Editorial
If ‘atypical’ neuroleptics did not exist, it wouldn’t be necessary to invent them: Perverse incentives in drug development, research, marketing and clinical practice
Bruce G. Charlton
Medical Hypotheses 2005; 65: 1005-1009
Summary
Perverse incentives in drug development, research, marketing and clinical usage can be illustrated by considering the example of the so-called ‘atypical’ neuroleptics which have grown to become a standard – indeed expanding – part of psychiatric practice despite their probable inferiority to older sedative agents. There is now ample evidence to suggest that neuroleptics (aka. anti-psychotics and major tranquillizers) are dangerous drugs, and patients’ exposure to them should be minimized wherever possible. This clinical imperative applies whether neuroleptics are of the traditional type or atypical variety, albeit for different reasons since the traditional agents are neurotoxic, while atypicals are mainly metabolic poisons. Usage of traditional neuroleptics seems indeed to be declining progressively, but the opposite seems to be happening for ‘atypicals’, and new indications for these drugs are being promoted. Yet the atypical neuroleptics are a category of pharmaceuticals which are close to being un-necessary since there are safer, cheaper and pleasanter substitutes, such as benzodiazepines and the sedative antihistamines (e.g. promethazine). If ‘atypical’ neuroleptics did not exist, it would not be necessary to invent them. Analysis of how such expensive, dangerous and inferior drugs as the ‘atypicals’ have nevertheless come to dominate clinical practice casts light on the perverse incentives which now motivate the pharmaceutical industry in an era of massive state regulation. The lack of positive incentives to deploy off-patent drugs is longstanding, but there is a new disincentive in the widespread but erroneous belief that only randomized controlled trials (RCTs) can provide valid ‘evidence’ of effectiveness. Consequently, those who control RCTs now control clinical practice. It sometimes makes commercial sense to develop and market new drugs that are inferior to existing agents, since new drugs are patent-protected and can be promoted on the back of a mass of new RCTs funded and ‘owned’ by the pharmaceutical corporations. The current regulatory and patenting situation, therefore, requires major reform if drug efficacy and patient safety are to become higher priorities. Given that psychiatric practice is apparently ‘locked-in’ to prescribing atypicals, and if (as seems likely) most informed individuals would wish to avoid neuroleptics for themselves and their loved-ones except as a last resort; then in the short-term it may be wise for patients and their families to explore the possibilities of increased self-management of psychiatric problems using over-the-counter drugs, such as the sedative antihistamines. In the long-term, there need to be legal reforms to change the regulatory and commercial framework of incentives relating to drug development. These might include new forms of short-term re-patenting of old drugs.
***
Neuroleptics to suppress agitated or psychotic behaviour
The perversity of currently operative incentives in the drug development, research, marketing and clinical use can be illustrated by considering the example of the so-called ‘atypical’ neuroleptics, which have grown to become a standard part of modern psychiatric practice. Atypicals represent a backward step in therapeutics, probably being no more effective, much more dangerous, and greatly more expensive than already-existing agents. The policy-implications derive from the fact that the atypical neuroleptics have emerged in the most advanced modern societies, and in a context of extreme pharmaceutical regulation. This situation highlights the unintended, but pathogenic, motivators which now operate to perpetuate undesirable trends in clinical practice.
The traditional neuroleptics (aka. anti-psychotics or major tranquillizers) are very powerful agents for controlling agitated and psychotic behaviour when compared with the drugs in use before their discovery [1]. The pre-neuroleptic behavioural control agents had been mainly sedatives (e.g. antihistamines, barbiturates, bromides and paraldehyde). However, the neuroleptics, such as chlorpromazine, haloperidol and fluphenazine decanoate – are now seen to be neurotoxic. They achieve their distinctively powerful behavioural control by making patients Parkinsonian in a dose–dependent fashion, with permanent neurological disease (tardive dyskinesia) as the long-term consequence of prolonged high dosages [1], [2], [3] and [4]. Furthermore, it is now apparent that neuroleptics induce dependence, so that withdrawal tends to provoke psychotic episodes, and patients who have never taken neuroleptics have a better prognosis than those who are maintained on them [1], [2] and [3]. Short-term benefits are achieved at the cost of long-term harm. On top of this, neuroleptics are notoriously unpleasant. Indeed, patients hate them, since the emotional effects of Parkinson’s disease include emotional blunting, mental dullness and demotivation; and sometimes agitated feelings of akathisia (inner turmoil and physical restlessness). These are not ‘side effects’ of neuroleptics, on the contrary they are the core therapeutic actions of neuroleptics. In a nutshell, neuroleptics work by causing the psychological symptoms of Parkinsonism [4].
