Anxiolytics, Adrenergic Agents, and Naltrexone
`
`MARK A. RIDOLE. M .D .. CAlL A. BER:-JSTEIN. M.D., EOWIN H. COOK. M.D .. HEN RIETIA !.. LEONARD. M.O.,
`JO H NS. MARCH. M. O .. A:"< ! I JMv1ES M . SWANSON. I'll. D.
`
`ABSTRACT
`Objective: To review extant data on lhe efficacy and safety of anxiolylic medications (benzodiazepines. buspirone. and
`other serolonin l A agonists). adrenergic agenls (~·blockers and u 2-adrenerg1c agonisls clonidine and guanlacine). and
`the opiate antagonist naltrexone !hat have been used to treat various psychopathologies in ch1ldren and adolescenls. To
`identify critical gaps in our current knowledge about these agents and needs for furlher research. Method: All available
`controlled !rials of these medications in children and adolescents published in English lhrough 1997 were reviewed. In
`add1tion. selected unconlrolled sludies are included. Results: The major finding. lhallhere are virtually no conlrolled
`data !hat support the efficacy of most of these drugs for the treatment of psychiatric disorders in children and
`adolescents, IS both surprising and unfortunate. For some drugs. e.g., buspirone and guanlacine. !his is because no con-
`trolled studies have been carried oul in children and/or adolescenls. For olher drugs. e.g., clonidine and naltrexone. most
`of lhe placebo-controlled studies have failed to demonstrate eHicacy. Conclusions: The strongesl recommendations for
`controlled studies of safety and eff1cacy in children and adolescents can be given for the following drugs: benzodiaze-
`p1nes for acule anxiety; buspirone (and newer serotonin 1 A agonists as they become available) lor anxiety and depression;
`!.!-blockers for aggressive dyscontrol: guanfacine lor anention-delicit/hyperactiv1ty disorder; and nallrexone lor hyperaclivify.
`inanention. and aggression in autislic disorder. J. Am. Acad. Child Ado/esc. Psychiatry. 1999. 36(5):546-556. Key Words:
`psychopharmacology, pediatric, drugs.
`
`T his review examines safety and efficacy data for several
`groups of mcdications rhat arc used to treat psychiatric
`disurd..:rs in children and adolescenrs. Classes of med-
`i<.:ations rcvicwc.:d arc the anxiolyti<.:s (bcnzodia:t.epim:s,
`buspirone, and other serotonin [5-HT] l A agonists) ,
`adrenergic agcnrs (the ~-blockers and thc a~-adrenergic
`agonists clonid ine and guanfaci ne), and the opiate
`antagonist naltrexone. O ther classes of d rugs that are
`used as anxiolytics, e.g .. the tricyclic antidcpressanrs and
`
`t ln rrrrd 1Jrm11ha .l. /'J'J/1.
`lJr. Hu/dlr u Am~dau l'rojf.mJr nj l~r' ht,ury '""' Pa!idlrio .md Oirraor.
`/ht•uwu af' f.'l•dd rnul Ado/n,·rlt/ /'oyd,llltf.'l· }r.!•u• H ttpl:lllf t\lo/,,.,,j
`lnstuutium. &tlumurr; I Jr. Ham ton IJ AJSf1•·;,,u Pm/tsior oj l~"f,.J,,,ury au,/
`[)j,.,·tur. lJir•isum n( Child ,,,/ lldolrs,·rm l'~tdJMtry. Untr•rni~y t{ Alimusnt,
`~lrdu·.,/ ."u.·hool. Aflllllt'llf'IIIU; I Jr. Cnok is l(,stJdrlft J'rr,/f'JitJT nf/'~r··hirllT)' rmd
`/hlllltflc'J. l hm·truty rJj ( .1titllfJ1 .~,/wul u/ ~ltrlidnr: /)r I ttm.trtf iJ l'ruji·S>ur o[
`/ Jtrr,·uu t{ l'nluuug. /)J,.j,wu nf (:J,j/d rtml
`l~'td'~tl/~'1 ll ml/'rtltrtfrlt J .uui
`Ado/rs, rm l'n·, ht11try. Hrml'lt llt~ll'trury -~~ J,,H,/ flf 1\frrli, mr. /lym1ulrm·r. HI;
`I Jr. Altm·b u t hw, t.l/t l'rofnJur of H.r« "''"'! tJmll'rthuru s .md I Jiu , 1ur of t ilt'
`/"ro,(ntlll m (},j/t{ 1/111/ Arln/r,aut AltXlrt)' /Jitflrdrn, /Jul.:r l 'mt't'TJif) Sdmo( nf
`r\lnlt, lilt'. I )ur/.,,m, Nf.": ,mt/ I Jr Sw.m sull t J l'roFiuJT uf' I'J)'d,fu,(.r in
`/'s.rthuiiY_'f. ( lm t·rnlly of l Ad(ftirmrlo/1 bt'IIJI'
`H,.f""' rrtlun tf to I Jr. Hit!tltt', C.1nMrrni l.'rmn Suitr 3·1f•. }uhw Ht,pkim
`Nu,ru,,/, (I(JIJ NtJrth \r'l,lj,. .~·trtn, H.tdrmmTt', ,\1/J 1/!l'I.-·.U!S.
