Full text of DOI:10.1592/phco.21.6.642.34545

_
C A S E R EP RTS
Flumazenil’s Reversal of Myoclonic-like Movements
Associated with Midazolam in Term Newborns
Win Zaw, MBBS, David C Knoppert, M.Sc.Phm., and Orlando da Silva, M.D.
Sedation is an important aspect of care for critically ill newborns. Proper
sedation reduces stress during procedures such as mechanical ventilation.
Midazolam, a short-acting benzodiazepine, is widely administered as a
sedative in newborn intensive care units but is not without side effects. Three
term newborns developed myoclonic-like abnormal movements after
receiving midazolam. In one, flumazenil controlled the abnormal movements.
Flumazenil is a potent benzodiazepine antagonist that competitively blocks
the central effects of benzodiazepines. It can reverse the sedative effects of
benzodiazepines occurring after diagnostic or therapeutic procedures or after
benzodiazepine overdose. Flumazenil may be considered in cases of abnormal
movements associated with midazolam. However, further studies are needed
to provide guidelines for the administration of this drug in newborns.
(Pharmacotherapy 2001;21(5):642–646)
reaction.13 In 76 premature infants, 6 (7.9%)
developed myoclonic activity during a continuous
infusion of midazolam.11 In a smaller series, 3 of
10 premature infants developed myoclonic
activity after an intravenous bolus of midazolam.12
Sedative and analgesic drugs are administered
to newborns experiencing pain and stress during
mechanical ventilation.1, 2 These drugs also are
used as preoperative agents3 and for various
procedures where sedation is required.4, 5 The
most common drugs are benzodiazepines
(midazolam, diazepam, and lorazepam), chloral
hydrate, and opioids (fentanyl and morphine).6
Midazolam is a water-soluble benzodiazepine
with sedative, amnesic, anxiolytic, muscle-
relaxant, and anticonvulsant properties.7 It has a
faster onset and shorter duration of action than
lorazepam.7 However, because of adverse effects
occasionally reported in children and newborns,
there are concerns about its safety. Adverse
effects include hypotension,8, 9 myoclonic
activity,8, 10–12 and a paradoxical behavioral
Case Reports
Patient No. 1
An infant weighing 2750 g was born by normal
vaginal delivery at 38 weeks’ gestation with
Apgar scores of 9 and 9 at 1 and 5 minutes,
respectively. The pregnancy was uneventful. The
infant was transferred to our institution within
24 hours of birth because of episodes of apnea
and bradycardia, which resolved later. At 4 days
of age, a magnetic resonance imaging (MRI) scan
was done to exclude intracranial abnormalities.
The infant was given an intravenous bolus of
midazolam 160 µg/kg for sedation before the
procedure. Forty-five minutes after the bolus, he
developed bicycling movements of both legs and
myoclonic-like movements of the upper limbs
that lasted for 90 seconds. After these episodes,
the infant became very irritable and unconsolable
for approximately 10 minutes before falling
From the Department of Child Health, Medical School,
University of Aberdeen, Aberdeen, Scotland (Dr. Zaw); the
Department of Pharmacy, St. Joseph’s Hospital, London,
Ontario, Canada (Mr. Knoppert); and the Division of
Neonatal-Perinatal Medicine, Department of Pediatrics,
University of Western Ontario, London, Ontario, Canada
(Dr. da Silva).
Address reprint requests to David C. Knoppert,
M.Sc.Phm., Department of Pharmacy, St Joseph’s Hospital,
268 Grosvenor Street, London, Ontario, Canada N6A 4V2.

FLUMAZENIL REVERSES MYOCLONIC-LIKE MOVEMENTS AFTER MIDAZOLAM Zaw et al 643
Table 1. Case Reports of Abnormal Movements after Administration of Midazolam
No. of
Gestational
Age (wks)
Birth Weight
(g)
Administration
Method
i.v. bolus
i.v. infusion
i.v. bolus
i.v. bolus
Patients
Dosage
Time of Onseta
Durationb
NP8
< 32
NP
200 µg/kg
30–60 µg/kg/hr
100 µg/kg
15–30 sec
2–48 hrs
Immediately
5 min
NP
611
27–34
25–30
24–26
970–2060
< 1501
A few hrs
5–60 min
5–10 min
312
322
671 ± 170c
100 µg/kg
NP = not provided.
aTime of onset of abnormal movements after administration of midazolam.
bDuration of abnormal movement activity after administration of midazolam.
cData are mean ± SD.
asleep. Investigations showed normal glucose,
calcium, and electrolytes. The MRI, an
electroencephalogram (EEG), and an auditory
brain stem evoked response (the latter two done
the day after the MRI) showed no abnormalities.
