Change in pregnancy-associated multiple sclerosis relapse rates over time: a meta-analysis

Background: Women with MS are advised that relapse rates fall during pregnancy and rebound post-partum. This advice originates from 1998; smaller, more recent, studies have not been previously pooled. Methods: All studies published since 1998 providing raw relapse data were considered for inclusion. Single arm meta-analysis was performed using a restricted maximum likelihood random effects model with inverse variance; secondary subgroup analysis and meta regression were then performed. Annualised relapse rates (ARR), or relapse numbers/rates suitable for conversion into ARR during pregnancy and the post-partum period were included. Secondary subgroup analysis examined year of data collection, DMT exposure, breastfeeding and data source. Results: 7034 pregnancies from 6430 women were included. ARR fell from 0.57 (95%CI 0.45-0.70) pre -pregnancy to 0.36 (0.28-0.44), 0.29 (0.21-0.36) and 0.16 (0.11-0.21) during trimesters 1,2, and 3, with a postpartum rebound (ARR 0.85, 95%CI 0.70-1.00). ARR reduced pre-pregnancy and post-partum over time (p<0.001). Relapse rates were lower in claims databases than elsewhere. Conclusions: Despite high heterogeneity, we confirm the historic assumption that ARR reduces during pregnancy, and demonstrate an overall reduction in ARR over time. Studies using data originating from claims databases demonstrated a lower relapse rate at all time points, which has not previously been demonstrated.