benefits of maintenance neuroleptictherapy. Options include individual ad­justment to a minimum effective dose,4-7as well as the use of very low or intermit­tent dosing,B-IO as the search for safer andmore effective antipsychotic agents con­tinues. ll ,12 Low or intermittent dosing in­volves the removal of neuroleptic drugs.Such procedures and, indeed, theresearch that supports 16!i:g~tetm neuro­leptic treatment, eVidently assume thatthe removal of a drug does not increasethe clinical risk above that associatedwith the natural history of untreated ill­ness, Critical reevaluation of thisassumption is encouraged by a recent

••.!.; ..:'

.. :..: .

ORIGINAl. AltTlCLE

', .. "

ARCH GEN PSYCH1ATRYNOL H. JAN 11)1)749

..... --,--_.._.__ _._-------------

allOWING THEIR introduc­tion in the 1950s, neurolep­tic drugs became the corner­stone of pharmacologicaltreatment of psychotic dis­

orders. The findings from many studiessupport their short-term efficacy and long­term benefits,l.3 Most studies of neurolep­tic maintenance have involved the inter­ruptfo-ri ohreatrnent to compare a placebowith continued medication. Meta­analyses of such studies have found highrates of relapse in the weeks after the in­terruption of active-treatmentY Gilbertand colleagues3recently concluded that therisk ofpsychotic relapse within 10 monthswas only 16% if antipsychotic medica­tion was continued, and 53% after discon­tinuation.

Late adverse effects (particularly tar­dive dyskinesia) encourage attempts to .minimize the ,risks without a loss of the

F

------------------

Clinical Risk:fQ~!.g~~g Abrupt and. Gradual ...Withdrawal ofi·~~~~tenal}.q.e NeurolepticTre~~,~~nt

, ;", .·".:{~f: ..;;);/~~;_.:~:'i<i~:;:,r ,,:'r! l. ~.- , '. ';';.:' . "', '.

Adele C. Viguera, MD; Ross]. BaI4es~arini, MD; James D. Hegarty, MD, MPH; .-..!. 'Daniel P. van Kammen, MD, PhD;:-¥auricio Tohen, MD, DrPH

From the ConsolidatedDepartment of Psychiatry andNeuroscience Program,Harvard Medical School,Boston, Mass (Drs Viguera,I3aldessQlini, Hegarty, andTohcn); the General AdultPsychiatry Program andLaboratories for PsychiatricRescarcll, Mailman ResearchCClller, MeLea II Division,Massachusetts GeneralHospital, Belmont(Dr Baldessarini); theDepartment of Epidemiology,Harvard ScllOol oj PublicHealth, Boston, Mass(Drs Viguera, Hegarty, andTohcn); and the Departmentoj Psychiatty, University ofPittsburgh, and Highland DriveVeterans Affairs MedicalCenter, Pittsbu rgh, Pa(Dr van Kam11lCll).

"

"

Backgro~iul::Abrupt diSC;~~ic ... ,<o"~:i~;i~l}.g-term,,, to 3.7 years; inp~tients relapsed m~;~\:i:~pI~{~~handidpsychotropic 'medication caiLb~:,[9nOwed;by,a''high.'; ;'. 'Outpatients (10 vslSweek5' to a25%:&¢1~pserisk).Inrisk of early relapse, This stU:dYilti;J1ea.'ioAuantifythe studies including subjectswhoseclrug:therapy was

. relapse riSkover time in pat~e~:~~~]t~c,hiz<iphrenia" withdrawn abrupqy·Cn=4.9) vsgr~C:ll!~}ly:,(n=58),following disc6ntinuation6fm:afrt~·'·'·-~ ""i-peuroleptiC'·: : _relapse was earlierafterabrupt discq'iitm#¥ti.on(25%

treatment.,~·'·A;~~·"j':·'~;;yn:\·_·.~:f';:·.:.:i~fe~~t~ :o~ut:e~)i..Wlt~ap'e~~¥tl'"'~' :!~~Ii~c~ for

.. ~~;~~~ff!:'~~~~~~i.Jlm~~;')'l:::=:;o;,:{~~U:;~'~:l ~~E;!t#~RUp~iJi$.r6search, combined with new:'d.~i:aW~ffcf¥~Mtiateaby ticularly in hospitalized patients. Mos~';ip¥1tients whosurvival analysis. '::\';: ';:;/;J;;';.'i,i",,':;":"!: remained stable for 6 months coni.iiiried~todosofor

;{ ·'.l-\:ji :~{;,f;~>;');j;:; .:" .'. IPl1gpqiods witl1o~tmedication.,ind.i~~tirigclinicalResults: Data were found,:J~t:,t2J9':~~h;zop?renk<.....chS~~rogeneit}vr)iug;:WithdrawaL~)'E~~~9!~.;J~!i~te.g,to

~:}:c::~ t~~~~~r?;;;J;~e;;'ts~aJ~-'~ .:,,~~~Y:;U@~:,i :;:;~1~ld.t~:~~~~~~?~~6~~ra~~~/*i~J~~il~~~1~~."therapy, and 204 discontin,1r ......;I~i _Jrii'~nt'gradually' ··;·'tinuing oral neuroleptic therapy<qri~s~QP-piIj.g'd~pot(~3 weeks) or stopped trditff[~I1Gwrili"'Cl.e.poT·neur6:-;:;";:lrijections,'earl}/relapse' -may he''Spltt'e'a:hya slowleptic drugs, After abruptdi~~ppf~~~~tioii;oforal . removal of drugs. . .medication, the risk of relapse reache'd-50% within 30.weeks, with remarkably little additional risk thereafter Arch Gen Psychiatry. 1997;54:49-55 .

