2004 Psychiatric Times. All rights reserved.

Atypical Antipsychotics Assessed at ACNP

by Kenneth J. Bender, Pharm.D., M.A.

Psychiatric Times February 1997 Vol. XIV Issue 2

(This article is based on selected presentations at the 35th Annual Meeting of

the American College of Neuropsychopharma-cology [ACNP], San Juan, P.R.,
Dec. 9 to 13, 1996.)

Do people like the term atypical? This question was asked by David Pickar,
M.D., chief of the National Institute of Mental Health's Experimental
Therapeutics Branch, who chaired the symposium "Tracking the Next
Generation of Antipsychotic Drugs."

"It's a term that has outlived its usefulness," rejoined the symposium
moderator, John Kane, M.D., department of psychiatry, Hillside Hospital, Glen
Oaks, N.Y. "What about 'novel?'" quipped Nina Schooler, Ph.D., professor of
psychiatry and director of the Psychosis Research Program, University of
Pittsburgh, amid discussion that quickly achieved consensus that new
antipsychotic agents which offer therapeutic advantages over older
neuroleptics will, themselves, soon become typical.

Kane then put semantics and categorizations aside, and challenged the
psychopharmacologists to conduct investigations of the new compounds that
will assist clinicians in choosing an optimal antipsychotic for a particular
patient at a particular stage in that patient's illness.

"I think we're going to be increasingly hard-pressed to develop a data base

that's going to assist us in answering those questions," Kane said. As the utility
of the "typical" neuroleptics is now being questioned, Kane indicated, so
should the typical clinical research paradigms be reconsidered.

One limitation of the current antipsychotic research data base, Kane recounted,
is that subject populations have invariably comprised patients with chronic,
entrenched illness.

"The patient population that we're studying in most of the antipsychotic drug
development programs is really one segment of the population," said Kane,
who is professor of psychiatry at Albert Einstein College of Medicine. "We
have to make some decisions and choices based on the data that are generated
by those particular trials."

Kane then recommended investigating the therapeutic potential of atypical

antipsychotic compounds in patients experiencing their first onset of illness,
before they are exposed to typical neuroleptics. "That has implications," Kane
said, "not only in looking at the broad and immediate clinical effects of these
drugs, but also setting the stage for actually attempting to study the impact of
antipsychotic drugs on the evolution of the illness itself." He elaborated, "The
evolution of negative symptoms, deficit state and a variety of other issues
really can't be studied in patients given putative atypical compounds when
they've already been treated with conventional drugs for 10 or 12 years."

Kane also asserted the continued need for placebo-control arms in

antipsychotic clinical trials, while acknowledging the growing concern about
the use of placebo in investigating medical conditions for which there is
proven, effective treatment (Rothman and Michels). Placebo control is a
valuable component of antipsychotic efficacy research, Kane argued, because
of the variable, often marginal response associated with schizophrenia.

"Without having these comparative groups," he said, "it's difficult to know the
responsiveness of the population you're studying, and we can draw erroneous
conclusions about a new compound."

The measures of drug response most frequently employed in antipsychotic

trials may also be inadequate, Kane suggested, to discern the scope of
therapeutic benefit or track fundamental aspects of a patient's progress.
Efficacy measures should extend beyond quantifying symptom reduction to
include functional outcomes, Kane recommended.

"We're ultimately interested in where patients end up," he said. "What patients
improve slightly or not at all? Were differences found between the new drug
and comparators clinically as well as statistically significant?"

Other areas that Kane hopes can be elucidated in future clinical trials include:
whether one antipsychotic offers sufficient therapeutic advantage to justify
switching from another agent; the extent to which such adverse effects as
akinesia and akathisia confound measures of clinical response; the optimal
doses of study drugs and comparators from investigations with multiple dosage
arms of each; and whether maintenance antipsychotic regimens can be
evaluated for long-term efficacy, safety and compliance.

Kane also questioned the use of typical neuroleptic comparators to repeatedly

demonstrate the efficacy of new, atypical antipsychotics. After initial studies
have shown the therapeutic advantages of a new compound, Kane suggested,
subsequent direct comparisons between atypical compounds could yield data
that is more helpful to clinicians.

To underscore the extent to which new antipsychotics have demonstrated

advantage over older neuroleptics, John M. Davis, M.D., Psychiatric Institute,
University of Illinois, Chicago, offered the results of his recent meta-analysis
of the olanzapine (Zyprexa)-haloperidol (Haldol) and the risperidone
(Risperdal)-haloperidol comparative trials (Davis and others). He included all
such studies published or presented prior to this ACNP meeting. He based the
drug effect size comparison on the reported Positive and Negative Syndrome
Scale (PANSS) and/or Brief Psychiatric Rating Scale (BPRS) score reductions
associated with haloperidol or the newer antipsychotics, calculated as: mean
reduction (new drug) less mean reduction (haloperidol comparator) divided by
pool standard deviation.

