.
`
`.
`
`. '
`
`..
`
`.
`
`·tn.. •.
`. . . ' .•. · .... · .. ·.
`... .................
`.· ' ·~··i';:·.;<~
`*I ME 0\NAOlc\~~''.JeURNAL OF
`~EUROLOGKAL SCIENCES
`tE]OURNAL CANADIEN DES
`~CIENCES NEUROLOGIQUES ·
`
`D'Anvtlle':s Doom - A Neurological Vignette from Historic Halifax
`srep11en F. Bedwell
`late-Lile Migraine Accompaniments as a Cause of Unexplained Transient
`Ischemic Attacks .................................................. c. Miller Fisher
`Postmortem Increases In Gaba Receptor Binding To Membranes of Cat
`· .Central Nervous System .................... Godfrey Tunnlclitl and G. Kair.ll MEJttleson
`Effects of Nocturnal Gamma·Hydroxybutyrate on Sleep/Waking
`Patterns In Narcolepsv·Cataplexy . • . . . . . . . . . . Roger Broughton and Mortimer Mamatak
`Hemispherectomy lor the Treatment of Epilepsy and Behavior Disturbance .. N.o. Amal!
`University of Toronto Neurosurgical Rounds No. 1 ·
`Masslvo Osteolysis In Association With Multiple Cerebrospinal Fluid
`Fistulae _ Harold J. Hoffman, Derek C. Harwood-Nash, T.P. Morley and M Barry Rewcaslle
`Relationships Belwaen Psychological Measuremnnts and Cerebral Organic
`Changos In Alzheimer's Disease ....... 1-1, Merslmy, M.J. Bllli, w. r. Rlum~;~, Allan .1. Fox.
`Hannah Fox, FL. 1-fersch, V.A. Kr!ll, fl.B. PalmAr
`Features of Croutzfoldt-Jakob Dlseasa In Brains of Pationts With Familial
`Demenlla of Alizhelm,er Typo ........................................... MJ. Ball
`The Effect of Variable Duratlen One Hertz lnterferem:e on Kindling . . . . . . . John Gaito
`Hornor's Syndrome, An Unusual Manifestation of
`Multlpfo Sclerosis . . . . . . . . . .. . Colin R. Bamford, Michael s. Smith and William A. Sibley
`Subdural Empyema With Negallve C.T. Scan:
`A Case Report ............. D.J. Wortl.man, w.s. Tucker,. D.C. Finlayson, R.S. McPht'Jdrrm
`and R Gershtiler
`
`Chronic Extradural Hematoma Presontlno
`With Subgaleal Mass ...... Gi/Del't R.C. Otmrrey, Youssef H. Gabriel and Sl!ml~t.V Tatumg
`Hereditary Sunsory Neuropathy: A Case Vfith Pain and
`Temperature Dissociation _ Betll o·anen. A. Jackson, R. rang-Wai, A.J. Lewis. E.A. Atack
`Nollees and Books ............................................................. .
`
`g ,,..
`
`19
`
`23
`33
`
`39
`
`45
`
`51
`59
`
`65
`
`61
`
`71
`
`73
`17
`
`._
`
`Ranbaxy Ex. 1009
`IPR Petition - USP 8, 772,306
`
`

`
`LE .IOU RN/\L CA :--!ADIEN OES SCIE".:CES NEU ROLOGlQUES
`
`Effects of Nocturnal Gamn1a-Hydroxy butyrate on
`Sleep I Waking Patterns in Narcolepsy-Cataplexy
`
`FR llROil(fHION
`
`NT ROD
`
`FEBR R\
`
`

`
`THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES
`
`Carskadon, 1976) or 3 hour (Weitzman
`et al.. 1974) "davs" in nol'mal3ubiccts.
`In the COUfliC of th·esc experinlents,
`the sustained
`which have
`involved
`fragmentation of sleep, polysomno(cid:173)
`gt·aphic patterns identical to those
`in narcolepsy have
`rapidly
`found
`emerged. Sleep onset REM periods
`and ol her manifcst<Hions of disso(cid:173)
`ciated sleep, such as multiple epochs of
`so-called "intcnncdiatc sleep" (Barros(cid:173)
`Perreira and Lairy, l 9'76), appeared
`within a few hours. A !though the full
`clinical syndrome wa.s never elicited, it
`is concci'o,.able that this might have
`to
`it been possible
`occurred had
`~:ontinue these studies for longer time~.
`Indeed. the cl.inical and polysomno(cid:173)
`graphic patterns Qf narcolepsy can
`develop in pathological conditions
`such as sleep apnea which arc typ1fied
`fragmentation
`sleep
`hy chronic
`(Guille.mirmult et at., 1976). Narcu~
`lepsy also appears to develop preferen(cid:173)
`tially in other individuals in whom
`l'or
`i~ chronically disrupted.
