Phototherapy for Neonatal

Jaundice—Therapeutic Effects on More Than One Level?

Thor Willy Ruud Hansen, MD, PhD*,†

Phototherapy for jaundice is a common treatment in neonatal medicine and is used to

prevent the neurotoxic effects of bilirubin. Studies have assessed the optimal wavelength
of phototherapy light, the importance of irradiance and spectral power, and the types of
light source, including the use of single versus multiple light sources. Outcome measures
have been duration of need for phototherapy or rate of reduction of serum bilirubin over a
given time. An apparent resurgence of kernicterus in recent years has forced us to focus on
the emergency management of severely jaundiced infants. Several studies have shown that
very rapid reductions of total serum bilirubin levels are possible. The speed with which
photoisomers are formed appears to be important both from this perspective and theoret-
ically may also be neuroprotective because of the more polar nature of the photoisomers.
This work reviews the evidence concerning the speed of photoisomer formation, as well as
the evidence regarding the relative neurotoxicity of bilirubin isomers.
Semin Perinatol 34:231-234 © 2010 Elsevier Inc. All rights reserved.

KEYWORDS bilirubin, bilirubin encephalopathy, bilirubin photoisomers, blood– brain barrier,

kernicterus, neurotoxicity, phototherapy

J aundice is very common in the neonatal period and has

been discussed in the medical literature for centuries.1 Al-
though a multitude of treatments were applied in the past,
the association between neonatal jaundice and brain damage
(kernicterus) appears not to have been made until the 19th
century.2-4 Truly rational and efficacious treatment of neona-
tal jaundice only became possible after the discovery of blood
groups in the 20th century, when exchange transfusion be-
came a viable procedure.5 Follow-up of infants who had un-
dergone exchange transfusions for erythroblastosis fetalis re-
vealed that the incidence of kernicterus was significantly
lower in these than in their untreated peers.6,7 Pursuant to
this discovery, exchange transfusions began to be used to
treat severe jaundice regardless of its etiology, and guidelines
were introduced on the basis of the recognition that ker-
nicterus was mostly seen in infants with total serum bilirubin
(TSB) levels ⬎ 20 mg/dL (⬇340 ␮mol/L).8
*Neonatal Intensive Care Unit, Women’s and Children’s Clinic, Oslo Uni-
versity Hospital-Rikshospitalet, Oslo, Norway.
†Institute of Clinical Medicine, Faculty of Medicine, University of Oslo,
Oslo, Norway.
Address reprint requests to Thor Willy Ruud Hansen, MD, PhD, Neonatal
Intensive Care Unit, Women’s and Children’s Clinic, Oslo University Hospital-
Rikshospitalet, N-0027 Oslo, Norway. E-mail: t.w.r.hansen@medisin.
uio.no
The story of the discovery of phototherapy provides a fas-
cinating insight into how serendipity may play a significant
role in scientific progress. Sister Ward in the premature baby
unit at the Rochford General Hospital in Essex, England, had
taken a jaundiced baby for a stroll in the sun and pointed out
to the pediatricians that an area of skin that had been covered
by clothing was visibly more yellow than the adjacent skin.9
A little later in the same hospital, the TSB value in a tube of
blood accidentally exposed to sunlight was noted to be much
lower than the preexposure value.10 The impression that light
could “bleach” bilirubin was subsequently strengthened by
further studies. Although photooxidation (“bleaching”) of
bilirubin does occur, it was later shown that this mechanism
contributes little to the clinical effect of phototherapy.11
After large trials in the 1960s and 1970s,12 phototherapy
for jaundice has arguably become the most commonly used
treatment for neonatal jaundice. We use this treatment to
prevent the neurotoxic effects of bilirubin— kernicterus.
Phototherapy has been studied extensively, leading to char-
acterization of its mechanisms of action. This consists of the
formation of polar isomers of bilirubin which can be excreted
in bile and urine without the need for conjugation.11 In the
clinical arena, studies of phototherapy have focused on the
effects of this therapy on TSB. Thus, researchers have asked
which factors are associated with optimal effects as far as
reducing or curtailing the rise of serum bilirubin values. To

