European Ne,~ropsychopharmacology, 1 (1991)89--95 © 1991ElsevierSciencePublishersB.V. / 0924-977X/91//$3.50 ADONIS 0924977X9100161M
89
ENP 00018
Review article
Psychopharmacology of anxiety Willy Haefely Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd., CH-4002 Basel (Switzerland)
(Received 15 October, 1990) (Accepted 10 December, 1990)
Key words: Anxiety; Animal model; Anxiolytic; Anxiogenic; Benzodiazepine receptor; Serotonin receptor
Summary This overview deals first with the actual views on the neurobiology of anxiety, focusing on the limbic system and epileptiform neuronal activities. No single neurotransmitter system is likely to be involved exclusively in prevention or generation of anxiety and, hence, to be a target for drugs affecting anxiety specifically. Next, animal tests currently used to identify drugs effective in the treatment of anxiety, in particular specific forms of anxiety, are discussed. Drugs currently used in the treatment of anxiety as well as anxiogenic agents and their mechanisms of action are compared. The future development in the field of anxiety-reducing drugs is seen in partial agonists of benzodiazepine receptors. Specific ligands of the various 5-HT receptors may replace benzodiazepine receptor ligands in specific forms of anxiety disorders, but do not seem to represent general alternatives, in particular, when an immediate attenuation of anxiety is required.
Drugs capable of reducing anxiety in its various forms, besides being of extreme practical therapeutic importance, are also of considerable scientific interest as probes for the study of the biological basis of anxiety and related emotional disorders, as well as brain function in general. For the last three decades, pharmacotherapy of anxiety has been virtually synonymous with the use of benzodiazepines (BZs). The elucidation of their molecular mechanism of action has indicated ways to improve on the therapeutic profile of ligands of the benzodiazepine receptor (BZR), mainly by reducing some undesired effects encountered with short or long lasting use. Recently, novel drugs have emerged that might be useful in the
Correspondence to: Dr. W. Haefely, Pharmaceutical Research Department, F. Hoffmann-LaRocheLtd., CH-4002Basel, Switzerland.
treatment of at least certain forms of anxiety and which act through mechanisms unrelated to the GABAA receptor and its associated modulatory sites (e.g. the BZR). Although it is too early for a judgement of the positive and negative properties of these drugs, it is reasonable to expect that our anxiolytic armamentarium will continue to be expanded in the near future. In this Context, terminological and conceptual problems arise. BZs immediately reduce anxiety in man and in most animal models presently available. Such an immediate or direct 'anxiolytic' action is not seen with other drugs that ameliorate chronic anxiety disorders with a considerable latency, e.g. MAO inhibitors and imipramine-like agents, and which are not called anxiolytics. Increasingly refined diagnostic criteria (classification of anxiety disorders on the basis of mainly phenomenological characteristics) and rating scales
90 enabling quantification of the intensity of anxiety will be useful in the evaluation of drugs for the treatment of anxiety disorders. A negative aspect of the huge interest in the various aspects of anxiety disorders and their treatment is that it detracts attention from less sophisticated, but not less relevant, problems related to the frequent states of anxiety in the diverse fields of medicine which require treatment with immediately effective anxiolytic drugs. This brief overview only touches on a few selected aspects of the psychopharmacology of anxiety which have to be considered at a point in time when intensive research efforts are being made to explore potential new avenues.
