ELEVATED AMNIOTIC-FLUID INTERLEUKIN-6 AS A PREDICTOR OF NEONATAL PERIVENTRICULAR LEUKOMALACIA AND INTRAVENTRICULAR HEMORRHAGE

Objective: To investigate the relationship between amniotic fluid inte rleukin-6 levels and the development of periventricular leukomalacia a nd intraventricular hemorrhage in the preterm neonate and to compare t he value of amniotic fluid interleukin-6 with amniotic fluid culture a nd histologic chorioamnionitis in the prediction of periventricular le ukomalacia and intraventricular hemorrhage. Methods: 119 women, betwee n 20 and 34 weeks gestation, in preterm labor with intact membranes, u nderwent transabdominal amniocentesis. Amniotic fluid was cultured for aerobic and anaerobic bacteria, Ureaplasma urealyticum and Mycoplasma hominis. Amniotic fluid interleukin-6 levels were determined by enzym e-linked immunosorbent assay. The placentas were examined for histopat hologic evidence of inflammation. Where the birth weight was <2,000 g, transfontanelle cranial sonography was performed on the 3rd and 7th d ays of life for diagnosis of periventricular leukomalacia and intraven tricular hemorrhage. Student's t test, the Mann-Whitney U test, likeli hood ratio chi(2), logistic regression, and receiver-operator characte ristic curve were used for analysis. Results: 33 women were excluded f rom the analysis because they delivered at other institutions. The neo nates of 33 women did not have sonography because they weighed >2,000 g at birth. Two neonates died before sonography was performed; four ne onates who weighed <2,000 g at birth did not have sonography. In the d efinitive study group of 47 women, those with neonates who developed p eriventricular leukomalacia and intraventricular hemorrhage (n = 14) h ad higher median amniotic fluid interleukin-6 levels (42,795 pg/ml ver sus 8,020 pg/ml; P = 0,009), more positive amniotic fluid cultures (64 % vesus 21%; P < 0.003), and a shorter median amniocentesis-to-deliver y interval (16 h versus 24 h; P = 0.045) than women (n = 33) who deliv ered neonates without periventricular leukomalacia or intraventricular hemorrhage. The groups did not differ in gestational age at admission (P = 0.15), birth weight (P = 0.09), or histologic chorioamnionitis ( P = 0.37). An amniotic fluid interleukin-6 level greater than or equal to 20,000 pg/ml had a sensitivity of 71% and a specificity of 70% com pared with a sensitivity of 69% and specificity of 79% for amniotic fl uid culture, and a sensitivity of 71% and specificity of 42% for histo logic chorioamnionitis in the prediction of periventricular leukomalac ia and intraventricular hemorrhage. Women with amniotic fluid interleu kin-6 levels greater than or equal to 20,000 pg/ml (n = 20) had more n eonates with periventricular leukomalacia or intraventricular hemorrha ge than women with amniotic fluid interleukin-6 levels <20,000 pg/ml ( n = 27) (50% versus 15%; P = 0.009). They also were of lower birth wei ght (P = 0.02), had more neonatal morbidity (P = 0.01), had more posit ive amniotic fluid cultures (P = 0.01), and more histologic chorioamni onitis (P = 0.02). Logistic regression analysis demonstrated that amni otic fluid interleukin-6 was an independent risk factor for the develo pment of periventricular leukomalacia and intraventricular hemorrhage (odds ratio, 5.81; 95% confidence interval, 1.02-33.16; P = 0.05) afte r controlling for gestational age, birth weight, histologic chorioamni onitis, and amniotic fluid culture (odds ratio, 7.94; 95% confidence i nterval 1.22-51.77; P = 0.03). Conclusions: In women in preterm labor with intact membranes amniotic fluid interleukin-6 is useful in predic ting neonatal periventricular leukomalacia and intraventricular hemorr hage.