`sleep
`example, in shift workers or in nurses
`and doctors whq must keep irregular
`hou.rs in the course of their duties
`ln 50-75% of
`(Broughton, 1971).
`idiopathic cases of narcolepsy-c:ata(cid:173)
`plexy a history of sever~\ sleep depriva(cid:173)
`tion or of irregular sleep habits
`preceded the onset of'the disc;ise, often
`by many years (Mitchell and Dement.,
`!968: Broughton and Ghanem, 1976).
`Moreov~~r. in estab!.ishcd narcoleptics
`the condition characteristically be(cid:173)
`comes llllt!SIIltlly difficult to control
`when there. is any disrupt.ion of the
`sleepjwaking rhythms hy shift work,
`jet lag, 1)r poor sleep habit~ (Broughton,
`1971; Zarcone, !9'73; Broughton &nd
`Ghanem, 1976).
`Although evidence therefore exists
`that. rreceding nocturr111l sleep distur(cid:173)
`bance may have an important mle in
`the genesis or the condition, and
`indeed some authors have included
`ordinary hypnotics .<.Js part of their
`treatment (Daniels, 1934; Zarcone,
`197:5), the major therapeutic appmach
`the daytime
`to suppress
`has heen
`symp·toms ·"sleep attach and drows(cid:173)
`iness with stimulants: and cataplexy
`aux.iliary
`(and other REM-based
`tricyclic or MAO
`symptoms) with
`inhibitory antidepressants.
`We decided to attempt to increase
`the continuity nnd duration or n(JC·
`
`turnal deep and to study the effect of
`thi~ on the symptom~ of the condition.
`To achieve this we have used nocturnal
`doses of gamnta-hydroxybutyratc
`(GHB), a central short chain fatty acid'
`(Doherty ct al., 1976) with hypnotic
`tlropcrtics (Laborit, 1964), We chose
`G H B because it had been shown til
`promote both REM and slpw-wavc
`in
`(Mamclak et aJ., 1977)
`sleep
`contrast to ordinary hypnotics which
`often ~up press these sleep statc.S'(Kale~
`ct al., 1970). GHB also posse~,sed an
`additional major advant<\gc over the
`usu11i hypnotics in that animal studies
`had failed to demonstratt: the develop(cid:173)
`the drug's
`to
`tolerance
`ment of
`hypnotic effects with pro(Qnged use
`(Vickers, 1969}.
`16
`treated
`To dnte, we have
`narcoleptic patients with G f1 B. In a
`preliminary communic:ation concern(cid:173)
`(Broug.hton u 11d
`ing 4 patients
`M<um:lak, !976) and in a companion
`article detailing the clinical asp~.~rts of
`the patients included in the present
`(Broughton and Mamelak,
`report.
`that GliB
`1979), we have shown
`murked!y improves nocturnal sleep
`and that nightmare~. hallucinations,
`and atta("h of sleep paralysis vanish.
`During the day, pressure for sleep
`becomes less imperative and cataplec(cid:173)
`tic attacks become milder and less
`frequent In many patients virtually all
`~ymptorm; of the disease disappear
`when small repeated daily doses of
`stimulants are u~cd in comhination
`with GHB at night. No tolerance ha~
`devdoped so far for this drug regimen,
`nor have there been any serious sltle
`effects, and patients generally find this
`treatment much more palalahle than
`the usual combination of ~timulants
`and tricyclic antidepressant drugs. In
`this papu, we fncus on the effects of
`GHB upon the n:corded sle.cp,!waking
`patterns of our pati1~nts.
`
`PATIENTS AND METHODS
`Fnurteen of the !6 patients (exclud(cid:173)
`technical
`for
`ing nos. 2 and 10,
`reasons), whose histories arc sum(cid:173)
`the previous
`report
`iu
`marized
`(Broughton anti Mamclak, 1979), have
`had complete studies of their 24 hour
`sh.:cp{waking patterns. They consisted
`of seven males and seven l'emal.cs
`betw1:cn the ages nf 21 and 57 (mcun
`
`41.8 ± 13.6). All showed one or several
`slee{) onset REM sleep periods during
`the recordings. Nine of the fourteen
`patients were seriously dcbilitalcd by
`their illness and four had not bcnclitcd
`much fmm the standard trent ments
`combining stimulants <Uld antidt~pres­
`~ant medication. Before starting G H B.
`all previous treatment for narcolepsy
`two
`least
`was dis1.:ontinued for at
`weeks. The pre-trial assessrnent in(cid:173)
`cl.uded a history and physical exami(cid:173)
`renal, and
`nation. hematological,
`hepatic 8tudies. a chest x-ray, ECG,
`EEG, and M MPlund a brief p~ycho­
`logicul assessment, repeated subjective
`assessment of sleepiness using the
`Stanford Skcpiness Scale (Hoddes et
`al., 1973), pupillometry in the Ottawa
`studies, am! baseline pol)'somnogra(cid:173)
`phic recordings. Afterthe investigative
`and purely volun!I.Hy nature of the
`study was expi.:l.lncd, informed and
`signed con~cm was obtained from each
`patient.
