Relation between viability of vaginal polymorphonuclear leukocytes and presence of histologic chorioamnionitis.

BACKGROUND

The number and percentage of viable vaginal polymorphonuclear leukocytes (vPMNs) are known to be increased in women who experience preterm labor. Whether those parameters may be associated with the presence of histologic chorioamnionitis (CAM) is not known.

METHODS

We investigated prospectively 39 women at 26.3 +/- 6.2 weeks of gestation. The following were determined in vaginal washings: total number of vPMNs and the percent that were viable, the pH, and the concentrations of granulocyte elastase and interleukin-8 (IL-8). In addition, the white blood cell count and the serum level of C-reactive protein were determined in peripheral blood. The placenta and the umbilical cord were examined histologically with special reference to the presence of CAM. The optimal cutoff value for prediction of histologic CAM was obtained for each variable using receiver operating characteristic curves. A multivariate logistic-regression model was used to determine the independent risk factors for this disorder.

RESULTS

Histologic CAM was present in ten women (37.1 +/- 3.8 weeks) and absent in 29 women (38.2 +/- 1.5 weeks). The total number of vPMNs, the percent of viable vPMNs, and the IL-8 level were all significantly increased in the women with CAM, in contrast to those without CAM. When the optimal cutoff value for each of seven covariates was entered into the model, only the percent of viable vPMNs, > or = 11%, demonstrated a significant relationship with histologic CAM (odds ratio 26.9; 95% confidence interval 1.3 to 545; p < 0.05).

CONCLUSIONS

Women with a % viability of vPMNs of > or = 11% were at a significantly higher risk for histologic CAM. Data suggest that an influx of PMNs into the vagina occurs continuously in patients with histologic CAM.