While the recent publication of Phase I results of the landmark Clinical
Antipsychotic Trials of Intervention Effectiveness (CATIE) study enabled
pharmaceutical companies and stock pundits to declare winners and losers among
the marketed antipsychotics, the big winners may be
clinicians who treat the 3.2 million Americans suffering from schizophrenia.

"The results of the CATIE study provide the most comprehensive set of data
on the pharmacologic treatment of schizophrenia ever assembled," Jeffrey A.
Lieberman, M.D., a Columbia
University psychiatrist
and lead author of the study, said at a press conference. "These [results] will
guide doctors in their selection of treatments and clinical management of
individual patients. This is because no study has ever examined all marketed
drugs in a controlled fashion for such a long time period using such extensive
measures of safety and efficacy, much less the cost data."

Thomas Insel, M.D., director of the National
Institute of Mental Health, which sponsored the $44 million study, described it
as the largest, longest and most comprehensive independent trial ever done to
examine existing therapies for this disease. The 18-month study involved 1,460
participants at 57 different clinical sites in 24 states. The effectiveness or "practical
trial" sought to be broadly representative of real-life settings and included
patients with physical or other mental health problems in addition to
schizophrenia, as well as patients from diverse ethnic and racial backgrounds.

By and large, the patients in the study were not refractory or acutely ill,
Lieberman told Psychiatric Times in
an exclusive interview. "They were people who were receiving medicines as
outpatients and were relatively stable. They went into the study seeking a
better treatment," he said.

Phase I of the study compared four of the second-generation or atypical antipsychotics (olanzapine [Zyprexa], quetiapine [Seroquel], risperidone [Risperdal] and ziprasidone
[Geodon]) and one first-generation antipsychotic, perphenazine (Trilafon). Aripiprazole (Abilify), Lieberman
noted, was not approved until November 2002, so it could not be in Phase I. It
was, however, added to Phase III of the study, so there will be descriptive
data on how it does compare with the other medicines. The dose range for each
medication was chosen based on the advice of experienced clinicians, clinical
practice patterns from national pharmacy databases and discussions with the
drugs' manufacturers. As described in the New
England Journal of Medicine, the primary outcome measure for Phase I of the
CATIE trial was all-cause treatment discontinuation, which reflects both
clinicians' and patients' judgments about efficacy and tolerability (Lieberman
et al., 2005). Secondary outcomes were the specific reasons for treatment
discontinuation (e.g., inefficacy or intolerability owing to such side effects
as weight gain, extrapyramidal signs or sedation as
judged by the patient's clinician) and scores on the Positive and Negative
Syndrome Scale (PANSS) and the Clinical Global Impression
(CGI) scale.

Robert Baker, M.D., medical director at Eli Lilly and Company, explained
that CATIE showed olanzapine "to be more effective on
the [all-cause] discontinuation rate than other medications studied," and it
had favorable findings for olanzapine "in terms of
duration of successful treatment and risks of rehospitalization."
The average time to discontinuation, according to Lilly's press statement, was
9.2 months for olanzapine, 4.6 months for quetiapine, 4.8 months for risperidone,
3.5 months for ziprasidone and 5.6 months for perphenazine. The differences were statistically
significant for olanzapine compared with risperidone and quetiapine, but
not for perphenazine or ziprasidone.
Total PANSS scores improved over time in all groups, according to Lilly, but
patients taking olanzapine had greater initial
improvement.

In Pfizer Inc.'s comment on the study results, Joseph Feczko,
M.D., chief medical officer, said there were small differences in efficacy
among the agents but important differences in potentially dangerous long-term
health risks. The company's press statement said its drug ziprasidone
"was the only medicine to reduce weight, cholesterol, lipids and measures of
glucose, while effectively improving patients' psychiatric symptoms." Patients
treated with the drug experienced an average weight loss of 2 lb, as well as
cholesterol and triglyceride reductions of 9.2 mg/dL
and 18.1 mg/dL, respectively.

Meanwhile, Janssen Pharmaceutica Products, L.P.,
questioned the results of the trial, stating that the efficacy results for risperidone "did not demonstrate the full efficacy of Risperdal because many patients in the CATIE trial received
doses that were too low."

Glenn Gormley, M.D., Ph.D., chief medical officer
of AstraZeneca, noted that the "study points to the
importance of balancing the risks and benefits to patients when choosing an
antipsychotic. The balance of efficacy and tolerability that Seroquel provides makes it an ideal choice in the treatment
of schizophrenia."

Schering-Plough, which provided perphenazine for
the study, discontinued production of the drug in May of 2002, but a company
representative noted the drug is available as a generic.

"The biggest surprise of the study was that the older medication, perphenazine, was comparably effective to at least three of
the new medications and not much worse than the new drug that did the best--olanzapine," said Lieberman, chair of the department of
psychiatry at the College of Physicians and Surgeons at Columbia University and
director of the New York State Psychiatric Institute. (He conducted most of the
study while still a professor of psychiatry at the University of North Carolina.)

