and tabulated results on a regular semiannual basis. Retrospective reports of exposed pregnancies for which the outcome is known at time of the first registry contact are not included in the registry. Because these reports are retrospective, their frequencies may not be representative of those in the exposed population.
Prospective data in Lamotrigine Pregnancy Registry
From September 1, 1992, through September 30, 2001, health care providers from 26 countries enrolled 492 pregnancies, 103 of which were lost to follow-up. The 389 pregnancies with available follow-up information resulted in 395 outcomes, 45 of which involved pregnancy losses without reported BDs and included one fetal death after 20 weeks of gestation. The analyses included 168 outcomes after first-trimester monotherapy exposure and 166 outcomes after first-trimester polytherapy exposure. Pregnancies were lost to follow-up for the following reasons: no response from the registering health care provider despite three or more attempts by the registry to obtain follow-up information (86%); the registering health care provider had left the practice with no forwarding address (4%); the patient had left the physician practice (6%); and the patient could not be identified by the registering health professional (4%). The majority of the reports originated from the United States (24%), the United Kingdom (21%), Sweden (9%), Denmark (8%), and Australia (6%). The remaining exposed pregnancies were reported from Austria, Belgium, Canada, Czech Republic, Finland, France, Germany, Greece, Holland, Iran, Ireland, Israel, Italy, Lebanon, Malta, New Zealand, Norway, Poland, South Africa, Spain, and Turkey.
Included in the analyses were a total of 13 prospectively identified major BDs exposed to LTG during the first trimester (Table 1). Of these, two resulted in elective abortions; three were exposed to LTG monotherapy; and 10 were exposed to LTG in combination with one or more AEDs during the first trimester. There also were three children with minor defects and dysmorphisms exposed to LTG and valproic acid (VPA). There was one additional anomaly after a first-trimester exposure to LTG and carbamazepine (CBZ) throughout the pregnancy. This child was described as having two X chromosomes and ambiguous genitalia and has yet to be more precisely classified. Because of its apparent chromosomal basis, this case was excluded from analysis. There were no reported defects detectable only on ultrasound.
Table 1. Major birth defects prospectively identified during follow-up in the Lamotrigine Pregnancy Registry
|Lamotrigine monotherapy|| || || |
| 1. Live male infant||40||Lamotrigine, 25 mg/day wk 0, 450 mg/day wk 6 through delivery. “Previously took carbamazepine”||Esophageal malformation repaired by surgery|
| 2. Live male infant||40||Lamotrigine, 200 mg/day week 0, 300 mg/day week 37 through delivery||Cleft soft palate|
| 3. Live male infant||40||Lamotrigine, 500 mg/day week 0 through delivery||Right club foot|
|Lamotrigine polytherapy|| || || |
| 4. Live male infant||40||Lamotrigine, 2,000 mg/day wk 0–7; carbamazepine, preconception and throughout pregnancy||One extra digit on one hand|
| 5. Live female infant||Unknown||Lamotrigine, 50 mg/day wk 0–40; valproic acid throughout pregnancy||Bilateral talipes|
| 6. Live male infant||37||Lamotrigine, 600 mg/day wk 0–37; phenytoin and primidone throughout pregnancy||Heart murmur and patent foramen ovale requiring banding around pulmonary artery; baby died at 3 mo after corrective surgery|
| 7. Live male infant||37||Lamotrigine, 400 mg/day wk 0–13, 600 mg/day wk 13–17, 800 mg/day week 17; gabapentin, preconception and throughout pregnancy||Skin tags on left ear; severe ear malformation, including no ear canal for one ear|
| 8. Live female infant||39||Lamotrigine, 100 mg/day wk 0–8, 50 mg/day wk 8–39, 100 mg/day wk 39; valproate, throughout pregnancy||Multiple congenital anomalies including cleft palate, hypertelorism, broad nasal bridge, transverse palmar creases, short proximal thumbs, and supraumbilical hernia|
| 9. Induced abortion||17||Lamotrigine, 700 mg/day wk 0–17; clobazam, through first trimester||Lumbar neural tube defect|
| 10. Induced abortion||17||Lamotrigine, 100 mg/day wk 0–9; valproate, preconception through first trimester||Spina bifida (myelomeningocele)|
| 11. Live female infant||38||Lamotrigine, 250 mg/day wk 7–11, 400 mg/day wk 11–22, 600 mg/day wk 22, through delivery; oxcarbazepine, preconception and throughout pregnancy||Patent (persistent) ductus arteriosis. Atrial septal defect|
| 12. Live male infant||40||Lamotrigine, 300 mg/day wk 8–33; valproate during first trimester||Atrial septal defect|
| 13. Live male infant||38||Lamotrigine, 200 mg/day wk 0 through delivery; valproate, 1,000 mg/day preconception and throughout pregnancy||Pulmonary stenosis after delivery; surgery performed; subsequent death|
Table 2 summarizes pregnancy outcomes after first-trimester exposures to LTG.
