The first wave of results from the National Institute of Mental Health's
Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) garnered
widespread media attention that may have contributed in part to widely
divergent conclusions about which antipsychotic medication is most
The media's extensive coverage of the CATIE results was indicative of the
study's importance. Launched by NIMH in 1999 with an initial five-year funding
of $44 million, CATIE is the largest and longest medication-treatment study of
patients with chronic schizophrenia. Researchers followed more than 1,400
patients for up to 18 months.
The results were published in the September 22 New England Journal of
Medicine and included data from phase I of CATIE's three planned
The methodology was both complex and comprehensive, as researchers
carefully designed the study to mirror real-world treatment conditions of
patients with chronic schizophrenia, but subject to stringent
Patients were randomly assigned to one of five antipsychotic medications:
the second-generation antipsychotics (SGAs) olanzapine (Zyprexa), quetiapine
(Seroquel), risperidone (Risperdal), and ziprasidone (Geodon) or the
first-generation antipsychotic (FGA) perphenazine (Trilafon). Neither the
patients nor their clinicians knew which drug patients were taking.
Patients who responded well to their assigned medication continued on that
drug for up to 18 months. Patients who discontinued their phase I drug were
randomly assigned to a different drug during phase II of the trial. Phase II
included the four SGAs from phase I and added clozapine (Clozaril); however,
it did not include perphenazine. In phase III, patients who discontinued their
phase II medication could choose to continue with open-label treatment with
another FGA, SGA, or a combination therapy.
The public announcement of the phase I results last month sparked a media
frenzy. Newspaper headlines reflected various—and sometimes
contradictory— interpretations of the study's conclusions.
For example, the Star-Ledger of northern New Jersey, home to
numerous pharmaceutical companies, ran the headline: "Study: Drug Built
in 1950s on Par With New Meds," while in Indianapolis, home of
olanzapine maker Eli Lilly and Co., the Indianapolis Star ran with"
Zyprexa Out-performs Rivals in U.S. Study."
At the same time, clinicians across the country were beginning to draw
their own widely varying conclusions. Clinicians told Psychiatric
News such disparate conclusions as "now all the standard
neuroleptic users can come out of the closet" and "olanzapine
could well become the drug of first choice" contrasted with"
olanzapine is likely to be used second line for patients who haven't
responded to other drugs."
The differing and even contradictory interpretations aren't just academic.
Within a week of the study's publication, some experts began to fear that
misinterpretation of the study's results and conclusions could have a negative
impact on regional and even national health care policy regarding access to
antipsychotic medications (see
The principal investigator of CATIE is Jeffrey Lieberman, M.D., chair of
psychiatry at Columbia University's College of Physicians and Surgeons and
director of the New York State Psychiatric Institute. CATIE included 1,432
patients with schizophrenia at 57 clinical sites (Psychiatric News,
May 18, 2001).
Lieberman said during a media briefing on the study that CATIE was born out
of the need to answer three main questions: Are the newer, second-generation,
or "atypical," antipsychotics (SGAs) more effective than the
older, first-generation, or "conventional," antipsychotics (FGAs)?
How do the newer SGAs compare with each other? Are the more expensive SGAs
"The results of the CATIE study provide the most comprehensive set of
data on the pharmacologic treatment of schizophrenia ever assembled,"
Lieberman added. "These data will guide doctors in their selection of
treatments and clinical management of patients."
The phase I findings indicate that "all these medications
work," Lieberman said, "but they also have substantial
The primary outcome variable measured was length of time patients took the
medication to which they were randomly assigned at the start of the study. The
measure of length of time to discontinuation, for any reason, was chosen by
the CATIE researchers "because stopping or changing medication is a
frequent occurrence and major problem in the treatment of
schizophrenia," the team wrote in the New England Journal of
Medicine report. The measure "integrates patients' and clinicians'
judgments of efficacy, safety, and tolerability into a global measure of
effectiveness that reflects their evaluation of therapeutic benefits in
relation to undesirable effects."
Three-fourths of the patients stopped taking their initially assigned
antipsychotic medication before the end of the 18-month study (see chart).
Fewer patients discontinued olanzapine (64 percent) than any of the other four
medications (between 74 percent and 82 percent). Moreover, patients taking
olanzapine stayed on their medication the longest (an average of 9.2 months),
while those on the other four medications discontinued their medication
significantly earlier (having stayed on them an average of between 3.5 and 5.6
Fewer patients stopped taking olanzapine because of a lack of perceived
efficacy compared with perphenazine, quetiapine, and risperidone. (The
difference between olanzapine and ziprasidone, however, was not statistically
significant.) Overall, there were no significant differences between the drugs
in the percentage of patients who stopped medication because of intolerable
side effects. Risperidone had the lowest rate of discontinuation due to
intolerable side effects (10 percent), while olanzapine had the highest rate
Differences in the medications were also seen in side-effect profiles, with
statistically significant differences found in extrapyramidal symptoms (EPS),
weight-gain and metabolic effects, insomnia, and elevations in prolactin. More
patients reported EPS with perphenazine than with any other drug, a
statistically significant finding, while more patients reported weight gain of
greater than 7 percent of body weight with olanzapine. Olanzapine was also
more likely to be associated with elevated measures of both glucose and
lipid/triglyceride metabolism, while risperidone was more likely to be
associated with elevated prolactin levels.
