Hyperbilirubinemia?
WHAT’S KNOWN ON THIS SUBJECT: Hyperbilirubinemia may be a risk factor for adverse neurodevelopmental outcomes for preterm infants. Ibuprofen can displace bilirubin from albumin binding in vitro. The subsequent increase in UB levels may enhance the risk of kernicterus and encephalopathy.
WHAT THIS STUDY ADDS: Ibuprofen was significantly associated with higher TSB levels in preterm infants. Ibuprofen may be linked to increased neurologic risk not only through direct albumin-bilirubin displacement but also through competition with the enzyme responsible for bilirubin conjugation.
abstract
OBJECTIVE:The aim of this study was to investigate whether ibuprofen exposure was associated with increased hyperbilirubinemia in pre-term infants.
METHODS:Since 2000, ibuprofen has been administered to all infants at⬍30 weeks of gestation who are admitted to our unit, to prevent patent ductus arteriosus. We retrospectively compared data for 418 infants subjected to ibuprofen prophylaxis (2000 –2007) and 288 in-fants not exposed to ibuprofen (1993–1999).
RESULTS:The ibuprofen group had a significantly higher peak total serum bilirubin level (9.0⫾2.5 mg/dL vs 7.3⫾3.3 mg/dL), more need for phototherapy (398 infants [95%] vs 254 infants [87.6%]), and a longer phototherapy duration (94.3⫾43.6 hours vs 87.2⫾38.6 hours). Groups did not differ with respect to gestational age, birth weight, gender ratio, glucose-6-phosphate dehydrogenase deficiency inci-dence, or hypoalbuminemia (⬍2.5 g/dL) incidence. Hemolytic isoimmu-nization was diagnosed with similar incidences (no-ibuprofen group: 7 of 288 infants; ibuprofen group: 8 of 418 infants). The rates of exchange-transfusion also were similar between the groups (no-ibuprofen group: 14 infants [4.8%]; ibuprofen group: 19 infants [4.5%]).
CONCLUSIONS:Ibuprofen administration was associated with higher peak total serum bilirubin levels, and the more-pronounced hyperbil-irubinemia led to longer phototherapy. The potential role of competi-tion between ibuprofen and bilirubin in the hepatic glucuronidacompeti-tion pathway is discussed.Pediatrics2009;124:480–484
CONTRIBUTORS:Enrico Zecca, MD, Costantino Romagnoli, MD, Maria Pia De Carolis, MD, Simonetta Costa, MD, Rosa Marra, MD, and Daniele De Luca, MD
Division of Neonatology, Department of Pediatrics, University Hospital “A. Gemelli,” Catholic University of the Sacred Heart, Rome, Italy
KEY WORDS
ibuprofen, patent ductus arteriosus, hyperbilirubinemia, preterm infants
ABBREVIATIONS
TSB—total serum bilirubin PDA—patent ductus arteriosus UB— unbound bilirubin
www.pediatrics.org/cgi/doi/10.1542/peds.2008-2433
doi:10.1542/peds.2008-2433
Accepted for publication Dec 18, 2008
Address correspondence to Enrico Zecca, MD, Istituto di Clinica Pediatrica, Divisione di Neonatologia, Policlinico Universitario “A. Gemelli,” Universita` Cattolica del Sacro Cuore, Largo Agostino Gemelli 8, 00168 Roma, Italy. E-mail: enrizecca@rm.unicatt.it
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275).
Copyright © 2009 by the American Academy of Pediatrics
Ibuprofen is a cyclooxygenase inhib-itor that is used to close a patent ductus arteriosus (PDA) in preterm infant; as an alternative to indometh-acin. A recent Cochrane meta-analysis indicated that available data support the use of both ibupro-fen and indomethacin and “clinicians need to weigh the potential side ef-fects of one drug versus the other” when choosing the therapy.1
Ibuprofen competes with bilirubin for albumin binding sites,2,3whereas
indo-methacin can perform the same action only at blood concentrations far in ex-cess of those usually reached for PDA closure.4 Moreover, competition
be-tween ibuprofen and bilirubin in the hepatic glucuronidation pathway has been demonstrated,5 and potentially
this could increase circulating biliru-bin levels. Preterm infants may be at increased risk for bilirubin encepha-lopathy, because the need for PDA treatment and the onset of hyperbiliru-binemia often overlap in the early neo-natal period. In vitro studies demon-strated that ibuprofen was able to displace bilirubin from albumin and this increased the levels of unbound bilirubin (UB) in neonatal plasma chal-lenged with the drug.6,7No clinical data
on the effect of ibuprofen on the course of hyperbilirubinemia and total serum bilirubin (TSB) levels are avail-able, and the only study that analyzed UB levels in vivo was that reported by Van Overmeire,8 who found no
differ-ences in UB levels after ibuprofen ad-ministration. The aim of this study was to provide data regarding the effect of ibuprofen on hyperbilirubinemia, com-paring 2 large populations of preterm infants who were exposed to the drug or not exposed.