Contrary to expectations, Lieberman said, the older, less expensive
medication did not cause substantially more Parkinsonian-type
side effects than the new drugs. He attributed this to the fact that perphenazine has a lower potency than the usual comparator,
haloperidol (Haldol), and that it was given at
moderate doses. Perphenazine, he said, "clearly is an
effective treatment and should not be disregarded just because it is older." He
added that clinicians might extrapolate from the study results that if they
used first-generation antipsychotics in moderate
doses, particularly those not of the highest potency, they could diminish the
substantial side-effect burden.

Robert Rosenheck, M.D., professor of psychiatry
and epidemiology at Yale
University and one of the
study authors, noted in the press conference that when the CATIE study was
being designed in 1998 and 1999, some researchers questioned the idea of
comparing second-generation antipsychotics to a
first-generation one. "The field had concluded that the new drugs were
superior. Because we were willing to ask a question that most people thought
they had the answer to, we were open to getting the surprising answer that
there wasn't that much difference," he said.

Adverse Effects

During the press conference, Lieberman noted that, although the treatments
used in the study are effective and are "far better than no treatment at all,"
patients and their physicians are looking for more in the way of symptom relief
and recovery, and "they want this with fewer side effects than current
treatments impose."

This desire for something better was reflected in the overall
discontinuation rates. Nearly three-quarters (74%) of the patients discontinued
the study medication before 18 months: 64% of those assigned to olanzapine; 75%, perphenazine;
82%, quetiapine; 74%, risperidone;
and 79%, ziprasidone. The rates of treatment
discontinuation due to intolerable side effects also differed between
treatments (p=0.04). Risperidone had the lowest percentage of patients discontinuing
due to intolerability (10%), followed by 15% each for patients taking perphenazine, quetiapine or ziprasidone, and 18% for those taking olanzapine.
Olanzapine, more than the other antipsychotics,
was associated with weight gain and increases in glycosylated
hemoglobin, total cholesterol and triglycerides--changes linked to the
development of metabolic syndrome (Lieberman et al., 2005).

"What we are seeing ... is the fact that there is a spectrum of propensity for
the newer medications to produce [weight gain] and a change in these metabolic
measures, but olanzapine does this the most. So we
see that quetiapine and risperidone
also, but to a lesser degree, produce changes in these measures," Lieberman
said. Thirty percent of patients in the olanzapine
group gained 7% or more of their baseline body weight during the trial compared
to 7% taking ziprasidone, 12% taking perphenazine, 14% taking risperidone
and 16% taking quetiapine (Lieberman et al., 2005).

In an editorial accompanying the NEJM
article, Robert Freedman, M.D., (2005) warned, "Even the most feared side
effect of first-generation drugs, tardive dyskinesia, seems less troubling than potentially fatal
metabolic problems." Responding to Freedman's comment, Rosenheck
said, "This study showed that there [were] increased weight and increased
triglycerides that put someone at risk, but you can't leap from increased risk
to increased mortality. That will take much larger and much longer studies to
answer whether there is actually increased mortality."

With regard to other side effects, patients in the olanzapine
and quetiapine groups had lower rates of insomnia
than patients in the other groups: olanzapine, 16%; quetiapine, 18%; risperidone,
24%; perphenazine, 25%; and ziprasidone,
30%. Quetiapine was associated with a higher rate of anticholinergic effects than other drugs (31% versus 20% to
25%), and risperidone was associated with hyperprolactinemia (Lieberman et al., 2005). "Concerns
about potential prolongation of the corrected QT interval with ziprasidone and of cataracts with quetiapine
were not realized in this study," the study authors said in the NEJM article. There were no significant
differences among the groups in the incidence of extrapyramidal
side effects, akathisia or movement disorders as
reflected by rating scale measures of severity.

Given the variations in efficacy and side effects, Liebermann said a "cut-and-dried
algorithm" would not work. Rather, the choice of the antipsychotic should be
governed by physicians' assessments of a patient's clinical status and past
history of response. As an example, he said, if a clinician is treating a
patient who has particularly severe symptoms and who has had an unresponsive or
less responsive history with other medications, the clinician might be more
inclined to go with a drug like olanzapine, if the
patient is not ready to move to clozapine (Clozaril), which is generally reserved for patients whose
symptoms are resistant to other medicines.

"You might want to use it because it looks like it is most effective, and
your priority is to control the psychiatric illness, even if there is a
potential for the patient getting more side effects," he said. "Then again, if
you have a patient who is a new patient or is a patient who has done well in
the past on other medications but has shown a particular sensitivity to side
effects, then drugs like ziprasidone or risperidone, as well as perphenazine,
are really good choices, because risperidone had the
lowest rate of discontinuation due to side effects and ziprasidone
had the least amount of weight and metabolic effects along with perphenazine."

In the PT interview, Lieberman
warned against using the Phase I results of CATIE "to restrict options or to
restrict formularies. If anything, they suggest that doctors and patients need
to have a range of choices, because of the variability among the antipsychotic
drugs with regard to efficacy and side effects, and among patients in terms of
their response characteristics and side effect sensitivities." Additionally, he
said, "it is premature to come to any kind of decision about possible policy
implications until the results of the rest of the study come out, particularly
the cost-effectiveness data."