1990 Notes to a History

http://jop.sagepub.comPsychopharmacology

Journal of

DOI: 10.1177/026988119000400310 1990; 4; 152 J Psychopharmacol

David T. Healy The psychopharmacological era: notes toward a history

http://jop.sagepub.com/cgi/content/abstract/4/3/152 The online version of this article can be found at:

Published by:

http://www.sagepublications.com

On behalf of:

British Association for Psychopharmacology

can be found at:Journal of Psychopharmacology Additional services and information for

http://jop.sagepub.com/cgi/alerts Email Alerts:

http://jop.sagepub.com/subscriptions Subscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

http://jop.sagepub.com/cgi/content/refs/4/3/152SAGE Journals Online and HighWire Press platforms):

(this article cites 62 articles hosted on the Citations

1990 British Association for Psychopharmacology. All rights reserved. Not for commercial use or unauthorized distribution. at The John Rylands University Library, The University of Manchester on September 3, 2007 http://jop.sagepub.comDownloaded from

152

The psychopharmacological era: notes toward ahistoryDavid T. HealyAcademic Subdepartment of Psychological Medicine, North Wales Hospital, Denbigh, Clwyd LL16, 5SS, UK

SUMMARY Cultural and economic influences on the psychopharmacological era are reviewed,in an attempt to bring into focus what has been happening in psychopharmacology for thepast thirty years. It is argued that the belief that clinical advances are made through the heroicachievements of disinterested scientists is a simplistic view that may militate against futuresignificant discoveries. Such discoveries, it is argued, still come about for the most partserendipitously, despite a widespread belief that psychopharmacology has become a rationalscience.

Introduction

The earliest histories of any branch of science,just as the histories of anything else, tend tofocus on chronologies of who discovered whatand when and the recollections of eminent dis-coverers. This is very much the case in psy-chopharmacology at present, as the recentvolumes on Discoveries in Pharmacology(Parnham and Bruinvels 1983) or Discoveries inBiological Psychiatry (Ayd and Blackwell, 1970)illustrate. Undoubtedly important as this is,especially for younger generations of workersentering the field, such individual heroicaccounts implicitly relegate to roles of minimalimportance other influences on developmentsuch as economic, or cultural factors. In somerespects this is not surprising as medical his-torians have generally contrived to ignore thehistory of the drug business, seemingly in thebelief that it is irrelevant to medical science,even though the production and use of drugs hasincreasingly underpinned medical business sincethe seventeenth century (Liebenau, 1987; Porterand Porter, 1989). Such neglect, however, seemsparticularly misplaced in psychopharmacology,which is a science that defines itself in termsof drug use and many of whose most eminentpractitioners work in or in close association withpharmaceutical companies. This article seeks tooffer some notes toward a future history, thatwill surely locate the dynamic of advance in

psychopharmacology other than solely with theinsights and oversights of individual scientists.

Science, business and drug development

One of the principal consequences of the intro-duction of antidepressants and neuroleptics wasthe development of the monoamine hypothesesof depression and the dopaminergic hypothesisof schizophrenia. However, far from thesehypotheses being an unambiguous advance inthe scientific understanding of mental illness, Ihave argued elsewhere (Healy, 1987a) that themonoamine hypotheses in particular were quitesimplistic; that they accounted for less of theclinical data and were as unscientific as the psy-chodynamic hypotheses before them, in that theyhave been in practice, incapable of disproof.They did, however, legitimize research on theneural substrates of behaviour in a way thatearlier hypotheses did not. They also providedopportunities for research, the benefits of which,as regards furthering our knowledge of brainprocesses and in bringing academic honours andresearch funding to contemporary neurosci-entists and psychiatrists are quite clear. But thebenefit in terms of increasing our knowledge ofclinical disorders is much less clear.From the start, drug company research and

promotional activity has had a dialecticalrelationship with academic interests. Initially, it


153

was the clinical use of reserpine and the labora-tory discovery that it depleted amine neuro-transmitters that led to the finding thatantidepressants blocked the effects of reserpineon mice (see Sulser and Mishra, 1983). Thisblockade and the notion that antidepressantsmust in some way increase brain catecholaminesbecame established in company laboratories as amethod for screening for antidepressants. Whilenot wishing to deny that some simple screeningmethod was needed to sift through the vast num-ber of compounds which have been synthesized,the disadvantages of this approach only becameevident with the essentially serendipitous dis-coveries of iprindole and mianserin. It thenbecame clear that the reserpine model and toostrict an adherence to catecholamine theorieswas leading to the development of modified tri-cyclics and MA01s, rather than to the discoveryof new agents, which might in turn lead todevelopment of or replacement of the orthodoxacademic hypotheses.The downgrading of the reserpine model and

the replacement of the initial formulations of themonoamine theories by amine receptor hypoth-eses, however, has not led to any change in drugcompany strategy. In the case of drugs developedduring the past decade, it is highly likely thatmany potential antidepressants have been missedbecause they did not bring about beta-receptordownregulation as Sulsers 1978 hypothesis sug-gested they should (Sulser and Mishra, 1983). Italso seems quite likely at the moment that drugsfailing to bring about potentiation of 5-hydroxy-tryptamine (5-HT) neurotransmission in linewith the proposal of de Montigny (de Montigny,Chaput and Blier, 1988) will not be promoted asvigorously as they otherwise might have been, inthe light of the current fashion for sensitizing 5-HT receptors.

This strategy of following academic dictatespersists despite the evidence that academicfashions are just that - fashions. Views areadopted and promoted vigorously with the claimthat they account for all the evidence, which theymay well appear to do at the time of initialpromulgation and for a honeymoon period after-wards. This happens inevitably as contrary evi-dence which may have existed previously wouldnot have been reported in the absence of anyreason to report it. For example, no one wouldhave reported a lack of correlation between CSF5-HIAA and suicidal behaviour until after claims

had been made that such correlations exist. Fur-thermore, new contrary evidence often receivesa poor reception and sometimes initial rejection,especially from the most prestigious journals.

Increasingly, however, this strategy, whichgoes under the name of a rational strategy ofdrug development, is leading us away from themethods that gave us the original antidepressantsiproniazid and imipramine - that is good clinicalobservation. Perhaps it is thought that the oppor-tunities for such pioneering/primitive researchdo not now exist. Yet it appears that millions ofpeople worldwide have been on calcium entryblockers for years, which have been enteringtheir brains and bringing about mental statechanges unnoticed by clinicians. Nothing wasdetected it seems until it was fortuitiously notedby Barrow and Childs in 1986, that verapamilappears to be beneficial for tardive dyskinesia.This discovery was delayed, despite academicawareness going back more than a decade thatdrugs active on calcium fluxes should bring aboutaltered brain functioning (Carlsson - discussion).This state of affairs is not confined to calciumentry blockers. There are some indications thatthe angiotensin-converting enzyme (ACE)inhibiting antihypertensives (Whalley, 1989) andaspirin (see discussion) have also been enteringthe brain for years and bringing about mentalstate changes.

Arguably part of the reason for this neglect isthat no investment is being put into fosteringclinical observation of the type that discoveredthat antitubercular agents could also be anti-depressant or that some anaesthetics were ata-ractic. Some reasons why this should be the caseare developed in the remainder of this article.

Science, business and drug marketing

One of these reasons stems from drug marketingstrategies. A common feature of the accounts ofhow individual clinicians discover new drugs isthe emphasis placed on how they not only recog-nized the value of the agent but at the same timethey recognized a condition it might treat (Kuhn,1990; Sandler, 1990). It is this act of makingvisible the invisible that is the truly creative act.But making visible the invisible could also welldescribe the process of niche marketing as cur-rently practised by drug companies.


154

Neuroleptic antidepressantsIn recent years, flupenthixol has been widelymarketed by Lundbeck as both a neuroleptic andan antidepressant. The basis for this marketingstrategy stems from a number of uncontrolledstudies reported from Denmark in the late 1960s(Sonne, 1966; Reiter, 1969), in which depressedsubjects showed a response to flupenthixol, thatwas comparable to that shown by other anti-depressants, that is 60-70% responded. It wasnoted that this response was brought about bylow doses of flupenthixol - lower than doses thenin use for the management of schizophrenia.