!"eripherdliii!iel!Ji!J,·synlpIl!,:·'b,temnl iheSt.poiginggiiiiiIk6)'"iCOle veins', .(entral :lid.'; .lers - (liquW:conf,jslol,. obnormnl gnit;-hypere~;\t1y,kineliiJ, dyiphiinici•.hW:Ot:.·

n, ~ruriltJs, e¥.hemntous.IOI~·;·:

!Oil lexlure, hypertrichnsis. poo"lisarders - ROle: exo~'Ji',erliculili" fe,"1 incnnlinelKe•.,.pl;c ulcer, prnctilis, sIOll1<l~Iil,

1; Infrequent: ITllllni'e. genen>:IS medio, nphthnus ,'nmaliijs.,,;or chnmber eye hemnrrhoge.ql'leemin. Musculoskelelol,,,; Rnre: hem~. psychia'rkdepmonolizntinn, d,pre,sion,,de ideolion nnd ottempl, ",j,.'e - Inflequent dysmennn1>lolit (2), femnle breo~ poin (1),S: (2) -%bused on fema~ ~ring. ~pnen, epi'ln~'; ROlflIl00n"on, conjuncti,ilis, d.,pl.""I. Urinary System DrsOf'·u. lenni pnin, urinary letenl~

: ond SGPI [0' All)) bnve b<!ll'pnlie enzyme tl"vnlions nwOItnlinunlinn. IOlOFT Ihernpy"gl:IIpplOximnlely 5~~J. ond a',moI

HC

RUG ABUSE AND DEPE •,lied ,ubslnnce. Physlcal a•i .IIlnn" fnl i~ pOlelllinllnr abUII,lid nnl revenl any lend,ncyf~D.:_i[~nI,hnuld cn"fully evolU<1I'.!.•lOLOFT mi,us, or nbuse (e:gJ;~Huma. bp.rIt... - As. 'j: ..

II 18 we" oven!oses of IOlO~lf :

~ ~~ .

··10.2:t0.6 'There weifirst 6 moure rate re'another I(Table 2

Oral I

inpatienl~,

removal eli[Figure :Iwithin 10patients n40.8% of!)years of I().

med~Iiliol

for inpalil"

Wi,l;$~f5etfings'§Mean durIIDepot estlIDoses art#Relapse c ­*,* Weighte,ttData frolttA. I, Gre

to R. J. B., Jt§§GradualIII1Gradual·;m1Gradual i ­

##Gradual.

'.~ ,

;;

'.1 ...~ -

1-

,.':,'.

The survival function after abrupt diSconrtmia:tibnoforalneuroleptic treatment in 1006 schiiophrenic'paiients(Figure -1-) indicated a rapid failure of clinical stabilityWithin 3 to 6 months, reaching a relapse risk of 25% within

itESUtTS

. . ".... ~ .

ings24 suggest a period of unusuallyhigh;ielapserisk....• after the removal.ormaintenance.tre~.~f~§;~~!i~~S<ln'. be reduced, andriotInereIy dday-ecl·;:.:b'}gnldtial

.'removal of medi~~~i,2~i;..>; . .· .... ~;;,;F·::.tf0;;.:·, .....,. Discontinuing ciralJjeuroleptic me,;,...;.,'n",;flrr~C~~;

:"larly abruptly, alsos~;rij.~a high riSk'ofp~yc::,lj,QticIiio,r-':', bidity.3,ls The hypoth.es~,,thatthis risk is,Y~tY:J#gJ:1wi,thin.

the first several montl1Safter discontinuaH'Bn>and less.thereafter is supported by results of a pre!i¥ip.aryartaly::sis that indicated a 13-fold increase in relaps¢"\vithin 3m'onths of discontinuing neuroleptictr~!ffi~prt#~'#sk .'

. fell below 2-foid in the second year.lsW~·pY~4~~~e.C:lthat

: slow discontinuation oforally admin.is'i~!~~.tE§;~r91~ptic;.: drugs or stopping injections of long-act!iig';ag~;tii,$}Y9uld

delay and perhaps even reduce this risk;JsSurviy~lanaly- .sis was used to quantify the risk over tini~aft~rthe in:' .terruption ofneuroleptic treatment in patients with schizo::phrenia, based on data from previous reports as well asfrom studies of new subjects.

..,~. -

MATERIALS AND METHODS diagnosed as having schizophrenia and followed up afterdiscontinuation of maintenance neuroleptic treatment.

Characteristics of the' 22 cohorts that were stud-We searched for studies utat involved the abrupt or gradual ied2~'i2 arc summarized (Table 1); some studies failed todiscontinuation of maintenance treatment with oral anti- specify drugs and doses but did indicate when only oral orpsychotic agents or stopping injections oflong-acting prepa- depot medication was involved. Definitions of relapse orralions in patients who were diagnosed as having schizo- exacerbation of illness varied but usually involved clinicalphrenia. MEDLINE-computerized searching and references assessment or the use of rating scale scores to indicate theobtained from the resulting reports yielded 11 studies with worsening of psychotic symptoms severe enough to war-data on the time to relapse for individuals, or survival analy- rant hospitalization or reinstitution of antipsychotic treat-ses of groups, and provided 1006 subjects (795 inpatients ment. Discontinuation and follow-up assessments wereand 211 outpatients) who were abruptly withdrawn from double-blind in 20 of the 22 cohorts (only 2 studies wereoral neuroleptic maintenance.25.35 Similar new data in- open). "Abrupt" discontinuation usually involved stop-volved 94 subjects with schizophrenia according to DSM- ping neuroleptic treatment within 1 day; "gradual" dis-lIl·R criteria ,who were rapidly discontinued from oral halo- continuation included the tapering of oral doses over atperidol therapy at the Highland Drive-Veterans Affairs least 3 weeks (mean::!::SD, 3.39::!::6.00 monlhs);br'nofuro "