METHODS
Since 2000, after the conclusion of a randomized, controlled trial,9 all
in-fants atⱕ30 weeks of gestation have
been subjected to intravenous ibupro-fen prophylaxis with a loading dose of 10 mg/kg followed by 2 maintenance doses of 5 mg/kg, 24 hours apart. We used data from our database, with the approval of our institutional review board, and we performed a retrospec-tive cohort study. The population was subdivided into 2 cohorts, according to the period in which the original trial9
was conducted, that is, 1993–1999 (no-ibuprofen group) or 2000 –2007 (ibu-profen group). We excluded from our analysis all infants with major congen-ital malformations, infants who died or underwent surgical procedures in the first 72 hours of life, and infants who developed cholestasis. Infants who had a diagnosis of PDA and received ibuprofen treatment also were ex-cluded from the no-ibuprofen group. More than 95% of mothers had re-ceived full prenatal care. Gestational ages were estimated by using post-menstrual dates and early-gestation sonographic findings. Parenteral nu-trition was initiated in the first 24 hours of life for all infants with birth weights of ⬍1250 g, according to a regimen that enabled fluid intake to be increased progressively from 60 to 150 mL/kg per day during the first week. No changes to this policy were made dur-ing the study period. Minimal enteral feeding was initiated as soon as possi-ble, generally within the first 72 hours of life.
TSB levels were measured in capillary blood with direct spectrophotometry10
(Microbilimeter Twin Beam Plus model 11144A73; Ginevri, Rome, Italy) at ad-mission, every 8 hours up to the begin-ning of phototherapy, and then every 12 hours until 48 hours after the end of phototherapy. Our protocol for photo-therapy uses a formal strict flowchart that was approved by the whole team; it has been subjected to periodic monitoring and discussion to verify its efficacy.11 Adherence to the protocol
has always been total, and no differ-ences in jaundice management be-tween the 2 study groups were noted. Phototherapy was initiated when TSB levels exceeded 6 mg/dL, and exchange-transfusion was performed if TSB levels increased to 14 mg/dL de-spite phototherapy. Phototherapy was stopped when TSB levels were reduced to 50% of the initial value (irrespective of the duration of phototherapy). Pho-totherapy was performed with stan-dard units composed of 4 fluorescent lamps and 4 blue lamps, and the jaundice management protocol11
re-mained unchanged throughout the study period.
All infants underwent blood group de-termination and direct Coombs testing to identify ABO/Rh factor isoimmuniza-tion. Hypoalbuminemia was defined as albumin levels of⬍2.5 g/dL during the first week. Glucose-6-phosphate dehy-drogenase activity was measured when indicated by any clinical or ana-mnestic data. No changes in diagnostic protocols were made during the study period. Fisher’s test, the 2test,
Stu-dent’s t test, or the Mann-Whitney U
test was used as appropriate, after the data distribution was checked with the Kolmogorov-Smirnov test. A retrospec-tive power analysis was conducted as described below.12,13
RESULTS
Complete data were available for 288 of 343 infants in the no-ibuprofen co-hort and 418 of 450 infants in the ibu-profen cohort (Fig 1). The 2 popula-tions were almost identical with respect to gestational age (27.8⫾1.8 vs 27.6⫾1.9 weeks;P⫽.152), birth weight (1016⫾384 vs 1024⫾496 g;
P ⫽ .819), male gender (53.4% vs 50.6%;P⫽.455), 5-minute Apgar score (8.1 [interquartile range: 7– 8.5] vs 8 [interquartile range: 7.5– 8.5]; P ⫽
.345), and Clinical Risk Index for Babies score (2.4⫾1 vs 2.2⫾1.2;P⫽.499).
