Kernicterus 

The term kernicterus literally means "yellow kern," with kern indicating the most commonly afflicted region of the brain (ie, the nuclear region). Historically, the term refers to an anatomic diagnosis made at autopsy based on a characteristic pattern of staining found in babies who had marked hyperbilirubinemia before they died.

Hervieux first described the condition in 1847, and Schmorl first used the term kernicterus as early as 1903. Regions most commonly affected include the basal ganglia; hippocampus; geniculate bodies; and cranial nerve nuclei, such as the oculomotor, vestibular, and cochlear. The cerebellum can also be affected. Bilirubin-induced neurologic dysfunction (BIND) refers to the clinical signs associated with bilirubin toxicity (ie, hypotonia followed by hypertonia and/or opisthotonus or retrocollis) and is typically divided into acute and chronic phases. The 2 terms are commonly used interchangeably, but this use is not technically accurate because one refers to clinical manifestations and the other to an anatomic diagnosis.

Prevalent in the 1950s and 1960s, kernicterus had virtually disappeared from the clinical scene, only to reappear during the 1990s. Early discharge of term infants (before their serum bilirubin concentration peaks) may be a factor in the reemergence of this devastating neurologic affliction.

Much of the traditional teaching regarding hyperbilirubinemia is now being questioned as more is learned about bilirubin metabolism and neurologic injury. Kernicterus is now recognized in the premature infant and, very rarely, in the term infant in the absence of profound hyperbilirubinemia; however, other problems (eg, acidosis or infection) are present in term infants without profound hyperbilirubinemia. Conversely, physiologic jaundice (sometimes to levels previously thought to be universally dangerous) has been recognized to be within the reference range in the first week of life in healthy term babies, particularly those who are breastfed. Jaundice of this type usually spontaneously resolves without sequelae.

Despite the lack of a clear-cut cause-and-effect relationship between kernicterus and the degree of hyperbilirubinemia, laboratory investigations have demonstrated that bilirubin is neurotoxic at a cellular level. Other in vitro studies have shown bilirubin to have more antioxidant capability than vitamin E, which is commonly assumed to be the most potent antioxidant in the human system. This possible role of bilirubin in early protection against oxidative injury, coupled with identification of multiple neonatal mechanisms to preserve and potentiate bilirubin production, has led to speculation about an as-yet-unrecognized beneficial role for bilirubin in the human neonate.

Bilirubin staining can be noted on autopsy of fresh specimens in the regions of the basal ganglia, hippocampus, substantia nigra, and brainstem nuclei. Such staining can occur in the absence of severe hyperbilirubinemia; in this situation, factors influencing permeability of the blood-brain barrier (eg, acidosis, infection) and the amount of unbound (versus albumin-bound) bilirubin may play a role.

Characteristic patterns of neuronal necrosis leading to the clinical findings consistent with chronic bilirubin encephalopathy are also essential in the pathophysiology of this entity. Bilirubin staining of the brain without accompanying neuronal necrosis can be observed in babies who did not demonstrate clinical signs of bilirubin encephalopathy but who succumbed from other causes. This staining is thought to be a secondary phenomenon, dissimilar from the staining associated with kernicterus.

Improved brain imaging modalities, such as MRI and ultrasonography, may be emerging as instrumental tools to help clarify the complex picture of kernicterus in contrast with asymptomatic bilirubin staining of brain tissues. Bilirubin staining has been suggested to be visualized on MRI as an increased signal in the posteromedial aspect of the globus pallidus. Despite its theoretical value, however, efforts to use cranial imaging in the clinical setting have remained unsatisfying. A 2008 case series by Gkoltsiou et al reported the inexplicable conclusion that, while all children with severe cerebral palsy and a history of hyperbilirubinemia had abnormal central grey matter on later scans, the characteristic central grey matter MRI features of kernicterus were not seen in early scans.^[1]

Frequency

United States

The exact incidence of kernicterus is unknown. A pilot kernicterus registry monitoring the cases of babies with kernicterus in the United States who have been voluntarily reported shows 125 babies with chronic kernicterus enrolled in the registry from 1984-2002.^{[2, 3]} All but 4 babies reported in the registry had been discharged from the hospital fewer than 72 hours after birth (97%). Five babies were born at home (4%). No sequelae were identified in 9 of 115 infants, and 1 was lost to follow-up.