`()WJO · H )(,7/')'l/.tHO~ OC\-16, ~ 1 9')') hy tht· A • ut·r i~.Jn At.:,h.lt·my 1,t' <:hild
`.ami Adull'\c..t·nt J •~yd 1u 1ry.
`
`the selective: serotonin n:uptake inhibitors, arc: reviewed
`elsewhere: in this Spccial Section (sec Emslie eta!., I 999;
`Geller er al., 1999).
`All w ntrolled trials of the sdcctcd medications in
`children and adolescents published in English through
`1997 are included in this review. In addition, selected
`uncontrollcd studies arc included.
`Each section follows a consistent format: background,
`efficacy, safery, and conclusions with recommendations
`for further research.
`
`BENZODIAZEPINES
`
`BACKGROUND
`Benzodia:t.epines have: muscle relaxant. anticonvul-
`sant, hypnotic, and antianxiety effects (Dant:t.er, 1')8)).
`Ben:t.adia:t.epines have been swdied widely in adults, bur
`only a fcw controlled studies in children and adolescents
`havc been reponed :md conclusions are limited by small
`sample si:t.es. short duration of medication trials. low dos-
`ages, and high placebo response rates. Benzodiazepines
`arc in general absorbed and meraboli1.cd more rapidly in
`children than in adults (Simeon. l ')')3). but no specific
`
`546
`
`I
`
`,\ ~1 AC,\11
`
`t ' H1 1ll A!l tlll·\t'. 1'\ \'( ' 111 A I ~\ . IK :'\ . ~l A \ l'J' I'J
`
`APOTEX EXHIBIT 1035
`Apotex v. Alkermes
`IPR2025-00514
`
`

`

`pharmacoki netic data in children and adolescents arc
`available: for a n y of the benzod ia7.epines except
`diazepam (Ciein and Riddle, 1995).
`
`EFFICACY
`
`Anxiety Disorders
`Opm-Lnbcl Smdil's. Of 18 children and adolescents
`with separation anxiety disorder u eared with alprazolam
`(0.5-6 mg/day), 89% were raced improved by psyd1ia-
`uiscs, 82% hy parents, 65% by self-reports. and 64% by
`reachers (R. Klein, personal communication. 1991, cited
`by Kurcher et al., 1992}. In another study, 4 adolescents
`with panic disorder improved on clonazcpam 0.5 mg
`twice a day (Kurcher and MacKenzie, 1988). Somatic
`symproms of anxiety improved more quickly rhan psy-
`chological symproms of anxiety.
`Plncebo-Comrollrd Studies. In an 8-week double-blind
`study comparing alprazolam (mean daily dosage of 1.4
`mg/day). imipramint: (mean dosage of 135 mg/day).
`and placebo in child ren and adolescents wirh anxiety
`and/or depressive disorders. there was a uend in favor of
`the active medication groups (Bernstein et al., 1989}.
`However, it was unclear whether the resulrs were affected
`by baseline difT~:r~:nces in symprom severity between the
`groups. In a double-blind, placebo-controlled srudy of
`alprazolam (mean dosage 1.6 mg/day, range 0. 5-3 .5
`mg/day} for 4 weeks in 30 children and ado lescen ts
`with overanxious disorder or avoidant disorder, 88% of
`the complerers on alprawlam improved versus 62% in
`rhe placebo group. hur chis difference was not statisti-
`cally significant (Simeon er al., 1992}. A double-blind
`crossover study evaluated 4 weeks of clonazepam (0. ) -
`2.0 mg/day} versus 4 weeks of placebo in 15 children
`with anxit:ty disorders, mainly separation anxiety dis-
`o rder (Graae et al.. 1994}, w ithour finding a significant
`difference between u catme nt arms. A double-blind,
`placebo-conrrolled study of clona1.epam fo r adolescents
`with panic disorder demonstrated benefi c wirh ;JCtivt>
`medication (Kutcher and Reiter. personal comm unica-
`tion, 1996). Those created with donazepam showt:d
`improvement on measures of generalized anxiety. fre-
`quency o f panic attacks, and school and social d isability.