Further support for this contention has comefrom a number of double-blind studies of flu-penthixol versus placebo (Predescu et al., 1973;Frolund, 1974; Ovhed, 1976) and versus nor-triptyline (Rosenberg, Ostensen and Fonnelo,1976; Johnson, 1979), amitriptyline (Young,Hughes and Lader, 1976) and phenelzine(Razak, 1980). In all these cases flupenthixolproved superior to placebo and at least equiv-alent to the antidepressants.Another pointer to the antidepressant effect

of flupenthixol that is cited is that when usedin schizophrenia it seems less likely to causedepression than fluphenazine (Carney andSheffield, 1975; Johnson and Malik, 1975). Onbalance, however, it would appear that thedepression induced by neuroleptics in schizo-phrenia is more an expression of drug-inducedtoxicity (demotivation and akinesia) rather thana true endogenous mood disorder (Bartels andDrake, 1988). Therefore the decreased like-lihood that flupenthixol causes such a syndromeis not strictly speaking evidence that it is anantidepressant.While accepting that there is sufficient evi-

dence to warrant further investigation of theantidepressant properties of flupenthixol, Rob-ertson and Trimble (1982) note several problemsin the way of accepting that this agent is anantidepressant in the conventional sense. In gen-eral, the studies cited above were conductedon patients with mixed anxiety-depressive statesrather than clearcut cases of endogenomorphicdepression. Second, the effect achieved has beenclearly present from the first week of treatment.For example, in Johnsons study both flu-penthixol and diazepam were as effective as nor-triptyline in the longer term (29 days) and moreeffective in the short term - but few would con-sider diazepam an antidepressant. In Reiters

study the effect of flupenthixol was most com-monly present within 24 h (Reiter, 1969). If notpresent by the end of the first week, he thoughtthat treatment could be discontinued, as the like-lihood of response was then minimal.Such rapidity of onset is characteristic of a

neuroleptic but not of an antidepressant.Whether one has schizophrenia, depression or issimply a control subject, all neuroleptics producean identifiable state of indifference within a fewhours of having had them (Kalinowsky andHoch, 1961; Swazey, 1974; Healy, 1989). Agi-tation is the target clinical feature across clinicalcategories (Baldessarini, 1980). The fact thatschizophrenic illnesses typically do not sub-stantially improve for several weeks after therapyhas been instituted has tended to obscure thisfact (Healy, 1990a, 1990b).Another problem emerges from Robertson

and Trimbles 1982 review. These authors alsoassessed the antidepressant potential of otherneuroleptics based on clinical reports. It wouldappear that all other neuroleptics possess com-parable antidepressant properties to those offlupenthixol. At least as convincing a case can bemade for thioridazine, chlorpromazine, per-phenazine, thiothixene and sulpiride. In the caseof all of these it would seem that their clinicalbenefits are apparent within the first few days oftreatment and that they are best in groups ofmixed anxiety-depressive states. Similar findingshad also been reported, as early as 1955, for thevery first neuroleptic - reserpine (Davies andShepherd, 1955). (To a historian the subsequentlack of citation of this Davies and Shepherdarticle, despite the eminence of its authors andits reporting of a controlled clinical trial, in con-trast to the widely reported uncontrolled obser-vations of reserpines propensity to causedepression is interesting - see Healy, 1987a).

Recent basic research on dopamine receptorsmay offer a rationale for an activating effect oflow dose neuroleptics. There is some evidencethat dopamine autoreceptors are more sensitiveto the effects of dopamine blocking agents thanare postsynaptic receptors (Leonard, 1984).Therefore low doses of a neuroleptic might beexpected to activate dopamine systems ratherthan inhibit neurotransmission, in much the waythat L-dopa does. It can be noted that L-dopahas been reported to have a similar activatingeffect in some cases of depression, but it wouldappear that this is clearly not an antidepressant


155

effect (Zis and Goodwin, 1982). While all neuro-leptics can be expected to do this, some of them,like chlorpromazine and thioridazine also sedateby virtue of an action on alpha, receptors. Incontrast the activating effects of flupenthixol arenot obscured by sedative side-effects.A further basis for a difference between flu-

penthixol and other neuroleptics lies in its dif-ferential binding to dopaminergic receptors.Most neuroleptics bind predominantly to D2receptors. Recent neuroleptics have been desig-ned to bind as specifically as possible to thesereceptors as clinical efficacy appears to correlatewith the ability to bind to D2 receptors (Wad-dington, 1989). However, thioxanthenes such asflupenthixol also bind significantly to D 1 recep-tors (Waddington and OBoyle 1989). Unfor-tunately, however, recent PET scan studiessuggest that whatever the in vitro ability of flu-penthixol to bind to D1 receptors, in vivo thereis negligible binding (Waddington, 1989).Whether or not any of these neurobiological

possibilities grounds an antidepressant action forflupenthixol, it remains clear that the marketingof flupenthixol as an antidepressant preceded thedevelopment of any neurobiological rationale forsuch marketing. The question, therefore, iswhether the primary difference between flu-penthixol and other neuroleptics lies in mar-keting strategy rather than in anything else? Amarket strategy that derives legitimacy from theability of flupenthixol to lower Hamilton ratingscores in a substantial number of depressed sub-jects.

This lowering of Hamilton rating scores with-out taking into account other aspects of the clini-cal picture is the critical issue. Max Hamiltonhimself did not see his scale as an instrumentfor measuring the severity of or changes in adepressive illness (Hamilton, 1967). Rather hesaw it initially as a checklist of questions clin-icians should be asking and observations theyshould be making. A great number of thesequestions and observations concern anxiety. Aconsequence of this is that any treatments orprocedures which reduce anxiety may bringabout relatively large and relatively rapidchanges in overall scores on the Hamilton scale.Despite this - or perhaps because of it - this scalehas become the supreme instrument of anti-depressant clinical trials and the supposedlyobjective results that stem from its use are givengreater weight than clinical judgements unbol-stered by rating scale support.

It would appear that a version of the flu-penthixol story is being replayed at the presenttime with the marketing of amoxapine in theUnited Kingdom. Amoxapine is being marketedas an antidepressant, with a particularly rapidonset of action. Again there is clinical trial evi-dence in the shape of reductions in Hamiltonrating scores to support this early onset of action(McNair, Rizley and Kahn, 1986). What is notbeing marketed is its ability to cause a neu-roleptic malignant syndrome and akathisia(Cocarro and Siever, 1985). An alternative storymight run that there was a drug company whohad two very similar molecules, loxapine andamoxapine, who thought it might be a good ideato target each of them at a different clinicalpopulation.The issue of whether drugs like flupenthixol

or amoxapine are antidepressant in the samesense as the canonical antidepressants,imipramine, amitriptyline and phenelzine, is onethat is of importance to more than general prac-titioners who might be concerned to avoid caus-ing tardive dyskinesia. It is also critical for basicresearchers, who try to develop animal modelsof depression and who are trying to elucidate themechanisms of action of antidepressants. Ref-erence to widely cited and apparently valid clini-cal trials will give them the impression that theyneed to produce behavioural tests and responsesor neurobiological changes, sensitive to both theinfluence of imipramine and phenelzine as wellas flupenthixol and amoxapine, but not to halo-peridol or chlorpromazine. Ambiguity on theissue of which drugs are and which are not anti-depressants, therefore, could potentially set backprogress.

5-HT Uptake inhibition and obsessive-compulsive disorderA further marketing strategy has seen the tar-geting of 5-HT uptake inhibitors for obsessivecompulsive disorders (OCD). In the early 1960s,in an attempt to get a stronger antidepressanteffect from imipramine, the basic molecule waschlorinated. The resulting compound appearedto have antidepressant activity (Brandner, 1964;Symes, 1967) but to be no more potent thanimipramine (Symes, 1967). This left Ciba-Geigywith three tricyclic antidepressants on theirbooks - imipramine, amitriptyline and chlo-rimipramine. On the basis that all three weremuch the same and in view of clomipramines less


156

favourable side-effect profile, the FDA resistedlicensing it in the United States.

Shortly afterwards a number of reportsappeared in which a possible beneficial effect ofclomipramine in obsessional states was suggested(Cordoba and Lopez-Ibor, 1967; Guyotat,Favre-Tissot and Marie-Gardine, 1968). Theseled to several open studies of clomipramine inanxiety states (Capstick, 1971, 1973; Rack; 1973;Marshall and Micev, 1973; Walter, 1973;Waxman, 1973). The results of these studies inseveral different centres appeared to support thenotion that clomipramine is useful in obsessionalstates. They further suggested that it was usefulin obsessional states that did not appear to besecondary to depression - in other words that thedrug seemed to have a primary anti-obsessionaleffect. It should be noted, however, that severalof these studies also indicated that clomipramineappeared even more useful for phobic thanobsessional states (Marshall and Micev, 1973).Furthermore, none of these studies were con-trolled by reference to another medication orto placebo and all involved small numbers ofpatients. Nevertheless by 1975, Ciba-Geigy wasactively promoting clomipramine for obsessive-compulsive disorder.However, in a later placebo-controlled study

of clomipramine and exposure therapy forobsessive-compulsive disorder, Marks et al.(1980) concluded that clomipramine appearedto act more as an antidepressant than as ananticompulsive agent. When depression wasminimal, they found clomipramine to be of noappreciable value, whereas exposure therapyspecifically affected rituals without appreciableeffects on mood. In 1988, Marks et al. replicatedtheir 1980 study and concluded that clom-ipramine had a limited adjuvant role to play inthe treatment of OCD. Reviewing the literature,they concluded that there is no evidence thatclomipramine is better than other tricyclic drugsin OCD. The impression that there is some speci-ficity of clomipramine to OCD they attributed toits being the drug that has been studied mostwidely for these conditions, with a dearth ofcomparisons between it and other antidepres-sants.These conclusions provoked a reply from Katz

et al. of Ciba-Geigy (1988), who argued thatsignificant effects could be demonstrated in OCDpatients on clomipramine, even in the Marks etal. (1988) study. To this Marks and Basoglu