Medical Center, Pittsburgh, Pa (methodological details have ther depol neuroleptic therapy after a final injection.been reported elsewhere36,37) and 6 similar subjects from a Treatment averaged 7.75 ::!::6.07 months and postwith-sludy at the Massachusetts Mental Health Center, Boston, drawal follow-up after the last dose· averaged 54::!:: 46on removing an average of 85% of the initial medication weeks (range, 10 weeks to 4 years), or 16, 20, and 17(A. l. Green, MD, S. V. Faraone, PhD, W. A. Brown, MD,). months after discontinuing oral medication treatmentGuttierez, MD, and M. T. Tsuang, MD, DSc, PhD, oral and abruptly or gradually, or stopping depot injections, respec-written communications [generously provided by Dr Green tively (Table 1).to R.].B. J. June 1995). Four studies (n= 107 subjects, in- The relapse risk over time after the discontinuationeluding 7 from Dr Green) provided additional data on of neuroleptiC lhe,.rapy was analyzed by Kaplan-Meier andgradual (>3 weeks) withdrawal of oral medication28,38,39 actuarial survival analysis, with variances, and comparedFive studies (n=83 cases, including 8 from Dr Green) in- statistically by Mantel-Cox nonparametric log raIik tech-valved stopping injections of a long-acting neuroleptiC niques to provide a X2.20.21.'I3.+'I These values, as well as thedrug]1:!o.H We excluded several studies that involved si- time to a defined percent reiapse::!::SE, were calculated withmul taneous or undefined mixtures o(oral and depot neu- commercially available microcomputer programs (~tatview/

roleptic medications, intennitt~ntIi~~roleptic therapy, Survival Tools, Abacus Concepts, Inc, Berkeley, CaliQ:Un-or imprecisely definedtiJD.ing;"i?f, ··\~~S?:mll,}uation .. :,.,;}ps otherwise st~ted,.~~~·~!t:: pr~~~tec!~st.h.;'~1~~~'~~~p." ..Analyses are based on finding~;r . or~ plus 2.i~;.orrates::!::SE;2-talledstaustJ,c~lslgniflcancer~qti . ,.:'S..,D:i

unPUb]iSh~d.~a,~ se.~, ..Yi~:'~!!:" ·"~~~?:~~~~~;;~'~ii1~~uignifiCance,. ~.~~1:J'~¢.t1'F:·~ '.':., ···.'''C::8~i~~~\j)i;(?~(~:~;;~~~~'~;:; ..;:"·.;

ARCH GEN PSYCHIATRYNOL 5'1, JAN 199750

reconsideration of the risks inv~iyMjliexperimentalprotocols that require interriip(16nJ:>tlIi:<li~tenancetreatments. I3-17 . '.. .,.

~igh early morbid risk.f9~~g,~?4~\,..~;~g,fiki\e~~tion .of mamtenance. treatments utsey~ral\Ils}'~.lj~atl:1.c.. dlsor- ..ders. In addition itO early withara.W~r;Ei~·¢'tibnsto theabrupt removal ofbenzodiazepiri4Wfp~i:i~ntswith anxi~'ety disorders, IS and physiologicalsyIll'p't9msassociatedwith the removal ofs~me antipsych~~c,a.g.dantidepres­sant agents,19 worsenmg of primary'dlSorders often fol­lows the rapid removal oflong-teiri(psy~hQti'opietreat­ments. This response includes recurrence'of affectiveepisodes in a majority of patien~~:Wth:bipolaiI Of IIdisorders within several monthS',bf~t9PJ>pigsuccessfullithium carbonate monotherapy after several years,20.22as well as in patients with major depressi6n after inter­rupting similarly prolonged imipramine hydrochloridemainlenance. 23 Moreover, after stopping lithiumtherapy abruptly in patients with bipolar disorders, thetime to a first recurrence was much briefer than theshortest intervals between spontaneous recurrencesbefore starting maintenance treatment. 20 The risk ofrecurrence of mania or depre~siohhfpadentswithbipolar disorders was reduced)il#'kedly by gradualdiSLontinuation of lithium therapy; even during sev­eral weeks,21,n These I7 ,21,22 and congruent new find-

ARCII (;I'N rSYCHIATRYNOl. 54, JAN Ill'l75\

ences between hospitalized and ambulatory patientsWere similar when those who were undergoinggradual removal of neuroleptic drugs were included(data not shown).