`
`Anxiety Associated With Medical Procedures
`In 13 pediatric oncolof,ry patients, an open-label study
`of low-dose alprazola rn (0. 125-1.0 mg} showed the
`drug to be effective in decreasing anticipatory and acute
`
`ANX I OLYTICS. A I >RE~ERCIC AI;E:-ITS. :\Nl> NALTREXO:-.IE
`
`s itua tiona l anxiety associated with bone marrow
`aspirations and spinal taps (PfefTerbaum er al.. 1987b).
`A double-blind, placebo-controlled study evaluated 0.2
`mg/ kg or o ral rnidawlam, a high-potency, shorr-acring
`bcnzod iazepine. in preschool child ren undergoing lacer-
`ation repai r (H ennes et al.. 1990}. Midazolarn is cur-
`rently available only as a parenteral injection solu tion.
`Seven ty percent in the midazolam group (2 I /30) im-
`proved versus 12% (3/25 ) in the control g roup (p <
`.0001 }. Then: were no respiratory or othtT adverse events.
`
`SAFETY
`
`As in adults, drowsiness and sedation are the most
`commo n side effects observed in childrt:n. These side
`effects are dose-related and generally resolve as tolerance
`develops (D u Po nr and Saylor, 1992}. O ther potential
`side effects include incoordination. diplo pi:1. tremor.
`and decreased me mal acuity (Biederman, I 99 I; Kutcher
`et al., 1992). Behavioral disinhibition in children is man-
`ifested by irritability, tantrums. and aggression (Graae
`et al., 1994), and in adolescents as irritability and behav-
`ioral outbursts (Reiter and Kutcher, 199 1 ). In a report
`of 4 children with behavioral d isinhibition on clonaze-
`pam, 3 of rhc children had underlying structural brain
`da magt: (Commander et al.. 199 1 ). These authors sug-
`gested that brain injury may be a risk facror for dewlop-
`ing chis adverse effect. Psychotic reactions or exacerbation
`of psych otic symptoms have a lso been reported.
`Pfeffcrbaum and colleagues (I 987a) described 2 cases
`of exposu re ro low-dose henzodiazepines which were
`associated wi th psychotic symptoms, w hich resolved
`upon discontinuation of benzodiazepines.
`Tolerance of and dependence o n benzodiazcpin<:s
`occur in ad ults (Salzman, 1989). No data have bet:n
`published regarding the risk of physiological and psy-
`chological d ependence in children and adolescents.
`H owever, it is recommended chat benzodiazcpint:s be
`prescribed for you th o n a short-term basis (i.e., weeks
`rather than mo mhs) because of rhe theoretical pmenrial
`fo r dependence. D iscontinuation of rhe bcnzodiazepines
`can be associated wirh rt:currence of anxiety. rt:bound
`anxiety. and w ithdrawal symptoms such as an xiety.
`malaise, irritabiliry, headache, swearing. gastroi nu:stinal
`sympto m s, insomn ia, an d m uscle rensio n (Coffey,
`1993; Salzman , 1990). Gradual tapering of rhe drug
`reduces rhe risk of developing these symptoms (Coflcy,
`1993; D uPo nt and Saylor, 1992; Kutcher et al.. 1992}.
`Abrupt disconrinuation of benzodiazcpines can resul t
`
`547
`
`

`

`RJ!J!JI.I: ET AI..
`
`in seizures. c.:spe~:ially in parierm with a hi~wry of sei-
`zures. r or clonazepam, :1 discontinuation r:tte of less
`th:tn 0.04 mg/kg per week was found to be safe in a pro-
`spective study (Sug:ti. 1993). Benzodiazepincs :ue rel-
`atively safe in overdose (Kutcht:r er al.. 1992). yer these
`drugs have additive: effects with ocher sedative :tnd hyp-
`notic drugs. including alcohol (Green, 1995). The rare
`of absorption o f tht: benzodiazepines and rhe mag-
`nitude of thc:ir CNS depression dfecrs arc also incre:tscd
`by alcohol (Rail. 1990).
`Unprescribed use of bcnzodiazepines occurs in ado-
`lescence. In a longitudinal study of I ,230 teenagers in
`Swcdt:n, 10% had rakc:n anxiolyric and/or hypnotic
`medications in the previous ye:tr (Pedersen and Lavik,
`1991 ). The majority gave sleep diswrbance, depression,
`or m inor life srressors as explanations for raking the
`drugs. '1\vo thirds of rhe teenagers received the benzo-
`diazcpincs from their parents, primarily their mothers.
