Pregnancy outcome in women exposed to leflunomide before or during pregnancy
Findings from animal studies have suggested that leflunomide may be a human teratogen. In the only human cohort study published to date, an increase in adverse outcomes in pregnancies after exposure to leflunomide was not detected. The aim of the present analysis was to expand on the previously published data with a description of birth outcomes among women who did not meet the previous cohort study criteria but who were exposed to leflunomide either during pregnancy or prior to conception.
Data on pregnancy exposures and outcomes were collected from 45 pregnant women who had contacted counseling services of the Organization of Teratology Information Specialists in the US or Canada between 1999 and 2009. Sixteen women were exposed to leflunomide during the first trimester of pregnancy and 29 women were exposed preconception.
All 16 of the pregnancies with leflunomide exposure during pregnancy and 27 (93%) of the pregnancies with exposure prior to conception resulted in liveborn infants. There were 2 infants with major malformations from mothers who were exposed during pregnancy, and no malformations reported in the preconception group. There was a potential known alternative etiology for at least some of the defects observed.
These data provide additional reassurance to women who inadvertently become pregnant while taking leflunomide and who undergo the washout procedure, as well as women who discontinue the medication prior to conception but have no prepregnancy documentation of drug clearance. However, until more conclusive data become available, women receiving leflunomide should be advised to use contraceptive methods and avoid pregnancy.
Leflunomide is a disease-modifying antirheumatic drug approved by the US Food and Drug Administration (FDA) for the treatment of rheumatoid arthritis (RA). Leflunomide inhibits dihydroorotate dehydrogenase (DHODH), and as a consequence, blocks the de novo pathway of pyrimidine nucleotide synthesis, which is crucial for T cell activation and proliferation (1). Mutations in the DHODH gene have been recently reported as the cause of Miller syndrome (also known as postaxial acrofacial dysostosis), a very rare autosomal-recessive genetic disorder (2). Leflunomide is also an inhibitor of protein tyrosine kinases, which play a fundamental role in the intracellular signal transduction triggered by cytokines (1).
Experimental animal studies in rats, rabbits, and mice have demonstrated that leflunomide is embryotoxic and teratogenic (1). Malformations of the head (including cranioschisis and exencephaly), rump, vertebral column, ribs, and limbs were observed in rats (1). Head and skeletal defects were noted in rabbits (1). Multiple malformations, including craniofacial defects and anomalies of the axial skeleton, heart, and great vessels, were observed in mice (3).
Based on these studies, leflunomide has been assigned a pregnancy category X in the US FDA classification system, indicating that, as demonstrated in these animal studies, fetal abnormalities and the risks involved in the use of the drug in pregnant women clearly outweigh the potential benefits. Women receiving such therapy are advised to use contraceptive methods and avoid pregnancy. Women who are planning pregnancy or who inadvertently become pregnant while they are taking leflunomide are advised to undergo a drug-elimination procedure in order to bring the plasma level of the active metabolite of leflunomide below 0.02 μg/ml, which has been regarded as safe in humans (4). Without the washout procedure, consisting of 1 or more courses of cholestyramine or, alternatively, charcoal, it may take up to 2 years to reach the above-mentioned plasma level.
Few data on the effects of leflunomide in human pregnancy are available. However, it is estimated that the rate of exposure in pregnancy is low. In a US insurance claims database study, among 34,169 women of reproductive age with a diagnosis of RA, 393 pregnancies were recorded, and of these women, 1.1% had received a prescription for leflunomide (5). Prior to 2010, 17 cases of exposure to leflunomide during gestation had been reported in the literature. Two pregnancies ended with an elective termination and 1 resulted in an infant born 9 weeks preterm with cerebral palsy and unilateral blindness. The other 14 pregnancies resulted in healthy infants (6–10).
The Organization of Teratology Information Specialists (OTIS) previously conducted a collaborative, prospective, controlled cohort study evaluating the outcome of pregnancies in women following exposure to leflunomide during the first trimester of gestation (11). After adjustment for confounders, this study did not reveal an increased risk of adverse pregnancy outcomes, including major malformations, in a group of 64 leflunomide-exposed women with RA when compared with a group of 108 women with RA who did not take leflunomide. In particular, no specific pattern of birth defects or similarities in defects compared to those noted in animal studies was detected following early pregnancy exposure to leflunomide. However, only a small subset of the sample was exposed to full doses of leflunomide beyond 3 weeks postconception; in fact, all women discontinued the medication upon recognition of pregnancy, and nearly all underwent the drug-elimination procedure.
