Total akinesia and trembling in place were phenomeno-logically different and distinct. Trembling in place wasaccompanied by acceleration of the cadence combinedwith a reduction of movement (festination).9,10 Totalakinesia was characterized by impairment of the heel-offmovement. Low doses of caffeine could improve theinability of heel-off of the first swing leg at gait initia-tion. However, the inability to regulate the stride-to-stride variations and impairment of toe-off at the stance–swing transition were not influenced by caffeine.

Caffeine and selective A2A receptor antagonists pro-duced locomotor stimulation in mice lacking the D2 recep-tors.11 Recently, a selective A2A receptor antagonist KW-6002 has been focused on as a novel nondopaminergicapproach to PD therapy and could prolong the efficacyhalf-time of levodopa.5 In animal studies, chronic exposureto caffeine results in tolerance to the motor-activating effectof caffeine and the selective A2A receptor antagonist.4,6 Inthe present study, consistent with previous animal studies,chronic administration of caffeine developed the toleranceto the effects on FOG within a few months. In addition, 150ml of coffee contains 100 mg of caffeine, and caffeineconsumption from all sources is more than 200 mg/day inthe United States and Europe.12 Therefore, it is possible thatchronic caffeine exposure may develop tolerance to theeffect of caffeine and may influence the motor effects of aselective A2A receptor antagonist.

In conclusion, this study revealed that different sub-types of FOG showed different therapeutic responses tocaffeine. Low doses of caffeine could improve the in-ability of heel-off at gait initiation in patients with totalakinesia, but chronic administration of caffeine devel-oped tolerance to the beneficial effect of caffeine on“total akinesia” FOG.

REFERENCES

1. Schaafsma JD, Balash Y, Gurevich T, Bartels AL, Hausdorff JM,Giladi N. Characterization of freezing of gait subtypes and theresponse of each to levodopa in Parkinson’s disease. Eur J Neurol2003;10:391–398.

2. OuchiY, Kanno T, Okada H, et al. Changes in dopamine avail-ability in the nigrostriatal and mesocortical dopaminergic systemsby gait in Parkinson’s disease. Brain 2001;124:784–792.

3. Le Moine C, Svenningsson P, Fredholm BB, Bloch B. Dopamine-adenosine interactions in the striatum and the globus pallidus:inhibition of striatopallidal neurons through either D2 or A2Areceptors enhances D1 receptor-mediated effects on c-fos expres-sion. J Neurosci 1997;17:8038–8048.

4. Quarta D, Ferre S, Solinas M, et al. Opposite modulatory roles foradenosine A1 and A2A receptors on glutamate and dopaminerelease in the shell of the nucleus accumbens. Effects of chroniccaffeine exposure. J Neurochem 2004;88:1151–1158.

5. Bara-Jimenez W, Sherzai A, Dimitrova T, et al. Adenosine A(2A)receptor antagonist treatment of Parkinson’s disease. Neurology2003;61:293–296.

6. Svenningsson P, Nomikos GG, Fredholm BB. The stimulatoryaction and the development of tolerance to caffeine is associated

with alterations in gene expression in specific brain regions. J Neu-rosci 1999;19:4011–4022.

7. Langston JW, Widner H, Goetz CG, et al. Core assessment pro-gram for intracerebral transplantations (CAPIT). Mov Disord1992;7:2–13.

8. Giladi N, Shabtai H, Simon ES, Biran S, Tal J, Korczyn AD.Construction of freezing of gait questionnaire for patients withParkinsonism. Parkinsonism Relat Disord 2000;6:165–170.

9. Nieuwboer A, Dom R, De Weerdt W, Desloovere K, Janssens L,Stijn V. Electromyographic profiles of gait prior to onset of freez-ing episodes in patients with Parkinson’s disease. Brain 2004;127:1650–1660.

10. Ueno E, Yanagisawa N, Takami M. Gait disorders in parkinson-ism: a study with floor reaction forces and EMG. Adv Neurol1993;60:414–418.