`On the other hand, 13% of the males and 20% of tht:
`females reponed inmxicarion as the purpose for raking
`these drugs. In this group. rhe benwdi:~zepines were
`obtained from peers and illegal sources. There was a
`strong association herwcen usc hy parents and unprc·
`scribed usc: by the adolescents, suggesting that the: teen-
`agers were modeling their parents' usc.
`
`RECOMMENDATIONS FOR FURTHER RESEARCH
`Future work should focus on controlled studies wirh
`adc:quare sample size, dosage:. and duration of trearmenr
`to addrc:ss efficacy of the bcnzodiazepines for anxiety
`disorders in childn:n :1nd adolescents. For those bcnzo-
`diazc:pim:s that demonstrate clinical dlical-y, pharmaco-
`k.inetic studies need to he conducted. In addition. studies
`that ~:valuate medication in combination with psycho-
`social trC'Jtmcnr arc desirJhle, as they more closely mimic
`treatmem in rhc: real world. It is also important to srudy
`tolerance and dependence so that clinicians will he guided
`regarding which youth arc c.·mdidates for benzodiazcpines
`and how long treatment should l:tsr. The long-rerm safety
`of this cl:tss of medication m:eds to bt: addresst:d.
`
`BUSPIRONE AND OTHER 5-HT14 AGONISTS
`
`BACKGROUND
`
`The S-HT 1A receptor agonises enhance the tOnic acti-
`vation of postsynaptic 5-HT receptors by acting to
`desensitize the 5-HT 111 receptor located on the somato·
`
`548
`
`dendritic portion uf the presynaptic neuron (Biicr c:r al.,
`1990). This receptor is part of a negative feedback loop
`that limits release of 5-HT from rhe presynaptic neuron
`as synaptic 5-HT co ncentrations rise. In studies of
`adults. buspirone and other azaperone partial agonises at
`the 5-HT11, receptor have been shown ro have both
`anxiolytic and antidepressant properties. Controlled
`trials have shown rhar buspirone is effective for major
`depression (Rickels et al., 199 1; Robinson et al.,l989)
`and generalized anxiety disorder (Ansseau er al., 1990;
`Enkclmann. 1991 ). Unlike gepironc (Pecknold et al.,
`1993) , buspirone docs not appear to be effective for
`panic disorder (Sheehan et al., 1993) or for obsessive-
`compulsive disordt:r as a primary agent (Pato er al.,
`199 1) or as an augmenror (McDougle er al. , 1993). Bus-
`pirone is the only 5-HT111 agonist currently marketed in
`the United Srarcs (for generalized anxiety disorder in
`adults). Despite lack of controlled studies, buspirone is
`used in children and adolescents for indications as
`diverse as oppositional behavior, anxiety, and depres-
`sion, in part because ir is remarkably free of side effects
`(Kutcher er al., 1995). Other compounds active ar pre-
`and postsynaptic 5-HT 1 receptors also are under devel-
`opment (Dubovsky, 1993; Mosconi er al., 1993). ror
`example, flesinoxan (Rodgers cr al., 1994), gcpironc
`(McGrath ct al., 19?4). ipsapirone (Curler et al.. 1994) ,
`and randospirone (Evans ct al., 1994) have shown prom-
`ise in adults.
`
`EFFICACY
`
`No pharmacokincric, dose-finding, or controlled safery
`and efficacy studies of buspirone or any orher 5-H T 1A
`agonist in mt:ntally ill children or adolescents have been re-
`poned (Hughes and Prcskorn, 1994; Kutcher er al., 1995).
`On the basis of open clara, clinical experience, and age·
`downward extension of studies in adults, buspirone has
`heo:n used for children with generalized anxiety (Coffey.
`1')90; Kutcher er al., 1992. 1995; Maleric et al., 1994;
`Popper, 1993). Moreover, it has been used in the following
`contexts: anxiety mixed with mild depression; affect·
`driven aggression in :tsSociation wirh oppositional symp-
`tOms; pervasive developmental disorders. where affect
`dysregularion, aggression, and cognitive rigidity are prob-
`lematic; and occasionally, attenrion-deficir/hypcractivity
`disorder (ADHD) refractory to more conventional treat-
`ments. However, until controlled studies arc available,
`rhe use of buspirone for these indications must be con-
`sidered preliminary.
`
`

`

`ANXIOLYTICS. ADRE NERGIC AGEI'TS. AN D NAI.TR EXONE
`
`In an open trial , Simeon (1993) rreared 15 patients
`(aged 6-14 years) with anxiecy disorders with buspirone
`for 4 weeks (1 8.6 rng mean maximum daily dose) and
`reported sign ificanr improvement in anxiecy, behavior,
`and hyperacrivicy. Adverse events were infrequent and
`mild. Case reports of children and adolescents also sug-
`gest benefit in overanxious disorder (Kranzler, 1988),
`depression and obsessive-compulsive disorder (Alessi and
`Bos, 1991), and social phobia (Zwier and Rao, 1994).
