2009 the significance of measurement of serum unbound bilirubin

8/10/2019 2009 the Significance of Measurement of Serum Unbound Bilirubin

1/5

Original Article

The significance of measurement of serum unbound bilirubin

concentrations in high-risk infants

Yong-Kye Lee, Yasuyo Daito, Yoshinori Katayama, Hirotaka Minami and Hirokuni Negishi

Department of Pediatrics & Neonatology, Takatsuki General Hospital, Takatsuki, Japan

Abstract Background: In the management of neonatal hyperbilirubinemia, total bilirubin (TB) concentration is not specific

enough to predict the brain damage caused by bilirubin toxicity. Unbound bilirubin (UB) easily passes the bloodbrain

barrier and causes neurotoxicity. We aimed to evaluate whether serum UB concentration would be a useful predictor of

bilirubin encephalopathy in high-risk infants.

Methods: We measured the serum TB and UB concentrations of 388 newborn infants treated with phototherapy or

exchange transfusion for their hyperbilirubinemia at Takatsuki General Hospital between January 2002 and October

2003. Peak serum TB and UB levels and UB/TB ratios were studied on each birthweight group: below 1500 g (very low

birthweight), 1500 g2499 g (low birthweight), above 2500 g (normal birthweight); and several clinical factors influ-

encing hyperbilirubinemia were also studied.Results: Peak serum TB and UB levels increased with increasing birthweight, while UB/TB ratios decreased. The very

low birthweight group showed higher UB levels and UB/TB ratios despite lower TB levels in intraventricular hemor-

rhage or severe infection compared to those in the others. The low birthweight and normal birthweight groups showed

higher TB and UB levels in cases of hemolytic disease of the newborn compared to non-hemolytic disease of the

newborn cases. Eight of 44 cases showed high UB levels accompanied by abnormal auditory brainstem responses, one

of whom subsequently developed ataxic cerebral palsy with hearing loss, whereas the other seven showed transient

abnormalities of auditory brainstem responses by the treatment of exchange transfusion or phototherapy.

Conclusion: The UB measurement was considered to be significant for the assessment of the risk of bilirubin neuro-

toxicity and the appropriate intervention for hyperbilirubinemia in high-risk infants.

Key words auditory brainstem response (ABR), bilirubin encephalopathy, kernicterus, unbound bilirubin (UB), very low

birthweight (VLBW).

Introduction

The occurrence of kernicterus is very rare, but recently it has

been revealed by autopsy despite low bilirubin levels in prema-

ture infants. In the management of neonatal hyperbilirubinemia,

total bilirubin (TB) concentration is not specific enough to

predict the brain damage caused by bilirubin neurotoxicity.1 The

development of kernicterus is thought to be associated with

increased unbound bilirubin (UB) levels in the blood, the disrup-

tion of the bloodbrain barrier or increased acidosis in the braintissue.2 Therefore, this study was aimed to evaluate retrospec-

tively whether UB measurement would be significant for the

management of hyperbilirubinemia in high-risk infants.

Methods

A total of 388 newborn infants were treated with phototherapy or

exchange transfusion (ET) for hyperbilirubinemia at Takatsuki

General Hospital between January 2002 and October 2003. Peak

serum TB and UB levels and UB/TB ratios were studied on each

birthweight group of newborn infants; below 1500 g (very low

birthweight [VLBW]), 1500 g2499 g (low birthweight [LBW]),

and above 2500 g (normal birthweight [NBW]). Peak TB and UB

values were recorded in each infant during the treatment for

hyperbilirubinemia. Clinical factors influencing hyperbilirubine-

mia such as hemolytic disease of the newborn (HDN), respiratory

disorders with mechanical ventilation, intraventricular hemor-

rhage (IVH) and severe infection (pneumonia or sepsis) were

studied. Auditory brainstem responses (ABR) were also studied

in order to evaluate bilirubin neurotoxicity in case the TB level

exceeded 20 mg/dL or the UB level exceeded 0.81.0 mg/dL.

