PEDIATRICS Vol. 107 No. 5 May 2001, pp. 1075-1080

Developmental Follow-Up of Breastfed Term and Near-Term Infants With Marked Hyperbilirubinemia

Mary Catherine Harris, MD^*, Judy C. Bernbaum, MD, Jessica R. Polin^*, Robert Zimmerman, MD^§, and Richard A. Polin, MD^*

From the ^* Division of Neonatology and  Division of General Pediatrics, Department of Pediatrics, and the ^§ Division of Neuroradiology, Department of Radiology, Children's Hospital of Philadelphia and Hospital of the University of Pennsylvania, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania.

	ABSTRACT

Top Abstract Methods Results Discussion Conclusion References

Objective.  In recent years, the increased prevalence of breastfeeding in conjunction with early discharge practices has increased the risk for marked hyperbilirubinemia in neonates. This has resulted in the potential for bilirubin brain injury in affected infants. The purpose of this study was to identify all infants 36 weeks' gestational age with bilirubin levels >25 mg/dL and evaluate them for early and late evidence of bilirubin brain injury.

Methods.  We reviewed the charts of all infants (from 1993-1996) 36 weeks' gestational age who were readmitted to the hospital during the first week of life with bilirubin levels >25 mg/dL. Readmission records were reviewed for early signs of bilirubin encephalopathy. Magnetic resonance imaging (MRIs) and Brainstem auditory-evoked responses (BAERs) were reviewed for evidence of bilirubin toxicity. At follow-up, study infants had a complete neurodevelopmental examination, repeat MRIs, and behavioral hearing evaluations.

Results.  From 1993 to 1996, we identified 6 term and near-term infants readmitted to the hospital within the first week of life with peak bilirubin values ranging from 26.4 mg/dL (451 µmol/L) to 36.9 mg/dL (631 µmol/L). Five of 6 infants had bilirubin values >30 mg/dL (513 µmol/L). All were exclusively breastfed or fed a combination of breast and bottle feedings. Five of 6 infants presented with abnormal neurologic signs. Four infants had initial MRIs, 3 of whom had increased signal intensity in the basal ganglia consistent with kernicterus. Two infants had abnormal BAERs; both also had abnormal MRIs. Five of 6 infants received exchange transfusions and all were treated with phototherapy and intravenous fluids. Follow-up examinations between 3 months and 2 years showed resolution of clinical signs in all but 1 infant. Four infants had a subsequent normal MRI and 1 had residual hearing impairment. One infant demonstrated severely abnormal developmental evaluations, as well as both an abnormal initial MRI and BAERs. Follow-up MRI showed evidence of encephalomalacia with changes not characteristic of kernicterus.

Conclusions.  We observed transient neurologic abnormalities in 5 of 6 infants readmitted to the hospital during the first week of life with marked hyperbilirubinemia. The abnormalities resolved following aggressive management using hydration, phototherapy, and exchange transfusion and may not correlate with long-term prognosis. Less aggressive therapy may be associated with residual neurologic abnormalities. We speculate that inadequate establishment of breastfeeding coupled with early discharge practices may play a role in the development of marked hyperbilirubinemia in these infants.  Key words:  bilirubin, brain injury, kernicterus.

In recent yars, the increased prevalence of breastfeeding in conjunction with early discharge practices has increased the risk for marked hyperbilirubinemia in term and near-term infants.^1,2 This has resulted in the potential for bilirubin brain injury in affected infants.³ Many of these infants are term or near-term with presumed nonhemolytic hyperbilirubinemia.