`The polysomnographic recordings
`the Ottawa patients (N=7) were
`in
`mude with portahle 4 channel Medilog
`recorders (Oxford Electrical Instm·
`ment Company). This permitted pa(cid:173)
`their normal
`in
`tient monitoring
`environment. and at their usual activity
`levels. The derivations used were C4 -
`t\;, C 1-A 2, a combined horizoma.l(cid:173)
`vertical oculogram ltndi a submental
`EMG. Twenty-four hours of data
`could be recorded <HI one regular C 120
`casscHe .. In the Toronto stltdies, the
`patients (N''7) were hospitalized and
`the recordings obl.ained with a Grass
`rnodel 78B pt)lygraph. None of the
`patients had histories of cxc.essive or
`i nlense snodng ~ugge:~tive. of :sleep
`apnea, and this symptom was formal!.)"
`the Toronto ~tmlies in
`excluded in
`\Vhich a sufficient number of recording
`channels made it possible to monitor
`nasal and thomeic respiration. Con(cid:173)
`tinuous 4B hour recordings of the
`sleep/waking patterns were obtaim~d
`in all patients in the pre-GliB baseline
`period apd then again after 7 to ll'
`nights on the dmg. During the 48
`hour Toronto laboratory recordings.
`lo
`1.11cre encoumged
`the patients
`rernain in bed exc:,t:pt for meals and
`bathroom breuks.
`An initial 1.5 gm to 2.25 grn (I0-15
`ml) dose c•f G H ll was given orally at
`bedtime and followed by one or two
`
`24- FEBRUARY 19!10
`
`Cill/1 and Slcepfwaking l'tmems h1 /'l'arcol .. pY.V
`
`f
`
`[
`t
`(
`(
`~
`[
`! (
`
`f
`(
`~·
`,t
`.4
`
`

`
`LE JOURNAL CAN AD! EN DES SCIENCES NE U ROLOGIQUES
`
`further 1.0 gm to t5 gm doses during
`the night with any major awakening, if
`more th;m 2.5 hours had passed from
`the previous dose. The patients were
`r;:;quired to fc.cl fully alert and clear
`headed before raking their next dose.
`The duration of 0 IH:I's hypnotic effect
`in man is about 2.:5 hours (Mamclak et
`al., 1977), which corresponds closely
`w that of its detectable presence in the
`blood (Hclrich et al., !964). In most
`patients, two or three dose~ were given
`each night in accord with our objective
`of maintaining as continuous a night'.>
`&!eep as possible. GHB was never given
`within two hours of the anticipated
`time of the morning awakening in
`order to avoid hanu-over effects. The
`total qmmtity given ead1 night ranged
`fmrn 3.75g,m to 6.25 gms, correspond(cid:173)
`ing to an <~vcrage patient dosage of
`nhout 50 mg/ kg.
`The polysomnographic data were
`analysed according to international
`uitcria (RechtschE!llcn and Kales,
`J 968) and scored using 40 sec epochs as
`\Vakefulncss, :;tagcs I. 2, 3, 4 and REM
`sh:cp, plus movement lime (MT, i.e.,
`epochs obscured by movement arti(cid:173)
`lacts for over 50% nf their duration
`with pn:vjous ami succeeding epochs
`''ontaining sleep paUcrns). The night
`and daytime portions nf the r~cordings
`were analysed separately. The former
`was arbitmrily defmed ;-u the time
`between the onset of night sleep to the
`time of the final awakt~ning for
`breakfast. Sleep during the rem;~inder
`of the 24 hours was scored as part of
`the daytime (Figs. !. and 2), T'he time of
`sleep onset was taken as the beginning
`of the first continuous 10 min of REM
`or of NREM sleep. ex.clu~ivc of stage
`I, whidl corres1'ondcd to the !Jaticnts's
`subjecti.vc appraisal of sleep onset for
`the night a» sron:d on the SSS forms,
`Si nee no formal bedtime existed in the
`laboratory .~t.ud ies, nor could one be
`e~tabllstu:d in the portable studies, the(cid:173)
`iatcncy from bedtimt: to sleep onset
`was not determined. For each record~
`ing t>eriod, nocturnnl and diurnnL we
`calculated the total sleep times indlld(cid:173)
`ing and exchtding stage ! (which
`corresponds to drowsiness and, most
`authors agree, not to actual sleep).
