Abstract
Multiple sclerosis (MS) is the most commonly acquired neurological disorder affecting young adults of reproductive age with approximately a 3:1 female-to-male ratio. Pregnancy is not contraindicated in MS but remains to be an issue that raises many questions. Although relapse rates tend to increase in the first 3 months postpartum, pregnancy does not seem to be a detriment to the long-term progression of MS and has a protective effect on reducing relapses, especially during the third trimester. MS does not appear to affect fertility or increase the risk of congenital anomalies or pregnancy complications. There has been some evidence that maternal treatment with β interferons, the most commonly used disease-modifying therapies in MS, may cause adverse reproductive outcomes, prompting the US FDA to issue warnings about their use at conception and during pregnancy.
Keywords
Multiple sclerosis (MS) is the most common acquired neurological disorder, other than trauma, affecting young adults of Northern European descent [1]. MS is characterized by demyelination of white matter in the CNS and axonal degeneration. Clinical signs and symptoms may include visual defects (e.g., diplopia or optic neuritis), motor and/or sensory symptoms (e.g., numbness and tingling in any or all four extremities), lack of coordination, tremor, incontinence and cognitive impairment [1,2]. There are four recognized disease patterns in MS:
Relapsing-remitting MS (RRMS): approximately 85% of MS patients initially follow a RRMS course in which a series of relapses and remissions (complete or partial) occur at intervals over a period of many years, even decades;
Secondary progressive (SPMS): over time (often decades), the majority of RRMS cases evolve into SPMS when relapses become less distinct and remissions no longer clearly occur;
Primary progressive: primary progressive MS is characterized by a progressive course from onset (i.e., there is no RRMS phase to the disease);
Progressive-relapsing MS (PRMS): PRMS is characterized with a progressive course from onset, but there are also episodes of clear relapses with or without full recovery [3].
The pathogenesis of MS remains unclear but genes, environment and interactions thereof, as well as epigenetic factors, are clearly involved [4–6]. Females are affected approximately three-times as often as males [7] and most patients are affected within their reproductive years [5]. While this article focuses on women with MS, one must remember that reproductive concerns also occur for males with an MS diagnosis.
Historically it was common to have a long lag time (i.e., time from the onset of clinical signs and symptoms to clinical diagnosis) for MS [3]. As a result, many pregnancies occurred after the onset of MS but before an actual diagnosis was made. Thus, the need for M S-related childbearing considerations was largely unrecognized. Current advances in the diagnostic procedure and revisions of diagnostic criteria to include MRI findings, have reduced this lag time to as short as a few weeks or months or even to an immediate diagnosis [8]. Considering also the increase in the use of disease-modifying therapies (DMTs), which are initiated early in the disease course and in fact are not approved for progressive disease, MS patients are now receiving a diagnosis and therapy at a younger age and thus are younger, less disabled and more likely to contemplate a pregnancy compared with their older counterparts [9].
In the past, women were usually discouraged to consider pregnancy following a diagnosis with MS, even to the extent that termination or tubal ligation were often recommended as appropriate, as it was presumed that pregnancy may worsen the course of the disease [10]. However, we now recognize that pregnancy is characterized by a downregulation of cellular immune responses, most likely to prevent fetal rejection as well as an increased production of estrogen and progesterone, which have been shown to have some neuroprotective properties. It would therefore seem reasonable that the combined anti-inflammatory and neuroprotective aspects of pregnancy would work together to prevent inflammatory attacks on the CNS in MS [11]. Nevertheless, it was not until 1998, when the PRIMS study launched the first large prospective study investigating the relationship between MS and pregnancy that led to changes in counseling MS women with regard to pregnancy [12].
Several studies have been reported on pregnancy and MS, but the data remain limited in some aspects, such as the effect of MS therapy on birth outcome. Finkelsztejn et al. published the first systematic review and meta-analysis of the literature on the subject that provided a summary of evidence-based data when pregnancy is considered in a family [13]. Here, we review the current literature on pregnancy-related issues in MS and discuss maternal considerations with regard to present knowledge and the need for more information.
The effect of pregnancy on MS
Relapses
In patients with established MS, early clinical observations indicated a period of low disease activity in terms of relapses throughout pregnancy and an increased risk of clinical relapses after delivery [14–17]. Supportive evidence came from a small study of two women with MS enrolled in a MRI study who underwent serial brain MRI scanning during pregnancy and postpartum. Their results showed a reduction in new or enhanced disease activity in the second half of pregnancy with a return in MRI lesion activity to prepregnancy levels after delivery [18].
In 1998, the PRIMS trial was the first large-scale study of 254 European women (269 pregnancies) who were followed for up to 1 year after delivery with the aim of assessing the effects of pregnancy and delivery on the clinical course of MS [12]. The results showed that overall, the frequency of relapses were slightly lower during the first and second trimester, but decreased significantly by 70% during the third trimester. Postpartum, the relapse rate during the first 3 months was substantially higher than that prior to pregnancy for these women but then returned to their prepregnancy rates. Several studies subsequent to the PRIMS trial have also had similar conclusions [19–23].