So, traditional neuroleptics are exceptionally nasty drugs whose use should be minimized as a matter of policy, and for which safer and less unpleasant replacements should be sought. So much is probably uncontroversial, at least at a personal level among informed psychiatric scientists, although for structural, professional reasons this conviction is seldom clearly stated as such.
Replacing traditional neuroleptics – the ‘atypicals’
The main response to this state of affairs has been the displacement of traditional neuroleptics by a group of drugs self-styled as ‘atypical’ neuroleptics, such as clozapine, risperidone and olanzapine [2]. Clozapine is not a neuroleptic, by classical definitions [1] and the others of this class are much weaker neuroleptics than traditional drugs. Because atypicals have less tendency to cause Parkinsonian symptoms and tardive dyskinesia than traditional agents, they are consequently less powerful at suppressing behaviour than traditional neuroleptics; but atypicals are powerful enough (it turns out) for most purposes. It seems that in their clinical effect, atypicals are essentially the pre-chlorpromazine sedative antihistamines, re-invented and re-packaged for modern times. Indeed, an old antihistamine such as cyproheptadine might well be counted as functionally an ‘atypical’ neuroleptic by today’s criteria i.e. it blocks serotonin 2 and cholinergic receptors, it is (probably, like many old antihistamines) only weakly anti-dopaminergic, it is powerfully sedative and it causes weight gain. Yet cyproheptadine, is an off-patent drug, and available cheaply over-the-counter.
But if most of the atypicals are only weakly neurotoxic they are instead ‘metabolic poisons’; having poorly-understood (this poor understanding itself being a damning indictment of the structural biases of contemporary psychiatric research) but clearly damaging effects on various aspects of energy metabolism [2] and [5]. Atypicals tend to cause physiological damage, such as glucose intolerance, diabetes and perhaps pancreatitis; gross weight gain and cardiac conduction problems. Atypicals are also associated with significantly increased mortality, from a variety of causes [6] and [7]. (Clozapine, notoriously, causes lethal blood dyscrasias, requiring continuous monitoring.) These side effects occur in the context of long-term prescription for psychiatric disorders which are often self-limiting and rarely fatal [1] and [2].
Perhaps the main useful lesson from the emergence of the ‘atypical’ (i.e. weak) neuroleptics is that psychiatrists did not actually need to make all of their agitated and psychotic patients Parkinsonian in order to suppress their behaviour. Atypicals are highly sedative agents. Apparently, the kind of sedation provided by the ‘atypicals’ is sufficient for behavioural control in most instances. This should not have been a surprise, since sedation was the standard method of controlling agitation and acute psychoses before the emergence of neuroleptics and in situations where neuroleptics were not available (e.g. in the Eastern Bloc and underdeveloped countries) [1], [2] and [8]. Indeed, the value of sedation should not be underestimated. Sedation is not merely a relatively safe way of controlling agitated behaviour; sedation also provides a potentially ‘curative’ benefit for psychotic patients with a causative element of delirium [4]. Sleep has profoundly restorative qualities where sleep disturbance is severe and prolonged, as is the case for many psychotic patients. In sum, drugs which promote sleep likely have an ‘anti-psychotic’ effect, as well as making patients both feel and function better [4] and [8].
In terms of therapeutic value, it therefore seems likely that ‘atypicals’ are merely an unusually dangerous way of sedating patients. In therapeutic terms these drugs therefore represent a significant backward step. Rationally, the atypicals should now be dropped and replaced with safer sedatives. Potential neuroleptic-substitutes which already exist would include benzodiazepines and sedative antihistamines, such as promethazine [4] and [8].
The situation therefore seems straightforward. Neuroleptics – whether traditional or atypical – are toxic and dangerous drugs, their use should be minimized, and safer alternatives must be sought as a matter of urgency. This should no longer be a matter for debate, but for remedial action.
Psychiatry ‘locked-into’ atypical use
Usage of traditional neuroleptics seems indeed to be declining – albeit without properly explicit transparency of the rationale [2]. But the opposite seems to be happening for ‘atypicals’ which have been increasing in use, and new indications for these drugs are being promoted – e.g. for affective disorders, and in very young children. Given the evidence for the dangers of atypical neuroleptics, the need to minimize exposure, and the availability of better substitutes – why is ‘nothing’ happening. Why indeed are ever-more new indications for atypical neuroleptics being pushed, when they almost always cause physiological damage and sometimes cause fatalities?
A full answer would need to include social attitudes to psychiatric illness, and the relative powerlessness (lack of autonomy) of psychiatric patients compared with other patient groups. But the situation is also illustrative of the incentives which operate generally in the pharmaceutical industry. Pharmaceutical companies must make money in order to survive, and this is usually done by researching and marketing new, relatively expensive, patent-protected drugs. These incentives lead to the production and promotion of new drugs even when they are less effective and/or more dangerous than existing agents. This clinically-damaging consequence is not simply a matter of ‘benign’ neglect of old, off-patent drugs – but actually a consequence of the driving imperatives of the system.