In the present report, we describe the pregnancy outcomes in an additional 45 women prospectively enrolled in the OTIS study who had exposure to leflunomide either during pregnancy or only prior to pregnancy but who did not meet the strict criteria for inclusion in the cohort component of the previously published study (11).
PATIENTS AND METHODS
From 1999 through 2009, the OTIS Collaborative Research Group conducted a study to address the fetal safety of exposure to leflunomide during pregnancy. Participants were recruited from among pregnant women who contacted any of the OTIS counseling services throughout the US and Canada. The primary focus of the study involved a prospective, controlled cohort design; the results of the cohort study have been described in detail in a previous publication (11). However, women who did not meet the inclusion criteria for the cohort were also also recruited into the study and followed up throughout pregnancy. Specifically, these women did not meet the cohort criteria because they were enrolled after 20 completed weeks of gestation, were treated with leflunomide for an indication other than RA, or discontinued the drug within 2 years prior to conception but had no documented confirmation that the drug had been eliminated prior to pregnancy.
Participants selected from the overall study for the current analysis comprised 2 groups: 1) pregnant women who did not meet the cohort criteria but who took at least one dose of leflunomide on or after the estimated date of conception; and 2) pregnant women whose last dose of leflunomide was within 2 years prior to the estimated date of conception but who had no record of drug plasma levels prior to conception that could document clearance of the drug, thus implying that these patients could still have been exposed to some level of leflunomide during pregnancy. These women were followed up from a central study coordinating center located at the University of California, San Diego, using the same methods as were employed with the participants in the cohort component of the study (11). Retrospectively ascertained cases (enrolled after the outcome of pregnancy was known) were not included in this analysis.
All women in the study initially provided oral consent for participation, and subsequently provided written consent. The study was approved by the University of California, San Diego Institutional Review Board.
Each woman enrolled in the study completed between 1 and 3 structured telephone interviews during pregnancy, depending on the gestational age at enrollment, that addressed history of previous pregnancies, family medical history, prepregnancy body mass index, socioeconomic and demographic information on the woman and her partner, and exposures during the current pregnancy. Socioeconomic status was classified using the Hollingshead categories based on maternal and paternal occupation and education (12). The exposure history reported by each subject included the dosages, dates, and indications for all medications, use of caffeine, use of supplemental vitamins, occupational exposures, a diagnosis of infectious or chronic disease, prenatal testing or other medical procedures, and use of recreational drugs, tobacco, and alcohol.
In addition, at the time of enrollment, women were asked to respond to questions about current disease activity or symptoms. Functional status was determined using the Health Assessment Questionnaire, with possible values from 0 to 60 (13). The severity of pain and patient's perception of global impact of the disease (over the last week) were evaluated with the use of 0–100-mm visual analog scales, modified for telephone administration.
Birth outcome was recorded on a standard form that was completed, via telephone interview of the mother, shortly after delivery or at the end of pregnancy. Measures included the outcome of pregnancy (livebirth, stillbirth, elective termination, or spontaneous abortion), the reported presence or absence of major structural defects in the newborn, gestational age at delivery, mode of delivery, length and type of hospital stay, maternal or newborn complications, maternal weight gain, infant Apgar scores, and infant birth weight, length, and head circumference. Medical records from the prenatal care provider, the hospital of delivery, the rheumatologist, and the pediatrician were examined for additional exposure and outcome data. In addition, the infant's physician was asked to return a form reporting postnatal growth measures and the presence or absence of any major structural defect noted up to that point.
Liveborn infants were also examined by a member of a team of 3 pediatric dysmorphologists (LKR, SRB, and KLJ), who traveled to see these infants, typically in their homes. As part of the overall study protocol, which included examination of unexposed comparison infants for the cohort component of the study, the examiner in each case was blinded with regard to the exposure and disease status of the mother–infant pair. The dysmorphology evaluations were completed for both major and minor structural anomalies; the latter were defined as structural defects that have no cosmetic or functional importance and that are known to occur in <4% of the general population (14). Infants who underwent this dysmorphology examination were evaluated using a standard checklist itemizing 132 such anomalies (15). Photographs were taken of each infant to aid in addressing possible issues of interrater reliability among multiple examiners.