For the olanzapine studies, Davis related finding the difference in effect size
favoring olanzapine to be highly statistically significant.

"It's established," said Davis, "that olanzapine is better than haloperidol."

Davis found the data for risperidone more complicated to evaluate because of
the large number of (and occasionally heterogeneous) studies. "But taken in
toto," Davis concluded, "I feel risperidone is significantly and quite
substantially better than standard neuroleptics."

Where there is such evident advantage over typical neuroleptics, Schooler

concurred with Kane that further distinctions between the atypical
antipsychotics and their relative advantages in particular subsets of patients can
be best ascertained through direct comparison. She suggested that an NIMH
collaborative effort could facilitate these comparisons; others indicated that
such studies are already underway, including a comparison of olanzapine and
risperidone scheduled for presentation later in the ACNP proceedings. The
interim report of this olanzapine-risperidone comparison is described below, as
are the studies presented during the ACNP meeting which addressed the issues
from this symposium of earlier treatment of schizophrenia, the switching of
antipsychotic agents to improve response and broadening efficacy measures to
include improvement of deficit states and function.

Directly Comparing Atypicals

Pierre Tran, M.D., clinical research physician with Lilly Research

Laboratories, presented interim results from Lilly's comparison of olanzapine
and risperidone in the treatment of patients with schizophrenia, schizophren-
iform disorder and schizoaffective disorder (Tran and others). Tran indicated
that his group intended not only to distinguish between the two drugs in
measures of efficacy and safety, but to judge the extent to which each met the
criteria for an "atypical" antipsychotic.

"With the introduction of these atypical agents," Tran said, "we have set a new
level of expectations for these [antipsychotic] compounds."

The three criteria set forth by the investigators to judge atypicality relative to
the effects of typical neuroleptics were: fewer extrapyramidal symptoms
(EPS); broader efficacy, especially in treating negative symptoms; and less
effect on the dopamine-sensitive hormone prolactin. Tran indicated that the
development of olanzapine and risperidone to achieve this atypical
antipsychotic profile has followed different paths, and resulted in the two
compounds having distinctly different chemical structures, receptor-binding
affinities, animal neuropharmacology, pharmacokinetics and clinical risk-
benefit profiles.

This first large-scale head-to-head comparison of two putative atypical

antipsychotics was conducted in randomized, double-blind parallel manner by
41 investigators at centers in nine countries. The study was statistically
powered to enable an interim comparative analysis of safety and efficacy after
at least 250 patients completed a washout period and began receiving either
study medication. At the time of this preliminary report, 297 had been
randomly assigned to receive either olanzapine or risperidone in a 1:1 ratio for
eight weeks of acute therapy and 20 weeks of maintenance. Two hundred and
twenty-six patients completed the eight-week acute phase, with no statistically
significant differences between the treatment groups in completion of this
phase. Of the 77 that completed the entire 28-week study period, statistically
significantly more olanzapine-treated patients completed than did risperidone-
treated patients.

The severity of the patients' psychotic disorders at their entry into the study
was quantified as a minimum BPRS score of 24 (BPRS extracted from the
PANSS, items 0 to 6). Approximately two-thirds of the study population was
male; the average age of the entire study population was 36 years; and their
illness onset had occurred at an average age of 23 years.

The starting dose of olanzapine was 15 mg once daily for the first week; and 1
mg risperidone twice daily, increased by 1 mg per dose daily to 3 mg twice
daily for days 3 through 7, in accordance with the approved labeling. Each
drug was thereafter titrated for optimal effect within the respective daily dose
ranges of 5 mg to 20 mg olanzapine and 4 mg to 12 mg risperidone. The
resultant daily mean modal dose of olanzapine was 16.9 mg at eight weeks and
17.1 mg by 28 weeks; and the daily mean modal dose of risperidone was 7.3
mg at both times. Critical discussion at the study presentation included an
observation that the risperidone dosage required by this study population was
higher than the average risperidone dose that marketing data indicate is used in
the United States.

At the doses employed, olanzapine best met the investigators' first criteria for
atypicality by having less EPS liability than risperidone; as monitored with the
Simpson-Angus Scale (SAS) for parkinsonism, the Barnes Akathisia Scale
(BAS), the Abnormal Involuntary Movement Scale (AIMS) and in the
frequency of use of anti-EPS medications. A statistically significantly smaller
proportion of patients receiving olanzapine (17.9%) requested one or more
doses of anticholinergic medication during the eight-week acute treatment
phase than the 31.5% of patients receiving risperidone who required anti-EPS
medication during that period.