Physiologically, the increase in relapses postpartum is not unexpected given that there is a sudden decrease in estrogen levels immediately after delivery and the loss of the immunosuppressive state of pregnancy [24,25]. In a 2-year follow-up of the PRIMS cohort [26], the following predictors were identified for the likelihood of postpartum relapses in at least some women: high relapse rate in the prepregnancy year, increased relapse rate during pregnancy and a higher MS disability score, as measured by the disability status scale, prior to pregnancy [27]. Of note, neither epidural analgesia nor breastfeeding were significant predictors. A recent prospective longitudinal study also showed that women with a relapse during pregnancy were more likely to have a postpartum relapse, but these findings did not reach statistical significance, most probably due to the small sample size [22]. An Argentinean study confirmed that relapses during the year before pregnancy predicted relapses during the postpartum period [20]. Exclusive breastfeeding for at least 2 months postpartum was significantly associated with a lower risk of postpartum relapse, but again this work had small numbers (n = 32) and disease course can be a confounding factor in the decision to breastfeed [28].
Recent literature is controversial with regard to the effect of vitamin D on postpartum relapses. Several studies have shown an inverse relationship of serum vitamin D levels and the risk of relapses in nonpregnant patients with MS [29,30], and that nursing mothers are at a higher risk of vitamin D deficiency and should therefore be supplemented [31]. However, in a small prospective cohort study of 28 pregnant MS women who were followed-up for exposures and symptoms through the postpartum year, the authors found that the lower vitamin D levels during pregnancy and lactation in their patients were not an important risk factor for postpartum MS relapses [32]. Other biomarkers in which their pregnancy-induced fluctuation levels are suggested to mediate the increase in postpartum relapses include leptin [33] and IL-8 [34].
Long-term MS progression & disability
Most studies have not found any adverse effect of pregnancy on long-term disease progression and disability, with some evidence even showing favorable effects for having had a pregnancy.
A Canadian retrospective population-based linkage survey of 185 women with MS found no association between disability and total number of term pregnancies, time of pregnancy relative to onset of MS or either onset or worsening of MS in relation to pregnancy [35] comparable to results from PRIMS [26]. By contrast, a Belgium study of 200 female MS patients, using time from onset to wheelchair dependence as an index of progression, found that women who had at least one pregnancy after onset were wheelchair dependent at an average of 18.6 years, significantly later than nulliparous women (12.5 years), even after correction for MS age of onset [36]. However, the latter [36] was published in 1994 and the natural history of MS, including time from onset to wheelchair use, has changed to a certain extent with earlier diagnosis and regular use of DMTs [37]. A recent Dutch study of a hospital-based cohort of 277 women with MS also found no evidence that parity affected the risk of secondary progression, defined as continuous worsening of neurological symptoms unrelated to relapse for at least 1 year [38].
A limitation of these studies, however, is the possibility of selection bias and/or conceptive behavior bias. For example, women with more severe disease may opt not to have children and some studies only focused on pregnancies resulting in live births.
A Swedish study looked at the effect of pregnancy on long-term prognosis through comparing the risk of transition from relapsing-remitting to a secondary progressive course or risk of reaching level 6 of disability status scale in women who became pregnant after MS onset and women who were not pregnant after MS onset [39]. In this study, the authors matched the two groups for neurological deficit, disease duration and age so as to eliminate the possibility for a selection bias to explain the findings; their findings were similar to previous studies with a significantly decreased risk for progressive course in the group that became pregnant during the study. A more recent Belgium study also corroborated the association of pregnancies with reduced progression of disability (measured by Expanded Disability Status Score [40]) in relapsing-onset MS [41].
MS onset after pregnancy
Inconclusive data are available on MS onset after conception or during gestation with some indication of pregnancy having a protective effect. A Swedish study using a near-complete MS incidence cohort with a 25-year follow-up showed a lower risk of MS disease onset among parous women compared with nulliparous women and that association of MS with nulliparity tended to increase with age [39]. Another study using data from the Oxford–FPA study, a prospective study of 17,032 white married women recruited between 1968 and 1974 at 17 clinics in Britain, also indicated a possible beneficial effect of pregnancy by reporting that nulliparous women or those who had two or fewer children developed MS twice as often as women with three or more children, but the difference was not statistically significant [42]. Another study, however, did not support a protective effect of pregnancy or oral contraceptive (OCP) use on the risk of MS [43]. Findings from a more recent study were also consistent with a cumulative beneficial effect of pregnancy [44]; the authors reported that a higher parity was significantly associated with a reduced risk of first clinical demyelination among women.
The effect of MS management on pregnancy outcome
MS itself does not appear to increase adverse pregnancy outcomes (e.g., spontaneous abortions, stillbirth, cesarean delivery, premature birth or birth defects) compared with women who do not have MS [16,17,45–52]. A large linkage study of the Norwegian MS registry and the Norwegian National Medical Birth Registry compared birth outcomes for women prior to the onset of their MS (‘pre-MS’), during the lagtime from MS onset and diagnosis (‘early MS’) and after the diagnosis of MS (‘manifest MS’) [53]. A significantly lower mean birth weight at term was found to be associated with ‘manifest MS’ compared with the ‘pre–MS’ group [53]. Similar findings were shown in other studies where the newborns of mothers with MS were premature [54] or smaller for gestational age compared with age-matched healthy controls [23,54]. It is important to note, however, that the effect observed in these studies could have been attributed to MS treatment. Only in a German study did the authors observe an association of MS with low birth weight and birth length even in the absence of treatment [23].
