Food and Drug Administration regulatory procedures assure that drugs approved for the treatment of schizophrenia have documentation of efficacy. What follows is complicated. The physician must decide which drug, at what dose, in what combination, for what period of time, for which aspect of illness, whether to use off-label and at what risk and cost. This decision is case specific, is modified over time, and must weigh patient wishes and alternative approaches. The empirical data is [sic] inadequate.
In a sophisticated meta-analysis, Chakos et al. (2001) found that, of 10 comparisons of second generation versus typical antipsychotics, six found significant differences that favored the second-generation antipsychotics on measures of treatment efficacy; four found no significant difference between treatments. This means that in 40% of the analyzed studies there was no evidence of therapeutic efficacy difference between the atypicals and the typicals. Even in areas where atypicals are purported to be superior to typicals, such as in effects on tardive dyskinesia, only two of five studies reported reduction in tardive dyskinesia prevalence; the remaining three studies showed no difference in tardive dyskinesia rates. The CATIE report (2005) substantiates the latter analysis.
As has been pointed out, above, in the CATIE report perphenazine performed equally as well or better than other atypical antipsychotics (ziprasidone, risperidone, quetiapine) but not as well as olanzapine; perphenazine was used at 200 mg chlorpromazine equivalent dose/day, that is, at a low dose, whereas olanzapine was used up to 30 mg/day, that is, above 800 mg chlorpromazine equivalency/day Moreover, perphenazine demonstrated less toxicity and a better tolerability than olanzapine.
There are an increasing number of publications which indicate that the therapeutic effectiveness difference between conventional and atypical medications is modest at best: no significant statistical differences in treatment response were found in the last observation carried-forward analysis between haloperidol and olanzapine in first psychotic episode treatment groups at 12 weeks. In acutely ill and severely symptomatic patients, relatively low doses of haloperidol or olanzapine were both effective in reducing symptom severity by the end of the 12-week period. When compared with patients given a low dose of haloperidol, risperidone-treated patients experienced similar improvements in positive and negative symptoms and similar risks in psychotic exacerbations. Even the gold standard treatment for schizophrenia, clozapine, does not always live up to its expected superiority: in a double blind, 52-week trial, Lieberman, Tollefson, et al. (2003) studied 164 patients in China who had experienced their first psychotic episode. The investigators found a 71% cumulative response rate at 12 weeks for clozapine-treated patients compared with 62% for the chlorpromazine patients, but no difference between the drugs at 1 year, with a cumulative response rate of 81% for clozapine and 79% for chlorpromazine. To reiterate, in the CATIE study, olanzapine demonstrated therapeutic superiority over the other studied medications (risperidone, quetiapine, ziprasidone, and perphenazine) at higher than recommended doses. There was no significant difference in therapeutic effectiveness among risperidone, quetiapine, ziprasidone, or perphenazine. For all the atypicals, in about 75% of patients, the initial medication was discontinued during the 18 months ofthe study for lack of effectiveness ortolerance problems. These results are comparable to the CAFE study. The apparent superiority of olanzapine other the other second generation antipsychotics, except for clozapine, was corroborated in the CATIE II results.
A meta-analysis ofthe relapse prevention role of antipsychotics shows that the magnitude ofthe advantage ofthe atypicals over conventionals is modest. Despite the proposed superiority ofthe new antipsychotics, the number of treatment failures in both groups was high. The available data do not allow for any conclusions about whether this modest superiority for the new antipsychotics (measured in relapse prevention) is related to enhanced efficacy, better adherence, or a combination of these factors.
The meta-analysis conducted by Davis, Chen, and Click (2003) reached different conclusions to that of Chakos et al. (2001), cited above. According to the new analysis, clozapine produced a better response than first generation antipsychotics (first generation antipsychotic), in the treatment of schizophrenia, with an effect size of d = 0.49, whereas amisulpride, risperidone, and olanzapine clustered around an effect size of d = 0.25 units. The effect of amisulpride, risperidone, and olanzapine over first generation antipsychotics is somewhat less than half the effect of first generation antipsychotics over placebo or clozapine over first generation antipsychotics. Large risperidone and olanzapine studies found consistent differences versus first generation antipsychotics. The authors disagree with prior assertions that the second generation antipsychotics (second generation antipsychotics) are equally efficacious as a homogeneous group; according to their analysis, aripiprazole, quetiapine, remoxipride, sertindole, and ziprasidone show similar efficacy to first generation antipsychotics. See the CATIE I and CATIE II reports commentaries above. Jibson (2003) disagrees with Davis, Chen, and Click’s (2003) conclusions: He asserts that there are aspects of the meta-analysis that preclude their being taken as the last word on the subject.
Most conspicuous was the decision to limit studies to those involving risperidone doses > 4 mg/day and olanzapine > 11 mg/day, but allow studies with quetiapine doses as low as 150 mg/day and ziprasidone and aripiprazole studies without any minimum dose. Clearly, these conditions are not equivalent, since appropriate clinical doses of risperidone and olanzapine were compared to potentially sub-therapeutic doses of quetiapine, ziprasidone, and aripiprazole. This bias would affect the outcome in precisely the direction noted,
Jibson added that recent clinical trials have a trend toward higher dropout rates and smaller effect sizes even for comparator medications, such as haloperidol, that are known to be effective. Newer drugs show smaller effect sizes than the older ones, reflecting a probable artifact of the increasingly treatment-resistant population that self-selects for these studies. It is unclear what the implications of this meta-analysis are for the field of child and adolescent psychiatry.