The only other drugs that the patient had
received were intravenous ampicillin and
gentamicin. No more midazolam was given, and
no further abnormal movements were noted. At
10 days of age, the infant was well and was
discharged.
intermittent bitemporal spikes of uncertain
significance. The patient did not receive any
other drugs. No more midazolam was given, and
no further abnormal movements were noted.
Patient No. 3
A baby girl weighing 2045 g was delivered by
spontaneous vaginal delivery at 38 weeks’
gestation. Before birth, a left-sided diaphragmatic
hernia and abnormal karyotype with terminal
deletion in the short arm of chromosome 8 had
been diagnosed. At birth, the infant’s Apgar
scores were 5 and 7 at 1 and 5 minutes,
respectively. The infant was intubated at 2
minutes of age and kept on a high frequency
oscillator. Initially, she was paralyzed with a
vecuronium infusion for 10 days and sedated
with a morphine infusion. Primary repair was
done successfully on day 7 with a silastic patch.
On day 12, a continuous intravenous infusion
of midazolam 60 µg/kg/hour was started in
addition to the ongoing morphine infusion.
Within an hour, the infant developed myoclonic
jerks of the extremities. The midazolam infusion
was stopped, and an intravenous bolus of
flumazenil 7.8 µg/kg was given. Abnormal
movements ceased within 5 minutes. Other
drugs at that time were morphine infusion 40
µg/kg/hour, heparinized saline (heparin 1 U/ml in
0.9% sodium chloride), and total parenteral
Patient No. 2
A full-term infant weighing 3770 g was born at
41 weeks’ gestation by spontaneous vaginal
delivery after an uncomplicated pregnancy.
Apgar scores were 7 and 9 at 1 and 5 minutes,
respectively. At 26 minutes of age, he was
intubated for episodes of apnea and cyanosis. He
then experienced generalized seizures. Diazepam
0.3 mg/kg was given intravenously. When
seizures persisted, a loading dose of intravenous
phenobarbital 20 mg/kg was added, with a
maintenance dosage of 5 mg/kg every 12 hours.
An EEG that day showed probable epileptiform
discharges in the frontal, left temporal, and right
midtemporal regions. Other investigations—
urea, electrolytes, glucose, calcium, magnesium,
cerebrospinal fluid, ammonia, amino acids,
lactate, long-chain fatty acid, and computerized
tomography of the head—did not reveal any
cause for seizures. Phenobarbital was stopped on
day 4 of age.
nutrition.
Arterial blood gases, serum
electrolytes, glucose, calcium, and magnesium
were within normal limits.
On day 6 of age, the infant was given
midazolam 500 µg/kg orally for sedation before
an MRI. Thirty minutes later, he became
increasingly agitated. He then displayed
myoclonic-like movements of both upper and
lower extremities. These abnormal movements
were controlled successfully with intravenous
phenobarbital 20 mg/kg. An EEG failed to show
seizure activity, with the exception of
Discussion
These three cases highlight the paradoxical
effects of midazolam in term newborns.