`Corresponding nocturnal sleep effi(cid:173)
`ciencies refer to the percentages of that
`portion of the recordings occupied by
`the relevant sleep pauerns. Delr.a .,;teep
`
`Broughton & Namdak
`
`Jattmcy was defined as the time front
`sleep onset to the firs! eonUmwus 3 m·
`more min of stage 3 or 4 sleep. REM
`sleep latency was defined as the time
`from the onset of 3 or more min in
`dmation of stage 2 to the first
`continuous 3 or more min of REM
`sleep. If REM sleep occurred before
`stage 2, its latency was determined by
`measuring the interval between the
`bq;inning of the J consecutive. min of
`REM sleep and
`the preceding 3
`con~ccutlve min of wakcfulnes~. REM
`density refers to the percentage of2:;ec
`mini-epochs containi11g one or more
`rapid eye movements. The values
`obtained for each REM period were
`uormali7ed for its durarion and an
`average value for {~ach oft he nocturnal
`and diurnal recording pcric1ds was
`determined.
`Two further parameters involving
`REM sleep were defined in order to
`measure the degree of REM sleep
`fragment:ttion. These were REM sleep
`efficiencies with and without ~tage 2.
`i.e. other patterns of definite sleep. For
`each REM ~deep· period, the number of
`epochs between the lint and the last 40
`sec REM sleep epoch of' that period
`was determim.~d. This was desi.gnatcd
`the "total REM sleep period duration".
`lleeause of fragmentation, it included
`I!JHH;hs of wakefulness, stage I, MT
`and, at time~. stage· 2. REM sleep
`efficiency without stage 2 rd!ors to the
`percentage of the REM ~leep period
`duration consisting of REM sleep
`epochs only. REM sleep efficiency
`inclttding: stage 2 refers to the percen(cid:173)
`tage ofthc REM sleep period dura.tion
`comisti11g of epochs of REM 8leep or
`of stage 2 sleep, i.e., of definite sleep.
`The two REM sleep efficiency values
`were normaiL,cd for each REM sleep
`period, and an overall average mean
`value for each of the nocturnal and
`diurnal recording periods was ob(cid:173)
`taiued. lu this srudy, a !<EM sleep
`epoch had to be ~eparated from the
`closet precedi.ng REM sleep epoch hy
`at least 15 min to he scored as pan of n
`scparulc REM sleep period. The
`number of REM sleep periods per
`night and their cycle duration, i.e., the
`time fmm the onset of one REM sleep
`period to the onset of t.he next period,
`were also calculated.
`A measm·c for determining the
`degree of overall fragmentation of
`
`night sleep wus also developed. We
`calculated lhe number of periods of
`sleep, be
`these NREM, REM, or
`comhinations of the two, which were
`separated from one another by one
`min or mon: <:lf either MT, wakefulness
`or s.tage 1. Depending upon their
`duration, the,~e nocturnal sleep periods
`were put into five cutegories; !5 min or
`less, 16-JO min,. 31-45 min, 46-60 min.
`and greater than 61 min. ln addition.
`we measured the fn:quellC)l of stage
`shirts out of stages 2, 3 and 4
`collectively (i.e., oUl of NREM sleep)
`and out of REM
`The number of
`shifts out of the former was cxpres:~cd
`per IOO min of the £mn of sl.ages 2, ,;
`and 4 per night, and out of the Iutter
`per !00 min of R F.M sleep per night.
`During the daytime portions of the
`rec.ording~, sleep was analysed for the
`duration of stages!., 2, J, 4, REM, and
`MT; and lhG total sleep times including
`and excluding stage I were eak.,ulated
`as above. The number of daytimt' sleep
`periods was also dctennined. A sleep
`period was defined a.> an episode of
`recorded sleep containing at lea~t 3
`min of ~tagcs 2, J, 4 orR EM sleep, and
`preceded and followed by at least 15
`min of wakefulne~s or stage
`l
`(drowsiness). Thcoe sleep period~ were
`divided into 3 grnups, those of J 1-45
`min, of 46-60 min •tnd of more thnn 6!
`min,. corresponding to
`the
`longer
`measures of conso!id.!LICd sleep at
`night.
`In this paper, the 48 hours baseline
`po!ysornnographi•• data for each pa(cid:173)
`tient is compared to dt~ta after 7 to lO
`nights on GHR treatment. The data of
`each patient for each of the two 24
`hour periods before and after G HR
`trearment were averaged before com(cid:173)
`parison. The two tailed Student t test
`was applied to each variable, unless
`otherwise stal<:d.
`
`RESULTS
`The data nhtai.ned using either the
`portable outpatient or the laboratory
`inpat.iem recorditig techniques were
`~imilar. The major diffcrcm:e was in
`the sleep putt•:rns which appeared just
`before sleep onset at ni.ght. The
`inpatient recording> usually ~howed a
`period of more or less sustained
`wnkfulness until sleep onset, which
`was then followed ~hortly by a REM
`
`FEBRUARY !980- 25
`
`

`
`HIE CANi\DIAN JOURNAl. OF NEUROtOOICAL SCIENCES
`
`ProcGH6
`d'
`
`3
`2
`
`Figure l - A 48-hour co!ltinu\ms baseline recording l.n a typical (hospitalized) patient. It illustrates the frequent awakenings during
`nocturnal sleep, multiple sleep onset REM periods, fragmental ion of REM sleep. and other features of ~leep in narcolepS}1-Cataplcxy
`(note: in Figs. I and Z the vertical a~i~ indicates. sleep stages, and thch.oriwmal ax.is the time in hours. REM sleep is shown as a horiwntal
`whitt bar at the level o! stage I, and movement by small triangles abov~: the sleep Htage line. Time zero hours in both figures was
`lO:.J.0-1 1;()0 ll.m.).