Anxiety: general aspects Anxiety is one of the most frequent human emotions which every individual experiences on occasion. It is the perception or anticipation of a danger threatening the integrity, security, the well-being and self-esteem of the individual. The threat may be a real, or only a potential danger from the outside world. However, increasingly important are doubts about one's own ability to cope with demands and expectations imposed by other people or by the individual himself, or doubts concerning achievement of a chosen goal. The symptoms of anxiety are well known and affect all levels of biological activity (affective-cognitive, vegetative, endocrine, motor, behavioural). For the experimental biologist working on animals, behavioural inhibition is of prime interest for obvious reasons. While definition and symptomatology of anxiety do not pose major problems, the difficulties become apparent when dealing with the various forms of anxiety. Anxiety in patients is encountered very differently by general practitioners, anaesthetists, doctors in intensive care units, internists or psychiatrists. Normal (adaptive), and mostly transient, anxiety, (Table 1) i.e. anxiety as an appropriate, understandable reaction to situations of psychic or physical stress, may not only be extremely painful and distressing, but also very dangerous, and requires appropriate drug treatment. Anxiety disorders proper (Table 2), as classified, e.g., in DSM-III-R, are longlasting states of anxiety in the absence of actual danger or stress or outlasting them or are excessively intense emotions induced by normal stimuli. How relevant is anxiety as a medical problem? Anxiety disorders are the most common psychiatric
TABLE 1 FORMS OF ANXIETY normal (adaptive) anxiety situational short-term long-term personality type associated with physical illness pathological anxiety symptomatic physical illness (cardiovascular respiratory, gastrointestinal endocrine) psychopathological (affective disorders schizophrenia) personality disorders idiopathic (anxiety disorders) DSM III-R
illnesses (Robins et al., 1984). Epidemiological studies performed in the U.S.A. yield a prevalence rate for anxiety disorders of 4--8%, the latter figure being the more realistic. The life-time prevalence for an anxiety disorder is estimated to be 16% of the population. Over half of the patients seeking a doctor's help are thought to show symptoms of anxiety. Fortunately, in addition to being frequent, anxiety disorders are also the psychiatric illnesses most responsive to treatment. Against this background of information, the high prescription rate of antianxiety drugs cannot be considered as merely due to doctor's laziness and patients' lack of psychic resistance.
TABLE 2 CLASSIFICATION OF ANXIETY DISORDERS ACCORDING TO (DSM III-R) (IN PARENTHESES DISORDERS WITH ANXIETY CLASSIFIED ELSEWHERE IN DSM-III-R) Generalized anxiety disorder Panic disorder with Agoraphobia without Agoraphobia Phobic disorders Agoraphobia (without history of panic disorders) Social Phobia Simple Phobia Obsessive compulsive disorder Post-traumatic stress disorder Anxiety disorder not otherwise specified (Separation anxiety disorder) (Adjustment disorder with anxious mood)
91
Neurobiology of anxiety The continuum of anxiety as a characteristic state with emotional, behavioural, motor and vegetative symptoms between normal fear as a vital signal of danger and free-floating, incapacitating disorder suggests a priori a rather stereotyped activity pattern of interconnected neuronal networks as the basis of anxiety. Different sources of information on the possible nature and localization of CNS systems producing the symptoms of anxiety are listed in Table 3. An impressive number of studies deal with localized brain abnormalities that are frequently connected with states and bouts of anxiety. Epileptic activity, either focal in limbic structures or reaching them secondarily, has frequently been associated with anxiety (e.g., Maurach, 1983; Roth and Harper, 1962), pointing very convincingly to paroxysmal neuronal activity in these structures as the biological basis of anxiety and introducing the kindling phenomenon into psychobiology. Results obtained by local electrostimulation are fully congruent with the above evidence from 'experiments of nature'. Further support for this notion derives from brain lesions (spontaneous and in psychosurgery) eliminating limbic structures or separating them from other brain regions. Drugs as probes for exploring brain function are a further, extremely valuable source of information on the neurobiology of anxiety. On the one hand, all drugs with immediate anxiolytic activity have anticonvulsant properties typical of drugs active in petit mal type epilepsies. Increasing evidence is also accumulating for anxiolytic activity of drugs primarily used in the treatment of epilepsies (e.g., valproate). On the other hand, subconvulsive doses of convulsantagents or appropriate doses of proconvulsant agents have long been known to induce acute TABLE 3 APPROACHES TO THE NEUROBIOLOGICAL NATURE OF ANXIETY localized disorders of the CNS associated with anxiety (e.g. 'limbic' epilepsy) electrostimulation brain lesions (psychosurgery) reducing anxiety drugs anxiolytic anxiogenic PET studies (localized changes in brain metabolism and cerebral circulation)
