Abstract

The aim of the present study was to evaluate whether bacterial vaginosis (BV) diagnosed in early pregnancy and treated with vaginal clindamycin affects pregnancy outcome, and to investigate the predictive value of interleukins-6 (IL-6) and -8 (IL-8), and insulin-like growth factor-binding protein-1 (IGFBP-1) in cervical secretions, separately and combined by cervical measurement with transvaginal ultrasonography, on preterm delivery. A further aim was to analyze retrospectively the significance of absent or reversed end-diastolic velocity (AREDV) in the umbilical artery on perinatal outcome, and to investigate the short- and long-term outcome of infants born prematurely as a result of various causes (indicated preterm birth, preterm premature rupture of the membranes=PPROM).

Bacterial vaginosis (BV) was screened in 1956 women in a low-risk population at the first antenatal visit, using Gram stain. One hundred and one of 143 BV-positive women were randomized to receive vaginal clindamycin or placebo. Seventy-seven women at 22–32 gestational weeks with premature uterine contractions, and 78 controls were recruited for assay of cervical IL-6, IL-8-, and IGFBP-1, and ultrasonographic measurements, which were repeated twice at two-week intervals. Eighty-three women with AREDV in the umbilical artery in high-risk pregnancies at less than 34 gestational weeks (e.g. pre-eclampsia, small-for-gestational age [SGA]) between the years 1988–95 were analyzed retrospectively as regards perinatal outcome. Further, for 103 women between the 24th and the 33rd week of pregnancy, delivered by cesarean section because of maternal or fetal indications, and for 103 matched women, between the years 1990–97, their infants were analyzed as regards neonatal mortality and morbidity, and the outcome at one year of corrected age. Similarly, 78 women with PPROM at gestational weeks 17–30, and 78 controls were also analyzed.

The prevalence of BV was 7.3% (143/1956) and the preterm birth rate in women with BV was 9.9%. Preterm birth occurred in 21% vs. 0% according to whether or not BV persisted. The preterm birth rate was 14% in the clindamycin group vs. 6% in the placebo group. Cervical IL-6 at a concentration of 128 ng/L had a 73% sensitivity and 77% specificity in predicting preterm birth (35% vs. 6%). The combination of IL-6 and a cervical index of > 0.2 increased the specificity to 97%, the sensitivity falling to 45%. Concentrations of IGFBP-1 were most elevated (> 21 μg/mL) in cases with neonatal infections (36% vs. 2%). In cases of absent end-diastolic velocity (AEDV) the perinatal mortality (PNM) rate was 9%, compared with 36% in the reversed end-diastolic velocity (REDV) group. Respiratory distress (RDS) and hypoglycemia, and chronic lung disease (CLD; 15% vs. 3%) occurred significantly more often in the indicated than in the spontaneously preterm infants. The PPROM infants had more limb contractures (8% vs. 0%) and pulmonary hypoplasia (12% vs. 5%) and more chronic lung problems up to one year of age than the spontaneously preterm born infants without PPROM.

The persistence of pregnancy BV is a risk factor for preterm birth, but vaginal clindamycin used in a low-risk population in early pregnancy is of no use in reducing the preterm birth rate in cases of BV. The level of IL-6 has a relatively low sensitivity and a limited role as a single method in clinical decision making but in combination with cervical examination by ultrasonography it seems to have a predictive role in cases of threatened preterm birth. A finding of AREDV in the umbilical artery is a warning signal of threatened fetal asphyxia. Infants born after indicated preterm delivery (for fetal or maternal reasons) or PPROM are at risk of later chronic lung disease.