`An interesting literature also has grown up around rhe
`use of buspirone in aggressive children (Gross, 1995;
`Mandoki, 1994; Sranislav er al., 1994), where speculation
`has it thar benefit may accrue from dopamine antagonist
`properties seen at high doses as well as from modulation
`of seroronergic activicy, and in autistic children (Realmuro
`et al., 1989), where am:ntion , impulse control, and
`hyperacrivicy have reportedly decreased in some patients.
`A recently published open-label study in 25 prepubertal
`children wirh anxiety and aggression rested doses of up to
`50 mg/day for up ro 9 weeks: 6 children showed increased
`aggression or mania, and of the 19 who completed rhe
`study only 3 had sufficient benefit ro continue buspirone
`after the study (Pfeffer et al., 1997). Buspirone is usually
`started at 5 mg 3 times per day and gradually increased ro
`30, 60, and 90 mg/day in 3 divided doses every 2 weeks.
`The need for rhrice-daily dosing limits feasibility and
`compliance. Time will tell whether compounds such as
`gepirone, with higher potency ar the 5-.HT,A rccept~r
`than buspirone; transdermal (parch) dcltvery of buspt-
`ronc, which allows much higher serum lcvds without
`excessive side effects and which one investigative group
`(Conners and March, personal communication, 1998) is
`srudying for the treatment of ADHD; or longer-lived 5-
`HT1A agonisrs may show greater benefit than the tablet
`form of buspirone in rhis regard.
`
`SAFETY
`Side effects across trials of parienrs wirh different dis-
`orders using differem 5-HT1A agonises have been uni-
`formly mild: light-headedness, stomach upset, dizziness,
`sedation, asthenia, or headaches. Furthermore, rhc 5-
`HT1A agonisrs cause no withd rawal symptoms even
`after prolonged administration (Rake!, 1990) and have
`no addictive potential (Murphy et al. , 1989).
`
`aisorders. In addirion, rhe newer 5-HT1,, agonises, such as
`flesinoxan and gepirone, should be assessed for safecy and
`efficacy in children with anxiety disorders. All of these
`agents are porenrially attractive for use in children be-
`cause of their favorable side effect profile.
`
`13-BLOCKERS
`
`BACKGROUND
`The ~-adrenergic blocking agents ("~-blockers") have
`been used for children and adolescents wirh anxiety dis-
`orders or aggressive dyscontrol, although sysremaric
`srudies have not been done. The largesr body of work
`actually exists for rheir use in children for treatment of
`nonpsychiatric disorders, such as migraine headache and
`neurally mediated syncope. For example, 36 children and
`adolescents with neurally mediated hypotension were
`rreared with ~-blockers, and rhe invesrigarors concluded
`that chey were safe and efficacious (Scott et al., 1995).
`T heir role in prophylaxis of migraine headaches has been
`reported since rhe early 1980s (Forsyrhe et al., 1984).
`There are essentially no pharmacokinetic dara in chil-
`dren. P-Biockers differ on type (specificicy) of ~-receptor
`blockade, lipophiliciry, elimination , and half-life.
`Propranolol and nadolol arc nonselective ~-blockers (at
`both p, and Pz receprors) , whereas arenolol and merop-
`rolol are selective for p, receptors. These drugs differ on
`exerting central and peripheral effects, although it is nor
`clear which may play a more important role in mod-
`erating anxiety symproms. Propranolol and metoprolol
`have both central and peripheral effects, whereas nadolol
`and atenolol have very litcle central action. Propranolol
`and meroprolol undergo hepatic metabolism. whereas
`atenolol and nadolol arc cleared by renal elimination.
`Propranolol is highly protein-bound, which has cl~nical
`import in terms of drug inreracrions. Drug-drug mrer-
`actions have been reported in which P-blockers may
`increase che levels and effects of certain drugs, as well as
`decrease those of orhers, generally through competitive
`inhibitio n mechanisms. G illerre and Tannery (1994)
`reported on 2 children wirh nearly toxic plasma levels of
`imipramine when raking concomitant propranolol.
`
`EFFICACY
`
`RECOMMENDATIONS FOR FURTHER RESEARCH
`There is a need for rigoro us, controlled studies of
`buspirone in children and adolescems with various anxiety
`
`Anxiety Disorders
`Studies in adults have not shown sign ificant effects of
`P-blockers over placebo in the rrearmem of social phobia,
`
`j . AM . ACAD. l' H i lll AIHli.I:SC. I' SYC HI ATHY . . IR :'. ~I.~Y 1 '1~'1
`
`549
`
`

`

`RIDDLE ET AI..