anxiety in man and anxiety-related behaviours and somatic responses in animals. Limbic structures have been shown to be the most sensitive to many of these agents. Indirect measurements of regional brain activity using positron emission tomography have shown abnormal activity in the temporal pole in patients with panic disorders and in healthy volunteers during anxious anticipation of aversive stimuli (Reiman et al., 1989). Similar studies in animals, estimating regional glucose utilization by the 2-deoxy[14C]glucose autoradiography, localized increased activity in mamillary bodies, septal nuclei and hippocampi of rats in response to the partial BZR inverse agonist carboline FG 7142 (Pratt et al., 1988), known to be anxiogenic in man. Thus, all five sources of information point to the limbic structures as the core regions involved in the generation of anxiety and suggest neuronal activity patterns similar to or identical with those occurring in epileptic disorders. It is difficult to avoid associating ictal- and interictal-like activities with paroxysmal (phasic) and lasting (tonic) anxiety states, respectively. Behavioural inhibition is an outstanding symptom of anxiety and the one most accessible to experimentation in animal studies. It is not surprising, therefore, that a 'behavioural inhibition system' (Gray, 1982) has been proposed as a synthetic construct of neuronal systems generating inhibition of ongoing behaviour and other symptoms of anxiety (Fig. 1). Intrinsic neuronal circuits of the limbic system and input as well as output systems into and out of the limbic system are proposed to form the complex neuronal network that generates anxiety (Gray, 1982, Gorman et al., 1989). Specific neurotransmitter systems play an important part in the anxiety Systems according to the hypothesis of Gray (1982) and other authors (see Gorman et al., 1989): the ascending noradrenergic locus caeruleus pathway and the ascending serotoninergic raph6 systems. The crucial role of the noradrenergic system is not supported by more recent studies, and more attention is now being paid to the serotoninergic system, mainly because the identification of a diversity of 5-HT receptors (Hartig, 1989) is expected to resolve serious discrepancies in the pharmacological support of the 5-HT hypothesis. The two monoaminergic systems are well known to be involved in diverse functions of the CNS and they are, therefore, unlikely to provide the basis for 'anxioselective' drugs. Clearly, the notion of specific neurotransmitter or modulator systems determining specific CNS functions is wishful thinking and a primitive, isolated view on the complex interaction of various neuronal systems in the control and
92
Outputs
Inputs Signalsof Punishment Signalsof N.....ard ~ [
i
Behavioural Inhibition
System
N~_~IncreasedAttention I Impair Anti-Anxiety Drugs Fig. 1. Simplified diagram of the behavioural inhibition systems (from Gray, 1982).
generation of emotional responses. Excitatory (e.g. glutamic acid) and inhibitory transmitters (mainly GABA) are as indispensable as elements in the anxiety-generating systems as they are in any complex function of the CNS and, hence, cannot serve as targets for anxioselective drugs. How, for practical purposes, selectivity of action can be achieved with drugs affecting the ubiquitous GABAA receptor, will be discussed below.
Behavioural animal tests used in anxiety research
The most common test procedures currently in use to identify anxiolytic agents (Haefely et al., 1986) are listed in Table 4. Inhibition of conditioned or spontaneous ongoing behaviour is the most frequently measured parameter. Behavioural inhibition is induced by response-contingent punishment, fear of response-contingent punishment, frustration, and aversive stimuli (one of which is thought to be novelty). The major debated aspects are whether behavioural inhibition is actually the result of conflict-induced anxiety and whether reversal of behavioural inhibition by drugs reflects an anxiolytic action proper or a more general behavioural disinhibition, e.g. inability to withhold responses, or even purely amnestic activity. The high degree of validity provided by the excellent correlation of antipunishment effect of drugs in these models and acute anxiolytic effect in man does not help to resolve these issues. Other critical arguments can be applied to aversive brain stimulation experiments and to the defensive burying test, which measures a coping response to an aversive object. While drug discrimination tests are subject to a great many possible interpretations, the precipitated withdrawal-induced convulsive cue
selection (Emmett-Oglesby et al., 1983) is highly interesting in view of the major problem of withdrawal anxiety in anxiolytic-dependent patients and the great need for drugs able to reduce this withdrawal symptom without the danger of replacing one type of dependency by another. Induction of anxiety-like behaviour by chemicals is possible using convulsive or proconvulsive agents as already mentioned above. The striking fact with these models is clearly that, if they involve anxiety at all, they represent acute situational, adaptive anxiety and that drugs are testTABLE 4 MAJOR ANIMAL BEHAVIOURAL TESTS CURRENTLY USED IN ANXIOLYTIC DRUG RESEARCH (1) 'Conflict-punishment tests' Punished instrumental behaviour response-contingent shock fear of punishment Punishment of spontaneous behaviour punished drinking (consumption) four-plate test (exploration) Conflict without punishment dark-light chamber elevated plus maze food consumption (novel environment, unfamiliar food) social interaction (2) 'Frustration' (non-reward) and aversive stimulation intermittent non-reward in a schedule of responding maintained by response-contingent reward defensive burying aversive PAG stimulation (3) Anxiety induced by anxiogenic chemicals (4) Withdrawal anxiety (5) Ultrasonic distress cries
9,3
ed for immediate anxiety-relieving activity. Animal models of chronic anxiety in general, as well as of panic disorders or phobic disorders in particular, if they exist, are not yet broadly usedin the drug discovery process. Interestingly, amine uptake inhibitors and an irreversible MAO inhibitor were found to be effective after chronic pretreatment in a conditioned drinking suppression test (Fontana et al., 1989).
