Use of B-Cell–Depleting Therapy in Women of Childbearing Potential With Multiple Sclerosis and Neuromyelitis Optica Spectrum Disorder

Overview of MS and NMOSD and Pregnancy

Studies to date have been largely reassuring that MS is not associated with difficulty with fertility/fecundity, with gestational complications, or with adverse pregnancy outcomes including early pregnancy loss, stillbirth, fetal malformations, low birthweight, and low gestational age.¹ Notably, pregnancy is felt to play a modulatory role on neuroinflammatory function. Pregnancy itself is considered an immunotolerant state, promoting tolerance to placental proteins, with a shift from Th1 to Th2 response. Perhaps as a result, relapse rates in MS have been reported across numerous studies to decrease during pregnancy and then rise postpartum. Up to about a third of women have been reported to experience an MS relapse in the first 3 months postpartum. Increased postpartum activity is evidenced by both increased rate of clinical relapse as well as new radiographic activity including new T2 hyperintense lesions or gadolinium-enhancing lesions.²

The effects of NMOSD on fertility are not clearly understood, possibly in part because of the more recent discovery of the anti–aquaporin-4 antibody, or because of its less common presentation in western populations. In a recent study, 8% of patients with NMOSD reported infertility of unknown etiology, and 6% reported infertility due to known causes including structural problems and hormonal dysregulation.³ As is the case with MS, women with NMOSD face an increased risk of postpartum relapse⁴; however, data on relapse rate during pregnancy are mixed. Prior studies have demonstrated stable or increased rate of relapse during pregnancy,^4,5 whereas a 2021 study of 46 women with aquaporin-4–seropositive NMOSD, 30 with MOG-associated disease, and 13 with double-negative NMOSD⁶ demonstrated a lower relapse rate during pregnancy compared with the prepregnancy period. In this study, only the patients with aquaporin-4–seropositive NMOSD demonstrated increased postpartum relapse compared with the prepregnancy period, but numbers were small. Unlike MS, women with NMOSD have an increased rate of pregnancy complications.⁷ One study demonstrated that NMOSD increases the risk of miscarriage, with 43% of pregnancies occurring after the onset of NMOSD ending in miscarriage compared with 7% before NMOSD onset. The risk of miscarriage was found to be independent of comorbid autoimmune conditions including antiphospholipid syndrome.⁷ NMOSD has also been associated with preeclampsia, at rates of 11.5% after disease onset compared with 3.1% in obstetric controls. Comorbid autoimmune diseases were noted to be a risk factor for preeclampsia in NMOSD.⁷

Use of B-Cell Depletion in MS and NMOSD

CD20 is a cell surface molecule expressed on B-cell subsets including pre-B, immature, mature, and memory B cells. The therapeutic efficacy of anti-CD20 therapies stems from significant depletion of circulating CD20⁺ B cells in the periphery.

Traditionally, the postulated pathophysiologic mechanism underlying MS was felt to be direct cytotoxicity mediated by T lymphocytes. Increasingly, however, the role of B cells has also been recognized. B cells function as antigen-presenting cells, thereby recruiting inflammatory cells and stimulating myelin-reactive T cells. They also produce antibodies, including IgG oligoclonal bands detectable in the spinal fluid of a majority of patients with MS. The high efficacy of B-cell–depleting therapy in MS further demonstrates that B cells are significant contributors to its pathophysiology. In comparison, monoclonal antibody treatment approaches targeting T cells uniquely have not demonstrated efficacy, although other T-cell treatment approaches have been efficacious. To date, for the treatment of MS, several monoclonal antibodies targeting CD20⁺ B cells have been approved by the U.S. Food and Drug Administration (FDA) and European Medicines Agency for treatment of MS including ocrelizumab and ofatumumab; rituximab has been used off-label for MS therapy for many years as well.

Aquaporin-4–seropositive NMOSD is mediated primarily by the humoral immune system leading to complement activation and resulting in CNS demyelination and axonal loss within the gray and white matter. For the treatment of NMOSD, inebilizumab, a B-cell–depleting monoclonal antibody targeting CD19, has been FDA approved; rituximab is also often used off-label for NMOSD.