0146-0005/10/$-see front matter © 2010 Elsevier Inc. All rights reserved. 231
doi:10.1053/j.semperi.2010.02.008
232
this end, studies have addressed the optimal wavelength of
photoherapy light, the importance of irradiance and spectral
power, and the types of light source, including the use of
single versus multiple light sources.11 Outcome measures
have typically been the duration of need for phototherapy or
the rate of reduction of serum bilirubin over a given time.
Treatment guidelines for neonatal jaundice have primarily
focused on the levels of serum bilirubin which indicate inter-
vention or cessation of the same.13
McDonagh and Lightner14 suggested almost a quarter of a
century ago that the immediate effect of phototherapy—
transforming the bilirubin molecule to a presumably less
toxic form— might be beneficial and deserved more recog-
nition. Very limited work has been done to follow up on this
suggestion, and the issue appears to be rarely discussed in
textbooks and guidelines, perhaps because of the dearth of
solid data. In their review, Maisels and McDonagh11 have
raised this issue again. Entry of bilirubin into brain is a sine
qua non for its neurotoxic effects, and occurs because biliru-
bin IX␣(z,z) is lipid soluble. Polar molecules, by contrast,
need transporters in the blood– brain barrier to gain access to
the neurons. Based on the physical characteristics of the bil-
irubin photoisomers, it may be hypothesized that these iso-
mers are less able to cross the blood– brain barrier.
An apparent resurgence of kernicterus in recent years
forces us to focus on the initial, emergency management of
severely jaundiced infants.15 Until now, the focus of manage-
ment has been on how to reduce the serum bilirubin levels as
quickly as possible. However, if bilirubin photoisomers do
not gain access to the brain, rapid isomerization may be brain
sparing even before changes in TSB can be detected. This
paper reviews the evidence for reversibility of acute bilirubin
encephalopathy, and discusses the possible role of aggressive
phototherapy and rapid bilirubin photoisomer formation in
this context.
Toxicity of Bilirubin
Yellow discoloration of the brain in infants dying with jaun-
dice may have been first described in 1847 by Hervieux.2,16 A
stronger staining of the brain nuclei was described by Orth,3
and the term kernicterus (jaundice of the nuclei) was coined
by Schmorl4 to distinguish this type of staining pattern from
the more diffuse yellow discoloration of the brain which he
observed in most of the infants he autopsied. With magnetic
resonance imaging, changes in the brain nuclei can now be
demonstrated in living infants with clinical evidence of ker-
nicterus.17,18
Toxicity in survivors with such brain changes is evidenced
by varying degrees of choreoathetosis, gaze palsy, sensori-
neural deafness, and (in some) developmental delays.19 Bili-
rubin entry into brain is sine qua none for such toxicity; the
issue extensively studied in animal models and reviewed.20
These studies have shown that bilirubin can pass through an
intact blood– brain barrier, although to a lesser extent than
might have been expected given the solubility characteristics
of the main isomeric form of bilirubin IX␣ (z,z), which be-
haves like a lipophilic molecule. Watchko et al21 have shown
T.W.R. Hansen
that membrane pumps, such as P-glycoprotein, may serve to
remove bilirubin from the central nervous system, and this
may perhaps explain why the brain-to-serum ratio of biliru-
bin is much lower than might have been expected. Increased
entry of bilirubin into brain is observed in increased brain
blood flow after osmotic opening of the blood– brain barrier,
and in the presence of substances which displace bilirubin
from its binding to albumin.20
Bilirubin, which is bound to albumin (which is true of
⬎99% of bilirubin circulating in serum), does not enter the
brain. Although this finding, as well as in vitro studies of
bilirubin toxicity, has led to the understanding that albumin-
bound bilirubin is nontoxic, it must be remembered that
there is a constant equilibration of unbound bilirubin in se-
rum with bilirubin bound to albumin and that this equilib-
rium can be shifted by a change in pH to acidotic range and
presence of binding competitors. Conjugated bilirubin (ie,
bound to glucuronic acid) is also most likely nontoxic, but
whether this is because it is water-soluble, is not absolutely
clear.
Toxicity of Bilirubin Photoproducts