The survival over time was very similar after thegradual discontinuation of oral medication over anaverage of 3.39±6.39 months and stopping depotinjections (n=113 and n=91, respectively; X2=0.12;P>.lO); thus, the data were pooled to provide a groupof patients with gradual discontinuation of treatment.There was no significant overall difference in theresulting survival functions for those who discontin­ued treatment abruptly vs gradually (n= 1006 andn=204, respectively; X2=1,J,,0; P>.10), although thetime to a 25% relapse risk tended to be shorter afterabrupt discontinuation (l1.0±0.3 vs 15.0± 1.0

".'. -' ..... ,,;._-- ~, ..-,:'~ -. . ." .•--' ..*'....-: '.:;.:,-';"~t'•.•~ ..•..1.:.... "".'*Thesestudies of(:oho'flS for which neuroleptic'm'afnteiianCiftreatfrien 1210 psychotic inpatients or outpatients'who ;were iec'iHvingng-tenn maintenance neuroleptic treatment in clinicaltj:sfable status prio .... .... iagnoses were based on unspecified clinical criteria,. 'Rese"!'!!h :iagnostic Criteria (ROC), or the American Psychiatric Association D~M~I1Lof;. " Mdt ..',~;;,'; ';", ,;>;:i': ;'" ,:<, ",;;\ "..::j;::(;n:'!i/i;,::;,c;tGender was defined as the percentage of men; (50%) indicates approximatelyeqiJa numbers of men and women sUbjects; approximately 69% were :ffien:::;'" 'tSettings of studies were inpatient (I)' units (41%) or outpatient (a) clinics (53%). :.,£ "::~.; , . '. . . ',' .,§Mean duration of treatment prior to discontinuation was 1.75±6.01 months or more.:'IIDepot esters (decanoate and palmitate) were injected intramuscularly..lIDoses are average chlorpromazine-equivalent milligrams per day (oral) or actual milligrams per 3 weeks (depot).#Relapse criteria: A, clearly worse clinically or by ratings; S, antipsychotic re-treatment required; and C, hospitalization required."Weighted (by n) mean follow-up time (months) was 15.6± 16.1 (abrupt, oral), 19.5±13.9 (gradual, oral), and 11.4± 11.1 (abrupt, depot).ttData from van Kammen and colleagues were previously unpublished, but their methods were presented elsewhere.36,37

ttA. I. Green, MD, S. V. Faraone, PhD, W. A. Brown, MD, J. Guttierel, MD, and M. T. Tsuang, MD, DSc, PhD, oral and written communications from Dr Greento R. J. B., June 1995. .

§§Gradual discontinuation time was 420 days (weighted average, 3.39±6.39 months).III1Gradual discontinuation time was 23 days (weighted average, 3.39±6.39 months).lMIGradual discontinuation time was 30 days (weighted average, 3.39±6.39 months).##Gradual discontinuation time was 60 days (weighted average, 3.39±6.39 months).

10.2±O.6 (±SE) weeks and 50% within 30.3±15.4 weeks.There were remarkably few additional relapses after the

.. first 6 months without medication: the computed fail~-"

ure rate reached 46.0% by 6 months, and it increased onlyanother 10.2% in the period from 6 to 24 months(Table 2).

Oral neuroleptic drugs were removed rapidly in 795inpatients and 211 outpatients. Their stability after drugremoval differed markedly (Mantel-Cox X2=28.4, P<.OOI[Figure 2]): 25% of inpatients vs outpatients relapsedWithin 1O.0±0.62 vs 18.0:!: 1.65 weeks, and 50% of in-

. patients relapsed within 18.0:!: 1.65 weeks, while only

.40.8% of outpatients relapsed within a maximum of3.69years of follow-up (Figure 2). Within 6 months withoutmedication, the relapse risk (:!:SE) was 49.6%± 1.8%

:for inpatients vs 31.4%:!:3.2% for outpatients. Differ-

)se riskhat canradual

ionmdred '.ch­thelithewlIn­SD.05

ld­[to

loror

calIhe'ar­:al­ereereJp­lis­. atL1r­In.lh­4617

ent,ec~

Jarticu­ic mor­1 withinmd less)' analy­,ithin 3.his riskled thatroleptic5 wouldLI analy­. the in­I schizo­,well as

i11 of oralJatients~tabi1ity{, within

lapse rate waafter the al:therapy. an,1, ljable 2)wh6se conclera' monthdrugs, the Iited\(Figuronly\40% 0:This\findiranalyses ofabruptly rcment, in wI­years of foil

Most 0

of oral nem,agnoses (Tacriteria for S

the clinicalstable with,might reOe,ity.46,47 How(=40%) ofdementia r ..century ha\Moreover, ­discontiml'schizophr("modern DSonly 62% (

. relapsed."Observed h,: medication

".. however, tJ. .'mates of th,

were diagnwhose COIle

rIsk period. An es:early morbuntreated s,911t pro tee

'relapse is gneuroleptic ­throughouadministralever,' an eaJcourse of Irelated to 11reflect phalprolongedinduction cbrain49 ; thilevels evenanimals,49may be COl

relapse ris!treatment (reOect the.albeit protmost vulne

The Ii,

fr

96

Gradual (n=58)

Abrupt (n=49)

16 ...... 32 -·-----48~ ..--.. ··- 64·-- --. 80·­

Weeks Alter Stopping Neuroleptic Therapy

20+---,.---,.---..,.......---,----.-----,o

40

80

90

100

~ 70'"·19

en. C> .

~;'600;'E

'. ~50....

30

Figure 2. Percent "survival" of schizophrenic inpatients (n=795) andoutpatients (n=211) whose conditions remained stable after abruptdiscontinuation of oral neuroleptic therapy. The risk was greater forinpatients (x'=28.4, P<.001), of whom 50% relapsed by 5 months;outpatients were followed up to 4 years without reaching a 50% relapserisk (data not shown).

