0959: Assessing the Association Between Hydroxychloroquine and Preeclampsia Risk in SLE Pregnancies Using Administrative Claims Data

Amadeia Rector¹, Ivana Marić², Yashaar Chaichian³, Eliza Chakravarty⁴, Miranda Cantu⁵, Michael Weisman⁶, Gary Shaw⁷, Maurice Druzin⁷ and Julia Simard¹, ¹Stanford University School of Medicine, Palo Alto, CA, ²Stanford University, Stanford, ³Stanford University, Stanford, CA, ⁴Oklahoma Medical Research Foundation, Oklahoma City, OK, ⁵Global Lupus Support Group, Portage, MI, ⁶Adjunct Professor of Medicine, Stanford University; Distinguished Professor of Medicine Emeritus, David Geffen School of Medicine at UCLA, Los Angeles, CA, ⁷Stanford University School of Medicine, Palo Alto

Background/Purpose: Preeclampsia risk is higher among pregnancies in patients with systemic lupus erythematosus (SLE). Hydroxychloroquine (HCQ), a common first-line treatment, is encouraged in SLE pregnancies as lupus-associated flares may increase risk of adverse pregnancy outcomes. HCQ may interrupt preeclampsia pathogenesis by mitigating oxidative stress and targeting toll-like receptors to prevent pro-inflammatory cytokine production. We investigated HCQ use and preeclampsia risk in patients with SLE.

Methods: We studied 1096 singleton liveborn pregnancies among 1043 patients with SLE continuously enrolled in private health insurance (IBM® MarketScan® Commercial Database) from 3 months before last menstrual period to pregnancy end (2007-2016). Modified Poisson models estimated risk ratios (RRs) and 95% confidence intervals (95% CIs) by parity. Multiple approaches used propensity scores (PS) to account for confounders (e.g., age, obesity, pregestational hypertension, prior renal manifestations, antiphospholipid antibodies/antiphospholipid syndrome (aPL/APS), pre-pregnancy steroids, and azathioprine use). Stratified analyses examined effect modification by multiple factors including flare during pregnancy, aPL/APS, prior renal manifestations, pregestational hypertension, and prior preeclampsia. Multiple sensitivity analyses redefined exposure to account for possible nonadherence.

Results: 15% of pregnancies were diagnosed with preeclampsia and 9% delivered preterm. 52% of pregnancies filled HCQ at least once. The HCQ-exposed pregnancies had more flare during pregnancy, pre-pregnancy steroid use, prior renal manifestations, and azathioprine use. We did not observe an association between HCQ and preeclampsia among nulliparous (RR 1.32, 95% CI 0.87, 2.00) nor parous pregnancies (RR 1.25, 95% CI 0.84, 1.85). Additional control for confounding via multiple methods decreased the RRs towards the null in primary and sensitivity analyses (nulliparous: PS-adj RR 1.12, 95% CI 0.71, 1.79 and parous: PS-adj RR 1.09, 95% CI 0.72, 1.64). Among pregnancies with prior preeclampsia, the RR was further attenuated (PS-adj RR 0.85, 95% CI 0.26, 2.80).

Conclusion: Growing evidence suggests that HCQ may protect against preeclampsia in SLE pregnancy as well as in patients with APS. Using a large insurance-based database we found a null association but noted evidence of higher disease activity among HCQ-exposed pregnancies. Data on parity and prior preeclampsia in claims data are often incomplete and prone to misclassification. Further studies leveraging large population-based data, prospective collection, and trials may better characterize how HCQ may influence preeclampsia risk in SLE. Acknowledgments: Data access for this project was provided by the Stanford Center for Population Health Sciences (PHS) Data Core. Disclaimer: IBM Watson Health and MarketScan are trademarks of IBM Corporation in the United States, other countries or both.
Figure 1. Flowchart of SLE pregnancy cohort, 2007-2016. SLE = systemic lupus erythematosus, LB = livebirth, SB = stillbirth, MB = mixed birth

Table 1. Demographics and clinical characteristics for liveborn pregnancies in patients with SLE by HCQ exposure (IBM MarketScan Research Databases, 2007-2016). HCQ exposure is defined as one or more fills from 3 months before LMP to the end of the first trimester.

Table 2. Relative risks of preeclampsia among nulliparous and parous singleton SLE pregnancies, presented as RR (95% CI).
Disclosures: A. Rector, None; I. Marić, None; Y. Chaichian, Amgen, Eli Lilly, Boehringer-Ingelheim; E. Chakravarty, None; M. Cantu, None; M. Weisman, None; G. Shaw, None; M. Druzin, None; J. Simard, None.