There is limited evidence regarding the toxicity of bilirubin
photoproducts. It is noteworthy that several different authors
present findings that suggest that the photoproducts of bili-
rubin may be less toxic. However, there are significant meth-
odological concerns associated with these studies. Thus, Sil-
berberg et al22 compared the toxicity of irradiated versus
nonirradiated bilirubin in cultures of cerebellar cells and
found the irradiated bilirubin solutions to be less toxic. How-
ever, it seems likely that given the procedure used for irradi-
ation, little of what was left is likely to have been present as
photoisomers. A similar caveat is likely to apply to the find-
ings of Thaler,23 Porto,24 and Kaul et al,25 who used bilirubin
isolated from the urine of jaundiced infants undergoing pho-
totherapy to test effects on red cell membranes. Although the
photoproducts thus isolated may contain some isomers, the
main constituent is likely to be photooxidation products,11
thus the interpretation of these results remains uncertain.
More data from Calligaris et al26 also show less toxicity of
light-irradiated bilirubin in cell cultures, although the impli-
cations of this finding are also limited by lacking documen-
tation of what the actual photoproducts were. Further, the
change in toxicity paralleled the change in unbound bilirubin
in the solution, and thus may not be directly related to the
photoproducts per se. In this context it should be noted that
phototherapy appears to reduce serum unbound bilirubin
more than TSB, reducing the UB:TSB ratio,27 and conse-
quently, the driving force for bilirubin to enter the cells.
Mechanism of Phototherapy
Bilirubin absorbs light most strongly in the blue region of the
spectrum near 460-nm wavelength. Phototherapy in the
jaundiced newborn is dependent on how far into the skin
light penetrates because only light that penetrates will reach
bilirubin molecules which can be converted. Light penetra-
Phototherapy for neonatal jaundice
tion increases with wavelength, thus phototherapy lights
with energy maxima in the 460-490 nm range (ie, blue light)
are most effective.11
When light is absorbed by the bilirubin IX␣ (z,z) isomer,
which predominates in the serum of jaundiced newborns,
the molecule rearranges to form structural (lumirubin) or
configurational (z,e) isomers. Although configurational
isomerization is very rapid and reversible, structural isomer-
ization is slower but irreversible. Both of these isomeric forms
are more polar than the parent compound and thus can be
excreted unconjugated in bile, whereas lumirubin can also be
excreted in urine. Formation of photooxidation products is a
slower process, and while these products are excreted in urine,
they are believed to contribute less to bilirubin elimination than
isomerization. For an in-depth discussion of phototherapy,
please see the review by Maisels and McDonagh.11
During the early trials of phototherapy, reduction in the
number of exchange transfusions was an important outcome
measure. In current clinical use the effect of phototherapy is
typically measured as rate of decline in TSB values per unit of
time. However, it should be remembered that in a situation in
which TSB values are expected to increase rapidly, stabiliza-
tion or reduction in the rate of rise may be very valid as a
measure of effect. It is in this kind of situation that the rate of
formation of photoisomers may be of potential therapeutic
value. The commonly used laboratory methods for clinical
bilirubin analysis do not distinguish between the isomers,
although they typically do not include the photooxidation
products. In other words, the laboratory TSB value includes
the sum of bilirubin IX␣ (z,z), lumirubin, and bilirubin IX␣
(z,e) concentrations. If, as might be expected based on the
physicochemical characteristics of the bilirubin photoiso-
mers, these are less prone to cross biological membranes, the
presence of 20%-25% of these isomers versus 3% to 4% of TSB
in the untreated infant, could translate into less driving force for
bilirubin IX␣ (z,z) to enter the brain. However, there is as yet no
experimental proof to back up this speculation.
Rapid Effect of
Intensive Phototherapy
The rate of decrease of TSB during phototherapy will depend on
the spectral power delivered, the peak wavelength of light, the
mechanism(s) which caused the jaundice, the TSB level, and the
postnatal age of the infant. With spectral irradiance greater than