International

In Denmark, 8 cases of kernicterus were reported from 1994-2002^[4] , whereas no cases had been reported for the preceding 20 years^[5] . Following this report, from 2002-2005, a more vigilant approach was taken to the management of newborn jaundice, and no more cases have been reported in Denmark.^[6] These combined data result in an overall incidence of kernicterus in Denmark of 1.1 in 100,000 live births from 1994-2005.

In June 2003, The Quarterly Bulletin of the Royal College of Paediatrics and Child Health announced the commencement of a surveillance program of cases of severe neonatal hyperbilirubinemia following anecdotal reports throughout Britain and Ireland of increasing observation of kernicterus.^[7] A 2004 UK surveillance study has reported kernicterus occurring at a rate of 1 in 100,000 live births.

A Canadian survey published in 2004 assessed the frequency of extreme hyperbilirubinemia (serum bilirubin >427 μmol/L or >25 mg/dL) as 1:2,840 live births, of which 13 (2 in 100,000) had abnormal neurological outcomes at the time of discharge.^[8]

Using these data, the risk of developing kernicterus in infants manifesting extreme hyperbilirubinemia (>25 mg/dL) can be estimated across populations. In Canada, this risk calculates to 1 in 17.6 infants, whereas in Denmark, the population risk is estimated as 1 in 16.2.^[9] When the threshold of extreme hyperbilirubinemia is increased to >30 mg/dL (>513 μmol/L), the risk of developing kernicterus increases to 1 in 5.5-7 live births, depending on the reports. It should be noted that kernicterus also occurs in infants in whom bilirubin levels remained < 25 mg/dL, and the population risk of this occurrence remains unknown.

Hispanic and Asian populations appear to have a greater propensity to develop hyperbilirubinemia, although the underlying explanation for this observation remains elusive. Genetic variants, such as Gilbert disease or G6PD deficiency that occur in sequestered populations, result in geographic and/or ethnic differences in the risk and frequency of kernicterus.

Mortality/Morbidity

Classic kernicterus has been defined in the term infant. Increasing experience with premature babies indicates that the clinical presentation in premature infants may be somewhat different. Identifying kernicterus as the specific cause of death in either the term or preterm infant may be difficult because of concomitant ongoing pathologic conditions. Neurologic sequelae due to bilirubin encephalopathy vary, and correctly attributing some of the long-term neurologic deficits frequently associated with prematurity alone to kernicterus may be problematic.

Review of the 125 cases reported to the Pilot Registry by August 2004 revealed 26 patients with underlying G6PD deficiency, 25 with hemolytic disorders, 18 with birth trauma contributing to their hyperbilirubinemia, and 53 with idiopathic hyperbilirubinemia. Of these patients, 30 were born at 35 to less than 37 weeks estimated gestational age (EGA), 24 at 37 weeks EGA, and 71 at greater than 37 weeks EGA.^[10]

In the kernicterus registry mentioned above, 6 of 125 patients (4.8%) died.^[10] Additional risk factors for mortality included late prematurity, G6PD deficiency, and sepsis. The number of patients who died from kernicterus without being reported to the registry is unknown, as is the experience in countries other than the United States, especially those with a high prevalence of hereditary hemolytic disorders. Of patients reported to the kernicterus registry, 111 of 119 survivors (93%) had severe sequelae due to BIND; 8 of 119 babies (6.7%) had no discernible sequelae after age 1 year; one baby was lost to follow-up prior to 3 months of age.

Race

Among infants reported in the US kernicterus registry, 58% were white.^[3] Asian and Hispanic babies born either in their native countries or in the United States and Native American and Eskimo infants have higher production levels of bilirubin than white infants. Black infants have lower production levels (see image below). The reasons for these racial differences have not been fully elucidated.

Sex

Male infants have consistently higher levels of serum bilirubin than do female infants. Among infants reported in the US kernicterus registry, 67% of the patients were male.^[3]

Age

Acute bilirubin toxicity appears to occur in the first few days of life of the term infant. Preterm infants may be at risk of toxicity for slightly longer than a few days. If injury has occurred, the first phase of acute bilirubin encephalopathy appears within the first week of life.

The pilot kernicterus registry data show that, of 122 infants (all >35 weeks' gestational age at birth), symptoms became apparent in 13 babies (10.6%) aged 3.5 days or younger and in 66 babies (54%) aged 4-7 days.^[10] In 36 babies (29.5%), symptoms did not appear until after the first week of life. Most of these babies (76%) were term infants (at least 37 completed weeks' gestation), and no infant was younger than 35 weeks' estimated gestational age.