`
`panic disorder, performance anxiery, or posnraumaric
`stress disorder (PTSD) (Liebowin ec al .. 1992; Turner
`et al., 1994), yet these agents arc commonly prescribed
`for such disorders. Data in chi ldren arc even more
`limited.
`Opm-Lnbrl Studies. Famularo and colleagues ( 1988)
`reported some improvement in II children with PTSO
`openly rrean:d with propranolol up to 2.5 mg/kg per
`day using an on-off-on design. Joorabchi ( 1977) re-
`ported char propranolol (up co 30 mg/day) hel ped 13 of
`14 adolescents wirh hyperventilation syndrome and
`suggested char this d rug might be effective in creating
`panic disorder.
`Plnubo-Comrollrd Srudil.'s. No sysrem;nic studies of a
`P-blocker have been completed for any pediacrit: anxiery
`disorder.
`
`Aggressive Dyscontrol
`Opm-Lnbel Studies. Williams and colleagues ( 1982)
`reponed that open treatment of propranolol in 30
`patients (age ranged from 7 co 35 years) with organic
`brain dysfunctio n resulted in moderate to marked
`improvement of the aggression using high dosages (50-
`1,600 mg/day). Subsequent open trials have reponed
`symptom improvement. Recently. a case report of a 14-
`year-old. multiply handicapped adolescent with severe
`self-injury repon ed a positive response to 300 mg of
`propranolol per day over a 12-month period (Lang and
`Remi ngto n, 1994). T he authors hypothesized rhar
`individuals with mental retardation whose symptoms
`are characterized by overacriviry. overarousal, poor frus-
`tration tolerance, and self-injurious behavior may be rhe
`target population, bur more studies arc needed.
`Plnubo-Controlled Studin No placebo-conrrollcd
`studies have been reponed.
`
`SAFETY
`Side effects repon ed in children are generally similar
`ro those in adults: sedation, mild hypotension. lowered
`heart rare, bronchoconst rictio n, hypoglycemia (i n
`diabetic patients), dizziness, Raynaud phenomenon,
`and sleep disruption (Coffey, 1990). Major concerns in
`children arc potential bradycardia, hypotension, and
`bronchoconsr ricrion in asthmatic patients. Rebound
`hypertension is reponed in adu lts upon abrupt with-
`drawal, so chis risk can be avoided by a gradual discon-
`tinuation.
`
`One possible effect rhar has received lirrle anention is
`that of P-blockers on growth hormone (GH) regu-
`lation. Catecholamines inhibit GH secretion through
`P-adrenergic receptors. P-Blockers do nor appear co
`stimulate GH when given alone, bur a controlled study
`fo und that lo ng-term adm inist ration of arcnolol
`potentiated the growth-promoting effects of GH-
`rclcasing hormone therapy in growth-deficient children
`(Cassorla er al., 1995). P-Biockers can also suppress
`mdaronin (Riddle er al. , 1988). This effect has provided
`rhe rationale to treat winrer depression with propranolol
`or arenolol (Schlager, 1994). T he long-term effects of
`these neuroendocrine manipulatio ns in children are
`unknown, and additional studies are needed.
`
`RECOMMENDATIONS FOR FURTHER RESEARCH
`This class of drugs needs further investigation regard-
`ing safery, efficacy, and pharmacokinetics. Controlled
`studies in patients wirh brain damage and aggression are
`particularly needed.
`
`a-ADRENERGIC AGONISTS: CLONIDINE
`AND GUANFACINE
`
`BACKGROUND
`Since the 1960s, clonidine has been used ro rrear
`hypertension in adults (see Wilber, 1980). In the late
`1970s, rhe psychiatric use of clonidine was iniriared by
`Cohen and colleagues ( 1979) for rhe treatment of chil-
`dren with Tourerre's and other tic disorders. Later this
`usc was extended by Leckman and Cohen ( 1983) and
`Hunt and colleagues ( 1985) as an alternative ro stim-
`ulant medications for the treatment of children wirh
`ADHD, alone or comorbid with Tourcrre's disorder. By
`rhc early 1990s, approximately 200,000 prescriptions
`per year were written in the United Scates (Swanson
`er al., 1995) for clonidine ar doses of 0.05 to 0. 10 mg
`administered multiple rimes during rhe day ro rrear
`children with ADHD (Hunt cc al., 1990) and some-
`rimes ar night to treat spontaneous or stimulant-related
`sleep problems (Rubinstein er al., 1994; Wilens ec al.,
`1994). At rhese doses, clonidine is considered ro have
`agonist effects on presynaptic CXradrenergic receptors,
`which result in a ncr negative effect on noradrenergic
`activiry by reducing its release (Svensson er al., 1975).
