An open label exploratory dose finding and pharmacokinetic clinical trial of bumetanide for the treatment of NEonatal seizure using Medication Off-patent (NEMO).

Perinatal asphyxia occurs in approximately 20 per 1,000 live births. A
proportion of these babies develop an early neonatal encephalopathy, or hypoxic
ischaemic encephalopathy (HIE) which is a major cause of perinatal mortality
(60% of perinatal mortality), and long-term severe neuro-disability (Volpe,
2008). The incidence of HIE is around 2-3/1000 births in the developed world,
but much higher in the developing world. The prognosis varies greatly, and
depends on the severity of the clinical encephalopathy, varying from mild
irritability to deep coma. Seizures are the hallmark of HIE, and EEG studies
have shown that many asphyxiated babies often have seizures that go unnoticed
and which is not reduced by current antiepileptic drug therapy.
Data from both human and animal studies suggest that seizures amplify neonatal
hypoxic-ischemic brain damage. In a recent study of term newborns with HIE
brain injury was independently associated with the severity of seizures (Miller
et al., 2002). Prolonged seizures cause progressive cerebral hypoxia and
changes in cerebral blood flow (Boylan et al., 1999). These findings support
the hypothesis based on data from animal models, that neonatal seizures are not
only a manifestation of acute ischemic brain injury, but also exacerbate tissue
damage (Wirrell et al., 2001) and provide evidence for the conjecture that
effective treatment of neonatal seizures could attenuate acute brain injury in
this setting.
Phenobarbitone remains the first line antiepileptic drug (AED) for seizures in
infants or babies world-wide despite the fact that it has limited efficacy. The
prognosis for neurodevelopmental outcome is poor in babies with seizures.
Furthermore, there is evidence that phenobarbitone may itself impair
neurodevelopmental outcome and may cause additional brain damage by increasing
neuronal death (apoptosis).
Better treatments for neonatal seizures have been identified as a high priority
for research worldwide. We propose to evaluate the use of bumetanide in babies
whose seizures are resistant to the standard first-line AED regimen
(phenobarbitone). We know that for these babies the outlook is poor. Hence a
trial of novel treatments, based on current knowledge gained in basic research,
is justified. Our research consortium has extensive experience in the study of
neonatal seizures as well as in clinical management and basic neuronal
mechanisms. This will enable us to use continuous EEG monitoring to measure
treatment responses in all infants.
The results of this trial should translate into better care for babies with
seizures and improved neurological outcomes.

The aim of the trial is to obtain data on the optimal dose, feasibility and
pharmacokinetics of bumetanide when given as an add-on treatment for seizures
in full term babies with hypoxic ischemic encephalopathy (HIE). Bumetanide will
be given in a range of doses from 0.05 to 0.3 mg/kg as an adjunct to the
standard treatment (phenobarbitone). The trial will consist of two stages:
Stage 1: a dose-finding and confirmatory stage and Stage 2: a pharmacokinetic
(PK) stage at the optimal dose. Overall this trial will also evaluate the
feasibility of a subsequent larger randomised controlled trial (NEMO2).

Primary Objectives

Stage 1
• To estimate the optimal dose of bumetanide for use as an adjuvant therapy to
phenobarbitone in neonatal seizures. This optimum
dose will be that which achieves the maximum seizure
reduction with an acceptable safety profile when used in addition
to standard therapy (second dose of phenobarbitone) in >
50% of patients
.
Stage 2
• To determine the pharmacokinetics of bumetanide when given
at the optimal dose as an add-on to phenobarbitone for neonatal
seizures not responding to the first dose phenobarbitone

Secondary Objectives
• To assess the feasibility of neonatal seizure treatment
with bumetanide in babies with HIE and seizures that are not responding to
a first dose of phenobarbitone alone.
• To assess whether bumetanide reduces the need for rescue
medication.
• To assess bumetanide as a diuretic in babies with HIE.

This multicentre clinical trial will consist of two stages: Stage 1: a
dose-finding and confirmatory stage and Stage 2: a PK stage, at the optimal
dose.
The dose-finding and confirmatory stage will be conducted using the continual
reassessment method requiring a Bayesian sequential design. Four different
dosage regimens will be tested: dosages of 0.05, 0.1, 0.2, 0.3 mg/kg followed
by three further doses at the same dosage at 12-h intervals. Efficacy will be
evaluated with continuous EEG monitoring.

