Hyperbilirubinemia is a natural and essentially ubiquitous transitional phenomenon among human newborns. Approximately 60% to 70% of all term infants, and nearly all premature infants, become visibly jaundiced during the first week of life after birth. For term infants, the serum or plasma total bilirubin (PTB) concentration typically peaks 3 to 4 days after birth in the range of 5 to 6 mg per dL (86 to 103 μmol per L). For premature infants, PTB levels peak later and higher after the first several days of life. Although much uncertainty and debate remains concerning the range of PTB considered as benign physiologic jaundice, the consensus places the maximal “safe” peak PTB threshold at approximately 17 mg per dL (291 μmol per L) for otherwise healthy term and late preterm babies. Because hyperbilirubinemia above this threshold is considered to be pathologic, the etiology of the hyperbilirubinemia should be investigated and appropriate therapy considered or initiated depending on the clinical circumstances (1). In immature infants, treatment to decrease PTB levels is often initiated at lower PTB levels because of lower albumin levels and diminished affinity of albumin for bilirubin in these infants. Moreover, binding to albumin is least avid in the early transitional period after birth and is influenced adversely by any confounding conditions, such as infection or acidosis, that increase free bilirubin in circulation and the likelihood of movement into tissues.

Historically, the main therapies for neonatal hyperbilirubinemia have been phototherapy and exchange transfusion. Light could be considered a drug for hyperbilirubinemia, but most physicians pay little attention to the radiometric qualities (effective spectral width and peak emission) and quantities [irradiance (μW per cm² per nm)] involved, or the factors that affect the dose of phototherapy (duration, body surface area exposure) (2,3). Finally, issues that are involved in producing phototherapy-related side effects (riboflavin destruction, erythema, and photosensitizing drugs) also deserve attention. A discussion of these topics should include reference to the qualities of light-emitting diodes because of their high intensity and narrowband light in the spectrum of choice with minimal heat generation (3,4,5,6), but these issues are beyond the scope of this chapter. In spite of the proven benefits of phototherapy and maximal spatial limitations, the understanding of the biology of newborn jaundice and the existence and further development of alternative pharmacologic therapies for neonatal hyperbilirubinemia are an integral part of the management of neonatal jaundice and its consequences.

Neonatal jaundice is the result of an imbalance between the production of bilirubin and its elimination (7,8,9). Bilirubin production on a body weight basis is increased in the newborn by approximately two to three times that of an adult (10,11). This relative increase in bilirubin production in the newborn is the result of an increased circulating red cell mass and a shortened red cell life span. Consequently, all newborn infants have increased bilirubin production as a contributing cause of their transitional or pathologic jaundice. The pattern of hyperbilirubinemia (its peak and duration) is influenced further by the efficiency with which the pigment is eliminated. The major factor contributing to impaired elimination in the transitional period after birth is decreased hepatic conjugation of bilirubin. The gradual induction of uridine diphosphoglucuronate glucuronosyltransferase (UGT) contributes most importantly to the pattern of hyperbilirubinemia after birth because changes in bilirubin production are slower and more gradual within the time frame of the rapid elevation in PTB levels after birth and the decline in the latter part of the first week and into the second week of life. Thus, because all newborn babies have temporarily impaired conjugation, any pathologic state associated with increased bilirubin production, such as hemolysis, represents a serious risk to the newborn infant, especially in the first several days of life. Even without pathologic elevations in bilirubin production, greater impairments in conjugation associated with conditions such as Gilbert syndrome (12,13,14) and the Asian G71R mutation in the UGT1A1 gene (14,15,16) can place infants at risk for kernicterus because of unexpected alterations in the pattern of hyperbilirubinemia, including its peak and duration. In particular, the coexpression of gene polymorphisms involved in bilirubin production, such as (GT)_n, repeats in the HO-1 promoter and glucose-6-phosphate dehydrogenase
(G6PD) mutations, and metabolisms, such as OATP1A1 and UGT1A1 and the TATAA box variants, may provide genetic markers for clinical risk assessments, as well as potential therapeutic targets (17,18).

The imbalances in the production of the pigment and its elimination have been well studied, and various methods have been proposed to identify infants at risk for severe hyperbilirubinemia. Because the predominant source of carbon monoxide (CO) in the body is the degradation of heme, which ultimately leads to the production of equimolar amounts of bilirubin, increased bilirubin production can be estimated by measuring the end-tidal CO in breath, or carboxyhemoglobin (COHb) in circulation, after these measurements are corrected for ambient CO (ETCOc or COHbc, respectively) (19,20). The normalization of COHbc to hemoglobin concentration (COHbc/Hb) can serve as an even more sensitive index of excessive red cell destruction (21). For example, the infant with hemolytic anemia would have a higher COHbc/Hb ratio than the infant with anemia caused by blood loss. By measuring the conjugated fraction of bilirubin, another index can be applied, which assesses the relative balance between bilirubin production and conjugation [COHbc/TCB (%)], where TCB is the total conjugated bilirubin (9). Finally, a nomogram plotting hour-specific bilirubin levels is also informed by the balance of bilirubin production and elimination over time (22). Because all infants have impaired conjugation during the transitional period, deviations from the percentile tracks in the first several days of life are most often related to a relative increase in bilirubin production, whereas deviations after the first week of life are more likely the result of persistent impairment in bilirubin conjugation and therefore elimination.