(1989) replied that the demonstration of an effectthat reaches statistical significance may yield verylittle information, and may in actual fact not beof significant clinical benefit to patients. Theyargued that a more discriminating approach andstudies to examine the long-term effects of psy-chotropic medications were needed.Has the OCD story then been a matter of

serendipitous marketing rather than the dis-covery of something different about the patho-physiology of this disorder? In favour of theformer explanation is the fact that Ciba-Geigyhad two other tricyclic antidepressants on themarket prior to clomipramine and would havefound it difficult to market another straight-down-the-middle agent. So also is the fact thatclomipramine appears to be broadly anxiolyticrather than specifically anticompulsive (Marshalland Micev, 1973; Marks and OSullivan, 1988)and the continuing paucity of conclusive evi-dence in favour of any specificity of 5-HT uptakeinhibitors to OCD. Nevertheless clomipramineis sold with a specific indication for OCD, asis the recently introduced specific 5-HT uptakeinhibitor fluvoxamine. The case of fluvoxamineis particularly interesting as, at the time of mar-keting, there was no evidence in favour of anyspecific efficacy it might have for OCD. Itappears, therefore, to be benefiting from theapparent establishment of the notion that 5-HTuptake inhibitors do something that no otherdrugs do for OCD.

Against this is some recent evidence that thenewly developed specific 5-HT uptake inhibitorsdo appear to have some - as yet incompletelycharacterized - beneficial effect in OCD(Goodman et al., 1989). There is also some evi-dence from Zohar et al. (1989) that serotonergicagonists such as trazodone and its derivativeMCPP in some way worsen symptoms of OCD.There is also the further intuition of Arvid

Carlsson and others (Carlsson, 1982) that tri-cyclics with a tertiary amine structure, such asimipramine, amitriptyline and clomipramine,differ phenomenologically in some importantway from their deaminated derivatives desi-pramine and nortriptyline. Far from the obviousconclusion that these derivatives must be thereal antidepressants, Carlsson and others had thehunch that the parent compounds were doingsomething else that was important. Biochemi-cally this difference appears to centre on theability of the original tricyclics to inhibit 5-HT


157

uptake, which their deaminated metabolites donot do to any significant extent (Carlsson, 1982).On the basis of this phenomenologically basedintuition, efforts were begun to synthesizespecific 5-HT uptake inhibitors, with some suc-cess it would appear. This phenomenologicaldifference may be pertinent to OCD as thereappears to be some evidence that imipraminemay be useful in OCD (Cottraux, 1988), whereasthere appears to be little evidence in favour ofdesipramine (Marks and OSullivan, 1988).There is a further piece of phenomenological

evidence that would appear worth pursuing. Inthe midst of recent investigations, the initialimpetus to the use of clomipramine in OCDappears to have been forgotten. This came fromreports from patients of a beneficial effect - asense that rituals and obsessions became lesscompelling (Cordoba and Lopez-Ibor, 1967).This lead has not been pursued since, despitean increasing awareness that it is probably notpossible to run a blind clinical trial of 5-HTuptake inhibitors in OCD as their effects on thepatient are quite distinctive (Marks et al., 1988).Do clomipramine, the older antidepressants orthe newer 5-HT uptake inhibitors have com-parable phenomenological effects? Is there anyaffinity between the effects induced by clo-mipramine in OCD and the ataractic effect(induction of psychic indifference - see Healy,1989) brought about by neuroleptics?Perhaps of note here is the fact that both

neuroleptics and clomipramine bring aboutincreases in plasma prolactin, whereas 5-HTuptake inhibitors in general do not (Tuomistoand Mannisto, 1985). Of related interest is thefact that there have been no controlled studiesof the use of neuroleptics in OCD (Marks andOSullivan, 1988) despite clinical impressionsgoing back to 1954 that neuroleptics helped inobsessive states (Kline, 1954). These phenom-enological issues appear to have got buriedbeneath an avalanche of rating scales and sub-scales, from which it is difficult for an outsiderto get any feel for whether the treatments inquestion make any real difference to patients.Reasons for the neglect of phenomenologicalissues in psychopharmacology will be exploredfurther in the final section of this article.

MAOIs and atypical depressionWith the introduction of amitriptyline in 1961and the discovery of the cheese effect associated

with MAOIs (Blackwell, 1970), drug companieswith MAOI antidepressants were faced with amarketing problem. This became particularlyacute after the MRC trial of 1965 in which ECT,imipramine, phenelzine and placebo were com-pared and phenelzine turned out to be no moreeffective than placebo.There were two possible responses to this

situation for both clinicians and drug companies.One was a rejection of the MRC findings; sub-sequent research suggests that this would havebeen the best option (Pare, 1985). The otherpossibility was to accept the findings but with thecaveat that as these drugs were known to beclinically useful, they must therefore be effectivefor something else other than straightforwarddepression. There was a ready something elseto hand in the concept of atypical depressivedisorders, first outlined by West and Dally(1959).Over the next 15 Years MAOIs were marketed

for atypical depressions of all sorts, whether Pol-litt and Youngs (1971) reversed functional shiftdepressions or the broader concept of non-endogenous depressions, or for anxiety and pho-bic disorders uncomplicated by personality dif-ficulties (Kelly et al., 1970; see also Paykel et al.,1983) or alternatively for hysteroid dysphoricpatients (Liebowitz and Klein, 1979). Inter-estingly it would appear that there has been adrift over the past two decades from seeing theMAOIs as useful for phobic patients with goodpremorbid personalities to the opposite end ofthe spectrum with the recent claim of a usefulnessfor borderline personality disorder (Cowdry andGardner, 1988).

Allied to this, the impression developed in thelate 1960S and early 1970s that tricyclic anti-depressants were only effective for the endo-genomorphic form of the illness and thatMAOIs were only effective for the non-endo-genomorphic forms of depression. However,there are now nine controlled trials of MAOIsin which the dose used was 60-90 mg/day phenel-zine, or the equivalent of other MA01s, and theresults suggest that MAOIs are as effective astricyclics for endogenomorphic forms of the dis-order (Pare, 1985). Conversely while it is unques-tionably the case as Roland Kuhn has argued,that tricyclic antidepressants are most clearlyefficacious in the classical endogenomorphicform of depression (Kuhn, 1970), they haveincreasingly been shown to be effective in a wide


158

variety of depressive states characterized byprominent anxiety (Paykel, 1989). Furthermore,there is the fact that both the tricyclics andMAOIs bind at least weakly to a multiplicity ofaminergic receptors (Healy, 1987b). Given thisit would be surprising if a number of anti-depressants did not have some independentanxiolytic action.Thus, while MAOIs and tricyclics differ some-

what in that some patients respond to one ratherthan the other (Pare, 1985) and the MAOIsappear more likely to have a transient psy-chostimulant effect, there nevertheless appearsto be little of substance in the difference betweenthem, other than their pasts and the impressionsthat have been created about what they do. Asnone of the various atypical affective disordershave ever been rigorously established as clearcutpsychopathological entities in their own rightresponsive to specific treatments (Paykel et al.,1983) and as biochemically there is little of sub-stance in the difference between MAOIs andtricyclic antidepressants, how have theseimpressions been created?

In the absence of a solid basis in empiricalobservation it can be suggested that theseimpressions have come about in part by virtue ofa Matthew effect (Merton, 1968, 1969). Thiseffect refers to the increased likelihood of factorssuch as the prior fame of one of the authorsaffecting the reception of the ideas in question,or the likelihood of influential journals acceptingthe piece of work. In other words, as theMatthew Gospel put it in the parable of thetalents, to him who has more shall be given.Many eminent names have been associated withvarious forms of atypical depression. This per-haps has unduly influenced the reception of theseideas.But as Paykel et al. (1983) have noted, the

various concepts have also been defined in largepart around a supposed responsiveness toMA01s. This focusing of issues on a response toa particular group of drugs was tailormade forthe promotional purposes of those drug com-panies who had MAOIs on their hands and werefacing a shrinking market. This in part has prob-ably helped produce what may be termed a Lukeeffect. Ideally the spread of ideas in science issupposedly one that is determined by the intrinsicquality of the ideas. But today in all branchesof the biomedical enterprise, drug companiesdisseminate large amounts of scientific literature.