DMTs
DMTs, also referred to as ‘immunomodulating therapy’ is effective in the treatment of RRMS by reducing the frequency and severity of relapses as well as the clinical and MRI findings [55]. There are currently seven approved DMTs available for the treatment of RRMS of which three belong to a class of medications called IFNβ: intramuscular IFNβ-1a (Avonex®, Biogen Idec, Inc., MA, USA), subcutaneous IFNβ–1a (Rebif®, Merck Serono, Geneva, Switzerland) and IFNβ-1b (Betaseron®, Bayer Healthcare, Leverkusen, Germany). Other approved DMTs are glatiramer acetate (GA; Copaxone®, Teva Pharmaceutical Industries, Petah Tikva, Israel), natalizumab (Tysabri®, Biogen Idec, Inc. and Elan Pharmaceuticals, Dublin, Ireland) and fingolimod (Gilenya®, Novartis, Basel, Switzerland).
IFNβ
IFNβ is currently most commonly used as a line of therapy for MS patients and works primarily through an anti-inflammatory mechanism of action in treating RRMS [56,57]. When taken during pregnancy, IFNβ could possibly have direct or indirect toxicity through its immunomodulating properties on pregnancy outcome [58,59,201]. It is largely because of these recommendations that there are only very few studies, mostly of small sample size, where there was exposure to IFNβ therapy at conception and/or during gestation, which displayed contradicting data regarding the effect of IFNβ treatment during pregnancy on pregnancy outcome (
Epidemiological studies of disease-modifying therapy use in pregnancy and risk to the fetus†.
Only studies discussed in this review are included.
AOR: Adjusted odds ratio; DMT: Disease-modifying therapy; GA: Glatiramer acetate; MS: Multiple sclerosis; OR: Odds ratio; TIS: Teratogen information services.
To date, all analytical studies conducted to assess the risks of maternal exposure to DMTs at conception or during gestation on the fetus have adopted a ‘cohort’ design. In prospective cohort studies, exposures are recorded during pregnancy without knowledge of the outcomes, which are obtained for either an entire population or for a defined study sample in a comprehensive and ongoing fashion. The main advantage of such studies is that the number of affected births accumulates over time, allowing for increased power to test the associations of rare outcomes without the need for the very high costs entailed by project-specific data collection. Cohort studies can also be retrospective in nature, where exposure (and in some cases even the outcome) may have already occurred in the past (e.g., studies based on linked administrative records). These studies usually have the benefit of being cheaper and less time-consuming. In contrast to cohort studies, ‘case–control’ studies measure the frequency of exposure in the pregnancies of mothers resulting in babies with (as controls) or without the specific outcome of interest. Exposure information in case–control studies is usually collected retrospectively through structured interviews or standardized, validated questionnaires given to mothers. These investigations are very useful for studying rare outcomes, such as specific birth defects, but may be limited by recall bias. They are also relatively expensive to perform because of the need to collect large amounts of data specifically for the study. As exposure to DMTs in pregnancy is not frequently encountered, it is not unexpected that there are no retrospective case–control studies in the literature assessing the risk of exposure to DMTs at conception or during gestation. The main studies, to date, on the risks of maternal exposure to DMTs are summarized in
Sandberg-Wollheim et al. published the first systematic analysis of pregnancy outcome from randomized clinical trials of subcutaneous and intramuscular IFNβ-1a [60]. The study showed no difference in pregnancy outcome between the IFNβ-1a in utero exposure group and the placebo group. Of the 63 pregnancies with in utero or less than 2 weeks prior to conception exposure (‘exposed group‘), 53 had the potential to go to term and 44 of those 53 women (83%) ended in live births and no dose effect was observed. Each group had a pregnancy that resulted in one child with a malformation. Although not significant, it is of interest that the rate of spontaneous abortions (26%; eight out of 31) and fetal deaths (3.2%; one out of 31) fell on the upper limit of what is expected for normal subjects [61,202]. The study is limited by the small sample size and the fact that the women studied were from clinical trials and, thus, are not representative of the general population given DMTs for MS therapy. Clinical trials also tend to lack potential confounders that bias findings as subjects tend to be exceptionally well screened compared with the general population of patients.
A more recent review of the outcomes of pregnancies from a global drug safety database containing individual case safety reports received in the postmarketing setting, and safety data from clinical trials of subcutaneous IFNβ-1a showed no difference in rates of spontaneous abortions (49 out of 425; 11.5%) or major congenital malformations in live births (four out of 425; 0.9%) from those of the general population [62]. This report was subject to several limitations: there was no control group for a proper analysis to take place, instead, general population data were used to compare the rates of outcomes observed. The authors also indicated that the outcomes for 111 pregnancies collected prospectively were not available, which may have introduced bias. In addition, factors such as previous pregnancy history, ethnicity, maternal age, family history, reasons for termination (e.g., social rather than medical) are not usually taken into account in safety data.