It is probable that the old antihistamines (for example) would often provide a safer and less-unpleasant alternative to both typical and atypical neuroleptics. The atypicals have confirmed that sedation without Parkinsonism can provide sufficient behavioural control for at least a significant number of patients – the clinical problem is simply to find the safest and best form of sedation for particular patients [4] and [8]. So, it is plausible that traditional sedating antihistamines, such as promethazine, would be sufficient to control acutely agitated behaviour in many psychotic patients, thereby avoiding the potential dangers of exposing these patients to either the traditional or atypical neuroleptics. The balance of potential harm would almost certainly favor the antihistamines over the neuroleptics.
Why not use sedative antihistamines? Because those who control RCTs control practice
Given the powerful reasons for avoiding or minimizing patient exposure to neuroleptics, it would not be unreasonable to consider immediately substituting antihistamines for neuroleptics wherever possible, especially by an initial trial of sedating antihistamines for newly admitted acutely disturbed psychotic patients. Since traditional neuroleptics are the most powerful agents for suppressing agitated behaviour, there may be rare situations in which their use is unavoidable. But the ‘atypical’ neuroleptics can probably always be substituted by safer sedatives, such as the benzodiazepines or antihistamines. Naturally, in the long-term, the balance of benefits and harm between sedatives and neuroleptics would need to be quantified by formal clinical trials in various situations and patient groups.
However, such a shift in practice – however desirable from the perspective of patients – cannot and will not happen without the development of new positive incentives to develop novel uses for old drugs. The lack of marketing effort, and consequent underusage, for old, off patent drugs is nothing new. But what is new is the erroneous belief that the only valid source of evidence regarding drug effectiveness comes from randomized controlled trials – particularly very large mega-trials [9] and [10]. This ignorant over-valuation of RCTs has reached the point where those who control randomized trials control clinical practice.
It is widely believed, and at the highest levels of policy, that RCT evidence is an uniquely ‘unbiased’ methodology because of the use of randomization. Yet mega-trial RCTs are not hypothesis-testing science, but instead a method of measuring the effect-size of therapies – and their meaning depends on what is being measured and how [10]. Therefore, RCTs do not constitute transparent, explicit information about therapies; but always require contextual interpretation [11]. Nowadays, the most influential contextual interpretation occurs within the constraints of marketing imperatives, which enable the biases of RCTs to be exploited by those who control them. The distorting effects of highly-selective subject recruitment, methodological specificity (e.g. choice of dosages and outcome measures), together with selective reporting and publicity (driven by marketing rather than scientific considerations), render RCTs just as bias-prone as any other clinical research methodology – and much more damagingly so since their universal generalizability is assumed [11].
An unintended-consequence of ignorantly over-valuing and misapplying RCTs is that drugs for which large quantities of recent randomized trial data are lacking have become – de facto – unusable, even when the use of these agents is supported by a coherent body of pharmacological, clinical and historical evidence as is the case for using sedatives in agitation and acute psychosis [8].
The pervasive misunderstanding and misinterpretation of RCTs among policy-shapers is another explanation why modern psychiatric practice is apparently ‘locked-in’ to the status quo of using ever-more atypical neuroleptics, no matter how harmful this situation may be, and how much preventable misery, disease and death this practice may be causing. Psychiatrists, as individuals, may be aware of the problems of ‘atypical’ neuroleptics, may indeed wish to stop using them and try something else; but other types of practice are all-but-indefensible under present conditions. Presumably this nightmarish situation will continue until the patents on atypicals lapse, and a new generation of expensive patent-protected agents has emerged with an accompanying mass of RCT data to backup their mass marketing. The irony is that, as things stand, any replacement drugs may be (even) worse than the atypicals.
Self-management for clinical self-defense, reshaping incentives for therapeutic progress
In the immediate term, it seems plausible that already-existing, cheap, sedative drugs offer realistic hope of being safer, equally effective and subjectively less-unpleasant substitutes for neuroleptics in many patients (although precise epidemiological measures of their relative safety and effectiveness are lacking). If we were to apply the test of choosing what treatment we would want for ourselves or our relatives with acute agitation or psychosis, knowing what we now know about neuroleptics, I think that many people (perhaps especially psychiatric professionals) would wish to avoid neuroleptics except as a last resort. And we would probably not wish to wait a decade or so for the accumulation of a mass of randomized trial data (which may never emerge) before making the choice of less dangerous and unpleasant drugs.