In addition, statistically fewer olanzapine-treated patients experienced elevated

prolactin levels after both eight weeks and 28 weeks than did those receiving
risperidone. The question of whether this difference, albeit statistically
significant, is clinically significant was among those arising in the discussion
of this study report. That question, along with unknown long-term effects of
relative levels of prolactinemia awaits further study.

In efficacy measures at week 8, olanzapine and risperidone were comparable in

reducing overall positive and negative symptoms of schizophrenia, although
olanzapine achieved numerically superior mean changes on six of eight
measures, and statistically significant superiority to risperidone in the PANSS
mood subscore at both eight and 28 weeks.

In addition, at 28 weeks, a statistically significant advantage with olanzapine

was evident among patients who had achieved at least a 30% improvement on
the PANSS total score. Fewer patients receiving olanzapine experienced
relapse (4% at three months and 9% at six months) than did those treated with
risperidone (9% at three months and 29% at six months). Relapse, a key
measure of maintenance efficacy, was defined as a 20% or greater worsening
in PANSS total score along with a Clinical Global Impression-Severity of
Illness (CGI-S) score equal to or greater than 3.

In seeking a broader measure of antipsychotic efficacy than reduction of overt

symptoms, the investigators also applied the Heinrichs-Carpenter Quality of
Life Scale (QLS) at baseline and week 28. The effect of the agents appeared
comparable by this measure, although the patients receiving olanzapine
demonstrated significantly better scores in the subscale for maintaining
positive interpersonal relations.

Tran summarized from this 28-week trial that olanzapine evidenced the key
criteria for atypical antipsychotic effects more often and to a greater degree
than risperidone. Olanzapine was associated with fewer side effects, less
elevation of prolactin and greater efficacy in mood symptoms. At the dosages
employed, he concluded, olanzapine was statistically significantly more likely
than risperidone to evoke greater initial therapeutic response and less likely to
be associated with relapse during maintenance therapy.

Risperidone and Haloperidol

Risperidone, in turn, was reported by other investigators at the ACNP to have

demonstrated favorable advantage over the typical neuroleptic comparator,
haloperidol. In an analysis of combined data from two eight-week, double-
blind, placebo-controlled trials of risperidone conducted in phase III, prior to
market approval, Stephen Marder, M.D., department of psychiatry at West Los
Angeles Veterans Affairs Medical Center, reported that patients treated with
risperidone in daily doses ranging from 6 mg to 16 mg evidenced significantly
greater response than those who had received 20 mg daily haloperidol (Marder
1996a).

The investigators' factor analysis of data from a total of 523 hospitalized

patients diagnosed with chronic schizophrenia revealed an advantage with
risperidone that was greatest for treatment of negative symptoms, uncontrolled
hostility/excitement and anxiety/depression. In addition, there was a
statistically significant advantage for risperidone in the treatment of positive
symptoms and disorganized thought. Haloperidol was not, however, found to
differ significantly from placebo in the treatment of either negative symptoms
or anxiety/depression. The specific PANSS items that improved most
significantly with risperidone in relation to the response with haloperidol were
passive social withdrawal, active social avoidance, depression, emotional
withdrawal and hostility. Since these are not the factors usually associated with
the psychotic component of schizophrenia, the investigators suggest that
risperidone showed a qualitatively different therapeutic effect than haloperidol.

Marder indicated that he finds relatively small differences between the new
antipsychotics and typical neuroleptics in treating positive symptoms of
schizophrenia, and posits that their therapeutic advantages arise from other
properties.

"I think, now, people are beginning to look in much more refined ways at the
activities of the drugs," Marder commented. "In the past, clinicians focused on
positive symptoms. Now there is a greater focus on negative symptoms,
cognitive symptoms and mood."

Treating the First Episode

The possibility mentioned by Kane that an atypical antipsychotic, with greater

efficacy and lower side-effect liability than typical neuroleptics, might exert
even greater benefit when used early in the course of schizophrenia illness was
examined further by Jeffrey Lieberman, M.D., professor and vice chairman of
psychiatry at the University of North Carolina, Chapel Hill, as chair of another
ACNP symposium, "Atypical Antipsychotic Drug Treatment of First-Episode
and Recent-Onset Schizophrenia."

Lieberman described the pattern and course of schizophrenia and the rationale
for early intervention with atypical anti-psychotics. Lieberman presented data
from a pilot study of risperidone and clozapine (Clozaril) in first-episode
schizophrenia, which he conducted with colleagues from Hillside Hospital
Research Department (poster presentation by Delbert Robinson, M.D., and
others).

