Original Articles
`Clozapine for the Treatment-Resistant
`Schizophrenic
`A Double-blind Comparison With Chlorpromazine
`
`John Kane, MD; Gilbert Honigfeld, PhD; Jack Singer, MD; Herbert Meltzer, MD;
`and the Clozaril Collaborative Study Group
`
`® The treatment of schizophrenic patients who fail to re-
`spond to adequate trials of neuroleptics is a major challenge.
`Clozapine, an atypical antipsychotic drug, has long been of
`sclentitic interest, but
`its clinical development has been
`delayed because of an associated risk of agranulocytosis.
`This report describes a multicenter clinical trial to assess
`clozapine’s efficacy in the treatment of patients who are
`refractory to neuroleptics. DSM-Ili schizophrenics who had
`falled to respond to at feast three different neuroleptics
`underwent a prospective, single-blind trial of haloperidoi
`(mean dosage, 61+ 14 mg/d) for. six weeks. Patients whose
`condition remained unimproved were then randomly as-
`signed,
`in a double-blind manner,
`to clozapine (up to
`$00 mg/d) or chlorpromazine (up to 1800 mg/d) for six weeks.
`Two hundred sixty-eight patients were entered In the double-
`blind comparison. When a priori criteria were used, 30% of the
`clozapine-treated patients were categorized as responders
`compared with 4% of chlorpromazine-treated patients. Cloza-
`pine produced significantly greater improvement on the Brief
`Psychiatric Rating Scale, Clinical Globai Impression Scale,
`and Nurses’ Observation Scale for Inpatient Evaluation; this
`Improvement Inciuded ‘‘negative” as well as positive symptom
`areas. Aithough no cases of agranulocytosis occurred during
`this relatively brief study, in our view, the apparenily increased
`comparative risk requires that the use of clozapine be limited
`to selected treatment-resistant patients.
`{Arch Gen Psychiatry 1988;45:789-796)
`
`he efficacy of antipsychotic drugs in short-term and
`maintenance treatment of schizephrenia has been well
`established in numerous double-blind placebo controlled
`trials over the past 30 years..? However, despite the
`considerable magnitude of the medication effect in this
`condition, most controlled trials continueto find a subgroup
`of 10% to 20% of patients who derive little benefit from
`typical neuroleptic drug therapy.’ The treatment of this
`
`Accepted for publication March 9, 1988,
`From the Department of Psychiatry, Hillside Hospital-Long Island
`Jewish Medica] Center, Glen Oaks, NY (Dr Kane); the Department of
`Psychiatry, State University of New York at Stony Brook (Dr Kane); the
`Department of Medical Research, the Sandoz Research Institute, East
`Hanover, NJ (Drs Honigfeld and Singer); the Department of Psychiatry,
`University of Medicine and Dentistry of New Jersey-Robert Wood Johnson
`Medical School, Piscataway (Dr Singer); and the Departmentof Psychiatry,
`Case Western Reserve Schoo! of Medicine, Cleveland (Dr Meltzer). The
`merbers of the Clozaril Collaborative Study Group are as follows: Joyce
`Small, MD, indianapolis; Richard Borison, MD, Augusta, Ga; Rob Conley,
`MD,Pittsburgh; Richard Wagner, MD, Providence, RI; Jan Volavka, MD,
`New York; John Rotrosen, MD, New York; Donald Seidel, MD, San Antonio,
`Tex; Larry Ereshefsky, PharmD, San Antonio, Tex; Jerome Casta, MD,
`Norwalk, Calif; John Herrera, PhD, Norwalk, Calif; Samuel Gershon, MD,
`Detroit; Neil Hartman, MD, Los Angeles; George Simpson, MD, Philadel-
`phia; Richard Abrams, MD, Chicago; Benjamin Graber, MD, Omaha; and
`Martha Martin, MD, Washington, DC.
`Read in part before the 140th Annual Meeting ofthe American Psychiatric
`Association, Chicago, May 14, 1987.
`Reprint requests to Department of Psychiatry, Hillside Hospital, Long
`Island Jewish Medical Center, PO Box 38, Glen Oaks, NY 11004 (Dr Kane).
`
`refractory subgroup remains a major public health prob-
`lem—these individuals require more intensive care and
`are subject to the persistent disabilities associated with
`chronic schizophrenia. In addition, the continued presence
`of psychotic signs and symptoms makesthese patients less
`available to psychosocial and vocational rehabilitation.