10 ␮W/cm2/nm, a reduction of TSB from 20% to 50% during
the first 24 hours may be expected.28 In extreme jaundice
(⬎500 ␮mol/L), reductions of 85 ␮mol/L/h (5 mg/dL/h) have
been documented during the first phase of treatment.29
The rate of photoisomer formation has been studied by
Onishi et al30 and Myara et al.31 Onishi et al found 17% to 18%
photoisomers after 2 hours of phototherapy, whereas Myara et al
found around 40% photoisomers after 3 hours. However, they
found around 17% to 18% photoisomers present even before
phototherapy was started, whereas Onishi et al found close to
12%. Although it is possible that the high values at time zero
may to some extent reflect bright nursery lights, it also raises
233
questions about the efficacy of their shielding of samples
during collection of processing.
Mreihil et al32 studied early photoisomer formation during
intensive phototherapy (⬇25-30 ␮W/cm2/nm). The percent-
age of photoisomers at time zero was around 4%, increasing
to about 10% at 15 minutes and 20% at 2 hours, at which
time TSB values were still not significantly reduced compared
with starting levels. Care was taken to shield samples from
light during collection, storage, and processing. Also, in the
current Newborn Individualized Developmental Care and
Assessment Program-influenced age, nursery light may have
been more dimmed than was common previously. The im-
portance of this observation is that photoisomer formation
starts almost as soon as the lights are turned on, and long
before changes in TSB can be detected.
Reversibility of Acute
Intermediate to Advanced
Stage Bilirubin Encephalopathy
Some central nervous system effects of bilirubin in the jaun-
diced newborn are reversible. Thus, infants who are lethargic
perk up when TSB is lowered. Changes in the brainstem
auditory response are normalized following treatment of
jaundice.33 However, in infants who exhibit signs of interme-
diate-to-advanced bilirubin encephalopathy, reversibility has
been thought to be improbable.19 Following the apparent
increase in reported cases of kernicterus in recent years, pro-
posals have been made for a “crash-cart approach” to severely
jaundiced infants.34 Intensive phototherapy should be part of
such rapid intervention.
Some reports have described infants who exhibited inter-
mediate-to-advanced acute stage bilirubin encephalopathy
yet were normal on follow-up.35,36 In the 6 cases reported by
Hansen et al,36 rapid institution of therapy, including photo-
therapy, seemed to have been a common theme. In one of the
infants, who exhibited opisthotonos and retrocollis, high-
performance liquid chromatography of serum had been done
in an attempt to resolve a discrepancy between the results of
TSB analyses with 2 different methods (cooximetry vs diazo
technique).37 After 5 hours of phototherapy, TSB had been
reduced by about 25%, whereas the more polar photo-iso-
mers of bilirubin constituted approximately 30% of the re-
maining bilirubin in serum. Thus, in the course of 5 hours of
intensive phototherapy the concentration of neurotoxic bili-
rubin IX␣ (z,z) had been reduced by almost 50%.
Conclusions
A rapid response to phototherapy has been documented,
particularly in situations with extreme neonatal jaundice. Re-
versal of acute intermediate-to-advanced phase bilirubin en-
cephalopathy has also been described, and phototherapy was
one of the tools used in these cases. Whether early photoi-
somerization plays a role in such reversal or could influence
blood-to-brain bilirubin transfer or bilirubin neurotoxicity is
a theoretically attractive hypothesis, which has yet to be stud-
234
ied with adequate methods. Such studies, although challeng-
ing to design, would appear to be well worth attempting.
However, even the absence of such evidence does not consti-
tute a viable argument against early and effective photother-
apy in severely jaundiced infants threatened by bilirubin neu-
rotoxicity.
References
1. Metlinger B: Ein Regiment der Jungen Kinder. Augsburg, Günther
Zainer, 1473
2. Hervieux JFE: De l’ictère des Nouveau-nés. Paris, France, Thèse, 1847
3. Orth J: Ueber das Vorkommen von Bilirubinkrystallen bei neuge-
borenen Kindern. Virchows Arch Pathol Anat 63:447-462, 1875
4. Schmorl C: Zur Kenntnis des Ikterus neonatorum, insbesondere der
dabei auftretenden Gehirnveränderungen. Verh Dtsch Pathol Ges
6:109-115, 1904
5. Diamond LK, Allen FH, Thomas WO: Erythroblastosis fetalis. VII.
Treatment with exchange transfusion. N Engl J Med 244:39-49, 1951