`Ar peak rimes, 2 to 6 hours after administration, this
`produces decreased sympathetic and increased para-
`
`550
`
`J A.\1 Al.Ail. Ulll.l! .~IHll t~C I'~YC III AI'ItY •. \M , ~ . MAY 1'1'1'1
`
`

`

`ANX IO LYTICS. ADREf'ERG I C AG ENTS. AND NALTREXONE
`
`sympathetic tone and results in decreased activity, wake-
`fulness, blood p ressure, heart rate, and saliva Aow.
`
`EFFICACY
`
`Attention-DeficiVHyperactivity Disorder
`Uncomrolled Swdies. Uncontrolled srudies o f the
`effects of clonidine in children wirh ADHD have been
`reponed. Hum ( 1987) rreared 8 children with ADHD
`with oral clonidine u p ro 5 J..lg/kg per day for 2 months,
`followed by a switch roan equivalent dose of rransdermal
`clonidine, and concluded that rhe effects of clonidine
`were significant and about rhe same magnitude as the
`effects o f m<!rhylphenidate (by comparison ro a prior
`investigation of these same cases). Steingard et al. ( 1993)
`conducted an open clinical trial of oral donidine (average
`dose of 0.19 mg/day) with 30 children with ADHD, half
`of whom had failed ro respond ro previous trials of stim-
`ulants o r antidepressants, and reported clinical benefits in
`16 (53%) of these cases. lnrerprerarion of rhesc studies is
`complicated by the different inclusion/exclusion criteria
`used in these studies.
`Controlled Studies. In an 8-week, double-blind,
`placebo-substitution study of clonidine 4 ro 5 J..lglkg
`per day (about 0.05 mg q.i.d.) in II children with
`ADHD, significant improvem ent with effect sizes
`greater than 1.0 was reported in 7 children, based on
`parent and teacher ratings of hyperactivity and conduct
`problems (Hunt er al., 1985). Gunning (1992) used a
`double-blind, parallel-groups design with 3 groups of
`24 children with ADHD, treated for 8 weeks wirh
`clonidine (0.03- 0.05 mg/kg per day), methylphenidate
`(0.4-0.6 mg/ kg per day), or placebo. Based on overall
`clinical improvement judged by a psychiatrist, rhc same
`percentage of subjects (50% , significantly greater than
`13% on placebo) were judged to have improved on
`clonidine and methylphenidate, and significantly greater
`than placebo decreases in parent and teacher ratings
`were reponed fo r clonidine (about 14%) and methyl-
`phenidate (about 20%). The interpretation of these
`srudic:s is complicated by rhe selectio n <.:riteria of the
`ADH D subjects, which limits gcneralizability. Also, the
`Gunning ( 1992) study was conducted in The Nether-
`lands, which has a different tradition for the recognition
`of ADHD and usc of stimulants, and a lower than
`expected response rare (SO%) to methylphenidate was
`observed in rhis sample of ADHD cases. Moreover. as
`discussed further. in 3 of 5 controlled studies of the
`
`efficacy of clonidine for the rrearment of Tourerre's dis-
`order, the effects of clonidine on behavior (symproms of
`ADHD) were not significant.
`Combination of Clonidine and Methylphenidate.
`C lonidinc is often used in combination with methyl-
`phenidate (and 01her stimulant medications) ro treat
`children with ADHD. Hunt ( 1987) and Hunt et al.
`( 1990) suggested rhar rhe combination may lead to ben-
`eficial effects on activity, aggression, and opposirionality
`at lower doses of methylphenidate than usually used, bur
`no experimental data in support of this hypothesis have
`been presented. Comings er al. (1990) made similar
`claims based on clinical experience with 4 I children with
`ADHD (and a large number of additional cases with
`Tourerre's disorder and conduct disorder), especially for
`rhe usc of clonidinc administered rransdermally. There
`are no controlled studies to support rhe combined use.
`Interactions of the opposing mechanisms of action of
`these 2 medications may result in rare bur serious side
`effects (Cantwell et al., 1997; and see below).
`
`Tic Disorders and Tourette's Disorder
`Uncontrolled Studies. Uncontrolled studies of the effi-
`cacy of clonidine for tics have produced conflicting re-
`sults. In a brieRy described. "open study" of transdermal
`clonidine in 210 patients with Tourerte's disorder, "some
`improvement" was reported in 6 1 o/o (Comings eta!.,
`1990), bur Sreingard et al. ( 1994) reviewed clinical charts
`of 7 patients with comorbid ric disorders and ADHD
`who were treated with clonidine 0.1 to 0.3 mg/day and
`found a significant reduction in ADHD symptoms bur
`nor tics.