`It is estimated that about 1 million Americans suffer
`from schizophrenia. While there are no definitive data
`available on how many do not respond to neuroleptics,
`extrapolations from clinical trial data suggest that there
`may be 100000 to 200000 such patients,
`Data from maintenance medication trials indicate that
`even among patients initially responsive to antipsychotic
`drugs, 20% to 30% may relapse during the first year or
`two of maintenance drug treatment.* A proportion of these
`patients contributes to the number in the subgroup of
`patients refractory to treatment. Since many of these
`patients remain ill, there is a cumulative increase in the
`numberof people in the treatment-refractory category.
`See also p 865.
`
`The recognition that some patients do not benefit from
`typical neuroleptics has resulted in research along two
`fronts: (1) to identify phenomenologic, demographic, and/
`or biologic factors that may be associated with poor
`treatment response and (2) to explore alternative treat-
`ment strategies that might be beneficial to this subgroup.
`With regard to the former,
`there are no consistently
`replicated findings providing clues about why patients are
`refractory to treatment. There are countless reports of
`anecdotal or pilot study experiences with a variety of
`alternative treatments for poor responders. However, no
`particular strategy has been found to be more than occa-
`sionally useful; with controlled studies, the usual result is
`that the experimental treatment proves to be no more
`effective than conventional treatments.
`Since the introduction of chlorpromazine, numerous
`other chemical classes and compoundswith antipsychotic ac-
`tivity have been used. Despite considerable differences in
`chemical structures, these agents seem to share an ability
`to bind to dopamine receptors. Whenin vitro binding assays
`are used, antidopaminergic (specifically, dopamine D, recep-
`tor antagonism) action and therapeutic potency are highly
`correlated. To a greater or lesser degree these are all
`“neuroleptics,” ie, associated with short-term extrapyram-
`idal side effects (including dystonias) and share the longer-
`term liability of inducing tardive dyskinesia. Despite nu-
`merous comparative trials, there are no consistent data
`suggesting that any specific antipsychotic drug or drug
`class is superior to any other in treating schizophrenia.
`Over the past decade, considerable effort has gone into
`the development and testing of potential antipsychotic
`compoundsdesignated atypical. The concept of atypicality,
`however, is a working concept rather than a well-delineated
`
`Arch Gen Psychiatry—Vol 45, Sept 1988
`
`Downloaded From: byJill Strand on 08/22/2018
`
`Clozapine—Kane etal
`
`789
`
`Exhibit 2046
`Slayback v. Sumitomo
`IPR2020-01053
`
`

`

`and validated classification. In general, this term has been
`used to describe drugs that appear to have limited short-
`term extrapyramidal effects in animals or human subjects.
`Most are more selective in their dopamine D, antagonist
`properties (eg, sulpiride or raclopride) and/or more broadly
`active, with marked antiserotonergic, antinoradrenergic,
`or other effects as well (eg, clozapine).
`Clozapine
`(8-chloro-11-(4-methyl-1-piperaziny|)-5H-di-
`benzo[b,e][1,4]diazepine) belongs to the chemical class of
`dibenzodiazepines, related chemically to the antipsychotic
`dibenzoxazepine drug loxapine. However, its pharmaco-
`logie characteristics are different from those of loxapine.
`Clozapine has serotonin (S,), adrenergic (a,), and hista-
`minergic (H,) blocking activity. It is also a potent musca-
`rinie acetylcholine receptor antagonist.*’ Its binding to D,
`and D, receptors is relatively weak and more equivalent
`than that of most typical neuroleptics.* The relationship
`between these characteristics and clozapine’s clinical ef-
`fects remains highly speculative, and a full review of this
`topic is beyond the scope ofthis report.®
`Unlike “typical” neuroleptics, clozapine produces only
`slight, transient elevations in serum prolactin levels in
`patients, even when moderate to high doses are given,”
`Its profile of extrapyramidal side effects appears to be
`very different from those of typical neuroleptics. In both
`US and foreign studies, it has been reported that clozapine
`does not induce dystonia when administered on a short-
`term basis, and although akinesia or akathisia develops in
`somepatients, the incidence appears to be low.”
