- Email: [email protected]
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants Luc Desfrere, MD1, Celine Thibaut, MD1, Yves Kibleur, MD2, Alexandre Barbier, MD3, Cecile Bordarier, MD3, and Guy Moriette, MD3 Objective To determine whether ibuprofen displaces bilirubin from albumin in preterm infants. Study design A total of 34 preterm neonates (<32 weeks gestation) treated by ibuprofen (10-5-5 mg/kg) were included in this prospective open-label study. Total bilirubin (TB), unbound bilirubin (UB), and ibuprofen concentrations were measured before, 1 hour, and 6 hours after the first dose; before and 1 hour after the second dose; and 72 hours after the beginning of treatment. The infants were screened by auditory brainstem responses and by neurologic examination at term. Results At baseline, TB, UB, apparent binding affinity of albumin (Ka), and albumin concentrations were 6.0 1.6 mg/dL, 1.9 2.2 mg/dL, 14.1 5.8 L$mmol1, and 28.7 2.3 g/L, respectively. Ibuprofen treatment had no effect on TB, UB, or Ka values. No correlation between UB or Ka and ibuprofen concentrations was found. No neurologic symptoms or significant modifications of auditory brainstem responses were observed at term. Conclusion Ibuprofen (10-5-5 mg/kg) did not displace bilirubin in preterm infants with a baseline TB concentration <8.8 mg/dL. (J Pediatr 2012;160:258-64).
N
ewborn jaundice occurs in almost all preterm infants and may be associated with acute neurologic injury (bilirubin encephalopathy) or death.1 Bilirubin-induced neurologic dysfunction, including developmental delays, may be associated with lesser degrees of hyperbilirubinemia.2 Phototherapy is guided by total bilirubin (TB) values; however, the fraction of TB not bound to albumin, known as unbound bilirubin (UB), is a better predictor of bilirubin-induced neurotoxicity.3 UB can easily cross the blood-brain barrier and more reliably reflects the bilirubin load in the brain.4 UB may vary considerably for a given TB level, because the apparent binding affinity of albumin (Ka) may be altered by many clinical and physiological factors, including postmenstrual age (PMA), acidosis, and hypoxia.5,6 Moreover, some substances (eg, free fatty acids, hematin, drugs) may affect the Ka value, increasing the UB level.7-9 The reported incidence of patent ductus arteriosus (PDA) in preterm infants is 40%-70% on the third day of life.10 PDA is involved in various neonatal comorbidities (eg, brain lesions, chronic lung disease, necrotizing enterocolitis). Although surgery for PDA has limited complications, severe adverse events and poor neurologic outcomes have been reported.11 Therapy with cyclooxygenase inhibitors remains the first-line treatment. Ibuprofen, widely used in the treatment of hemodynamically significant PDA (HsPDA), is as effective as indomethacin and has fewer harmful effects on renal, mesenteric, and cerebral perfusion.12,13 Preferably, HsPDA should be treated during the first week of life, typically at the time when the bilirubin pool is increased. Ibuprofen is 95% bound to albumin and displaces bilirubin from albumin-binding sites at high concentrations in vitro, and there is conflicting evidence regarding the ibuprofen and TB levels required to induce this effect.14-18 Van Overmeire et al19 reported reassuring results with no increase in UB levels during treatment with ibuprofen in vivo. However, Aranda et al20 found significantly higher UB levels in newborns treated with ibuprofen, with no correlation between ibuprofen concentrations and UB. No trial of the potential bilirubin-displacing effect of ibuprofen has yet been published, and none of the previous studies calculated the Ka during ibuprofen treatment. The main objective of the present study was to evaluate whether ibuprofen induces an increase in UB or a change in Ka when used to treat HsPDA in preterm newborns. Secondary objectives included to determine ibuprofen plasma concentrations,
ABR AUC HsPDA IV Ka MR PDA PMA TB UB
Auditory brainstem response Area under the curve Hemodynamically significant patent ductus arteriosus Intravenous Apparent binding affinity of albumin Total bilirubin to albumin molar ratio Patent ductus arteriosus Postmenstrual age Total bilirubin Unbound bilirubin
From the 1Neonatology Service, Louis Mourier Hospital, ^ pitaux de Paris (AP-HP), Assistance Publique-Ho Colombes, and Denis Diderot University, Paris, France; 2 Orphan Europe, Paris, France; and 3Neonatology Service of Port-Royal, Cochin hospital, Assistance ^ pitaux de Paris (AP-HP) and ParisPublique-Ho Descartes University, Paris, France Supported by a grant from Orphan Europe. The authors declare no conflicts of interest. This trial has been registered at (2004-002458-77), a European clinical trials database. 0022-3476/$ - see front matter. Copyright ª 2012 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2011.07.014
258
Vol. 160, No. 2 February 2012 evaluate ibuprofen efficacy and safety, and detect any possible neurotoxicity through both neurologic and auditory brainstem response (ABR) evaluation.
Methods All preterm newborns admitted to the neonatal intensive care unit of Port Royal Hospital were eligible for this open-label, noncomparative trial if they met the following inclusion criteria: PMA strictly <32 weeks, postnatal age between 12 and 72 hours, dependence on mechanical ventilation or nasal continuous positive airway pressure, HsPDA assessed by echocardiographic criteria, umbilical venous catheter positioned in the inferior vena cava already being used for serial blood sampling following the local guidelines, TB level 2.98.8 mg/dL assessed by standard hospital assay, and informed consent signed by both parents/legal guardians. The study protocol was approved by the regional Ethics Committee. This trial was registered at (2004-002458-77), a European clinical trials database (https://eudract.ema.europa.eu/). Exclusion criteria were ductus-dependent congenital heart disease, right-to-left ductal shunt, hydrops fetalis, necrotizing enterocolitis, hepatic insufficiency and/or albumin concentration less than the mean reference values 1 SD for PMA21 or administration of human serum albumin during the 2 hours before inclusion, severe hyperbilirubinemia with TB >8.8 mg/dL (150 mmol/L), platelet count <30 109/L, grade 3 or 4 intraventricular hemorrhage, neurologic dysfunction, urine output #1 mL/kg birth weight/ hour during the previous 12 hours, and serum creatinine $140 mmol/L. Echocardiographic assessments were performed during the first 72 hours of life and at 24 hours after the last ibuprofen injection (T72) using color and pulsed Doppler echocardiography (ATL HDI 3500, with a 7.5- to 10-MHz transducer; Philips Medical Systems, Suresnes, France). HsPDA was defined as evidence of blood flow through the ductus arteriosus (left-to-right shunt) and/or retrograde blood flow in the main pulmonary artery with at least one of the following: ductus arteriosus diameter >1.5 mm, left atrium/aortic ratio >1.4, and left-right ductal shunt with velocity <1 m/s. Infants received the recommended dosage regimen of intravenous (IV) ibuprofen (Pedea; Orphan Europe SARL, Puteaux, France), with a loading dose of 10 mg/kg birth weight followed by 2 maintenance doses of 5 mg/kg at 24-hour intervals. The drug was infused continuously over 15 minutes through the umbilical venous catheter, and the IV line was rinsed with 2 mL of sterile water to ensure complete administration. No concomitant medication was administered during the period 2 hours before and 1 hour after ibuprofen except for continuously infused drugs such as inotropics, insulin, and sufentanyl. Infants were nursed in humidified incubators. Fluid intake was initially 60-80 mL/kg and was increased by 10-20 mL/kg/ day, depending on body weight loss and sodium concentra-
tion. Respiratory distress was treated by respiratory support (conventional mechanical ventilation or high-frequency oscillatory ventilation), supplemental oxygen, and surfactant administration. Inotropics and/or volume expansion (normal saline solution) were used at the attending neonatologist’s discretion. Phototherapy was indicated according to the guidelines for the management of hyperbilirubinemia in preterm infants. Because the modified peroxidase method provides higher UB values than the conventional method,3 and because no population references are yet available for this method, experimental data were not used to guide clinical management of the infants. TB and conjugated bilirubin (standard hospital assays) were measured in the hospital’s biochemistry laboratory using a colorimetric assay (Roche Diagnostics, Mannheim, Germany) from 12 hours of life until values were consistently below thresholds for phototherapy.22 Phototherapy was delivered by standard fluorescent tubes (Photo-therapy 4000, with 4 fluorescent blue light lamps; Dr€ager Medical, L€ ubeck, Germany). Individual serial blood samples (0.5 mL) were obtained before (T0), 1 hour (T1), and 6 hours (T6) after the loading dose; before (T24) and 1 hour (T25) after the second dose; and 72 hours (T72) after starting ibuprofen. Each sample was collected in heparin-lithium tubes, immediately centrifuged, decanted, and stored at 75 10 C protected from light until analysis. TB and UB were assayed with a modified peroxidase method as described previously23 using minimal 2-fold sample dilution at 2 different peroxydase concentrations. The automated FloPro spectrophotometer (LW Ligand, Carlsbad, California) used requires small sample volumes (50 mL). TB is calculated from initial light absorbance of bilirubin at 460 nm, and UB is calculated from the decreased absorbance over time occurring after the addition of horseradish peroxidase and peroxide, which catalyzes UB to colorless compounds. During the study, all samples for each infant were determined at the same time, and a distinct prestudy assay qualification was carried out before each analysis of study samples to check peroxidase activity. The mean concentrations determined during these prestudy assays were set to be the actual concentrations of the quality control samples. Three quality control samples at 3 different TB and UB concentrations were systematically run during the analysis of each set of study samples. Blood pH also was measured at each time point. Because serum albumin level remains relatively constant during the first weeks of life,24 it was measured only at inclusion. TB, UB, and albumin concentrations were used to calculate the TB-to-albumin molar ratio (MR) and Ka (L$mmol1) using the following equation: Ka ¼ ðTB UBÞ=½UBðalbumin TB þ UBÞ Individual plasma concentrations of both optical isomers of ibuprofen (R- and S-ibuprofen) were measured by chiral high-pressure liquid chromatography.25 Laboratory, demographic, and clinical data were recorded up to 36 weeks PMA or transfer from the facility. Maternal 259
THE JOURNAL OF PEDIATRICS
www.jpeds.com
treatments, antenatal glucocorticoids, premature rupture of membranes, PMA, birth weight, and sex were recorded. The efficacy of the first course of ibuprofen (ie, PDA closure, defined as the absence of flow though the ductus arteriosus and absence of retrograde flow in the pulmonary trunk at T72) and the need for another course of ibuprofen or surgical ligation were recorded. Safety was assessed up to 36 weeks PMA. Renal adverse events were detected by daily recordings of weight; fluid and sodium intake; urine output measured by weighing diapers; and serum creatinine, serum sodium, and blood urea nitrogen levels. Phototherapy and other medications were recorded. At 36 weeks PMA, ABRs were recorded by a specialized neurologist. Bilirubin toxicity was defined by a retrocochlear impairment of responses with slowed conduction time between I and V waves and decreased V/I amplitude ratio. Neurologic examinations were performed at term by a single pediatric neurologist (C.B.) using the Amiel-Tison scale.26 Statistical Analysis Qualitative results are expressed as number and percentage, and quantitative results are presented as mean SD or as median (range) if nonnormally distributed. Variations in TB, UB, MR, and Ka values were analyzed by repeatedmeasures ANOVA. A P value <.05 was considered significant. Single linear regression analyses were carried out on each variable theoretically influencing UB and Ka, that is, maximum total (R + S) ibuprofen concentration and area under the plasma concentration time curve of ibuprofen. Statistical analyses were performed with SAS 8.2 (SAS Institute, Cary, North Carolina). As determined by the statistical analyses and based on previously reported clinical data,19 30 preterm infants were required to detect a mean change in UB of 0.09 mg/dL for an SD of 0.15 mg/dL or a mean change of 0.06 mg/dL for a SD of 0.09 mg/dL, with 90% power and a 5% significance level.
Vol. 160, No. 2
Table I. Baseline perinatal characteristics of the study infants (n = 34) PMA, weeks, mean SD Birth weight, g, mean SD Growth restriction, n (%)* Males/females, n Use of antenatal steroids, n (%) Premature rupture of membranes, n (%) Preeclampsia, n (%) Cesarean section, n (%) Apgar score at 5 minutes, median (range) CRIB score, median (range) Ventilation at inclusion, n (%) Conventional mechanical ventilation/high-frequency oscillation Nasal continuous positive airway pressure Surfactant administration, n (%) Inspired oxygen at inclusion, %, mean SD Intracranial hemorrhage, n Grade 1 Grade 2 Grade 3 or 4
27.3 1.5 982 251 4 (11.8) 18/16 30 (88.2) 8 (23.5) 4 (11.5) 21 (61.8) 8 (3-10) 4 (0-13) 21 (61.8)/11 (32.4) 2 (5.9) 34 (100) 24.8 5.3 2 16 0
CRIB, Clinical Risk Index for Babies. *Defined as birth weight <10th percentile on the Lubchenco chart.42
by continuous phototherapy, which was maintained during ibuprofen treatment in 28 of the 31 infants. IV lipids were started during ibuprofen treatment in 9 infants; lipid intake was <2 g/kg/day throughout the study period. Pharmacokinetic Results Total ibuprofen (ie, R + S enantiomers) concentrations for 158 samples are shown in Figure 2. Median (range) concentrations were 50.1 mg/L (18.9-122.7 mg/L) at T1, 33.3 mg/L (6.7-47.3 mg/L) at T6, 25.1 mg/L (3.3-48.6 mg/ L) at T24, 46.4 mg/L (22.1-80.6 mg/L) at T25, and 25.0 mg/L (0.9-41.3 mg/L) at T72. Higher ibuprofen concentrations were observed at the end of the loading dose infusion in 21 infants and after the first maintenance
130 120
Results
260
Ibupr ofen concentr ations (mg/l)
Between December 2004 and December 2007, a total of 387 preterm infants of PMA #32 weeks were admitted to the neonatal intensive care unit (Figure 1; available at www. jpeds.com). Eighty-three infants with PDA were eligible, 34 of whom met all of the study inclusion criteria. One infant died from severe Candida albicans septicemia before 36 weeks PMA (at 33 days postnatal age), and a total of 33 infants completed the study. Demographic characteristics of the study group are presented in Table I. The infants were included in the study at a mean postnatal age of 58 12 hours. At inclusion and using the standard hospital assays, TB, conjugated bilirubin, and albumin concentrations were 5.9 1.6 mg/dL, 0.8 0.3 mg/dL, and 28.7 2.3 g/L, respectively, resulting in an MR of 0.24 0.06. All infants were treated
110 100 90 80 70 60 50 40 30 20 10 T0 T1
T6
T24 T25
T72
Time (hours)
Figure 2. Individual plasma concentrations of total ibuprofen (ie, R + S enantiomers) during the course of ibuprofen therapy. The solid line and T-bar represent the median concentrations and IQRs, respectively, at the various times. Desfrere et al
ORIGINAL ARTICLES
February 2012
Table II. Blood pH, TB, UB, and Ka values during the course of ibuprofen therapy Blood pH
TB, mg/dL
UB, mg/dL
Ka, L$mmol1
Mean SD Median (range)
7.24 0.06 7.24 (7.13-7.36)
6.0 1.6 6.1 (1.5-9.8)
1.9 2.2 1.3 (0-10.0)
14.1 5.8 13.6 (1.0-24.1)
Mean SD Median (range)
7.25 0.05 7.26 (6.98-7.32)
6.0 1.5 5.9 (3.0-8.9)
1.8 1.6 1.4 (0.4-9.1)
14.3 6.9 12.7 (1.5-32.1)
Mean SD Median (range) T24 Mean SD Median (range) T25 Mean SD Median (range) T72 Mean SD Median (range)
7.27 0.08 7.27 (7.14-7.41)
5.8 1.9 5.6 (2.5-9.4)
1.5 1.1 1.2 (0.5-5.7)
14.9 7.1 13.0 (1.6-34.5)
7.27 0.07 7.27 (7.12-7.42)
5.9 2.0 6.1 (2.6-9.8)
1.7 1.8 1.2 (0.3-10.1)
15.7 10.2 12.4 (1.9-48.2)
7.26 0.06 7.28 (7.15-7.45)
5.7 1.9 5.8 (2.1-10.4)
1.7 2.1 1.3 (0.4-11.7)
14.4 8.9 12.0 (1.4-41.4)
7.25 0.06 7.25 (7.12-7.40)
5.8 2.2 5.9 (0.1-9.4)
1.5 1.7 1.2 (0-9.5)
17.2 10.1 14.4 (2.7-49.1)
Time* T0 T1 T6
*Blood samples were obtained before (T0), 1 hour after (T1), and 6 hours after (T6) the loading dose of ibuprofen; before (T24) and 1 hour after (T25) the first maintenance dose; and 72 hours after (T72) the beginning of treatment.
dose in 12 infants. Although marked interindividual variability in ibuprofen concentrations was observed, most values were between 10 and 70 mg/L. Only 2 infants had an ibuprofen concentration >100 mg/L after the loading dose (112 and 122 mg/L). Ibuprofen–Bilirubin–Albumin Interactions Blood pH remained stable over time (Table II). No increase in UB or TB from baseline values was observed after any injection of ibuprofen. Inversely, both TB and UB concentrations tended to decrease over time, although none of these changes was statistically significant. No significant decrease in Ka value was observed after any ibuprofen injection. Linear regression analyses demonstrated no significant correlation between changes from baseline in UB or Ka values and total peak concentration or area under the curve (AUC) of ibuprofen globally over time (Figure 3) and at each sampling time (data not shown). Efficacy and Safety of Treatment PDA was closed after the first course of ibuprofen in 11 infants (32%) (Table III; available at www.jpeds.com). PDA subsequently reopened in 1 of these infants and another 8 infants had spontaneous delayed closure of PDA. Sixteen infants received a second course of ibuprofen, resulting in successful closure in 7 infants. Thus, the total treatment success rate was 73.5% (25 of 34), and 9 infants required surgical ligation of the ductus arteriosus. Audiometric assessment was performed in 30 of 34 infants, 27 of whom were evaluable at 36 weeks PMA; 3 infants were excluded because of a bilateral transmission defect. ABRs were normal in 19 of these infants. Two infants had endocochlear disorders, and 6 had prolonged I-V latency with no significant alteration of the I/V amplitude ratio. Linear regression analyses showed the absence of any significant cor-
relation between I-V latencies and peak UB concentration (R2 = 0.015; P > .05). Neurologic examination was performed in 29 infants at term and was normal in 22 of them. Seven infants exhibited minor abnormalities frequently observed in preterm newborns. No clinical signs of bilirubin neurotoxicity were detected, and routine cerebral ultrasound detected no signs of damage to the basal ganglia from bilirubin neurotoxicity.