Small-for-gestational age infants were equally represented in the 2 groups (28.1% vs 28.5%;P⫽.92). Given theP
values of⬎.10, none of these variables was suitable for a multivariate analy-sis; therefore, such an analysis was not performed. Table 1 shows the main jaundice data. Ibuprofen-treated in-fants had higher peak TSB levels, more need for phototherapy, and a longer duration of phototherapy, compared with non–ibuprofen-treated infants. A retrospective power analysis,12,13
con-sidering an␣error of .05, the size of the cohorts, and the described results, yielded a power of⬃99% for the differ-ence in TSB levels,⬃96% for the differ-ence in phototherapy duration, and
⬃70% for the phototherapy need. Although we did not measure UB levels, the TSB-UB correlation curve
pub-lished by Ahlfors14in 2000 allowed us
to estimate mean UB levels of⬃0.19⫾ 0.08 g/mL and ⬃0.253 ⫾ 0.012
g/mL for the no-ibuprofen and ibu-profen groups, respectively. The usual dosing scheme for ibuprofen adminis-tration provides serum drug concen-trations of 116 to 180 mg/dL,2,3which
can produce 1.5- to 2-fold increases in UB concentrations, according to Ahl-fors.7Therefore, we estimated mean UB
levels of⬃0.38 to⬃0.51g/mL for the ibuprofen group. Taking into account the 95th percentile (13.05 mg/dL) of the dis-tribution of TSB levels in the ibuprofen group, we estimated a corresponding UB level of⬃0.80g/mL.
DISCUSSION
We provide preliminary findings re-garding the effect of ibuprofen on
neo-tions, in which 92.9% and 83.9% of ne-onates were exposed or not exposed, respectively, to the drug. The retro-spective power analysis proved that our study was sufficiently powered, because it represented almost the en-tire population of preterm infants of
ⱕ30 weeks of gestation who were admitted to our unit in those years. Moreover, because jaundice and fluid management did not change and no other drugs interfering with albumin-bilirubin binding were introduced, our study represents an almost-ideal situ-ation for evalusitu-ation of the effect of the changing variable (ibuprofen adminis-tration). The different findings for the 2 study groups seemed to be unrelated to isoimmunization, albumin levels, or glucose-6-phosphate dehydrogenase deficiency. We emphasize that, al-though our incidence of glucose-6-phosphate dehydrogenase deficiency should be considered an estimation of the real disease incidence, it is consis-tent with the mean incidence reported in Italy and in other Mediterranean countries.15
The potential neurologic risks of using ibuprofen for very preterm infants are still being debated. Attention is partic-ularly focused on UB levels, which are associated with kernicterus more strongly than are TSB levels. In a pre-liminary report, Van Overmeire8
de-scribed no difference in UB levels after ibuprofen administration; but these data are limited to infants with TSB lev-els of⬍10 mg/dL.
To date, there has been no report on the use of TSB levels as a biomarker of adverse neurodevelopmental out-comes for preterm infants treated with ibuprofen for PDA closure. Data from the National Institute of Child Health and Human Development Neo-natal Research Network demonstrated that TSB levels in the first 2 weeks FIGURE 1
Population enrollment flowchart. Studied infants represented 83.9% (no-ibuprofen [NoIBU] group) and 92.9% (ibuprofen [IBU] group) of the whole eligible population.