It is probable that literature from such sourcesmakes up a significant proportion of what is readby many clinicians. (It made up a significantproportion of the reference list of this article.)Not unreasonably the material that is passed onrelates favourably to company market concerns.In this manner many concepts that might other-wise be retired early to inhabit the backshelvesof libraries may these days be given an extendedlease of life. Put another way, drug companiesobviously make drugs but less obviously theymake views of illnesses by selectively reinforcingcertain possible views.The notion of a Luke effect is taken from the

parable of Luke the physician about the sowersowing the seed, some of which fell on stonyground, some of which fell on fertile ground butgrowing up was choked by weeds. This parableends with an exhortation to those who have earsto listen. Listening is arguably what drug com-panies have been doing very successfully wherethe MA01s, flupenthixol and clomipramine areconcerned, as well as for alprazolam and calciumentry blockers as outlined below.

In marked contrast a radically different agent,lithium, has not been promoted in anything likethe same way. It has been argued that this isbecause there has been little money to be madefrom it (Amdisen, 1984; Johnson, 1984). As aconsequence lithium therapy has taken a longtime to get established and even now is a poorcousin to the other psychotropics.Alprazolam and panic-disorderThe recent study of alprazolam in panic disorderprovides perhaps one of the best possibleexamples of a Luke effect. When Upjohn pro-duced a new triazolo-benzodiazepine in the early1980s, they had a marketing problem as benzo-diazepines were becoming increasingly unpopu-lar. One response was to market the newcompound as an antidepressant. As the case offlupenthixol shows it is relatively easy to produceantidepressant credentials in the shape ofreduction of Hamilton rating scale scores - andthis has been done for alprazolam (Rickels,Feighner and Smith, 1985) - but also for otherbenzodiazepines (Tiller et al., 1989). This option,however, has been neglected almost certainlybecause of the sudden opportunity provided bythe creation of panic-disorder in 1980.

Panic-disorder only came into existence for-mally in 1980 with the publication of DSM 111,


159

in which it was controversially separated fromagoraphobia as a diagnostic entity (Klerman etal., 1989). This led, to the mutual benefit of theproponents of panic-disorder and the makersof alprazolam, to a large multicentre placebo-controlled study of alprazolam in the treatmentof agoraphobia and panic-disorder starting in1983 (Klerman et al., 1989). In 1988 the resultsof this study were published, claiming that alp-razolam was effective for both agoraphobia andpanic-disorder. However, it has been argued byMarks et al. (1989) that this claim is misleading;that one can legitimately read the results as indi-cating that there was no significant effect of alp-razolam. Alternatively, even if one reads theresults in the manner favoured by the principalinvestigators, the effects were minimal, of briefduration and liable to lead to a worsening of theclinical state of the patient on discontinuation oftreatment. As the results were presentedhowever, these latter issues are glossed over andthe casual reader is liable to be left with theimpression that alprazolam is a specific therapyfor both panic-disorders and agoraphobia, andthat panic-disorder is an autonomous biologicalentity. Quite apart from any effectiveness ofalprazolam for panic-disorder or any autonomyof panic disorder as an entity, the focus of thisarticle is on Upjohns investment in panic-disorder. So great has this been that panic-disorder was commonly referred to in the mid-1980s as Upjohn illness.

Making visible the invisible - paradoxicalserendipity

The history of varying associations between theMAOIs and a range of exotic clinical syndromes,as well as the targeting of neuroleptics for depres-sion and 5-HT uptake inhibitors for OCD, pointsto the existence of a gap between the patho-genetic bases of affective disorders and thepathoplastic effects that they may give rise to,and gap between depression and the anxiety itmay give rise to. There is a comparable gapbetween the pathogenetic basis of schizophreniaand the delusional and neurotic behaviours itmay stimulate (Healy, 1990b). Through thesegaps, propelled by their respective marketforces, academic investigators and drug com-panies have all too quickly attempted to march.The tramp of their passing feet has for the most

part, it would seem, only served to deepen psy-chopathologys classificatory quagmire and toblurr the outlines of what Nathan Kline (Sandler,1990), Roland Kuhn (1990) and Henri Laborithad first made visible.

But, just as Columbus sailed in search of theIndies and profits but found instead somethingunsuspected, so also it can be argued that thesiren call of the market has also led, or is leadingto a revolution in the perception of mentalillness. Early work on the epidemiology of men-tal illness revealed that affective disorders arevery widespread in the community (Shepherd etal., 1966). This can be taken to support eitherthe view that the greater part of the affectivedisorders are mild and self-limiting or that thereare two forms of the illness: one a mild psy-chological problem, and the other a severe medi-cal illness only correctable by physicaltreatments. While the latter view was possiblythe conclusion most commonly drawn from earlycommunity surveys, a significant new element inthe current debate on mental disorders in thecommunity has come from the use of anti-depressants by general practitioners (Fahy,1989).Where earlier surveys looked at the degree of

misery and anxiety in community samples, morerecent work has had to address itself to thequestion of who is getting antidepressants fromtheir general practitioners and what effects thesedrugs are having. The reply would appear to bethat many relatively mild disorders, that mightotherwise have been taken to be psychologicalproblems (distress rather than disease), respondto antidepressants (Sireling, et al., 1985; Blackerand Clare, 1987). One possible interpretation ofthis is that biological depression is a mild illnessfor the most part and that those who end upbeing hospitalized for the disorder are an unrep-resentative minority.

Furthermore, there would seem to be a gooddeal of evidence that many subjects with verysimilar clinical conditions get well spontaneouslywithout antidepressants and in relatively shortperiods of time - within a matter of weeks(Blacker and Clare, 1987). If this is the case,the therapeutic agenda at least for the affectivedisorders would shift from one dominated by thequestion of how do we get depressives well toone of why do a few patients not get well spon-taneously. This is a far more optimistic thera-peutic agenda than the one that facedpsychiatrists in the early 1950s.


160

Similarly, notwithstanding the evidence thatcan be marshalled in favour of the argument thatthe emptying of the asylums did not dependon the introduction of these agents (Shepherd,1990) or the evidence in favour of the argumentthat neuroleptics are not directly antischizo-phrenic (Healy, 1990a), the advent of the neu-roleptics has altered the perception ofschizophrenia. Thus while the promotionalactivities of drug companies may sometimesobscure the barely visible outlines of a rationalpsychopathology, so also the aggressive mar-keting of psychotropics is leading just as muchas the insights of any detached scientists to unex-pected discoveries - and just as serendipitously.

Science, business and history

It was noted in the introduction that historiesof the psychopharmacological era focus on theachievements of eminent discoverers and in gen-eral neglect other factors. Such a focus is typicalof the earliest approaches to the history of abranch of medicine. In general such approachesgive way to increasing interest in medical politicsand in the politics surrounding medicine (Porterand Porter, 1989). But even in the more devel-oped branches of medical history, there appearsto be a blind spot for the economics or businessof medicine (Liebenau, 1987; Porter and Porter,1989). Such a focus would seem indicated asmodern medicine unlike medieval medicine hasfrom about 1600 onwards involved the sale of arapidly increasing number of drugs alongside thetraditional sale of skills. This latter developmentgave rise to the pharmaceutical industry, to con-cern about the monstrous profits of druggists andto an increasing consumption of medicaments(Liebenau, 1987; Porter and Porter, 1989), allof which remain themes that characterize thepsychopharmacological era.Some further examples to add to those above

may indicate the need to take the policies ofpharmaceutical companies into account whenwriting any history of this era. As was notedabove there has been widespread use in recentyears of calcium entry blockers but a seemingblindness to any mental state changes they mayhave been bringing about. While the currentbias against using the observations of individualclinicians or patients as a basis for therapeuticdevelopment is undoubtedly one reason for this,

other factors can be cited. One such is that drugcompanies marketing calcium entry blockers suc-cessfully for angina have effectively had an inter-est in denying that their drugs crossed the blood-brain barrier. This at least was the response to acolleague of mine interested to pursue the ques-tion of calcium entry blockers for tardive dys-kinesia (Dinan, personal communication, Dinanand Capstick, 1989). The underlying thinkingappears to be one of not risking a certain marketby opening to clinical scrutiny an area that mightas quickly lead to curtailments on the marketingof the product as it would to further applications.

Following the observations of Barrow andChilds (1986) that calcium entry blockers seemedto be of benefit in tardive dyskinesia, it wouldappear that there has been an explosion of inter-est in possible therapeutic applications. Is this acase of academics striking off shackles put onthem by business? Against this possibility, afurther factor can be set - the development oflaser surgery for coronary artery disease. Thisdevelopment potentially could substantiallyreduce the market for calcium entry blockers,which at present have angina as their primaryindication. Could it be therefore that recent clini-cal interest reflects business concerns more thanany other?