`
`43
`
`48
`
`19
`
`2!
`
`l'l'gr.tre 2 --···A 48-hour mwrding nfthe~amc patient on days 9 and 10 of nocturnal G HR. Times of administration am noted by arrows beloW
`the hnrb:ontal a.xis. The figure illusl.ml.es the increased continuity of nocturnal REM sleep. lhe decrca~e in number of nocturnal
`awakenings, and the nduction (If daytime sleep (dc.~pite the subjects having remained quietly in tht: hospital laboratory while un (l H B).
`
`2h- FEBRUARY 1980
`
`GliB and Slet'p/waking Pallt'I"'M in Narcolep.W
`
`[
`
`~)
`{
`
`! ( !
`I
`I f
`~ j
`
`~
`
`f
`
`

`
`LEJOURNA Ci\NADIEN DFS SCIENCES NEI.:FWLOGlQl!ES
`
`Sl
`s:
`(min)
`
`Stage l (min)
`Siagr..M 2
`S!age ] 1 4
`RFM (mini
`
`TABLF .I
`
`GH.R
`
`16. ± 6.1
`1.3
`llli.l ....
`;
`8\1 ()
`7 . .7
`S4.l
`11.0
`9 .. 1 ± l4
`16.0
`7.7
`
`li.O
`5.9
`I.J
`2.7
`I
`_._ 1
`09:t
`L2 ±I
`
`·······················
`
`..........
`
`TABLE l
`
`llasdhw
`
`CUB
`
`50.2
`69 8 ± 47 3
`IH.9 ± !6.6
`l ± 21
`9.9
`I 1.6
`19.5
`
`S2
`
`Slg.
`
`.005
`
`005
`
`.02.'5
`
`Slg,
`
`.015
`
`.005
`.01
`
`

`
`THE CANADIAN JOURNAL OF NElJROLOGICAL SCIENCES
`
`of this, although the total sleep time
`(including stage I patterns of drowsi(cid:173)
`ness) during the duy remained un·
`changed, <tctual sleep (excluding stage
`!) was decreased and the individual
`daytime sleep pcriQds became &hortcr.
`The overall major cffl:ct of the drug,
`then, was to impr<1VC the c<mtitwily of
`not:t.urnal :deep and to reduce long
`periods of daytime sleep and diurnal
`slow wave and REM sleep. Subjective(cid:173)
`ly. the daytime sleep was pcrccivc:d as
`being less imperative.
`finally, although there is evidence
`that GH B can produce EEG and
`behavioral manife-stations similar to
`petit mal epilepsy in rats (GodM:halk et
`aL, 1977) and in cats (Snead ct al.,
`1976), no potentially epileptogenic
`EEG dischurgcs were present in these
`very J>to!ongcd recordings or in later
`follow-up recordings, and no dink-a]
`~ci:r.un:~ have occurred ..
`
`DISCUSSION
`The clinical and polysomnographk
`changes produced by GHB during the
`7- J0 d!l)' J)Criod followed ll parallel
`COUI'SC. Cliniclllly, as previously re-
`ported (Broughton and Mamelak,
`1979), there was reduction both in the
`duration of duytime >leep and in the
`incidencc ar1tl intensity of cataplectic
`attat:ks; and, corresponding to tl1is, the
`daytime portions of polygraphic re-
`eordlngs showed less actual total sleep
`lime, and less time in ~low wave sleep
`and in REM sk'!!.'P· Subjective drows)..
`to be a
`ness, howev•:•·, continued
`pmblem. It was ref1ected in the lad; of
`any significant change in daytime ~tage
`I sleep. which in fact was somewhat
`increased. (Drowsin,c.s~ was subs.e-
`quently improved with methyl-pheni-
`date.) Night sleep was perceived as
`being de~eper and less restles&. There
`was loss of nightmares and hai.lucina-
`tions, although dreaming, in a more
`pleasant manner, continued. Corrcs-
`pondl ngly, the nighttime portion of
`the re<:onlings showed lhat sleep was
`consolidatt:d into longer periods, there
`were fewer stage shifts and sleep,
`particularly REM sleep, was more
`integrated <md less fragmented. Al-
`though sleep <m~ct REM periods still
`occurred, 1md in fact were even more
`frequent on GHR, these differed fmm
`
`sleep period (Fig. IJ. Patients recorded
`at home tended to d_rift from wakeful·
`ness in and out of brief 1-3 rrJin periods
`of REM ~d<>ep or stage I for scvcml
`minutes or even dozens of minutes,
`before f11lling into a consolidated sleep
`period of at least 10 min; and tll<.'Jl
`usually thcfl had much longer or even
`normal REM sleep latencies~ The
`REM sleep latencies recorded in the
`outpatient sltldies wert: thus signifi(cid:173)
`cantly longer than in the inpatient
`studie~ (Chi squared test, p < 0.005)~
`Other REM sleep measures did ll()t
`differ signific<mtly between the two
`laboratories.