TABLE 1 Jaundice Evaluation
No-Ibuprofen Group (N⫽288)
Ibuprofen Group (N⫽418)
P
Peak TSB level, mean⫾SD, mg/dL 7.3⫾3.3 9⫾2.5 ⬍.001
TSB levels, range, mg/dL 1.2–14.8 1.8–15.4
Need for phototherapy,n(%) 254 (88.2) 398 (95.2) ⬍.001
Phototherapy duration, mean⫾SD, h 87.2⫾38.6 94.3⫾43.6 .034
Exchange-transfusion,n(%) 14 (4.8) 19 (4.5) .870
Isoimmunization,n(%) 7 (2.4) 8 (1.9) .646
Hypoalbuminemia,n(%) 47 (16.3) 71 (16.9) .795
G6PDH deficiency,n(%) 6 (2) 8 (1.9) .9
were correlated directly with death or neurodevelopmental impairment in ex-tremely low birth weight infants.16
Our data showed that ibuprofen treat-ment was significantly associated with higher TSB levels by ⬃1.5 mg/dL, an amount that might be clinically relevant for very preterm and sick infants, as con-firmed by the greater need for photo-therapy and the longer duration of pho-totherapy in the ibuprofen group. Our findings provide insights into a new mechanism in the complex interactions between bilirubin, albumin, and ibupro-fen. Ibuprofen is known to be partially metabolized by uridine diphosphate-glucuronosyltransferase, which is also responsible for bilirubin glucuronida-tion.5Competitive inhibition of 15%–30%
between ibuprofen and bilirubin was demonstrated in human liver micro-somes,5and our findings are likely to be
the result of this inhibition. This in vitro effect is probably enhanced in vivo by the poor maturity of liver metabolism. Very preterm infants have reduced hepatic glucuronidation activity, which ibupro-fen can further decrease, potentially modifying a weak equilibrium. In this way, ibuprofen may be linked to in-creased neurologic risk not only through direct albumin-bilirubin displacement but also through competition with the enzyme responsible for bilirubin conju-gation. At higher TSB levels, additional bilirubin displacement and increasing UB concentrations may occur when suf-ficiently high ibuprofen levels are
circu-lating simultaneously.14It is of some
con-cern that ⬃5% of infants in the ibuprofen group might have reached UB concentrations of ⱖ0.80 g/mL, a threshold value associated with clinical symptoms of bilirubin neurotoxicity,17
whereas a threshold value of 0.5g/mL was indicated to be predictive of abnor-malities in auditory brainstem re-sponses in preterm infants.18Our study
suggests that the effect of ibuprofen on hepatic glucuronidation may counter-balance the TSB level reduction attribut-able to albumin displacement. An inter-esting case of kernicterus in which reduced hepatic conjugation associated with glucose-6-phosphate dehydroge-nase deficiency was thought to be in-volved and the infant presented with ex-tremely high TSB and UB levels has been described.19It is conceivable that
circu-lating ibuprofen levels that are not high enough to cause significant albumin dis-placement may be sufficient to interfere with the glucuronidation pathway.
We acknowledge that one limitation of our study is that the UB values pro-vided are estimates calculated on the basis of free bilirubin kinetics, by us-ing the bilirubin-albumin bindus-ing con-stant,17 and the estimates may be
affected by some degree of error. However, because the constant is lower in smaller and sick neonates, which leads to higher levels of UB,17the
error (if any) should yield an underes-timation of real UB levels in our popu-lation. A second limitation of our study
is its retrospective design and the pos-sible historical bias. Nevertheless, our study has some characteristics pro-viding power and reliability; jaundice management remained unchanged during the study period and between the 2 groups and the baseline charac-teristics of the groups were not differ-ent. These characteristics allow our results to be considered usefully reli-able for further studies.
Our study suggests that ibuprofen treatment for the closure of PDA in pre-term infants results in increased hyperbilirubinemia. The interactions between bilirubin, albumin, and ibu-profen are complex, and our data suggest that they may occur through-out different pathways. The evaluation of possible ibuprofen toxicity is partic-ularly worthwhile when new treatment strategies with higher doses of ibupro-fen are proposed for very low birth weight infants,20and such evaluation
requires well-designed, prospective studies evaluating both TSB and UB lev-els during ibuprofen administration. UB assays are still reserved for re-search purposes, and TSB levels are the only parameters that are widely used in clinical practice. This was a pragmatic study to advise clinicians to balance the benefits of ibuprofen against its possible side effects, among which more-pronounced hy-perbilirubinemia, with a subsequent impact on global newborn care, should be considered carefully.