" ;i::....'...·~\~iV~~¥i:J·i,:·((,:ti::"',,~',,: .; ..~: .... :."~!~',:::i::~~d-'~;·.·~·:i,· ':"Figure 3. Percent "surviva(.".''ofschizophre'nlcpatiiJntS''whrj'seconditionsremained stable after tliii ii~rup . "o!itinuiitiiiiiot'neiiriiiiipiic"tlierapy(n=49) vs the gradual disij'(]"!!~. :h;rW .Stoppag~'ot~ep'o{inJections[n=58]), based on reports withdafa for.both condltlonsm the samestudy28.34 (A. I. Green, MD, S. VJ~raone, PhD, W. A. Brown,MD,J. Guttierez. MD, and M",'{,.'[sua,ng,,MD. {)SC, PhD, 0~1 aM written .communications, June 1995/;" The dsf('Wasgreatelfoi'paiiiirits' who werewithdrawn abruptly (x'=11.1;·.P'<;001)i Of wliom50% relapsed: within 2.5months, while outpatientsA1flflPtreapir that JereJ.,: ,'; ),.::'!'

TImiito 50% Risk

5O+--.--.,.---r~-,....-""":~...,

o 4 8 12 16 20 24 28 32

Weeks Aller Slopping Neuroleptic Therapy

100

24 "'48--'-"72""-96 '... 120······ -'144 168 192Weeks Alter Slopping Oral Neuroleptic Therapy Abruptly

20+---.----'".,"'·-;.....,--,---.,.-----r--,.---..----,o

30

40

100

~ 70~co·Z 60'iiIE'"a: 50....

Figure 1. Computed "survival" functions based on findings from studiesthat discontinued maintenance oral neuroleptic drugs in patients withschizophrenia (Table 1). Data are the percentage of patients whoseconditions remained st11ble vs the weeks after the abrupt stoppage oftreatment (n= 1006). Dashed lines indicate 95% confidence intervals.Inset, The time to a 50% relapse risk (1.5 months).

COMMENT

ARCH GEN PSYCHIATRYIVOL 54, JAN 199752

•CI indicates confidence Interval. Based on Kaiilan:M~ier survivalanalysis, with 95% Cls.a~shown in Agure 1.

<.: "; :; ..>~.;~ ,.... . .. :' : j..~.;: ":,.' ..,' <;\~.~ ~ ..~.:··l .'

weeks). Gh7f?o!:i~1~1:~Hg~~~Hy¢\gj5~~~~·tionand theprobable het~fdg~ti,~ity,::aIIl-prig:~tpdi,~si"we furthercompared;re§~l~yW)Iii'thei3~stiidj;f~·,'i1Thtwerefoundwith abrupt(n·~;t~r~n~gradu:lr.(ri~,$8)removal ofneuroleptic agenrs\B.thm the sanie:trial~.3'I (A.I. Greenet aI, oral and Written commuIlic"tions, June 1995)(Table 1). With data pooled from these. better-matchedcohorts, there was a significant difference in survivalfunctions (X2=1l~1{R<.00l [Figu..e3lr After abruptvs gradual discoritiiluation; respectively; the computedtime to a 25~~tela'pse riskwas:q;OO± 1.50 v~10.0±~.73 ,,:eel)sian,d the comPl.lted probability ofrelapsmg wlt~m~(jmonthswas:32.5%±3.0% vs64.90,6±6.50,6 (a 2-fold difference).,:·, ..

The present findings from 1960 to 1995 should beinterpreted cautiously owing to the variability in diag­nostic criteria, lengths and methods of follow-up, anddefinitions ofrellips.e,· as well as. the types, duration,and doses of neuroleptiC drugs (Table:!). Most reportsalso provided little information about possibly rel­evan t aspects of clinical history, current state, and

ARCH GEN I'SYCHlATRYNOL 5-+, JAN 19975t

chlorprOIl123:221-,

29. EngelhartlI tlon of ps'I evaluation

30~ Prien RF, I

i withdraw,3\ left JP, ~

\ chlatry, 1~

32, Hogarty E\ the alterc,"chiatry, l'

33. Clark ML.phrenic 0

34. Levine J,R, Steinbphrenic pchopharn

35. Cheung ~

stop or CI

36. van Kam,adrenergiwithdraw;

37. van Kam,Peters JLdrawai in

38. Brancheyon clinicachiafry, 1

39, Crow TJ.trial of p'

40, Denckermuniiy "1980;61\

41. Wistedl F" drawal (.1

The AI

offici" ioffice­

If(',each II

;;;'Ameri,fAsubs<trainin

',.;

"~

, 'do L, Tohen M. Disc:onlinuing main­iession: ,risks and, implications. Har-

;:p~es T, Tohen M. Outcome after rapid'1ar mood disorde'fi.Arch Gen PSy-

i?f.. :" "0; ~ :;.(/.,"'-'"

, IlErEIU:NCJ:S

1. Balde~sarinl RJ. Chemotherapy In Psychiatry: Principles and Practice. Cam­bridge, f:v1ass;Harvard Urilversliy Press; 1985.

2. BaldessarlnlRJ,'Drugs and the treatment of psychiatric disorders: antlpsy­chollcarid' antlanxlety agents.'ln: Harden W, Rudin W, Mollnoff PB, Rail T,eds: Goodman and Gllmari's the Pharmacologic Basis of Therapeutics. 9th ed.New York, 'NY: McGraw-Hili Book Co: 1996:399-430,

3. Gilbert PL, Harris MJ, Jeste OV. Neuroleptic withdrawal In schizophrenic pa­tients: a review of the literature. Arch Gen Psychiatry. 1995;52:173-188.

4. Baldessarlnl RJ, Cohen BM, Tslcher MH, Significance of neuroleptic dose andplasma levsl In the pharmacological treatment of psychoses. Arch Gen Psy­chiatry. 1988;45:79-91.