Evaluation of outcomes.
Major structural defects were defined in general according to the Metropolitan Atlanta Congenital Defects Program classification system (16). In addition to those defects noted in the physical examination or medical record or reported by the mother, information on functional abnormalities was collected. These functional abnormalities were defined as developmental abnormalities that did not represent clearly defined congenital defects in structure.
Based on the suspected predictive value of having ≥3 minor structural defects to indicate the presence of a major abnormality (17), the number of minor anomalies was tabulated for each child who underwent the dysmorphology examination. In addition, because of the known association of a pattern of specific minor anomalies with prenatal exposure to many known human teratogens, clustering of specific minor structural defects in leflunomide-exposed infants was evaluated for evidence of a pattern, defined as at least 3 specific minor anomalies occurring in at least 2 children in the exposed group.
Prematurity was defined as spontaneous delivery at <37 completed weeks of gestation. Small for gestational age and prematurity-adjusted chronologic age were defined as less than or equal to the tenth percentile for sex and age, determined using standard National Center for Health Statistics 2000 growth curves (18) for full-term infants and Lubchenko curves (19) for preterm infants.
As these data represent an exposure series and there is no appropriate comparison group, descriptive statistics were used to report means, standard deviations, and frequencies of the maternal characteristics and outcomes using standard measures. SAS version 9.1 software was used to calculate these values.
A total of 45 women exposed to leflunomide were prospectively followed up. Sixteen of the women were exposed to the drug during pregnancy (identified as group 1, the during-pregnancy exposure group) and 29 were exposed to the drug only prior to conception (group 2, the preconception exposure group). There were no pregnancies lost to followup.
The characteristics of the mothers are shown in Table 1. The majority of the women in both groups had a diagnosis of RA or juvenile RA. Women whose last dose of leflunomide was in the preconception period tended to have a higher socioeconomic status, to have started multivitamins or folate supplements before conception, to be smokers and alcohol drinkers, and to have a higher global impact score on the measure of RA disease severity/symptoms in the most recent week prior to enrollment.
Table 1. Maternal characteristics of the women treated with leflunomide*
|Maternal age, mean ± SD years||29.9 ± 7.9||29.9 ± 4.6|
|Ethnicity|| || |
| White/non-Hispanic||6 (37.5)||22 (75.9)|
| Hispanic||4 (25.0)||5 (17.2)|
| Black||5 (31.3)||2 (6.9)|
| Asian||1 (6.3)||0|
|Low socioeconomic status||7 (46.7)||4 (13.8)|
|Primigravid||5 (31.3)||10 (34.5)|
|Primiparous||4 (25.0)||13 (44.8)|
|Any previous spontaneous abortions||5 (31.3)||6 (20.7)|
|Any previous elective terminations||1 (6.3)||5 (17.2)|
|Prepregnancy BMI, mean ± SD kg/m2||27.8 ± 10.5||26.7 ± 6.9|
|Multivitamins or folate at conception||4 (25.0)||10 (34.5)|
|Any alcohol during pregnancy||5 (31.3)||19 (65.5)|
|Any tobacco use during pregnancy||2 (12.5)||7 (24.1)|
|Gestational age at time of enrollment, weeks|| || |
| Mean ± SD||15.8 ± 10.0||10.6 ± 4.8|
|Diagnosis|| || |
| Rheumatoid arthritis||6 (37.5)||21 (72.4)|
| Juvenile rheumatoid arthritis||2 (12.5)||7 (24.1)|
| Psoriatic arthritis||1 (6.3)||0|
| Systemic lupus erythematosus||4 (25.0)||1 (3.4)|
| Scleroderma||1 (6.3)||0|
| Rheumatoid arthritis + systemic lupus erythematosus||2 (12.5)||0|
|Symptom/activity score, mean ± SD|| || |
| Activity, 0–60 points||4.4 ± 4.8||11.0 ± 10.3|
| Pain, 0–100 points||28.8 ± 31.6||43.2 ± 28.4|