Even so, the average duration of treatment was the longest in those
randomly assigned to olanzapine. In addition, significantly fewer patients
taking olanzapine required hospitalization during the study for exacerbation
of their schizophrenia (see chart on page 1).
Overall, the CATIE researchers reported that "olanzapine was the most
effective in terms of the rates of discontinuation, and the efficacy of the
conventional antipsychotic agent perphenazine appeared similar to that of
quetiapine, risperidone, and ziprasidone. Olanzapine was associated with
greater weight gain and increases in measures of glucose and lipid
Thus, it turns out that the headlines in both the Star-Ledger and
Indianapolis Star were not entirely off the mark, even though they
appeared to contradict each other. CATIE had indeed shown that olanzapine
appeared to be superior (on certain measures), while perphenazine appeared to
be comparable (on certain measures) to three of the newer SGAs.
Within hours of the release of the CATIE findings, clinicians and
researchers not involved with the study raised significant questions about the
study's methodology, including issues that could affect interpretation of the
The most significant appears to involve the dose ranges used in the study.
The dose ranges were designed to be a best estimate, relying on input from
FDA-approved labeling, clinical experience, and the pharmaceutical companies
that manufacture the medications.
"Our general feeling," said Robert Rosenheck, M.D., a professor
of psychiatry at Yale University and a co-investigator on CATIE, "was
that the actual dosing in the end was somewhat higher for olanzapine [compared
with doses used in other studies], and the dosing for risperidone was somewhat
Other researchers were quick to point out that the differences in dosing
could have biased the overall effectiveness of the medications. The
antipsychotic effect of the medications studied in CATIE is thought to be
related to their ability to bind to and block D2 dopamine receptors
in the brain, noted John Newcomer, M.D., a professor of psychiatry at
Washington University at St. Louis.
"In CATIE, olanzapine was dosed at or above its D2
receptor-binding threshold," Newcomer said. "At higher doses,
D2 binding is increased, and presumably, you might see better
However, Newcomer continued, the other four medications were dosed at or
below their D2 receptor-binding thresholds, potentially reducing
their efficacy trial. "Ideally," he said, "you would want to
see each of the drugs used at doses that produce equivalent D2
Other researchers and clinicians questioned the use of perphenazine as the
comparator representing the older FGAs. Many believed that a primary
difference between the FGAs and the SGAs was the relative absence of EPS such
as akathisia, parkinsonism, and other movement disorders associated with the
newer medications. In general, other studies and clinical experience have
shown that over half of all adult patients taking FGAs for five or more years
developed significant EPS, and some 30 percent to 50 percent developed tardive
dyskinesia (TD). Rates are thought to be even higher for elderly patients (see
Among the older antipsychotics, perphenazine is considered to have moderate
potency. The more commonly used FGA, haloperidol (Haldol), is considered a
high-potency FGA. Higher-potency FGAs have long been thought to have higher
efficacy in treating psychotic symptoms, but they are more frequently
associated with EPS.
The CATIE researchers chose perphenazine precisely because of its moderate
potency, Lieberman explained, and they administered the drug at moderate doses
to help avoid any EPS. Indeed, patients who entered the study with tardive
dyskinesia (TD) were not assigned to the perphenazine group because the risk
was thought to be unacceptable.
"Because people who started the study with TD were selectively
randomized," said John Davis, M.D., a professor of psychiatry at the
University of Illinois, Chicago, "it makes any comparison of TD rates
between these five medications problematic."
As a result, comparisons of CATIE patients who discontinued their
medication because of EPS are "very difficult to interpret," said
Davis, who was not involved in the study.
Yale's Rosenheck said that the study "was not a horse race—it
didn't produce a winner," he said. "Each drug has benefits and
risks. Doctors will have to judge what works best for a particular
In the end, most of the experts interviewed for this article agreed that
the history of patients who participated in CATIE has a bearing on the
application of the results to patients in clinical practice.
With an average age of around 40 years and an average length of time on
antipsychotic medication of over 14 years, the patients followed in CATIE are
likely to have taken multiple medications prior to entering the study. These
are not patients who have acute or new-onset illness, noted Darrel Regier,
M.D., M.P.H., executive director of the American Psychiatric Institute for
Research and Education and director of research at APA.
Indeed, Regier added, "the 70 to 75 percent of patients on medication
at baseline needed to be tapered off prior to treatment before they could be
started in one of the study treatment arms."
Lieberman emphasized that the overall message from the initial CATIE phase
I report should be that "each of these medications is effective, and
each medication has its limitations."
He added, "Despite the overall comparability of the treatments in the
study, there were variations in the therapeutic and side-effect profiles of
each of the drugs. And these can be substantial for patients. It indicates
that diverse response characteristics and treatment needs of individual
patients require a wide range of medication choices."
Lieberman noted that several more reports are to be published regarding
full results of the CATIE study, including further analyses of phase I data.
In addition, phase II and III data will be reported, as well as analyses of
differential effects of the medications on the cognitive and negative symptoms
associated with schizophrenia. The team is also working on a
The first of those reports should be submitted for publication within six
to nine months, Lieberman said.
An abstract of "Effectiveness of Antipsychotic Drugs in
Patients With Chronic Schizophrenia" is posted at<http://content.nejm.org/cgi/content/abstract/353/12/1209>.▪