Writing a history also involves accounting forwhy certain questions do not get properly addres-sed. For the conventional mythologies of science,the notion that certain questions are being sys-tematically avoided is unbelievable (Healy,1987a) - for example, the use of antidepressantsfor mania and the administration of anti-depressants every few days rather than a fewtimes every day. Any of the clinical trials doneto date of tricyclics in mania have indicated thatthey may have antimanic properties, whichwould be in accord with the fact that both lithiumand ECT are both antimanic and antidepressant(Healy and Williams, 1989). Yet given the weightof popular expectation that far from curingmania, tricyclic antidepressants are liable tocause it, expectations that have possibly survivedthe demise of the original catecholamine hypoth-esis in the way that grins survive cheshire cats,drug companies are reluctant to fund a study ofthe issue (personal observation). The principalreason seems to be a certain nervousness aboutjeopardizing their markets - something com-parable to the thinking, which denied that cal-cium entry blockers enter the brain.


161

There are further studies that drug companiesare not enthusiastic to fund. For example, thereis little evidence that antidepressants act in thesame way as other psychotropic agents - thatis acutely and for a limited duration, therebynecessitating b.d, t.d.s or q.d.s regimes (Bau-mann et al., 1988). Increasingly, companies seemto be moving toward single-dose regimes, orclinicians seem to be giving single-dose regimes,whether or not this is advocated by the company.There are furthermore a number of studies inwhich antidepressants have been given every fewdays, with seemingly as good results (Pollock etal., 1989; Montgomery et al., 1986). Indeed afurther point that can be noted is that owing tothe differences between rat and human metab-olism the antidepressant regimes that laboratoryrats are given to bring about 5-HT receptor sen-sitization or beta-receptor downregulation or toreverse depressive behaviours in animal models,typically produce large-dose pulsatile profilesrather than the more even steady state regimessupposedly being aimed at in man (Baumann etal., 1988). Thus there is some evidence thatpoints toward the possibility that antidepressantsmight work optimally, in the same way as ECT -that is once every few days rather than a fewtimes per day.

Such findings, if upheld, would pose seriousquestions for any theories about the mode ofaction of antidepressants. But there are problemsdoing the research involved in that drug com-panies are not likely to willingly support it, forthe very good reason that the volume of theirsales might be threatened by the results. Thiscould provide at least two barriers to progress:difficulty in getting placebo tablets; difficulty ingetting funding for the research. The outcome ofthis will inevitably be a lack of research in theseareas as it is easier to research issues for whichfunding is readily available - research that isdifficult to fund remains undone by default.To answer these and other questions, a future

historian would need access to company records.Far from some future history being simply achronicle of the perceptions of scientists of whatthey were doing and why, it would also need toinclude a chronicle of marketing and companyresearch decisions. Books such as the recentDiscoveries in Pharmacology (Parnham andBruinvels, 1983) urgently need supplementingwith volumes on the same period as seen fromthe perspective of the pharmaceutical industry.

There needs to be some assessment of how thevarious companies stand as regards scientificissues. Are they genuinely concerned to promotedevelopments or do they simply use the languageof science for marketing purposes? Historically,the major companies have differed in this regard(Liebenau, 1987).

Science, business and politics

There is a further factor whose influence a his-torian might wish to assess. At the start of thepsychopharmacological era, the interactionsbetween industry and academia were relativelyimmediate. Drugs could be synthesized andadministered to patients within a matter ofmonths. Now that interval is more likely to beof the order of 10-15 years. As a consequenceany rational basis there might bave been to themanufacture of the drug in the normal course ofevents is increasingly likely to be superseded bythe time that rationality is put to the test.

This delay has much to do with the operationsof governmental agencies, the relation of whoseinterests to those of the pharmaceutical-aca-demic complex is uncertain. These agencies actboth to delay the clinical use of a drug by increas-ingly detailed premarketing surveillance, and toaffect the reception of a drug through post-mar-keting surveillance. The latter is exerted in partthrough national schemes instituted in the 1960sfor the reporting of adverse drug reactions.To appreciate how such factors have altered

the psychopharmacological landscape, considerthe latter schemes. The 1970s saw the intro-duction of agents such as nomifensine,mianserin, and the 1980s the specific 5-HTuptake inhibitors zimelidine, fluvoxamine andfluoxetine for the treatment of depression. As aconsequence of schemes for the reporting ofadverse drug reactions, the past 5 years haveseen the first removal of psychotropics from themarket. This has almost exclusively affected thenewer agents, with nomifensine and zimelidinebeing the prominent casualties, and mianserinbeing more recently under threat.The basis for withdrawal has been the occur-

rence of fatal adverse reactions, as reported tothe Committee on Safety of Medicines (CSM).From recently released data it would seem thatthese were quite considerable in the case of zime-lidine, far less marked in the case of nomifensine,


162

but not unknown for all other antidepressants(nomifensine was comparable to clomipramine,for example; (see Pinder, 1988; Beaumont,1989). However there are ambiguities in theCSM figures in that the older antidepressantswere introduced in an era before the reportingof adverse reactions received the attention it nowdoes (Girard and Biscos-Garreau, 1989). Thereis now a particular onus on clinicians to reportall adverse reactions to a drug during its first 3years on the market. A further problem withantidepressants is that the newer agents appearto be more likely to be given to populations whomay be particularly at risk for adverse reactions,such as resistant depressives, the elderly, thephysically ill or those extensively treated pre-viously with other drugs (Pinder, 1988).The cases of lofepramine and mianserin illus-

trate the effects of the present operating of theCSM. Lofepramine seems least likely of all thetricyclics to lead to fatalities in overdose. Yetrecently the CSM has drawn the attention ofprescribers to the propensity of tricyclics to raiseliver enzymes, in a manner that appeared tospecifically target lofepramine (CSM update, 231988). This is despite the fact that liver toxicitywas a notable feature of the early use of tri-cyclics - running to 10% of subjects showingincreased SGOT levels on amitriptyline(Holmberg and Janssen, 1962; Klerman andCole, 1965; Davies, 1981; Dukes, 1988).A similar story appears to apply to mianserin.

From being one of the most widely prescribedantidepressants, this agent is now used much lessfrequently, in part almost certainly because ofconcerns about the agranulocytosis it has beenreported to produce. But in a recent review ofleucopenia after antidepressant use, Moller,Meier and Wernicke (1988) found that ami-triptyline was no less likely to produce leu-copenia than mianserin. In a survey of publisheddata from clinical trials, Girard and Biscos-Gar-reau (1989) concluded that the data does notexist to say that mianserin is more haematoxicthan the older tricyclics. In the case of bothmianserin and lofepramine, these drugs are suf-fering from an inbuilt bias in the CSM againstnovel agents. Indeed it can be suggested thatneither amitriptyline or imipramine would getproduct licences if Ciba-Geigy had to apply forthem today. Given that the toxicity programmesthat the more recent antidepressants have togo through prior to launch, it would be most

surprising if they were not at least on balance assafe as the older agents, which were not subjectto such thorough assessment.However, the population of depressed patients

who are at by far the greatest risk of drug-inducedfatalities are those who are suicidal or who takeoverdoses. At present 15% of deaths by fatalpoisoning involve antidepressants (Office ofPopulation Censuses and Surveys (OPCS), 1977-86). Indeed antidepressant overdose is the com-monest life-threatening drug ingestion world-wide (Pinder, 1988). With the current cloud overthe prescription of benzodiazepines for anxietyand the targeting of the treatment of anxiety bytricylic antidepressants now being undertaken bysome drug companies, this figure seems set torise.

In this particular at-risk population, the majorrisk of fatal adverse reactions comes from theolder tricyclic antidepressants rather than thenewer agents. If one adds the fatal adverse reac-tions occurring during therapeutic intake of anti-depressants to those occurring after overdosage,then agents such as nomifensine, mianserin andlofepramine emerge as clearly safer than theolder tricyclics (Pinder, 1988). That is, if allantidepressants were prescribed equally, fewerpeople would die on nomifensin or mianserinthan on imipramine or dothiepin. One mightquestion on this basis which agents should bewithdrawn.