`Thirteen cases of “dyskinesia” were reported from a
`sample of 12 000 patients in Europe, but the nature of these
`cases is not clear (unpublished results, P. Krupp, MD, and
`C. Monka, Sandoz Ltd, Basel, Switzerland, 1987). There
`has been one report of clozapine apparently exacerbating
`preexisting tardive dyskinesia.“ One case of apparent
`neuroleptic malignant syndrome has been reported in a
`patient receiving clozapine and lithium,
`Previous controlled clinical trials have been conducted
`with clozapine. Claghorn et alreported a six-center
`double-blind comparison of clozapine and chlorpromazine
`in 151 hospitalized schizophrenic patients who had experi-
`enced either extrapyramidal side effects or tardive dyski-
`nesia with at least two different neuroleptics. Clozapine
`was significantly superior to chlorpromazine aecording to
`the major efficacy measures, and it produced fewer side
`effects. The dosage ratio of chlorpromazine to clozapine in
`this study was approximately 2:1. Fischer-Cornelssen and
`Ferner“ conducted a five-center double-blind comparison
`of clozapine and chlorpromazine in 223 hospitalized schiz-
`ophrenic patients; they found clozapine to be superior in
`efficacy, particularly among the more severely ill patients.
`In this study, however, the mean chlorpromazine dose at
`six weeks was only 360 mg compared with 310 mg of
`clozapine.
`In a similar
`two-center European study,'*
`clozapine was compared with haloperidol in a sample of 79
`schizophrenic inpatients. The average dosage of clozapine
`was 397 mg/d at day 40 compared with a dosage of
`7.6 mg/d of haloperidol. Though clozapine was found to be
`more efficacious,
`the latter two comparisons could be
`criticized on the basis ofinadequate dosing of the reference
`drug. The results of these clinical trials suggested that
`clozapine is an effective antipsychotic drug and also pro-
`vided some suggestionsof potential benefit in patients who
`are more severely jll or refractory to treatment.
`However,
`in 1975, granulocytopenia developed in 16
`patients in Finland, and agranulocytosis developed in 13
`of these patients (eight fatalities resulted from secondary
`infection).'"* Worldwide experience now reveals over 100
`
`790
`
`Arch Gen Psychiatry—Vol 45, Sept 1988
`
`Downloaded From: byJill Strand on 08/22/2018
`
`cases of agranulocytosis in patients receiving clozapine.
`Becauseof this, the use of clozapine was curtailed in many
`countries, and the drug was withdrawn for a time from
`clinical research by its US sponsor. For humanitarian
`reasons, some countries (including the United States)
`allowed continued use of the drug for carefully selected
`patients who were resistant to treatment, sensitive to
`extrapyramidal side effects, or dyskinetic; these patients
`underwent intensive precautionary monitoring of white
`blood cell and differential counts. Since the introduction of
`restrictions in use and intensive hematologic monitoring,
`the overall incidence of agranulocytosis has declined, as
`has the lethal risk for patients in whom this reaction
`develops. Overall estimates continue to indicate that the
`risk of agranulocytosis with clozapine exceeds that asso-
`ciated with other antipsychotic drugs. In the United States,
`this problem developed in ten patients of 894 treated, and
`all of these patients recovered without any apparent long-
`term effect. Using the life-table method of calculating risk,
`data from the US experience indicate a 2% cumulative
`incidence after 52 weeks of clozapine treatment
`(95%
`confidence limits, 0.2% and 4%). Based on US and world-
`wide experience, the risk of this adverse effect does not
`appear to be related to age, sex, or dose. The risk of
`“benign” neutropenia, however, does not appear to be any
`higher than with marketed neuroleptics.
`Given clozapine’s apparently greater risk and its promise
`of benefit for patients unresponsive to neuroleptics, the
`decision was madetoinitiate a controlled trial in carefully
`selected treatment-resistant patients. In considering the
`benefit-to-risk ratio of a therapeutic trial of clozapine, the
`time course of the development of agranulocytosis was also
`considered. The majority of agranulocytosis cases world-
`wide have occurred between the sixth and 18th weeks of
`clozapine treatment. Previous data also suggest that six
`weeks would provide a reasonably accurate test of the
`drug's therapeutic potential in individual patients. Expo-
`sure beyond that time was therefore limited in the present
`study to only those patients who had already shown
`significant therapeutic benefit from clozapine.
`METHODS
`Study Design
`
`This study was designed to test the comparative efficacy of
`clozapine in schizophrenic inpatients who by history and prospec-
`tive study would be considered to be resistant to treatment.