Drugs used for the treatment of anxiety
Drugs used currently for the treatment of various forms of anxiety (Table 5) clearly fall into two very different categories: those with immediate anxiolytic effect in animal models and anxious patients (agonists of the B Z R and, in the past, barbiturates and propandiolcarbamates) and those effective in anxiety disorders after a rather long latency and lacking immediate anxiolytic activity (MAO inhibitors, imipramine-like drugs). Buspirone belongs to this last category based on clinical results; immediate effects in the common acute models of anxiety are either absent or inconsistent, weak, and restricted to a narrow dose range, except for an antipunishment effect in the pigeon. The anxiolytic effect of antipsychotics in low doses is controversial in man and absent in animal models. There is obviously an urgent need for novel animal tests of anxiety capable of picking up drugs that might ameliorate anxiety by yet unknown mechanisms differing from those of currently used 'classical' drugs. The mechanisms of action of drugs used in the treatment of anxiety are indicated in Table 6. Barbiturates and BZs act by enhancing the effectiveness of GABAergic synaptic transmission through activation of two different allosteric modulatory sites on the GABAA receptor/chloride channel. With barTABLE 5 DRUGS CURRENTLY USED IN THE TREATMENT OF ANXIETY (1) barbiturates meprobamate BZR agonists and partial agonists antipsychotics (2) antidepressants (MAO-I, amineuptake-I) buspirone (novel 5-HT receptor ligands) ' (1) with immediate efficacy (2) with delayed efficacy
TABLE 6 MECHANISMS OF ACTION OF MAJOR DRUGS EFFECTIVE IN VARIOUS ANXIETY STATES Barbiturates
modulation of GABAA-R inhibitionof various excitatory synapses inhibitionof various ion channels
Meprobamate Benzodiazepine receptor agonists
modulation of GABAA-R (increased sensitivity to GABA)
Buspirone
agonist of 5-HT1A-R other mechanisms?
biturates this G A B A potentiation is very marked and, at high doses, is accompanied by a direct GABA-independent opening of chloride channels'. Moreover, the effect of barbiturates on the GABAA receptor/chloride channel is only one of many effects on other ion channels and excitatory transmitter receptors, which readily accounts for the restricted anxiolytic usefulness of barbiturates. The BZR, the site at which BZs act with high specificity, is unique for various reasons. It is the first allosteric modulatory site on a transmitter receptor found to mediate a very important therapeutic .action. Even more surprising is the fact thati the B Z R mediates two opposite effects, facilitation and inhibition of GABA-mediated chloride channel gating, which can be competitively blocked by ligands of the B Z R that lack modulatory activity by themselves. The facilitatory effect of full and partial agonists and the inhibitory effect of full and partial inverse agonists of the B Z R on the effectiveness of G A B A in inducing a chloride conductance of the neuronal cell membrane, is shown in the diagram of Fig. 2. This diagram illustrates another important aspect of B Z R agonists, namely that the sensitization of neurones to the inhibitory effect of G A B A mediated through GABAA receptors is achieved.without increase of the maximal inhibitory efficacy of GABA. Alt.hough the B Z R mediates both anxiolytic and anxiogenic effects, it is no more an 'anxiety receptor' than adrenergic or 5-HT receptors. The potent and efficacious anxiolytic activity of B Z R agonists and the ease with which inverse B Z R agonists induce anxiety derives from the strategic location of the GABAA-BZR Complex in neuronal circuits. Abnormal neuronal activity of critical circuits in the generation of anxiety makes the involved neurones more responsive to the GABA enhancing action of B Z R agonists than circuits with normal activity. Conversely, tuning down
94 anxiety in healthy subjects (Murphy et al., 1989). The anxiogenic effect of peripheral sympathetic over-stimulation (e.g., injection of adrenaline) is likely to be due to excitatory stimuli originating in the periphery. The attack-inducing effect of lactate in patients with panic disorder is a special case and the mechanism is yet unknown. The suggested existence of endogenous anxiogenic agents, e.g. inverse agonists of the BZR, is still an intriguing possibility, although it only defers the problem to the mechanism and situation triggering the release of such a compound (Haefely, 1988a).
g Cl-
i '1
/
,ii
'
'
/\D
!