Discussion This study’s main finding is the lack of effect of a conventional course of ibuprofen on bilirubin–albumin association in preterm infants. No significant variations in TB, UB, or Ka values from their respective pretreatment values were observed. Moreover, no correlation was found between plasma ibuprofen concentration and UB or Ka values. Despite advances in care, bilirubin encephalopathy remains a serious problem, especially in preterm infants. The major determinant of neurotoxicity is the degree of hyperbilirubinemia present. In extremely low birth weight infants, cases of kernicterus have been reported with TB values as low as 8.8 mg/dL.27 UB is a neurotoxic agent because of its propensity to cross the blood-brain barrier.5 UB increases occur either when the blood-binding capacity is exceeded or when other substances compete for its binding sites on albumin28; thus, the displacement capacity of drugs used in preterm infants merits further study. Because ibuprofen is known to displace bilirubin from albumin at high concentrations in vitro, the ibuprofen dose commonly administered to preterm newborns with jaundice is questionable, warranting a thorough evaluation. An original feature of the present study is the extensive study of ibuprofen in preterm infants specifically in terms of its potential effects on Ka values.18 Our findings confirm those of a previous study reported by Van Overmeire
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
261
www.jpeds.com
11 9 7 5 3 1
R2 = 0.006; P >.05
-1 -3 -5 -7 -9 -11 20
11 9 7 5 3 1
R2 = 0.001; P >.05
-1 -3 -5 -7 -9 -11 0
50 40 30 20
1000
2000
3000
4000
AUC (R + S) ibuprofen (h.µg/mL)
R2 = 0.017; P >.05
10 0 -10 -20 -30 -40
40 60 80 100 120 140 (R + S) ibuprofen (mg/L)
0
Change of Ka from Baseline (L.µmol-1)
Change of UB from Baseline (µg/dL)
0
Vol. 160, No. 2
Change of Ka from Baseline (L.µmol-1)
Change of UB from Baseline (µg/dL)
THE JOURNAL OF PEDIATRICS
20
40 60 80 100 120 140 (R + S) ibuprofen (mg/l)
50 40 30 20
R2 = 0.003; P >.05
10 0 -10 -20 -30 -40 0
1000
2000
3000
4000
AUC (R + S) ibuprofen (h.µg/mL)
Figure 3. Single linear regression analysis between the changes from T0 (before ibuprofen treatment) of A and B, UB or Ka and ibuprofen concentrations at each time ibuprofen or C and D, AUC globally over time.
et al.19 However, that study, as well as a study reported by Aranda et al,20 included very few UB measurements in infants treated with ibuprofen, insufficient to provide an accurate indication of the maximum drug effect because of the rapidly changing UB and TB concentrations14 and the lack of calculation of Ka during ibuprofen treatment.18 In vivo, a steady state is established between plasma and tissue UB levels. In the presence of reduced bilirubin–albumin affinity, the increased plasma UB diffuses into tissues, and a new equilibrium is established. A potential UB increase may be attenuated by this tissue diffusion, which does not occur in vitro.18 In animal models, UB concentration returned to its pretreatment value after a sharp and transient increase despite continued infusion of the displacing drug.29 Therefore, the present study in262
cluded serial and simultaneous determinations of ibuprofen, TB, and UB concentrations during the course of ibuprofen therapy. TB should decrease under these conditions,30 and the increased TB observed in retrospective studies31,32 was likely related to confounding clinical variables.18 Another strength of the present study is its use of the current best available method for UB determination,23 in contrast to most other previous studies, including the most recent study,18 which used a method that might not be sufficiently sensitive to detect slight increases in UB.7 The method used in the present study avoids underestimation of UB due to rate-limiting dissociation of the bilirubin–albumin complex or to binding disorders associated with dilution, such as masking of bilirubin-binding competitors.7 The use of Desfrere et al
February 2012 multiple peroxidase concentrations also avoids underestimation of the UB equilibrium. Nevertheless, a doubling of the UB levels was probably necessary to produce neurotoxicity, and such an increase may be detected by other methods using standard dilutions.7 Indeed, a significant increase in UB concentration has been reported using a less-sensitive method after IV lipid administration in preterm infants.33 The Ka value for very dilute human serum albumin is >108 L$mol1, whereas the Ka value found in the present study was closer to 107 L$mol1. These values are in agreement with the range reported in undiluted serum in previous studies,34,35 bearing in mind that the modified method used to determine UB concentration provides 5- to 10-fold higher values. The results of the present study are more relevant than those derived from in vitro studies at high ibuprofen concentrations,14,16-18 because the potential impact of ibuprofen on bilirubin–albumin dissociation did not result in any clinical manifestations. The present study was prospective, and most of the potential confounding factors that might have masked a change in bilirubin affinity during ibuprofen therapy (eg, albumin, lipids and administration of other drugs, blood pH, phototherapy) were controlled and monitored for. We observed a marked interindividual variability in UB concentrations before and during ibuprofen treatment. UB concentration may vary widely for a given TB value, given that the binding capacity of albumin can be decreased by many clinical and physiological variables.36 In clinical studies, ibuprofen concentrations >100 mg/ L have been reported in the context of ibuprofen overdose and other rare circumstances.37-40 In the present study, isolated ibuprofen concentrations >100 mg/L were observed once each in 2 infants and were not associated with an increase in UB, a decrease in TB, or any change in Ka. Furthermore, a thorough neurologic examination and ABR assessment found no significant abnormal findings possibly associated with increased UB concentration, as highlighted by serial ABR measurements.41 Finally, the hypothesis of competitive inhibition of bilirubin glucuronidation by ibuprofen, as has been proposed elsewhere,32 is not supported by the stable conjugated bilirubin levels before and after ibuprofen therapy (data not shown) during the observation period of the present study; this study was not design for this specific evaluation, however. For ethical as well as safety reasons, our study approved by ethics committee did not allow the inclusion of preterm infants with a baseline TB value >8.8 mg/dL. No displacement of bilirubin by ibuprofen was detected at these TB concentrations. The study provides no information regarding situations of higher baseline TB values or on the possible risks of using higher ibuprofen doses, however. In conclusion, we found no displacement of bilirubin by ibuprofen at a dosage of 10-5-5 mg/kg in preterm infants with a baseline TB concentration <8.8 mg/dL. However, despite these reassuring results, caution should always apply in cases of elevated TB in preterm infants during ibuprofen
ORIGINAL ARTICLES treatment for PDA or when higher doses of ibuprofen are administered. n We thank the families of the study infants for their continuous support of this study, and the medical and nursing staff of the neonatal intensive care unit of Port Royal Hospital for patient care. We also thank Anthony Saul for a stylistic review of the manuscript and Michele. Mayer, MD for the ABR recordings. Submitted for publication Feb 1, 2011; last revision received May 19, 2011; accepted Jul 12, 2011. ^ pital Louis Mourier, 178 rue des Reprint requests: Luc Desfrere, MD, Ho Renouillers, 92700 Colombes, France. E-mail: [email protected]
References 1. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003;88:F455-8. 2. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005;25: 54-9. 3. Wennberg RP, Ahlfors CE, Bhutani VK, Johnson LH, Shapiro SM. Toward understanding kernicterus: a challenge to improve the management of jaundiced newborns. Pediatrics 2006;117:474-85. 4. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C. New concepts in bilirubin encephalopathy. Eur J Clin Invest 2003;33:988-97. 5. Hansen TW. Mechanisms of bilirubin toxicity: clinical implications. Clin Perinatol 2002;29:765-78. 6. 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. Pediatrics 2007;120:1067-73. 7. Ahlfors CE. Benzyl alcohol, kernicterus, and unbound bilirubin. J Pediatr 2001;139:317-9. 8. Brodersen R. Competitive binding of bilirubin and drugs to human serum albumin studied by enzymatic oxidation. J Clin Invest 1974;54: 1353-64. 9. Robertson A, Karp W, Brodersen R. Bilirubin-displacing effect of drugs used in neonatology. Acta Paediatr Scand 1991;80:1119-27. 10. Investigators of the Vermont-Oxford Trials Network Database Project. The Vermont-Oxford Trials Network: very low birth weight outcomes for 1990. Pediatrics 1993;91:540-5. 11. Clyman RI, Chorne N. Patent ductus arteriosus: evidence for and against treatment. J Pediatr 2007;150:216-9. 12. Patel J, Roberts I, Azzopardi D, Hamilton P, Edwards AD. Randomized double-blind controlled trial comparing the effects of ibuprofen with indomethacin on cerebral hemodynamics in preterm infants with patent ductus arteriosus. Pediatr Res 2000;47:36-42. 13. Pezzati M, Vangi V, Biagiotti R, Bertini G, Cianciulli D, Rubaltelli FF. Effects of indomethacin and ibuprofen on mesenteric and renal blood flow in preterm infants with patent ductus arteriosus. J Pediatr 1999; 135:733-8. 14. Ahlfors CE. Effect of ibuprofen on bilirubin-albumin binding. J Pediatr 2004;144:386-8. 15. Ambat MT, Ostrea EM, Jr., Aranda JV. Effect of ibuprofen L-lysinate on bilirubin binding to albumin as measured by saturation index and horseradish peroxidase assays. J Perinatol 2008;28:287-90. 16. Cooper-Peel C, Brodersen R, Robertson A. Does ibuprofen affect bilirubin–albumin binding in newborn infant serum? Pharmacol Toxicol 1996;79:297-9. 17. Diot C, Kibleur Y, Desfrere L. Effect of ibuprofen on bilirubin–albumin binding in vitro at concentrations observed during treatment of patent ductus arteriosus. Early Hum Dev 2010;86:315-7. 18. Soligard HT, Nilsen OG, Bratlid D. Displacement of bilirubin from albumin by ibuprofen in vitro. Pediatr Res 2010;67:614-8. 19. Van Overmeire B, Vanhagendoren S, Schepens P, Ahlfors C. The influence of ibuprofen-lysine on unbound bilirubin plasma levels in preterm
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
263
THE JOURNAL OF PEDIATRICS
20.