`Sixteen participating centers contributed data on a total of 319
`patients, Patients had to meet DSM-11/criteria for schizophre-
`nia. The criteria for being classified as refractory to treatment
`included the following: (1) at least three periods of treatment in
`the preceding five years with neuroleptic agents (from at least
`two different chemical classes) at dosages equivalent to or greater
`than 1000 mg/d of chlorpromazine for a period of six weeks, each
`without significant symptomaticrelief, and (2) no period of good
`functioning within the preceding five years.
`Subjects had to meet the following psychopathologic severity
`criteria: total Brief Psychiatrie Rating Scale (BPRS) score of at
`least 45 (18-item version, in which | indicates absent and 7 indicates
`severe} plus a minimumClinical Global Impressions (CGI) Seale
`rating of 4 (moderately ill). In addition, item scores of at least 4
`(moderate) were required on two of the following four BPRS
`items: conceptual disorganization, suspiciousness, hallucinatory
`behavior, and unusual thought content.
`All patients who met both the historical criteria for treatment
`resistance and the initial severity criteria and gave their informed
`consent entered a prospective period of treatment with haloperidol
`{up to 60 mg/d or higher) and benztropine mesylate (6 mg/d) for a
`period of six weeks to confirm the lack of drug responsiveness.
`improvementin this context was defined a priori as a 20% decrease
`in the BPRS total score phis either a post-treatment CGI Seale
`
`Clozapine—Kaneet al
`
`

`

`rating of mildly ill (=3) or a post-treatment BPRS score of 35 or
`less. Any haloperidol responders(ie, those who met the improve-
`mentcriteria) were dropped from further study.
`Patients who met the multiple psychiatric symptom criteria
`were then randomly assigned to a six-week double-blind treatment
`trial with either clozapine (up to 900 mg/d) or chlorpromazine and
`benztropine mesylate (up to 1800 mg/d of chlorpromazine hydro-
`chloride and up to 6 mg/d of benztropine mesylate). All medications
`were coded and administered under double-blind conditions; in
`addition to coded active antipsychotic medication in blue capsules,
`patients received either white benztropine tablets (chlorpromazine
`group) or identical white placebo tablets (clozapine group). The
`use of prophylactic benztropine mesylate (up to 6 mg/d) for all
`patients receiving chlorpromazine was designed to enhance the
`double-blind condition,
`in light of clozapine’s previously estab-
`lished profile of reduced extrapyramidal side effects. In addition,
`this strategy was thought to minimize the potential for behavior-
`ally manifest adverse effects to confound assessmentof the relative
`clinical efficacy of the two drugs.
`Before the start of the study, a priori criteria for supporting the
`superiority of clozapine in this patient population were deter-
`mined. These criteria required proof of statistical superiority in
`all of three predetermined areas: the CGI Scale, changes in BPRS
`total score, and significant improvement in at least two of the
`following four BPRS items (or the cluster score derived from
`summing these four items): conceptual disorganization, halluci-
`natory behavior, suspiciousness, and unusual thought content.
`Treatment
`
`Patients entering the double-blind phase of the study were
`treated for six weeks. During the first two weeks, the dosage was
`titrated upward,
`if well tolerated,
`to dosage levels of either
`500 mg/d of clozapine or 1000 mg/d of chlorpromazine (plus
`6 mg/d of benztropine mesylate for chlorpromazine patients only).
`Dosing during the final four weeks was flexible,
`to maximum
`allowable dosages of 900 mg/d of clozapine and 1800 mg/d of
`chlorpromazine (plus up to 6 mg/d of benztropine mesylate). The
`numberof patients entering each study period was as follows:
`No. of
`Patients
`319
`305
`272
`268
`
`Period No.
`1
`2
`3
`4
`
`Description
`Baseline placebo
`Haloperidol
`Placebo washout
`Double-blind
`
`Duration, d
`=14
`=42
`<7
`=42
`
`Of the patients who entered period 4, 126 were randomized to
`clozapine, and 142 were randomized to chlorpromazine and benz-
`tropine mesylate.
