Assessment of cervical antibody concentrations fails to enhance the value of cervical length as a predictor of preterm delivery.

Abstract

The purpose of this study was to determine if cervical fluid antibody concentrations can enhance the value of cervical length in predicting risk of preterm delivery. We obtained cervical fluid samples with preweighed cellulose wicks from a prospective cohort of women 23 to 32 weeks' gestation with signs and symptoms of preterm labor and intact membranes. Total immunoglobulin A and G (IgA and IgG) concentrations were determined by enzyme-linked immunosorbent assay. Bacterial vaginosis was diagnosed by Gram stain, and cervical length was measured with endovaginal ultrasound. For subjects with term (n = 77) and preterm (n = 24) deliveries, median IgA and IgG concentrations were 736 vs 643 microg/mL (P = .33) and 1528 vs 1769 microg/mL (P = .85). For subjects with normal flora (n = 71), intermediate flora (n = 14), and bacterial vaginosis (n = 16), median IgA and IgG concentrations were 717, 624, and 774 microg/mL (P = .90) and 1383, 1553, and 2731 microg/mL (P = .02). In a forward stepwise logistic regression model, cervical length was the only factor associated with preterm delivery (P < .001). Measuring the concentrations of IgA and IgG in cervical fluid does not enhance the value of cervical length in predicting risk of preterm delivery.