Relapse occurrence in women with multiple sclerosis during pregnancy in the new treatment era

Discussion

Women with MS usually plan their pregnancies when they have been stable without any disease activity and tend to consider pregnancy early in their disease course while they have no or minimal neurologic disabilities. In addition, pregnancy is an important stratification factor in the decision to initiate or switch DMTs, especially for therapies requiring strict washout periods such as fingolimod.

At baseline, heterogeneity of the studied women receiving different DMTs was observed, especially with respect to the clinical characteristics (disease duration, relapse, and EDSS score), which might have affected the findings. There was a 4-fold increase in the relapse occurrence in our cohort, driven mostly by patients who were on natalizumab and fingolimod before conception. We expected to find quiescent disease in interferon-treated patients and possible reactivation of disease in natalizumab-treated patients. This was a potential source of bias because most natalizumab-treated patients were escalated after having breakthrough disease while on first-line therapies such as interferons or had an initial highly active course necessitating initiation of natalizumab as a first-line therapy. On the other hand, the effect of the washout period in the fingolimod group and to a lesser extent in the natalizumab group was unexpected given the faster disease reactivation in the first trimester.

Most of the previous reports did not assess the effect of treatment or evaluated the effects of only IFN-β/GA because of the unavailability of high-efficacy DMTs or because of the lack of data on pregnancies in patients using second-line therapies. Most of the available reports were driven by retrospective, observational, and open-label studies or were extracted from postmarketing pharmacovigilance data collected by registries or pharmaceutical companies.^6,–,9,17 A retrospective study evaluated 152 pregnancies, of which 61 pregnancies were exposed to immunomodulators (41 to GA, 17 to IFN-β, 2 to IVIG, 2 to IVMP) with a mean period of 18.4 ± 13.2 weeks.⁸ The largest study was performed with an international MS registry (MSBase) that included 893 pregnancies (674 women). DMTs were used by 39.0% of women in the 2-year period before conception. Discontinuation of DMT occurred at a median of 7.1 months before the onset of pregnancy. The 3 period epochs examined were 1967 to 2002, 2002 to 2006, and 2006 to 2010, which mostly did not include any second-line or new DMTs. The authors acknowledged the potential bias in that women who have milder disease were more likely to stop a DMT to become pregnant.¹⁷ In a prospective cohort study from the German Multiple Sclerosis and Pregnancy Registry, 445 pregnancies in women with MS were exposed to IFN-β (n = 251), and 194 patients had no exposure to DMTs.⁹

The interval between diminished/no treatment effect and the beginning of pregnancy is difficult to assess without proper pharmacokinetics studies. The serum concentration of a drug does not reflect its sustained effects. However, there are no established blood levels to determine the cessation of treatment effects. Some of the practical decisions with respect to determination of cessation of treatment effect were based on clinical or laboratory observations. For example, the normalization of lymphopenia in fingolimod-treated patients may indicate the diminution of effect, whereas a rebound/reactivation phenomenon might be seen in natalizumab-treated patients after 12 weeks of treatment cessation, suggestive of a possible lack of effect. Teriflunomide is the only drug for which the serum level can be measured after the accelerated clearance procedure is performed; however, it would be difficult to definitely determine that the treatment has no clinical effect.

Disease reactivation on discontinuation of DMTs played a significant role in our study. This was evident when the relapse occurrence decreased in women who continued natalizumab after implementation of the practice recommendation. The effect of washout duration was more pronounced in patients who were on fingolimod given the unpredictability of pregnancy after fingolimod discontinuation. Concerns regarding a risk of reactivation after DMT discontinuation have increased with the advent of recent drugs.¹⁸ Disease reactivation after natalizumab discontinuation is an established phenomenon, especially when natalizumab is stopped in patients with highly active MS because of concerns about progressive multifocal leukoencephalopathy.^10,19 Even in what it was presumed to be a protective period, clinical and radiologic disease activity reappeared after natalizumab withdrawal for pregnancy planning with rare consequences of accumulation of disabilities after a relapse during pregnancy.^11,20 Similarly, disease reactivation during pregnancy was observed several weeks after cessation of fingolimod.¹²