Benzodiazepines are known to cause paradoxical
stimulation of the central nervous system and
thus may increase the frequency of nightmares,
euphoria, restlessness, hallucinations, and

644
PHARMACOTHERAPY Volume 21, Number 5, 2001
hypomanic behavior in adults14 and children.13, 15
Anticonvulsant benzodiazepines can also increase
the frequency of seizures in patients with
epilepsy.16 In fact, lorazepam has caused seizures
in newborns.17–21
parameters after intravenous infusion of
midazolam.24, 26–28 Mean clearance rate is lower
in newborns younger than 39 weeks’ gestation
(1.17 ml/kg/min) than in those older than 39
weeks (1.84 ml/kg/min).24 Moreover, infants
weighing less than 1000 g at birth have
significantly lower clearances than those
weighing more.28 The low clearance rate is
probably related to immature hepatic enzyme
activity rather than differences in renal
function.24 Therefore, individual dosage
adjustment may be required.27
Benzodiazepines, including midazolam, exert a
clinical effect by binding to ␥-aminobutyric acid
A (GABAA) receptors in the nervous system. The
GABAA receptors are a family of ligand-gated ion
channels that mediate the principal actions of
GABA, the principal inhibitory neurotransmitter.
Benzodiazepine-binding regions may be
homologous to agonist-binding regions but are
situated at different subunit interfaces.29 In a
hypothetical model of the GABAA receptor, there
are agonist (for GABA) and modulatory (for
benzodiazepine-type ligands) binding sites.30
The receptor is composed of five subunits (2-1,
Abnormal movements after midazolam
administration in preterm infants (Table 1)8, 11, 12,
22
and adults have been reported.23 Involuntary
epileptiform movements lasting 15–30 seconds in
premature newborns with gestational ages under
32 weeks appeared after intravenous boluses of
midazolam 200 µg/kg.8 Myoclonus in six
premature infants (gestational ages 27–34 weeks)
was observed after intravenous infusions of
midazolam in dosages ranging from 30–60
µg/kg/hour.11 In these infants, myoclonus was
noticed 2–48 hours after beginning the
midazolam infusion, ceased a few hours after
discontinuing the infusion, and never recurred.
Three infants with very low birth weights
(average 671 ± 170 g) experienced accentuated
myoclonic jerks resembling clonic seizures
within 5 minutes of receiving slow bolus
administration of midazolam 100 µg/kg for
sedation.22 In another study, 3 of 10 preterm
newborns (25–30 weeks’ gestation) with birth
weights under 1500 g developed myoclonus after
midazolam administration.12
2
␤1 and 1␥2 peptides), which are arranged around
a central chloride ion-selective pore that opens as
a consequence of agonist (GABA) binding. These
receptors are found in the greatest density in the
olfactory bulbs, cerebral cortex, cerebellum,
hippocampus, substantia nigra, and inferior
colliculus.31
Benzodiazepines and GABA together regulate
the opening of chloride channels in the cell
membrane.32 Stimulation of this ionophore has
the effect of hyperpolarizing postsynaptic
neurons. These hyperpolarized cells are more
difficult to depolarize in response to excitatory
neurotransmitters.7
Myoclonus was observed after the use of
midazolam in 11 neonatal intensive care units in
France.24 In these units a loading dose of
100–200 µg/kg was administered followed by a
continuous infusion of 30–200 µg/kg/hr. No
subsequent neurologic abnormality was seen in
any of the infants.
Another study11 recorded EEG patterns in six
preterm infants who developed myoclonus after
midazolam infusion. Investigators did not
observe ictal activity in these infants. Five of the
six displayed excessive discontinuity on their
EEGs. Disappearance of physiologic fluctuation
and excessive theta waves were evident in three
patients. An EEG performed on our second
patient during the episode of abnormal
movements showed no ictal activity.
The mechanism responsible for abnormal
movements after midazolam administration is
unclear. Adverse reactions may result from
inadequate or excessive dosages, improper
administration, cerebral hypoxia, or they may be
simply paradoxical reactions.25
␥-Aminobutyric acid
synthesis in
experimental animals is low during the neonatal
period. The activity of the synthetic enzyme for
GABA, L-glutamate decarboxylase, in the cortex
of a newborn rodent is low but rises rapidly to
adult levels by 3 weeks of age.33 Also, binding
sites for GABA are present at relatively low
concentrations at birth and increase slowly in the
postnatal period.34 Therefore, there may be less
neuronal inhibition in children than would be
expected in adults. In contrast, benzodiazepine
receptors are found at relatively high levels in
newborns.32 It is unclear if this imbalance
between relatively low numbers of GABA
receptors (agonist) and relatively high levels of
benzodiazepine receptors (modulators) create an
Dosages used in our patients were within the
recommended range, and the drug was properly
administered. Among newborns, however, large
interpatient variability exists in pharmacokinetic

FLUMAZENIL REVERSES MYOCLONIC-LIKE MOVEMENTS AFTER MIDAZOLAM Zaw et al 645
behavioral reactions in children.13
excitatory, rather than inhibitory, influence.