| Global impact, 0–100 points||23.4 ± 25.9||40.9 ± 30.1|
|Dosage of leflunomide, mg/day|| || |
| Mean ± SD||17.5 ± 4.5||18.3 ± 5.1|
|Last dose of leflunomide received preconception|| || |
| 1–15 weeks preconception||–||19 (65.5)|
| 15–30 weeks preconception||–||3 (10.3)|
| >30 weeks preconception||–||7 (24.1)|
|Last dose of leflunomide received postconception, weeks|| || |
| Mean ± SD||7.1 ± 7.9||–|
|Cholestyramine washout procedure||13 (81.3)||21 (72.4)|
|Methotrexate use||3 (18.8)||0|
| First-trimester exposure||3 (18.8)||0|
|Systemic steroid use||11 (68.8)||19 (65.5)|
| First-trimester exposure||6 (37.5)||15 (51.7)|
|Nonsteroidal antiinflammatory drug use||7 (43.8)||12 (41.4)|
| First-trimester exposure||6 (37.5)||10 (34.5)|
|Weight gain in pregnancy for live births, mean ± SD kg||11.7 ± 8.4||14.2 ± 11.1|
In the during-pregnancy exposure group, the mean ± SD dosage of leflunomide was 17.5 ± 4.5 mg/day, and the mean ± SD gestational timing of the last dose of leflunomide was 7.1 ± 7.9 weeks postconception. All women in the during-pregnancy exposure group discontinued leflunomide upon recognition of pregnancy. Most of the 16 women in the during-pregnancy exposure group (81.3%) underwent at least one washout procedure with cholestyramine after leflunomide therapy was discontinued (Table 1).
In the preconception exposure group, the majority of women (65.5%) discontinued leflunomide therapy during the last 15 weeks before conception. Most of the women in the preconception exposure group also underwent the cholestyramine washout procedure (72.4%), which was performed after conception in all but one case (Table 1).
The overall percentage of women exposed to systemic steroids and nonsteroidal antiinflammatory drugs was comparable in the 2 groups, but women who discontinued leflunomide therapy before conception more frequently reported taking systemic steroids during the first trimester of pregnancy (51.7%). The therapeutic regimen also included methotrexate in 3 women exposed to leflunomide during pregnancy (18.8%).
Birth outcomes in both groups are shown in Tables 2, 3, and 4, and a detailed listing of each of the outcomes among the women exposed to leflunomide during pregnancy is shown in Table 5. All 16 pregnancies in group 1 and all but 2 of the 29 pregnancies in group 2 (93.1%) resulted in liveborn infants. The absence of spontaneous abortions in group 1 may be a reflection of the relatively late gestational age (mean 15.8 weeks' gestation) in these women at the time of study enrollment.
Table 2. Birth outcomes of the pregnant women treated with leflunomide*
|Liveborn infant||16 (100)||27 (93.1)|
|Spontaneous abortion||0||2 (6.9)|
|Characteristics of live births|| || |
| Male sex||5 (31.3)||18 (66.7)|
| Twin gestation†||1 (6.3)‡||3 (11.1)|
| Delivery by cesarean section||6 (37.5)||10 (37.0)|
| Preeclampsia||0||2 (7.4)|
| Diabetes any||2 (12.5)||1 (3.7)|
Table 3. Gestational age, birth size, and postnatal growth in liveborn infants of women treated with leflunomide*
|Gestational age at delivery, weeks|| || |
| Mean ± SD||37.8 ± 2.3||37.7 ± 2.6|
| Range||35.3 – 41.7||31 – 41|
|Preterm delivery (<37 weeks)||8 (50.0)||7 (25.9)|
|Full-term infants|| || |
| Weight, mean ± SD gm||3,172.3 ± 302.7||3,311.7 ± 617.4|
| Length, mean ± SD cm||49.6 ± 2.0||49.7 ± 3.1|
| Head circumference, mean ± SD cm||34.4 ± 1.2||34.1 ± 2.1|
|Liveborn infants in ≤10th percentile at birth|| || |
| By weight||1 (6.3)||3 (11.1)|
| By length||3 (18.8)||4 (15.4)|
| By head circumference||1 (9.1)||4 (21.1)|
|Liveborn infants in ≤10th percentile postnatally|| || |
| By weight||2 (16.7)||5 (22.7)|
| By length||2 (16.7)||2 (9.1)|
| By head circumference||2 (16.7)||5 (25.0)|
Table 4. Major and minor anomalies in infants born to women treated with leflunomide*
|Major structural defects in live births||2 (12.5)||0|