We have confirmed that, as shown previously, the postpartum period was associated with a high risk of relapses. A higher relapse rate in the prepregnancy year is believed to be a predictor of a higher relapse rate in the postpartum period.^1,3,8,17 Pregnant women who were exposed to immunomodulators were less likely to have postpartum relapses than women who were not exposed to any DMT during the pregnancy or in the 3 months before conception.⁸ Similarly, the use of DMTs in the 2 years preceding conception was independently associated with a decreased postpartum relapse rate.¹⁷ It is unclear how DMTs may attenuate the relapse rate several months after discontinuation, yet this observation could be an argument for maintaining DMTs until conception whenever it is safe.²⁰ Most treating physicians tend to advise the resumption of DMTs after delivery, although the optimal time to restart DMTs is unclear.²¹

Several reports addressed possible measures to prevent relapse occurrence in the postpartum period. The administration of different doses of IVIG was evaluated by 2 prospective studies, and there were no statistical differences between the dose regimens in reducing postpartum relapses.^22,23 A recent retrospective study that evaluated 134 pregnancies (47 in women who received IVIG after delivery) in women with MS did not find any beneficial effect of the use of IVIG on relapse rates.²⁴ In a nonrandomized study of monthly IVMP in 20 patients, the authors suggested a modest beneficial effect in reducing the risk of postpartum relapses. However, the results were not replicated.²⁵ Breastfeeding has also been evaluated for possible protective effects in the postpartum period, but the results were conflicting in that 1 study that included 32 patients showed beneficial effects while another with a larger cohort of 298 women showed the opposite.^26,27

In view of our findings, there are few practical points that need to be discussed. Women who are in remission or actively using DMT for at least 2 years have a significantly decreased risk for postpartum attacks but not necessarily during pregnancy.¹⁷ Women contemplating pregnancy should be counseled on several issues. First, a woman might have discontinued treatment for a long period of time before conception, resulting in a prolonged washout period. Hence, it might be possible for the woman to continue her DMTs such as IFN-β and GA until conception on the basis of systematic evidence of postmarketing studies.^18,28 Second, an extended washout period before conception in patients with highly active disease might be associated with an increased risk of relapse especially beyond several weeks/months. Thus, continuing natalizumab until conception and possibly during the first trimester could decrease this risk because the teratogenic effects are low except in the third trimester when there is a risk of thrombocytopenia in the fetus.^29,30 It was shown that treatment with natalizumab in women with highly active MS during their third trimester of pregnancy was associated with the suppression of disease activity.³¹ Fingolimod is associated with teratogenic effects, and it is recommended to withdraw the drug 2 months before pregnancy planning. Similar issues could be raised with alemtuzumab and teriflunomide because the former requires at least a 4-month washout period while the latter needs to be washed out with an elimination procedure. A definitive washout period for dimethyl fumarate is yet to be established but is ≈2 weeks because of its short half-life.³⁰ The possibility of switching to a safer drug such as IFN-β/GA until pregnancy begins can be discussed in a case-specific manner.¹⁸ Third, the appropriate measure to prevent postpartum relapse remains controversial. In a woman who had a relapse during pregnancy or had highly active disease before pregnancy, it is recommended to resume her maintenance therapy after delivery. The decision regarding the potential benefits of breastfeeding vs treatment reinitiation to reduce the risk of postpartum relapse should be individualized.³

Our study has several limitations. Apart from the retrospective nature of the study, the population of women was relatively young at the onset of MS. Second, the prepregnancy relapse rate was low because the majority of patients were on DMTs and half of women were on high-efficacy therapies. Furthermore, a subgroup analysis could not be performed given the low number of relapses in the interferon group. On the other hand, the study was conducted with a registry, eliminating the bias of it being conducted in a tertiary center. In addition, this is the first study to be conducted with a high proportion of pregnant women who were previously on natalizumab/fingolimod.

In this study, the risk of relapse in women who were on fingolimod and natalizumab was high during the first and third trimesters. Pregnancy in women contemplating conception should be carefully planned with an emphasis on shortening washout periods.