Management of women with multiple sclerosis through pregnancy and after childbirth

Abstract

Multiple sclerosis (MS) is a major acquired neurologic disease of young adults. The prototypic patient is a young woman of reproductive age. Gender preference is becoming more pronounced, since MS is increasing specifically among women. Any healthcare provider who deals with MS must be prepared to discuss pregnancy issues, and provide appropriate counseling. This is now complicated by the availability of multiple treatment options. There is growing literature on which to base recommendations, particularly regarding washout periods. After a brief background introduction, this review will discuss state-of-the-art family planning counseling in the treatment era, divided into prepregnancy, pregnancy, and postpartum MS issues.

Background

Pregnancy is a major concern for many patients with multiple sclerosis (MS) [Bove et al. 2014; Carvalho et al. 2014; Miller et al. 2014]. MS is the most common acquired neurologic disorder of young adults, short of trauma, with at least 2.5 million individuals affected worldwide. The disease shows a female predominance that now approximates 3 to 1. This gender bias will become more pronounced, because MS is on the rise among young women. The prototypic patient is a woman of reproductive age. Specific management questions and concerns surround the prepregnancy, pregnancy and postpartum periods (Table 1). This is complicated by the fact that MS has entered a treatment era. Unfortunately there are no recognized and implemented guidelines [Borisow et al. 2014; Wundes et al. 2014].

Table 1.

Pregnancy-related management issues in multiple sclerosis.

Prepregnancy

• Fertility and fetal development

• Impact on the female versus male patient

• Genetic risk

• Basic counseling

• Pregnancy impact on MS prognosis (early and late)

• Contraception choice

• Impact of MS on ability to care for children

• Economic and social burden of having children

• DMT use and recommended washout periods

• Assisted reproductive technology/in vitro fertilization

Pregnancy

• Disease activity

• DMT use

• Evaluation and treatment of relapses and disease symptoms

• Delivery and anesthesia choices

Postpartum

• Disease activity

• Breastfeeding

• DMT use

• Impact on infant/child development

MS, multiple sclerosis; DMT, disease-modifying therapies.

The major current pregnancy issue is pharmacotherapy [Lu et al. 2012; Vukusic and Marignier, 2015]. There are now multiple disease-modifying therapies (DMTs) for relapsing forms of MS, but they are not recommended for use in patients who are pregnant, trying to become pregnant, or who are breastfeeding. Recent articles have called for these initial rules to be liberalized [Coyle, 2014a; Hellwig, 2014]. After a brief introduction covering the DMTs, this review will discuss management recommendations for MS patients around the prepregnancy, pregnancy, and postpartum periods in the MS treatment era. This is based on an up-to-date review of the literature.

Disease-modifying therapies

There are 13 distinct approved DMTs for relapsing forms of MS. They encompass eight different mechanisms of action (Table 2). The European Medicines Agency (EMA) has a rating scale for drug pregnancy risk. The United States Food and Drug Administration (FDA) previously used ratings that involved letters (A, B, C, D, X). For new drugs, this system has been replaced with descriptions of fetal risk summary, clinical considerations and data. The FDA plans to rewrite existing drug ratings over the next several years.

Table 2.

Approved multiple sclerosis (MS) disease-modifying therapies (DMTs) and pregnancy, washout, and breastfeeding recommendations.

DMT	Dosing and brands	Pregnancy ratings (USA/Europe)*	Prepregnancy washout	Pregnancy use	Breastfeeding
Interferon beta (IFNβ)	SC IFNβ-1b 250 µg every other day (Betaseron, Extavia) IM IFNβ-1a 30 µg weekly (Avonex) SC PEG IFNβ-1a 125 µg every 2 weeks (Plegridy) SC IFNβ-1a 44 or 22 µg 3 times weekly (Rebif)	C/Category 2	Probably not necessary (0–1 month)	Probably acceptable	Probably acceptable (levels 0.006% of maternal dose) [Hale et al. 2012]
Glatiramer acetate	SC 40 mg 3 times weekly, or SC 20 mg daily (Copaxone, Glatopa)	B/Category 2	Not necessary	Probably acceptable	Probably acceptable
Fingolimod	PO 0.5 mg daily (Gilenya)	C/Category 2	2 months	Not used	Avoid (no data)
Teriflunomide	PO 14 mg (also 7 mg in USA) daily (Aubagio)	X/Category 1 (pregnancy must be ruled out prior to use)	Accelerated elimination procedure (until blood level < 0.02 µg/ml)	Not used	Avoid (no data)
Dimethyl fumarate	PO 240 mg twice a day (Tecfidera)	C/Category 2	Probably not necessary (0–1 month)	Not used	Avoid (no data)
Natalizumab	IV 300 mg monthly (Tysabri)	C/Category 2	Probably not necessary (0–1 month)	Can be considered for very active patients	Present in breast milk
Alemtuzumab	IV 12 mg daily for 5 days in the first year, 3 days in the second year (Lemtrada)	C/Category 1	4 months	Not used	Avoid (no data)
Mitoxantrone	IV 12 mg/m² every 3 months (lifetime max 140 mg/m²) (Novantrone)	D/Category 2 (pregnancy must be ruled out prior to use)	6 months	Not used	Avoid (still present in breast milk at 4 weeks)

