Exposure to hydroxychloroquine in early pregnancy and incidence of pre-eclampsia and pre-term delivery in patients with systemic lupus erythematosus in Sweden: a nationwide population-based cohort study.

Abstract

Background: Pregnant women with systemic lupus erythematosus (SLE) have elevated risks of pre-eclampsia and pre-term delivery. Hydroxychloroquine, the mainstay treatment during pregnancy in women with SLE, is currently a promising agent for pre-eclampsia prevention. We aimed to examine associations between hydroxychloroquine use and pre-eclampsia and pre-term delivery in pregnant women with SLE.

Methods: In this nationwide population-based cohort study, we included all singleton pregnancies of women with prevalent SLE that lead to a delivery (livebirths and stillbirths) between Jan 1, 2007, and Dec 31, 2022, diagnosed in secondary or tertiary care centres in Sweden. Hydroxychloroquine exposure was defined as two or more dispensations from 3 months pre-pregnancy until the end of the first trimester. The primary outcomes were pre-eclampsia (diagnosed from 20⁺⁰ weeks of gestation to 6 weeks postpartum) and pre-term delivery (delivery before 37⁺⁰ weeks of gestation). Inverse probability of treatment weighting adjusted for measured confounders (eg, maternal smoking, BMI, reproductive characteristics, pre-gestational hypertension, glucocorticoid use) and modified Poisson models estimated risk ratios and 95% confidence intervals. We involved a person with lived experience of pregnancy with SLE in all aspects of the study.

Findings: Between Jan 1, 2007, and Dec 31, 2022, we included 959 singleton pregnancies from 685 women with prevalent SLE in Sweden. 404 (42%) of 959 pregnancies were nulliparous pregnancies (232 [57%] were unexposed and 172 [43%] were hydroxychloroquine-exposed) and 555 (58%) were parous pregnancies (333 [60%] were unexposed and 222 [40%] were hydroxychloroquine-exposed). The mean maternal age was 32 years (SD 4·7). Pre-eclampsia was recorded in 19 (11%) of 172 hydroxychloroquine-exposed pregnancies and 30 (13%) of 232 unexposed pregnancies in the nulliparous group and 12 (5%) of 222 hydroxychloroquine-exposed pregnancies and 20 (6%) of 333 unexposed pregnancies in the parous group. Pre-term delivery was recorded in 33 (19%) of 172 hydroxychloroquine-exposed pregnancies and 34 (15%) of 232 in unexposed pregnancies in the nulliparous group and 26 (12%) of 222 hydroxychloroquine-exposed pregnancies and 41 (12%) of 333 unexposed pregnancies in the parous group. The adjusted risk ratio for pre-eclampsia in SLE pregnancies with hydroxychloroquine exposure versus those without exposure was 0·49 (95% CI 0·31-0·79) overall, 0·59 (0·33-1·08) in the nulliparous group, and 0·44 (0·22-0·89) in the parous group. Associations between hydroxychloroquine and pre-term delivery were unclear in the overall (risk ratio 0·95 [95% CI 0·67-1·34]), nulliparous (1·10 [0·68-1·80]), and parous (0·75 [0·47-1·24]) groups. Stratification by antiphospholipid syndrome, renal diseases, and hypertension showed similar results.

Interpretation: In this large cohort of pregnant women with SLE, hydroxychloroquine exposure in early pregnancy was associated with a lower pre-eclampsia risk. However, the association with pre-term delivery was unclear.

Funding: US National Institutes of Health and the Ingegerd Johansson Donation.