Midazolam is highly protein bound (96%).
Low serum albumin predisposes newborns to
potentially elevated serum concentrations of free
midazolam. Neurologic sequelae after prolonged
intravenous administration of midazolam in
infants were significantly associated with low
serum albumin concentrations.35 Therefore, the
uninhibited movement of free, unbound drug
across the blood-brain barrier may be related to
the paradoxic stimulation of neural synapses in
newborns who already have decreased GABA
levels.
It also was suggested that myoclonus-like
abnormal movements may be related to
hemodynamic changes that may affect cerebral
blood flow.8, 22 Hypotension is a well-known side
effect of midazolam, particularly in premature
infants. The mechanisms responsible for
hypotension are not clear, but they might involve
peripheral vasodilatation, a sudden decrease in
circulating catecholamines, or direct myocardial
depression.36 In ventilated newborns, midcerebral
blood flow velocity decreased by 12–43% after
transient drops in systemic blood pressure after
midazolam administration.10, 12, 37 However, as in
an earlier study,22 we did not observe clinically
significant changes in blood pressure.
In our first case, the infant fell asleep within
approximately 10 minutes of the onset of
myoclonic-like movements. Intravenous
phenobarbital was used successfully to treat the
abnormal movements in our second patient.
Flumazenil was used successfully in our third
patient to reverse the myoclonic effects of
midazolam. There was no clinical or pharmaco-
logic preference between the two agents. We are
not suggesting that flumazenil be used based on
one case report; we are simply reporting our
initial experience with this drug in a newborn.
Further studies will provide guidance as to
whether phenobarbital, flumazenil, or, possibly,
no pharmacologic intervention is best in specific
clinical situations.
Flumazenil is a potent antagonist that displaces
benzodiazepines linked to receptors and blocks
their central effects.44 We observed that
abnormal movements stopped within 5 minutes
of flumazenil injection. One study13 reported
that the amount of time to reverse the
paradoxical behavior reaction in children with
flumazenil was variable (3–50 min) and was
dependent on the dosage of midazolam. One
case report noted that the paradoxical reaction
improved within 1 minute of flumazenil
administration in an adult.45 Although these are
anecdotal case reports, they suggest a role for
flumazenil in the control of myoclonus induced
by midazolam. It should be emphasized that
flumazenil can cause seizures in children,46, 47
therefore, caution needs to be exercised in its use.
We used an intravenous bolus of flumazenil 7.8
µg/kg arbitrarily based on the dose recommended
for managing benzodiazepine overdose in
children (10 µg/kg).48
In summary, we have described three cases of
midazolam-induced, myoclonic-like abnormal
movements in term infants. Flumazenil was
effective in reversing these abnormal movements
in one infant, indicating that flumazenil may be
considered in cases of abnormal movements
associated with midazolam. However, further
studies are needed to provide guidelines for the
administration of this drug to newborns.
References
1. Rosen DA, Rosen KR. Midazolam for sedation in the pediatric
intensive care unit. Intensive Care Med 1991;17(suppl
1):S15–19.
2. Notterman DA. Sedation with intravenous midazolam in the
pediatric intensive care unit. Clin Pediatr 1997;36:449–54.
3. Holloway AM, Jordaan DG, Brock-Utne JG. Midazolam for the
intravenous induction of anaesthesia in children. Anaesth
Intens Care 1982;10:340–3.
4. Diament MJ, Stanley P. The use of midazolam for sedation of
infants and children. Am J Roentgenol 1988;150:377–8.
5. Rosen DA, Rosen KR. Intravenous conscious sedation with
midazolam in paediatric patients. Int J Clin Pract
1998;52:46–50.
6. Wolf AR. Neonatal sedation: more art than science. Lancet
1994;344:628–9.