SC, subcutaneous; IM, intramuscular; PEG, pegylated; IV, intravenous; PO, taken orally.

^*USA ratings have been replaced by description of animal/human data.

The interferon betas have shown dose-dependent first trimester abortofacient effects in primate models at 2.8 to 40 times the recommended human dose; this effect has not been noted in human studies [Bayer HealthCare Pharmaceuticals, Inc. 2012; Coyle et al. 2014; Sandberg-Wollheim et al. 2011]. After injection, serum levels peak between 8 to 24 hours, then decline [Khan et al. 1996]. Interferon beta is not believed to cross the placenta; a sister type-1 interferon (alpha) was shown by two assays not to penetrate placental tissue [Waysbort et al. 1993]. Some reports suggest maternal interferon beta exposure is associated with lower infant birth weights and length, and higher incidence of premature births [Amato et al. 2010; Boskovic et al. 2005; Patti et al. 2008; Weber-Schoendorfer and Schaefer, 2009]; but other analyses do not confirm these findings [Coyle et al. 2014; Hellwig et al. 2012; Romero et al. 2015]. This could be a disease-specific rather than drug-specific issue, since other studies report lower birth weights and birth lengths with MS pregnancies in general [Dahl et al. 2006, 2008; Ebrahimi et al. 2015]. However, a recent large study found no link between MS and birth weight, or preterm delivery [Van der Kop et al. 2011].

Glatiramer acetate is also not believed to cross the placenta [Cree, 2013]. There are no measurable blood levels, and it is rapidly hydrolyzed at the site of injection [Ziemssen et al. 2001]. Glatiramer acetate has not been associated with negative pregnancy effects in either animal or human studies, and has the best pregnancy rating among the DMTs.

Because the interferon betas and glatiramer acetate have been available for years, they have the largest accumulated pregnancy exposures. No significant issues have been identified. Recent commentaries suggest these DMTs do not require a washout prior to attempting pregnancy [Amato and Portaccio, 2015; Coyle, 2014a]. Glatiramer acetate, and to a lesser extent interferon beta, have been used during pregnancy for women in whom there is concern about extended treatment-free periods [Dung and Panda, 2014; Fragoso et al. 2010; Hellwig and Gold, 2011; Salminen et al. 2010]. For the pregnant MS patient, the appropriate counseling remains that the preferred option is no DMT. However, selected pregnant women may choose to continue glatiramer acetate or interferon beta, when they feel this is their best option. Both glatiramer acetate and interferon beta are probably compatible with breastfeeding. They would not be expected to enter breast milk at high levels, due to their negligible or low blood levels and higher molecular weights. Any trace amounts would likely be depolymerized in the neonate’s gastrointestinal tract after oral intake. In a small study, breast milk contained only 0.006% of the maternal dose of intramuscular interferon beta-1a [Hale et al. 2012].

None of the oral DMTs are recommended for use while pregnant or breastfeeding. Teriflunomide has the most profound warning. It shows selected teratogenic and embryo lethal effects in multiple animal species, at doses below those used clinically. Human pregnancy exposures to teriflunomide or leflunomide (its precursor used in rheumatoid arthritis) have not demonstrated teratogenicity, however. A recent teriflunomide series reported on 83 female and 22 male pregnancy exposures, without identified issues [Kieseier and Benamor, 2014]. There is a rapid elimination protocol (using oral cholestyramine over several days) to quickly lower teriflunomide levels to less than 0.02 µg/ml; otherwise it can persist in the body for up to 2 years. Mean half-life is 16–18 days. European product information on teriflunomide does not note any implications for male MS patients. The drug is detected in semen in very low concentrations, and could be inoculated into the female partner if a barrier contraceptive technique is not used. Estimated exposure would be over 100 fold less than from direct oral dosing [Kieseier and Benamor, 2014]. Teriflunomide has been detected in rat milk following a single oral dose.