11. Aoyama S, Kase H, Borrelli E. Rescue of locomotor impairment indopamine D2 receptor-deficient mice by an adenosine A2A recep-tor antagonist. J Neurosci 2000;20:5848–5852.

12. Fredholm BB, Battig K, Holmen J, Nehlig A, Zvartau EE. Actionsof caffeine in the brain with special reference to factors thatcontribute to its widespread use. Pharmacol Rev 1999;51:83–133.

Orofacial Automatisms Inducedby Acute Withdrawal from

High-Dose Midazolam MimickingNonconvulsive StatusEpilepticus in a Child

David Epstein, MD,1 and Marc Difazio, MD2*1National Naval Medical Center, Bethesda, Maryland;2Uniformed Services University of the Health Sciences,

Bethesda, Maryland

Abstract: Nonconvulsive Status Epilepticus (NCSE) is notuncommon in children, and can be challenging to diagnoseand treat. Etiologies vary widely and include infection,trauma and acute withdrawal from medications such asanticonvulsants. We report a child who experienced orofa-cial dyskinesias concerning for NCSE after withdrawal

This article includes Supplementary Video, available online at http://www.interscience.wiley.com/jpages/0885-3185/suppmat

*Correspondence to: Dr. Marc Difazio, 9715 Medical Center Drive,Suite 233, Rockville, MD 20850. E-mail: mddifazio@aol.com

The views expressed in this article are those of the authors and do notreflect the official policy of the Department of the Navy, Department ofDefense, or U.S. Government.

Received 5 April 2006; Revised 15 August 2006; Accepted 16August 2006

Published online 20 March 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mds.21260

712 EPSTEIN AND DIFAZIO

Movement Disorders, Vol. 22, No. 5, 2007

from high dose benzodiazepines andopiates. Automonicsigns typically associated with sedative withdrawal wereabsent and treatment with benzodiazepines did not improvehis symptoms. Diagnostic testing was negative, includingelectroencephalogram, and resolution was complete withinfive days. Our case demonstrates the orofacial dyskinesiasthat may occur during sedative medication withdrawal, andhighlights potential confusion with non-convulsive statusepilepticus. © 2007 Movement Disorder Society

Key words: fentanyl; midazolam; dyskinesias; nonconvul-sive status epilepticus; withdrawal

Nonconvulsive status epilepticus (NCSE) is not un-common in children, although its actual incidence re-mains unknown.1 It remains difficult to diagnose, andbecause clinical presentations may be variable, it is oftennot recognized immediately.2 Most commonly, it occursin those with preexisting epilepsy and subsequent toclinically overt status epilepticus. It may also incur incomatose patients with significant medical or neurologicdisorders.2,3 It is also a recognized complication of sed-ative medication withdrawal, most notably benzodiaz-epines.4 In some situations, recognition of NCSE is fur-ther complicated by concurrent illnesses that producevery similar clinical characteristics. Because of this find-ing, an electroencephalogram is needed for confirmationof NCSE and to determine response to treatment, espe-cially in children who are critically ill or receiving med-ications that alter the level of consciousness. We report a4-year-old child with acute onset of mental statuschanges, in association with orofacial automatisms, sub-sequent to acute withdrawal of high-dose midazolam andnarcotics. These movements were clinically consistentwith complex partial seizures, indicating a potential forconfusion with nonconvulsive status epilepticus afteracute withdrawal of benzodiazepines in children.

