Bilirubin beyond the blood-brain barrier.

We evaluated the utility of recording neural signals, such as auditory brainstem responses, as potential new criteria for treatment of neonatal jaundice. Findings from recent studies of auditory brainstem responses and behavior of jaundiced infants reinforce the notion that asymptomatic and even symptomatic neurotoxicity caused by bilirubin may occur relatively frequently at relatively low serum bilirubin concentrations. Evidence suggests that this is transient and reversible. Whereas current criteria for the prevention of kernicterus are based on the idea that bilirubin entry to the CNS should be averted, the existence of transient subclinical and clinical bilirubin-induced neurotoxicity suggests that the focus of prevention might be shifted to events beyond the blood-brain barrier. With the use of "finer tools" to detect neurotoxicity, it may be possible to recognize a prior stage of neurotoxicity, described here as "transient subclinical bilirubin-induced neurotoxicity." Recordings of neural signals might be used as either predictors of kernicterus or as immediately available "outcomes" or end points with which biochemical predictors (serum bilirubin and "unbound" bilirubin concentrations) might be correlated, to determine their relative value as predictors of entry of bilirubin to the CNS. It is suggested that the results of studies to determine the relative risks associated with the various predictive criteria, namely, a prospective cohort analysis, may not be available for many years and that randomized controlled trials of new criteria for exchange transfusion are even further removed from reality.