In rabbit and rat models, fingolimod is associated with fetal malformations, death, growth retardation, and learning deficits. Depending on the animal model, problems were seen at doses below the clinically used amount, as well as with much higher doses. The S1P receptor is involved in vascular formation during embryogenesis. The Fingolimod Pregnancy Registry reported on 66 human exposure cases, and noted a 7.6% rate of abnormal fetal development [Karlsson et al. 2014]. Fingolimod has a half-life of 6–9 days [Novartis Pharmaceuticals Corporation, 2015]. It is known to cross the placenta, is present in rat milk, and requires a 2-month washout before attempting pregnancy [Novartis Pharmaceuticals Corporation, 2015]. There is a concern for rare rebound activity when fingolimod is discontinued [Faissner et al. 2015].

In rats, dimethyl fumarate is associated with embryo lethality at doses two times higher than the approved human dose. Testicular toxicity was noted in mice, rats, and dogs. Male rats showed increase in nonmotile sperm. Human pregnancy exposures to date have not detected any deleterious signal. It is not clear that dimethyl fumarate requires any washout, due to its short half-life (approximately one hour) [Biogen Idec, Inc., 2014]. Currently all the oral agents are avoided in breastfeeding patients.

Natalizumab is an IgG4 humanized monoclonal antibody. In guinea pigs it has been associated with decreased pup survival at 7 times the human dose, and in primates with reversible fetal hematologic abnormalities, at doses 2.3 times the human dose. Alpha 4-integrin, its antigen target, is expressed on human uterus epithelia as well as embryonic tissue. This antigen plays a role in fertilization, implantation, placental, and cardiac development [Duquette and Prat, 2015]. However, no human implantation or embryonic defects have been reported, and no cardiac issues have been found [Cree, 2013]. In a recent prospective, controlled, observational study, 101 MS women exposed to natalizumab in the first trimester were compared with 78 pregnant MS women without DMT exposure, and 97 healthy control women [Ebrahimi et al. 2015]. There was no difference in major malformations, low birth weight, or premature births. Higher miscarriage rates and lower birth weights were noted in the two MS groups, which were not significantly different. Natalizumab washout recommendations have ranged from 1 to 3 months, but all agree it should be as short as possible. There is risk not only for return of disease activity, but also rebound activity [De Giglio et al. 2015; Verhaeghe et al. 2014]. A recent report suggested one could justify no washout, since monoclonal antibodies do not cross the placenta until the second trimester [De Giglio et al. 2015]. IgG is the only antibody isotype to significantly cross the placenta (IgG1, followed by IgG4 and 3). This transfer does not start until week 13, and peaks in the third trimester [Palmeira et al. 2012]. Early on, the theoretic concern is that natalizumab could interfere with implantation. Natalizumab has been detected in human breast milk, with relative infant exposure as high as 5.3% of the maternal dose, and concentrations increasing over time [Baker et al. 2015]. It is not typically used by nursing mothers. Natalizumab has been used during pregnancy without notable teratogenicity, but newborns may experience transient hematologic abnormalities including anemia and thrombocytopenia [Haghikia et al. 2014]. One study noted an immune disturbance (abnormal lymphocyte chemotaxis) in two exposed neonates [Schneider et al. 2013].

Alemtuzumab (an IgG1 humanized monoclonal antibody) is a recently approved DMT. It is an induction agent, with prolonged suppression of disease activity after 2 cycles of treatment (5 days in year 1, and 3 days in year 2). Alemtuzumab is embryo lethal in mice, and can lower offspring lymphocyte counts. It is detected in the milk of lactating mice. All pregnant women with exposure should be checked for hypothyroidism. Placental transfer of antithyrotropin receptor antibodies has been associated with neonatal Graves’ disease with thyroid storm, in an infant born 1 year after DMT dosing. Before trying to become pregnant, a 4-month washout is recommended. This may be excessive, since blood levels are negligible or undetectable by 30 days after the last dose, and IgG does not cross the placenta until after the first trimester anyway. However this is a new DMT with limited pregnancy data, and the recommended washout should be followed for now.