CASE REPORT

A previously healthy 4-year-old boy was admitted tothe intensive care unit (ICU) with a diagnosis of epiglot-titis. Initial manifestations included fever, pain, anddrooling. He underwent elective intubation and mechan-ical ventilation for airway protection. Sedation was ini-tiated with midazolam and fentanyl infusions. Agitationand excessive movement necessitated an escalation ofsedative medications to adequately protect the airwayand lessen orotracheal trauma. The initial midazolaminfusion was maintained at a rate of 0.5 mg/kg per hr for3 days along with a fentanyl infusion at up to 8 �g/kg perhour, and later a morphine infusion at up to 0.8 mg/kgper hour. There were also intermittent additional doses offentanyl, morphine, and chlorohydrate. Cumulativedoses of 26 mg/kg of midazolam, 130 �g/kg of fentanyl,

and 18 mg/kg of morphine were eventually delivered.The patient remained intubated and sedated for 4 days.

Medications were discontinued over approximately 12hours, and the patient was extubated. After extubationand transfer from the pediatric ICU (PICU), a lucidperiod was observed, with normal play, speech, andparental interaction. Within several hours, however, hebegan to experience agitation, punctuated by periods ofsleep. Upon awakening, the child appeared alert but didnot respond to verbal commands. There were frequentintermittent tongue protrusion, inappropriate laughter,and lipsmacking behaviors. There were also occasionaltruncal myoclonic jerks. The neurologic exam at thattime was otherwise nonfocal. An electroencephalogramobtained at this time demonstrated diffuse slowing andtriphasic waves, without epileptiform activity (Fig. 1).Due to the high degree of clinical suspicion for NCSE, atest dose of 1 mg of lorazepam was given without clin-ical effect. Subsequent evaluation with brain magneticresonance imaging and lumbar puncture were also unre-markable. The following 24 hours were characterized bycontinued encephalopathy and insomnia. Additional be-haviors were also noted, including upper extremity ste-reotypies, apparent visual hallucinations, and continuedorofacial dyskinesias (see Video). During this period,there were intermittent periods of more appropriate re-sponsiveness and lucidity, which were characterized byincreased interaction with parents and appropriatespeech. He returned to baseline slowly over the next 4 to5 days without further treatment. He experienced noresidual neurologic deficits or involuntary movements,and he has since returned to his premorbid level offunctioning.

DISCUSSION

The manifestations of NCSE are protean, with a vari-ety of behavioral syndromes associated with varyingdegrees of alteration of consciousness. Reported findingsinclude delirium, staring, inappropriate laughter, crying,and oral automatisms.5 A high index of suspicion is,therefore, required for diagnosis, and familiarity withclinical conditions that may predispose to this entity iscrucial for rapid recognition. These conditions includeconvulsive status epilepticus, which not uncommonlyprogresses to NCSE.3 Importantly, unrecognized NCSEmay last for hours or days and has been associated withsignificant morbidity or mortality.3,6 Other neurologicconditions may predispose to seizure or even NCSE,including central nervous system trauma and postopera-tive neurosurgical care. These conditions often requirecontinuous intravenous sedatives or narcotics for seizurecontrol, agitation, pain relief, or control of intracranial

ACUTE WITHDRAWAL FROM MIDAZOLAM MIMICKS NCSE 713

Movement Disorders, Vol. 22, No. 5, 2007

pressure. Additionally, significant increases in the dosingof sedative medications may be necessary, due to therapid development of tolerance. Subsequent to high-dosesedative infusions, symptoms of withdrawal are a com-mon occurrence and are estimated to be as high as 35%in the pediatric intensive care setting.7 Additionally,acute withdrawal from midazolam therapy has been re-ported to occur in up to 17% of children followingprolonged sedation in an ICU setting.8 Symptoms ofwithdrawal include agitation, anxiety, muscle twitching,sweating, and tremor. However, in contrast to more overtabstinence syndromes, more subtle behavioral manifes-tations have also been attributed to benzodiazepine ornarcotic withdrawal.8–10 Typically, these manifestationsoccur within 24 hours of drug discontinuation and in-clude visual and auditory hallucinations, language im-pairments, inappropriate laughter, and facial grimac-ing.8,11 More complex choreoathetotic and dyskineticmovements including repetitive tongue thrusting, chew-ing movements, and stereotyped movements of the ex-tremities have also been reported.8,10 In all of the cases,there was a return to normal neurologic function. Aspecific inciting drug dose or combination has not beenidentified, although the majority of reported patientshave received high cumulative doses of the benzodiaz-epine, often in conjunction with fentanyl. Hughes andcolleagues reported typical doses of midazolam of 0.9 to