`
`Evaluationof Efficacy
`
`Patients were interviewed by physicians or psychologists weekly
`during the course ofdouble-blind treatment, and their assessments
`were recorded on the BPRS and on a seven-point CGI Scale (in
`which 1 indicates no mental illness and 7 indicates severe mental
`illness). In addition, patients were regularly evaluated in terms of
`ward behavior by the nursing staff, using the 30-item Nurses’
`Observation Scale for Inpatient Evaluation (NOSTE-30).2
`Evaluation of Safety
`
`Adverse reactions were evaluated by systematic patient query
`and observation by both medical and nursing personnel. Reactions
`were graded for severity and evaluated as to attribution to study
`drug, and the course of the reaction was documented. Regular
`clinical laboratory tests were performed, as were physical exami-
`nations, an electrocardiogram, and vital sign determinations.
`Systematic assessments of extrapyramidal symptoms and abnor-
`mal involuntary movements were made weekly using the Simpson-
`Angus Seale for Extrapyramidal Side Effects® and the Abnormal
`Involuntary Movements Scale (AIMS).*
`SUBJECTS
`
`Three hundred nineteen inpatients entered this study; their
`demographic and treatment history characteristics are summa-
`rized in Tables 1 and 2. Only 20% of the patients were female,
`largely due to the high proportion of Veterans Administration
`
`Arch Gen Psychiatry—Vol 45, Sept 1988
`
`Downloaded From: byJill Strand on 08/22/2018
`
`Table 1.— Sex, Race, and Diagnosis
`of Patients Entering the Study (N= 319)
`
`of Patients
`
`No.(%)
`
`Characteristic
`Sex
`
`
`M
`.
`256 (80)
`
`
`F
`63 (20)
`
`
`Race
`White
`
`
`Black
`
`
`Hispanic
`
`
`Oriental
`Other
`
`
`Diagnosis (OSM-Hi schizophrenic subtypes)
`
`
`160 (50)
`Undilferentiated
`107 (34)
`Paranoid
`
`
`25 (8)
`Disorganized
`
`
`11 (3)
`Residual
`Unspecified
`10 {3)
`
`
`Catatonic
`6 (2)
`
`
`
`208 (65)
`74 (23)
`31 (10)
`2 (1)
`4 (1)
`
`medical centers among the participating institutions and possibly
`also because women were less likely to have received 1000-mg
`chlorpromazine equivalents of three different neurolepties.
`The typical patient was a 35-year-old male chronic undiffer-
`entiated schizophrenic first hospitalized for psychosis at age 20
`years, after which seven or eight additional periods of hospitali-
`zation ensued. The median duration of the current hospitalization
`was about two years.
`
`RESULTS
`
`Over 80% of the patients completed the six-week prospective
`haloperidol phase of the study. A complete tabulation of patient
`outcomesafter haloperidol treatment is provided in Table 3.
`Of those patients who completedthefull six weeks of haloperidol
`treatment (dosages up to 60 mg/d and greater; mean {SD}, 61
`{14] mg/d), 80% were nonresponders. Fewer than 2% were classified
`as haloperidol responders. In the balance of the patients, haloper-
`idol was terminated early for a variety of reasons,
`the most
`prominent of which was intolerance to haloperidol. On average,
`haloperidol-treated patients showed no change during the course
`of six weeks of treatmentin any areas of the BPRS or NOSIE-30.
`Twenty-two patients were unable to tolerate the complete halo-
`peridol phase due to adverse effects, but since they met all
`retrospective criteria for treatment resistance, they were allowed
`to continue into the double-blind comparison. (Thirteen of these
`patients received chlorpromazine, and nine received clozapine.
`Efficacy analyses excluding these patients were also carried out
`and did not alter the results.)
`Two hundred sixty-eight patients entered the critical clozapine
`vs chlorpromazine and benztropine double-blind phase. The diag-
`nostic composition of each treatment subgroupin the double-blind
`phase was similar to that seen initially: approximately half of the
`patients in each treatment group were in the “undifferentiated”
`category and about one third were in the “paranoid” category.
`From the point of view of psychiatric history, the subgroups did
`not differ in any significant way in major characteristics of patient
`history and treatment, including age at first hospitalization for
`psychosis, numberof hospitalizations, durationofillness, duration
`of current episode, and duration of present hospitalization.
`Average daily doses of active antipsychotic medication received
`during double-blind treatment are shown by treatment week in
`Fig 1. Adequate dose levels of each drug were attained with
`mean peak dosages exceeding 1200 mg/d of chlorpromazine and
`600 mg/d of clozapine. The decrease in average dosage for both
`treatment groups at week 6 reflects the mandated taper-down at
`the end of the treatment period for all patients, designed to avoid
`abrupt discontinuation.