Mitoxantrone is no longer used to treat MS in the United States. It is associated with growth retardation and premature delivery in animal models, and there is one case of Pierre Robin syndrome in human exposures [Hellwig et al. 2011]. A 6-month washout has been recommended prior to pregnancy. Mitoxantrone is excreted in human milk with concentrations of 18 ng/ml as late as 28 days after dosing [Azuno et al. 1995]. Drug treatment can produce amenorrhea, typically in those over age 35.

Among off-label therapies for MS, anti-CD20 and anti-CD25 monoclonal antibodies may become approved agents in the near future. These IgG antibodies would be actively transported across the placenta beginning during the second trimester. Anti-CD20 monoclonals include chimeric (rituximab) and human (ofatumumab) antibodies, as well as an experimental humanized (ocrelizumab) antibody. The greatest experience is with rituximab. In monkey models, exposure during pregnancy was only linked to decreased lymphoid tissue B cells in the newborn, with immunosuppression that returned to normal by 6 months [Biogen Idec, Inc. and Genentech Inc., 2014]. In humans, preconception or first-trimester exposure has not been associated with negative outcomes; use in the second and third trimester results in detectable cord blood levels [Østensen, 2014]. With regard to anti-CD25 monoclonal, which is under consideration for MS by the FDA, there are no available data on pregnancy risk.

Prepregnancy issues

Fertility and fetal development

Patients should be counseled that MS has no significant impact on the ability to conceive, on fetal development, or the ability to carry to term. Fertility does not appear to be affected by MS phenotype or DMT use [Roux et al. 2015]. Definitive data are lacking, although a recent study suggested MS patients showed decreased ovarian reserve [Thöne et al. 2015]. There is no disease-associated increase in spontaneous abortions, stillbirth, cesarean delivery, premature birth, or birth defects [Alwan et al. 2013; Ramagopalan et al. 2010]. There are conflicting data on whether babies born to MS mothers may be a little bit smaller in weight and length. This is not a definite association, and should not be a major concern.

Female versus male patients

Pregnancy issues and management concerns are for the most part limited to women. Two recent studies evaluated MS men who fathered children. A British Columbia database of 202 MS men versus 981 matched controls found no impact of paternal MS on birth weight or gestational age [Lu et al. 2014]. They concluded that paternal MS, and factors such as disease duration and disability, had no impact on birth outcomes. A second, smaller Italian study found no paternal MS association with adverse fetal outcome, even when exposure to DMTs was included as a factor [Pecori et al. 2014]. Analysis of semen quality in MS men is limited, with one study reporting lower total sperm counts, sperm motility, and percent normal sperm morphology [Safarinejad, 2008].

Genetic risk

MS is a complex disease involving interactions between environmental and genetic factors, with epigenetic mechanisms likely involved [Harbo et al. 2013; Küçükali et al. 2015]. Patients are concerned about risk to their children. One can counsel that MS is not an inherited disease; most patients (80%) have no affected relative, but there are a number of identified risk or susceptibility genes. Having a first-degree relative with MS increases disease risk from 0.13% to 2–2.5%. Risk is slightly higher when the first-degree relative is a sib versus a parent, supporting the impact of environmental factors [O’Gorman et al. 2013]. Risk is at least 30% in two instances: monozygotic twins, and when both parents have MS.

Basic counseling

Vitamin D deficiency increases risk for MS, and MS women were shown to have lower vitamin D concentrations during pregnancy and postpartum compared with controls [Jalkanen et al. 2015]. Vitamin D deficiency should be evaluated and treated prior to pregnancy. Any MS woman seeking to become pregnant should be counseled to take prenatal vitamins and folic acid, avoid alcohol and smoking, and assure good sleep hygiene and diet.

Pregnancy impact on prognosis

Up until the 1950s MS women were counseled to avoid pregnancy because it would make their disease worse. This turned out to be untrue. Some studies suggest pregnancy conveys a long-term benefit [Keyhanian et al. 2012; Masera et al. 2015; Runmarker and Andersen 1995; Verdru et al. 1994], while others find no long-term impact [Karp et al. 2014; Ramagopalan et al. 2012]. The most accurate current counseling would be to inform patients that pregnancy has no negative effect on long-term prognosis. This statement is valid for relapsing MS women, who make up 87–97% of all pregnancies [Coyle, 2014a]. One study did suggest pregnancy impact might be different for progressive MS women [D’Hooghe et al. 2012]. However, the number of pregnant progressive MS patients is so limited it would take a multicenter and probably multinational pregnancy registry to evaluate this. For now, progressive pregnant MS patients should be managed like relapsing patients, and if possible entered into an ongoing pregnancy database.

