The value of amniotic fluid interleukin-6 determination in patients with preterm labor and intact membranes in the detection of microbial invasion of the amniotic cavity.

OBJECTIVE

Our purpose was to (1) determine the value of amniotic fluid interleukin-6 determination in the detection of microbial invasion of the amniotic cavity and (2) compare interleukin-6 to other rapid tests in the evaluation of preterm labor.

METHODS

Amniotic fluid interleukin-6 was determined quantitatively by enzyme-linked immunosorbent assay in 91 amniotic fluid specimens obtained by amniocentesis in 89 patients with preterm labor. Amniotic fluid cultures for aerobes, anaerobes, and mycoplasma species were performed. Receiver-operator characteristic curve analysis, logistic regression analysis, and Cox's proportional-hazards model were used to explore the relationship between several explanatory and outcome variables. Diagnostic index values of interleukin-6, glucose level, Gram stain, leukocyte esterase, and limulus amebocyte lysate assay for prediction of a positive amniotic fluid culture, preterm delivery, clinical infection, and neonatal sepsis were calculated.

RESULTS

The prevalence of positive amniotic fluid cultures was 13% (12/89). The median interleukin-6 concentration in patients with positive cultures was 241.8 ng/ml, in contrast to 0.291 ng/ml in patients with negative cultures (p. < 0.005). Sensitivity and specificity of an interleukin-6 level > or = 6.17 ng/ml was 75% and 79%, in contrast to that of glucose, < or = 12 mg/dl (83% and 86%) for a positive amniotic fluid culture and sensitivity (p = 0.26, z test). All patients with an interleukin-6 level > 6.17 ng/ml were delivered preterm, in contrast to 85.2% of patients with a glucose level < or = 12 mg/dl. When all rapid tests and clinical parameters were considered simultaneously in the logistic model, only interleukin-6 maintained a significant relationship to preterm birth (odds ratio 35, p = 0.003). Cox's proportional analysis demonstrated a strong relationship between interleukin-6 and the amniocentesis-to-delivery interval after clinical variables were controlled for (hazard ratio 3.01, p < 0.00001).

CONCLUSIONS

Amniotic fluid interleukin-6 determination may be a useful adjunct to our armamentarium of rapid tests to exclude infection and predict delivery in patients with preterm labor and intact membranes.