`Review of dispositions at the end of each patient’s double-blind
`participation indicated high overall completion rates for both
`clozapine- and chlorpromazine-treated patients (88% and 87%,
`respectively). Early terminations occurred for the following rea-
`sons: adverse reactions (6%), illness not related to drugs (1%),
`
`Clozapine—Kane etal
`
`791
`
`

`

`Table 2.—General Characteristics of Patients Entering the Study (N= 319)
`
`Characteristic
`
`Pationter
`318
`
`35.0
`
`Mean (SD)
`35.7 (8.87)
`
`Range
`20-59
`
`
`
`
`Age, y
`
`Duration of current
`
`212.0
`314.7 (316.76)
`5-1976
`symptoms, wk
`307
`
`
`
`
`20.4 (4.61)
`Ageatfirst hospitalization, y
`
`9.2 (7.26)
`No. of hospitalizations
`
`
`Ouration of current
`
`hospitalization, wk
`
` 215.9 (321.41)
`
`*The numberof patients varies because of “missing” or “unknown" data elements.
`
`ms&88
`
`g
`
`\
`/
`
`/
`
`7
`
`“
`
`Z B00
`>
`a 600 “
`
`25
`
`&=
`
`4004
`i
`
`2004toa
`
`4
`
`5
`
`6
`
`0
`
`1
`
`2
`
`3
`Weekof Study
`Fig 1.—Meandaily doses of clozapine (solid tine) and chlorprom-
`azine (broken line) during double-blind phase of study (period 4).
`For clozapine, at week 1, n= 126; week 2, n= 126; week 3, n= 122:
`week 4, n=120; week 5, n=119; and week 6, n=116. For
`chlorpromazine, at week 1, n=141; week 2, n= 140; week 3,
`n= 137; week 4, n= 133; week 5, n= 128; and week 6, n=125.
`
`changes, withdrew consent)
`
`Table 3.— Patient Classification After Treatment With
`Haloperidoi and Benztropine
`
`No.(%) of
`Patients (n= 305)
`
`Patient Classification
`Haloperido! responder
`Haloperidol nonresponder
`Terminated early
`Intolerant of haloperidol
`Uncooperative
`Protocol violated
`Physical conditions not related to drug
`Other (6g, seizure,
`eiectrocardiographic
`
`uncooperativeness (2.9%), protocol violations (1%), symptom ex-
`acerbation (1%), and other causes (1%), Ratesof early termination
`for all reasons were comparable for patients in both treatment
`groups.
`
`Clinical Etficacy
`
`Analyses of covariance of posttreatment change scores con-
`ducted for week 6 vs baseline (using pretreatment scores as
`covariates) were performed forall efficacy variables. An “intent
`to treat” analysis™ was carried out for all patients who had a
`baseline assessment and at least one assessment following ran-
`domization, with the last observation carried forward, yielding
`essentially equal numbersof patients in each cell.
`Figures 2 and 3 display findings for two of the predetermined
`critical variables, the two overall indexes of improvement: BPRS
`total score and the CGI Seale. The improvementin both the BPRS
`total score and the CGI Scale was approximately three times
`greater in the clozapine-treated patients. Differences favoring
`clozapine were statistically significant by the first week of treat-
`ment and continuedto be present each week over the entire course
`ofstudy. Similarly, four “positive” BPRS items determineda priori
`to be central to the assessmentof therapeutic response (conceptual
`disorganization, hallucinatory behavior, suspiciousness, and unu-
`sual thought content) ali demonstrated significant differences
`favoring clozapine over chlorpromazine and benztropine. These
`items were combined into a cluster score, which also yielded
`significant differences favoring clozapine (Fig 4). The mean scores
`at baseline and end point are presented in Table 4. Clozapine was
`superior to chlorpromazine in the treatment of negative signs and
`symptoms as well, as evidenced by statistically significant differ-
`ences on the BPRSitems of emotional withdrawal, blunted affect,
`psychomotor retardation, and disorientation. These items in
`eombination form the BPRS “anergia” factor, displayed in Fig 5.