Measurement of TB and UB concentrations was performed

using a UB-Analyzer UA-2 (Arrows Co., Ltd, Osaka, Japan). The

principle of UB assay is that UB is rapidly oxidized by peroxi-

dase in the presence of hydrogen peroxide and changes to a

colorless compound.3 UB concentration is automatically deter-

mined by measuring the initial velocity of bilirubin oxidation

photometrically.4 ABR were examined using a Neuropack-2

(Nihon Kohden Corporation, Japan).

Correspondence: Yong-Kye Lee, MD PhD, 1-3-13 Kosobe, Takatsuki,

Osaka 569-1192, Japan. Email: [email protected]

Received 21 June 2007; revised 22 October 2008; accepted 18

March 2009.

Pediatrics International(2009) 51, 795799 doi: 10.1111/j.1442-200X.2009.02878.x

2009 Japan Pediatric Society


2/5

anovawas used to assess the difference between peak biliru-

bin levels in the three birthweight groups, and the Mann

Whitney U-test was used to assess the difference between

bilirubin variables in clinical factors influencing hyperbilirubine-

mia. StatView5.0 was used for the analyses.

Results

The clinical characteristics of subjects are shown in Table 1.

Respiratory disorders requiring mechanical ventilation were

complicated in 88% (76/86) of VLBW infants, 36% (66/183) of

LBW infants and 12% (14/119) of NBW infants. IVH and severe

infection were complicated in 15% (13/86) and 7% (6/86) of

VLBW infants, respectively. There were 24 HDN cases including

four LBW infants and 20 NBW infants.

Peak serum TB and UB levels increased with increasing birth-

weight, while UB/TB ratios decreased (Table 2). In the VLBW

group, 18 cases (21%) were complicated with IVH or severe

infection, and their UB levels and UB/TB ratios were statistically

higher compared to those of the others despite lower TB levels

(Table 3). In LBW and NBW groups, respiratory disorders had

little effect on TB or UB levels. In the LBW and NBW groups,

peak serum TB and UB levels of HDN were higher than those of

non-HDN, whereas UB/TB ratios showed no significant differ-

ence between the two groups. Among 24 HDN cases, one was

RhE-incompatible and the others were ABO-incompatible.

Forty-four cases (11%) showed high UB levels; 30.8mg/dL in

the VLBW group or 31.0mg/dL in the LBW and NBW groups

(Table 4). Fourteen cases in the VLBW group and 30cases in the

LBW and NBW groups showed high UB levels. ABR were

examined during therapy against 10 cases in the VLBWgroup and

26 cases in the LBW and NBW groups. ABR findings showed no

response in two VLBW infants (Case-1 and Case-2), poorresponse in a LBW infant, and the prolonged latency of Wave-I in

two VLBW infants, an LBW infant and two NBW infants. The

other 28 infants were within normal limits of ABR.

These eight cases had abnormal ABR findings and

their UB values all exceeded 1.0mg/dL. Firstly, Case-1

(GA:25w, BW:705 g) developed persistent hyperbilirubinemia

(TB:12.6 mg/dL, UB:1.45mg/dL) associated with late-onset

pneumonia on day 15, and a high UB level exceeding 1.0 mg/dL

prolonged for more than 48 h. She slowly improved by three-

day of intensive phototherapy along with replenishment of

immunoglobulin and administration of antibiotics (piperacillin

+ amikacin). Although she did not deteriorate neurologically

during the intensive phototherapy, her ABR showed no response

at 5 months after birth and she subsequently developed ataxic

cerebral palsy with hearing loss. Her magnetic resonance

images showed the cerebellar hypoplasia and the atrophy of the

basal ganglia and the cerebral white matter at 15 months of life.