21.
22. 23.
24. 25.
26.
27.
28. 29.
30.
31.
264
www.jpeds.com
neonates [abstract]. In: PAS Annual Meeting. San Francisco, CA; 2004:2687. Available from http://www.pas-meeting.org/. Accessed Aug 23, 2011. Aranda JV, Wrong RJ, Thomas R, Vreman HJ, Steinhilber G, Ahlfors C, et al. plasma unbound bilirubin in placebo and ibuprofen treated preterm neonates [abstract]. In: PAS Annual Meeting. Honolulu, Hawaii; 2008:5844.9. Available from http://www.pas-meeting.org/. Accessed Aug 23, 2011. Reading RF, Ellis R, Fleetwood A. Plasma albumin and total protein in preterm babies from birth to eight weeks. Early Hum Dev 1990; 22:81-7. Cockington RA. A guide to the use of phototherapy in the management of neonatal hyperbilirubinemia. J Pediatr 1979;95:281-5. Ahlfors CE, Marshall GD, Wolcott DK, Olson DC, Van Overmeire B. Measurement of unbound bilirubin by the peroxidase test using zone fluidics. Clin Chim Acta 2006;365:78-85. Cartlidge P, Rutter N. Serum albumin concentrations and oedema in newborn. Arch Dis Child 1986;61:657-60. Gregoire N, Desfrere L, Roze JC, Kibleur Y, Koehne P. Population pharmacokinetic analysis of ibuprofen enantiomers in preterm newborn infants. J Clin Pharmacol 2008;48:1460-8. Amiel-Tison C. Update of the Amiel-Tison neurologic assessment for the term neonate or at forty weeks corrected age. Pediatr Neurol 2002; 27:196-212. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, WeisglasKuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003;112:1256-63. Shapiro SM. Bilirubin toxicity in the developing nervous system. Pediatr Neurol 2003;29:410-21. Brodersen R, Ebbesen F. Bilirubin-displacing effect of ampicillin, indomethacin, chlorpromazine, gentamicin, and parabens in vitro and in newborn infants. J Pharm Sci 1983;72:248-53. Brodersen R, Friis-Hansen B, Stern L. Drug-induced displacement of bilirubin from albumin in the newborn. Dev Pharmacol Ther 1983;6: 217-29. Rheinlaender C, Helfenstein D, Walch E, Berns M, Obladen M, Koehne P. Total serum bilirubin levels during cyclooxygenase inhibitor treatment for patent ductus arteriosus in preterm infants. Acta Paediatr 2009;98:36-42.
Vol. 160, No. 2 32. Zecca E, Romagnoli C, De Carolis MP, Costa S, Marra R, De Luca D. Does ibuprofen increase neonatal hyperbilirubinemia? Pediatrics 2009; 124:480-4. 33. Amin SB, Harte T, Scholer L, Wang H. Intravenous lipid and bilirubinalbumin binding variables in premature infants. Pediatrics 2009;124:211-7. 34. Ahlfors CE. Measurement of plasma unbound unconjugated bilirubin. Anal Biochem 2000;279:130-5. 35. Weisiger RA, Ostrow JD, Koehler RK, Webster CC, Mukerjee P, Pascolo L, et al. Affinity of human serum albumin for bilirubin varies with albumin concentrations and buffer composition: result of a novel approach. J Biol Chem 2001;276:29953-60. 36. Oh W, Stevenson DK, Tyson JE, Morris BH, Ahlfors CE, Bender GJ, et al. Influence of clinical status on the association between plasma total and unbound bilirubin and death or adverse neurodevelopmental outcomes in extremely low birth weight infants. Acta Paediatr 2010; 99:673-8. 37. Aranda JV, Clyman R, Cox B, Van Overmeire B, Wozniak P, Sosenko I, et al. A randomized, double-blind, placebo-controlled trial on intravenous ibuprofen L-lysine for the early closure of nonsymptomatic patent ductus arteriosus within 72 hours of birth in extremely low-birth-weight infants. Am J Perinatol 2009;26:235-45. 38. Desfrere L, Zohar S, Morville P, Brunhes A, Chevret S, Pons G, et al. Dose-finding study of ibuprofen in patent ductus arteriosus using the continual reassessment method. J Clin Pharm Ther 2005;30: 121-32. 39. Gregoire N, Gualano V, Geneteau A, Millerioux L, Brault M, Mignot A, et al. Population pharmacokinetics of ibuprofen enantiomers in very premature neonates. J Clin Pharmacol 2004;44:1114-24. 40. Van Overmeire B, Touw D, Schepens PJ, Kearns GL, van den Anker JN. Ibuprofen pharmacokinetics in preterm infants with patent ductus arteriosus. Clin Pharmacol Ther 2001;70:336-43. 41. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001;107:664-70. 42. Lubchenco L, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from live-born birth weight data at 24 to 42 weeks of gestation. Pediatrics 1963;32:793-800. 43. Jobe A, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med 2001;163:1723-9.
Desfrere et al
ORIGINAL ARTICLES
February 2012
All NICU Admissions n = 1189
PMA < 32 wks n = 387
PDA between H12 and H72 of life n = 83
Patients excluded n = 49 Early death n = 6 No umbilical cord catheter n = 7 No HsPDA n = 12 Hypoalbuminemia n = 3 Parental refusal n = 8 Enrollment in another study n = 3 Enteropathy n = 2 Congenital heart disease n = 2 TB < 2.9 or > 8.8 mg/dL n = 4 Neurological dysfunction n = 2
Patients included n = 34
Completed the study n = 33
Drop-outs n=1
Figure 1. Population enrollment flowchart.
Table III. Postnatal patient outcomes (n = 34) PDA status, n (%) Closed ductus at 72 hours Need for backup treatment Surgical ligation Renal adverse effects during treatment, n (%) Oliguria (<1 mL/kg/h) Increased creatinine >140 mmol/L Decreased sodium #130 mmol/L Outcome variable, n (%) Death before 36 weeks PMA Moderate or severe bronchopulmonary dysplasia* Sepsis Necrotizing enterocolitis Intestinal perforation Severe intraventricular hemorrhage (grade 3-4) Periventricular leukomalacia Retinopathy of prematurity Abnormal ABR† Abnormal neurologic examinationz
11 (32.3) 16 (47.1) 9 (26.5) 2 (5.9) 0 4 (12.1) 1 (2.9) 7 (26.5) 8 (24.2) 4 (12.1) 1 (2.9) 2 (5.9) 0 1 (2.9) 8 (26.6) 7 (24.1)
*Bronchopulmonary dysplasia was defined as oxygen requirement >21% for at least 28 days assessed at 36 weeks PMA.43 †ABR assessment was performed in 30 infants at 36 weeks PMA. zNeurologic examination was performed in 29 newborns at term.
