Pediatric Pharmacology
Pilot Evaluation of the Population Pharmacokinetics of Bumetanide in Term Newborn Infants With Seizures
The Journal of Clinical Pharmacology 2016, 56(3) 284–290 © 2015, The American College of Clinical Pharmacology DOI: 10.1002/jcph.596
Vincent Jullien, PharmD, PhD1,2, Ronit M. Pressler, MD, PhD3, Geraldine Boylan, MD, PhD4, Mats Blennow, MD, PhD5, Neil Marlow, MD, PhD3, Catherine Chiron, MD, PhD1, and Gerard Pons, MD, PhD1 for the NEMO Consortium Neonatal Seizure Treatment With Medication Off-Patent
Abstract Recent experimental data suggest bumetanide as a possible therapeutic option in newborn infants with seizures after birth asphyxia. Because pharmacokinetic (PK) data are lacking in this population, who very often benefit from therapeutic cooling, which can modify the PK behavior of a drug, a PK study was conducted in term infants with seizures caused by hypoxic-ischemic encephalopathy. Fourteen infants were included, 13 of them being cooled. Forty-nine blood samples were available for the determination of the plasma concentration of bumetanide. Concentration-time data were analyzed by the use of a population approach performed with Monolix Software. Bumetanide was found to follow a 2-compartment model. The mean values were 0.063 L/h for clearance, 0.28 and 0.44 L for the central and peripheral distribution volumes, respectively, and 0.59 L/h for the distribution clearance. Birth body weight explained the interindividual variability of bumetanide clearance via an allometric model. No relationship was found between bumetanide exposure and its efficacy (reduction in seizure burden) or its toxicity (hearing loss). This study describes the first PK model of bumetanide in hypothermia-treated infants with seizures.
Keywords population pharmacokinetics, neonatology, central nervous system
Seizures are the most common neurological emergency in the neonatal period and are associated with poor clinical outcome (mortality, postnatal epilepsy, motor and cognitive deficits). Phenobarbital is the first-line antiepileptic drug (AED) worldwide, although benzodiazepines and phenytoin are also sometimes used initially. Secondline treatment varies between and within countries.1 Both a Cochrane review and a WHO review found that any of the drugs currently used are effective in only 40%– 60%.2,3 Because of the lack of alternatives, phenobarbital remains the preferred first-line treatment for this condition,1 and new trials of potentially more appropriate drugs are required.1,4 Bumetanide is a loop diuretic whose mechanism of action includes inhibition of the isoforms 1 and 2 of the NKCC Naþ/Kþ/Cl– cotransporter. Recently, isoform 1 of NKCC (NKCC1) has been considered as a promising target for the treatment of neonatal seizures.5 NKCC1 is overexpressed in the immature brain, leading to an accumulation of chloride anion within the neuron. Because of this, the neurotransmitter g-aminobutyric acid (GABA) displays a paradoxical depolarizing, and therefore excitatory, effect in the immature neuron, as its fixation to its GABAA receptor induces the diffusion of the chloride anion toward the extracellular space. This
contrasts with the mature neuron, in which GABA provokes the entrance of the chloride anion, resulting in hyperpolarization of the cellular membrane, explaining the inhibitory effect of GABA on the mature CNS.5–7 There are several experimental models that support the use of bumetanide in this context.5,8 Bumetanide pharmacokinetics (PK) have been previously investigated in term and preterm newborn infants by classical approaches, but these studies involved infants 1
INSERM U1129 “Infantile Epilepsies and Brain Plasticity,” Paris, France; Paris Descartes University; CEA, Gif sur Yvette, France 2 Service de Pharmacologie, H^ opital Europ�een Georges Pompidou, Assistance Publique H^ opitaux de Paris, Paris, France 3 Neuroscience Unit (ICH) and Neonatal Unit (IWH), University College London, London, UK 4 INFANT Research Centre & Neonatal Intensive Care Unit, University College Cork, Cork, Ireland 5 Neonatology, Karolinska Institutet and University Hospital, Stockholm, Sweden Submitted for publication 3 March 2015; accepted 14 July 2015. Corresponding Author: Vincent Jullien, PharmD, PhD, Service de Pharmacologie, H^ opital Europ�een Georges Pompidou, 20 rue Leblanc, 75015 Paris, France Email: [email protected]
