Measuring free bilirubin: the clinical perspective.

In 1959, Gerald Odell first proposed that albumin and tissue compete for binding bilirubin and that this competition is mediated by a very low concentration of free bilirubin (Bf) (bilirubin not bound to plasma proteins). He argued that displacement of albumin-bound bilirubin by binding competitors could explain the reported high incidence of kernicterus in premature infants receiving sulfonamide prophylaxis (1 ). This sentinel publication led to a flurry of proposed binding tests, most of which were nonspecific competitivebinding assays, until 1974, when Jacobsen and Wennberg presented a method for measuring Bf in infant plasma that was based on the observation that only unbound bilirubin serves as a substrate for enzymatic oxidation by horseradish peroxidase (2 ). Several studies from Japan, where the peroxidase assay had been adapted to a semiautomated instrument, identified a clear relationship of Bf with acute neurotoxicity in premature infants (3 ) and with acute auditory impairment in term infants (4 ). Enthusiasm was tempered in the US by a perceived difficulty in performing the assay and by confusion regarding the poor correlation between Bf and outcome in premature babies who died with kernicterus at very low total bilirubin (Bt) concentrations (5 ). The reported high incidence of death of infants with kernicterus was isolated to specific care centers: Very high Bf/Bt ratios in affected infants indicated very poor binding, and the phenomenon of lethal kernicterus at low Bt concentrations largely disappeared after cessation of benzyl alcohol use as a preservative in multiple-dose vials or solutions for parenteral therapy (6 ). Kernicterus is still occasionally observed at low Bt concentrations in babies with sepsis or renal failure, conditions associated with impaired albumin binding of bilirubin and other ligands. Interest in measuring Bf declined during the 1970s after the development of passive immunization to prevent Rh hemolytic disease (7 ) and the introduction of phototherapy as an effective treatment for hyperbilirubinemia (8 ). With the aggressive use of phototherapy, frank kernicterus became a rare event in affluent nations. Concern about jaundice was placed on the back burner until a perceived resurgence of kernicterus in term/near-term infants led the American Academy of Pediatrics (AAP) to issue new guidelines in 2004 to improve early recognition of babies at risk for severe hyperbilirubinemia (9 ). Intervention guidelines based on the Bt concentration (adjusted downward with a number of coexisting conditions such as sepsis, hemolysis, low albumin, low gestational age) were little changed, and the AAP panel acknowledged that the uncertainty in risk assessment required treating a large number of babies to prevent a single adverse outcome. Evidence to support the rather specific recommendations was surprisingly sparse, being limited to 120 patients with a variety of acute and/or persistent adverse auditory or neurologic outcomes reported in scattered case studies and reports (10 ). There are good theoretical reasons and an increasing body of clinical evidence to suggest that risk assessment based on Bf or a combination of Bt, albumin, and Bf will improve the identification of babies in need of treatment (11 ). In this issue of Clinical Chemistry, Huber and colleagues describe a new approach to Bf measurement based on competitive binding of bilirubin to a high-affinity probe coupled to a sensitive fluorophore (12 ). In contrast to the peroxidase assay, which requires measuring a reduction in the Bt concentration and extrapolating that to the initial Bf-dependent oxidation rate, the high-affinity modified free fatty acid receptor can be used at a very low concentration so that the equilibrium Bf concentration is minimally perturbed. The developers have carefully documented the potential pitfalls in Bf measurement and have used model sera in validating the method by comparing it with the peroxidase method. The simplicity of the method and the potential for point-of-care instrumentation are very attractive and worthy of clinical trials that compare proposed markers for intervention. The clinical application of predictive markers is complicated by vagaries in the behaviors of albumin binding and bilirubin chemistry, notably the effects of plasma dilution and photoisomerization on Bf assays and the interpretation of results. Plasma dilution in1 Department of Pediatrics, University of Washington, Seattle, WA. * Address correspondence to the author at: 23304 Locust Way, Bothell, WA 98021. Fax 206-543-8926; e-mail [email protected]. Received February 19, 2012; accepted February 28, 2012. Previously published online at DOI: 10.1373/clinchem.2012.183962 2 Nonstandard abbreviations: Bf, free bilirubin; Bt, total bilirubin; AAP, American Academy of Pediatrics. Clinical Chemistry 58:5 811–813 (2012) Editorials