A small prospective study performed through teratogen information services (TIS) was conducted to compare outcomes of patients who conceived while on IFNβ, patients who conceived after discontinuing IFNβ and healthy control subjects [63]. TIS studies typically identify pregnant women when they call to request counseling about the teratogenic potential of medications or other products and then follow the women to determine pregnancy outcome. In this study, mothers contacted the Motherisk Program help lines at the Hospital for Sick Children in Toronto (ON, Canada) and after the expected date of confinement, interviewers followed-up on the pregnancy outcome following the participating women's informed consent [63]. The authors found an increased risk for spontaneous abortions and stillbirths (nine out of 23; 39.1%) and a decreased mean birth weight in the exposed groups compared with the healthy controls, but no difference with regard to prematurity. However, it is important to note the large difference in the gestational age at first contact: the enrollment of IFNβ mothers was at week 4.2, while the non-MS therapy control group was at 9.3 weeks, which may have contributed to the difference in rates of spontaneous abortion.
Another more recent and larger German TIS study also showed an increased rate of spontaneous abortions among pregnancies specifically exposed to IFNβ-1b in early pregnancy (five out 18; 28%), but not in those exposed to IFNβ–1a (two out of 42; 4.8%), compared with the treatment-naive MS control and the non-MS control groups [64]. There were no major birth defects in either of the IFNβ cohorts. Although the adjusted mean birth weight was in normal range for all groups, newborns who were exposed to IFNβ medications had a significantly lower birth weight (p < 0.001). It is difficult, however, to extrapolate the results of TIS studies to the population as a whole, mostly because of the small sample size (even though this is the largest prospective cohort study on IFNβ) and the fact that women who call a teratogen information service tend to be from higher socioeconomic groups and are therefore less likely to have babies with many birth defects [65].
In Italy, a multicentric prospective cohort study compared the rates of spontaneous abortions and other outcomes including birth weight, birth length, preterm delivery and cesarean delivery between women who discontinued IFNβ therapy less than 4 weeks from conception (exposed), and those who had discontinued the drug at least 4 weeks from conception or who had never been treated with DMTs (unexposed) [66]. No increased rates of spontaneous abortions were noted between exposed and unexposed groups, but a trend for exposure with a lower birth weight and length was observed (but not with low birth weight per se (i.e., intrauterine growth restriction). However, the study lacked control for confounders or longitudinal data on live births.
An Argentinean multicentric retrospective cohort study was the first to show a significantly increased rate of birth defects in pregnancies exposed to any DMT at conception or less than 15 days before conception (18.8%, three out of 16) compared with (2.3%, two out of 88) (odds ratio: 10.8; 95% CI: 1.6–71.0) [20]. The following birth defects were noted: ureteral stenosis following a 30-day exposure to G A, ductus and ovale foramen after 15-day exposure to IFNβ-1b and hip dysplasia after 2-month exposure to IFNβ-1a, while birth defects in pregnancies not exposed to DMTs included retinopathy and IgA deficiency. No associations were found with premature delivery, spontaneous abortions or low birth weight between exposed and unexposed groups. The study suffered from its retrospective collection of data through surveys, limited size and lack of family and previous pregnancy history data. Another recent Canadian retrospective cohort study linking two provincial population-based databases was the first to find an increased incidence of assisted vaginal delivery in women exposed to DMTs prior to conception and/or during pregnancy compared with women who had never been exposed to DMTs (OR: 3.0; 95% CI: 1.0–9.2), but not with other delivery or neonatal outcomes [67]. Other retrospective cohort studies with smaller exposure numbers did not show any significant differences in rates of spontaneous abortions or malformations between exposed and unexposed individuals [68,69].
In total, seven analytical studies, including two clinical trial reviews have looked specifically at the effect of maternal IFNβ exposure during pregnancy on the fetus (
Glatiramer acetate
GA, another DMT, has also been used to treat RRMS [70]. Animal teratology studies, performed only by the manufacturer and not published or peer reviewed, report no risk to fetal development in rats and rabbits with doses up to 18- and 36-times the maximum human therapeutic dose [203]. Few studies have been conducted so far in human pregnancies to assess the safety of GA on the fetus. The manufacturer's postmarketing surveillance data (presented only as an abstract) showed 215 live births among 277 known outcomes, the majority of which reported to be exposed to GA in the first trimester [71].
Their data showed a 17% spontaneous abortion rate, which approximates the rate observed in the general population, and six (3.1%) of 193 liveborn infants were found to have congenital anomalies, which were different in each case. Again, there was no information on potential confounders such as maternal age, ethnicity, previous pregnancy history or family history.
In a German TIS study of 31 women exposed to GA in early pregnancy (median duration of 6.9 weeks, with 25% taking G A beyond week 7), no increased risks of spontaneous abortions, premature birth or low birth weight were reported when compared with either pregnant MS patients who had not taken any DMTs or to healthy controls [65]. There were two cases of major birth defects in the exposed group, (one club foot, one atrioventricular canal defect), but no significant association was found.
