Risk of preterm delivery and early pregnancy hydroxychloroquine use from a Californian lupus cohort.

Abstract

Objective: Pregnant patients with systemic lupus erythematosus (SLE) have 2-3 times higher risk of preterm delivery (PTD). Hydroxychloroquine (HCQ) is recommended during pregnancy and may reduce PTD risk. This study investigates whether early pregnancy HCQ-use reduces PTD risk in a diverse SLE cohort.

Methods: We included singleton pregnancies reaching ≥20 weeks' gestation (2011-2020) among patients with SLE aged 18-50 receiving care at Kaiser Permanente Northern California. HCQ exposure was defined as ≥2 prescriptions filled from 3 months before the last menstrual period through the first trimester. PTD was defined as delivery <37 weeks and continuously as gestational weeks for time-to-delivery analyses. Propensity scores (PS) based on demographics, comorbidities and medication use were calculated to address confounding. Risk ratios (RR) and HRs, including 95% CIs, were estimated using PS-adjusted Poisson regression with robust SEs and Cox regression, stratified by parity. To investigate effect modification, we stratified by prepregnancy comorbidities and pregnancy corticosteroid use.

Results: Among 399 pregnancies in 324 patients, 21% were preterm. The PS-adjusted RR was 1.08 (95% CI 0.52 to 2.23) and 0.88 (95% CI 0.50 to 1.57) for nulliparous and multiparous pregnancies exposed to HCQ, respectively. The PS-adjusted HRs were similar, and results remained consistent across analyses stratified by potential effect modifiers.

Conclusions: Although periconceptional HCQ-use was not associated with reduced PTD nor appreciably altered gestational age at delivery, we found no increased risks for these specific adverse outcomes. Consistent with other work, we found potentially protective associations in subsets stratified by parity. However, we had limited statistical power to test this.