|Major structural defects in pregnancy losses||–||0|
|Functional problems in live births||1 (6.3)||1 (3.7)|
|Dysmorphology examination completed||14 (87.5)||21 (77.8)|
|Minor structural anomalies|| || |
| 0 – 1||3 (21.4)||6 (28.6)|
| 2||4 (28.6)||6 (28.6)|
| ≥3||7 (50.0)||9 (42.9)|
|Similar pattern of minor anomalies||2 (14.3)||0|
| Type of anomaly||Short nose, flat nasal bridge, long philtrum|| |
Table 5. Maternal characteristics and pregnancy outcome in women exposed to leflunomide during pregnancy*
|1/24||NA||PsA||15||35||30||Dosage unknown, from −2 to 2.9 weeks||Pred. (second, third); NSAID (first, second, third)||No||Live birth; female; 41.7 weeks; severe sensorineural hearing loss|
|2/32||High||RA, depression||10||85||70||Dosage unknown, from −2 to 4.1 weeks||MTX (first); Pred. (second, third)||Yes||Live birth; male; 36.9 weeks|
|3/37||High||RA, asthma, depression, hypothyroidism, type 1 diabetes||0||10||5||20 mg/day, from −2 to 2.3 weeks||MTX (first); NSAID (first)||Yes||Live birth; male; 35.7 weeks (twin pregnancy)|
|4/32||High||SLE, Sjögren's syndrome, vasculitis||12||100||50||20 mg/day, from −2 to 1.7 weeks; no washout||Pred. (first, second, third)||No||Live birth; female; 40.3 weeks|
|5/24||High||RA||0||0||0||20 mg/day, from −0 to 0.7 weeks; no washout||–||No||Live birth; female; 40.9 weeks (twin pregnancy)|
|6/23||High||Juvenile RA||4||25||15||20 mg/day, from −2 to 6.9 weeks||NSAID (first)||Yes||Live birth; female; 39.9 weeks|
|7/36||High||SLE, gestational diabetes||3||40||25||20 mg/day, from −2 to 3.3 weeks||Pred. (first, second, third); NSAID (first)||Yes||Live birth; female; 36.3 weeks|
|8/23||Low||Juvenile RA||3||50||75||20 mg/day, from −2 to 6.9 weeks; no washout||MTX (first); Pred. (first, second, third)||No||Live birth; female; 35.9 weeks|
|9/43||Low||Scleroderma||0||0||0||20 mg/day, from −2 to 26.4 weeks||Pred. (first, second, third)||Yes||Live birth; female; 35.3 weeks|
|10/30||Low||RA, SLE||0||0||0||20 mg/day, from −0.9 to 7.9 weeks||Pred. (first)||No||Live birth; female; 40 weeks|
|11/28||High||SLE||7||45||50||10 mg/day, from −2 to 3.1 weeks||Pred. (first, second, third)||No||Live birth; female; 35.4 weeks; pattern of 3 minor malformations (short nose, flat nasal bridge, long philtrum)|
|12/45||Low||RA||9||0||0||10 mg/day, from −2 to 24.6 weeks||NSAID (first,second)||No||Live birth; female; 39.4 weeks; pattern of 3 minor malformations (short nose, flat nasal bridge, long philtrum)|
|13/35||High||SLE||0||0||0||10 mg/day, from −2 to 2.3 weeks||–||No||Live birth; male; 37 weeks|
|14/27||Low||RA||0||0||0||Dosage and weeks of exposure unknown||Pred. (third); NSAID (first)||No||Live birth; female; 35.4 weeks; aplasia cutis congenita (surviving member of a twin pregnancy; the other twin was spontaneously aborted)|
|15/23.8||Low||RA, SLE||4||50||25||20 mg/day, from −0.4 to 3.7 weeks||Pred. (second, third)||No||Live birth; male; 35.6 weeks; Pierre-Robin sequence, spina bifida occulta, patent ductus arteriosus, chondrodysplasia punctata, congenital heart block|
|16/15.5||Low||RA||4||20||30||Dosage unknown, from −2 to 9.6 weeks||NSAID (second)||No||Live birth; male; 39.7 weeks|
Women exposed to leflunomide during pregnancy had a higher rate of preterm delivery (50%) than those whose last dose was prior to conception (25.9%) (Table 3), but no deliveries took place before 35.3 weeks' gestation. In group 2, there were 5 deliveries before 35 weeks of pregnancy (18.5%), one of these at 31 weeks. The rate of preterm birth was higher when leflunomide therapy was stopped within the last 15 weeks before conception (33.3%) than when it was stopped earlier (11.1%). However, women who discontinued leflunomide within the last 15 weeks also had a higher mean global impact score on the measure of RA disease severity/symptoms obtained from the mother at the time of enrollment (data not shown).