Secondly, Case-2 (GA:30w, BW:1468 g) developed hyperbiliru-

binemia (TB:15.3 mg/dL, UB:1.51 mg/dL) with no response of

ABR on day 4. Immediately thereafter, when he was treated

with ET, he showed no abnormal neurological signs, and his

ABR improved to be within normal on day 36 (Fig. 1). Lastly,

the other six infants showed transient abnormalities of ABR and

Table 1 Clinical characteristics of subjects

BW group VLBW LBW NBW

n = 86 n = 183 n = 119

Male/Female 45/41 105/78 73/46

GA(weeks), mean (SD) 28.3 (3.1) 34.3 (2.2) 38.6 (1.9)

BW(g), mean (SD) 1075 (298) 2003 (272) 3071 (374)

Apgar 1 min, median (range) 7 (19) 9 (110) 9 (410)Respiratory Disorders 76 (88%) 66 (36%) 14 (12%)

IVH 13 (15%) 1 (0.5%) 1 (0.8%)

Infection 6 (7%) 0 0

HDN 0 4 (2%) 20 (17%)

BW, birthweight; GA, gestational age; HDN, hemolytic disease of the newborn; IVH, intraventricular hemorrhage; LBW, low birthweight; NBW,

normal birthweight; VLBW, very low birthweight.

Table 2 Peak bilirubin levels in the three BW groups

BW group VLBW LBW NBW

n = 86 n = 183 n = 119

Peak bilirubin levels

Age (days) 4 (135) 4 (220) 4 (19)

TB (mg/dL) 8.2 (2.115.4) 12.7 (8.323.7) 17.1 (10.828.9)

UB (mg/dL) 0.47 (0.081.79) 0.60 (0.241.73) 0.74 (0.361.96)

UB/TB (10-4) 0.60 (0.211.77) 0.47 (0.201.02) 0.44 (0.210.91)

Data are expressed as median (range). TB and UB levels and UB/TB ratios also had significant differences among the three BW groups assessed

using anova. BW, birthweight; LBW, low birthweight; NBW, normal birthweight; TB, total bilirubin; UB, unbound bilirubin; VLBW, very low

birthweight.

796 Y-K Leeet al.



3/5

were treated with intensive phototherapy, improving without

apparent neurological signs.

Discussion

Albumin-unbound bilirubin; UB easily penetrates the blood

brain barrier and causes neurotoxicity. Albumin-bound bilirubin

enters the brain through the bloodbrain barrier disrupted by

hypoxemia, hypercarbia, hyperosmolarity and septicemia. It has

been shown in vitro and in vivo that bilirubin toxicity uniquely

derives from UB.5,6 Diamond & Schmid revealed that the biliru-

bin toxic to the central nervous system was UB; albumin-

unbound bilirubin based on the finding that bilirubin dissolved insaline, was easily transferred to the brain, while that added to

human serum albumin was not when it was infused to the jugular

vein of newborn guinea pigs.6 In the full-term infant with marked

hyperbilirubinemia secondary to hemolytic disease, a clear cor-

relation can be discerned between the occurrence of kernicterus

and the recorded maximal level of serum bilirubin, whereas ker-

nicterus without marked hyperbilirubinemia in the premature

infant has been demonstrated at autopsy.7

The automated method using the UB-Analyzer correlates with

the conventional manual method extremely well (correlation

coefficient = 0.974) and its coefficient of variation by repeated

measures showed good reproducibility of 2.25% in TB and

3.27% in UB.4 The limitation of the UB-Analyzer is that it

occasionally shows an apparently higher value when strong

hemolysis is visible or direct bilirubin exceeds 2 mg/dL in the

specimen, but an apparently lower value when a large quantity of

vitamin C exists in the reaction system; its antioxidative effect

interferes with bilirubin oxidation by peroxidase. In cases with

cholestasis or bronze baby syndrome, the UB value shifts higher,

influenced by increased conjugated bilirubin or accumulation of

the photoisomer, EZ-cyclobilirubin.8 It has also been reported

that standard sample dilution using a UB-Analyzer significantlydecreased the UB concentration reading compared to that of