During the last several years, approaches to the therapy of neonatal hyperbilirubinemia have also changed. Notably, there has been a movement away from the age old concept of vigintiphobia to a kinder, gentler approach to the management of neonatal jaundice.^4-6 This nonaggressive approach was recommended because of the lack of observed toxicity when bilirubin values were conservatively managed, as well as the paucity of evidence that elevated bilirubin values in term infants without hemolysis were hazardous.⁶ However, these more relaxed guidelines for therapy, in conjunction with both the increased frequency of breastfeeding and early discharge practices, may be factors that place infants at risk for the development of marked hyperbilirubinemia and bilirubin brain injury.^3,7 In a recent study of infants discharged at <72 hours, the most common cause for readmission within the first 2 weeks of life was hyperbilirubinemia.⁸

The purpose of this study was to evaluate term and near-term infants with marked hyperbilirubinemia for evidence of bilirubin brain injury at the time of readmission to the hospital during the first week of life and at follow-up between 3 and 24 months old.

	METHODS

Top Abstract Methods Results Discussion Conclusion References

Study infants were selected from a retrospective chart review of all infants from September 1993 to September 1996 readmitted to the hospital during the first week of life with a diagnosis of hyperbilirubinemia. Criteria for entry into the study included: 1) bilirubin levels >25 mg/dL; and 2) gestational age 36 weeks.

During this time period, we identified 6 infants who met study criteria (Table 1). The infants studied had a mean birth weight of 2.9 kg and a mean gestational age of 39 weeks. All of the infants had been initially discharged from the hospital within 48 hours of birth. The mean age at readmission was 5 days. All of the infants were either exclusively breastfed or fed a combination of breast and bottle feedings. Two infants had mild jaundice noted before discharge (patients 4 and 5); patient 5 had a predischarge bilirubin of 12.7 mg/dL on day 2. The mothers of 5 of the 6 patients noted jaundice at home often accompanied by decreased activity or poor feeding in the 24 hours before admission.

None had any laboratory evidence of hemolytic disease as evidenced by anemia, reticulocytosis, or abnormality of the blood smear. Infants were also screened for perinatal factors associated with an increased risk of hyperbilirubinemia, including ethnic influences, maternal diabetes mellitus, hypertension, administration of oxytocin during labor, as well as neonatal bruising, central nervous system bleeding or injury, or perinatal infection. Two infants were of Asian extraction and 1 was black. One was born of a mother with gestational diabetes, however, maternal dietary and drug history were unremarkable otherwise. None of the infants was suspected of having an infection during the initial hospital stay.

At the time of readmission, study infants were evaluated for clinical signs of acute bilirubin encephalopathy manifested as changes in level of consciousness, tone/movement, and alterations in brainstem function. Infants were screened for lethargy, irritability, poor feeding, abnormal cry, muscle tone abnormalities, arching, opisthotonus, and/or seizures.⁹ Other causes of hyperbilirubinemia including bruising, dehydration, and sepsis were investigated. Laboratory studies included complete blood counts, electrolytes, glucose, and cultures. Magnetic resonance imaging (MRIs) were performed on a 1.5 Tesla Siemens Vision MR Imager (Siemens Medical Systems, Inc, Iselin, NJ) using conventional proton density and T1-weighted (T1WI) and T2-weighted pulse sequences. T1WI are with a repetition time (TR) of 3500 msec and an echo time (TE) of 90 msec. Conventional spin echo proton density are with TE of 30 msec. T1WI are with a TR of 650 msec and a TE of 15 msec. Matrix size are 256 × 256 for T1 and T2. Brainstem auditory evoked responses screening (BAERs) was completed using a Natus Algo 2e automated auditory brainstem response system (Natus Medical, Inc, San Carlos, CA). Stimuli were presented at a level of 35 dB nHL and responses measured using a nape-vertex electrode array with shoulder ground. Pass or fail criteria for results were determined per the standard equipment algorithm (yielding a published sensitivity and specificity of 96%-99%).

Study infants were seen in follow-up between 3 and 24 months old. All children underwent a complete neurologic examination¹⁰ including evaluation of muscle tone and reflexes. Repeat MRIs and follow-up behavioral hearing evaluations were performed when possible pending parents' agreement. Developmental assessment was performed by a licensed psychologist using the Bayley Scales of Infant Development yielding a mental developmental index (MDI) and psychomotor developmental index (PDI), which have a mean of 100 and one standard deviation of 15 points. Interim history was also obtained specifically related to feeding regimen, significant illnesses, medications, and intervention services.