Table 2. Pregnancy outcomes, stratified by earliest trimester of exposure of lamotrigine monotherapy: prospectively identified outcomes worldwide from September 1, 1992, through September 30, 2001
|Lamotrigine monotherapy|| || || || || |
| Infants without birth defectsb||165e||c6c||0||d5d||176|
| Infants with birth defectsb||3||0||0||0||3|
| Live birth|| || || || || |
Lamotrigine monotherapy during the first trimester
LTG monotherapy was used in 166 pregnancies resulting in 168 pregnancy outcomes (excluding spontaneous or induced abortions without BDs) exposed during the first trimester (Table 2), and three infants had major BDs. LTG doses for the three BDs ranged between 25 and 500 mg/day. The percentage with BDs in pregnancies exposed to LTG monotherapy (three of 168) was 1.8% (95% CI, 0.5–5.5%).
Specific lamotrigine polytherapy regimens during the first trimester
The overall rate of malformation for fetuses with first-trimester exposure to polytherapy with LTG was 10 (6.0%) of 166. However, the rate varied according to whether VPA was included in the regimen. The percentage with BDs after first-trimester use of combinations involving VPA and LTG, with or without other AEDs, was five (10.0%) of 50 (95% CI, 3.7–22.6%). All of the BDs cases exposed to LTG/VPA combinations were from Europe and Australia, and 82% of LTG/VPA exposures were from these regions. The frequency of major BDs in pregnancies exposed to first-trimester LTG polytherapy combinations without VPA was five (4.3%) of 116 (95% CI, 1.6–10.3%). No distinctive pattern of major malformations was identified in the infants enrolled in the Registry.
This study is one of the largest published studies of major BDs after LTG use during pregnancy. We noted no distinctive pattern of abnormalities that could have suggested a common cause such as medication exposure. Although the sample size is too small to rule out small elevations in the frequency of major BDs, the frequency of malformations after LTG monotherapy exposure observed in this study did not suggest a large elevation. In a U.K. pregnancy registry monitoring women exposed to any AED during pregnancy, the frequency of major BDs in women using LTG monotherapy was 4.0% (95% CI, 0.6–7.6%) (19).
Published studies reported rates ranging between 3.6 and 9.6% after monotherapy with other AEDs during pregnancy (1,3,12–15). Holmes et al. (5) recently reported that the birth prevalence of major BDs in offspring of women receiving common monotherapies was 4.5%, and in women with no history of seizure or AED therapy was nine (1.8%) of 508.
Because epilepsy is almost always treated during pregnancy, it is not clear whether there are independent effects of epilepsy on risk of major BDs. If the baseline risk of all major BDs in women with epilepsy is 4%, then a denominator of 200 monotherapy pregnancies is needed to be able to detect a twofold or greater excess from baseline. However, if the expected frequency independent of medications is closer to the 2.2% frequency in the general population (20), then a sample size of 402 monotherapy pregnancies would be needed to detect a twofold elevation in the rate of major BDs.
Between-study variation in frequency of major BDs for women using AED therapy to treat epilepsy depends on a number of factors, including inclusion and exclusion criteria for major BDs, the geographic regions included, how early in the pregnancy women are enrolled into the registry, source of pregnancy-outcome information, timing of follow-up, whether elective abortions are included, and the medication(s) they receive during the first trimester. Therefore one should use caution when comparing risk estimates across studies, and when possible, we limit references to large studies that initiate follow-up before prenatal testing and that obtain BD information on voluntary abortions.
The frequency of major BDs after LTG polytherapy in this registry was similar to the 6.5% reported in a retrospectively ascertained cohort of U.K. women taking LTG before LTG was approved for monotherapy (21).
The higher frequency of major BDs in the LTG/VPA combination was similar to the 9% observed in one of the largest (n = 184) published studies of VPA monotherapy (22). However, it cannot be determined from cross-study comparisons what the relative roles of the two AEDs are in contributing to the risk of major malformations in the combination. In general, the percentage with BDs has been shown to be higher in women with AED polytherapy (3). Therefore the International League Against Epilepsy (ILAE) Guidelines recommend that doses of combined AEDs be carefully reviewed during pregnancy (23).
The International Lamotrigine Pregnancy Registry is designed to facilitate data collection with minimal effort on the part of reporting physicians, who come from diverse specialties and many countries and health systems; there are a number of methodologic limitations to this approach. One of the principal limitations is lack of standardization in outcome ascertainment. As follow-up information is collected during the first months of infant life, reporting of only the most obvious defects occurs, and some anomalies likely to involve subjective judgment may be over- or underascertained.