With regard to whether pregnancy affects risk for development of MS, a large Australian study looked at clinically isolated syndrome (CIS) patients and matched controls [Ponsonby et al. 2012]. They reported a cumulative benefit of pregnancies to decrease risk for CIS. Giving birth, or even having a terminated pregnancy seems to protect women against MS over the next 5 years [Hedström et al. 2014; Magyari, 2015]. Another small study of radiologically isolated syndrome (RIS) patients (n = 60) found that pregnancy (n = 7) was associated with increased risk for subsequent clinical attack in the postpartum period, and MRI disease activity [Lebrun et al. 2012]. This observation requires verification. For now, patients can be told that the data indicate pregnancy protects against developing MS. RIS patients who become pregnant should have more intense follow up.

Contraception choice

MS women capable of pregnancy should be counseled on contraception before they initiate a DMT. Failure rates differ based on chosen technique [Coyle, 2014b]. Abstinence is most effective (0% failure rate). Long-acting reversible contraception (subdermal rod or intrauterine devices) and tubal sterilization have failure rates of less than 1%. Oral, patch, injection, ring-hormonal contraceptives (5% failure rate), and barrier methods (11–32% failure rate) are much less effective. Among the DMTs, teriflunomide may increase levels of ethinylestradiol and levonorgestrel contained in oral contraceptives [Genzyme Corporation, 2012]. Of all available contraception techniques, the long-acting reversible methods should be encouraged based on efficacy, safety, and convenience [Committee on Gynecologic Practice, 2015].

Ability to care for children; economic and social burden

MS patients often have concerns that their disease will have a negative economic and psychosocial impact on bearing and raising children. In a survey of 5949 MS patients diagnosed during their reproductive years, 79% did not become pregnant after diagnosis [Alwan et al. 2013]. In 34.5%, this choice involved MS-related issues, in particular concerns that their disease would interfere with parenting, followed by concerns about burdening their partner and passing MS on to their children. Patients can be told pregnancy does not produce disability, and the modern DMT era appears to be changing the natural history of MS in a positive way. Choosing to have children should not be directed by a diagnosis of MS.

DMT use and washout

None of the DMTs are approved for use in women who are actively trying to become pregnant. A frequent controversy involves newly diagnosed young women, and whether they should forego DMT use to start a family. A large study from the MSBase global registry reported that prior DMT use, any time in the 2 years before pregnancy, resulted in a 45% decreased risk for postpartum relapse [Hughes et al. 2014]. The authors interpreted their study as providing evidence in favor of choosing to delay pregnancy, in order to commit to a DMT to control disease activity first. This would be particularly pertinent for MS women with high disease activity, greater disability, or a poor prognostic profile.

The DMT washout period should be as short as possible (Table 2). There is increasing consensus that glatiramer acetate and the interferon betas do not require washouts, and can be continued until pregnancy is confirmed. It is possible that natalizumab and dimethyl fumarate do not need washouts. When a washout is used, one consensus group proposed monthly pulsed corticosteroids until pregnancy is achieved in very active MS women, or those with a history of delayed conception [Bove et al. 2014].

Assisted reproductive technology and in vitro fertilization

About 10% of women have difficulties getting or staying pregnant. Assisted reproductive technology or in vitro fertilization can be used, with success rates as high as 39% in those under age 35. Estimates are that at least 1% of live births in developed countries use such technology [Bove et al. 2014]. MS women who undergo in vitro fertilization and do not conceive may be at increased risk for clinical and MRI disease activity in the 3 months post procedure [Correale et al. 2012; Hellwig and Correale, 2013; Michel et al. 2012]. This appears to be associated with use of gonadotropin-releasing hormone agonists, as opposed to antagonists, to produce multiple eggs. Although the number of patients studied is less than 100, MS patients should be counseled about this possible risk. Fertility experts must be aware of this data as well. One could avoid using an agonist, or elect to continue use of glatiramer acetate or interferon beta during the procedure.