	RESULTS

Top Abstract Methods Results Discussion Conclusion References

Table 2 shows the peak serum bilirubin values as well as detailed information regarding the 6 study infants at the time of readmission to the hospital for hyperbilirubinemia. Peak bilirubin levels ranged from 26.4 mg/dL (451 µmol/L) to 36.9 mg/dL (631 µmol/L) measured between day 3 and day 6 of age. Five of 6 infants had bilirubin values >30 mg/dL. Five of 6 infants presented with 1 or more abnormal neurologic signs including 4 with lethargy, 1 with irritability, 2 with poor feeding, 2 with a high pitched cry, 4 with abnormal muscle tone, and 2 with arching, opisthotonus, or seizures. The 1 infant with a normal neurologic examination presented with documented urosepsis.

Four of the 6 infants received antibiotic therapy for suspected sepsis, only 1 of whom had documented Escherichia coli urosepsis. All other infants demonstrated negative evaluations and cultures for bacterial infection. All infants received phototherapy and 5 received 1 or more exchange transfusions. One infant (patient 3) was treated with intravenous fluids and phototherapy alone. One infant (patient 6) demonstrated severe dehydration with an 18% weight loss and an admission serum sodium level of 160 mg/dL. The other 5 infants had a mean weight loss of 4% since birth (range 0% to 10%) with serum sodium values on readmission between 134 and 142 mg/dL. All infants had a history of normal urination and stooling patterns, except patient 6 who had both diminished urine and stool output. All 6 infants received fluid therapy for suspected dehydration. Glucose6-phosphate dehydrogenase levels were measured on patients 2 through 6 and were normal; thyroid function tests were measured on patients 1 through 5 and were normal as well. All of the other laboratory results were unremarkable including serum electrolytes, glucose, complete blood counts, and cultures.

Four infants had MRI studies performed as part of their evaluation during readmission (days 5-21). Three infants had abnormal studies, demonstrating increased T1WI intensity in the basal ganglia consistent with kernicterus. One infant had a normal MRI study. Five infants also had BAERs screening performed between days 8 and 26. Of these, 2 infants demonstrated neurosensory hearing abnormalities, both of whom also had abnormal MRIs. Figure 1A shows a representative abnormal initial MRI from 1 of our study infants performed at 9 days of age (patient 3).

View larger version (109K): [in this window] [in a new window]   Fig. 1.   Initial and follow-up T1WI MRI studies on patient 3. A, Initial MRI. Abnormal increased signal intensity is seen in the globus pallidus bilaterally (arrows). Normal signal intensity is seen in the posterior limb of the internal capsule (arrow heads). B, Follow-up MRI. Normal signal intensity is seen in the posterior limb of the myelinated internal capsule (arrow heads).

Table 3 shows the results of follow-up evaluations of study infants with peak serum bilirubin levels listed at the top as a point of reference. Neurodevelopmental examinations demonstrated resolution of clinical signs in all but 1 infant (patient 6) who continued to manifest residual neurologic abnormalities. The other 5 infants had entirely normal neurodevelopmental evaluations. The mean MDI score in the normal infants was 99.7 and the mean PDI was 100.3. 

Four infants who were normal on follow-up had a subsequent normal MRI with no evidence of increased signal intensity in the basal ganglia. Figure 1B shows the resolution of MRI abnormalities in patient 3. The same infant (patient 3), however, has residual hearing loss bilaterally requiring hearing aids (Table 3). Two infants had normal follow-up behavioral hearing evaluations. In the remaining 2 otherwise normal infants who were not formally tested, there was no clinical suspicion of hearing loss. Parents of these infants refused hearing tests, stating that there was no evidence of hearing dysfunction clinically based on screening performed at their primary care centers. All of these children developed normal language acquisition.