Flumazenil was approved by the United States
Food and Drug Administration in 1991 for use in
adults for complete or partial reversal of the
sedative effects of benzodiazepines occurring
after diagnostic or therapeutic procedures or after
benzodiazepine overdose.38 It was administered
in children to reverse the sedative action of
midazolam.38–40 Flumazenil also reversed
diazepam-induced neonatal apnea and
hypotonia,41, 42 midazolam-induced laryngospasm
in adults,43 and midazolam-induced paradoxical
7. Blumer JL. Clinical pharmacology of midazolam in infants and
children. Clin Pharmacokinet 1998;35:37–47.
8. van den Anker JN, Sauer PJ. The use of midazolam in the
preterm neonate [letter]. Eur J Pediatr 1992;151:152.
9. Burtin P, Daoud P, Jacqz-Aigrain E, Mussat P, Moriette G.
Hypotension with midazolam and fentanyl in the newborn.
Lancet 1991;337:1545–6.
10. van Straaten HL, Rademaker CM, de Vries LS. Comparison of
the effect of midazolam or vecuronium on blood pressure and
cerebral blood flow velocity in the premature newborn. Dev
Pharmacol Ther 1992;19:191–5.
11. Magny JF, d’Allest AM, Nedelcoux H, Zupan V, Dehan M.
Midazolam and myoclonus in neonate. Eur J Pediatr
1994;153:389–90.

646
PHARMACOTHERAPY Volume 21, Number 5, 2001
12. Harte GJ, Gray PH, Lee TC, Steer PA, Charles BG.
Hemodynamic responses and population pharmacokinetics of
midazolam following administration to ventilated, preterm
neonates. J Pediatr Child Health 1997;33:335–8.
receptors. Trends Pharmacol Sci 1997;18:425–9.
31. Reves JG, Glass PS, Lubarsky DA. Nonbarbiturate intravenous
anaesthetics. In: Miller RD, ed. Anesthesia. New York:
Churchill-Livingstone, 1994:247–89.
13. Massanari M, Novitsky J, Reinstein LJ. Paradoxical reactions
in children associated with midazolam use during endoscopy.
Clin Pediatr 1997;36:681–4.
14. Hall RC, Zisook S. Paradoxical reactions to benzodiazepines.
Br J Clin Pharmacol 1981;11(suppl 1):99S–104.
15. Davies FC, Waters M. Oral midazolam for conscious sedation
of children during minor procedures. J Acc Emerg Med
1998;15:244–8.
16. Hobbs WR, Rall TW, Verdoorn TA. Hypnotics and sedatives;
ethanol. In: Hardman JG, Limbird LE, Molinoff PB, Ruddon
RW, Gilman AG, eds. Goodman and Gilman’s the
pharmacologic basis of therapeutics, 9th ed. New York:
McGraw-Hill, 1996:361–96.
32. Johnston MV, Silverstein FS. Development of
neurotransmitters. In: Polin R, Fox WW, eds. Fatal and
neonatal physiology. Philadelphia: W.B Saunders,
1998:2110–17.
33. Coyle J, Enna S. Neurochemical aspects of the ontogenesis of
GABAnergic neurons in the rat brain. Brain Res 1976;111:
119–33.
34. Johnston MV. Biochemistry of neurotransmitters in cortical
development. In: Peters A, Jones EG, eds. Cerebral cortex. New
York: Plenum Press, 1988:211–36.
35. Bergman I, Steeves M, Burckart G, Thompson A. Reversible
neurologic abnormalities associated with prolonged
intravenous midazolam and fentanyl administration. J Pediatr
1991;119:644–9.
17. Cronin CM. Neurotoxicity of lorazepam in premature infant.
Pediatrics 1992;89:1129–30.
18. Lee DS, Wong HA, Knoppert D. Myoclonus associated with
lorazepam therapy in very low birth weight infants. Biol
Neonate 1994;66:311–15.
36. Heikkila J, Arola M, Kanto J, Laaksonen V. Midazolam as
adjunct to high dose fentanyl anaesthesia for coronary artery
bypass grafting operation. Acta Anaesthesiol Scand
1984;28:683–9.
19. Reiter PD, Stiles AD. Lorazepam toxicity in a premature infant.
Ann Pharmacother 1993;27:727–9.