There are several potential sources of bias in this study that could lead to an overestimate or an underestimate of the true percentage with BDs. Any tendency by individual health care providers to report patients perceived as high risk could skew the results. In addition, patients lost to follow-up might experience rates of BDs different from those monitored, and this could affect the overall estimates of percentage with BDs. Finally, prenatal tests performed before enrollment provide information that could bias a physician's decision to enroll the pregnancy. Health care providers can enhance the validity of the data by enrolling each exposed pregnancy into the registry as early as possible before any prenatal testing is performed. Unfortunately, the registry does not collect information on the dates of prenatal testing, and 36% of the registered pregnancies are registered after 20 weeks. It is more likely that some or most of these have already had an ultrasound that do not indicate the presence of a major malformation. Although false-negative ultrasounds are not uncommon (24), the inclusion of these pregnancies may result in an underestimate of the overall frequency of major malformations. There were three major BDs in 107 pregnancies enrolled before week 21. Thus using this stricter definition of prospectively reported pregnancies does not generate an alarming estimated frequency of major malformations, given what is seen in the general population and in other AED monotherapies.
Three minor anomalies were reported for the LTG/VPA combinations: (a) facial asymmetry, especially of the mandible; (b) wide-set eyes, no epicanthic folds, infraorbital creases, broad nose, large mouth, bilateral undescended testes, sinus-bridge of nose, right facial palsy, persistent patent ductus arteriosus; and (c) light facial dysmorphy: no philtrum, and syndactyly of toes two and three, bilateral. Such facial anomalies have been observed in children born of women who use VPA for the treatment of epilepsy (25,26). Although the three infants are classified as not having major defects, it is recognized that the presence of multiple craniofacial dysmorphisms can be associated with underlying developmental or neurologic deficits. Because such deficits frequently require subspecialty evaluation and may not be diagnosed until months to years after birth, monitoring their frequency and impact among children exposed prenatally to LTG is beyond the scope of the registry methods.
A number of ongoing pregnancy registries are designed to quantify pregnancy outcomes in women with epilepsy using AEDs during pregnancy (Table 3). The International Lamotrigine Pregnancy Registry collects prospective data on LTG exposures from physicians throughout the world; the U.K. AED Pregnancy Register prospectively collects pregnancy-outcome data on pregnancies exposed to any AED from U.K. neurologists; and the North American AED Pregnancy Registry collects 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 coordinates the collection of data on pregnancies exposed to AEDs throughout the world (Table 3). To facilitate comparison of results across registries, it is preferred that individual pregnancies not be intentionally enrolled into more than one registry, and we encourage all clinicians and patients to register pregnancy exposures to any AED into an appropriate exposure registry as early in the pregnancy as possible.
Table 3. Ongoing pregnancy registries prospectively quantifying frequency of major birth defects in pregnancies exposed to antiepileptic drugs
|Lamotrigine|| || |
| Lamotrigine Pregnancy Registry||Reporting health care providers can call 1-800-336-2176 in North America, or outside North American, contact the Medical Director at local GlaxoSmithKline companies||GlaxoSmithKline Worldwide Epidemiology Dept and the International Lamotrigine Pregnancy Registry Scientific Advisory Committee|
|Lamotrigine or other AED|| || |
| North American AED Pregnancy |
|Pregnant women with epilepsy in North America can call 1-888-233-2334||Dr. Lewis Holmes, Boston, MA, U.S.A.|
| United Kingdom Epilepsy and Pregnancy Register||Any physician, nurse, midwife, or pregnant woman can enroll at the address given or free phone 0800 3891248 or download registration forms from www.epilepsyandpregnancy.co.uk||Dr. James Morrow, Belfast, Northern Ireland, and the U.K. Epilepsy and Pregnancy advisory committee|
| EURAP||Physicians enroll patients through local investigators in participating countries throughout the world||Dr. Torbjörn Tomson, Stockholm, Sweden|
Although definitive conclusions are not available, we hope these data are useful for considering benefits and risks when addressing the optimal management of pregnancies in women with epilepsy. The observed percentage with all major malformations within the pregnancies exposed to LTG monotherapy during the first trimester has not raised any obvious safety concern. However, to date, the sample size is too small to make conclusions about frequency of specific BDs in pregnancies exposed to LTG monotherapy or specific polytherapy regimens. The frequency of BDs in LTG/VPA combination appears somewhat higher than that in monotherapy. However, the sample size is relatively small, and attribution of cause to one or the other AED or a specific combination effect cannot be assessed with this study design.