39 mg/kg, with maximum infusion rates from 48 to 720�g/kg per hour and durations of infusion from 22 to 356hours.8 Lane and associates reported 5 patients whomanifested acute onset myoclonus, ataxia, and choreo-athetosis after the withdrawal of prolonged sedation withfentanyl, without concurrent benzodiazepine dosing.12

Of interest, the characteristic autonomic symptoms ofopioid abstinence were absent, suggesting that the with-drawal of fentanyl may be an important component in themore complex behavioral manifestations. At present, theunderlying pathophysiology remains unclear. Our patientmanifested prominent orofacial automatisms, stereo-typed movements of the extremities and behavioralchanges, in the context of high-dose benzodiazepine/narcotic withdrawal. There were none of the more com-mon signs of abstinence syndromes. All of his behavioralmanifestations were consistent with previous reports ofbizarre behavior during withdrawal of these agents.However, his movements were also very concerning forcomplex partial status epilepticus. In fact, the degree ofclinical suspicion was so high that, despite the absence ofcharacteristic electroencephalogram findings, he re-ceived empiric treatment for NCSE. No other infectiousor traumatic etiology was evident during the investiga-tion for his behaviors, and recovery over days withoutspecific intervention or residua made an encephalitic orischemic cause less likely.

FIG. 1. The electroencephalogram obtained during ongoing orofacial automatisms. Diffuse slowing with triphasic waves only are evident. No changeoccurred subsequent to lorazepam administration.

714 EPSTEIN AND DIFAZIO

Movement Disorders, Vol. 22, No. 5, 2007

CONCLUSION

In conclusion, withdrawal from midazolam and pos-sibly other sedative agents is associated with a distinctclinical syndrome closely mimicking NCSE in children.Given that high-dose benzodiazepines are often used forstatus epilepticus, either subtle or nonconvulsive, clinicalconfusion is a distinct possibility if these manifestationsoccur in the setting of medication tapering. Autonomicmanifestations may be absent, further confusing the clin-ical picture. Recognition of the syndrome and urgentdifferentiation from NCSE has important managementand prognostic implications. Once the diagnosis is madeof a nonepileptic movement disorder, expectation can befor full recovery without intervention or treatment.

LEGEND TO THE VIDEO

The patient is noted to have nearly constant orofacialautomatisms, characterized as tongue thrusting, chewingmovements and grimacing.

REFERENCES

1. Stores G. Nonconvulsive status epilepticus in children. In: PedleyT, Meldrum B, editors. Recent advances in epilepsy. Edinburgh:Churchill Livingstone; 1986. p 295–310.

2. Stores G, Zaiwalla Z, Styles E, Hoshika A. Non-convulsive statusepiletpicus. Arch Dis Child 1995;73:106–111.

3. Delorenzo R, Waterhouse E, Towne A, et al. Persistent noncon-vulsive status epilepticus after the control of convulsive statusepilepticus. Epilepsia 1998;39:833–840.

4. Sury M, Billingham I, Russell G, Hopkins C, Thorington R, VivoriE. Acute benzodiazepine withdrawal syndrome after midazolaminfusions in children. Crit Care Med 1989;17:301–302.

5. Kaplan P. Nonconvulsive status epilepticus in the emergencyroom. Epilepsia 1996;37:643–650.

6. Krumholz A. Epidemiology and evidence for morbidity of non-convulsive status epilepticus. J Clin Neurophysiol 1999;16:314–322.

7. Fonsmark L, Rasmussen Y, Carl P. Occurrence of withdrawal incritically ill sedated children. Crit Care Med 1999;27:196–199.