37. Jacqz-Aigrain E, Daoud P, Burtin P, Desplanques L, Beaufils F.
Placebo-controlled trial of midazolam sedation in mechanically
ventilated newborn babies. Lancet 1994;344:646–50.
38. Shannon M, Albers G, Burkhart K, et al. Safety and efficacy of
flumazenil in the reversal of benzodiazepine-induced conscious
sedation. J Pediatr 1997;131:582–6.
39. Collins S, Carter JA. Resedation after bolus administration of
midazolam to an infant and its reversal by flumazenil.
Anaesthesia 1991;46:471–2.
40. Rodriguez Nunez A, Martinon Sanchez JM, Rivas Pumar P,
Martinez Soto I, Relova Quinteiro JL, Pena Guitian J. Use of
flumazenil to reverse the effects of midazolam in children.
Anales Espanoles de Pediatria 1991;35:332–4.
41. Richard P, Autret E, Bardol J, Soyez C, Barbier P, Jonville AP.
The use of flumazenil in a neonate. J Toxicol Clin Toxicol
1991;29:137–40.
42. Cone AM, Nadel ASB. Flumazenil reverses diazepam-induced
neonatal apnoea and hypotonia [letter]. Eur J Pediatr
1993;152:458–9.
43. Davis DP, Hamilton RS, Webster TH. Reversal of midazolam-
induced laryngospasm with flumazenil. Ann Emerg Med
1998;32:263–5.
44. Votey SR, Boose GM, Bayer MJ, Hoffman JR. Flumazenil: a
new benzodiazepine antagonist. Ann Emerg Med
1991;20:181–8.
45. Rodrigo CR. Flumazenil reverses paradoxical reaction with
midazolam. Anesth Prog 1991;38:65–8.
20. Sexon WR, Thigpen J, Stajich GV. Stereotypic movements
after lorazepam administration in premature infants: a series
and review of the literature. J Perinatol 1995;15:146–9.
21. Chess PR, D’Angio CT. Clonic movements following
lorazepam administration in full-term infants. Arch Pediatr
Adol Med 1998;152:98–9.
22. Waisman D, Weintraub Z, Rotschild A, Bental Y. Myoclonic
movements in very low birth weight premature infants
associated with midazolam intravenous bolus administration
[letter]. Pediatrics 1999;104:579.
23. Engstrom RH, Cohen SE. A complication associated with the
use of midazolam [letter]. Anesthesiology 1989;70:719.
24. Burtin P, Jacqz-Aigrain E, Girard P, et al. Population
pharmacokinetics of midazolam in neonates. Clin Pharmacol
Ther 1994;56:615–25.
25. Anonymous. Benzodiazepines. In: McEvoy GK, ed. AHFS drug
information. Bethesda, MD: American Society of Health-System
Pharmacists, 1998:1934–60.
26. Jacqz-Aigrain E, Wood C, Robieux I. Pharmacokinetics of
midazolam in critically ill neonates. Eur J Clin Pharmacol
1990;39:191–2.
27. Jacqz-Aigrain E, Burtin P. Clinical pharmacokinetics of
sedatives in neonates. Clin Pharmacokinet 1996;31:423–43.
28. Lee TC, Charles BG, Harte GJ, Gray PH, Steer PA, Flenady
VJ. Population pharmacokinetic modeling in very premature
infants receiving midazolam during mechanical ventilation:
midazolam neonatal pharmacokinetics. Anesthesiology
1999;90:451–7.
46. McDuffee AT, Tobias JD. Seizure after flumazenil
administration in a pediatric patient. Ped Emerg Care
1995;11:186–7.
29. Galzi J-L, Changeux J-P. Neurotransmitter-gated ion channels
as unconventional allosteric proteins. Curr Opin Struct Biol
1994;4:554–65.
47. Davis CO, Wax PM. Flumazenil associated seizure in an 11-
month-old child. J Emerg Med 1996;14:331–3.
48. Taketomo CK, Hodding JH, Kraus DM. Pediatric dosage
handbook. Cleveland: Lexi-Comp Inc., 2000–2001:423.
30. Sigel E, Buhr A. The benzodiazepine binding site of GABA(A)