Vaccine Considerations in Women of Childbearing Age

Although, ideally, neurologic and obstetrical clinicians regularly communicate once a woman with MS or NMOSD has decided to conceive or become pregnant, many times the decision to start B-cell–depleting therapy precedes—sometimes by years—the decision to conceive. Furthermore, women may also accidently become pregnant, even when counseled to avoid pregnancy; rates of unplanned pregnancy, representing almost half of pregnancies in the general US population, are unknown in MS and NMOSD. Therefore, before initiating anti-CD20 therapy, neurologists should broach the subject of vaccination, including vaccines relevant to pregnancy management, in women of childbearing potential with MS and NMOSD because they may not mount as effective a response once therapy is initiated.

There are general guidelines surrounding the vaccination of women of childbearing potential against preventable diseases before conception. Vaccination during pregnancy is considered appropriate when the vaccine is safe and when infection confers an increased risk to the mother and/or fetus. Immunization is reported to be equally effective in pregnant and nonpregnant women.¹⁹

Although patients should be up to date on appropriate vaccinations before starting B-cell–depleting therapy, some vaccinations may be needed once on treatment, including annual influenza vaccination and COVID-19 vaccination and booster series. In patients on B-cell–depleting infusion therapy, there is evolving guidance on the timing of immunization relative to treatment in an effort to optimize appropriate response to vaccination.¹⁶ There are no clear data on the optimal timing of vaccination in the setting of monthly ofatumumab injections; however, in our practice, we recommend obtaining immunization at least 4 weeks after the last injection and resuming ofatumumab 2–4 weeks thereafter. The timing of immunizations with B-cell–depleting therapies can be challenging, and as such, assessment of the patient's individualized risk is important when guiding patients with ongoing vaccination requirements.

Current vaccination guidelines from both the American College of Obstetricians and Gynecologists (ACOG) and the Infectious Diseases Society of America (IDSA) are summarized in Table 1. This table includes a list of vaccinations, which are considered safe during pregnancy for a woman of childbearing age and which ones should be considered before conception and/or before B-cell–depleting therapy, and the optimal timing of vaccinations in patients currently on B-cell–depleting therapy.

Because the live attenuated measles-mumps-rubella and varicella vaccinations are to be avoided both after B-cell depletion and during pregnancy, they should be considered in all women of childbearing potential at least 4 weeks before the initiation of B-cell–depleting therapy. The Gardasil vaccine, protecting against certain strands of human papillomavirus, is an important consideration in women of childbearing age given the risk of cervical dysplasia in women on highly effective disease-modifying therapies.²⁰ Guidelines about age eligibility have evolved, and Gardasil is now approved up to age 45 years. Finally, though not routinely recommended in women of childbearing age, according to the IDSA guidelines, the zoster and pneumococcal vaccines should also be administered for all immunocompromised patients regardless of age, including those on or planning to initiate B-cell–depleting therapies.

Safety of B-Cell–Depleting Therapy During Pregnancy

To evaluate the safety of B-cell depletion exposures during pregnancy, consideration of physiology of placental transfer, possible impacts of B-cell depletion, and data from clinical series can be helpful. Overall, there are several mechanisms for the transport of both small and large molecules across the placenta, including passive transport, simple diffusion, active transport through ATPase channels, facilitated diffusion, and receptor-mediated endocytosis. Although many medications are small molecules that passively diffuse across the placenta, monoclonal antibodies (including anti-CD20 and CD19 therapies) are larger and do not efficiently passively diffuse across the placenta.²¹ Although few studies focus on placental transfer of large biomolecules including monoclonal antibodies, research on placental transfer of human immunoglobulin (Ig) is a useful model. The fetus itself synthesizes minimal immunoglobulin, and thus, the majority of fetal Ig before birth is obtained from the mother. Ig molecules are large (∼150 kDa), and placental transport occurs via receptor-mediated active transport, predominantly via Fc receptors.²² Other subclasses of immunoglobulins including IgM, IgE, and antibody fragments do not contain the Fc domain and are thus minimally transported across the placenta. One exception to this is IgA, which does have some minimal transfer from mother to fetus, but not via Fc-mediated active transport.²² Maternal transfer of IgG is minimal during the first trimester, increases significantly around gestational weeks 13–18, and peaks around gestational weeks 22–26.²¹ IgG1 transfers through the placenta most efficiently, followed by IgG3 and IgG4; IgG2 transfers only minimally through the placenta.²² Maternal transfer of immunoglobulin subclasses is summarized in Table 2.