In a prospective observational case-series of 14 pregnancies in 13 women exposed to GA (nine conceived under GA) that were systematically followed-up after pregnancy, most mothers continued taking GA until after pregnancy, except for three cases who stopped just before week 5; and one pregnancy where intake was stopped at week 19. There were 13 live births (one twin pregnancy) and two spontaneous abortions. No birth defects were identified among the 13 babies born. In most cases, birth weights recorded were within the normal range, except in two cases; one was of the woman who delivered twins prematurely, and the other stopped GA at 19 weeks who had a relapse in the third trimester and delivered the baby at 38 weeks [72].
Another Brazilian case series followed 11 pregnant women who were exposed to GA for at least 7 months into pregnancy [73]. All pregnancies resulted in live births (one pregnancy with twins delivered at 31 weeks) with no cases of malformations, but there was one case of neonatal death with an unknown cause. The twin babies were premature and had low birth weight (not unusual for multiple gestations), while the birth weights recorded in the rest of the cohort were of normal range. There were six cases of cesarean delivery (54.5%), but this is compatible with the rates for the general Brazilian population.
Natalizumab
Natalizumab is usually indicated for RRMS patients with high disease activity or when there is a failure to respond to first-line agents. It is not a first–line agent because of severe and potentially lethal side effects, specifically progressive multifocal leukoencephalopathy [74].
Thus, data on its effect in human pregnancy are very scarce. Animal studies on guinea pigs show no teratogenic effects when treated with one- to ten-times the recommended human therapeutic dose [75]. However, studies on cynomolgus monkeys show an increased rate of miscarriage with a high dose treatment (more than 190-times the recommended human therapeutic dose) [76].
In a small German prospective cohort study following 35 pregnancies of MS patients with accidental exposure to natalizumab in very early pregnancy (starting from 8 weeks prior to last menses), the rates of adverse outcomes in this group were compared with 23 pregnant MS patients without DMT exposure. In this report, the first of its kind, the rate of adverse outcomes, including low birth weight, prematurity or congenital anomalies, was not statistically different from that of the control group. Five (14.3%) of the 35 pregnancies resulted in a spontaneous abortion, which was a slightly but not significantly higher rate than that observed in the non-DMT-exposed control group [77].
Biogen Idec, Inc. and Elan Pharmaceuticals have implemented the Tysbari® Pregnancy Exposure Registry (TPER), which is an ongoing registry following-up on pregnant women with MS or Crohn's disease who are exposed to natalizumab within 3 months prior to conception or during gestation [78]. With 277 pregnancy outcomes reported to date, malformations occurred in 23 outcomes. The authors note, however, that a rate for major birth defects was not calculated as the sample size of enrolled patients was small and had limited power to detect an increased risk of major birth defects. The rate of spontaneous abortions was consistent with background rates [78].
Reports on the outcome of three exposed pregnancies while on natalizumab have been published [79,80]. In one pregnancy, natalizumab treatment was accidentally administered during pregnancy through seven infusions until 31 weeks gestation [79], while the medication was administered in the periconceptional period and until 20 weeks gestation in the other two cases [80]. The outcomes of all three pregnancies showed normal prenatal and postnatal development. However, possible long-term effects in those cases could not be ruled out.
It is important to note that current data provide no strong evidence that either IFNβ, GA or natalizumab are likely to pose a substantial teratogenic risk, but the data remain insufficient to state absolute safety and to confirm the absence of later-onset problems in live births.
Fingolimod
Fingolimod is a S1P-receptor modulator that has been recently approved as a new DMT for MS therapy [81]. No epidemiological studies of birth defects among infants born to women who were treated with fingolimod during pregnancy have been reported. Animal teratology studies of this drug conducted by the manufacturer have not been published in peer-reviewed literature.
Immunosuppressive agents
MS is an inflammatory, B-cell- and T-cell-mediated autoimmune disease affecting the CNS. Immunosuppressive agents were widely used over the past 30 years before DMTs were approved by the FDA. In addition, these are currently being offered as a combination therapy or monotherapy to ameliorate the effect of the disease process, especially when DMTs fail to be as effective in some patients [82]. The most common immunosuppressive agents used for MS are mitoxantrone (Novantrone®, Serono Laboratories, Inc., MA, USA), cyclophosphamide (CPA; Cytoxan®, Bristol-Myers Squibb, NY, USA), azathioprine (Imuran®, GlaxoSmithKline, London, UK), methotrexate (MTX) and mycophenlate mofetil (MMF; CellCept®, Roche, Basel, Switzerland).
Immunosuppressive agents used in treating multiple sclerosis and their possible fetal risks.
TERIS: the Teratogen Information System and the online version of Shepard's Catalog of Teratogenic Agents (internet database) [203].
The US FDA Pregnancy Risk Categories are currently being revised and will be replaced. A new rule has been proposed with major revisions to prescription drug labeling to more completely inform the use of medicines during pregnancy and breastfeeding [204].