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
264.e1
N
ewborn jaundice occurs in almost all preterm infants and may be associated with acute neurologic injury (bilirubin encephalopathy) or death.1 Bilirubin-induced neurologic dysfunction, including developmental delays, may be associated with lesser degrees of hyperbilirubinemia.2 Phototherapy is guided by total bilirubin (TB) values; however, the fraction of TB not bound to albumin, known as unbound bilirubin (UB), is a better predictor of bilirubin-induced neurotoxicity.3 UB can easily cross the blood-brain barrier and more reliably reflects the bilirubin load in the brain.4 UB may vary considerably for a given TB level, because the apparent binding affinity of albumin (Ka) may be altered by many clinical and physiological factors, including postmenstrual age (PMA), acidosis, and hypoxia.5,6 Moreover, some substances (eg, free fatty acids, hematin, drugs) may affect the Ka value, increasing the UB level.7-9 The reported incidence of patent ductus arteriosus (PDA) in preterm infants is 40%-70% on the third day of life.10 PDA is involved in various neonatal comorbidities (eg, brain lesions, chronic lung disease, necrotizing enterocolitis). Although surgery for PDA has limited complications, severe adverse events and poor neurologic outcomes have been reported.11 Therapy with cyclooxygenase inhibitors remains the first-line treatment. Ibuprofen, widely used in the treatment of hemodynamically significant PDA (HsPDA), is as effective as indomethacin and has fewer harmful effects on renal, mesenteric, and cerebral perfusion.12,13 Preferably, HsPDA should be treated during the first week of life, typically at the time when the bilirubin pool is increased. Ibuprofen is 95% bound to albumin and displaces bilirubin from albumin-binding sites at high concentrations in vitro, and there is conflicting evidence regarding the ibuprofen and TB levels required to induce this effect.14-18 Van Overmeire et al19 reported reassuring results with no increase in UB levels during treatment with ibuprofen in vivo. However, Aranda et al20 found significantly higher UB levels in newborns treated with ibuprofen, with no correlation between ibuprofen concentrations and UB. No trial of the potential bilirubin-displacing effect of ibuprofen has yet been published, and none of the previous studies calculated the Ka during ibuprofen treatment. The main objective of the present study was to evaluate whether ibuprofen induces an increase in UB or a change in Ka when used to treat HsPDA in preterm newborns. Secondary objectives included to determine ibuprofen plasma concentrations,
ABR AUC HsPDA IV Ka MR PDA PMA TB UB
Auditory brainstem response Area under the curve Hemodynamically significant patent ductus arteriosus Intravenous Apparent binding affinity of albumin Total bilirubin to albumin molar ratio Patent ductus arteriosus Postmenstrual age Total bilirubin Unbound bilirubin
From the 1Neonatology Service, Louis Mourier Hospital, ^ pitaux de Paris (AP-HP), Assistance Publique-Ho Colombes, and Denis Diderot University, Paris, France; 2 Orphan Europe, Paris, France; and 3Neonatology Service of Port-Royal, Cochin hospital, Assistance ^ pitaux de Paris (AP-HP) and ParisPublique-Ho Descartes University, Paris, France Supported by a grant from Orphan Europe. The authors declare no conflicts of interest. This trial has been registered at (2004-002458-77), a European clinical trials database. 0022-3476/$ - see front matter. Copyright ª 2012 Mosby Inc. All rights reserved. 10.1016/j.jpeds.2011.07.014
258
Vol. 160, No. 2 February 2012 evaluate ibuprofen efficacy and safety, and detect any possible neurotoxicity through both neurologic and auditory brainstem response (ABR) evaluation.
Methods All preterm newborns admitted to the neonatal intensive care unit of Port Royal Hospital were eligible for this open-label, noncomparative trial if they met the following inclusion criteria: PMA strictly <32 weeks, postnatal age between 12 and 72 hours, dependence on mechanical ventilation or nasal continuous positive airway pressure, HsPDA assessed by echocardiographic criteria, umbilical venous catheter positioned in the inferior vena cava already being used for serial blood sampling following the local guidelines, TB level 2.98.8 mg/dL assessed by standard hospital assay, and informed consent signed by both parents/legal guardians. The study protocol was approved by the regional Ethics Committee. This trial was registered at (2004-002458-77), a European clinical trials database (https://eudract.ema.europa.eu/). Exclusion criteria were ductus-dependent congenital heart disease, right-to-left ductal shunt, hydrops fetalis, necrotizing enterocolitis, hepatic insufficiency and/or albumin concentration less than the mean reference values 1 SD for PMA21 or administration of human serum albumin during the 2 hours before inclusion, severe hyperbilirubinemia with TB >8.8 mg/dL (150 mmol/L), platelet count <30 109/L, grade 3 or 4 intraventricular hemorrhage, neurologic dysfunction, urine output #1 mL/kg birth weight/ hour during the previous 12 hours, and serum creatinine $140 mmol/L. Echocardiographic assessments were performed during the first 72 hours of life and at 24 hours after the last ibuprofen injection (T72) using color and pulsed Doppler echocardiography (ATL HDI 3500, with a 7.5- to 10-MHz transducer; Philips Medical Systems, Suresnes, France). HsPDA was defined as evidence of blood flow through the ductus arteriosus (left-to-right shunt) and/or retrograde blood flow in the main pulmonary artery with at least one of the following: ductus arteriosus diameter >1.5 mm, left atrium/aortic ratio >1.4, and left-right ductal shunt with velocity <1 m/s. Infants received the recommended dosage regimen of intravenous (IV) ibuprofen (Pedea; Orphan Europe SARL, Puteaux, France), with a loading dose of 10 mg/kg birth weight followed by 2 maintenance doses of 5 mg/kg at 24-hour intervals. The drug was infused continuously over 15 minutes through the umbilical venous catheter, and the IV line was rinsed with 2 mL of sterile water to ensure complete administration. No concomitant medication was administered during the period 2 hours before and 1 hour after ibuprofen except for continuously infused drugs such as inotropics, insulin, and sufentanyl. Infants were nursed in humidified incubators. Fluid intake was initially 60-80 mL/kg and was increased by 10-20 mL/kg/ day, depending on body weight loss and sodium concentra-
tion. Respiratory distress was treated by respiratory support (conventional mechanical ventilation or high-frequency oscillatory ventilation), supplemental oxygen, and surfactant administration. Inotropics and/or volume expansion (normal saline solution) were used at the attending neonatologist’s discretion. Phototherapy was indicated according to the guidelines for the management of hyperbilirubinemia in preterm infants. Because the modified peroxidase method provides higher UB values than the conventional method,3 and because no population references are yet available for this method, experimental data were not used to guide clinical management of the infants. TB and conjugated bilirubin (standard hospital assays) were measured in the hospital’s biochemistry laboratory using a colorimetric assay (Roche Diagnostics, Mannheim, Germany) from 12 hours of life until values were consistently below thresholds for phototherapy.22 Phototherapy was delivered by standard fluorescent tubes (Photo-therapy 4000, with 4 fluorescent blue light lamps; Dr€ager Medical, L€ ubeck, Germany). Individual serial blood samples (0.5 mL) were obtained before (T0), 1 hour (T1), and 6 hours (T6) after the loading dose; before (T24) and 1 hour (T25) after the second dose; and 72 hours (T72) after starting ibuprofen. Each sample was collected in heparin-lithium tubes, immediately centrifuged, decanted, and stored at 75 10 C protected from light until analysis. TB and UB were assayed with a modified peroxidase method as described previously23 using minimal 2-fold sample dilution at 2 different peroxydase concentrations. The automated FloPro spectrophotometer (LW Ligand, Carlsbad, California) used requires small sample volumes (50 mL). TB is calculated from initial light absorbance of bilirubin at 460 nm, and UB is calculated from the decreased absorbance over time occurring after the addition of horseradish peroxidase and peroxide, which catalyzes UB to colorless compounds. During the study, all samples for each infant were determined at the same time, and a distinct prestudy assay qualification was carried out before each analysis of study samples to check peroxidase activity. The mean concentrations determined during these prestudy assays were set to be the actual concentrations of the quality control samples. Three quality control samples at 3 different TB and UB concentrations were systematically run during the analysis of each set of study samples. Blood pH also was measured at each time point. Because serum albumin level remains relatively constant during the first weeks of life,24 it was measured only at inclusion. TB, UB, and albumin concentrations were used to calculate the TB-to-albumin molar ratio (MR) and Ka (L$mmol1) using the following equation: Ka ¼ ðTB UBÞ=½UBðalbumin TB þ UBÞ Individual plasma concentrations of both optical isomers of ibuprofen (R- and S-ibuprofen) were measured by chiral high-pressure liquid chromatography.25 Laboratory, demographic, and clinical data were recorded up to 36 weeks PMA or transfer from the facility. Maternal 259
THE JOURNAL OF PEDIATRICS
www.jpeds.com
treatments, antenatal glucocorticoids, premature rupture of membranes, PMA, birth weight, and sex were recorded. The efficacy of the first course of ibuprofen (ie, PDA closure, defined as the absence of flow though the ductus arteriosus and absence of retrograde flow in the pulmonary trunk at T72) and the need for another course of ibuprofen or surgical ligation were recorded. Safety was assessed up to 36 weeks PMA. Renal adverse events were detected by daily recordings of weight; fluid and sodium intake; urine output measured by weighing diapers; and serum creatinine, serum sodium, and blood urea nitrogen levels. Phototherapy and other medications were recorded. At 36 weeks PMA, ABRs were recorded by a specialized neurologist. Bilirubin toxicity was defined by a retrocochlear impairment of responses with slowed conduction time between I and V waves and decreased V/I amplitude ratio. Neurologic examinations were performed at term by a single pediatric neurologist (C.B.) using the Amiel-Tison scale.26 Statistical Analysis Qualitative results are expressed as number and percentage, and quantitative results are presented as mean SD or as median (range) if nonnormally distributed. Variations in TB, UB, MR, and Ka values were analyzed by repeatedmeasures ANOVA. A P value <.05 was considered significant. Single linear regression analyses were carried out on each variable theoretically influencing UB and Ka, that is, maximum total (R + S) ibuprofen concentration and area under the plasma concentration time curve of ibuprofen. Statistical analyses were performed with SAS 8.2 (SAS Institute, Cary, North Carolina). As determined by the statistical analyses and based on previously reported clinical data,19 30 preterm infants were required to detect a mean change in UB of 0.09 mg/dL for an SD of 0.15 mg/dL or a mean change of 0.06 mg/dL for a SD of 0.09 mg/dL, with 90% power and a 5% significance level.