6. Hsia DY, Allen F, Gellis SS, et al: Erythroblastosis fetalis. VIII. Preven-
tion of kernicterus. N Engl J Med 246:668-671, 1952
7. Mollison PL, Cutbush M: Haemolytic disease of the newborn, in Gaird-
ner D (ed): Recent Advances in Pediatrics. New York, NY, P Blakiston
and Son, 1954, pp 110-132
8. Watchko JF, Oski F: Bilirubin 20 mg/dl ⫽ Vigintifobia. Pediatrics 71:
660-663, 1983
9. Dobbs RH, Cremer RJ: Phototherapy. Arch Dis Child 50:833-836,
1975
10. Cremer RJ, Perryman PW, Richards DH: Influence of light on the hy-
perbilirubinaemia of infants. Lancet 1:1094-1097, 1958
11. Maisels MJ, McDonagh AF: Phototherapy for neonatal jaundice. N Engl
J Med 358:920-928, 2008
12. Brown AK, Kim MH, Wu PY, et al: Efficacy of phototherapy in preven-
tion and management of neonatal hyperbilirubinemia. Pediatrics 75:
393-400, 1985
13. AAP: Subcommittee on hyperbilirubinemia: Management of hyperbil-
irubinemia in the newborn infant 35 or more weeks of gestation. Pedi-
atrics 114:297-316, 2004
14. McDonagh AF, Lightner DA: Like a shriveled blood orange. Bilirubin,
jaundice and phototherapy. Pediatrics 75:443-455, 1985
15. Ebbesen F: Recurrence of kernicterus in term and near-term infants in
Denmark. Acta Paediatr 89:1213-1217, 2000
16. Hansen TWR: Pioneers in the scientific study of neonatal jaundice and
kernicterus. Pediatrics 106:e15, 2000
17. Penn AA, Enzmann DR, Hahn JS, et al: Kernicterus in a full term infant.
Pediatrics 93:1003-1006, 1994
18. Yokochi K: Magnetic resonance imaging in children with kernicterus.
Acta Paediatr 84:937-939, 1995
19. Volpe JJ: Neurology of the Newborn (ed 5). Philadelphia, PA, Elsevier,
2008
T.W.R. Hansen
20. Hansen TWR: The pathophysiology of bilirubin toxicity, in Maisels MJ
and Watchko JF (eds): Neonatal Jaundice. London, United Kingdom,
Harwood Academic Publishers, 2000, pp 89-104
21. Watchko JF, Daood MJ, Hansen TWR: Brain bilirubin content is in-
creased in P-glycoprotein deficient transgenic null mutant mice. Pedi-
atr Rev 44:763-766, 1998
22. Silberberg DH, Johnson L, Schutta H, et al: Effects of photodegradation
products of bilirubin on myelinating cerebellum cultures. J Pediatr
77:613-618, 1970
23. Thaler MM: Toxic effects of bilirubin and its photodecomposition
products. Birth Defects Orig Arts Series 6:128-130, 1970
24. Porto SO: In vitro and in vivo studies on the effect of phototherapy
upon bilirubin. Birth Defects Orig Arts Series 6:83-89, 1970
25. Kaul R, Bajpai VK, Shipstone AC, et al: Bilirubin-induced erythrocyte
membrane cytotoxicity. Exp Mol Pathol 34:290-298, 1981
26. Calligaris SD, Bellarosa C, Giraudi P, et al: Cytotoxicity is predicted by
unbound and not total bilirubin concentration. Pediatr Rev 62:576-
580, 2007
27. Nakamura H, Lee Y, Uetani Y, et al: Effects of phototherapy on serum
unbound bilirubin in icteric newborn infants. Biol Neonate 39:295-
299, 1981
28. Tan KL: The pattern of bilirubin response to phototherapy for neonatal
hyperbilirubinemia. Pediatr Rev 16:670-674, 1982
29. Hansen TWR: Acute management of extreme neonatal jaundice—The
potential benefits of intensified phototherapy and interruption of en-
terohepatic bilirubin circulation. Acta Paediatr 86:843-846, 1997
30. Onishi S, Isobe K, Itoh S, et al: Metabolism of bilirubin and its photoi-
somers in newborn infants during phototherapy. J Biochem 100:789-
795, 1986
31. Myara A, Sender A, Valette V, et al: Early changes in cutaneous bilirubin
and serum bilirubin isomers during intensive phototherapy of jaun-
diced neonates with blue and green light. Biol Neonate 71:75-82, 1997
32. Mreihil K, McDonagh A, Nakstad B, et al: Early isomerization of bili-
rubin in intensive phototherapy of neonatal jaundice. Pediatr Acad Soc
Baltimore MD USA, May, 2009
33. Tan KL, Skurr BA, Yip YY: Phototherapy and the brain-stem auditory
evoked response in neonatal hyperbilirubinemia. J Pediatr 120:306-
308, 1992
34. Johnson L, Brown AK, Bhutani VK: System-based approach to manage-
ment of neonatal jaundice and prevention of kernicterus. J Pediatr
140:396-403, 2002
35. Harris MC, Bernbaum JC, Polin JR, et al: Developmental follow-up of
breastfed term and near-term infants with marked hyperbilirubinemia.
Pediatrics 107:1075-1080, 2001
36. Hansen TWR, Nietsch L, Norman E, et al: Apparent reversibility of
acute intermediate phase bilirubin encephalopathy. Acta Paediatr 98:
1689-1694, 2009
37. McDonagh AF: Ex uno plures: The concealed complexity of bilirubin
species in neonatal blood samples. Pediatrics 118:1185-1187, 2006