Summary of the US Pregnancy Risk Categories (A: no evidence of fetal harm in human studies; B: no evidence of fetal harm in animal studies; C: evidence of fetal harm in animal studies or no data available; D: evidence of fetal harm in humans; use may be justified in some circumstances; X: evidence of fetal harm in human; not indicated for use in pregnancy)
Mitoxantrone is the only immunosuppressive agent that is currently approved by the FDA for MS treatment for SPMS, PRMS and worsening RRMS [83]. Mitoxantrone is contraindicated in pregnancy and is considered a potential human teratogen because of its mechanism of action. Animal studies reported an increase in fetal resorptions and decreased fetal weights in rats [84], as well as retarded development of the fetal kidney and increased incidence of premature delivery in unpublished studies performed by the manufacturer. In human pregnancy, only one case report has been published on accidental maternal exposure to mitoxantrone in pregnancy (from conception until 29 weeks gestation) [85]. The pregnancy was complicated by oligohydramnios and intrauterine growth restriction until delivery through a cesarean section. The baby had no malformations but was born at a low birth weight (less than a third percentile).
CPA is an immunosuppressive alkylating agent that suppresses rapidly dividing cells, such as B and T cells. It has been used off label to treat active relapsing and progressive forms of MS [83]. CPA is shown to be teratogenic in animal studies when given to rats, mice, chicks and rabbits, producing similar CNS and skeletal anomalies [86–89] as well as craniofacial malformations in rhesus monkeys [90]. Human data are scarce with no available prospective observational studies of CPA exposure in pregnancy and its effect on fetal outcome. A recent review on the teratogenic potential of CPA gathered information on nine confirmed cases of teratogenesis induced by inadvertent exposure to CPA during early pregnancy [91]. The malformations reported were similar and included hydrocephalus, micrognathia, ectrodactyly, cleft palate and exencephaly. From these data, a syndrome of CPA embryopathy was described [92]. However, data on maternal CPA exposure during the second and third trimesters of pregnancy have indicated that it is considered relatively safe [93].
MTX has been suggested to have some therapeutic benefit to the progressive forms of MS and has been used off label to treat MS [94]. MTX acts as a folate antagonist and thus is highly contradicted during pregnancy in lieu of its teratogenic potential that has been demonstrated in animal studies and multiple human reports [95,96]. The outcome in at least 24 cases of mothers exposed to MTX during the first trimester of pregnancy has been reported with a very uncommon and characteristic pattern of congenital anomalies, including abnormal head shape, large fontanelles, craniosynostosis, ocular hypertelorism and skeletal defects, as well as mental retardation [97,203]. The teratogenic risk of MTX may be dose related, as the frequency of malformations occurring in pregnancies exposed to MTX appears to be lower with lower doses [97,98]. No information is available on exposure during the second or third trimester of pregnancy.
Azathioprine, indicated for rheumatoid arthritis and patients with renal transplants, has been used off label for the treatment of RRMS and SPMS [83]. Increased frequencies of congenital anomalies have been reported in animal studies [99]. Two population-based cohort studies [100,101] and one TIS study [102] indicated a higher prevalence of congenital anomalies in infants born to mothers who had taken azathioprine in the first trimester of their pregnancy, but the increased rates observed did not reach statistical significance in comparison with their control groups. No recurrent pattern of congenital anomalies was apparent in more than 40 clinical series or among published case reports of infants with malformations whose mothers were treated with azathioprine during pregnancy [99]. Azathioprine exposure in later pregnancy may also be associated with increased frequencies of prematurity, fetal growth retardation or fetal or neonatal death [99–101]. In general, women treated with azathioprine are at an increased risk of having children with congenital anomalies, but it is not clear to what extent this is related to the treatment versus the illness for which it is used as a treatment or to concomitant use of other medications [203].
MMF is an oral immunosuppressant used in patients with solid organ transplants and in treating autoimmune diseases. At least 16 children with a recurrent pattern of malformations have been reported after maternal MMF exposure during the first trimester of pregnancy [103,104]. The most common features reported were unilateral or bilateral microtia or anotia, cleft lip with or without a cleft palate, micrognathia, congenital heart defects, diaphragmatic hernia, agenesis of the corpus callosum and ocular coloboma. However, the absolute risk of fetal malformations associated with maternal MMF treatment during the first trimester of pregnancy has not been determined, and the natural history of this embryopathy is still unknown.
Other management methods used for the treatment of disabling acute MS relapses that affect function include high-dose oral or intravenous corticosteroids. Corticosteroids are regarded to be generally safe when used in later pregnancy, but a slightly increased risk for orofacial clefts have been reported to be associated with first trimester corticosteroid exposure (most often prednisone or prednisolone) in two case–control studies [105,106]. Use of corticosteroids in the postpartum period has been shown to have a slight beneficial effect on decreasing postpartum relapses [107]. For MS patients with acute relapses who cannot tolerate steroids, intravenous immunoglobulin (IVIG) therapy is usually offered off label as an alternative [83]. It appears that IVIG therapy during pregnancy is not associated with severe adverse reproductive outcomes, including malformations, and is well tolerated during pregnancy [108]. In several studies, IVIG therapy has been shown to significantly lower the risk of postpartum relapses in women with MS [108–110].
In conclusion, most immunosuppressant agents are contraindicated during pregnancy because of their likely teratogenic potential, except for azathioprine, where its risk of pregnancy exposure is thought to be minimal, and for mitoxantrone, where its teratogenic risk is still undetermined. It is recommended that women anticipating pregnancy or who become pregnant while on an immunosuppressant stop taking it as there are not sufficient data to clearly state that any therapy is entirely safe at conception or during gestation.