2Jullien et al with to 50fluid mL of overload plasma.due Thetosamples cardiopulmonary were thendisease, precipitated who 9–11 These received with 200 bumetanide mL acetonitrile as aand diuretic. vortex-mixed for populations 30 seconds. are different infantsatwith seizures in that (1) Samples werefrom centrifuged 2200g for 30 minutes, and the doses used for diuretic at effect lower, (2) heart supernatant wasthe evaporated 40°Careunder a stream of 13 disease PKreconstituted in infants12 and nitrogen.impacts The drybumetanide residues were withadults, 200 mL and (3) mobile infants with hypoxic-ischemic of the phaseseizures (25:75 due [v/v]to mixture of water encephalopathy areacetic oftenacid treated by therapeutic containing 0.1% (HIE) [v/v] of and methanol), and hypothermia, which can PK of drugs.14,17 To 10 mL was injected into alter the the chromatographic system. provide PK data this specific population of infants with Bumetanide and in flunitrazepam were separated on a C8þ seizures, a PK study was a prospective satisfaction column (250 � 3nested mm, 5within mm), with a mobilemulticenter to phase flow European rate of 0.5Clinical mL/min,Trial and(NEMO), detected designed by UV and evaluate the efficacy of bumetanide to determinewas the Fluo detectors mounted in series. and Flunitrazepam optimal dose for the wavelength treatment ofofneonatal detected bumetanide by UV with an absorption 210 nm, seizures due to HIE. and bumetanide was detected by fluorescence with excitation and emission wavelengths of 228 and 418 nm, respectively. Intra- and interday inaccuracy and CV were Methods less than 10% over the calibration range of the method Patients Treatment (0.005–2and mg/L). The protocol15 was approved and registered by the regulatory authority of each participating member state, Pharmacokinetic Analysis and the relevant ethical committee using at each center Population PK analysis was performed MONOLIX approved the protocol (ie, 7 neonatal intensive care units software (version 4.2) implementing the stochastic approxacross imationEurope). expectation maximization (SAEM) algorithm. Infants HIE and seizures not responding to a One-, 2-, with and 3-compartment models with first-order loading dose of 20 mg/kg IV phenobarbital were included elimination and 0-order input were investigated as structural in study.Nonlinear The other elimination inclusion criteria were gestational PKthe models. (ie, Michaelis-Menten) age between 37 and 43 weeks, postnatal age belowwas 48 was also investigated. Interindividual (IIV) variability hours, evidence of perinatal asphyxia (defined as Apgar assumed to arise from a log-normal distribution of the PK score of 5 A at correlation 5 minutes after birthrandom or severe acidosis parameters. between effects waswith also umbilical cord or first arterial blood sample pH 7.10 or a investigated. Several error models were tested: additive, base deficit of 16 mmol/L or a continued need for proportional, and a combination of additive and multiplicaresuscitation at 10 minutes after birth).the Exclusion criteria tive error models. For each model, log-likelihood, were major congenital infecestimated by congenital importance abnormalities, sampling and the Bayes information, inborn errors of metabolism, genetic syndromes, and tion criterion (BIC), was computed. Model selection was laboratory abnormalities (plasma sodium < 120 mmol/L, based on the likelihood ratio test (LRT) when models were potassium 3.0 and mmol/L, total models bilirubinwere > nested (P 200 mmol/L), administration nonnested. Goodness-of-fit (GOF) plots and precision of of other diuretics or AED prior to inclusion parameter estimation were also taken into(exception account of in midazolam given as a bolus prior to intubation). selecting the best basic model. Therapeutic hypothermia, rescue covariates AED, and (GA, any other The influence of continuous BW, drugs permitted. consent was blood were pH) on the IIVWritten of theinformed PK parameters was obtained from parents prior to recruitment followed by investigated according to the following equation: ongoing confirmation during treatment using a “continual ucovdaily verbal confirmation of consent” including ðCOVi =COV ui ¼ u � process, std Þ consent with parents. Bumetanide was administered as a 5-minute IV where u is the standard value of the parameter for a subject infusion every 12 hours value for 48(COV), hours.and Four different COVi with the median covariate ui and unitary dose levelsand were a priori 0.1, 0.2,i. are the parameter thedefined covariate value (0.05, for a subject andThe 0.3 mg/kg) allocated the included possibleand influence of tocooling on theinfants IIV ofby PKa 16 continuous reassessment parameters was investigatedmethod. as follows:Briefly, a priori probabilities of efficacy/safety were defined for each dose The dose þ u2 corresponding � ð1 � ucooled Þ to the best a priori ui ¼ ulevel. 1 � ucooled efficacy/safety ratio was administered to the first 2 included The between and u2 are therelationship standard values of thebumetanide parameter where u1 infants. dose its efficacy/safety was then on the for a and cooled patient and a patient notassessed cooled, based respectively
The Journal of Clinical Pharmacology / Vol XX No XX 2015 3 285
(ucooled obtained being equal to 1these for hypothermia-treated results with 2 patients, and thepatients newly and to 0 foroptimal normothermic patients). determined dose among the 4 possible doses was wereto selected based of on 2ainfants. Wald The test thenCovariates administered the next cohort followingunitary a forward (P < .05) andwas backward procedure optimal bumetanide dose reassessed after (P