`Analysis of variance and analysis of covariance results for all
`BPRSvariables, including the a priori criteria, are shown in Table
`5. Therapeutic response was assessed by the nursing staff as well,
`whorated patients’ ward behavior on the NOSIE-30 (Table 5). For
`all six factors (social competence, social interest, personal neat-
`ness, irritability, manifest psychosis, and retardation), the nursing
`staff, blind to treatment assignment, judged clozapine effects
`superior to those of chlorpromazine and benztropine. Weekly
`
`20
`
`~ a
`
`~ o
`
`o
`
`
`Total
`
`ScoreonBPRS,MeanChangeFromBaseline
`
`
`aa
`
`
`0
`
`1
`
`2
`
`4
`
`5
`
`6
`
`3
`Week of Study
`Fig 2.—Meanchange from baseline in total score on Brief Psychi-
`atric Rating Scale (BPRS) for patients treated with clozapine (solid
`line, n= 126) or chlorpromazine and benztropine mesylate (broken
`line, n= 139). P<.001 during each weekof study.
`
`changes on the composite score, “total patient assets,” are pre-
`sented in Fig 6.
`Concerning the onset of therapeutic effects, Figs 2 to 6 indicated
`significant differences favoring clozapine over chlorpromazine as
`early as the first week. Analysis of variance of the comparative
`rates of improvement for these treatment groups (analysis of
`slopes) found that clozapine produced morerapid onset of activity
`
`782
`
`Arch Gen Psychiatry—Vol 45, Sept 1988
`
`Clozapine—Kaneetal
`
`Downloaded From: byJill Strand on 08/22/2018
`
`

`

`Chlorpromazine
`
`Table 4.—Comparative Drug Efficacy and Neurologic Ratings
`
`Score
`Two-Tailed
`(Mean+SD)
`No. of aN Analysis of
`Patientst
`Baseline
`End Point
`Covariance, P
`61212
`45+13
`61+11
`56+12
`1924
`14+5
`19+4
`17+4
`
`BPRSiota! score
`
`BPRSciusterof four key items
`
`CGI Scale
`
`AIMStotal score
`
`Simpson-Angus Scaie for Extrapyramidal Side Effects
`
`Clozapine
`Chlorpromazine
`Clozapine
`Chlorpromazine
`Clozapine
`Chlorpromazine
`Clozapine
`Chlorpromazine
`Clozapine
`
`*BPRSindicates Brief Psychiatric Rating Scale; CGI, Clinical Global impressions; and AIMS, AbnormaiInvoluntary Movements Scale.
`tThree patients were exciuded from these analyses. Onepatient did not undergo rating after randomization, and one study site had only two patients, both
`of whom received chlorpromazine.
`
`Drug(s)
`Proved
`Effectivet
`
`Drug Proved
`Superior/Pt
`
`Week of
`Onset of Superior
`Drug Activityt
`
`Drug
`Proved
`Faster§
`
`Criterion
`Varlable*
`BPRSpositive symptoms
`Conceptuai disorganization
`Mannerisms/posturing
`Hostility
`Suspiciousness
`Hallucinatory behavior
`Excitement
`Unusual thought
`Grandiosity
`BPRSnegative symptoms
`Emotional withdrawal
`Uncooperativeness
`Biunted affect
`Disorientation
`Motorretardation
`BPRS general symptoms
`Somatic concern
`Anxiety
`Guilt
`Tension
`Depressed mood
`
`Clozapine and chlorpromazine
`Clozapine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`Clozapine and chiorpromazine
`Clozapine and chlorpromazine
`Clozapine
`
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<,001
`Clozapine/<.001
`
`Clozapine
`Clozapine
`Clozapine
`Clozapine
`
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<.001
`Clozapine/<.001
`
`wee
`Clozapine/<.05
`
`Clozapine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`
`Clozapine/<.01
`
`
`a
`Clozapine/<.001
`a
`
`1
`1
`1
`2
`2
`3
`1
`
`2
`1
`3
`2
`6
`
`6
`
`:
`1
`cae
`
`1
`
`Clozapine
`Clozapine
`Clozapine
`
`—
`Clozapine
`
`Ctozapine
`Cfozapine
`Clozapine
`Clozapine
`
`Clozapine
`
`ces
`
`. Clozapine
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`BPRStotal score
`
`Clozapine and chlorpromazine
`
`Clozapine/<.001
`
`Clozapine and chlorpromazine
`Clozapine
`Clozapine
`Clozapine and chlorpromazine
`Clozapine and chlorpromazine
`
`NOSIE-30 factors
`2
`Clozapine/-<.001
`Social competence
`1
`Clozapine/-<.001
`Social interest
`2
`Clozapine/<.001
`Personal neatness
`2
`Clozapine/<.01
`Irritability
`eee
`2
`Clozapine/<.001
`Manifest psychosis
`Clozapine
`2
`Clozapine/<.05
`Motor retardation
`Clozapine
`2
`Clozapine/<.004
`Clozapine and chlorpromazine
`NOSIEtotal assets
`“BPRSindicates Brief Psychiatric Rating Scale; CGI, Clinical Global mpression; and NOSIE-30; 30-item Nurses’ Observation Scale for Inpatient Evaluation.