Lieberman explained that most patients with schizophrenia experience their

acute onset by early adulthood; and many, after initial improvement, then
follow an undulating, deteriorating course with increased, seemingly
accumulating morbidity of both persisting symptoms and declining function.

Lieberman pointed out that the deterioration process invariably commences in

the early phase of illness, often prior to psychotic symptoms and within five
years of the first episode.

The treatment response during the first episode has been found to be distinct
from that observed later in the course of illness, Lieberman explained, with
patients in this early stage often having both a better therapeutic response to an
antipsychotic and more sensitivity to its side effects (Lieberman 1996;
McEvoy and others). The apparently greater therapeutic response at this stage
of illness has then been associated with improved levels of recovery, as well as
better long-term functioning (Wyatt). The atypical antipsy-chotics may have
particular utility at this early stage, Lieberman suggested, not only for their
increased efficacy, but because they are more likely to be tolerated in
therapeutic doses than neuroleptics with greater EPS liability.
Lieberman posits an association between the duration of psychosis and poorer
treatment response and outcome. He suggested that the pathophysiologic
process underlying the psychiatric illness, when left unabated by incomplete
antipsychotic effect or delayed treatment, progresses to cause the persistent
morbidity of schizophrenia. The first episode of the illness is then, he
proposed, the most opportune time for therapeutic intervention.

Lieberman suggested that the question of "whether there is any added value"
from employing atypical antipsychotics earlier in illness onset should be
assessed with measures of antipsychotic efficacy beyond drug effects on
positive and negative symptoms. He urged assessment of the possible
antipsychotic benefit for cognitive deficits, affective symptoms, functional
impairment, as well as their effectiveness in preventing relapse and subsequent
deterioration.

In their pilot study of atypical agents in first-episode, 16 patients (eight men

and eight women, mean age 25.8 years) were treated with risperidone; and 30
patients (17 men and 13 women, mean age 22.6 years) were treated with
clozapine. Inclusion criteria included a lifetime treatment history of previous
anti-psychotic medication treatment of 12 weeks or less, a diagnosis of
schizophrenia or schizoaffective disorder and current severity of psychotic
symptoms to achieve a rating of 4 or more on at least one psychotic item on the
Schedule for Affective Disorders and Schizophrenia-Change Version with
psychotic and disorganization items (SADS-C+PD). Initially in the pilot study,
patients were offered risperidone, and as the study progressed, patients were
then offered clozapine first, and risperidone if clozapine was refused. Robinson
related that there was wide variation from the clozapine mean daily dose of
230 mg, with some patients requiring less than 100 mg per day as a result of
these first-episode patients being particularly sensitive to its side effects.

"For clozapine [these reactions include] sedation and orthostatic hypotension.

The end result is that many people [first episode] responded at dosages that are
much lower than what we're used to with clozapine for treatment-refractory
patients," Robinson said. "So you have to start at a very low dose and
gradually build up."

Robinson reported that the onset of response to either risperidone or clozapine

was similar to that of fluphenazine (Prolixin), as was the degree of
improvement in such positive symptoms as hallucinations and delusions.
Fifteen of the 22 patients treated with clozapine (68%) and nine of the 13
patients receiving risperidone (69%) for more than six weeks met the response
criteria of a rating of mild or less on the four psychotic symptom items of the
SADS-C+PD, and a much or very much improved rating on the CGI. Robinson
reported that clozapine-treated patients improved more on depression and
negative symptoms than did patients treated with risperidone or fluphenazine.

Patients receiving either risperidone or clozapine experienced, as expected, less

EPS than those who had received fluphenazine. Robinson indicated that a more
rigorous study design with parallel comparison of first-episode patients
receiving clozapine or a typical neuroleptic should be conducted to confirm the
apparent advantage of clozapine.

Olanzapine and Haloperidol

Lieberman had also collaborated with Lilly Research Laboratories scientists

Todd Sanger, Ph.D., and Gary Tollefson, M.D., Ph.D., in extracting and
examining the data for olanzapine and haloperidol in treatment of first-episode
psychosis from the larger data base generated in the international, multicenter
double-blind parallel trial in 1,996 patients with schizophrenia, schizophren-
iform disorder and schizoaffective disorder. Sanger, lead author of that study,
briefly described it to the participants of the symposium chaired by Lieberman,
and presented it in the subsequent poster session (Sanger and others).

