Skip to main content

Transforming the understanding
and treatment of mental illnesses.

Celebrating 75 Years! Learn More >>

Questions and Answers About the NIMH Clinical Antipsychotic Trials of Intervention Effectiveness Study (CATIE) — Phase 2 Results

April 1, 2006

1. What is the goal of phase 2 of the CATIE study?

The primary purpose of the CATIE study was to provide information to guide the everyday treatment of people with schizophrenia. The goal of phase 2 of CATIE was to provide guidance for doctors and patients facing the dilemma of choosing which antipsychotic medication to try next if the first antipsychotic medication was not satisfactory.

Two articles in the April 1, 2006 issue of the American Journal of Psychiatry (1, 2) describe results from phase 2 of CATIE, the National Institute of Mental Health’s (NIMH) Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) study.

In phase 1 of CATIE, people with schizophrenia were randomly assigned to receive treatment with one of the newer (introduced in the last decade), “atypical” antipsychotic medications: olanzapine (Zyprexa®), quetiapine (Seroquel®), risperidone (Risperdal®), or ziprasidone (Geodon®), or an older “conventional” medication, perphenazine (Trilafon®). Approximately one-quarter of all the participants were satisfied with the level of symptom relief they experienced from this first antipsychotic medication, were able to tolerate its side effects, and stayed on it for the entire 18 months of the study. However, three-quarters of the participants stopped taking their first antipsychotic medication before the end of 18 months. The study investigators recorded why a participant stopped taking a medication: if the medication did not control symptoms, if the side effects were not tolerable or, if the patient chose to stop treatment for some other reason.

2. What treatments were available in phase 2 of CATIE?

Two different treatment pathways were available to the participants who stopped medication for any reason during phase 1 yet wanted to continue with the study. Participants in phase 2 collaborated with their doctors to determine the pathway that was best for them.

  • The efficacy pathway was designed for participants who discontinued their phase 1 medication because of inadequate symptom control. This pathway examined the question: if a patient stops taking an atypical antipsychotic because it was not effective enough, what are the benefits of clozapine (Clozaril®) versus another atypical antipsychotic medication as the next treatment? The efficacy pathway compared clozapine to the other newer atypical antipsychotic medications. Participants who chose this pathway were randomly assigned to receive either clozapine or an atypical antipsychotic (olanzapine, risperidone, or quetiapine) different from the one they took in phase 1.
  • The tolerability pathway was designed for participants who discontinued their phase 1 medication because of side effects. This pathway examined the question: if a person with schizophrenia stops taking an atypical antipsychotic because of intolerable side effects, which medication is the best next choice? The tolerability pathway compared ziprasidone to the other atypical medications. Participants who chose this pathway were randomly assigned to receive either ziprasidone or an atypical medication different from their phase 1 medication.

3. Why was clozapine chosen as the comparison medication for the efficacy pathway?

Clozapine is the only antipsychotic medication that has been shown to be more effective than other antipsychotics in controlling psychotic symptoms. Unfortunately, clozapine is associated with serious side effects, including life-threatening blood and heart complications such as agranulocytosis (decreased white blood cell count) and myocarditis (inflammation of the heart); as a result, people who take clozapine must be monitored closely, which includes blood tests. Due to these safety concerns, current practice guidelines limit the use of clozapine for treating people with schizophrenia who have not gotten better on other antipsychotic medications. However, many doctors and patients find clozapine’s potential side effects and monitoring requirements too burdensome and are reluctant to consider using clozapine even where it might be helpful.

4. Why was ziprasidone chosen as the comparison medication for the tolerability pathway?

Phase 1 demonstrated that the atypical antipsychotics differ considerably in their side effects. However, patients and doctors did not know whether a person who had intolerable side effects with one atypical medication might respond to and tolerate a different atypical antipsychotic medication. When the CATIE study began, ziprasidone was the newest antipsychotic and the least familiar to doctors. It was known that the side effects of ziprasidone were very different from the side effects of the other atypical antipsychotic mediations; in particular, ziprasidone was known not to cause weight gain. Therefore, it was important to test whether switching to ziprasidone versus switching to some other atypical antipsychotic medication would be beneficial to participants who stopped taking their first antipsychotic medication due to side effects.

5. How were participants in phase 2 assigned to the two treatment pathways?

Participants who discontinued their first antipsychotic medication because of inadequate management of symptoms were encouraged to enter the efficacy (clozapine) pathway, while participants who discontinued their first treatment because of intolerable side effects were encouraged to enter the tolerability (ziprasidone) pathway. However, participants were free to choose either pathway regardless of the reason for discontinuing their initial treatment. This freedom to choose mimics real-world practice where patients might or might not accept a doctor’s recommendation of one treatment over another.

Of the 1,052 participants who discontinued phase 1 treatment before 18 months, 509 left the study entirely, 99 participants entered the efficacy (clozapine) pathway, and 444 entered the tolerability pathway. Most of the 99 participants who entered the efficacy pathway (86 percent) had stopped their phase 1 medication because of inadequate symptom management; very few participants who stopped due to side effects chose the efficacy pathway.

There were 184 participants who stopped their phase 1 medication because of inadequate symptom control who did not enter the efficacy pathway but instead entered the tolerability pathway, presumably because they were reluctant to try clozapine. Thus, the 444 participants in the tolerability pathway included a mix of participants: 41 percent discontinued their first treatment due to lack of efficacy, 38 percent discontinued due to intolerable side effects, and 21 percent discontinued for other reasons.