Since psychiatrists cannot or will not stop using neuroleptics, the implication may be for more widespread and better informed self-management of psychosis [12] – including more informal management by families of their members with psychiatric problems. Unless prescription-only agents are obtained using the internet, this probably suggests a greater use of ‘over-the-counter’ (OTC) drugs which can be purchased from a pharmacist. For instance, there are many sedative antihistamines which are available OTC and without prescription. Potentially, it might become as common for people to self-manage or prevent milder episodes of psychotic agitation or hypomania using sedating antihistamines, as it currently is to self-manage milder forms of depression and anxiety using St John’s Wort/Hypericum [12].
But self-management – while probably useful – is insufficient. We also need to think in terms of optimizing future therapeutic discoveries. Drug development, research, marketing and therapy are all subject to perverse structural incentives which – in the medium-term – require substantial re-orientation. The magnitude of the problem facing policy-makers can be encapsulated by the observation that – even if (hypothetically) it happened it be the proven case that sedating antihistamines were in every way superior drugs to atypicals – under current conditions there is no effective professional motivation in the system of psychiatry to stimulate the huge shift in clinical practice and clinical reasoning entailed by their adoption. In particular, with present incentives, it is extremely unlikely that there will ever be a significant body of randomized trials of – say atypical neuroleptics versus sedating antihistamines in calming agitation, the treatment of acute psychosis, or for the prevention of manic or schizophrenic breakdown. And without such a mass of trials (and the marketing power to push them), then a change to safer sedatives is very unlikely.
In policy terms, there need to be new incentives whereby it becomes possible to make money from marketing new indications for old, off-patent drugs and technologies. These incentives need to be financial if they are to mobilize sufficient resources to change clinical practice. It will probably be necessary to consider introducing new forms of short-term licensing and re-patenting of old drugs, with similar arrangements for clinical procedures and other medical technologies [13]. This is uncharted territory – various ideas will need to be proposed, critiqued and trialed. But there should be a sense of urgency about tackling the problem. We are all patients, potential patients and relatives of patients; so it is in everyone’s interests that drug development and research should continue, and that the incentives which shape it should be orientated towards optimizing therapeutic benefits.
References
[1] D. Healy, The creation of psychopharmacology, Harvard University Press, Cambridge, MA, USA (2002).
[2] R. Whitaker, Mad in America, Perseus Publishing, Cambridge, MA, USA (2002).
[3] R. Whitaker, The case against antipsychotic drugs: a 50 year record of doing more harm than good, Med Hypotheses 62 (2004), pp. 5–13. SummaryPlus | Full Text + Links | PDF (189 K) | View Record in Scopus | Cited By in Scopus
[4] B. Charlton, Psychiatry and the human condition, Radcliffe Medical Press, Oxford, UK (2000).
[5] D. Healy, Psychiatric drugs explained (3rd ed.), Churchill Livingstone, Edinburgh (2002).
[6] M.G. Morgan, P.J. Scully, H.A. Youssef, A. Kinsella, J.M. Owens and J.L. Waddington, Prospective analysis of premature mortality in schizophrenia in relation to health service engagement: a 7.5 year study within an epidemiologically complete, homogenous population in rural Ireland, Psychiat Res 117 (2003), pp. 127–135. SummaryPlus | Full Text + Links | PDF (98 K) | View Record in Scopus | Cited By in Scopus
[7] C. Montout, F. Casadebaig, R. Lagnaoui, H. Verdoux, A. Phillipe and B. Begaud et al., Neuroleptics and mortality in schizophrenia: a prospective analysis of deaths in a French cohort of schizophrenic patients, Schizophr Res (2002), pp. 147–156. Abstract | Abstract + References | PDF (113 K) | View Record in Scopus | Cited By in Scopus
[8] J. Moncrieff and D. Cohen, Rethinking models of psychotropic drug action, Psychother Psychosom 74 (2005), pp. 145–153. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus
[9] D. Healy, Let them eat Prozac, New York University Press, NY, USA (2004).
[10] B.G. Charlton, Fundamental deficiencies in the megatrial methodology, Curr Control Trials Cardiovas Med 2 (2001), pp. 2–7. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus
[11] B.G. Charlton, P.R.A. Taylor and S.J. Proctor, The PACE (population-adjusted clinical epidemiology) strategy: a new approach to multi-centred clinical research, QJM 90 (1997), pp. 147–151. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus
[12] B.G. Charlton, Self-management of psychiatric symptoms using over-the-counter (OTC) psychopharmacology: The S-DTM therapeutic model – self-diagnosis, self-treatment, self-monitoring, Med. Hypotheses 65 (2005), pp. 823–828. SummaryPlus | Full Text + Links | PDF (96 K) | View Record in Scopus | Cited By in Scopus
[13] B.G. Charlton and P. Andras, The future of ‘pure’ medical science: the need for a new specialist professional research system, Med Hypotheses 65 (2005), pp. 419–425.