Of the 1,996 patients enrolled in the comparative trial of olanzapine and

haloperidol, 83 were determined to be in their first episode of psychosis. The
random allocation of medication had originally been in the ratio of two
olanzapine to one haloperidol, with 59 of the first-episode patients receiving
olanzapine, 24 receiving haloperidol. The illness characteristics and
demographics of the groups were similar: 67.8% of the olanzapine patients
were male; 32.2% were female. The average age of those receiving olanzapine
was 29 years and had been 27.5 years at illness onset; the average length of
illness episode was 390 days. In the olanzapine group, 28.8% had not
previously received antipsychotics. The haloperidol group was 70.8% male,
29.2% female, with average age of 27 years and 26.3 years age at illness onset;
the average length of episode was 389 days; and 29.2% had not previously
received antipsychotics.

Dosing in this trial was somewhat naturalistic, each drug initiated at 5 mg

daily, and subsequently titrated to optimal effect within the range of 5 mg to 20
mg in single daily dose. The resultant mean modal dose of olanzapine found
most effective for treating first-episode patients was 11.6 mg, lower than the
12.7 mg utilized for the multiepisode cohort. In contrast, the 10.8 mg
haloperidol dose found necessary for the first-episode patients was essentially
the same as the 11 mg required by multiepisode patients. Tollefson suggested
in the symposium on first-episode treatment that the greater sensitivity to
antipsychotic drug effect than is anticipated in these patents relative to those
with chronic, multiepisode schizophrenia was demonstrated by this disparate
dose requirement of olanzapine, but was not evident with haloperidol.

A possible indication that olanzapine exerted greater efficacy and was better
tolerated than haloperidol was the higher number of olanzapine-treated patients
completing the six-week acute phase of the trial (72.9%) than did the 37.5% of
those treated with haloperidol. Analysis of the reasons patients dropped out of
the study revealed that 33.3% of haloperidol patients left the study due to lack
of drug efficacy, in comparison to 13.6% of those receiving olanzapine; and
16.7% of haloperidol-treated patients left because of adverse events, compared
to 1.7% of those receiving olanzapine.
The adverse events prompting patients receiving haloperidol to leave the study
were one report each of akathisia, hypertonia, abnormal thoughts and EPS,
while depression was the adverse condition reported by the single olanzapine-
treated patient to leave the study for any treatment-emergent event. Analysis of
all treatment-emergent adverse events in the first-episode cohort during the
six-week acute phase revealed that 18.6%, 15.3% and 15.3% of olanzapine-
treated patients, respectively, experienced somnolence, asthenia and headache
while no haloperidol-treated patients experienced these events. In turn, 8.3% of
haloperidol patients experienced agitation, 8.3% experienced movement
disorder and 16.7% experienced hypokinesia, reactions that did not occur in
patients receiving olanzapine. An approximate 29% incidence each of
hypertonia and of akathisia was associated with haloperidol compared to an
8.5% incidence of hypertonia and 5.1% incidence of akathisia associated with
olanzapine.

In efficacy measures, olanzapine was found to be significantly superior to

haloperidol in reduction of BPRS total and negative scores and in the PANSS
total and negative score. The investigators concluded that olanzapine was both
more effective than haloperidol and better-tolerated by these patients with first-
episode psychosis. Patients receiving olanzapine were 2.3 times more likely to
respond than haloperidol-treated patients, and were significantly more
improved on total, positive and negative psychopathology. The response rate to
olanzapine was higher, and the superiority of that response to that associated
with haloperidol was greater in first-episode patients than in patients with
multiepisodes. The response rate to haloperidol was approximately the same in
both first-episode and multiepisode patients.

"This tells me," said Sanger, "that the first-episode [patients, in contrast to
those having multiple episodes] have a much greater advantage in going on the
atypical olanzapine, than a typical [neuroleptic]."

Treatment-Resistant Patients

Evidence that increased therapeutic benefit can occur from either changing
from a typical neuroleptic regimen to an atypical anti-psychotic or switching
one atypical antipsychotic compound for another is most pertinent for patients
who have not previously responded to treatment. Both olanzapine and
risperidone were investigated for their efficacy in treatment-resistant patients,
and these results were reported at the ACNP.

Stephen Erhart, M.D., research fellow at the West Los Angeles Veterans Affairs
Medical Center, presented the poster of a comparison of risperidone and
haloperidol in patients with treatment-resistant schizophrenia conducted by his
colleagues (Ames and others). In their trial, 67 patients with schizophrenia
were identified with a history of nonresponse to typical neuroleptics. After
administering placebo during a one-week washout period, these patients were
randomly assigned in a double-blind manner to two four-week treatment
phases. In the initial month, patients received fixed daily doses of either 6 mg
risperidone or 15 mg haloperidol. In the subsequent month, their clinicians
could adjust their dose within ranges of between 3 mg and 15 mg of
risperidone and 5 mg to 30 mg of haloperidol. The patients were then
evaluated weekly during the two months of treatment with measures of
psychopathology, neurotoxicity, mood, obsessive-compulsive symptoms and
subjective response.