There were 2 newborns with major malformations, both born to women in the during-pregnancy exposure group. One of the infants was a liveborn female twin with aplasia cutis congenita involving both thighs. The other twin from this pregnancy was spontaneously aborted and not examined. Aplasia cutis is a congenital defect involving absence of a portion of the skin. The other infant with a major malformation in the during-pregnancy exposure group was a liveborn male singleton with multiple anomalies (Pierre-Robin sequence, spina bifida occulta, patent ductus arteriosus, chondrodysplasia punctata, and congenital heart block). This infant was born to a mother with a diagnosis of both RA and systemic lupus erythematosus (SLE). There were no major congenital malformations reported in the preconception exposure group (Table 4).
In each of the 2 groups, we observed 1 liveborn infant with functional anomalies, one involving a female infant with severe sensorineural hearing loss (group 1), and the other being a female infant with intrauterine growth restriction and cerebral palsy (group 2). The latter infant was delivered by cesarean section at the 31st week of gestation, to a mother with juvenile RA, diabetes, hypertension, and asthma.
Thirty-five infants underwent the dysmorphology examination, 14 in group 1 and 21 in group 2. The frequency of infants with at least 3 minor anomalies was similar in the 2 groups, involving 50% of infants in the during-pregnancy exposure group and 42.9% of infants in the preconception exposure group. Two children in group 1 had the same 3 minor anomalies (flat nasal bridge, short nose, and long philtrum). This same pattern was not observed in any of the infants in group 2 (Table 4), nor was this pattern observed in the previously published cohort study among any of the 51 infants with prenatal exposure to leflunomide who underwent the dysmorphology examination (11).
Preclinical animal studies in rats, rabbits, and mice demonstrated leflunomide embryotoxicity and teratogenicity. Fukushima at al (20) showed that supplementation with exogenous uridine in pregnant mice reduced the rate of most of the congenital malformations caused by leflunomide and suggested that the teratogenicity of leflunomide was related to the inhibitory effect of DHODH activity. Despite these results, the only prospective study in humans published to date showed no differences, after adjustment for confounding, in the frequencies of adverse outcomes in pregnancies exposed to leflunomide during the first few weeks of gestation, relative to a disease-matched comparison group (11).
Mutations in the DHODH gene have been reported in 6 kindred infants with Miller syndrome, an autosomal-recessive disease whose characteristic features are craniofacial anomalies (malar hypoplasia, coloboma of eyelids, micrognathia, cleft lip, cleft palate, hypoplastic cup-shaped ears), limb abnormalities (absence of fifth finger, syndactyly, shortening and in-curving of forearms), accessory nipples, and, occasionally, other skeletal or visceral malformations; cognitive development is usually normal (2, 21). These malformations are similar to those observed in experimental animal studies and suggest that DHODH activity is important in normal human embryo-fetal development. All patients with Miller syndrome analyzed for DHODH genotype have been reported to be compound heterozygous for mutations, predicted to be deleterious, in the DHODH gene (2).
In the present study, we observed 2 infants with major birth defects who were born to women exposed to leflunomide during pregnancy, and no malformed infants in those exposed only prior to pregnancy. None of these specific defects were reported in the previously published cohort study, nor are they similar to those reported in the animal studies (11). In addition, there is a potential known alternative etiology, unrelated to leflunomide exposure, for at least some of the defects observed.
In the first case, a female infant with aplasia cutis congenita of the thighs was the surviving member of a twin pregnancy in which the cotwin was spontaneously aborted. There are several case reports in the literature of this skin defect found in body areas other than the scalp in the surviving member of a monozygotic twin pair. This suggests that the defect may have been the result of the twinning/cotwin demise as opposed to exposure to leflunomide (22–29).