undiluted serum, therefore accurate UB measurement required

minimal sample dilution.9

Human albumin has a single, tight, high affinity primary

binding site for bilirubin and one or more (probably two),

weaker, lower affinity secondary binding sites. In vitro study,

when the bilirubin/albumin (B/A) molar ratio is less than 1, UB

concentration linearly correlates with TB concentration. And

when the B/A molar ratio is more than 1, UB concentration

Table 3 Clinical factors influencing hyperbilirubinemia

VLBW group IVH or Infection non IVH & Infection P

n = 18 n = 68

Peak bilirubin levels

Age (days) 6 (215) 4 (135)


4/5

increases more rapidly compared to the increasing TB concen-

tration. Wennberget al.reported that the bilirubin-binding affin-

ity was 7 107 L/M in sick infants and 25 107 L/M in normal

infants.5 Recently, Bender et al. reported that the bilirubin-

binding capacity was larger in the low-risk group (20.8 1 4.6 mg/

dL) than in the moderate-risk group (17.8 1 3.5 mg/dL) and the

high-risk group (17.3 1 3.4 mg/dL), that was, the bilirubin-binding capacity was smaller in unstable infants than stable

infants, whereas bilirubin binding affinity showed no difference

depending upon clinical risk status or gestational age.10 Theoreti-

cally, the UB/TB ratio might be proportional to serum albumin

concentration, and influenced by endogenous and exogenous

competitors such as free fatty acids and drugs in high-risk infants.

Therefore, the UB/TB ratio is assumed to be one of the bilirubin

binding variables, and a high UB level with a high UB/TB ratio

potentially shows an extremely high risk of acute bilirubin

encephalopathy.

In this study, we compared peak serum TB and UB levels and

UB/TB ratios by each birthweight group and investigated various

clinical factors influencing hyperbilirubinemia such as HDN,

respiratory disorders, IVH and severe infection. Peak serum TB

and UB levels increased with increasing birthweight, while

UB/TB ratios decreased. Hence, it was assumed that with

decreasing birthweight, the serum albumin level decreased, the

UB levels relative to TB levels were high, and UB/TB ratios

increased. Furthermore, peak serum TB and UB levels in HDN

cases were significantly higher compared to those in non-HDN

cases, whereas UB/TB ratios showed no significant difference

between them. The measurement of UB levels was not affected

unless strong hemolysis with serum hemoglobin of more than

1.0 g/dL was visible in the experiment with hemolysate added.4

Although serum TB and UB levels rapidly increase as early-onsetjaundice in HDN, UB/TB ratios might show no significant dif-

ference from those in non-HDN cases, usually at the B/A molar

ratio


5/5

dystonia, and athetosis as extrapyramidal disturbance, hearing

loss, gaze palsy, and enamel dysplasia of deciduous teeth. Iso-

lated kernicterus shows symptoms limited to a single system.

Strictly, isolated kernicterus is not generally observed, but is

classified into mixed cases: auditory predominant or motor pre-

dominant. Preliminary evidence suggests that auditory predomi-

nant kernicterus is more likely to occur in premature infants.15

Hence, further studies of clinical symptoms, ABR findings, mag-

netic resonance images and long-term follow up of neonatal

hyperbilirubinemia in high-risk infants are essential. As for ker-

nicterus in VLBW infants in particular, of which neurological

symptoms are scarcely demonstrated, large cohort studies and

precise examinations for prevention of it are necessary.