The investigators found that risperidone was associated with greater

improvement than haloperidol from baseline on measures of overall
psychopathology. While the subjective response to haloperidol was
predominantly dysphoria, it was largely positive to risperidone, with
risperidone showing no significant liability for either obsessive-compulsive or
depressive symptoms. As expected, a significantly greater percentage of
patients treated with haloperidol (61%) required medication to treat EPS side
effects than did those receiving risperidone (21%).

"There was a notable difference, which we didn't expect at all," commented

Erhart, "in eight risperidone patients of a total of 32, who showed a diminution
of their baseline BPRS scores of between 40% and 80%."

Two investigations of olanzapine in treatment-refractory patients with

schizophrenia employed particularly stringent inclusion criteria defining
treatment resistance. In one study (Conley and others), Robert Conley, M.D.,
and Carol Tamminga, M.D., of the Psychiatric Research Center, University of
Maryland, Baltimore, and Charles Beasley, M.D., Lilly Research Laboratories,
applied the same rigorous criteria that Kane had developed (Kane and others)
to test clozapine's effectiveness. These investigators contrasted the effect of
olanzapine and chlorpromazine (Thorazine) in patients who had failed to
achieve greater than 20% improvement on BPRS in two trials with adequate
dose and duration of different neuroleptics exclusive of haloperidol (1000 mg
daily of chlorpromazine equivalent for six weeks each); and then failed to
improve in their own six-week prospective trial on haloperidol at daily doses
of 10 mg to 40 mg, along with benztropine (Cogentin) up to 4 mg daily.

At the time Conley reported their interim results to the ACNP, 89 patients had
met the inclusion criteria for treatment resistance and had been titrated over
one week to either olanzapine 25 mg daily or chlorpromazine 1200 mg daily,
and maintained on these fixed- dose regimens for seven additional weeks.
Fifty-six patients completing the blinded trial then opted to continue, after a
washout period, in an open study of olanzapine treatment at an initial daily
dose of 10 mg for the first week, then titrated for optimal benefit and tolerance
up to 25 mg daily at the discretion of the investigator clinician. If responsive
after seven weeks of open olanzapine therapy, patients could continue on
olanzapine maintenance. The results were that 48% of the patients were found
improved, with a 20% or greater reduction in total BPRS scores or CGI scores
of less than or equal to 3 or a 1 point fall on the CGI.

While the data on the olanzapine-chlorpromazine comparison remained

blinded at the time of his interim report, Conley indicated that there was
improvement in the overall study group, which was lost in the washout period
preceding the open olanzapine study phase. Improvement was resumed,
however, by week 4 of the open olanzapine treatment, with significant
reduction in BPRS total and positive symptom scores and anxiety factor. This
improvement on olanzapine appeared persistent throughout a 27-week follow-
up period during which olanzapine was well-tolerated. Dry mouth was
reported by approximately 20% of the patients, dizziness in approximately 5%,
and there was an occasional report of blurred vision. EPS was infrequent, with
only one patient experiencing treatment-emergent EPS associated with
olanzapine.

Another study evaluated the response to olanzapine in patients who had

histories of being unresponsive to typical neuroleptics and who had also failed
a well-defined, four-month course of treatment with clozapine (Birkett and
others). The study population was drawn from patients who failed to improve
during their participation in the Cooperative Demonstration Project of
Clozapine for Neuroleptic Nonresponsive Patients, administered at five centers
in Israel under the auspices of the Israeli Ministry of Health.

After a washout period of five to nine days, 50 patients were enrolled for open-
label treatment with olanzapine initiated at 15 mg per day, and titrated for
optimal benefit within a daily dose range of 5 mg to 25 mg for a treatment
period of up to 18 weeks. Forty-five patients continued participation (64%
male, 36% female, mean age 38.7 years) and were evaluated for drug efficacy
weekly for the first six weeks and monthly for the remaining 12 weeks.

On average, these patients experienced a 13% reduction in PANSS total score

and a 14% reduction in BPRS total score during the first six weeks.
Approximately half of these previously unresponsive patients showed at least
minimal improvement on the CGI scale at endpoint. Seventeen patients
(37.8%) evidenced response to olanzapine in meeting the criteria of at least a
20% reduction in PANSS total score; and 20 patients (44.4%) were deemed
responsive by achieving that same reduction in their BPRS total score.

Tollefson commented on his colleagues' study, pointing out that olanzapine

was well-tolerated by the study population with few treatment-emergent events
or patients leaving the study.

"Neutrophils weren't a problem," he pointed out. "Even in patients who had

previously taken clozapine and had some hematological complications, they
were able to tolerate olanzapine." Tollefson added, "They actually improved
from baseline in their EPS scores, so the drug [olanzapine], even up to 25 mg,
appeared not to be inducing a lot of motor side effects."