In reply it may be said that it is difficult to takesuicide into account in this kind of calculation.[Whatever about deaths from overdose, anti-depressants should not be agents which are liableto kill subjects during routine use.] It is not clear,however, that this is the thinking of the CSM orthe true grounds on which drugs are removedfrom use. There are three other factors that maybe influencing current events. One is the cost ofnewer drugs, which may be up to 25 times morecostly than the older agents. In terms thereforeof burden on the exchequer, there is a very goodreason why antidepressant prescribing should berestricted to the older agents such asdesipramine, amitriptyline and imipramine. Aseparate reason may be the belief that drug com-panies are being allowed to make too muchmoney - a belief that has been present for overa century (Liebenau, 1987; Porter and Porter,1989), although the current evidence in its favourhas been called into doubt (Cantopher, Edwardsand Olivieri, 1988). Another may be a need to


163

be seen to be responding to consumer pressure.Similar political forces have in the past led to theoutlawing of ECT in some American states andhence the responses to such pressures need care-ful consideration.However, it can also be noted that historically

it would appear that there has always been anexchange of personnel between regulatorybodies and drug companies and that the intro-duction of regulations, if not actively sponsoredby the major drug companies, have alwaysfavoured their interests - although superficiallythis may not have been apparent at the time(Liebenau, 1987). Is this still the case? Someassessment of these issues will be needed if ahistory of the psychopharmacological era is to becomprehensive.An assessment of the impact of regulatory

agencies will also be needed. For example, thepresent approach, which is progressivelyrestricting the agents available to treat mentalillnesses to the older psychotropics will in effectreserpinize the database for laboratory, clinicaland phenomenological studies of psychotropicsand the illnesses they treat. As has been noted,the newer 5-HT reuptake inhibitors were syn-thesized on the basis that it seemed that theymight differ phenomenologically from desi-pramine and nortriptyline (Carlsson, 1982).While it seems clear that these agents are anti-depressants in the sense of reducing HamiltonRating scale scores, it has not yet become clearjust what phenomenological differences thereare between them and the tricyclics. If the opera-tions of the CSM continue as before, we may notbe given the time to find out.

Science, business and culture

There is something seriously missing in a field ofmental illness that does not attend closely and broadlyto patients subjective experiences ... And yet muchof the contemporary scene in disciplines that focus onmental illness reflects this neglect. Driven by varioustheoretical models and the quest for being scientific inonly a narrow sense, clinicians neglect many aspects ofpatients reports, their implications for understandingillness and healing processes and the need to developimproved methods for studying subjective experience... (Strauss and Estroff, 1989).One of the themes that runs through the vari-

ous examples detailed above is the neglect of the

subjective effects of drug ingestion. They areneglected as a means of discovering new psy-chotropic agents as subjective impressions arenot methodically sought. If offered they are alltoo likely to be dismissed as neurotic. I haveargued above that good clinical observation hasa better track record in discovering new drugsthan current rational drug development pro-grammes.But does good clinical observation need to be

the preserve of men such as Roland Kuhn, JeanDelay or Nathan Kline? Cannot the patients whotake these agents also be observers? At presentwe appear to assume that having a mental illnessprecludes an ability to observe empirically(Healy, 1990b). There is no evidence in favourof this presupposition. Indeed it is conceded thatdrug abusers may be very discriminating of theeffects of many of the same medications. Theyappear able to distinguish agents from each otherand different phases in the effects of the oneagent, such as cocaine, and these different phasescan be independently manipulated (Scherer,1988).There is, furthermore, some elegant work by

Philip May and colleagues to show that tailoringneuroleptic regimes according to the subjectiveresponses of the takers produces the best clinicaleffects (May, Van Putten and Yale, 1976). Onemight wonder about our willingness to ignorethe often heard statements of patients that theirparticular neuroleptic regimes have not beenhelping them (Healy, 1990b).

Paying heed to subjective effects might helpto resolve the issue of whether flupenthixol is anantidepressant or a neuroleptic and might clarifythe role of clomipramine in obsessive-compulsivedisorders. It can also be argued that paying heedto the subjective effects of neuroleptics affordsperhaps the most cogent piece of evidenceagainst the dopamine hypothesis of schizo-phrenia (Healy, 1989, 1990a). Briefly, neuro-leptics induce an ataractic effect (an indif-ference) in everyone who has them, whethercontrol or schizophrenic, within hours of takingthem, provided that 60-80% of D2 receptors areblocked by the dose given. This effect parallelsin time course the blockade of D2 receptors theyalso bring about. If this effect is mediatedthrough D2-receptor blockade, its occurrence inboth schizophrenics and controls would point toward a normal functioning of the dopaminergicsystem in schizophrenia.


164

Similarly if 85% or more of D2 receptors areblocked, rapid behavioural control in the formof the experience of internal straitjacketing isachieved - and can be achieved in personalitydisordered patients, manic patients and schizo-phrenics as well as in volunteers. This drug com-panies have always known and have accordinglyfilled many of the advertisements for neu-roleptics with images of aggression being rapidlybrought under control. This also argues for anormal functioning of dopaminergic systems inpsychiatric patients. It has suited drugcompanies, however, to support the idea of adopamine hypothesis of schizophrenia, as ifpatients fail to get well the obvious remedy ismore drugs. It also suits almost everybody elseas well, as if the dose of chlorpromazine were tobe restricted to around 400 mg/day or that ofhaloperidol to 30 mg/day - the amounts neededto block up to 80% of D2 receptors - psychiatrichospitals would become ungovernable and men-tal illness even more of a pressing political issue.

Perhaps symbolic of the present orientation isthe shift that has occurred from the term phar-macopsychology, which was coined by Kraep-elin, to the modern term, psychopharmacology.The former suggests an exploration of the psycheby means of drugs. The latter conjures up con-cerns with plasma drug levels and receptor num-bers, with the psyche being of secondary concernand then principally by virtue of the quantitativecomplexity that it presents, when compared toorgans such as the heart, rather than because ofany qualitatively different problems that emergein this particular branch of pharmacology.There are historical reasons for this neglect

of subjective impressions (Healy, 1990b). Theneglect, however, is not inevitable or necessary.This current period began with the demise ofintrospection at the end of the nineteenth centuryand the rise of behaviourism at the start of thetwentieth. However, there are indications thatwith the development of neuropsychology andcognitive psychology that a willingness to workwith the subjective experiences consequent onaltered neuropsychological functioning is re-emerging (see Strauss and Estroff, 1989).A further cultural factor that is of significance

here relates to cultural conceptions of the role ofmedicine. At present the public perception ofthe medical enterprise is one that credits thedramatic improvements in health over the pasttwo centuries to medical intervention and to the

increasing complexity of medical biotechnology(McKeown, 1979). The evidence however doesnot support this view. While medical dev-elopments have been important, it would seemthat social and economic factors have been muchmore influential in improving health and arelikely to remain so for the forseeable future(McKeown, 1979). This has led to a neglect ofsocial and behavioural inputs to the origin ofdisease and also to its treatment.Perhaps this background of overvaluing

biotechnical contributions to solution of healthproblems, leads to relatively minor effects suchas that of clomipramine and other drugs in OCDthat do, however, reach the 0.05 confidenceintervals, being seen as significant. The waysuch data are presented in academic journals,and indeed the language used, obscures thelarger question of whether such effects areindeed significant. Technically the term signifi-cant is often misused in current practice andshould in many cases be replaced by referenceto confidence intervals (Gardner and Altman,1986). It is all too easy to produce effects, regard-ing whose replicability one can speak with speci-fiable amounts of confidence but should sucheffects dominate clinical practice if they offerpatients little of tangible benefit? (Marks et al.,1988, 1989; OSullivan and Mark, 1990). Theanswer in practice at present would appear tobe that if these effects are brought about bybiotechnical means rather than by efforts to mod-ify individual behaviour, then the weight of cur-rent cultural beliefs regarding the role ofmedicine will lead to their widespread prom-ulgation and to the impression that the effectsare significant (Marks, 1989).

Concluding remarks

The thrust of this paper has been that whileindividual scientists may make discoveries, thereare larger cultural and economic forces whichalso have a bearing on the likelihood of signifi-cant discoveries being made in psychiatry andthe brain sciences. Regarding these forces, accessto the data that reveal their influences is atpresent restricted. Accordingly one has to askwhether a history (as opposed to a mythology)of the psychopharmacological era can be written.The only work that approaches being a historyof this area is Josephine Swazeys 1974 study of


165

chlorpromazine, in which the contributions ofindividuals, the academic community, the phar-maceutical industry, state agencies and the zeit-geist are all weighed in the balance (Swazey,1974). Perhaps what is needed is further volumeson individual drugs rather than on individualscientists.More generally where economic factors are

concerned, their operation can be described, atleast in part, in terms of a Luke effect; deter-mining which scientific seeds will fall on barrenground, which will be choked by thistles andweeds on growing and which will yield up theirbounty. The commercial underpinning of thepsychopharmaceutical industry has almost cer-tainly helped to prepare the soil for selectedscientific seedlings. Whether it also fosters thedevelopment of choking weeds is a question thatneeds addressing - one on which a certain his-torical perspective can now be brought.Aside from such major issues in the sociology

of science, can any more modest conclusions bedrawn from the data of the psychopharm-acological era? At present two would seem to bepossible. The first is that nothing of note hashappened as regards the development of new(legal) psychotropic drugs since the 1950s. Inparticular we still await a truly antischizophrenicagent, and we have no idea why the anti-depressants we have seem as limited in theireffectiveness as they do. One might suggest thatthe current fashion for operational criteria mightbe usefully extended to cover antidepressantsand neuroleptics, although this might not beeasily achieved (Shepherd, 1990). Perhaps partlybecause criteria of the type that restricted theuse of the term antidepressant, for example,to compounds whose effects closely resembledthose of imipramine, would be bad for business.Whatever the reason, the current profusion ofincreasingly sophisticated research instrumentsand methodologies only seems to make it easierfor investigators to claim anything they want. Itwould seem to paraphrase an old dictum, thatthere is fraud, damned fraud and there isresearch methodology.