5. McEvoy JP, Hogarty GE, Stelngard S. Optimal dose of neuroleptic In acuteschizophrenia: controlled stUdy of the neuroleptic threshold and higher halo­peridol dose. Arch Gen Psychiatry. 1991 ;48:739-745"

6. Inderbitzen LB, Lewine RRJ, Scheller-Gilkey G. Swofford CO, Egan GJ.GloersenBA, Vidanagama BP, Watemaux C. A double-blind dosage reductiontrial of fluphenazine decarioate in chrolifc-'-unstabie schizophrenic patients. AmJ Psychiatry. 1994;151':1i53~lj59: '

7, Smith RC. Lower-dose therapy with traditional neuroleptics in chronically hos­pitalized schizophrenic patients. Arch Gen Psychiatry. 1994;51:427-429.

8. Carpenter WT, Hanlon TE, Heinrichs OW, Summerfelt AT, Kirkpatrick B. levineJ, Buchanan RW. Continuous vs targeted medication in schizophrenic outpa­tients: outcome results. Am J Psychiatry. 1990;147:1138-1148.

9. Herz MI, Glazer WM, Mostert MA, Sheard MA, Szymanski HV. Hafez H. MirzaM, Vana J. Intermittent vs maintenance medication in schizophrenia: two-yearresults, Arch Gen Psychiatry. 1991;48:333-339.

10. Jolley AG, Hirsch SR. Contirllidiisvs' iriiermitlent neuroleptic therapy in schizo­phrenia. Drug Sat. 1993;8:331-339.

11, Baldessarln; RJ, Frankenburg FA. Clozapine-a novel antipsychotic agent N Eng!J Med. 1991;324:746-754.

12. Meltzer HY. Atypical antipsychotic_drugs. In: Bloom FE, Kupfer DJ. eds. Psy­c~op~armacology:the roiJithGeneration of Progress. New York, NY: Raven

,1}~~~i~;~~~~~i~l(:~J~';'"2:;>illr~~~:hical use of Place~o controls,

; 14. la, ~~.,9 )rials disputed.,.science.) 995;

:~(~l~essanm " 'afth'd;~~1i inst~i~h~~~enic patient!l.

16.'1;~r~th;~~~~riPt" •"'c,:; " ,'~';;tena'nce~dtlpero~ "il'm:'needfor individual rather than dogmaticapproach.Arcll GenPSYChiiltr{: 995;52:209-212; , ,

17. BaldesSarini RJiSu~""~<"'" '!il:Li1liium withdrawal in bipoiar disorder:•• impi'iciiuons fo(clirif '~~~elimental therapeutics researcti. Am J1il){!i~1~~~' ,, , , "", '

"'19" riiiSaver eler noamine oXidasei~~ibit~~and neu-

',;:'.;;!,~~~~~~n~~~~:,.. .... ,OPSYc.hoPharm~?,o':fiO!;PSYChiatry.20. 'SlippeS'T; Baldassari " ,,:Tohen M. Risk of recurrence following

discontinuaiion Oflit6iu#!r~~@ie~jfn,biPolar disorder. Arch Gen Psychiatry.1991"48:1082~1088> "

21. i=aeiidaGL,TondoC~

·2;.,~:~'1~i:l:~r§N~f ".lenancetreatment,hilip

" .,~ardRev f'sychlaui..:i9923. Kupfer OJ;' Frank E; P~rer

, AB,Gioctioclnski iJJ FIV"me'for maintenance therapies in recur­rent depression. ArCh Gen Ps'whiaii,Y/1992;49:769-773.

24. Baldessarlnl RJ, Tondo,L, Pcieddii,G;Jlorls G, Suppes T, Rudas N. Effects ofthe rate of discontinuing HthlulTlmalnienance treatment in bipolar disorders.J Clln Psychiatry. 1996;57:44j:.448/ ,"

25. olscin GW, Peterson DB, Suddei{removal of tranquiliZing drugs from chronicpsychiatric patients. J Neiv Msnl'DI,i>1960;131:252-255.

26. Whittaker CB, Hoy AM>WlthdrawaJii(tlliiphenazlne In chronic schizophrenia." BO psyCh!a(ry,1963:1.09;4~?;1g7,~:F,;;. , ,27. Gaffey EM, PlalJlo~d4i;JrMM~; !iriis~erger JC, Herman L, Klett CJ, Roth~

steln C. D1scontlnualfon 'ofred~'Ct1ori pI chemotherapy In chronic schizophren-Ics. J Chronic Dis. 1964;1'7:347-358.' --'-''''',

26. Greenberg LM, Roth S. Olfterentialeffects 01 abrupt vs gradual withdrawal 01

s1mptJmalic differences between hospitalized and, ambulatory psychotic patients,2,iS,"7

'T,ne full implicatioriS'of the present findings are lim­ited bythetypitally coinpl~andirregular course of un­treated schizophreriia46,i7 to be compared with the coursefollOwing discontinuation of treatment. Although clini­cal and possible ethical considerations now severely con­strain observations ofmitreated psychotic patients for pro­longed periods,l3,H treatment noncompliance as well aselective clinical practice often interrupt antipsychotic treat­ment for varying periods. In psychotic affective disor­ders, intermittent-interrupted exposure to neurolepticdrugs is particularly common, and its potential contri­bution to clinical instability isunknown.I,2,s4,ss In addi­tion, U!o_~,L~~~tLmentalprotocols that evaluate thelong-term efficacy oTneuroleptic agents and manyother drugs have compared continued vs withdrawntreatment. I-3,IS-17 The results of such studies that sup­port a therapeutic effe'ct usually have not consideredcontributions of drug withdrawal to observed drug­placebo contrasts. Moreover, protocols that have beendesigned to study patients in a presumed drug-freestate often have employed a drug "washout" that wasprobably too short to provide a drug-free and physi­ologically basal state. I,2,4S