The investigators concluded that olanzapine demonstrated efficacy in some

patients who had previously been unresponsive to typical neuroleptics and to
clozapine; and that olanzapine was both safe and well-tolerated. Tollefson
observed, "It would suggest that if you've done two [neuroleptic] treatments
and you've moved on to a trial with clozapine, and the patient is either not
tolerating clozapine or hasn't responded, then you've got approximately a 2 out
of 5 chance that you'll see some improvement with olanzapine."

Future Treatment Targets

Cognitive deficits associated with schizophrenia have proved to be elusive
targets in the treatment of the illness. On the closing day of the ACNP meeting,
the progress and prospects for mitigating neurocognitive deficits and
improving functional outcome in schizophrenia were considered in a
symposium chaired by William Carpenter, M.D., Maryland Psychiatric
Research Center, Baltimore, and Richard Mohs, Ph.D., professor, Mt. Sinai
School of Medicine, New York.

The dilemma, as Carpenter described it, is that neurocognitive deficit appears

to be independent, or has minimal relationship to any of the symptom
complexes which have, to some extent, responded to traditional neuroleptics.

"At least [there is] the lack of relationship in the context of clinical trials or
clinical change," he explained. "That is, you can get large changes in
symptoms of schizophrenia without getting changes in the various cognitive
measures."

In his address to the symposium, Michael Green, Ph.D., UCLA Research

Center, Camarillo State Hospital, Camarillo, Calif., credited Lieberman with
making the comparison, and said, "neurocognition of schizophrenia is to the
1990s what negative symptoms were to the '80s."

Barbara Cornblatt, Ph.D., Hillside Hospital, Glen Oaks, N.Y., described her
research findings which indicate that attentional deficits are independent of the
clinical symptoms of schizophrenia (Cornblatt and others). Her evidence,
which counters the traditional view that impaired attention is a component of
the schizophrenia clinical syndrome, includes a longitudinal study of more than
350 subjects at high and low risk for schizophrenia in which attentional
dysfunctions followed a different developmental course from the onset of
psychotic symptoms.

Cornblatt also cited a number of additional studies which indicate that

impaired attention is a core feature of schizophrenia, from its earliest stages in
adolescence through chronic illness in adults.

Furthermore, in many studies, attention does not appear to improve with

neuroleptic treatment, even though clinical symptoms diminish--again
supporting the independence of attentional functioning from clinical state. In
contrast with typical neuroleptics, however, Cornblatt suggested that attention
may improve with amphetamines as indicated in a preliminary study of
patients with schizotypal personality disorders (Kirrane and others).

Cornblatt has focused on developing measures to identify subjects at high risk

for schizophrenia and is hopeful that early detection of deficits that appear well
before clinical symptoms may enable intervention and favorably affect illness
progression.

"Standard neuroleptic treatment is quite beneficial for the clinical symptoms,"

she said, "resulting in a lower amount of hospital time and, hopefully, reduced
relapses. It therefore has a beneficial effect on outcomes." This responsiveness
of clinical symptoms contrasts, however, with the lack of effect of neuroleptics
on the neurocognitive domain which is critical to how well a patient can
function in a community setting. "What I'm calling for," Cornblatt said, "is an
effort to develop a treatment that will focus on the neurocognitive pathway of
the illness."

While there is little correlation between neurocognitive deficits and the clinical
syndrome of schizophrenia, Green summarized several studies which have
ascertained associations between specific deficits and patients' lack of
functioning (Green). For example, verbal memory is associated with all types
of functional outcomes; the level of vigilance is related to social problem
solving and skill acquisition. Measures of executive functioning relate to
community functioning, but not social problem-solving. And the negative
symptoms of schizophrenia are associated with social problem-solving, but not
skill acquisition.

Green emphasized the importance of addressing specific neurocognitive

capacities such as verbal memory and vigilance, which are correlated with
functional outcomes, noting that this will require novel therapies, since
conventional neuroleptics have minimal effect in these domains.

"If we're going to start talking about the treatment of neurocognition and
psychotic symptoms--then it challenges the definition of treatment efficacy,"
he said. "It raises the question of whether or not 'symptom reduction,' narrowly
defined, should be broadened to include something more akin to 'disability
reduction.'"

