An infectious optimism has infused the field of schizophrenia with the availability of the new “atypical” antipsychotics. However, an article by Geddes et al in this issue provides sobering evidence which questions this (p 1371).1 By pooling data from 52 controlled studies comparing atypical antipsychotics with the old typical antipsychotics in 12000 patients, Geddes et al. failed to find any clinically significant evidence of superiority in efficacy, or for that matter tolerability, for atypical antipsychotics as a group, once the inappropriately high doses of the comparator were taken into account.
The results are even more sobering given the fact that most of these trials were conducted in patients who had already had a history of partial response to typical antipsychotics—thus having an inherent bias against the older typical antipsychotics. On the other hand, prescription data suggest that atypical antipsychotics account for nearly three out of four new prescriptions for antipsychotics in North America. So, how can we reconcile this large shift in prescribing practices (at least in North America) with the sobering evidence provided by Geddes et al? Is this shift largely a victory of clinical hope and marketing hype over hard evidence, or, is there a truth that is missed in the meta-analysis. The answer lies somewhere in between.
Schizophrenia has long been neglected, by society and by pharmaceutical companies. For most patients partial remission of symptoms is the best that they can hope for. As a result, patients and their caregivers are always searching for something new. Any new medication, whether substantially better or not, is embraced with great enthusiasm, so it is not surprising that even before all the facts are in, atypical antipsychotics have become synonymous with progress and hope for patients with schizophrenia.
However, that is not the entire story. The data from Geddes et al clearly show a much lower incidence of motor side effects with all the atypical agents, and so do other meta-analyses.2 Extrapyramidal side effects are not just incidental “side” effects. For many patients extrapyramidal effects are the central factor in their treatment. Extrapyramidal effects by themselves have been related to a poor outcome, a compromised compliance, secondary negative symptoms, cognitive parkinsonism, and depression as well as long term risk of tardive dyskinesia.3 While the gain on extrapyramidal effects is unequivocal, the mechanism by which the atypical drugs achieve this is not. It was originally thought that the reduction in extrapyramidal side effects reflected a special multireceptor profile of the newer drugs.4 This concept has been questioned recently.5 Improvement in extrapyramidal side effects may result mainly from a more physiologically appropriate level of blockade on the dopamine D2 receptors at a system level, rather than a new therapeutic principle. While a scientific explanation of the mechanism is a matter for debate, there is no doubt that the improvement in extrapyramidal effects is an important advance for patients.
Meta-regression analysis (and such data pooling techniques) often miss in precision what they gain in perspective. For example, carefully controlled clinical trials, designed to specifically look at the issue of enhanced efficacy, have clearly shown the superiority of clozapine in well selected, homogenous populations of patients with seriously refractory disease.6 A meta- analysis that lumps together different trials, diverse populations, and different drugs can miss such a “niche” effect. Also, it may well turn out that the real superiority of atypical antipsychotics is not in their antipsychotic abilities but in their ability to control ancillary symptoms related to mood, cognition, hostility, and a higher level of compliance. If so, these drugs may turn out to be much more effective in the real world, even though their efficacy in controlled trials is not that different. While this real world outcome remains a fervent hope, there are as yet few, well controlled trials that unequivocally support this position. Moreover, the new atypical agents also have their own new side effects (weight gain, diabetes) which will have to be taken into account in any balanced reckoning.7,8
How much is the improvement in extrapyramidal side effects really worth? There is no one answer to such a question. The answer depends on who you ask and what else is at stake. We can only share our experience in our setting. We practice in an environment where the new atypical agents are available to patients at no cost to them, the cost being borne by the public purse. We make patients aware of the relative risks and benefits of low dose typical versus atypical antipsychotics—almost all patients choose atypical drugs. Their choice is mainly influenced by the chance of lower extrapyramidal effects, and the novelty of the atypical antipsychotic drugs. Our impression is that if costs were borne personally the decisions would be different. Which atypical drug to start with is largely a tradeoff between expected side effects; there is little reason to believe there are significant differences in efficacy between the atypical drugs. If one atypical drug fails, we try another or suggest a typical antipsychotic in low doses or as a depot. If all these efforts are unsatisfactory, as they often are, we always suggest a trial of clozapine, for it is as yet unequalled in refractory cases. Thus, the paper by Geddes et al and our thoughts leave the clinician on a tightrope act between the persuasiveness of the marketing claims, the precise but somewhat myopic results of idealised clinical trials, and the complex realities of clinical practice.