RESUMEN
BACKGROUND: Maternal markers of intestinal immune activation may be used to predict preterm birth (PTB) in pregnant women living with HIV. METHODS: This study used de-identified samples from the International Maternal Pediatric Adolescent AIDS Clinical Trials Group (IMPAACT) Protocol P1025 study. Singleton pregnancies with ≥3 ml plasma available and HIV viral load ≤400 copies/ml within 4 weeks of specimen collection were included. Frequency matching of PTB cases and term birth controls was performed on basis of maternal race, number of available plasma specimens, and timing of plasma sample collection in a 1:1 ratio. Plasma progesterone, 25-hydroxy vitamin D, soluble CD14, intestinal fatty acid binding protein (I-FABP), Lipopolysaccharide (LPS)-binding protein, and inflammatory cytokines (IL-1B, IFN-gamma, IL-6, TNF-alpha) were measured. Generalized mixed linear regression modeling was used to examine the association between PTB and biomarkers, adjusting for covariates and confounders. Data analyses were performed using SAS 9.4 (Cary, NC). RESULTS: We included 104 PTB compared to 104 controls. Third trimester log2 IL-1B was lower among PTB versus term birth controls by univariate analysis (-1.50 ± 2.26 vs. -.24 ± 2.69, p = .01) though this association was no longer significant by regression modeling. In an uncontrolled, exploratory sub-analysis, subjects with prior PTB had increased odds of PTB with higher I-FABP [aOR 2.72, 95% CI 1.18-6.24] and lower IFN-gamma [aOR .23, 95% CI .12-.41] after adjustment for covariates and confounders. CONCLUSIONS: Intestinal immune activation measured by soluble CD14 or intestinal fatty acid binding protein was not associated with preterm birth among pregnant women with low-level HIV viremia.