One infant (patient 6) demonstrated MDI and PDI scores in the severely abnormal range. On initial evaluation this infant had both an abnormal MRI and BAERs. On follow-up MRI, this infant had evidence of encephalomalacia with changes that were not characteristic of kernicterus. There was no history of perinatal asphyxia. Follow-up hearing evaluation, however, did not show neurosensory hearing loss. In addition to ongoing seizure activity, this infant has hypertonic cerebral palsy and blindness. We believe that these findings are not solely attributable to this infant's marked hyperbilirubinemia and kernicterus.

Because 5 of 6 infants had normal follow-up evaluations, recommendations were made that these infants be followed clinically and referred back for subsequent developmental evaluations as needed. Telephone follow-up has confirmed that these children continue to develop normally at 3 to 6 years.

	DISCUSSION

Top Abstract Methods Results Discussion Conclusion References

Hyperbilirubinemia is felt to be a benign condition for infants born at term or near-term gestation. In ~5% of healthy term infants, however, serum bilirubin values exceed 17 mg/dL (291 µmol/L), a value which the American Academy of Pediatrics deems significant.⁴ Levels exceeding 20 mg/dL (342 µmol/L) occur in 1.2% of healthy newborn infants.¹¹ The vast majority of infants with serum bilirubin values in excess of 20 mg/dL remain well. They need minimal care other than the occasional use of phototherapy and careful monitoring of serum levels. Before 1990, kernicterus in the previously healthy-term infant was extraordinarily rare and for most pediatricians, it was a disease they were unlikely to see in their practice lifetimes. Since 1990, there has been an increase in the number of reported cases of kernicterus in the United States.⁷ Thirty-one cases have been reported in term infants who were well at the time of hospital discharge, and several additional cases have occurred in near-term infants. Although it is unknown whether there is an actual increase in the incidence of kernicterus in the United States, reported cases are thought to be attributable to a variety of events including: shortened hospital stays with inconsistent follow-up beyond discharge; an increase in the frequency of breastfeeding; and a lack of concern about high bilirubin levels among pediatric care providers.

This study describes 6 infants ( 36 weeks' gestation) with bilirubin levels >25 mg/dL who were admitted to the newborn intensive care unit at the Children's Hospital of Philadelphia. All of the infants were breastfed and had been discharged within 48 hours of birth. Peak serum bilirubin levels ranged from 26.4 mg/dL (451 µmol/L) to 36.9 mg/dL (631 µmol/L). Although 3 infants had MRI studies initially read as abnormal, 4 had a follow-up MRI that was read as normal. Furthermore, neurodevelopmental evaluations performed between 3 and 24 months of life demonstrated complete resolution of clinical signs in all but 1 infant.

Five of the 6 infants in this study presented with neurologic signs previously described in infants with kernicterus including lethargy, arching, muscle tone abnormalities, and seizures.⁹ Although it is currently unknown what degree of kernicterus is reversible and what may be irreversible, our findings would encourage physicians to treat aggressively even in the face of signs such as arching and questionable seizures. All but 1 of our infants is ultimately developing normally and none of the infants exhibited sequelae associated with chronic post kernicteric bilirubin encephalopathy. The patient population reported by Maisels and Newman¹² was similar to ours in that all of the infants who developed kernicterus were otherwise healthy, breastfed term newborn infants. Unlike our series, however, the inclusion criteria of Maisels and Newman required all infants to have signs compatible with postkernicteric bilirubin encephalopathy. Furthermore, none of the infants in our study cohort had a peak serum bilirubin concentration in the range reported by those authors (39.0-49.7 mg/dL). Many of the signs of post kernicteric bilirubin encephalopathy can appear late, and it is unusual for athetosis to appear before 18 months old.^9,10 In addition, sensorineural hearing loss has been reported to develop outside of the neonatal period.¹³ Additional follow-up has not demonstrated abnormal neurologic findings in the infants in our study, although patient 3 demonstrates residual hearing loss. It is speculated that the less aggressive treatment of this infant with only phototherapy may have contributed to the persistent hearing loss.

