935. Repeat Pregnancies Among Women Living With HIV: Evaluating Temporal Changes in HIV Disease Status and Exploring the Association Between Preterm Birth and Protease Inhibitor Use in Pregnancy

Background With improved treatment, women living with HIV (WLHIV) are increasingly becoming pregnant. Studies have shown suboptimal viral suppression following pregnancy. In addition, protease inhibitors (PI) have been associated with preterm birth (PTB). Methods We studied WLHIV with at least 2 live births while on the PHACS SMARTT study. We first compared CD4 counts and viral loads (VL) between the first and second pregnancies using Wilcoxon rank-sum tests. We then examined trends in these measures over all reported pregnancies using mixed-effect linear regression models adjusting for maternal age and birth year, with a random effect to account for repeated measures in the same woman over time. Finally, we evaluated the association of PI or non-PI use during pregnancy with PTB, using GEE logistic regression models to adjust for pregnancy number, maternal age, and birth year. Results Between 2007 and 2018, 699 women had >1 pregnancy while on study, with a total of 1642 children. Their mean CD4 counts remained stable over repeat pregnancies. Their mean log10 VL decreased between the first and second pregnancies, both early and late in pregnancy (–0.42 cp/mL and –0.16 cp/mL respectively, P < 0.001 for each), but increased by 0.61 cp/mL (P < 0.001) between the end of the first pregnancy and early in the next pregnancy. Most women had VL suppression during pregnancy with VL rebound by the next pregnancy (figure). A majority of women (55%) received a PI in both their first and second pregnancies, with an increase in PTB rate of 4.3%, whereas those who changed from a PI to a non-PI had a decrease of 4.7% (table). Changing to a PI resulted in a stable rate, whereas remaining on a non-PI resulted in a drop of 2%. In adjusted models including all pregnancies, first trimester PI use was associated with an increased rate of PTB (adjusted OR 1.35; 95% CI 1.02, 1.97). Conclusion Most WLHIV achieved VL suppression during pregnancy, but many had a VL rebound after pregnancy. First trimester PI use was associated with higher risk of PTB.Table: Paired Group and PTB Paired Pregnancy Regimen Group Percent of PTB First/second pregnancy first pregnancy second pregnancy Difference Non-PI/Non-PI (n = 103) 11.7% 9.7% -2.0% Non-PI/On PI (n = 91) 15.4% 15.4% 0% On PI/Non-PI (n = 86) 16.3% 11.6% -4.7% On PI/Stayed on PI (n = 351) 14.8% 19.1% +4.3%Figure: Percent of women with unsuppressed VL (>400 cp/mL) Disclosures E. Chadwick, Abbott Labs: Shareholder, stock dividends. AbbVie: Shareholder, stock dividends. R. B. Van Dyke, Giliad Sciences: Grant Investigator, Research grant.