8. Hughes J, Gill A, Leach H. A prospective study of the adverseeffects of midazolam on withdrawal in critically ill children. ActaPaediatr 1994;83:1194–1199.

9. Ducharme M, Munzenberger P. Severe withdrawal syndrome pos-sibly associated with cessation of a midazolam and fentanyl infu-sion. Pharmacotherapy 1995;15:665–668.

10. van Engelen B, Gimbrere J, Booy L. Benzodiazepine withdrawalreaction in two children following discontinuation of sedation withmidazolam. Ann Pharmacother 1993;27:579–581.

11. Bergman I, Steeves M, Burckhart G, Thompson A. Reversibleneurologic abnormalities associated with prolonged intravenousmidazolam and fentanyl administration. J Pediatr 1991;119:644–649.

12. Lane J, Tennison M, Lawless S, Greenwood R, Zaritsky A. Move-ment disorder after withdrawal of fentanyl infusion. J Pediatr1991;199:649–651.

Antiparkinsonian Activity of L-Propyl-L-leucyl-glycinamide orMelanocyte-Inhibiting Factor in

MPTP-Treated CommonMarmosets

Regina Katzenschlager, MD,1,2

Michael J. Jackson, BSc,3 Sarah Rose, PhD,3

Kim Stockwell, PhD,3

Kayhan A. Tayarani-Binazir, BSc,3

Mohammed Zubair, BSc,3 Lance A. Smith, PhD,3

Peter Jenner, DSc,3 and Andrew J. Lees, MD, FRCP1*1Reta Lila Weston Institute of Neurological Studies,

University College London, United Kingdom and NationalHospital for Neurology and Neurosurgery, London,

United Kingdom; 2Department of Neurology, Donauspital/SMZ-Ost, Vienna, Austria; 3Neurodegenerative Diseases

Research Centre, King’s College London, United Kingdom

Abstract: The neuropeptide melanocyte-inhibiting factor(MIF) or L-propyl-L-leucyl-glycinamide (PLG) has been re-ported in some studies to improve the motor signs of Parkin-son’s disease (PD) and in rodent models of PD. In this study oforal and intravenous MIF in N-methyl-4-phenyl-1,2,3,6-tetra-hydropyridine (MPTP)-lesioned marmosets, a wide range ofdoses of MIF administered alone (0.25, 1, 2, 5, 10, 20 mg/kgorally) did not increase locomotor activity, relieve motor dis-ability, or induce dyskinesias. When MIF (1.0 and 5.0 mg/kgorally or 10 and 20 mg/kg intravenously) was administeredconcomitantly with levodopa/benserazide, no significant dif-ferences in motor function or dyskinesias were observed com-pared with levodopa/benserazide alone. The results of this firststudy of MIF in the marmoset MPTP model provide no en-couragement for the reinvestigation of MIF in the clinicalmanagement of the motor signs of PD. © 2007 MovementDisorder Society

Key words: Parkinson’s disease; dyskinesia; marmoset; me-lanocyte-inhibiting factor (MIF)

L-propyl-L-leucyl-glycinamide (PLG) or melanocytestimulating hormone release-inhibiting factor (MIF) is ahypothalamic tri-peptide that controls the release of mel-anocyte-stimulating hormone (MSH). Circulating levelsof MSH were found to be elevated in patients withParkinson’s disease (PD),1 and MSH has been observed

*Correspondence to: Dr. Andrew J. Lees, Reta Lila Weston Instituteof Neurological Studies, Windeyer Building, 46 Cleveland Street,London W1 T 4JF, United Kingdom. E-mail: a-lees@ion.ucl.ac.uk

Received 19 May 2006; Revised 1 August 2006; Accepted 19August 2006

Published online 20 March 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mds.21256

ANTI-PD ACTIVITY OF PLG OR MIF IN MARMOSETS 715

Movement Disorders, Vol. 22, No. 5, 2007