Gavoille Antoine, Rollot Fabien, Casey Romain, Mathey Guillaume, Le Page Emmanuelle, Ciron Jonathan, De Sèze Jérôme, Ruet Aurélie, Maillart Elisabeth, Labauge Pierre, Zephir Hélène, Kwiatkowski Arnaud, Papeix Caroline, Defer Gilles, Lebrun-Frenay Christine, Moreau Thibault, Laplaud David-Axel, Berger Eric, Dubessy Anne-Laure, Clavelou Pierre, Thouvenot Eric, Heinzlef Olivier, Pelletier Jean, Al-Khedr Abdullatif, Casez Olivier, Bourre Bertrand, Wahab Abir, Magy Laurent, Moulin Solène, Camdessanché Jean-Philippe, Doghri Inès, Sarov-Riviere Mariana, Hankiewicz Karolina, Pottier Corinne, Dos Santos Amélie, Nifle Chantal, Subtil Fabien, Vukusic Sandra
Hospices Civils de Lyon, Service de Neurologie, sclérose en plaques, pathologies de la myéline et neuro-inflammation, F-69677 Bron, France.
Université de Lyon, Université Lyon 1, CNRS, Laboratoire de Biométrie et Biologie Évolutive UMR 5558, F-69100 Villeurbanne, France.
JAMA Neurol. 2025 Aug 4. doi: 10.1001/jamaneurol.2025.2550.
In women with multiple sclerosis (MS), disease-modifying therapy (DMT) management during pregnancy might impact relapse risk.
To estimate the effect of DMT management during pregnancy on MS relapse rate and compare different therapeutic strategies.
DESIGN, SETTING, AND PARTICIPANTS: This was a multicenter retrospective cohort study using data from January 1990 to December 2023. Data were extracted in December 2023 from the French MS registry. Among 52 955 women in the registry, we included pregnancies identified through childbirths in patients with relapsing-onset MS who were monitored for at least 18 months before delivery and 9 months after. Pregnancies occurring less than 18 months apart or with missing month of birth were excluded.
Mediation analysis was used to estimate the total, direct, and indirect (mediated by DMT management) effects of pregnancy. Different therapeutic strategies were compared: DMT interruption, switching to or maintaining interferon β or glatiramer acetate, switching to or maintaining natalizumab until the third trimester, and switching to or maintaining intravenous anti-CD20 and interrupting it 3 months before conception.
The primary outcome was the annualized relapse rate (ARR) during the preconception, gestation, and postpartum periods. Within a causal inference framework, counterfactual ARRs were estimated using longitudinal g-computation, combining a random forest algorithm for predicting DMTs, and a mixed-effects Poisson model for relapses.
We included 6341 pregnancies occurring in 4998 women (mean [SD] age at conception, 31.5 [4.5] years). DMT management during pregnancy significantly increased ARR during gestation (causal rate ratio [cRR], 1.13; 95% CI, 1.06-1.22) and postpartum (cRR, 1.08; 95% CI, 1.01-1.16) periods. This led to a deleterious total effect of pregnancy on ARR, particularly in women receiving natalizumab before pregnancy with prolonged interruption (ie, interruption before the second trimester or resumption more than 3 months after delivery; cRR, 2.18; 95% CI, 1.76-2.69), and in women receiving fingolimod (cRR, 2.15; 95% CI, 1.60-2.93). Compared to DMT interruption, anti-CD20 strategy was the most effective (cRR, 0.38; 95% CI, 0.25-0.52), followed by the natalizumab strategy with short interruption (cRR, 0.80; 95% CI, 0.71-0.90), whereas interferon β (cRR, 0.93; 95% CI, 0.86-0.99) and glatiramer acetate strategies (cRR, 0.91; 95% CI, 0.84-0.99) were less effective.
In this study, DMT management during pregnancy significantly increased relapse risk, particularly in patients receiving natalizumab with prolonged interruption or fingolimod. The strategy based on the use of anti-CD20 before pregnancy was the most effective to mitigate this risk.
在患有多发性硬化症（MS）的女性中，孕期疾病修饰治疗（DMT）管理可能会影响复发风险。
评估孕期DMT管理对MS复发率的影响，并比较不同的治疗策略。
设计、设置和参与者：这是一项多中心回顾性队列研究，使用了1990年1月至2023年12月的数据。数据于2023年12月从法国MS登记处提取。在登记处的52955名女性中，我们纳入了复发型MS患者分娩时确定的妊娠，这些患者在分娩前至少监测18个月，分娩后监测9个月。间隔少于18个月或出生月份缺失的妊娠被排除。
采用中介分析来估计妊娠的总效应、直接效应和间接效应（由DMT管理介导）。比较了不同的治疗策略：DMT中断、改用或维持干扰素β或醋酸格拉替雷、改用或维持那他珠单抗直至孕晚期，以及改用或维持静脉注射抗CD20并在受孕前3个月中断。
主要结局是孕前、孕期和产后期间的年化复发率（ARR）。在因果推断框架内，使用纵向g计算估计反事实ARR，结合用于预测DMT的随机森林算法和用于复发的混合效应泊松模型。
我们纳入了4998名女性发生的6341次妊娠（受孕时的平均[标准差]年龄为31.5[4.5]岁）。孕期DMT管理显著增加了孕期（因果率比[cRR]，1.13；95%置信区间，1.06-1.22）和产后（cRR，1.08；95%置信区间，1.01-1.16）期间的ARR。这导致妊娠对ARR产生有害的总体效应，特别是在孕前接受那他珠单抗且中断时间延长的女性中（即孕中期前中断或分娩后3个月以上恢复；cRR，2.18；95%置信区间，1.76-2.69），以及接受芬戈莫德的女性中（cRR，2.15；95%置信区间，1.60-2.93）。与DMT中断相比，抗CD20策略最有效（cRR，0.38；95%置信区间，0.25-0.52），其次是中断时间短的那他珠单抗策略（cRR，0.80；95%置信区间，0.71-0.90），而干扰素β（cRR，0.93；95%置信区间，0.86-0.99）和醋酸格拉替雷策略（cRR，0.91；95%置信区间，0.84-0.99）效果较差。
在本研究中，孕期DMT管理显著增加了复发风险，特别是在接受那他珠单抗且中断时间延长或接受芬戈莫德的患者中。孕前使用抗CD20的策略是降低这种风险最有效的方法。