`Controlled Swdies. Several controlled studies have
`been conducted to evaluate the hypothesis (sec Cohen
`et al. , 1979) that clonidine is an effective treatment for
`To urerre's disorder, either alone or with comorbid
`ADH D. Two early studies did not confirm this hypoth-
`esis (Borison ct al., 1982; Goetz er al., 1987). bur
`Leckman et al. (1991 ) partially confirmed the hypoth-
`esis about tics in a parallel-groups, double-blind study of
`children with Tourcrte's diso rder (half of whom had
`comorbid ADHD) treated with up to clonidine 0.25
`mg/day (n = 24) or placebo (n = 23). C linician ratings
`suggested a significant decrease in severity of tics but not
`in severity of hyperactivity/impulsivity. bur parent rat-
`ings suggested the opposite pattern (a significant decrease
`in severity of sym ptoms of hyperactivity/impulsivity bur
`not tics). In Gunning's (1992) double-blind, parallel-
`
`j . A:VI. ACA il . Cll ll.ll A I H li. ESC. PSYCHIATRY . .IX : ~. MAY I'J'l'J
`
`551
`
`

`

`RIDDLE ET AL.
`
`groups srudy, groups of children wirh ADHD plus
`Tourerte's disorder were rreared for 8 weeks wirh doni-
`dine 0.03 to 0.05 mg/kg per day (n = 16) or placebo (n =
`16), bur rhe hypothesis abour rics was not confirmed:
`fewer subjects improved in the clonidine-m:ated group
`(2So/o) than in rhe placebo-treated group (3 1 o/o). Singer
`et al. (1995) conducted a double-blind. crossover study
`of children with Tourette's disorder and ADHD (n = 34)
`treated with clonidine (0.05 mg q.i.d.), desipramine
`(100 mg/day), and placebo. Treatment with clonidine
`did not result in a significant decrease in parenr ratings of
`tics or hyperactivity relative to ratings on placebo, bur
`treatment with desipramine did produce a significant
`reduction in ratings of both tics and hyperactivity. Thus,
`in 4 of 5 placebo-controlled studies of treatment of
`Tourette's disorder, clonidine was no more effective than
`placebo in reducing the severity of tic symptoms. Over-
`all, these controlled srudies do not support the hypoth-
`esis that clonidine is an effective treatmenr for tics.
`
`Aggression
`Uncontrol&d Studies. Improvement has bc.:c:n reponed
`at doses up ro 0.4 mg/day both in outpatients (by
`Dawson et al., 1989, in a case reporr of a child with
`explosive disorder; by Comings et al., 1990, who rreated
`41 children with conduct disorder; and by Kemph eta! ..
`1993, who treated 17 children in crisis due to serious
`aggression) and in inpatients, ofn:n in combination
`with other psychotropic medications (by Schvehla er al.,
`1994, who treated 18 children with uncontrolled anger
`and aggression who had failed to respond to trials of
`stimulant medications, and by Chandran, 1994, who
`treated 60 children with uncontrolled anger and physical
`aggression). These resulrs may be due (at least in part) to
`nonspecific sedarive effects of clonidine or Hawthorne
`effects on rhe subjective measures of response, which
`cannot be discounted by the unconrrollcd designs.
`
`Sleep Disorders
`Uncontrolled Studies. Rubinstein era!. ( 1994) reported
`rhat 0.05 to 0.1 mg of clonidine at night decreased sleep
`latency in 15 children who had sleep difficulties during
`treatment with methylphenidate for ADHD. Wilens
`er al. (1994) described the effects of 0.05 to 0.4 mg of
`clonidine administered at night to treat sleep problems in
`more than 100 children with ADHD (wirh spontaneous
`or stimulant-induced sleep problems) and reported
`sedation wi thin 30 minutes which persisted until
`
`morning. Prince et al. ( 1996) described similar dfects in a
`systematic chart review of 62 cases in which clonidine was
`used to rreat sleep disturbances associated wirh ADHD.
`
`Autism and Fragile X Syndrome
`Uncontrolled Studies. Uncontrolled studies have sug-
`gested beneficial effects: Fankhauser et al. ( 1992) reported
`on the use of transdermal clonidine to treat 9 children
`with autism and reported decreases in "hyperarousal"
`noted by parents and clinicians. Hagerman et al. (1998)
`reported on the use of clonidine (alone or in combination
`with methylphenidate) to treat 35 children with fragil