Pregnancy issues

Disease activity

Pregnancy is an immunotolerant state. Clinical and MRI disease activity is suppressed, most clearly during the third trimester. Although precise mechanisms continue to be studied, it is coincident with increasing concentrations of a number of hormones (estrogens, progesterone, prolactin, glucocorticoids, leptin). This observation has fostered investigation of sex hormone therapy as a treatment for MS [Durand-Dubief et al. 2014; Sicotte et al. 2002]. In a published phase I open-label study, six relapsing MS patients benefited from oral estriol 8 mg daily, while four secondary progressive MS women did not [Sicotte et al. 2002]. In a phase II, 2-year combination trial (glatiramer acetate with add-on estriol/progresterone or placebo), confirmed annual relapse rate at 2 years was 0.25 versus 0.37 (p = 0.077), with mixed MRI effects [Voskuhl et al. 2016]. Unfortunately, the associated editorial emphasized that the study must be viewed as negative [Langer-Gould, 2016].

DMT use

The current standard of care is to avoid DMT use during pregnancy, and to stop a DMT as soon as pregnancy is recognized. Exposed patients should be encouraged to participate in a formal pregnancy registry. Among the DMTs, glatiramer acetate, interferon beta, and natalizumab have been used in pregnant patients, and there is increasing awareness that extensive glatiramer acetate and interferon beta data suggest they are safe to use [Dung and Panda, 2014; Fragoso et al. 2010; Salminen et al. 2010]. The decision to continue a DMT during pregnancy should involve formal documentation of the conversation involving the risks and benefits, and a clear indication the patient has made this informed decision.

Relapse

Clinical attacks are less common during pregnancy, but can occur and are more likely in the first two trimesters. There is no conclusive evidence that MRI scans up to 3T are associated with fetal harm [Bove and Klein, 2014]. Contrast should not be used without informed consent from the mother; typically it is avoided because gadolinium compounds cross the placenta and enter the fetal bloodstream. To treat the unusual relapse that occurs during pregnancy, there is no contraindication to conventional short-term corticosteroids (most often 1 g of intravenous methylprednisolone daily, given for several days without an oral taper). There is controversy as to whether first trimester steroid exposure is associated with cleft lip and palate abnormalities [Gur et al. 2004; Hviid and Mølgaard-Nielsen, 2011], but obstetricians feel short-term steroids can be used safely in all trimesters. Dexamethasone (which crosses the placenta with minimal metabolism) should always be avoided in favor of prednisone or methylprednisolone.

Disease symptoms

Pharmacotherapy for MS symptoms should be evaluated carefully and limited during pregnancy [Miller et al. 2014]. Drugs should be used at their minimum effective doses, for as short a time as possible. When there are options, drugs with the most favorable pregnancy ratings should be chosen. Since most pregnant MS patients are not disabled and are relatively early in their disease, they do not typically require significant pharmacologic symptom management.

Delivery and anesthesia choices

Having MS does not create a high-risk pregnancy, and the birth hospitalization period is not increased [Lu et al. 2013]. Any anesthetic choice or delivery method is acceptable [Pastò et al. 2012]. These are purely obstetric decisions. The one exception is the rare, more disabled pregnant MS patient, where it may be necessary to consider assisted vaginal delivery or even cesarean section.

Postpartum issues

Disease activity

The first 3 months postpartum are a recognized high-risk period for increased clinical and MRI disease activity. Activity then reverts back to the prepregnancy level [Paavilainen et al. 2007]. This may relate to rapid reversal of late pregnancy hormone levels. Factors associated with postpartum relapses have included higher relapse rate and disability prior to pregnancy, lack of prior DMT use, and relapse during pregnancy [Coyle, 2014b; Hughes et al. 2014; Vukusic et al. 2004]. In the original pregnancy in MS (PRIMS) cohort, almost 30% of MS women experienced a postpartum relapse during this period [Confavreux et al. 1998]. In the more modern MSBase registry, only 14% relapsed [Hughes et al. 2014]. In a recent small study (44 pregnant relapsing MS women and 20 controls), low maternal vitamin D levels were not associated with postpartum relapse risk [Runia et al. 2015]. The authors interpreted this as not supporting the argument for additional vitamin D supplements in pregnant MS women. As noted earlier, ideally this issue is dealt with before pregnancy occurs. Small-scale studies have evaluated postpartum prophylactic intravenous immune globulin, as well as pulse corticosteroids, to minimize relapses with reported success [De Seze et al. 2004; Haas and Hommes, 2007]. Hormonal therapy (percutaneous 17-β estradiol 100 µg weekly along with oral nomegestrol acetate 10 mg daily) was evaluated in the POPARTMUS trial, but showed no ability to suppress postpartum relapses or MRI activity [Durand-Dubief et al. 2014].