In conclusion, rapid discontinuation of mainte­nance treatment with short~acting oral neuroleptic agentsin schizophrenia,capi~da 50% risk of relapse within 30

relapse. Ing~ner~I~i,W~':present findings add to grow­ing' e~denc~e' that'''ilqfupt: discontin'uation of mainte-

~t~C~ir~'e~t~m~r~~~~~~P:~Y~~~f::a~o~o~~;;~t;,atriC'dis6rd~ ,"enomerioii'shbuldbe consid-

~~~11li;~l!li~~~~t~:n~:E·counsdiIig:'I"iitient~7JoGiadualdoser-eduction (withclose cliniC'alfC?lloW:iHpj'inay limit relapse risk associ­ated ~thi~t~@li~.9Ji:8fmaiiltena.ncejtEeatment, andis reconiriierided~r1z;J~1::~Y!);:' ',"•"',': ,;' ,; ,' ;,

. ~-".n}}~_;( ::: ";'." i. :.... :

Accepted for;~W?,_." }flanuary3" ifJ96;.ThisstuaY<Wg$f'Sflpjiorted in part by grants MH­

31154 and MH-47J70!fom the National Institute of Men­tal Health, Rockvm~LMd, and by awards from the Bruce]. Anderson Foundation and the Adam Corneel, Anony­mous Donor, Vivian'Temte, and Richard Whalin Psychi­atric Research FihfdS~ Boston, Mass,

Alan Greeii:'Mb/ge~erously provickd data from hisstudies: Mary K~11~V1v.fS;. SHvina Zarate, and MichelleWets.S; MS,assis[t4'-'wUh' data analyses:

CorrespoTlding'aiithor: Ross]. Ba!dessarinl; MD, Mal/­man Research Center 316, McLean Division, Massachu­setts General Hospltal,lI5 MI/l St, Belmont, MA 02178.

chronic S"chlzophrenlc patients. Acta Psychlalr Scand. 1981;64:65-84.42, Sampath G, Shah A, Krska J, Sont SO, Discontinuation in the very stable Sclli

phrenIc patient: relapse rates and serum neuroleptic levels. Hum Psycllopllmacol, 1992;7:255-264,

43. Cox DR. Regressldn models and life fables. J R Stat Sac SfJr B. 1972;34:1/220 ", "".: .,: ".' . '

44. Kahri;~II.*~mp~sj~T" Slatlstlcal Methods {n EpldemlolollY. New York, ~Oxford UnlversltY.press .Inc; 1989. .,

45.· HegartY JOiBaldessarlnl RJ, Tohen M, Waternaux eM, Depen G, One hundlyears 01 sehl~ophrenla: a meta-analysis 01 the outcome literature. Am J P:chlatO': 1994;151:1409-1416.

46, Fenton WS, McGlashan TH. Natural history of schizophrenia SUbtypes. AiGen Psychiatry. 1991 ;48:969-977.

47. Ram R, Bromel EJ, Eaton WW, Pato C, Schwartz JE. The natural courseschfiophrenia:Yieview of first-admission studies. Schizophr Bull. 1992;1185:20iJ '.T:;>( :.', .

48. Cotie{BM.'T~bnelzuml T, Baldessarini RJ, Campbell A, Babb S. Differenc'betWeeriailtiiisYChotic drugs in persistence of brain levels and behavioral tfects:.psyciiB/iharmacoiogy (Berl). 1992;108:338-344. .

@9. Baldess~rtriL6J;;tarsy D. Relationship of theaetions of neuroleptic drugsthepathoP~yStpii:igyoftardive dyskinesia. Int Rev Neurabial. 1979;21:1-45

5°'~~~:ft~:~~~tfu:~ ~~;:~;~~~m~:~~~~n;;~~i 1a:;~;;~:~~~I~~tiCS: tim,51. Eklund K,Fo'rs;n;rn; A. Minimal effective dose and relapsl7-dollble-blind triJ

haloplMit&l.decaHiiate vs piacebo. Clin Neuropharmacol. 1991;14(suppI2):SiS17·"fJ:~;;L:> . :

52. BaldessannrRJ~;Gardner OM, Garver DL. Conversions from clozapine to otht.anti .,' . . ));}n. Psychiatry. 1995;52:0DO-000.

53., . > ,-- ... ,Geisler s. Ramos-Lorenzi J, Woerner M, Novacenko •~iio "'hylphenidateresponse, psychopathology and tardivdYs'. of relapse in schizophrenia. Neuropsychapharmaca,ogf)994;L 7;H!{:;·· .

54. Choii'jJ:'-:,'(Ji!i'JC'"ent"advances in the treatment of acute mania. J Clin Psychopharmiicii(j991;11;3~21 :

55. Sernyk Mj,Wo6ds~·sW. Chronic neuroleptic use in manic depressive illnes,PsychdPh~rtiia~ojiJiJli '1993;29:375-381.

. ::';':~~J:.~:~~~{ -·~·;:h., -: ;

: :. .

:. t·,·..:,'.

"',.'('::"

chlorpromazine In hospitalized schizophrenic patients. Am J Psychlalry. 1966;123:221-226.

29. Engelhardt OM, Rosen B, Freedman N, Margolis R. Phenothlazines In preven­tion of psychiatric hospltal/zatlon: delay or prevenllon of hospitalization, aJe-

" evaluation. Arch Gen Psych/aIry: 19~7;16:~8:1.QL_ ",30. Prfen RF, Cole JD, Belkin NF. Relapse In chronic schizophrenics (ollowlng abrupt. . withdrawal of tranquilizing medication. Br J Psychlalry. 1969;115:679-686.3L LeffJP, Wing JK. Trial of malnteriance therapy In schizophrenia: BrJ PSY-, ~hlatry. 1971 ;11 :599-604.32: Hogarty GE, Goldberg SC, Schooler NR, UlriCh RF. Drug and soclotherapy In