Symptom-specific therapies
Several symptoms that women with MS experience when they are not pregnant (e.g., urinary urgency and frequency, fatigue, lower extremities paresthesias and gait problems) may be aggravated during pregnancy. Most of the medications that are used to treat these symptoms are either contraindicated during pregnancy or have a minimal or undetermined adverse effect on the fetus (
Symptom management agents used in multiple sclerosis patients and their pregnancy risk.
TERIS: the Teratogen Information System and the online version of Shepard's Catalog of Teratogenic Agents (internet database) [203].
The US FDA Pregnancy Risk Categories are currently being revised and will be replaced. A new rule has been proposed with major revisions to prescription drug labeling to more completely inform the use of medicines during pregnancy and breastfeeding [204].
Summary of the US Pregnancy Risk Categories (A: no evidence of fetal harm in human studies; B: no evidence of fetal harm in animal studies; C: evidence of fetal harm in animal studies or no data available; D: evidence of fetal harm in humans; use may be justified in some circumstances; X: evidence of fetal harm in human; not indicated for use in pregnancy).
Data taken from [11].
Other maternal considerations in MS patients
Contraceptive methods
Estrogen and progesterone-based OCP agents are well tolerated in MS patients and may even be beneficial. The findings of a cohort study conducted in the UK using diagnostic criteria supplied by pediatricians showed no greater risk of MS incidence with OCP use [111], while in another nested case–control study, the authors found a 40% reduction in the risk of MS among OCP users [112]. In yet another prospective cohort study in the USA, the authors also did not find a lasting protective effect of OCP use on the risk of MS [43].
It is important to note that an increased risk of deep venous thrombosis has been associated with OCP use in patients with low mobility and spasticity. Factors contributing to such an increased risk include being aged over 35 years and smoking. Therefore, these patients should be counseled accordingly about taking OCPs or HRT and alternative birth control methods should be offered [113]. Furthermore, the effectiveness of OCPs may be reduced with concomitant use of other medications used to treat common MS symptoms, including carbamazepine, some broad–spectrum antibiotics and rifampicin [114].
Fertility & sexual dysfunction
In theory, it was thought that MS may affect fertility due to abnormalities in the hypothalamic– pituitary–adrenal axis or the hypothalamic– pituitary–gonadal axis [115]. However, there is no evidence to support this. Women (and men) with MS do not suffer from infertility problems as part of the natural history of the disease, but some DMTs, such as IFNβ, which are associated with menstrual irregularities, could be associated with infertility [n]. Immunosuppressants such as mitoxantrone and cyclosphosphamide, have been documented to affect fertility [116].
Thus, although the natural history of untreated MS may not affect infertility, it appears that female MS sufferers are more likely to be childless compared with the general population [39]. In a large study aimed at determining reproductive attitudes of North Americans diagnosed with MS, almost 80% did not become pregnant following diagnosis with MS. The most common MS-related reasons for this decision was fear that their symptoms may interfere with parenting and taking care of the baby followed by concerns of burdening their partner and of their child inheriting MS [117].
A large prospective study in Finland showed that women with RRMS need artificial insemination to initiate pregnancy more often than the general population [52]. One of the factors that may potentially affect fertility in MS is sexual dysfunction, which is a frequent MS symptom affecting 50–90% of men and women. It is usually described in women as reduced libido, difficulty in achieving orgasms, changes in sensation and vaginal dryness causing dyspareunia. Therefore, it is important to bring up these issues during reproductive counseling as effective therapies in the form of medications or counseling techniques can be sought to symptomatically manage this condition [113].
Inheritance & genetic counseling
MS is a complex disease with both genetic and environmental factors involved. No prenatal test exists for MS inheritance and transmission is thought to be epigenetic in nature with more than 70 susceptibility genetic loci identified to date [6]. Children born to parents where one has MS have approximately 20–50-times the risk of the general population of developing MS, but the absolute rate itself is low (3–5%) [118]. It is estimated that the risk of developing MS if one parent has MS is 4%, assuming noone else in the family has MS, while if both parents have MS, the risk can be as high as 30% [4]. These figures need to be discussed with patients planning a pregnancy through their reproductive counseling or genetic counseling sessions.
Breastfeeding
Initial studies suggested that postpartum relapse rates may be reduced in MS mothers who have breastfed exclusively for at least 2 months postpartum [12,28,119]. Several other studies, however, did not support such findings [26,120–122]. A large-scale Finnish prospective investigation did not find a statistically significant difference in postpartum relapse rates between breastfeeding or bottle-feeding mothers, although a higher but insignificant relapse rate was shown for mothers who did not breastfeed or breastfed for less than 2 months [120]. Another relatively small study of 35 women with RRMS also failed to show a protective effect of breastfeeding on postpartum relapses [121]. In a larger Italian prospective cohort study following-up on a sample of 302 deliveries of 298 women, postpartum relapses were predicted only by relapses before and during pregnancy, while breastfeeding was not shown to have an effect [122]. The authors of this study suggest that the reported association between breastfeeding and a lower risk of postpartum relapse reported in other studies could be due to different patient behavior, biased by the disease activity [122].
