Cord Blood Proteins and Multichannel-Electroencephalography in Hypoxic-Ischemic Encephalopathy*

Abstract

To explore the association between multiple umbilical cord blood proteins and severity of hypoxic-ischemic encephalopathy as defined by continuous multichannel electroencephalography. A prospective case-control cohort study, which was divided into separate exploratory and validation cohorts. A single tertiary neonatal intensive care facility. The study recruited full-term infants with perinatal asphyxia and controls. Identical procedures were used to recruit a representative exploratory sample (n = 30) and a subsequent validation cohort (n = 100). All had umbilical cord blood drawn and biobanked at delivery, continuous multichannel electroencephalography commenced in the first 24 hours, and a modified Sarnat score assigned. Analysis of 37 potential cord blood protein markers of hypoxic-ischemic encephalopathy was performed using Luminex multiplex assays. Cord blood from 130 infants was analyzed. Interleukin-16 and interleukin-6 significantly differentiated between a moderate-severely abnormal and normal-mildly abnormal electroencephalography background in both exploratory (p = 0.005 and p = 0.016, respectively) and validation cohorts (p = 0.039 and p = 0.024, respectively). To develop a predictive model for a moderate-severely abnormal electroencephalography, stepwise regression analysis was used to combine these analytes with current standard clinical markers of asphyxia (pH, base deficit, and 10-min Apgar). Only Apgar score and interleukin-16 remained in the model, which was highly predictive of an abnormal electroencephalography (area under the curve [AUC] = 0.956, p < 0.001, positive predictive value = 89%, and negative predictive value = 94%). Cord blood interleukin-6 and interleukin-16 were associated with electrographic grade of hypoxic-ischemic encephalopathy. To predict an abnormal electroencephalography, interleukin-16 and 10-minute Apgar used in combination performed better than current markers.