Vol. 160, No. 2
Table I. Baseline perinatal characteristics of the study infants (n = 34) PMA, weeks, mean SD Birth weight, g, mean SD Growth restriction, n (%)* Males/females, n Use of antenatal steroids, n (%) Premature rupture of membranes, n (%) Preeclampsia, n (%) Cesarean section, n (%) Apgar score at 5 minutes, median (range) CRIB score, median (range) Ventilation at inclusion, n (%) Conventional mechanical ventilation/high-frequency oscillation Nasal continuous positive airway pressure Surfactant administration, n (%) Inspired oxygen at inclusion, %, mean SD Intracranial hemorrhage, n Grade 1 Grade 2 Grade 3 or 4
27.3 1.5 982 251 4 (11.8) 18/16 30 (88.2) 8 (23.5) 4 (11.5) 21 (61.8) 8 (3-10) 4 (0-13) 21 (61.8)/11 (32.4) 2 (5.9) 34 (100) 24.8 5.3 2 16 0
CRIB, Clinical Risk Index for Babies. *Defined as birth weight <10th percentile on the Lubchenco chart.42
by continuous phototherapy, which was maintained during ibuprofen treatment in 28 of the 31 infants. IV lipids were started during ibuprofen treatment in 9 infants; lipid intake was <2 g/kg/day throughout the study period. Pharmacokinetic Results Total ibuprofen (ie, R + S enantiomers) concentrations for 158 samples are shown in Figure 2. Median (range) concentrations were 50.1 mg/L (18.9-122.7 mg/L) at T1, 33.3 mg/L (6.7-47.3 mg/L) at T6, 25.1 mg/L (3.3-48.6 mg/ L) at T24, 46.4 mg/L (22.1-80.6 mg/L) at T25, and 25.0 mg/L (0.9-41.3 mg/L) at T72. Higher ibuprofen concentrations were observed at the end of the loading dose infusion in 21 infants and after the first maintenance
130 120
Results
260
Ibupr ofen concentr ations (mg/l)
Between December 2004 and December 2007, a total of 387 preterm infants of PMA #32 weeks were admitted to the neonatal intensive care unit (Figure 1; available at www. jpeds.com). Eighty-three infants with PDA were eligible, 34 of whom met all of the study inclusion criteria. One infant died from severe Candida albicans septicemia before 36 weeks PMA (at 33 days postnatal age), and a total of 33 infants completed the study. Demographic characteristics of the study group are presented in Table I. The infants were included in the study at a mean postnatal age of 58 12 hours. At inclusion and using the standard hospital assays, TB, conjugated bilirubin, and albumin concentrations were 5.9 1.6 mg/dL, 0.8 0.3 mg/dL, and 28.7 2.3 g/L, respectively, resulting in an MR of 0.24 0.06. All infants were treated
110 100 90 80 70 60 50 40 30 20 10 T0 T1
T6
T24 T25
T72
Time (hours)
Figure 2. Individual plasma concentrations of total ibuprofen (ie, R + S enantiomers) during the course of ibuprofen therapy. The solid line and T-bar represent the median concentrations and IQRs, respectively, at the various times. Desfrere et al
ORIGINAL ARTICLES
February 2012
Table II. Blood pH, TB, UB, and Ka values during the course of ibuprofen therapy Blood pH
TB, mg/dL
UB, mg/dL
Ka, L$mmol1
Mean SD Median (range)
7.24 0.06 7.24 (7.13-7.36)
6.0 1.6 6.1 (1.5-9.8)
1.9 2.2 1.3 (0-10.0)
14.1 5.8 13.6 (1.0-24.1)
Mean SD Median (range)
7.25 0.05 7.26 (6.98-7.32)
6.0 1.5 5.9 (3.0-8.9)
1.8 1.6 1.4 (0.4-9.1)
14.3 6.9 12.7 (1.5-32.1)
Mean SD Median (range) T24 Mean SD Median (range) T25 Mean SD Median (range) T72 Mean SD Median (range)
7.27 0.08 7.27 (7.14-7.41)
5.8 1.9 5.6 (2.5-9.4)
1.5 1.1 1.2 (0.5-5.7)
14.9 7.1 13.0 (1.6-34.5)
7.27 0.07 7.27 (7.12-7.42)
5.9 2.0 6.1 (2.6-9.8)
1.7 1.8 1.2 (0.3-10.1)
15.7 10.2 12.4 (1.9-48.2)
7.26 0.06 7.28 (7.15-7.45)
5.7 1.9 5.8 (2.1-10.4)
1.7 2.1 1.3 (0.4-11.7)
14.4 8.9 12.0 (1.4-41.4)
7.25 0.06 7.25 (7.12-7.40)
5.8 2.2 5.9 (0.1-9.4)
1.5 1.7 1.2 (0-9.5)
17.2 10.1 14.4 (2.7-49.1)
Time* T0 T1 T6
*Blood samples were obtained before (T0), 1 hour after (T1), and 6 hours after (T6) the loading dose of ibuprofen; before (T24) and 1 hour after (T25) the first maintenance dose; and 72 hours after (T72) the beginning of treatment.
dose in 12 infants. Although marked interindividual variability in ibuprofen concentrations was observed, most values were between 10 and 70 mg/L. Only 2 infants had an ibuprofen concentration >100 mg/L after the loading dose (112 and 122 mg/L). Ibuprofen–Bilirubin–Albumin Interactions Blood pH remained stable over time (Table II). No increase in UB or TB from baseline values was observed after any injection of ibuprofen. Inversely, both TB and UB concentrations tended to decrease over time, although none of these changes was statistically significant. No significant decrease in Ka value was observed after any ibuprofen injection. Linear regression analyses demonstrated no significant correlation between changes from baseline in UB or Ka values and total peak concentration or area under the curve (AUC) of ibuprofen globally over time (Figure 3) and at each sampling time (data not shown). Efficacy and Safety of Treatment PDA was closed after the first course of ibuprofen in 11 infants (32%) (Table III; available at www.jpeds.com). PDA subsequently reopened in 1 of these infants and another 8 infants had spontaneous delayed closure of PDA. Sixteen infants received a second course of ibuprofen, resulting in successful closure in 7 infants. Thus, the total treatment success rate was 73.5% (25 of 34), and 9 infants required surgical ligation of the ductus arteriosus. Audiometric assessment was performed in 30 of 34 infants, 27 of whom were evaluable at 36 weeks PMA; 3 infants were excluded because of a bilateral transmission defect. ABRs were normal in 19 of these infants. Two infants had endocochlear disorders, and 6 had prolonged I-V latency with no significant alteration of the I/V amplitude ratio. Linear regression analyses showed the absence of any significant cor-
relation between I-V latencies and peak UB concentration (R2 = 0.015; P > .05). Neurologic examination was performed in 29 infants at term and was normal in 22 of them. Seven infants exhibited minor abnormalities frequently observed in preterm newborns. No clinical signs of bilirubin neurotoxicity were detected, and routine cerebral ultrasound detected no signs of damage to the basal ganglia from bilirubin neurotoxicity.
