Abstract
Paul Fitzgerald, Department of Psychological Medicine, Monash University, Melbourne, Australia:
It was with considerable concern that I read the recent letter by White in this journal [1]. The letter outlines legal implications of the continued prescribing of typical antipsychotics and reflects similar concerns raised over recent years with regard to the prescribing of tricyclic antidepressants. These concerns appear to be based on the assumption that atypical antipsychotics are ‘as effective or more effective’ than typical antipsychotics and that the ‘typicals’ have a ‘more severe side-effect profile’. This latter statement is based upon considerations of the incidence of tardive dyskinesia (TD) produced by the two classes of drugs. Both of these assumptions warrant comment.
First, the evidence basis for the generalization of the increased efficacy of atypical agents is unclear and questionable. Certainly there appears to be evidence that clozapine has greater efficacy for patients considered as treatment non-responsive, but the utility of the other atypicals in this patient group has not been convincingly established [2]. In treatment responsive patients, increased effectiveness could be seen in several domains. Regarding positive symptoms, there is minimal evidence to suggest an advantage for atypicals, especially when considering that the majority of comparison studies bias towards the atypical by randomizing patients only partially responsive to typical antipsychotics [3]. The interpretation of these comparison studies has also been widely questioned as the doses of typical medications used are often in excess of that now recognized to best balance efficacy and side-effects [4]. This difficulty must be considered when comparing the effectiveness of typicals and atypicals in the treatment of negative symptoms and cognition as well. Our capacity to infer that the atypicals are more effective in this area is markedly limited by the likelihood of the production of secondary negative symptoms by the excessive dosing of the typical agent in these treatment trials [3]. Where treatment doses are taken into account, typical antipsychotics also have established efficacy against negative symptoms [5]. Despite these biases towards the atypical antipsychotics in most of the trials conducted to date, the degree of increased benefit for either positive or negative symptoms is quite small [3].
These considerations also affect our capacity to judge whether atypicals truly have a safer side-effect profile. Certainly rates of extrapyramidal side-effects (EPSE) are dose-and dopamine-D2-occupancy-dependent. Some atypicals may have a particular advantage in this regard due to fast dissociation from the D2 receptor but this is not necessarily a property of all atypicals [6, 7]. Tardive dyskinesia risk has been inferred from the rates of EPSE seen with these drugs, and reduced TD risk seems well established for clozapine [8]. Based upon the incidence of EPSE and the results of studies to date (for example [9]), the risk of TD with a number of the other atypicals would appear also appear likely to be reduced. However, TD is not the only side-effect on which safety judgements should be made, as may be inferred from the letter by White. All antipsychotics have a range of side-effects of varying severity and clinical importance and the atypical antipsychotics are no exception in this regard. Agranulocytosis, weight gain, glucose intolerance and the effects of neuroendocrine disruption appear with a number of the commonly prescribed atypicals and some of these side-effects appear more commonly in these medications than in some of the traditional agents. Therefore any sensible discussion of potential clinical or indeed medico-legal risks must take these into account. Similarly, there would appear to be no differences between typical and atypical agents in the processes required for obtaining informed consent as would appear to have been suggested in the letter.
These concerns are not just theoretical but of critical importance in shaping the restraints that could easily form around clinical practice. Inferences about treatment effectiveness can be easily drawn from studies that may be later questioned, but the ‘acquired knowledge’ somehow takes on a life of its own, especially with the backing of considerable industry support. If an environment of medico-legal concern falls around the use of typical agents, it is likely to accelerate the diminishing use of these treatments, the failure of new doctors to be trained in their use, and could easily lead to the withdrawal from the local market of many potentially valuable treatment tools. Instead I would suggest that we require balanced education as to the role of low dose rationally prescribed typical agents and long-term independently funded research as to the long-term safety of all of these drugs. The status of either of these categories of agents is no different when considering issues of informed consent. The advent of atypical antipsychotics has been one of great importance for patients with psychotic disorders but we are in increasing danger of throwing the ‘typical antipsychotic’ baby out with the bathwater.