Conclusion

Peak serum TB and UB levels increased with increasing birth-

weight, while UB/TB ratios decreased. The VLBW group

showed higher UB levels and UB/TB ratios despite lower TB

levels in IVH cases or severe infection compared to those in the

others. LBW and NBW groups showed higher TB and UB levels

in HDN cases compared to those in non-HDN cases. Eight of the

44 cases had high UB levels exceeding 1.0 mg/dL accompanied

by abnormal ABR. One of these patients subsequently developed

ataxic cerebral palsy with hearing loss, whereas the other seven

showed transient abnormalities of ABR by the treatment of ET or

intensive phototherapy. In conclusion, the UB measurement was

considered to be significant for the assessment of the risk of

bilirubin neurotoxicity and the appropriate intervention for

hyperbilirubinemia in high-risk infants.

Acknowledgment

We are grateful to Hajime Nakamura, Emeritus Professor of

Kobe University Graduate School of Medicine, for helpful com-

ments, who originally devised the peroxidase and glucose

oxidase method in UB determinations.

References

1 Nakamura H, Yonetani M, Uetani Y, Funato M, Lee Y. Determi-

nation of serum unbound bilirubin for prediction of kernicterus in

low birthweight infants. Acta. Paediatr. Jpn. 1992; 34: 6427.

2 Bratlid D. How bilirubin gets into the brain.Clin. Perinatol. 1990;

17: 44965.

3 Nakamura H, Lee Y. Microdetermination of unbound bilirubin in

icteric newborn sera: An enzymatic method employing peroxidase

and glucose oxidase. Clin. Chim. Acta 1977; 79: 41117.

4 Shimabuku R, Nakamura H. Total and unbound bilirubin determi-

nation using an automated peroxidase micromethod.Kobe J. Med.

Sci. 1982; 28: 91104.

5 Wennberg RP, Ahlfors CE, Rasmussen LF. The pathochemistry of

kernicterus. Early Human Development 1979; 35372.

6 Diamond I, Schmid R. Experimental bilirubin encephalopathy. The

mode of entry of bilirubin-14C into the central nervous system.J. Clin. Invest. 1966; 67889.

7 Volpe JJ. Bilirubin and brain injury. In Volpe JJ ed. Neurology of

the Nwborn, 5th edn. Saunders, Philadelphia, PA, 2008; 61951.

8 Itoh S, Kawada K, Kusaka Tet al.Influence of glucuronosyl biliru-

bin and (EZ)-cyclobilirubin on determination of serum unbound

bilirubin by UB-analyser.Ann. Clin. Biochem. 2002; 39: 5838.

9 Ahlfors CE, Vreman HJ, Wong RJet al.Effects of sample dilution,

peroxidase concentration, and chloride ion on the measurement of

unbound bilirubin in premature newborns. Clin. Biochem. 2007;

40: 2627.

10 Bender GJ, Cashore WJ, Oh W. Ontogeny of bilirubin-binding

capacity and the effect of clinical status in premature infants born

at less than 1300 grams. Pediatrics2007; 120: 106773.

11 Nakamura H, Takada S, Shimabuku R, Matsuo M, Matsuo T,

Negishi H. Auditory brainstem responses in newborn infants with

hyperbilirubinemia. Pediatrics1985; 75: 7038.

12 Funato M, Tamai H, Shimada S, Nakamura H. Vigintiphobia,

unbound bilirubin, and auditory brainstem responses. Pediatrics

1994; 93: 503.

13 Amin SB, Ahlfors CE, Orlando MS, Dalzell LE, Merle KS, Guillet

R. Bilirubin and serial auditory brainstem responses in premature

infants. Pediatrics 2001; 107: 66470.

14 Rapin I, Gravel J. Auditory neuropathy: Physiologic and patho-

logic evidence calls for more diagnostic specificity. Int. J. Pediatr.

Otorhinolaryngol. 2003; 67: 70728.

15 Shapiro SM, Bhutani VK, Johnson L. Hyperbilirubinemia and

kernicterus. Clin. Perinatol. 2006; 33: 387410.

Unbound bilirubin in high-risk infants 799


2009 the significance of measurement of serum unbound bilirubin

Documents