The physiology of human IgGs suggests that the highest exposure to IgG1-based monoclonal antibodies, including B-cell–depleting therapies, is during the second and third trimesters, particularly after 32 weeks gestation, but is minimal during the first trimester.²³ The molecular weights of rituximab and ocrelizumab are ∼143 kDa, of ofatumumab ∼146 kDa, and of inebilizumab ∼149 kDa.^24,-,26 These molecular weights are similar to human IgG, and the large molecular size prevents significant placental transfer during the first trimester of organogenesis.²⁷

To weigh the likelihood of exposure, it is important to consider the kinetics of elimination of a compound. After an initial dose, assuming first-order pharmacokinetics, 94%–97% of a drug is eliminated after 4–5 half-lives. The elimination half-life of rituximab is about 18–22 days, and thus, by about 110 days, the drug has been effectively eliminated in a majority of cases, although the maximal half-life of up to 77 days has been reported.²⁴ Similarly, the elimination half-lives of ocrelizumab, inebilizumab, and ofatumumab are 26, 18, and 15 days, respectively.^25,26 In a study of patients with rheumatoid arthritis on rituximab, reconstitution of peripheral B cells was variable and did not necessarily correlate with disease relapse.²⁸

Given that the elimination of B-cell–depleting therapy occurs within 3–6 months after exposure and that these drugs transfer minimally across the placenta during the first trimester, it could be hypothesized that minimal IgG1 might be transferred across the placenta if conception coincides with the timing of B-cell–depleting infusion or self-injection. In this same vein, if conception occurs 3 months after the most recent exposure of B-cell–depleting therapy, even less IgG1 might be transferred across the placenta. The fetal exposure based on the terminal half-life and IgG1 placental transfer characteristics is summarized in Figure 1. Importantly, these considerations are based on 4-5 mean, and not maximal, half-lives. To date, studies of reproductive risk have considered any treatment within 6 months of conception as exposed pregnancies.

• Download figure
• Open in new tab
• Download powerpoint

Figure Fetal Exposure to B-Cell–Depleting Therapy Based on Terminal Half-Lives as Well as IgG1 Placental Transfer Characteristics

(A) Schema illustrating elimination of B-cell–depleting therapy. Average t₁/₂ of anti-CD20: 18–26 days. Of note, this schema is based on usual half-lives of 18 (rituximab) to 26 (ocrelizumab) days; maximal half-lives of up to 77 days have been reported. (B) Schema illustrating fetal exposure to B-cell–depleting therapy.

In a systematic review of 102 pregnancies in women who became pregnant within 6 months of exposure to rituximab used for a range of neurologic and nonneurologic maternal indications, 78 resulted in live births and 12 in spontaneous abortions, similar to that seen in the general population.²⁷ That article also reported on a case series of a further 10 women with MS or NMOSD treated with rituximab. Altogether, in the 63 live births with reported gestational age, 40 (63%) occurred at term (37 + weeks), and 2 (3%) occurred before 32 weeks. The primary adverse effect was low neonatal B-cell count, occurring in 39% of newborns, with the lymphopenia normalizing within 6 months in all cases. Notably, the specific timing between the patient's last infusion and conception was variable. Among the studies included in this review was 1 large study that included 153 pregnancies that occurred at a variable timeframe after maternal exposure to rituximab, of which 90 (59.0%) resulted in live births.²⁹ In this study, first trimester pregnancy loss occurred in 21% of these pregnancies, slightly higher than the general population (10–15%), though similar to rates in women with chronic disease. Of the live births, 19% were premature, similar to the rate of prematurity in women with certain chronic medical conditions. Two congenital malformations were reported, consistent with the rate in the general population.²⁹ Adverse events in neonates included hematologic abnormalities (lymphopenia) and neonatal infection. Further reviews have also described reassuring outcomes in women treated at more variable intervals before conception for both MS and other conditions.^30,31 These data provide some guidance regarding the possible implications for women with refractory disease who may need to be treated during pregnancy.