`tSignificant pre-post change by within-group ¢ tests.
`tSignificant pre-post change by between-group analysis of covariance.
`§Analysis of variance of rates of improvement.
`
`Clozapine
`Clozapine
`Clozapine
`
`
`
`
`
`
`
`
`
`
`
`Arch Gen Psychiatry—Vo! 45, Sept 1988
`
`Downloaded From: byJill Strand on 08/22/2018
`
`Clozapine—Kane etal
`
`793
`
`

`

`14
`
`a
`
`
`
`CGIScale
`
`Score,MeanChangeFromBaseline
`
`1
`
`Week of Study
`Fig 3.—Mean change from baseline in score on Clinical Global
`Impressions (CGI) Scale for patients treated with clozapine (solid
`line, n= 126) or chlorpromazine and benztropine mesylate (broken
`line, n= 139) For week 1, P.003; weeks 2 through 6,
`P<,001.
`
`°to 3
`_
`
`ny
`
`
`BPRSAnergia
`
`Score,MeanChangeFromBaseline
`
`
`
`
`
`nmw>
`
`Week of Study
`Fig 4.—Mean change from baseline in cluster score on four key
`items from Brief Psychiatric Rating Scale (BPRS) for patients
`treated with clozapine (solid line, n= 126) or chlorpromazine and
`benztropine mesylate (broken line, n= 139). For week 1, P= .011;
`week 2, P= .001; weeks 3 through 6, P<.001.
`
`
`
`10Be2
`
`ClusterScoreonFourKeyItemsFromBPRS,
`
`MeanChangeFromBaseline
` 12
`
`0
`
`1
`
`2
`
`on
`gD 4
`asoOFE
`we *
`Qs
`S=o
`j
`-2
`
`4
`
`5S
`
`6
`
`3
`Week of Study
`Fig 6.—Mean change from baselinein score on total patient assets
`item from Nurses’ Observation Scale for Inpatient Evaluation
`(NOSIE)for patients treated with clozapine (solid line, n= 126) or
`chlorpromazine and benztropine mesylate (broken line, n= 139).
`For week 1, P = .356, weeks 2 through 6, P<.001.
`
`Weekof Study
`Fig 5.—Mean change from baseline in score on anergia item from
`Brief Psychiatric Rating Scale (BPRS) for patients treated with
`clozapine (solid line, n= 125) or chlorpromazine and benztropine
`mesylate (broken line, n= 139). For week 1, P<.544; week 2,
`P = .002; weeks 3 through 6, P<.001.
`
`in 16 of 27 tests performed; this was never true for chlorpromazine
`(Table 5).
`To test for differential effects among centers, mean improvement
`seores {total BPRS) bytreatment group were individually arrayed
`for each of the 16 centers. The data were homogeneous:in 14 of 16
`centers, greater improvement was found for clozapine-treated
`patients.
`The interpretations allowed by the parametric data are limited
`by the fact that clinically unimportant changes in rating-scale
`scores can be statistically significant if a large enough sample of
`patients is studied. The critical test from a clinical perspective is
`the extent to which a treatment producesa clinically meaningful
`response; ie, is the patient believed to have truly benefited from
`the medication? This issue underscores the importance of the a
`priori criteria for clinical improvement that provide the critical
`outcome measuresin this investigation.
`Patients were classified as having “improved” to a clinically
`significant extent or not over the course of double-blind treatment.
`The a priori criteria for defining a patient as improved included a
`reduction greater than 20% from baseline in the BPRStotal score
`plus either a posttreatment CGI Scale score of 3 (mild) or less or
`a posttreatment BPRStotal score of 35 or lower. When these
`eriteria were applied to all patients who completed