It is reassuring that 4 of the infants who had normal follow-up neurodevelopmental examinations had a normal follow-up MRI with no evidence of kernicterus. Although previous studies have shown resolution of clinical signs and BAERs findings with aggressive, early treatment, ours is the first study to report changes in MRI studies from positive to negative. It is possible that the earlier MRI studies showing abnormalities in the basal ganglia represented transient neurologic dysfunction rather than permanent injury. Reversible disruption of brain metabolism has been described in experimental animals studied with nuclear magnetic resonance spectroscopy.¹⁴ Furthermore, studies by Nakamura et al¹⁵ have demonstrated reversibility of behavioral and hearing abnormalities in infants with high levels of serum bilirubin after exchange transfusion. It is unclear, however, how long bilirubin levels must remain persistently high for brain injury to become permanent. It is also possible that the disappearance of the abnormal MRI findings represented death and dissolution of the damaged cells. In that case, one would need to assume that the brain injury was below the level of clinical detection in our study infants. Because there have not been systematic investigations of infants with moderate degrees of hyperbilirubinemia using MRI, both hypotheses remain unproven. Moreover, although imaging studies have been performed previously on a small number of infants during the acute phase of bilirubin brain injury, the pathologic basis for the abnormal signal intensity on MRI remains unknown.¹⁶

The dose-response relationship between hyperbilirubinemia and brain injury in healthy term infants has been recently questioned.¹⁷ In a reevaluation of the data from the Collaborative Perinatal Project, Newman and Klebanoff¹⁰ could not demonstrate a relationship between bilirubin levels >26 mg/dL (>342 uM/L) and an abnormal neurologic examination. However, because more than half of these infants were treated with exchange transfusions, the lack of neurologic abnormalities may relate to the aggressive management of hyperbilirubinemia.³ Furthermore, when abnormal and suspicious neurologic findings were grouped together, a significant dose- response relationship was demonstrated. Such abnormalities may be more difficult to prevent with exchange transfusion. Brown and Maisels have recently recommended that a registry be developed for infants with total serum biliriubins >20 mg/dL (personal communication, conversation with Drs L. Johnson and A. K. Brown, Society for Pediatric Research Meeting, May 2000). Such a registry may be the only way to determine if subtle neurologic abnormalities occur in healthy term infants with modest elevations of serum bilirubin.

Although the American Academy of Pediatrics practice parameter guidelines suggest a return visit within 2 to 3 days for neonates discharged <48 hours after birth,⁴ the mean age of readmission in our study infants was 5 days. Breastfeeding was probably an important variable in these cases, as it has been for the other cases of kernicterus reported in the literature.^3,7,17 In a case-controlled study, Maisels⁸ recently found that hyperbilirubinemia was the major reason for hospital readmission during the first 2 weeks of life. Moreover, the majority of the jaundiced infants in their study were breastfed and had a greater mean weight loss since birth than nonjaundiced controls, suggesting that even mild degrees of dehydration in conjunction with breastfeeding impacts on the severity of hyperbilirubinemia.

	CONCLUSION

Top Abstract Methods Results Discussion Conclusion References

We have identified 6 breastfed term or near-term infants with marked hyperbilirubinemia who were managed aggressively. Despite neurologic dysfunction at the time of readmission, including abnormal clinical examinations, MRI studies, and BAERs, 5 of 6 of these infants subsequently demonstrated resolution of neurologic abnormalities and are developing normally, although 1 infant demonstrates residual hearing loss. In these cases, the findings of bilirubin encephalopathy and increased signal intensity in the basal ganglia on MRI may represent a transient phenomenon that may not correlate with long-term prognosis. We speculate that inadequate establishment of breastfeeding coupled with early discharge practices and, therefore, less opportunity for breastfeeding counseling, may play a role in the occurrence of marked hyperbilirubinemia in these infants. We further speculate that less aggressive treatment of these infants may contribute to residual neurologic abnormalities. Future studies should investigate the frequency of such findings in infants with moderate degrees of hyperbilirubinemia and possible modes of prevention.