Discussion This study’s main finding is the lack of effect of a conventional course of ibuprofen on bilirubin–albumin association in preterm infants. No significant variations in TB, UB, or Ka values from their respective pretreatment values were observed. Moreover, no correlation was found between plasma ibuprofen concentration and UB or Ka values. Despite advances in care, bilirubin encephalopathy remains a serious problem, especially in preterm infants. The major determinant of neurotoxicity is the degree of hyperbilirubinemia present. In extremely low birth weight infants, cases of kernicterus have been reported with TB values as low as 8.8 mg/dL.27 UB is a neurotoxic agent because of its propensity to cross the blood-brain barrier.5 UB increases occur either when the blood-binding capacity is exceeded or when other substances compete for its binding sites on albumin28; thus, the displacement capacity of drugs used in preterm infants merits further study. Because ibuprofen is known to displace bilirubin from albumin at high concentrations in vitro, the ibuprofen dose commonly administered to preterm newborns with jaundice is questionable, warranting a thorough evaluation. An original feature of the present study is the extensive study of ibuprofen in preterm infants specifically in terms of its potential effects on Ka values.18 Our findings confirm those of a previous study reported by Van Overmeire
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
261
www.jpeds.com
11 9 7 5 3 1
R2 = 0.006; P >.05
-1 -3 -5 -7 -9 -11 20
11 9 7 5 3 1
R2 = 0.001; P >.05
-1 -3 -5 -7 -9 -11 0
50 40 30 20
1000
2000
3000
4000
AUC (R + S) ibuprofen (h.µg/mL)
R2 = 0.017; P >.05
10 0 -10 -20 -30 -40
40 60 80 100 120 140 (R + S) ibuprofen (mg/L)
0
Change of Ka from Baseline (L.µmol-1)
Change of UB from Baseline (µg/dL)
0
Vol. 160, No. 2
Change of Ka from Baseline (L.µmol-1)
Change of UB from Baseline (µg/dL)
THE JOURNAL OF PEDIATRICS
20
40 60 80 100 120 140 (R + S) ibuprofen (mg/l)
50 40 30 20
R2 = 0.003; P >.05
10 0 -10 -20 -30 -40 0
1000
2000
3000
4000
AUC (R + S) ibuprofen (h.µg/mL)
Figure 3. Single linear regression analysis between the changes from T0 (before ibuprofen treatment) of A and B, UB or Ka and ibuprofen concentrations at each time ibuprofen or C and D, AUC globally over time.
et al.19 However, that study, as well as a study reported by Aranda et al,20 included very few UB measurements in infants treated with ibuprofen, insufficient to provide an accurate indication of the maximum drug effect because of the rapidly changing UB and TB concentrations14 and the lack of calculation of Ka during ibuprofen treatment.18 In vivo, a steady state is established between plasma and tissue UB levels. In the presence of reduced bilirubin–albumin affinity, the increased plasma UB diffuses into tissues, and a new equilibrium is established. A potential UB increase may be attenuated by this tissue diffusion, which does not occur in vitro.18 In animal models, UB concentration returned to its pretreatment value after a sharp and transient increase despite continued infusion of the displacing drug.29 Therefore, the present study in262
cluded serial and simultaneous determinations of ibuprofen, TB, and UB concentrations during the course of ibuprofen therapy. TB should decrease under these conditions,30 and the increased TB observed in retrospective studies31,32 was likely related to confounding clinical variables.18 Another strength of the present study is its use of the current best available method for UB determination,23 in contrast to most other previous studies, including the most recent study,18 which used a method that might not be sufficiently sensitive to detect slight increases in UB.7 The method used in the present study avoids underestimation of UB due to rate-limiting dissociation of the bilirubin–albumin complex or to binding disorders associated with dilution, such as masking of bilirubin-binding competitors.7 The use of Desfrere et al
February 2012 multiple peroxidase concentrations also avoids underestimation of the UB equilibrium. Nevertheless, a doubling of the UB levels was probably necessary to produce neurotoxicity, and such an increase may be detected by other methods using standard dilutions.7 Indeed, a significant increase in UB concentration has been reported using a less-sensitive method after IV lipid administration in preterm infants.33 The Ka value for very dilute human serum albumin is >108 L$mol1, whereas the Ka value found in the present study was closer to 107 L$mol1. These values are in agreement with the range reported in undiluted serum in previous studies,34,35 bearing in mind that the modified method used to determine UB concentration provides 5- to 10-fold higher values. The results of the present study are more relevant than those derived from in vitro studies at high ibuprofen concentrations,14,16-18 because the potential impact of ibuprofen on bilirubin–albumin dissociation did not result in any clinical manifestations. The present study was prospective, and most of the potential confounding factors that might have masked a change in bilirubin affinity during ibuprofen therapy (eg, albumin, lipids and administration of other drugs, blood pH, phototherapy) were controlled and monitored for. We observed a marked interindividual variability in UB concentrations before and during ibuprofen treatment. UB concentration may vary widely for a given TB value, given that the binding capacity of albumin can be decreased by many clinical and physiological variables.36 In clinical studies, ibuprofen concentrations >100 mg/ L have been reported in the context of ibuprofen overdose and other rare circumstances.37-40 In the present study, isolated ibuprofen concentrations >100 mg/L were observed once each in 2 infants and were not associated with an increase in UB, a decrease in TB, or any change in Ka. Furthermore, a thorough neurologic examination and ABR assessment found no significant abnormal findings possibly associated with increased UB concentration, as highlighted by serial ABR measurements.41 Finally, the hypothesis of competitive inhibition of bilirubin glucuronidation by ibuprofen, as has been proposed elsewhere,32 is not supported by the stable conjugated bilirubin levels before and after ibuprofen therapy (data not shown) during the observation period of the present study; this study was not design for this specific evaluation, however. For ethical as well as safety reasons, our study approved by ethics committee did not allow the inclusion of preterm infants with a baseline TB value >8.8 mg/dL. No displacement of bilirubin by ibuprofen was detected at these TB concentrations. The study provides no information regarding situations of higher baseline TB values or on the possible risks of using higher ibuprofen doses, however. In conclusion, we found no displacement of bilirubin by ibuprofen at a dosage of 10-5-5 mg/kg in preterm infants with a baseline TB concentration <8.8 mg/dL. However, despite these reassuring results, caution should always apply in cases of elevated TB in preterm infants during ibuprofen
ORIGINAL ARTICLES treatment for PDA or when higher doses of ibuprofen are administered. n We thank the families of the study infants for their continuous support of this study, and the medical and nursing staff of the neonatal intensive care unit of Port Royal Hospital for patient care. We also thank Anthony Saul for a stylistic review of the manuscript and Michele. Mayer, MD for the ABR recordings. Submitted for publication Feb 1, 2011; last revision received May 19, 2011; accepted Jul 12, 2011. ^ pital Louis Mourier, 178 rue des Reprint requests: Luc Desfrere, MD, Ho Renouillers, 92700 Colombes, France. E-mail: [email protected]
References 1. Watchko JF, Maisels MJ. Jaundice in low birthweight infants: pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003;88:F455-8. 2. Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). J Perinatol 2005;25: 54-9. 3. Wennberg RP, Ahlfors CE, Bhutani VK, Johnson LH, Shapiro SM. Toward understanding kernicterus: a challenge to improve the management of jaundiced newborns. Pediatrics 2006;117:474-85. 4. Ostrow JD, Pascolo L, Shapiro SM, Tiribelli C. New concepts in bilirubin encephalopathy. Eur J Clin Invest 2003;33:988-97. 5. Hansen TW. Mechanisms of bilirubin toxicity: clinical implications. Clin Perinatol 2002;29:765-78. 6. 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. Pediatrics 2007;120:1067-73. 7. Ahlfors CE. Benzyl alcohol, kernicterus, and unbound bilirubin. J Pediatr 2001;139:317-9. 8. Brodersen R. Competitive binding of bilirubin and drugs to human serum albumin studied by enzymatic oxidation. J Clin Invest 1974;54: 1353-64. 9. Robertson A, Karp W, Brodersen R. Bilirubin-displacing effect of drugs used in neonatology. Acta Paediatr Scand 1991;80:1119-27. 10. Investigators of the Vermont-Oxford Trials Network Database Project. The Vermont-Oxford Trials Network: very low birth weight outcomes for 1990. Pediatrics 1993;91:540-5. 11. Clyman RI, Chorne N. Patent ductus arteriosus: evidence for and against treatment. J Pediatr 2007;150:216-9. 12. Patel J, Roberts I, Azzopardi D, Hamilton P, Edwards AD. Randomized double-blind controlled trial comparing the effects of ibuprofen with indomethacin on cerebral hemodynamics in preterm infants with patent ductus arteriosus. Pediatr Res 2000;47:36-42. 13. Pezzati M, Vangi V, Biagiotti R, Bertini G, Cianciulli D, Rubaltelli FF. Effects of indomethacin and ibuprofen on mesenteric and renal blood flow in preterm infants with patent ductus arteriosus. J Pediatr 1999; 135:733-8. 14. Ahlfors CE. Effect of ibuprofen on bilirubin-albumin binding. J Pediatr 2004;144:386-8. 15. Ambat MT, Ostrea EM, Jr., Aranda JV. Effect of ibuprofen L-lysinate on bilirubin binding to albumin as measured by saturation index and horseradish peroxidase assays. J Perinatol 2008;28:287-90. 16. Cooper-Peel C, Brodersen R, Robertson A. Does ibuprofen affect bilirubin–albumin binding in newborn infant serum? Pharmacol Toxicol 1996;79:297-9. 17. Diot C, Kibleur Y, Desfrere L. Effect of ibuprofen on bilirubin–albumin binding in vitro at concentrations observed during treatment of patent ductus arteriosus. Early Hum Dev 2010;86:315-7. 18. Soligard HT, Nilsen OG, Bratlid D. Displacement of bilirubin from albumin by ibuprofen in vitro. Pediatr Res 2010;67:614-8. 19. Van Overmeire B, Vanhagendoren S, Schepens P, Ahlfors C. The influence of ibuprofen-lysine on unbound bilirubin plasma levels in preterm
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
263
THE JOURNAL OF PEDIATRICS
20.