Breastfeeding

Although breastfeeding is considered to offer superior health benefits to using formula, a recent very large analysis comparing healthy sibs (one breastfed, the other bottle fed), did not confirm any advantage [Colen and Ramey, 2014]. MS women should be told it is acceptable to choose not to breastfeed.

Typically, the major postpartum issue is whether to breastfeed, or to start, or restart, a DMT. Until recently, these options were considered mutually exclusive. As noted earlier, both glatiramer acetate and the interferon betas can be considered with breastfeeding. This cannot be said for any of the other DMTs. For treated postpartum relapses, methylprednisolone levels in breast milk may be as high as 1.45% of maternal dose, but are much lower with a 2- to 4-hour delay following the intravenous infusion [Cooper et al. 2015]. This should be taken into account when using steroids to treat postpartum attacks in breastfeeding patients.

The value of breastfeeding in MS is controversial [Langer-Gould and Hellwig, 2013; Vukusic and Confavreux, 2013]. The literature on breastfeeding and MS has never suggested a deleterious effect. Studies have shown breastfeeding either had no effect, or more commonly was associated with suppression of disease activity [Hutchinson, 2013; Langer-Gould and Beaber, 2013; Langer-Gould and Hellwig, 2013; Langer-Gould et al. 2009, 2011; Pakpoor et al. 2012; Portaccio et al. 2011; Vukusic and Confavreux, 2013]. These conflicting results have been attributed to mixing patients who exclusively breastfed, with those who did not. Exclusive breastfeeding is defined as no regular formula feedings (less than one bottle a day). This results in a prolonged lactational amenorrhea with ovarian suppression, high prolactin levels, and low nonpulsatile luteinizing hormone levels. MS patients should be told that if they breastfeed it should be exclusive, because this is more likely associated with decreased MS disease activity. It is not clear if disease suppression equals that of an effective DMT, however. Some studies suggest that breastfeeding, particularly when prolonged (⩾4 months), reduces risk for MS in the child [Conradi et al. 2013; Ragnedda et al. 2015].

DMT use

For women at higher risk of postpartum relapse (very active disease, poor prognostic profile, relapse during pregnancy, no prior DMT use), it may be advisable to plan on starting a DMT. Treatment can be instituted very quickly after delivery. In a recent study, natalizumab started within 8 days of delivery was able to prevent postpartum relapses in five of six highly active patients [Vukusic et al. 2015].

Impact on infant/child development

Maternal MS is not recognized to have a negative impact on long-term development of the child. There have been limited analyses of development of children exposed to DMTs in utero; to date, no negative impact has been found [Coyle et al. 2014; Fragoso et al. 2013].

Neuromyelitis optica spectrum disorder

For many years neuromyelitis optica spectrum disorder (NMOSD) was considered a variant of MS. It is now recognized as a distinct neuro-immune channelopathy, with most patients showing diagnostic anti-aquaporin 4 IgG. Aquaporin 4 is expressed maximally by the human placenta during midgestation; and in animal models, antibodies are capable of causing placental inflammation and fetal death [Saadoun et al. 2013]. NMOSD shows a distinct pregnancy impact from MS. The attack rate is not significantly reduced during pregnancy, but then is also elevated during the 3 months postpartum [Shimizu et al. 2015]. NMOSD is also associated with higher risk for miscarriage [Nour et al. 2016].

Unmet needs

Pregnancy will continue to be a major focus for MS patients, and the treatment era makes decision making more complex. It would be very helpful to have accurate biomarkers for disease activity and prognosis, to guide appropriate recommendations. This could help decide whether a young woman with MS should commit to therapy, or attempt to become pregnant. Accumulating accurate data on fetal risks and breast milk exposure should allow the development of better recommendations on when it is possible to: eliminate DMT washouts, treat during pregnancy, and treat while breastfeeding. Finally, identifying why MS disease activity decreases late in pregnancy, then temporarily increases for 3 months postpartum could focus development of novel therapeutic strategies.