If seizure burden is not reduced by >80% after the firts dose of bumetanide,
the rescue medication can be given according to the preferences of each
participating recruitment centre. There should be however a delay of at least
two hours before the rescue medication is given, to allow time to see a
potential effect of bumetanide. Only when a status epilepticus is present, a
rescue medication may be given earlier. However it is recommended to use either
midazolam or lignocaine (see section 9.12 Rescue Medication for dose
recommendations).
Due to a relatively low incidence of hypoxic encephalopathy and seizure it is
necessary to recruit form several trial sites, hence this will be a multicentre
clinical trial.

Sample Size Determination

Infants will be studied in nine centres in Europe (Utrecht, Rotterdam, Cork,
Uppsala, Stockholm, Paris, London, Leeds, Helsinki)
• Stage 1 (Dose-finding and confirmatory stage): 24 patients
• Stage 2 (PK stage): Minimum 25 patients

In the dose-finding and confirmatory stage, cohorts of two consecutive patients
will receive the same dose regimen, as determined by the statistician on the
basis of the preceding cohort results. The primary efficacy endpoint will be
the reduction of seizure burden of >=80% on EEG after phenobarbitone and
bumetanide treatment defined as >= 80% reduction of seizure burden within hours
3 and 4 after the first bumetanide administration, compared to the baseline; a
2 hour epoch immediately prior to the first Bumetanide administration.
A decision rule will be set up before initiating the trial in order to avoid
further exposure to side effects in subsequent patients and consisting in
situations where the calculation would lead to increase the dose either to
maintain the previous dose or to decrease it according to a pre-established
intensity simple 3-steps ranking of the side effect of interest.
In the PK stage, patients will receive bumetanide at the optimal dose
established in the dose-finding and confirmatory stage

Trial design

Dose-finding and confirmatory design

The primary aim of the trial is to determine the efficacy and toxicity of four
dose levels of bumetanide (0.05 -0.3 mg/kg).
A dose-finding combined phase I/IIa open sequential trial will be performed.
Patient response and tolerance will be jointly modelled as dual binary
endpoints. The recommended dose levels for future experimentation will the one
satisfying both efficacy criteria and toxicity restriction.
The trial will consist of the sequential treatment of groups of two patients
until a total number of at least 24 patients had been reached. The first group
will be treated at the first dose level (i.e. 0.05 mg/kg), whereas the doses
levels for the subsequent groups will be determined according to the model
estimates of the dose-efficacy and dose-toxicity relationships.
At each dose level, data will be recorded in a computer program that performed
an analysis of both efficacy and safety. The trial is planned to end before the
fixed maximal number of patients, if all the dose levels had an efficacy lower
than the defined target, and/or if all the dose levels had toxicity higher than
the target (Figure 1). An underlying mathematical model expressed the
probabilities of response and tolerance as independent functions of dose
(O*Quigley et al. 2001, Zohar and O*Quigley 2006).
The treatment allocation will be performed by the eCRF following this
methodology base on safety and efficacy outcomes from previous patients. This
means that it is crucial to enter efficacy and safety criteria for all babies
in real time. To secure the process, an e-mail will be automatically sent by
the server (ClinInfo) specifying the current treatment allocation and the next
treatment allocation depending on the various simulated hypothesis.

All infants recruited for the study will be admitted to a neonatal intensive
care unit and will be monitored continuously for most vital signs, such as
respiratory rate, heart rate and bloodpressure. EEG will be used to monitor
brain activity continuously as well. Compared to care given before the study
period, no additional monitoring will be used and no invasive tests will be
performed. Blood samples will be taken, but as all children will have an
arterial line inserted while being in the NICU, this will be not an
invasive/painful procedure. The bloodsamples will be small ( 4 x 0.5 mL) and as
the infants will be full term, the amount sampled is not likely to lead to
additional blood transfusions.
The drug that will be used, has been used in neonates as a diuretic, especially
in the United States. Side effects that can occur in adults are • Muscle cramps
(seen in 1.1% of treated patients)
• Dizziness (1.1%)
• Hypotension (0.8%)
• Headache (0.6%)
• Nausea (0.6%)
• Encephalopathy in patients with pre-existing liver disease (0.6%)
• Dehydration
• Tachycardia

Note: assessment of the symptoms and signs listed above will be limited to
quantifiable signs in the neonatal patient; i.e. dehydration and secondary
hypotension and tachycardia.
The benefit of particpation in the study is that infants will be monitored
carefully for neonatal seizures and these will be treated. In our unit this is
not really different from the care at present and the benefit may be that
bumetanide will be more effective than drugs used at present in this group of
infants.