The possibility that new antipsychotics may have greater utility in the
neurocognitive domain of schizophrenia than older neuroleptics was posed at
this symposium by separate studies related, respectively, by Marder and by
Dilip Jeste, M.D., professor of psychiatry and neuroscience, University of
California, San Diego, and the San Diego VAMC. Marder's group at the West
Los Angeles VAMC and at Camarillo State Hospital reported differential
effects of risperidone and haloperidol on both the spatial and verbal working
memory of patients with schizophrenia (Marder and others 1996b). The study
concluded that benztropine, which is more frequently administered to patients
taking haloperidol, has a negative effect on spatial working memory. The
relative benefit from risperidone appeared greatest among those patients
having the poorest initial performance on verbal working memory measures.

Jeste reported comparing the results of a global cognitive test--i.e., the Mini-
Mental State Examination (MMSE) in middle-aged and elderly patients with
schizophrenia and other psychotic disorders receiving risperidone in one study
and haloperidol and other conventional neuroleptics in another (Jeste and
others). While acknowledging that this comparison was not double-blinded,
Jeste indicated that there was a significant improvement in MMSE score with
risperidone which did not occur with older neuroleptics. The investigators
wrote in their conclusion, "It is not clear, however, whether atypical
antipsychotics such as risperidone produce cognitive enhancement directly, as
the effect might have been secondary to a greater improvement in
psychopathology along with fewer extrapyramidal symptoms compared to
typical neuroleptics."

Tollefson had earlier described the expanded efficacy of antipsychotics that

fulfill the "atypical" pharmacological criteria as potentially preventing further
cognitive deterioration in schizophrenia and/or improving performance
(Tollefson). He related that the clinical effects of olanzapine on cognitive and
psychomotor functions are now being investigated in elderly subjects. In one
study in progress, the effects of olanzapine and haloperidol were compared in
14 healthy, elderly volunteers. While both antipsychotics were associated with
some impaired efficiency in cognitive function after the first day's dose, those
subjects receiving olanzapine returned to pretreatment levels by the next day,
while cognitive deficits associated with haloperidol increased during
subsequent days of treatment. While it is uncertain whether these disparate
effects of olanzapine and haloperidol can be related to relative benefit for
primary cognitive deficits of schizophrenia, Tollefson pointed to the relative
benefit of olanzapine not exhibiting a sustained or increased impairment of
cognition or psychomotor function.

Although little evidence exists of the direct effects of atypical antipsychotics

on neurocognition, Tollefson described one experimental model for this
domain, and the demonstrated activity of olanzapine and clozapine. A
diminished prepulse inhibition (PPI) of startle response has been found to be
characteristic of patients with schizophrenia, and was recently modeled in
animals by administration of N-methyl-D-aspartate (NMDA) antagonists,
phencyclidine (PCP) or the investigational compound MK-801. Clozapine has
previously been found to block the attenuation of normal PPI by NMDA
antagonists, while the typical neuroleptic haloperidol did not. Tollefson related
that olanzapine was also recently found to antagonize this PPI disruption by
NMDA antagonists, and did so to a comparable degree to clozapine (Bakshi
and others). Tollefson concluded, "Cognitive impairment in schizophrenia
greatly impedes psychosocial performance and eventual reintegration into
society." The cognitive features of schizophrenia, then, "become key targets in
the development of new therapeutic modalities."

Evaluate This Article

References

1. Ames D, Wirshing WC, Barringer DM, et al. Risperidone versus haloperidol in treatment-
resistant schizophrenia. Presented at the 35th Annual Meeting of the American College of
Neuropsychopharmacology, Dec. 13, 1996; San Juan, P.R.

2. Bakshi VP, Geyer MA. Antagonism of phencyclidine-induced deficits in prepulse inhibition

by the putative atypical antipsychotic olanzapine. Psychopharmacology (Berl).
1995;122(2):198-201.

3. Birkett M, Dossenbach M, Shoshani D, et al. Treatment failure with clozapine: can

olanzapine be the alternative therapy? Presented at the 35th Annual Meeting of the American
College of Neuropsychopharmacology, Dec. 13, 1996; San Juan, P.R.
4. Conley RR, Tamminga CA, Beasley C. Olanzapine versus chlorpromazine in treatment-
resistant schizophrenia. Presented at the 35th Annual Meeting of the American College of
Neuropsychopharmacology, Dec. 13, 1996; San Juan, P.R.

5. Cornblatt B, Obuchowski M, O'Brien JD, Siever L. Are attentional deficits independent of

clinical symptoms in schizophrenia? Presented at the 35th Annual Meeting of the American
College of Neuropsychopharma-cology, Dec. 13, 1996; San Juan, P.R.

6. Davis JM, Janicak PG, Sharma RP. Does serotonin blockade augment antipsychotic drug
response? Presented at the 35th Annual Meeting of the American College of
Neuropsychopharmacology, Dec. 13, 1996; San Juan, P.R.

7. Green MF. Neurocognitive deficits and prediction of functional outcome in