Elevated amniotic fluid interleukin-6 levels at genetic amniocentesis predict subsequent pregnancy loss

OBJECTIVE: Our purpose was to determine the proportion of pregnancy loss after genetic amniocentesis that is related to preexisting subclinical intrauterine inflammation. STUDY DESIGN: We accessed our bank of stored second-trimester amniotic fluid and maternal serum samples obtained from women undergoing genetic amniocentesis at our institution from 1988 to 1995 (N = 11,971). Interleukin-6 levels were measured by enzyme-linked immunosorbent assay in samples from every case resulting in spontaneous postprocedure loss (excluding fetal aneuploidy and anomalies) within 30 days after the procedure (n = 66) and from 66 normal control women delivered at term and matched for year of test, gestational age, maternal age, and indication for amniocentesis. RESULTS: Mean maternal serum interleukin-6 levels were the same in each group (0.02 +/- 0.07 ng/ml for cases and 0.06 +/- 0.25 ng/ml for controls, p = 0.45). Mean amniotic fluid interleukin-6 levels were higher in cases (4.0 +/- 13.1 ng/ml) than in controls (0.5 +/- 0.7 ng/ml, p = 0.04). The higher mean amniotic fluid interleukin-6 levels in the cases resulted from the inclusion of eight very high values (greater than or equal to 3 SD or greater than or equal to 2.5 ng/ml). When these samples were excluded, the means and range of values were the same in each group (0.4 +/- 0.4 ng/ml for cases and 0.5 +/- 0.7 ng/ml for controls, p = 0.58). Twelve percent (8/66) of the cases and 3% (2/66) of the controls had amniotic fluid interleukin-6 levels greater than or equal to 2.5 ng/ml (p = 0.048, odds ratio 4.1, 95% confidence interval 1.0 to 31.2). Although the overall correlation between maternal serum and amniotic fluid interleukin-6 levels was good (r = 0.50, p <0.002), only one of the eight cases would have been identified by a maternal serum interleukin-6 level greater than or equal to 3 SD above the mean (greater than or equal to 0.8 ng/ml). CONCLUSION: Analysis of our complete unselected group of postamniocentesis pregnancy losses indicates that up to 12% may result from preexisting subclinical intrauterine inflammation. This inflammation is most likely localized and may not be identified by a maternal serum interleukin-6 level before the procedure.