Maternal cardiovascular events in autoimmune rheumatic diseases and antiphospholipid syndrome pregnancies
Rashmi Dhital,Rebecca J Baer,Gretchen Bandoli,Monica Guma,Dilli R Poudel,Kenneth Kalunian,Michael H Weisman,Christina Chambers
DOI: https://doi.org/10.1016/j.ajogmf.2024.101319
Abstract:OBJECTIVE:: Antiphospholipid syndrome (APS) and autoimmune rheumatic diseases (ARDs) are known to increase the risk for cardiovascular events (CVEs) in the general population., However, research in pregnancy is limited. We compared the occurrence of acute CVEs in pregnant women with and without ARDs or primary APS using a Californian population-based cohort. STUDY DESIGN:: This was a retrospective cohort study of pregnant individuals who delivered singleton infants in California between 2005 and 2020. Birth certificates were linked by the Study of Outcomes of Mothers and Infants to maternal records. The study was approved by institutional review boards of the State of California and the University of California San Diego. Pregnancies with ARDs were identified by ≥1 International Classification of Diseases (ICD) code for rheumatoid arthritis (RA), systemic lupus erythematosus (SLE), spondyloarthritis, Sjogren’s syndrome, and other ARDs (systemic sclerosis, inflammatory myositis, and vasculitides) in maternal hospital discharge, emergency, or ambulatory surgery records. Lupus nephritis (LN) was defined as SLE plus glomerulonephritis, renal failure, nephritic or nephrotic syndrome, or proteinuria. Primary APS included women with APS without concurrent ARDs. Women without ARDs or APS comprised the reference group. Acute CVEs during pregnancy and up to 6 weeks postpartum were identified by ICD codes and compared between groups. CVEs were classified as myocardial infarction, cardiovascular accident, heart failure and peripartum cardiomyopathy, inflammatory heart diseases, cardiac dysrhythmias, and venous thromboembolism (VTE). Covariates identified and adjusted for include age, race and ethnicity, insurance, education, prepregnancy body mass index, preexisting hypertension, diabetes, hyperlipidemia, depression, and substance use disorders. A log linear regression with a Poisson distribution was used to estimate the adjusted relative risks (aRRs) and 95% confidence intervals (CIs) for the associations of ARDs and APS with CVEs. Analyses were performed using Statistical Analysis Software (SAS), version 9.4 (SAS Institute, Cary, NC). Causal mediation analyses were performed for potential mediators, such as gestational diabetes, gestational hypertension, and preeclampsia. RESULTS:: Overall, 19,340 women had ARDs, 7758 had primary APS, and 7,004,334 had neither condition. The ARD and APS groups had more traditional cardiovascular risk factors. Acute CVEs were observed in 2.0% of ARD cases (388/19,340), 7.0% of primary APS cases (540/7,758), and 0.4% of reference group cases (24,402/7,004,334). The aRR for CVEs in the ARD and APS groups in comparison with the reference group was 4.1-fold (95% CI, 3.7–4.5) and 14.7-fold (95% CI, 13.5–16.0), respectively (Table). Of the 388 CVEs that occurred in women with ARDs, 164 (42%) were VTEs, 96 (25%) were heart failure or peripartum cardiomyopathy, and 92 (24%) were cardiac dysrhythmias. In primary APS cases, 453 of 540 (83%) CVEs were VTE. Acute CVEs occurred in 3.1% (159/8,422) of patients with overall SLE, in 10.7% (55/513) of patients with SLE with APS, and in 8.5% (77/903) of patients with SLE with LN (Table). The aRR for CVEs was 6-fold higher among those with SLE (95% CI, 5.3–6.8) and was notably increased with concurrent APS or LN, with aRRs of 18.1 (95% CI, 13.9–23.6) and 12.7 (95% CI, 10.1–15.9), respectively (Figure). The risks for both VTE and non-VTE CVEs was increased among pregnancies affected by ARD and APS (Figure). Five of 6 in-hospital deaths in ARD pregnancies (5/388 or 1.3%) occurred among women with acute CVEs (Table). The risks for CVEs were higher during all perinatal periods. In mediation analyses, 11.2% of the excess risk for CVE in ARD pregnancies was mediated by preeclampsia, whereas <0.5% was mediated by gestational diabetes or hypertension. CONCLUSION:: Pregnancies affected by ARDs and APS had both more traditional cardiovascular risk factors and about a 4-fold and 15-fold higher CVE risk, respectively, than the reference group. Pregnancies affected by SLE with APS (18.1-fold) and LN (12.7-fold) had substantial increased risks for CVEs. VTEs were the most frequently observed CVEs. The in-hospital mortality rate for any acute CVE in ARD pregnancies was 1.3% (Table).