`rhc nocturnal pre-G HB baseline
`recordings (Table I) showed a number
`of features when compared to pub(cid:173)
`lished data (Williams L:l at., 1974}, and
`confirmed the findings of other:s for
`thi~ condition ( Rechtschaffen ei ;ll.,
`1962; Harms-Ferreira and Lairy, 1976;
`1978). Thest:
`Montplaisir et a!.,
`included early or direct sleep 011set
`REM periods, frequent awnkenings
`and period~ of relatively prolonged
`low sleep efficiencies.
`wakefulncs.s,
`and frequent stage shifts. In short,
`sleep was characterized by
`night
`marked fragmentation. which was also
`in our mc<tsures showing
`reflected
`frequent short (i ,e., 15 min or less)
`periods of sleep and low REM sleep
`efficiencies (with and without stnge 2).
`The daytime sleep measures before
`Gi-l B arc given in Table 2. Fig. l shows
`a 48-hour pre~GHB recording in a
`typical patient.
`GHB (Table t, Fig. 2) significantly
`increased the dm-alion ()f nocturnal
`slow wave sleep at !:he cxp!:nse ofstage
`sleep eff1ciency
`the
`I,
`increase<i
`measures. nnd decreased the number
`of sleep periods less than 15 min in
`duration. The total amount of REI-1
`~leep was unchanged, hut it bcc<lmc
`indicated by
`fragmented, as
`less
`significantly fewer stage shift~ out of
`REM sleep and by an increase in the
`REM sleep efficiency. GHB signifi(cid:173)
`cantly decreased both the latency to
`REM ~leep and the density ofthe rapid
`eye movements themselves. The day(cid:173)
`time data (Table 2) indicated that
`nocturnal GHB resulted in a signifi(cid:173)
`cunt decrease in the durution of both
`diurnal slow wave sleep and REM
`! patterns, however,
`sleep. St:agc
`increased (non-significantly). Hec:au~e
`
`2!1- FEliRU/\.RY 1980
`
`..
`...
`
`[
`t
`( I
`r l
`
`'~
`
`f·
`
`1;
`(·~
`
`~
`·
`
`''l
`
`··_
`
`:1~·
`
`that they were not frightening, they
`never reached halludnatory intensity~
`control over mentation was lo~t rather
`thun mnintaincd, and the presence of
`concomit<Hif <1warencs.s of ones' sm·.
`roundings, whid1 <:an occur in thi;;
`C\'mdition ( Hishikawa, 1976; Vogel
`!976), was no longer pre>ent.
`Like other investigator~ surh as Bar(cid:173)
`ros~Fcniera and Lairy ( !976) and
`Montplaisir and colleagues ( 1978), we
`were imprcsol:d by ihe marked di~soci­
`ation and fragmentation of nocturnal
`(and diurnal) sleep whit:h we found in
`our patients' baseline recordings. In
`addition to frequent sleep onset REM
`sleep periods, thl:re were numerous
`qwchs of "intermediate sleep" (i.e"
`simultaneous features (Jf stugc: 2 and
`REM sleep), multiple brief sleep frag ..
`ment~ a.nd prok•nged periods with
`mixed features cof sb.:p and wakefui(cid:173)
`ness. Sleep and it• subcomponents ap(cid:173)
`t.o have become dispersed
`peared
`aro11 nd I. he 24 hours and the barriers
`between sleep and wakefulncs.s to have
`been breached, as exemplified hoth by
`the chronic davtim1: drowsiness and
`the wakeful awareness during po.ly-
`graphically monitored REM sleep,
`especially at sleep onset.
`GHR tended to reverse these rea-
`tun:~. lt I>YDduced increased mnsolid-
`ation and re-integration of ~!eep and
`increasingly synchroniJ~cd sleep with
`the nocturnal period. Each dnse
`<u;sured a 2 to 3 hour period of steep at
`uhoul the same time cad1~ night. In
`each of these periods, REM ~leep
`usually oc.curn:d at sleep onset and was
`followed by a period of ~;low wave
`rc·
`the
`(Fig. 2). Although
`slt:cp
`nnrmnlit.at.ion of night sleep clearly
`therefore not complete, each
`was
`period of drug·i:nduct~d sleep consisted
`of sleep which w<ts more continuous,
`having fr'lrT.r awakenings and fewt:~·
`stage shifts. The subjective assessment
`of patients on medication was tlH't
`they were truly asleep during each two
`three hour drug-induced sleep
`to
`period agd did not experience "twi·
`light" states of mixed sleep and
`wakefulness. Although the total dun.\·
`tion per se of nocturnal sleep was not
`incrca~cd by GHB., the drug's nocHH..