The logic of removing bilirubin from circulation after it has been produced is clear, but it is also only reactionary and may not avoid potential neurologic injury in every circumstance. Another, more rational, treatment strategy would be to inhibit bilirubin production, thus ameliorating the primary contributing factor in neonatal hyperbilirubinemia and the risk for kernicterus. If a safe drug for inhibiting bilirubin production could be identified, its use could be universalized. However, another alternative preventive strategy could involve the early rapid and accurate identification of infants at risk for increased bilirubin production or at least with an imbalance between the production and the elimination of bilirubin. One direct approach would be to measure ETCOc or COHbc as an index of bilirubin formation and, thus, identify high producers of the pigment for targeted therapy. Yet another approach would be to plot hour-specific bilirubin levels and be cognizant of early deviations from the nomogram suggestive of increased bilirubin production or the combination of relatively insufficient conjugation for a given bilirubin load (23). Whether the approach is targeted would depend, at least in part, on the safety, efficacy, and cost of the chemotherapeutic agent.

Heme is degraded in a two-step enzymatic pathway, which requires molecular oxygen and NADPH (see Fig. 18.1). The first step is rate limiting and catalyzed by heme oxygenase (HO), a membrane-bound enzyme (24). In this first step, the porphyrin macrocycle is broken at the 9-α-meso carbon bridge after a series of oxidations and reductions, and liberates CO, iron, and biliverdin in equimolar amounts. In the second enzymatic step, biliverdin is immediately reduced in the cytosol by biliverdin reductase in an NADPH-dependent reaction to generate bilirubin. Because HO is the rate-limiting enzyme in the pathway, its inhibition results in decreased production of CO, iron, and biliverdin, thus leading to decreased bilirubin production (25,26).

D-Penicillamine is a chelating agent in use since the 1950s in the treatment of Wilson disease and heavy metal intoxication. It has also been used to treat cystinuria and rheumatoid arthritis. In the early 1970s, D-penicillamine was described as a therapy for neonatal hyperbilirubinemia in Europe (59). Further studies provided a likely mechanism of action for its use in the treatment of hyperbilirubinemia, indicating that a 3-day course of the drug significantly reduced HO activity levels in neonatal, but not adult, rats (60). A study of D-penicillamine to reduce severe hyperbilirubinemia was conducted in 120 full-term infants with ABO hemolytic disease over a 5-year period, using 60 untreated infants from the first part of the study period as historical controls (59). If initiated within the first 24 hours of life, treatment with D-penicillamine was associated with significantly reduced PTB levels and a decreased need for exchange transfusion in this population. Although the proposed mechanism in its action in decreasing PTB levels may be through inhibition of HO activity, this finding has not been confirmed through direct measurement. Its administration (300 to 400 mg per kg per day) has been associated with amelioration of neonatal hyperbilirubinemia; however, its efficacy has not been well proven, especially in light of the considerable risks associated with its use in neonates.

Numerous cutaneous lesions, including urticaria, macular or papular reactions, pemphigoid lesions, and dermatomyositis, have been reported with long-term use of D-penicillamine (61). Furthermore, significant renal, hepatic, and hematologic complications, including nephrotic syndrome, Goodpasture syndrome, elevation of liver enzymes, aplastic anemia, and thrombocytopenia, have been associated with prolonged therapy for rheumatoid arthritis. The effect of short-term therapy on liver function has been reported in 20 term or near-term infants and on renal function in 30 term or near-term infants (62). Liver function tests were found to be unchanged after 4.3 ± 1.7 days of treatment with D-penicillamine at a dose of 300 mg per kg per day. Cholesterol, blood urea nitrogen, and creatinine levels were also unaffected after 2.8 ± 1.1 days of D-penicillamine treatment at the same dose. The in vitro effect of the drug on human peripheral granulocytes was also investigated. Superoxide anion generation and β-glucuronidase release were both found to be significantly increased at all concentrations of D-penicillamine; however, phagocytic or killing activity of the granulocytes appeared to be unaffected by the drug.

A decreased incidence of retinopathy of prematurity (ROP) was unexpectedly noted among very low-birth-weight infants treated with D-penicillamine in later studies (63,64). Recently, a meta-analysis of the effect of prophylactic D-penicillamine on the incidence of ROP in infants of less than 2,000 g birth weight was undertaken (65). This review concluded that treatment was unlikely to affect survival and may reduce the incidence of ROP among survivors. It is important to note, however, that these conclusions were based on the findings of only two randomized trials, and that no conclusion could be reached regarding the effect of D-penicillamine on severity of ROP. Further studies are required to fully evaluate the possible efficacy and potential side effects of D-penicillamine in the neonatal population before including this drug in the therapeutic armamentarium.