REFERENCES
1. Ohlsson A, Walia R, Shah SS. Ibuprofen for the treatment of patent ductus arteriosus in preterm and/or low birth weight infants.Cochrane Database Syst Rev.2008;(1):CD003481
2. Aranda JV, Varvarigou A, Beharry K, et al. Pharmacokinetics and protein binding of intravenous ibuprofen in the premature newborn infant.Acta Pediatr.1997;86(3):289 –293
3. Van Overmeire B, Touw D, Schepens PJC, Kearns GL, van den Anker JN. Ibuprofen pharmacokinet-ics in preterm infants with patent ductus arteriosus.Clin Pharmacol Ther.2001;70(4):336 –343
4. Rasmussen LF, Ahlfors CE, Wennberg RP. Displacement of bilirubin from albumin by indomethacin.
J Clin Pharmacol.1978;18(10):477– 481
5. Kuehl GE, Lampe JW, Potter JD, Bigler J. Glucuronidation of non-steroidal anti-inflammatory drugs: identifying the enzymes responsible in human liver microsomes.Drug Metab Dispos.2005;33(7): 1027–1035
7. Ahlfors CE. Effect of ibuprofen on bilirubin-albumin binding.J Pediatr.2004;144(3):386 –388 8. Van Overmeire B. Ibuprofen, bilirubin and albumin: the importance of a balance.Eur J Pediatr.
2006;165(suppl 1):113
9. De Carolis MP, Romagnoli C, Polimeni V, et al. Prophylactic ibuprofen therapy of patent ductus arteriosus in preterm infants.Eur J Pediatr.2000;159(5):364 –368
10. Kazmierczak SC, Robertson AF, Catrou PG, Briley KP, Kreamer BL, Gourley GR. Direct spectropho-tometry method for measurement of bilirubin in newborns: comparison with HPLC and an auto-mated diazo method.Clin Chem.2002;48(7):1096 –1097
11. Romagnoli C, Zecca E, Papacci P, Vento G, Girlando P, Latella C. Which phototherapy system is most effective in lowering serum bilirubin in very preterm infants?Fetal Diagn Ther.2006;21(2): 204 –209
12. Dupont WD, Plummer WD. PS power and sample size program available for free on the Internet.
Control Clin Trials.1997;18(3):274
13. Dupont WD, Plummer WD. PS power and sample size calculations: a review and computer pro-gram.Control Clin Trials.1990;11(2):116 –128
14. Ahlfors CE. Unbound bilirubin associated with kernicterus: a historical approach.J Pediatr.2000; 137(4):540 –544
15. Cappellini MD, Fiorelli G. Glucose-6-phosphate dehydrogenase deficiency.Lancet.2008;371(9606): 64 –74
16. Oh W, Tyson JE, Fanaroff AA, et al. Association between peak serum bilirubin and neurodevelop-mental outcomes in extremely low birth weight infants.Pediatrics.2003;112(4):773–779 17. Wennberg RP, Ahlfors CE, Bhutani VK, Johnson LH, Shapiro SM. Toward understanding kernicterus:
a challenge to improve the management of jaundiced newborns.Pediatrics.2006;117(2):474 – 485 18. Amin SB, Ahlfors CE, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory
responses in preterm infants.Pediatrics.2001;107(4):664 – 670
19. Ahlfors CE, Herbsman O. Unbound bilirubin in a term newborn with kernicterus.Pediatrics.
2003;111(5):1110 –1112
DOI: 10.1542/peds.2008-2433 originally published online July 20, 2009;
2009;124;480
Pediatrics
Marra and Daniele De Luca
Enrico Zecca, Costantino Romagnoli, Maria Pia De Carolis, Simonetta Costa, Rosa
Does Ibuprofen Increase Neonatal Hyperbilirubinemia?
Services
Updated Information &
http://pediatrics.aappublications.org/content/124/2/480
including high resolution figures, can be found at:
References
http://pediatrics.aappublications.org/content/124/2/480#BIBL
This article cites 19 articles, 6 of which you can access for free at:
Subspecialty Collections
b
http://www.aappublications.org/cgi/collection/hyperbilirubinemia_su Hyperbilirubinemia
sub
http://www.aappublications.org/cgi/collection/fetus:newborn_infant_ Fetus/Newborn Infant
following collection(s):
This article, along with others on similar topics, appears in the
Permissions & Licensing
http://www.aappublications.org/site/misc/Permissions.xhtml
in its entirety can be found online at:
Information about reproducing this article in parts (figures, tables) or
Reprints
http://www.aappublications.org/site/misc/reprints.xhtml
DOI: 10.1542/peds.2008-2433 originally published online July 20, 2009;
2009;124;480
Pediatrics
Marra and Daniele De Luca
http://pediatrics.aappublications.org/content/124/2/480
located on the World Wide Web at:
The online version of this article, along with updated information and services, is
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.