`nal d'fC;cts did alter the duratiNI and
`organization Df daytime sleep. There
`was significant decrease in the dura-
`tion of both REM sleep and ~low wave...· ... ··~ •I.
`
`their pre-treatment cnunterparts
`
`in
`
`G Jill ami Sleep f wakil'lg Pal/ems in Nat<"o/eJMY ...I
`
`

`
`LEJOURNALCANADIEN DES SCIENCES NEUROLOGIQUES
`
`sleep during the d;iy and the individual
`~Jeep periods became shorter and more
`fragmented. This effect might have
`been more impressive statLstieally, had
`not half our patlents (the Toronto
`inpatients) remained in bed during the
`day.
`\Vhi!c on the drug our patients
`reported that, although they were still
`drowsy and even slept during the day,
`thev wen: now better able to resist
`r!c~p and could ~tay &w<dte, when this
`was necessary. Before starting treat(cid:173)
`ment they averaged about 9 to !0
`hours of sleep in a 24 hour period (of
`which 6 to 7 hours occurred at night),
`Thes!: total figures, which were not
`t:ha nged much by G HI~ treatment, are
`not very different from those recorded
`in ad lib sleep of normals, who will also
`sleep for about 10 to ! 2 hours in a 24
`hour period. when freely permitted to
`do so (Hishikawa et aL 1976). Yet.
`under most circumstances, normals
`remain fully awake during the day with
`S(tVen to eight hours of sleep at night or
`even
`less
`(\'./ebb and Cartwright,
`!978). What make:;
`this pattern
`possible for them but not for narcolep(cid:173)
`tics'! We suggest that it is because the
`night
`sleep of normals
`is more
`\ll
`int(:grutcd
`than
`is
`tlw
`sleep
`narcoleptics. That is. in normal ~leep
`the component suhsystems run their
`cour~e for the most part in seven to
`eight conl!ecutive hours usually syn(cid:173)
`chronin:d with the nocturnal period.
`In narcoleptic sleep, on the othCI" hand,
`the di:~sociati<'m and temporal disper(cid:173)
`sion of the ~leep sub~components
`prevents this and leads to daytime
`occurrence of sleep or of chronic
`drowsiness -- a mixture of sleep and
`wakefulness. At the same time, the
`is
`nighttime sleep of nan:olcptics
`rendered shallow and fragmented and
`loses its stable circadian pattern of
`deep NRilM s.!eep concentrated in the
`first third of the nighL
`It is our thesis that the 6 to 7 hours of
`sleep facilitated by G I-IH has greater
`circadian ~tabilily und is a more fullv
`i.ntegrated sleep-, especially of the
`REM sleep state, than is that which
`occurs in narcoleptics in the absence of
`the drug. As evidence, we can cite the
`drug-induced decrease in the number
`of nocturnal sleep stage shifts. as well
`<is the overall improvement in sleep
`efficiency nt night. Nocturnal GHB
`
`appears 1 o "glue" together the c:om(cid:173)
`poncnt subsystems of sleep and to
`impede
`their
`temporal dispcr;;ion
`around the 24-h(HII'S. As a result,
`daytime sleep become~ less consolid(cid:173)
`ated, with stage I sleep, i.e, drowsi(cid:173)
`ness, increalling at the expense ofslow
`wave sleep ;md REM sleep.
`!'his
`ac.counts for the patients' subjective
`imprc:~sion tlml the:~ are better able 1.0
`re~l:~t ~lceping during the day on GHB.
`When small divided doses (Stu 10 rng
`Li.d.) of methylphenidate weJe later
`added to the drug regimen during t.he
`day, diurnal sleep and drowsiness
`virtually disappeared in many patients
`(Broughton and l'vfamelak, 1979). lt
`can be <Juestioned whether me1hyl~
`phenidate would be neces-'<ary ~tl all, if
`the duration of action of GHB c;ould
`he extended l<:l' integrate sleep at night
`for a full seven to eight hours.