6. Was perphenazine included in phase 2?

Perphenazine was not studied in phase 2. Phase 1 found that the older typical antipsychotic medication, perphenazine, generally performed as well as the four newer atypical medications and at the low doses used in CATIE study, perphenazine was just as well tolerated as the atypical medications. The study design did not anticipate this unexpected result that challenged the widely accepted (but never proven) belief that the newer atypical antipsychotic medications are better than all older antipsychotic medications. Thus, the CATIE study did not evaluate, and therefore cannot answer, whether the older antipsychotic medications, such as perphenazine, would be good treatment choices after discontinuing one of the atypical antipsychotic medications.

7. What are the main results of the efficacy (clozapine) pathway?

Most of the participants who chose the efficacy (clozapine) pathway had not benefited from their first phase 1 treatment. Their psychotic symptoms at the start of phase 2 were worse than at the start of phase 1. Furthermore, these participants had worse symptoms than the participants who chose to tolerability pathway.

In this group of very ill patients, clozapine was remarkably effective and was substantially better than all the other atypical medications: 20 out of 45 patients (44 percent) who received clozapine were able to stay on clozapine for the rest of the study, whereas only eight out of 45 patients (18 percent) who received another atypical antipsychotic medication were able to stay on that medication to complete the study. The participants taking clozapine remained on it for an average of 10 months compared to an average of three months for those taking any of the three other medications. Those taking clozapine also had greater symptom reduction than those who took any of the other medications. Only one patient developed agranulocytosis (and was taken off clozapine).

8. What are the main results of the tolerability (ziprasidone) pathway?

The tolerability (ziprasidone) pathway of phase 2 studied the same antipsychotic medications as phase 1 of CATIE (except perphenazine was not included). As in phase 1, there was again a high rate (74 percent) of patients who stopped taking their medication for any reason. However, also as in phase 1, there were differences between the four atypical antipsychotic medications. In phase 2, about 35 percent of the participants who took olanzapine or risperidone were able to continue on their medication until the end of the 18 months of the study. This compares to only 23 percent of those who took ziprasidone and 16 percent of those who took quetiapine that were able to continue.

It was important to examine the results in phase 2 separately for those participants who had stopped their phase 1 medication for different reasons. For those who had stopped phase 1 medication due to inadequate management of psychotic symptoms, those taking olanzapine or risperidone in phase 2 stayed on their medication for a significantly longer duration than those taking quetiapine or ziprasidone, results that are similar to the overall results above. However, for the participants who had stopped phase 1 medication due to side effects, there were no significant differences among the four phase 2 medications. Thus, which medication works best depends in large part on why a patient was switched to it.

The main difference between the phase 1 and phase 2 results is that in phase 2, those who took risperidone received similar benefit as those who took olanzapine. In phase 1, olanzapine’s management of symptoms was better than all of the three other atypical medications.

9. How will the results from the two CATIE phase 2 pathways help doctors and people with schizophrenia?

Doctors now have more information to guide the selection of the next antipsychotic medication to try in treating people with schizophrenia if treatment with a first antipsychotic is not successful. The results have shown that the reason why the first medication was stopped is an important consideration in choosing the next medication. Specifically, the success of symptom management and side effects experienced by the patient on the first medication may help predict which medication may be more successful next.

The CATIE results show that for patients whose symptoms are not wholly responsive to other antipsychotic medications, clozapine is an effective choice for the next step. Almost half the patients who took clozapine in phase 2 continued on treatment to the end of the study. Clozapine was more effective at preventing discontinuation because of inadequate control of symptoms than the other three atypicals in this phase: olanzapine, risperidone, or quetiapine. It was also shown to be more effective at reducing the actual level of symptoms than two of the atypicals in this phase: risperidone and quetiapine. The evidence confirms that clozapine is an effective medication in schizophrenia and the CATIE study results provide further support for efforts directed at both patients and doctors to increase the use of clozapine in the treatment of schizophrenia.

Even with increased support for the use of clozapine, there will still be numbers of patients who need a medication change because their symptoms are not well managed yet do not want to try clozapine. For those patients, CATIE phase 2 results show that olanzapine and risperidone are more effective than ziprasidone and quetiapine. But side effects must be considered.

For patients who stop their first medication because of side effects, there is no clear evidence for switching to one atypical medication over another. Instead, the patient and doctor will best be guided by the nature of the side effects and will have to balance this with the level of symptom control achieved. Olanzapine was associated with substantial weight gain and metabolic problems, more so than the other medications. Ziprasidone was consistently associated with reduction in weight and improvement in metabolic indicators. Risperidone showed the lowest rate of intolerable side effects.


1. McEvoy JP, Lieberman JA, Stroup TS, Davis SM, Meltzer HY, Rosenheck RA, Swartz MS, Perkins DO, Keefe RSE, Davis CE, Severe J, Hsiao JK, for the CATIE Investigators. Effectiveness of Clozapine Versus Olanzapine, Quetiapine, and Risperidone in Patients With Chronic Schizophrenia Who Did Not Respond to Prior Atypical Antipsychotic Treatment. Am J Psychiatry 163: 600-610, 2006.

2. Stroup TS, Lieberman JA, McEvoy JP, Swartz MS, Davis SM, Rosenheck RA, Perkins DO, Keefe RSE, Davis CE, Severe J, Hsiao JK, for the CATIE Investigators. Effectiveness of Olanzapine, Quetiapine, Risperidone and Ziprasidone in Patients With Chronic Schizophrenia Following Discontinuation of a Previous Atypical Antipsychotic. Am J Psychiatry 163: 611-622, 2006.