Second, the most substantial development ofrecent years would appear to be in the realm ofattitudes to mental illness - a change that drugcompanies have helped to bring about - butparadoxically not because of any great effec-tiveness of current psychopharmacologicalagents!

ReferencesAmdisen A 1984 Lithium treatment of mania and depression

over one hundred years, in Current Trends in Lithiumand Rubidium Therapy, ed Corsini G U. MTP Press,Lancaster, 11-26

Ayd F J, Blackwell B 1970 Discoveries in biologicalpsychiatry, Lippincott, Philadelphia.

Baumann P, Jonzier-Perey M, Koeb L, Le P K, Tinguely D,Schopf J 1986 Amitriptyline pharmacokinetics and clinicalresponse: 1 free and total plasma amitriptyline and nor-triptyline. Int Clin Psychopharmacol 1: 89-101

Baldessarim R 1980 Drugs used in the treatment of thepsychoses, in The pharmacological basis of therapeutics,eds Gilman A G, Goodman L S, Gilman A. Basingstoke,Macmillan, 395-418

Barrow N, Childs A 1986 Anti-tardive-dyskmesia effect ofverapamil. Am J Psychiat 143: 1485

Bartels S T, Drake R E 1988 Depressive symptoms in schizo-phrenia : comprehensive differential diagnosis. Comp Psy-chiat 29: 467-483

Beaumont G, 1989 The toxicity of antidepressants. Brit JPsychiat 154: 454-458

Blacker R, Clare A 1987 Depressive disorder in primarycare. Brit J Psychiatry 140: 737-751

Blackwell B 1970 The process of discovery, in Discoveries inbiological psychiatry, eds Ayd F J, Blackwell B, Lippin-cott, Philadelphia, 11-29

Brandner M 1964 Schweiz Arch Neurol Psychiat 93, 137Cantopher T, Edwards J G, Olivieri S 1988 The government,

the pharmaceutical industry and the economics of psy-chotropic drug prescribing. Human Psychopharmacol 3:289-300

Capstick N 1971 Chlorimipramine in obsessional states.Psychosomatics 12: 332

Capstick N 1973 The Graylingwell Study. J Int Med Res 1:392-396

Carlsson A 1982 Rationale and design of a selective inhibitorof 5HT reuptake. Brit J Clin Pract (Symp Suppl) 19: 19-22

Carney M W P, Sheffield B F 1975 Forty-two months experi-ence of flupenthixol decanoate in the maintenance treat-ment of schizophrenia. Curr Med Res Opin 3: 447-452

Cocarro E F, Siever L J 1985 The second generation ofantidepressants: a comparative review. J Clin Psy-chopharmacol 25: 241-260

Cordoba E F, Lopez-Ibor J J 1967 Monochlorimipramine inpsychiatric patients resistant to other forms of treatment.Actas Luso-espanolas Neurol Psiquiat 26: 119-147

Cowdry R W, Gardner D L 1988 Pharmacotherapy of bor-derline personality disorder: alprazolam, carbamazepine,trifluoperazine, tranylcypromine. Arch Gen Psychiat 45:111-119

Cottraux J 1988 Obsession-compulsion: pharmacologicaland/or behavioural treatment. Actual Psychiat 22-27

Davies D L, Shepherd M 1955 Reserpine in the treatment ofanxious and depressed patients. Lancet ii: 117-120

Davies D M 1981 Textbook of adverse drug reactions, OxfordUniversity Press, Oxford, 226-227

De Montigny C, Chaput Y, Blier P 1988 Lithium aug-mentation of antidepressant treatments: evidence for the


166

involvement of the 5HT system? in New concepts indepression eds Briley M, Fillion G. Macmillan Press, Bas-ingstoke 144-160

Dinan T G, Capstick C 1989 A pilot study of verapamil in thetreatment of tardive dyskinesia. Hum Psychopharmacol 4:55-58

Dukes M N 1988 Meylers side effects of drugs, Elsevier,Amsterdam

Fahy T J 1989 Depression in general practice: the Newcastleconnection, in contemporary themes in psychiatry, edsDavison K, Kerr A, Gaskell Publications, Alden Press,Oxford

Frolund F 1974 Treatment of depression in general practice;a controlled trial of flupenthixol (Fluanxol). Curr Med ResOpin 2: 78-89

Gardner M J, Altman D G 1986 Confidence intervals ratherthan P values: estimation rather than hypothesis testing.Brit Med J 292: 746-750

Girard M, Biscos-Garreau M 1989 Reliability of data onhaematotoxicity of antidepressants: a retrospective assess-ment of haematological monitoring in clinical studies ontricyclics. J Affect Disord 17: 153-158

Goodman W K, Price L H, Rasmussen S A, Delgado P L,Hemnger G R, Charney D S 1989 Efficacy of fluvoxaminein obsessesive-compulsive disorder. Arch Gen Psychiat 46:36-44

Guyotat J, Favre-Tissot M, Marie-Gardine M 1968 A clinicaltrial with a new antidepressant G 34586. Congress depsychiatrie et de neurologie, Dijon 1967, 717, Masson,Paris

Hamilton M 1967 Development of a rating scale for primarydepressive illness. Brit J Soc Clin Psychol 6: 278-296

Healy D 1987a The structure of psychopharmacological rev-olutions. Psychiat Dev 5: 349-376

Healy D 1987b The comparative psychopathology of affectivedisorders in animals and humans. J Psychopharmacol 1:193-210

Healy D 1989 Neuroleptics and psychic indifference. J RoySoc Med 82: 615-619

Healy D, Williams J M G 1989 Moods, misattributions andmania. Psychiatric Dev 7: 49-70

Healy D 1990a Schizophrenia: basic, release, reactive anddefect processes. Human Psychopharmacol 5: 105-121

Healy D 1990b The suspended revolution: psychiatry andpsychotherapy re-examined. Faber and Faber Ltd, London

Holmberg M B, Jansson B 1962 A study of blood count andserum transaminase in prolonged treatment with ami-trityline. J New Drugs 2: 361-365

Johnson D A W 1979 A double-blind comparison of flu-penthixol, nortriptyline and diazepam in neurotic depres-sion. Acta Psychiat Scand 59: 1-8

Johnson D A W, Malik N A 1975 A double-blind comparisonof fluphenazine decanoate and flupenthixol decanoate inthe treatment of acute schizophrenia. Acta Psychiat Scand51: 257-267

Johnson F N 1984 The history of lithium therapy. Macmillan,London

Kalinowsky L B, Hoch P H 1961 Somatic treatment inpsychiatry. Grune and Stratton, New York

Katz R, Landau P L, Moroz G, DeVeaugh-Geiss J 1988Obsessive-compulsive rituals. Brit J Psychiat 153: 845

Kelly D, Guirguis W, Frommer E, Mitchell-Heggs N, SargantW 1970 Treatment of phobic states with antidepressants:a retrospective study of 246 patients. Brit J Psychiat 116:387-398

Klerman G L, Cole J O 1965 Clinical pharmacology ofimipramine and related compounds. Pharmacol Rev 17:101-141

Klerman G L, Ballenger J C, Burrows G D, DuPont R L,Noyes R, Pecknold J C, Rifkm Q A, Rubin R T, SwinsonR P 1989 In reply to Marks et al. Arch Gen Psychiat 46:670-672

Kline N S 1954 Use of reserpine serpentina benth in neu-ropsychiatric conditions. Ann N Y Acad Sci 59: 107-132

Kuhn R 1970 The imipramine story, in Discoveries in bio-logical psychiatry, eds Ayd F J, Blackwell B, Lippincott,Philadelphia 205-217

Kuhn R 1990 Artistic imagination and the discovery of anti-depressants. J Psychopharmacol 4(3): 00-00

Leonard B E 1984 Pharmacology of new antidepressants.Prog Neuro Psychopharmacol Biol Psychiat 8: 97-108

Liebenau J 1987 Medical science and medical industry, Mac-millan Press, Basingstoke