In as-yet unpublished data from ocrelizumab postmarketing surveillance,³² a total of 608 pregnancies were reported in women with MS exposed to ocrelizumab; 104 of these both were defined as likely fetal exposure in utero (defined in this case as conception occurring within 3 months of the last ocrelizumab infusion) and had known pregnancy outcomes. These outcomes included 62 (59.6%) live births, of which 90.3% were healthy, and 3 (4.8%) were preterm with an abnormal findings. Elective termination/therapeutic abortion occurred in 24 (23.1%), spontaneous abortion occurred in 15 (14.4%), 1 pregnancy (1%) was ectopic, and 2 were still births (1.9%). The published epidemiologic range for these outcomes is as follows: therapeutic abortion (2.9%–36.1%), spontaneous abortion (9.1%–17.2%), ectopic pregnancy (0%–1.5%), and still birth (0%–1.3%).^33,34 Of all 156 live births reported among the 608 women (regardless of exposure), 6 congenital malformations and 1 adverse pregnancy outcome were reported. There were very limited data on infant outcomes according to maternal treatment or potential breastmilk exposure.

Although the underlying physiology of IgGs and clinical experiences to date are reassuring, evidence remains limited, and there is a notable dearth of prospectively designed pharmacy-sponsored studies in this arena to guide care. The product labels remain cautious. FDA labels of ocrelizumab and ofatumumab state that these therapies may cause fetal harm based on animal studies and thus recommend that contraception be used during treatment and for 6 months after medication discontinuation.^25,26 Though not approved for MS or NMOSD, rituximab is commonly used off-label to treat these conditions. The FDA label for rituximab states that based on the limited human data available, rituximab can cause fetal harm from B-cell lymphocytopenia in infants exposed to the drug in utero. It advises females of reproductive potential to use effective contraception for at least 12 months after the last dose.²⁴ The European Medical Agency offers similar guidance, with recommendations to use contraception during and through 12 months after exposure to these monoclonal antibody therapies³⁵; this would de facto place some patients (e.g., with NMOSD or active MS) to elevated risk of inflammatory activity with B-cell repopulation.

Clinical Efficacy of B-Cell–Depleting Treatment Before and During Pregnancy

Several case series of women with MS and NMOSD have suggested that B-cell–depleting therapy before pregnancy is associated with clinical stability both before conception and during pregnancy. A systematic literature review including case series documented that in 10 pregnant women (7 with MS and 3 with NMOSD) treated with rituximab within 6 months of conception, no maternal relapses occurred before or during pregnancy.²⁷ Additional studies have corroborated these findings.³⁶ In fact, in the general MS population, discontinuation of rituximab therapy for reasons including pregnancy is associated with a very low risk of rebound activity, suggesting that this can be a stabilizing approach when planning a pregnancy.³⁶

Because B-cell reconstitution is variable and may occur more than 6 months after the last infusion, it may be reasonable to trend lymphocyte subsets in women attempting conception and delay infusion if CD19 levels remain undetectable. These could be monitored prospectively (e.g., every 6 weeks). As inflammatory attacks in NMOSD tend to be more severe with less recovery than in MS, we recommend close attention to maintaining undetectable CD19 levels in NMOSD to prevent disease relapse. There may be less urgency in maintaining undetectable CD19 levels in MS during a period of attempting conception, particularly in those patients with mild disease. Areas of notable uncertainty include lack of knowledge about the long-term effect of a lack of maternal B cells, as well as lack of maternal immunoglobulins—particularly IgA—in long-term B-cell–depleted mothers, on the developing fetus' immune system.