`The decrease in the frequency and
`intensity of cataplectic attacks is one of
`the earliest and most
`impressive
`clinical benefits of GHB treatment As
`the duration of' the direct action of
`GHB (Mamelak et uL, 1977) and its
`detectable presence
`in
`the blood
`(Belrich et aL, 1964) lasl nnly some
`2. 5-.3.0 hours, the dayt !me. changes
`must be '~"plaincd mainly or only by
`noctumal cfl'ecb. when the substance
`h given. Again, it is :mggcstcd that. t.hi8
`re~ults from
`the nocturnal skcp
`integrating and synchronizing actions
`of the drug. Cataplexy has been
`attributed to the dissociated selective
`aclivati nn nf thL: motor inhibitory
`eomponent of REM sleep (Dement ct
`a!., 1976). Our data indicate that
`nocturnal GHH significantly decreases
`the total amount of REM sleep during
`the day: the dec!im; in the number of
`diurnal cataplectic attacks may be due
`to thiN, Other studies, moreover, have
`shown that daytime administration uf
`GHB in nan:olepsy-cataplexy has the
`apparently unique effect of heing ahle
`to induce sleep paralysis (!VIame!ak e1
`al., 1977). This suggests
`that,
`in
`addition to its facilitating or activating
`effe<:t on REM slee1' per ~c. the drug
`c:an also selectively activate th<.: motor
`inhibitory component of th1: REM
`sleep state
`in
`such patients. The
`sensitivity of this motor process to
`GHR was further demonstrated re(cid:173)
`cently by 1\hmnlek, Sowden and
`Curuso (in press) in studic~ on the
`
`Broughton & l'•'mrwlak
`
`L_
`
`effect of this drt1g on monosynaptic
`tran:~misNion in the spinal conj, using
`the H·rclkx technique, Monosynaptic
`transmis:;ion in known to be suppress(cid:173)
`ed during REM sleep (Hnd<;~ and
`Dement, 1964: Hishikawa. and Kaneko,
`1965) as part of the motor inhibitory
`process during this sleep state (1-'mn(cid:173)
`peiano, 1976)_ nut
`the studies of
`Mamelak, Sowden, and Caruso (in
`press) show that with G H B morwsyn(cid:173)
`a.ptk transmission is blocked dming
`borh REM and slow wave otatcs
`l'oll.<lwing drug <tdministrut inn. Since a
`refractory period occurs afler REM
`sleep (Jouvet. 1962; Pompeiano, !976),
`an analogous state may also prevail
`after the iS(l[atcd activation of thc
`motor !nhibiiory pwccss. The ameli(cid:173)
`oration of daytime cataplexy may be
`related in this W<IY to prolonged noc(cid:173)
`turnal activation by the drug of !he
`motor inhibitory mechanisms of REM
`sleep.,
`Why do ordinary hypnotic drugs
`11ot benefit narcoleptic patients'? It
`should be noted first that at times they
`can do !>O, Many n<trcok:ptic patients
`use such drugs at night to tmf>mve
`!.heir sleep and obtain eon,identhl~
`relief from
`their diurnal ~ymptom>
`(IJunicls, 1934; Zarcone, 1973). Bt.~­
`causc of their lung duration or action,
`however, tlu:se drugs may increase
`daytime drow~oiness (Daniels, 19.34);
`and, in addiuon, lheir consolidating
`cff(a;ts on nighttime sleep tend to wane
`tolerance develops_ Moreover,
`!IS
`ordinary hypnotic drugs often sup(cid:173)
`press both REM slcer' and slow wave
`sleetl, and can create incrcuscd pres(cid:173)
`sure, a1 least for REM sleep, later in
`the night as the drugs wear off (Kales ct
`a!., 1970). GH.B hilS none of these
`disaclvaniagcs. lt io rnpi.dly rnetubo(cid:173)
`Iiz.ed and iN deared rmm the blqod
`two
`three hours
`stream after
`lo
`(Helrich el at; 1964), tolerance fails to
`develop lo its hypnotic effect (Vickers,
`1969), and most import:::~nt, it does nol
`appear to suppress either REM sleep
`or slow wav(: sleep or sub-components
`of them_ ln fact, in direct contrast to
`the synthetic hypnotics, it generally
`increases the duration of slow wave
`sleep and
`facilitates REM
`sleep
`(Mamc!ak. ct aL, 1977).
`It must be emphasized, however,.
`that
`the
`increase
`in ddta activitv
`pnxh1ccd by GHB may not represent it
`
`FEBRUARY 1980-29
`
`

`
`THE CANADIAN JOURNAL OF NEUROLOGICAL SCIENCES
`
`true increase in physiological slow
`wave sl<~l:p. G Hl3 can paradoxically
`induce della activity witb the subject
`either awake or asleep (Metcalfe cl aL,
`966; Yamad<t ct al., !967). Thcover~ll
`increase in delta activity recorded In
`our subjects may therefore represent a
`drug eflht. rather than an increase of
`physiological ~low wave >!ccp. REM
`sleep facilitation
`is a more certain
`property of the drug .. Not only were the
`psychological attributes of G H B(cid:173)
`induced REM sleep simihu to those of
`natundly oceurriog REM sle.ep, but its
`poly~omnographic and motor charac(cid:173)
`teristics are similar as weU (Mamelak
`et aL, !977; Mamelak, Sowden, and
`Caruso (in press). For these reasons, it
`is int