In the second case, a male infant was diagnosed as having Pierre-Robin sequence, spina bifida occulta, patent ductus arteriosus, chondrodysplasia punctata, and congenital heart block. This child was born to a mother who had a diagnosis of RA and SLE. Chondrodysplasia punctata has been reported to occur with increased frequency in infants born to mothers with SLE and other systemic autoimmune diseases (30–36). Moreover, the placental transport of maternal anti-SSA/Ro or anti-SSB/La antibodies in women with autoimmune diseases is a well-known cause of congenital heart block (37, 38). While the chondrodysplasia punctata and the congenital heart block might be explained by the underlying maternal disease, the Pierre-Robin sequence, which is characterized by micrognathia, glossoptosis, and breathing problems (with or without cleft palate), is a feature of Miller syndrome. It is possible that leflunomide could have played a causative role in the pathogenesis of this subset of malformations in a single infant.
One-half of the children who were born to women exposed to leflunomide during pregnancy and who underwent the dysmorphology examination in this series had ≥3 minor anomalies, similar to the prevalence of ≥3 minor anomalies in both the exposed and comparison groups in the previously published cohort study. Two children in the during-pregnancy exposure group in the current analysis (14.3% of the 14 who were examined) had the same 3 defects: flat nasal bridge, short nose, and long philtrum. The commonality of this specific pattern may be due to chance, particularly since this same pattern was not seen in any of the exposed infants in the cohort study of RA patients, and moreover, was not associated with additional major malformations in this exposure case series.
Birth outcomes in the group of women whose last dose of leflunomide was received prior to conception are also reassuring. There were no major malformations, there was no pattern of minor anomalies, and the spontaneous abortion rate was within the normal range seen in the general population.
The rates of preterm delivery in both groups were higher compared to that in the general population. However, as suggested by multivariate analysis in the previous controlled cohort study, the higher rate of preterm delivery is likely to be a consequence of the underlying maternal diseases, including RA and SLE, and/or the concomitant use of oral steroids (11, 39). In the second group, women who discontinued leflunomide within the last 15 weeks before conception were more likely to have a preterm delivery than women who stopped therapy earlier, but they also had a higher mean global impact score, reflecting greater disease severity/symptoms, at the time of enrollment, suggesting once again a possible role of the underlying maternal disease.
Limitations of our study include the relatively small number of pregnancies analyzed and the lack of an unexposed comparison group. The use of a volunteer sample may also limit generalizability. However, as an expansion of the previously published cohort study, the results from this series help to fill the gap in information about pregnancy outcomes when exposure to leflunomide continues beyond the first 3–4 weeks after conception. In this series, 40% of the women continued the therapy until or after 6.9 weeks postconception. Moreover, this is the first study to analyze prospectively collected data on the outcome of pregnancies when the mother has been exposed to full doses of leflunomide but has discontinued its use prior to conception without documentation of blood levels of the drug that would confirm washout prior to conception.
In conclusion, even though preclinical animal studies and the genetic defect underlying Miller syndrome, by analogy, suggest that leflunomide has potential as a human teratogen, the results of this analysis and those in the previously published cohort study have not demonstrated an increase in the rate of major malformations or a specific pattern of major malformations in children of women exposed to leflunomide prior to or after conception, at the doses currently used, or among women who had undergone the cholestyramine washout procedure after discontinuation of therapy. The number of women exposed to full doses of leflunomide during the entire first trimester of pregnancy is, however, too small to draw firm conclusions.
These data provide additional reassurance that leflunomide is not a major human teratogen in women who inadvertently become pregnant while taking leflunomide and who undergo the washout procedure. However, until more conclusive data become available, women receiving such therapy should still be advised to use contraceptive methods and avoid pregnancy.
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be published. Dr. Chambers had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Cassina, Johnson, Robinson, Braddock, Jones, Chambers.
Acquisition of data. Johnson, Robinson, Braddock, Jimenez, Mirrasoul, Salas, Jones, Chambers.
Analysis and interpretation of data. Cassina, Johnson, Robinson, Braddock, Xu, Jimenez, Mirrasoul, Salas, Luo, Jones, Chambers.
ROLE OF THE STUDY SPONSOR
Sanofi-Aventis had no role in the study design or in the collection, analysis, or interpretation of the data, the writing of the manuscript, or the decision to submit the manuscript for publication. Publication of this article was not contingent upon approval by Sanofi-Aventis.