21.
22. 23.
24. 25.
26.
27.
28. 29.
30.
31.
264
www.jpeds.com
neonates [abstract]. In: PAS Annual Meeting. San Francisco, CA; 2004:2687. Available from http://www.pas-meeting.org/. Accessed Aug 23, 2011. Aranda JV, Wrong RJ, Thomas R, Vreman HJ, Steinhilber G, Ahlfors C, et al. plasma unbound bilirubin in placebo and ibuprofen treated preterm neonates [abstract]. In: PAS Annual Meeting. Honolulu, Hawaii; 2008:5844.9. Available from http://www.pas-meeting.org/. Accessed Aug 23, 2011. Reading RF, Ellis R, Fleetwood A. Plasma albumin and total protein in preterm babies from birth to eight weeks. Early Hum Dev 1990; 22:81-7. Cockington RA. A guide to the use of phototherapy in the management of neonatal hyperbilirubinemia. J Pediatr 1979;95:281-5. Ahlfors CE, Marshall GD, Wolcott DK, Olson DC, Van Overmeire B. Measurement of unbound bilirubin by the peroxidase test using zone fluidics. Clin Chim Acta 2006;365:78-85. Cartlidge P, Rutter N. Serum albumin concentrations and oedema in newborn. Arch Dis Child 1986;61:657-60. Gregoire N, Desfrere L, Roze JC, Kibleur Y, Koehne P. Population pharmacokinetic analysis of ibuprofen enantiomers in preterm newborn infants. J Clin Pharmacol 2008;48:1460-8. Amiel-Tison C. Update of the Amiel-Tison neurologic assessment for the term neonate or at forty weeks corrected age. Pediatr Neurol 2002; 27:196-212. Govaert P, Lequin M, Swarte R, Robben S, De Coo R, WeisglasKuperus N, et al. Changes in globus pallidus with (pre)term kernicterus. Pediatrics 2003;112:1256-63. Shapiro SM. Bilirubin toxicity in the developing nervous system. Pediatr Neurol 2003;29:410-21. Brodersen R, Ebbesen F. Bilirubin-displacing effect of ampicillin, indomethacin, chlorpromazine, gentamicin, and parabens in vitro and in newborn infants. J Pharm Sci 1983;72:248-53. Brodersen R, Friis-Hansen B, Stern L. Drug-induced displacement of bilirubin from albumin in the newborn. Dev Pharmacol Ther 1983;6: 217-29. Rheinlaender C, Helfenstein D, Walch E, Berns M, Obladen M, Koehne P. Total serum bilirubin levels during cyclooxygenase inhibitor treatment for patent ductus arteriosus in preterm infants. Acta Paediatr 2009;98:36-42.
Vol. 160, No. 2 32. Zecca E, Romagnoli C, De Carolis MP, Costa S, Marra R, De Luca D. Does ibuprofen increase neonatal hyperbilirubinemia? Pediatrics 2009; 124:480-4. 33. Amin SB, Harte T, Scholer L, Wang H. Intravenous lipid and bilirubinalbumin binding variables in premature infants. Pediatrics 2009;124:211-7. 34. Ahlfors CE. Measurement of plasma unbound unconjugated bilirubin. Anal Biochem 2000;279:130-5. 35. Weisiger RA, Ostrow JD, Koehler RK, Webster CC, Mukerjee P, Pascolo L, et al. Affinity of human serum albumin for bilirubin varies with albumin concentrations and buffer composition: result of a novel approach. J Biol Chem 2001;276:29953-60. 36. Oh W, Stevenson DK, Tyson JE, Morris BH, Ahlfors CE, Bender GJ, et al. Influence of clinical status on the association between plasma total and unbound bilirubin and death or adverse neurodevelopmental outcomes in extremely low birth weight infants. Acta Paediatr 2010; 99:673-8. 37. Aranda JV, Clyman R, Cox B, Van Overmeire B, Wozniak P, Sosenko I, et al. A randomized, double-blind, placebo-controlled trial on intravenous ibuprofen L-lysine for the early closure of nonsymptomatic patent ductus arteriosus within 72 hours of birth in extremely low-birth-weight infants. Am J Perinatol 2009;26:235-45. 38. Desfrere L, Zohar S, Morville P, Brunhes A, Chevret S, Pons G, et al. Dose-finding study of ibuprofen in patent ductus arteriosus using the continual reassessment method. J Clin Pharm Ther 2005;30: 121-32. 39. Gregoire N, Gualano V, Geneteau A, Millerioux L, Brault M, Mignot A, et al. Population pharmacokinetics of ibuprofen enantiomers in very premature neonates. J Clin Pharmacol 2004;44:1114-24. 40. Van Overmeire B, Touw D, Schepens PJ, Kearns GL, van den Anker JN. Ibuprofen pharmacokinetics in preterm infants with patent ductus arteriosus. Clin Pharmacol Ther 2001;70:336-43. 41. Amin SB, Ahlfors C, Orlando MS, Dalzell LE, Merle KS, Guillet R. Bilirubin and serial auditory brainstem responses in premature infants. Pediatrics 2001;107:664-70. 42. Lubchenco L, Hansman C, Dressler M, Boyd E. Intrauterine growth as estimated from live-born birth weight data at 24 to 42 weeks of gestation. Pediatrics 1963;32:793-800. 43. Jobe A, Bancalari E. Bronchopulmonary dysplasia. Am J Respir Crit Care Med 2001;163:1723-9.
Desfrere et al
ORIGINAL ARTICLES
February 2012
All NICU Admissions n = 1189
PMA < 32 wks n = 387
PDA between H12 and H72 of life n = 83
Patients excluded n = 49 Early death n = 6 No umbilical cord catheter n = 7 No HsPDA n = 12 Hypoalbuminemia n = 3 Parental refusal n = 8 Enrollment in another study n = 3 Enteropathy n = 2 Congenital heart disease n = 2 TB < 2.9 or > 8.8 mg/dL n = 4 Neurological dysfunction n = 2
Patients included n = 34
Completed the study n = 33
Drop-outs n=1
Figure 1. Population enrollment flowchart.
Table III. Postnatal patient outcomes (n = 34) PDA status, n (%) Closed ductus at 72 hours Need for backup treatment Surgical ligation Renal adverse effects during treatment, n (%) Oliguria (<1 mL/kg/h) Increased creatinine >140 mmol/L Decreased sodium #130 mmol/L Outcome variable, n (%) Death before 36 weeks PMA Moderate or severe bronchopulmonary dysplasia* Sepsis Necrotizing enterocolitis Intestinal perforation Severe intraventricular hemorrhage (grade 3-4) Periventricular leukomalacia Retinopathy of prematurity Abnormal ABR† Abnormal neurologic examinationz
11 (32.3) 16 (47.1) 9 (26.5) 2 (5.9) 0 4 (12.1) 1 (2.9) 7 (26.5) 8 (24.2) 4 (12.1) 1 (2.9) 2 (5.9) 0 1 (2.9) 8 (26.6) 7 (24.1)
*Bronchopulmonary dysplasia was defined as oxygen requirement >21% for at least 28 days assessed at 36 weeks PMA.43 †ABR assessment was performed in 30 infants at 36 weeks PMA. zNeurologic examination was performed in 29 newborns at term.
Unbound Bilirubin does not Increase during Ibuprofen Treatment of Patent Ductus Arteriosus in Preterm Infants
264.e1