Insight Matters
Winter 2006

THE CATIE STUDY FINDINGS:

The Good, the Bad, and the Unexpected
 By Henry A. Nasrallah, M.D., Editor

The CATIE schizophrenia study comparing first and second generation antipsychotics has created considerable controversy about potential impact since the appearance of the initial findings . I received the following letter about the CATIE to which I will respond as well as summarize what I believe the CATIE findings do or do not imply for use of antipsychotics in schizophrenia. As one of the CATIE investigators, my perspective is not just as Editor, but also as a researcher and clinician.

Dear Editor:

I am responding to your call for opinions regarding the CATIE study. I teach the Journal Club at the NEOUCOM Department of Psychiatry and we recently evaluated the CATIE resultsr published in NEJM.

These are some of the conclusions that I have reached regarding CATIE;

1. None of the antipsychotics is very good. The most effective medication, olanzapine, also conferred the highest risk of weight gain, metabolic changes, and risk factors for vascular disease. This study should be a wake-up call for researchers to go back to the drawing board to seek antipsychotic medications that are both effective and do not shorten the life of the patient.

2. The first generation antipsychotic (FGA), perphenazine performed comparably to the second generation antipsychotics (SGAs). Newer and more expensive does not necessarily mean better when it comes to medications.

3. The dosing of perphenazine was lower than commonly employed. Are low doses of FGAs similar to regular doses of SGAs?

4. Since none of the antipsychotics was highly effective, psychiatrists and their patients need to have the flexibility to try the full range of available treatments to optimally match medication to each individual.

5. There needs to be more research that is not beholden to pharmaceutical company sponsorship.

6. It is important to include aripiprazole in a similar study or the next phase of CATIE.

Dennis Helmuth, MD, PhD

Clinical Assistant Professor of Psychiatry

NEOUCOM

I will respond to Dr. Helmuth’s points and address some additional issues as well.

1. It is not true that the CATIE schizophrenia outpatient trial demonstrated that none of the antipsychotics is very good. What the CATIE demonstrated is that this serious brain disease ,schizophrenia ,is associated with a high discontinuation rate (74%) which reduces the effectiveness of any antipsychotic used to treat it.

It should be emphasized that discontinuation includes not only dropping out of treatment but also switching to another antipsychotic, which mirrors what happens in the “ real world” of a community mental health center. Schizophrenia is associated with several features that may disrupt outpatient adherence to treatment such as anosognosia (lack of insight), cognitive dysfunction (especially memory and executive function deficits), negative symptoms (apathy, amotivation, avolition), and substance abuse. The fact that different antipsychotics have side-effects that may be intolerable for some patients is no different from any class of drugs for any other medical disease. Discontinuation can be reduced by matching the patient with the appropriate antipsychotic and then avoiding discontinuation due to lack of efficacy by titrating that antipsychotic to an optimal dose. Controlled past efficacy studies that preceded the CATIE showed very similar efficacy for all the atypicals when given at equivalent doses for chronic schizophrenia .

In the first phase of the CATIE (comparing four randomly assigned atypicals plus perphenazine), olanzapine did show statistically better efficacy but also had the worst tolerability (mainly due to metabolic adverse effects). Olanzapine was the most appropriately dosed antipsychotic in phase I of the CATIE (20.1 mg/day) compared to relatively lower mean doses for the others (3.9 mg of risperidone, 543.4 of quetiapine, 112.8 of ziprasidone and 20.8 of perphenazine.) Most clinicians would agree that an equivalent dose of the other antipsychotics should have been 5 mg. of risperidone, 800 mg. of quetiapine, 160 mg. of ziprasidone, and 40 mg. of perphenazine for a chronic schizophrenia population like the CATIE where the duration of illness was 16 years.

Dr. Helmuth’s “wake-up call” to develop more effective antipsychotics is valid, but not just because of the CATIE findings. Long before the CATIE was even conceptualized psychiatry,researchers have been diligently seeking better pharmacological agents with broader efficacy on negative and cognitive deficits and a better safety profile. The atypicals were a step in that direction (they do have broader efficacy while avoiding the neurological movement disorders of the first generation). Medical progress occurs in gradual increments and rarely solves all the problems at once. As with every disease, the risk-to-benefit ratio of the treatment should be considered and most atypicals arguably help persons with schizophrenia more than they harm them. There is no evidence that antipsychotic drugs reduce the life-span of those afflicted with schizophrenia, but as with most other medications, they can occasionally have serious side-effects or fatalities such as laryngeal dystonia with the first generation neuroleptics or diabetic ketoacidosis with some of the second generation atypicals.

2. While perphenazine appeared to perform comparably to the atypical antipsychotics in the CATIE with regard to discontinuation, that does not mean it is “equal” in other ways, and does not mean that the low cost of generic perphenazine justifies its use as first-line. Several decades of experience with perphenazine have established its limited efficacy, serious neurotoxicity (acute EPS and TD), and failure to improve negative, cognitive, and mood symptoms. Further, the emerging evidence for the neuroprotective effects of atypicals compared to the older antipsychotics is a very important neurobiological dimension not assessed in the CATIE trial. It is worth mentioning that perphenazine had the highest EPS in the CATIE despite the fact that none of the 231 subjects with TD at enrollment in the CATIE was assigned to perphenazine but was randomized to one of the four atypicals. This was done for ethical reasons and it would have been very revealing (but ethically untenable) to observe what perphenazine would have done to patients with TD had they been assigned to it. Finally, two points about the cost issue: a) medications constitute only 3-4% of the total annual cost of schizophrenia disease management according to many studies, and b) one case of TD would be far more expensive than the differential costs of old and new antipsychotics.

3. It is true that perphenazine was dosed at a lower level in the CATIE than in the “real world” during the era of the first generation antipsychotics. Despite that, perphenazine still had the highest EPS rate in the CATIE trial. The high serotonin antagonism by the atypicals differentiates them from older antipsychotics like perphenazine. As mentioned earlier, the high discontinuation rate for all atypicals in the CATIE is not related to efficacy but probably due to multiple factors inherent in the disease itself regardless of what drug is administered.

4. Many psychiatrists ,including myself, would agree that optimizing the dose of antipsychotics is a key factor in efficacy and effectiveness, as long as the tolerability threshold is not breached. The upper limit of the FDA approved doses for all antipsychotics is often exceeded in treatment-resistant chronic schizophrenia who responds only partially to the usual dose rang. Thus, in the real world, clinicians usually do not switch to another antipsychotic until they had tested the higher dose ranges , unlike the CATIE where only 40% of patients received the highest dose range.

5. I completely agree that more clinical trials not funded by pharmaceutical companies themselves would be welcome. NIMH and private sponsors (e.g. Stanley Foundation) do support some psychopharmacology research but the vast majority of clinical trials trials [registration trials for FDA approval and subsequent phase IV trials] are conducted with industry support. However, while some are tilted in design and interpretation, most industry-sponsored studies are not biased and their results are often accepted and published by top-tier peer-reviewed journals.

6. Aripiprazole was included in the CATIE trial, but only in phase 3, which is open -label. Enrollment in the CATIE was practically 90% complete when aripiprazole arrived on the market and it was too late for it to be incorporated in the first two [blinded] phases of the CATIE. I certainly agree that aripiprazole should be included in any future effectiveness studies or replications of the CATIE, but I also anticipate that intramuscular long-acting atypical formulations will also be included in future controlled effectiveness trials because they may ameliorate the discontinuation rate in some patients.

Back to newsletter