Pilot Evaluation of the Population Pharmacokinetics of Bumetanide in Term Newborn Infants With Seizures
The Journal of Clinical Pharmacology 2016, 56(3) 284–290 © 2015, The American College of Clinical Pharmacology DOI: 10.1002/jcph.596
Vincent Jullien, PharmD, PhD1,2, Ronit M. Pressler, MD, PhD3, Geraldine Boylan, MD, PhD4, Mats Blennow, MD, PhD5, Neil Marlow, MD, PhD3, Catherine Chiron, MD, PhD1, and Gerard Pons, MD, PhD1 for the NEMO Consortium Neonatal Seizure Treatment With Medication Off-Patent
Abstract Recent experimental data suggest bumetanide as a possible therapeutic option in newborn infants with seizures after birth asphyxia. Because pharmacokinetic (PK) data are lacking in this population, who very often benefit from therapeutic cooling, which can modify the PK behavior of a drug, a PK study was conducted in term infants with seizures caused by hypoxic-ischemic encephalopathy. Fourteen infants were included, 13 of them being cooled. Forty-nine blood samples were available for the determination of the plasma concentration of bumetanide. Concentration-time data were analyzed by the use of a population approach performed with Monolix Software. Bumetanide was found to follow a 2-compartment model. The mean values were 0.063 L/h for clearance, 0.28 and 0.44 L for the central and peripheral distribution volumes, respectively, and 0.59 L/h for the distribution clearance. Birth body weight explained the interindividual variability of bumetanide clearance via an allometric model. No relationship was found between bumetanide exposure and its efficacy (reduction in seizure burden) or its toxicity (hearing loss). This study describes the first PK model of bumetanide in hypothermia-treated infants with seizures.
Keywords population pharmacokinetics, neonatology, central nervous system
Seizures are the most common neurological emergency in the neonatal period and are associated with poor clinical outcome (mortality, postnatal epilepsy, motor and cognitive deficits). Phenobarbital is the first-line antiepileptic drug (AED) worldwide, although benzodiazepines and phenytoin are also sometimes used initially. Secondline treatment varies between and within countries.1 Both a Cochrane review and a WHO review found that any of the drugs currently used are effective in only 40%– 60%.2,3 Because of the lack of alternatives, phenobarbital remains the preferred first-line treatment for this condition,1 and new trials of potentially more appropriate drugs are required.1,4 Bumetanide is a loop diuretic whose mechanism of action includes inhibition of the isoforms 1 and 2 of the NKCC Naþ/Kþ/Cl– cotransporter. Recently, isoform 1 of NKCC (NKCC1) has been considered as a promising target for the treatment of neonatal seizures.5 NKCC1 is overexpressed in the immature brain, leading to an accumulation of chloride anion within the neuron. Because of this, the neurotransmitter g-aminobutyric acid (GABA) displays a paradoxical depolarizing, and therefore excitatory, effect in the immature neuron, as its fixation to its GABAA receptor induces the diffusion of the chloride anion toward the extracellular space. This
contrasts with the mature neuron, in which GABA provokes the entrance of the chloride anion, resulting in hyperpolarization of the cellular membrane, explaining the inhibitory effect of GABA on the mature CNS.5–7 There are several experimental models that support the use of bumetanide in this context.5,8 Bumetanide pharmacokinetics (PK) have been previously investigated in term and preterm newborn infants by classical approaches, but these studies involved infants 1
INSERM U1129 “Infantile Epilepsies and Brain Plasticity,” Paris, France; Paris Descartes University; CEA, Gif sur Yvette, France 2 Service de Pharmacologie, H^ opital Europ�een Georges Pompidou, Assistance Publique H^ opitaux de Paris, Paris, France 3 Neuroscience Unit (ICH) and Neonatal Unit (IWH), University College London, London, UK 4 INFANT Research Centre & Neonatal Intensive Care Unit, University College Cork, Cork, Ireland 5 Neonatology, Karolinska Institutet and University Hospital, Stockholm, Sweden Submitted for publication 3 March 2015; accepted 14 July 2015. Corresponding Author: Vincent Jullien, PharmD, PhD, Service de Pharmacologie, H^ opital Europ�een Georges Pompidou, 20 rue Leblanc, 75015 Paris, France Email: [email protected]