the' aftercare of schizophrenic patients: two-year relapse rate. Arch Gen PSy­chlatry.1974;31 :603·608.

33. Clark ML, Huber WK, Hill 0, Wood F, Costiloe JP. Pimozide In chronic schizo­phrenic outpatients. DIs Nerv Syst, 1975;36:137-141,

34. Levine J, Schooler NR, Severe J, Escobar J, Gelenberg Ai, Mandel M, Sovner':,R; :Steliibook R. Discontinuation of oral and depot fluphenazine In schizo-

,,' _::/p~iep)cpa1ientsafter one year of continuous medication. Adv Bioctiem Psy-i<c'>';';;:i:0CChoph'armiicoi. 1980;24:483-495. ..

t.~,..'.·.:,I,I,:;.•,;,',.~,',.'?:35}c~liinO''HK:SChizoplirenics fully remitted on neuroleptics for 3-5 years~to, ;'g'~:~p~4tcontinuedrugs? Br J Psychialry. 1981;138:490-494. .

36F'Vaj1iKii"lninim OP, Agren H, Yao JK, O'Connor OT, Gurklis J, Peters JL. Nor­;;;};~drenergic activity and prediction of psychotic relapse following haioperldoli'ifl~~ffdra~~lJn schizophrenia. Ani J Psychiatry. 1994;51:379-384.:"" '.3Ic)~Q.l<amrj1en OP, Kelley ME, Gurklis JA, Gilbertson MW, Yao JK, Condray R,:·~::)eteiS)LPredieting duration ofel/nieal stability tollowing haloperidol with­i'";!;!ilrawaf,iil schizophrenia. .Neuropsychopharmacology. In press.' ". ,..3~.~fBiirlth~fMH, Branchey LB, Richardson MA. Effects of neuroleptic adjustment

..,J~~i ". ':conditlon in tar~ive dyskinesia in schizophrenic patients. AniJ P,sy-. '_'·..:,w ,,~1)38:608-612,' .. .,','" , .., "., '"

J::MacMllian JF; Johnson AL: JOhnstone EC. Randomized controlled. '... Il1l(fpIiYlaCtiCrieuroleptic'treatmerit. Br J Psychialry. 1986;148:12b~ 127­'40> ellceri;J;Jepp M, Maim U. Do schizophrenics well adapted in the com­

.: ~; J)rt:uHi'tY"ileed niwroleptics? a depot withdrawal stUdy. Acta Psychiatr Seand.':!":·'J980;1l1'(suppl 279):64-76.41 .•Wisledt'lf Adepot neuroleptic withdrawal study: acontrolled stUdy of the with­e •\ifawal afdepot fluphenazine decanoate and depot flupenthixol decanoate in. ·-~·:·~r:··:·:::~~~~·~~;~o.- .' . ..

':t"· •. ". .', ' .. <"::::.~~);:::.:",:~:

-.. , ,: ;..{ ,:" :", ,.:.:.~.: " :

•hrenle patients.. ~:i':;"::}~ ;.,~;: .

'llroleptic 'main~

than dogmatic

lacebo controls.

,erapy in schizo-t'; ';~,i'

ltic agent NEn~1 ,.

er OJ, eds. Psy~

'ork, NY: Raven

I chronically h()s­51:42HZ9... ,.;(patrick B, levinezophrenic outpa-48, .I, Hafez 1'1. Mirza,hrenia: tvi0-year .

d CD, Egan GJ,dosage reductionenic patients; Am

'I roleptlc In acuteand higher halo-

nd Practice. Cam-

usorders: antlpsy­olinoff PB, Rail T,erapeutlcs. 9th e.d.

schizophrenic pa­;;52:173-188.Jroleptlc dose and~. Arch GenPsy-

ipolar disorder:research. Am J

$. 1994;1:377-

libitor and neu­1ia/ Psychiatry. ARCHIVES Ci,-odatiolJ

ence following;en Psychiatry.

)Ole after rapidArch Gen Psy-

ntinuing main­lications. Har-

IE, McEachran,pies in recur-

The ARCHIVES is available by request to nonfederal physicians in the United States (50 states and Washington, DC) whoseofficial American Medical Association masterfile record shows a primary specialty of psychiatry or child psychiatry in anoffice- or hospital-based practice as a staff physician, resident in training beyond the first year, or clinical fellow.

If you meet the above qualification criteria and are not currently receiving the ARCHIVES and would like to receive iteach month, you must complete a free subscription request card. To receive a request card, please write to Kathryn Osten,Aui.erka,n Medical Association, Circulation Processing Department, 515 -N-State-St;Chicago, IL 60610 (fax 312-464-5831).Asubscription request card will be sent to you in response. Ifyou are a resident or fello'w, please include verification of yourtraining program and a complete mailing address.

; N. Effects ofliar disorders.

from chtonlc

;chizoPhrJ~la.lett CJ, R~th­SChizoPhrtn.

withdrawaliof

.....'

..IROi GEN I'SYCIlIATRYNOI. 54, j,\N 19CJ755

--_....--------_._--