It remains generally advisable for new mothers with MS to consider breastfeeding their infants unless their MS disease course is severe enough that being medicated is a necessity or fatigue is a major issue. As most mothers resume DMTs at some point postpartum, there is not enough data to understand the effect of DMTs taken postpartum on breastfed babies and it is not known whether they even pass through the mother's milk, therefore DMTs remain contraindicated for breastfeeding mothers.
Executive summary
Multiple sclerosis (MS) is the most common acquired neurological disorder other than trauma affecting young adults of Northern European descent.
There are four recognized disease patterns in MS: the relapsing-remitting; the secondary progressive MS; primary progressive MS; and progressive-relapsing MS.
Females are affected approximately three-times as often as males and most patients are affected during their reproductive years.
With advances in diagnostic procedures, MS patients are now receiving a diagnosis and therapy at a younger age and thus are younger, less disabled and more likely to contemplate a pregnancy compared with their historical counterparts.
Overall, the frequency of relapses appears slightly lower during the first and second trimester but decreases significantly by 70% during the third trimester.
Postpartum, the relapse rate during the first 3 months is substantially higher than that to prior pregnancy but then returns to their prepregnancy rates.
Some suggested predictors identified for the likelihood of postpartum relapses, in at least some women, are high relapse rate in the prepregnancy year, increase relapse rate during pregnancy and a higher MS disability score.
Most studies have not found any adverse effect of pregnancy on long-term disease progression and disability, with some even showing favorable effects for having had a pregnancy.
Inconclusive data are available on MS onset after conception or during gestation with some indication of pregnancy having a protective effect.
MS itself does not appear to increase adverse pregnancy outcomes (e.g., spontaneous abortions, stillbirth, cesarean delivery, premature birth, birth defects) compared with women who did not have MS.
IFNβ is currently the most commonly used of all DMTs as a line of therapy for MS patients.
In total, seven analytical studies, including two clinical trial reviews, have looked specifically at the effect of maternal IFNβ exposure during pregnancy on the fetus. Because these medications are contraindicated during pregnancy, the amount of data available to study this question remains limited. Therefore evidence is insufficient to make conclusions about the safety of these drugs and whether or not such contraindication is actually warranted.
For both glatiramer acetate and natalizumab, the current data suggest that exposure during pregnancy is unlikely to pose a substantial teratogenic risk, but the data remains insufficient to state absolute safety and the absence of later onset problems in live births.
Most immunosuppressant agents are contraindicated during pregnancy because of their likely teratogenic potential, except for azathioprine, where its risk of pregnancy exposure is thought to be minimal, and for mitoxantrone where its teratogenic risk is still undetermined.
Likewise, most symptom-specific therapies are either contraindicated during pregnancy or have a minimal or undetermined adverse effect on the fetus.
Estrogen and progesterone-based oral contraceptive agents are well tolerated in MS patients and may even be beneficial.
Some immunosuppressants such as mitoxantrone and cyclosphosphamide have been documented to affect fertility.
Current evidence suggests that postpartum relapse rates may be reduced in MS mothers who have breastfed exclusively for at least 2 months postpartum.
Larger population-based studies that have sufficient power to assess the teratogenic potential of DMTs and any long-term effects on neurobehavioral function in children are needed to evaluate the magnitude and the nature of risks associated with maternal DMT treatment during pregnancy.
It remains inconclusive whether breastfeeding produces beneficial effects in reducing MS relapses and more studies should therefore be directed to estimate the range of benefits from exclusive breastfeeding.
Future perspective
The benefit of relapse reduction during pregnancy seems to be greater than that seen with currently available standard therapies for MS. Most patients can therefore expect an improvement in their MS during pregnancy despite having to withdraw from their usual MS treatment. However, having said this, it must be remembered that achieving a pregnancy may take some time and the effect of no therapy during this period is unknown. It remains inconclusive whether breastfeeding produces beneficial effects in reducing MS relapses and more studies should therefore be directed to estimating the range of benefits from exclusive breastfeeding.
Immunosuppressant agents are contraindicated at conception, during gestation or while breastfeeding. The teratogenic risk of DMTs, the most commonly used medications to target MS relapses, is not fully determined due to insufficient data in the literature. A small increase in the risk of IFNβ has been suggested and IFNβ therapy is contraindicated during pregnancy. It is important that the potential risks and benefits of discontinuing (or continuing) IFNβ treatment during pregnancy be assessed thoroughly and discussed with each patient on an individual basis, while offering the possibility to switch to other DMTs, such as GA, which does not point to any major concern, but there is too little available data to make any definite conclusions on the safety of this drug. Larger population-based studies that have sufficient power to assess the teratogenic potential of DMTs and any long-term effects on neurobehavioral function in children are needed to evaluate the magnitude and the nature of risks associated with maternal DMT treatment during pregnancy.
Footnotes
S Alwan is the recipient of a postdoctoral fellowship from the Multiple Sclerosis Society of Canada. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.
No writing assistance was utilized in the production of this manuscript.