Liebowitz M R, Klein D F 1979 Hysteroid dysphoria. Psy-chiat Clin N Amer 2: 555-575

McKeown T 1979 The role of medicine, Basil Blackwell,Oxford

McNair D M, Rizley R, Kahn R J 1986 Amoxapine andspeed of onset: new antidepressant faces old methodology.J Clin Psychiat Mono, 4: 18-22

Marks I M 1989 The gap between research and policy inmental health care. J Roy Soc Med 82: 514-517

Marks I M, Stern R S, Mawson D, Cobb J, McDonald R1980 Clomipramine and exposure for obsessive-compulsiverituals. Brit J Psychiat 136: 1-25

Marks I M, Lelliott P, Basoglu M, Noshirvani H, MonteiroW, Cohen D, Kasvikis Y 1988 Clomipramine, self-exposure and therapist-aided exposure for obsessive-com-pulsive rituals. Brit J Psychiat 152: 522-534

Marks I, Basoglu M 1989 Obsessive-compulsive rituals. BritJ Psychiat 154: 565-567

Marks I, OSullivan G 1988 Drugs and psychological treat-ments for agoraphobia/panic and obsessive-compulsivedisorders: a review. Brit J Psychiat 153: 650-658

Marks I M, De Albuquerque A, Cotteaux J, Gentil V, GreistJ, Hand I, Liberman R L, Relvas J S, Tobena A, Tyrer P,Wittchen H-U 1989 The efficacy of alprazolam in panic-disorder and agoraphobia: a critique of recent reports.Arch Gen Psychiat 46: 668-670

Marshall W K, Micev V 1973 Clomipramine (Anafranil) inthe treatment of obsessional illnesses and phobic anxietystates. J Int Med Res: 403-412

May P R, Van Putten T, Yale C 1976 Predicting individualresponses to drug treatment in schizophrenia: a test dosemodel J Nerv Ment Dis 162: 177-183

Medical Research Centre Trial 1965 Clinical trial of thetreatment of depressive illness. Brit Med J i: 881-886

Merton R K 1968 The Matthew effect in science. Science159: 56-63

Merton R K 1969 Behaviour patterns of scientists. Am Sci57: 1-23


167

Moller H J, Meier K, Wernicke T 1988 Empiricial inves-tigation on the risk of agranulocytosis/leucopenia undermedication with antidepressants. Pharmacopsychiatry 21:304-305

Montgomery S A, James D, De Ruiter M, Egglestone J,Hawley C, Birkett P, Baldwin D, Chu C, Montgomery D1986 Weekly oral fluoxetine treatment of major depressivedisorder. 15th CINP, Puerto Rico

Office of Population Censuses and Surveys 1977-1984. Mor-tality statistics: accidents and violence. Series DH4, Table10. Deaths from poisoning by solid or liquid substances.HMSO, London

OSullivan G, Marks I 1990 Long-term outcome of phobicand obsessive-compulsive disorders after treatment, inHandbook of anxiety disorders, eds Noyes R et al, ElsevierPress, Amsterdam

Ovhed I 1976 A double-blind study of flupenthixol (Fluanxol)in general practice. Curr Med Res Opin 4: 144

Pare C M B 1985 The present status of monoamine-oxidaseinhibitors. Brit J Psychiat 146: 576-584

Parnham M J, Bruinvels J 1983 Discoveries in pharmacology.vol 1 Psycho and neuropharmacology, Elsevier Press,Amsterdam

Paykel E S, Parker R R, Rowan P R, Rao B M, Taylor C N1983 Nosology of atypical depression. Psychol Med 13:131-139

Paykel E S 1989 Antidepressants: their efficacy and place intherapy. J Psychopharmacol 2: 105-118

Pinder R 1988 The benefits and risks of antidepressant drugs.Human Psychopharmacol 3: 73-86

Pollitt J, Young J 1971 Anxiety state or masked depression?A study based on the action of monoamine oxidase inhibi-tors. Brit J Psychiat 119: 143-149

Pollock B G, Perel J M, Nathan S, Kupfer D J 1989 Acuteantidepressant effect following pulse loading with intra-venous and oral clomipramine. Arch Gen Psychiat 46: 29-35

Porter R, Porter D 1989 The rise of the English drugsindustry: the role of Thomas Corbyn. Med Hist 33: 277-295

Predescu V, Ciurezu T, Timofte G, Roman I 1973 Symp-tomatic relief with flupenthixol (Fluanxol) of the anxious-algetic-depressive syndrome complex in neurotic states.Acta Psychiat Scand 49: 15-27

Rack P H 1973 Clomipramine (Anafranil) in the treatmentof obsessional states with special reference to the LeytonObsessional Inventory. J Int Med Res 1: 397-402

Razak 1983 Flupenthixol versus phenelzine in the treatmentof neurotic depression: a double-blind trial cited in Rob-ertson and Trimble as in press

Reiter P J 1969 On flupenthixol, an antidepressant of a newchemical group. Brit J Psychiat 115: 1399-1402

Rickels K, Feighner J P, Smith W T 1985 Alprazolam,amitriptyline, doxepin and placebo in the treatment ofdepression. Arch Gen Psychiat 42: 134-141

Robertson M M, Trimble M R 1982 Major tranquillisers usedas antidepressants. J Affect Disord 4: 173-193

Rosenberg I U, Ostensen A I, Fonnelo H 1976 Flupenthixol-nortriptyline in the treatment of patients with anxiety-depression-asthenia. Tidssk Nor Laegeforen 96: 229-233

Sandler M 1990 The discovery of monoamine oxidase inhibi-tors : history and mythology. J Psychopharmacol 4(3): 00-00

Scherer M A 1988 Intravenous cocaine: psychiatric effects,biological mechanisms. Biol Psychiat 24: 865-885

Shepherd M, Cooper B, Brown A C, Kalton G 1986 Psy-chiatric illness in general practice, Oxford University Press,London

Shepherd M 1990 The Neuroleptics. J Psychopharmacol4(3) 00-00

Sireling L I, Paykel E S, Freeling P, Rao R M, Patel S P1985 Depression in general practice: case thresholds anddiagnosis. Brit J Psychiat 147: 113-119

Sonne 1 M 1966 Flupenthixol (fluanxol) in the treatment ofdepressive states. Nord Psychiat Tidssk 20: 322-324

Strauss J S, Estroff S E 1989 Foreword. Schizophrenia Bull15: 177-178

Sulser F, Mishra R 1983 The discovery of tricyclic anti-depressants and their mode of action, in Discoveries mPharmacology vol 1, Psycho and neuropharmacology, edsParnham M J, Bruinvels J, Elsevier Press, Amsterdam,233-247

Swazey J P 1974 Chlorpromazine and psychiatry. MIT Press,Cambridge, Mass

Symes M H 1967. Monochlorimipramine: a controlled trialof a new antidepressant. Brit J Psychiat 113: 671-672

Tiller J, Schweitzer I, Maguire K, Davies B 1989 Is diazepaman antidepressant? Brit J Psychiat 155: 483-489

Tuomisto J, Mannisto P 1985 Neurotransmitter regulation ofanterior pituitary hormones. Pharmacol Rev 37: 249-332

Waddington J L 1989 Implications of recent research ondopamine D1 and D2 receptor subtypes in relation toschizophrenia and neuroleptic drug action. Curr Opin Psy-chiat 2: 89-92

Waddington J L, OBoyle K M 1989 Drugs acting on braindopamine receptors: a conceptual re-evaluation 5 yearsafter the first selective D1 antagonist. Pharmacol Ther 43:1-52

Walter C J S 1973 Clinical impressions on treatment ofobsessional states with intravenous clomipramine (anaf-ranil). J Int Med Res 1: 413-416

Waxman D 1973 A general practitioner investigation on theuse of clomipramine (anafranil) in obsessional and phobicdisorders. J Int Med Res 1: 417-420

West E D, Dally P J (1959) Effects of iproniazid in depressivesyndromes. Brit Med J 1: 1491-1499

Whalley L J 1989 Drug treatments of dementia. Brit J Psy-chiat 155: 595-611

Young J P R, Hughes W C, Lader M H 1976 A controlledcomparison of flupenthixol and amitripytline in depressedout-patients. Brit Med J i: 1116-1118

Zis A P, Goodwin F K 1982 The amine hypothesis, inHandbook of the affective disorders, ed Paykel, E S.Churchill-Livingstone, Edinburgh 175-190.

Zohar J, Insel T R, Zohar-Kadouch R C, Mueller E A,Murphy D L 1989 Serotonergic challenges before and aftertreatment in obsessive-compulsive disorder. J Psych-opharmacol Proc Summer Meeting of Brit Assoc for Psy-chopharmacology


1990 Notes to a History

Documents

psychopharmacological

medical science

british association

futuresignificant discoveries

economic influences

drug business

branch of science

termsof drug use