2Jullien et al with to 50fluid mL of overload plasma.due Thetosamples cardiopulmonary were thendisease, precipitated who 9–11 These received with 200 bumetanide mL acetonitrile as aand diuretic. vortex-mixed for populations 30 seconds. are different infantsatwith seizures in that (1) Samples werefrom centrifuged 2200g for 30 minutes, and the doses used for diuretic at effect lower, (2) heart supernatant wasthe evaporated 40°Careunder a stream of 13 disease PKreconstituted in infants12 and nitrogen.impacts The drybumetanide residues were withadults, 200 mL and (3) mobile infants with hypoxic-ischemic of the phaseseizures (25:75 due [v/v]to mixture of water encephalopathy areacetic oftenacid treated by therapeutic containing 0.1% (HIE) [v/v] of and methanol), and hypothermia, which can PK of drugs.14,17 To 10 mL was injected into alter the the chromatographic system. provide PK data this specific population of infants with Bumetanide and in flunitrazepam were separated on a C8þ seizures, a PK study was a prospective satisfaction column (250 � 3nested mm, 5within mm), with a mobilemulticenter to phase flow European rate of 0.5Clinical mL/min,Trial and(NEMO), detected designed by UV and evaluate the efficacy of bumetanide to determinewas the Fluo detectors mounted in series. and Flunitrazepam optimal dose for the wavelength treatment ofofneonatal detected bumetanide by UV with an absorption 210 nm, seizures due to HIE. and bumetanide was detected by fluorescence with excitation and emission wavelengths of 228 and 418 nm, respectively. Intra- and interday inaccuracy and CV were Methods less than 10% over the calibration range of the method Patients Treatment (0.005–2and mg/L). The protocol15 was approved and registered by the regulatory authority of each participating member state, Pharmacokinetic Analysis and the relevant ethical committee using at each center Population PK analysis was performed MONOLIX approved the protocol (ie, 7 neonatal intensive care units software (version 4.2) implementing the stochastic approxacross imationEurope). expectation maximization (SAEM) algorithm. Infants HIE and seizures not responding to a One-, 2-, with and 3-compartment models with first-order loading dose of 20 mg/kg IV phenobarbital were included elimination and 0-order input were investigated as structural in study.Nonlinear The other elimination inclusion criteria were gestational PKthe models. (ie, Michaelis-Menten) age between 37 and 43 weeks, postnatal age belowwas 48 was also investigated. Interindividual (IIV) variability hours, evidence of perinatal asphyxia (defined as Apgar assumed to arise from a log-normal distribution of the PK score of 5 A at correlation 5 minutes after birthrandom or severe acidosis parameters. between effects waswith also umbilical cord or first arterial blood sample pH 7.10 or a investigated. Several error models were tested: additive, base deficit of 16 mmol/L or a continued need for proportional, and a combination of additive and multiplicaresuscitation at 10 minutes after birth).the Exclusion criteria tive error models. For each model, log-likelihood, were major congenital infecestimated by congenital importance abnormalities, sampling and the Bayes information, inborn errors of metabolism, genetic syndromes, and tion criterion (BIC), was computed. Model selection was laboratory abnormalities (plasma sodium < 120 mmol/L, based on the likelihood ratio test (LRT) when models were potassium 3.0 and mmol/L, total models bilirubinwere > nested (P 200 mmol/L), administration nonnested. Goodness-of-fit (GOF) plots and precision of of other diuretics or AED prior to inclusion parameter estimation were also taken into(exception account of in midazolam given as a bolus prior to intubation). selecting the best basic model. Therapeutic hypothermia, rescue covariates AED, and (GA, any other The influence of continuous BW, drugs permitted. consent was blood were pH) on the IIVWritten of theinformed PK parameters was obtained from parents prior to recruitment followed by investigated according to the following equation: ongoing confirmation during treatment using a “continual ucovdaily verbal confirmation of consent” including ðCOVi =COV ui ¼ u � process, std Þ consent with parents. Bumetanide was administered as a 5-minute IV where u is the standard value of the parameter for a subject infusion every 12 hours value for 48(COV), hours.and Four different COVi with the median covariate ui and unitary dose levelsand were a priori 0.1, 0.2,i. are the parameter thedefined covariate value (0.05, for a subject andThe 0.3 mg/kg) allocated the included possibleand influence of tocooling on theinfants IIV ofby PKa 16 continuous reassessment parameters was investigatedmethod. as follows:Briefly, a priori probabilities of efficacy/safety were defined for each dose The dose þ u2 corresponding � ð1 � ucooled Þ to the best a priori ui ¼ ulevel. 1 � ucooled efficacy/safety ratio was administered to the first 2 included The between and u2 are therelationship standard values of thebumetanide parameter where u1 infants. dose its efficacy/safety was then on the for a and cooled patient and a patient notassessed cooled, based respectively
The Journal of Clinical Pharmacology / Vol XX No XX 2015 3 285
(ucooled obtained being equal to 1these for hypothermia-treated results with 2 patients, and thepatients newly and to 0 foroptimal normothermic patients). determined dose among the 4 possible doses was wereto selected based of on 2ainfants. Wald The test thenCovariates administered the next cohort followingunitary a forward (P < .05) andwas backward procedure optimal bumetanide dose reassessed after (P