	FOOTNOTES

Received for publication Apr 9, 1999; accepted Jun 19, 2000.

Address correspondence to Mary Catherine Harris, MD, Division of Neonatology, Department of Pediatrics, Children's Hospital of Philadelphia, 34th and Civic Center Blvd, Philadelphia, PA 19104

	ABBREVIATIONS

MRI, magnetic resonance imaging; T1WI, T1-weighted; TR, repetition time; TE, echo time; BAERs, brainstem auditory-evoked responses; MDI, mental developmental index; PDI, psychomotor developmental index.

	REFERENCES

Top Abstract Methods Results Discussion Conclusion References

1. Liu LL, Clemens CJ, Shay DK, Davis RL, Novack AH The safety of newborn early discharge: the Washington State experience. JAMA 1997; 278:293-298 [Abstract]
2. Maisels MJ, Newman TB Jaundice in full-term and near-term babies who leave the hospital within 36 hours. Clin Perinatol 1998; 25:295-302 [Medline]
3. Johnson L, Bhutani VK Guidelines for management of the jaundiced term and near-term Infant. Clin Perinatol 1998; 25:555-574 [Medline]
4. American Academy of Pediatrics, Provisional Committee for Quality Improvement and Subcommittee on Hyperbilirubinemia Practice parameter: management of hyperbilirubinemia in the healthy term newborn. Pediatrics 1994; 94:558-565 [Abstract/Free Full Text]
5. Dennery PA, Rhine WD, Stevenson DK Neonatal jaundice -what now? Clin Pediatr 1995; 34:103-107
6. Newman TB, Maisels MJ Evaluation and treatment of jaundice in the term newborn: a kinder and gentler approach. Pediatrics 1992; 89:809-818 [Abstract/Free Full Text]
7. Brown AK, Johnson L. Loss of concern about jaundice and the reemergence of kernicterus in full-term infants in the era of managed care. In: Faranoff AA, Klaus MH, eds. The Year Book of Neonatal and Perinatal Medicine. St Louis, MO: Mosby-Year Book; 1996:17-28
8. Maisels MJ, Kring E Length of stay, jaundice, and hospital readmission. Pediatrics 1998; 101:995-998 [Abstract/Free Full Text]
9. Volpe JJ. Neurology of the Newborn. Philadelphia, PA: WB Saunders; 1995:490-514
10. Newman TB, Klebanoff MA Neonatal hyperbilirubinemia and long-term outcome: another look at the Collaborative Perinatal Project. Pediatrics 1993; 92:651-657 [Abstract/Free Full Text]
11. Maisels MJ, Gifford K, Antle CE, Normal serum bilirubin levels in the newborn and the effect of breast feeding. Pediatrics 1986; 78:837-843 [Abstract/Free Full Text]
12. Maisels MJ, Newman TB Kernicterus in otherwise healthy, breastfed term newborns. Pediatrics 1995; 96:730-733 [Abstract/Free Full Text]
13. Worley G, Erwin CW, Goldstein RF, Delayed development of sensorineural hearing loss after neonatal hyperbilirubinemia: a case report with brain magnetic resonance imaging. Dev Med Child Neurol 1996; 38:271-278 [Medline]
14. Palmer C, Smith MB Assessing the risk of kernicterus using nuclear magnetic resonance. Clin Perinatol 1990; 17:307-329 [Medline]
15. Nakamura H, Satoshi T, Shimabuku R Auditory and brainstem responses in newborn infants with hyperbilirubinemia. Pediatrics 1985; 73:703-708
16. Gourley GR Bilirubin metabolism and kernicterus. Adv Pediatr 1997; 44:173-229 [Medline]
17. Newman TB, Maisels MJ Does hyperbilirubinemia damage the brain of full-term infants? Clin Perinatol 1990; 17:331-358 [Medline]