Postpartum Considerations

Summary: Practical Guidance for Use of B-Cell–Depleting Monoclonal Antibodies in Women of Childbearing Potential

The use of B-cell–depleting monoclonal antibodies as highly effective therapy for MS, NMOSD, and other neuroinflammatory conditions is becoming increasingly common. To date, clinical experience has suggested that these can be effective agents to prevent relapses both while awaiting conception (including when discontinuing therapies with risk of discontinuation rebound) and during the period of heightened postpartum risk. Counseling around risk is always individualized and can be informed based on several considerations.

For all women of childbearing potential, regardless of their current family planning goals, it is important to counsel on both age-appropriate vaccinations and those usually recommended by obstetricians peripregnancy—before starting B-cell–depleting therapy because its potential for attenuating the immune response to vaccination could have implications for both the mother and her future offspring.

In women planning conception, optimal timing of B-cell–depleting infusions is informed by both maternal risk and preference. The biologic effects of B-cell depletion are long lasting and persist beyond the drug's pharmacokinetic elimination, allowing for a period of attempting conception during which there is no detectable drug and theoretical risk to fetus. Ideally, there would be a period of 3 months between the last infusion and conception to ensure that by the second trimester, when placental transfer of IgG1 is expected, there has been near complete elimination of the product based on maximal half-lives reported.⁴⁷ However, some women with a higher risk of relapse or fertility concerns may prefer to attempt conception immediately following infusion given the low likelihood of transfer of the product in the first trimester when organogenesis occurs and when levels would be detectable. If there are delays in conception, CD19 levels should be monitored beginning at 6 months postinfusion to evaluate for peripheral B-cell reconstitution, and a pregnancy test should be obtained before each infusion in patients attempting conception.⁴⁷ Notably, in contrast to the B-cell–depleting infusion therapies timed every 6 months, the cumulative effect of monthly ofatumumab injections is unknown, raising concerns about whether women will stay protected during pregnancy with prenatal ofatumumab exposure.

In neonates born to mothers exposed to B-cell–depleting therapy either before conception or during pregnancy, we recommend evaluating lymphocytes and B-cell subsets in newborn cord blood. Some newborns have been reported to have a reduction in B cells during the first few weeks of life, with normalization of circulating B cells by 6 months. The cord blood can be sent clinically by the delivery team; counseling the patient to request this, and recommending to the delivery team that order sets and tubes be prepared in advance, can reduce delays and thereby the likelihood that this blood is clotted once it reaches the laboratory. In the event of transient neonatal B-cell depletion, vaccines could be delayed until peripheral B-cell reconstitution is observed.⁴⁸ Once vaccination efforts are initiated, pediatricians could consider evaluating vaccine titers during the first year of the infant's life to ensure that an optimal immune response to vaccination has been mounted.

In the postpartum period, women on B-cell–depleting therapies could be encouraged to breastfeed if they desire because breastfeeding has numerous established advantages for mothers and infants and can also reduce the risk of postpartum relapses in MS in particular. Because these IgG-based monoclonal antibodies have a large molecular size, B-cell–depleting therapies are transferred minimally into breastmilk and can safely be continued during this period.

A number of questions remain for the care of pregnant and postpartum patients with MS and NMOSD on B-cell–depleting therapies. Although the literature available suggests the safety and efficacy of these therapies in women of childbearing age, there are limitations to the data, particularly as the studies available use small sample sizes with a considerable number of biases. Long-term safety data of maternal B-cell depletion during pregnancy, long-term outcomes of infants born to women exposed to this class of monoclonal antibody therapy, and the impact of prepregnancy maternal B-cell depletion on postpartum immunoglobulin levels, including IgA, in maternal circulation and breastmilk are not fully known. Future studies elucidating these clinical queries are needed for the optimal care of women of childbearing potential with MS and NMOSD.