I
//I / / ,,'/ // //
/ / " " : ' ~" /"
///// I
I
Io
I
[GABA]
Fig. 2. Schematic diagram of the allosteric modulation of the GABAA receptor/chloride channel by BZR ligands. C: theoretical dose-response curve for the chloride conductance inducing effect of G A B A in a single neurone. A: leftward shift of the GABA dose-response curve induced by a maximally effective concentration of a full BZR agonist; B: shift produced by a submaximal concentration of a full agonist or by the maximally effective concentration of a partial BZR agonist. D: effect of a full inverse BZR agonist.
the G A B A A receptor function by inverse BZR agonists increases activity in anxiety-generating circuits more easily than in others. Drugs with a delayed effect on anxiety apparently do not affect directly the balance between excitation and inhibition in anxiety related circuits. Antidepressants and, possibly, buspirone-like compounds are more likely to induce slowly developing and longer lasting changes in neuronal sensitivities (plasticity changes), e.g. reduced responsiveness to excitatory inputs or increased responsiveness to inhibitory influences.
Anxiogenic agents
Future developments in the psychopharmacology of
anxiety A question that arises immediately when considering possible future developments in the drug treatment of anxiety is whether ligands of the BZR will soon be replaced as anxiolytics by other drugs. I think that BZR ligands will be replaced as soon as true alternatives emerge, i.e. drugs with at least as good therapeutic efficacy and clearly less undesired effects. Such alternatives for a replacement of BZR ligands in their broad spectrum of indications do not
open field
• 300 > 50
horis.wit'e
tel
ataxia
ral
alexia
dol
sedation
rl/
/o
monkell
Iocom.acf,
re1
atasla
eel
max.(S
lal
conflict
ral
PTZ seizures conflict
Agents with generally accepted anxiogenic activity are all convulsants or proconvulsants (i.e., agents reducing GABAA receptor function by any of the many steps from GABA synthesis to chloride channel gating modulation, antagonists of adenosine receptor, such as caffeine). Similarly, convincing evidence for an anxiogenic action of agents affecting central noradrenergic and serotoninergic transmission (yohimbine) is lacking..To complicate the suggested role of 5-HT receptors in the anxiolytic activity of novel drugs, the 5-HT agonist mchlorophenylpiperazine has been found to induce
tel
cont,avotd, monkel
I
el.
•/,,,i>"
u/=
ral
monke , 0.01
, 01
, 1
, 1 mll/kg p,o.
, 100
, 1000
Fig. 3. Pharmacological profiles of a full B Z R agonist (diazepam) and partial agonist (Ro 16-6028). Plotted on the ordinate are effects with increasing inhibition of CNS activity. As can be seen, diazepam produces anticonflict (conflict), anticonvulsant (pentetrazol seizures, maximal electroshock seizures), muscle relaxation-ataxia, reduction of locomotor activity and severe sedation (horizontal wire test, continuous avoidance activity, open field behaviour) over a dose-range of approximately one log unit. Ro 166028 is more potent than diazepam in anticonflict and anticonvulsent tests, but is clearly less potent or even ineffective in producing severe sedation.
95
seem to be in sight. There will remain the need for anxiolytics with an immediate onset of action, for anxiolytics with sedative (even amnesic) effects; a muscle relaxant effect is not undesired in all situations. What will change - and the process is already under way - is a more correct use of the available drugs, which means, in particular, finding the balance between uncritical overuse and the hysterical damnation, underprescription, and withholding (Kratipl Taylor, 1989). Nevertheless, improvement on the available drugs is possible and has been initiated in three directions. One is to exploit the accumulated knowledge about the GABAA-BZR structure and function (Haefely, 1987). Partial agonists of the BZR, i.e. ligands with an intrinsic efficacy between that of full agonists and pure antagonists, have been shown to retain full anxiolytic and anticonvulsant efficacy with greatly reduced sedative and muscle relaxant-ataxic properties (Fig. 3). Even more important, the liability for development of tolerance and physical dependence has been shown to be greatly reduced or almost abolished with partial agonists (Martin et al., 1988; Pieri et al., 1988). The theoretical basis for this change in the pharmacological profiles has been discussed elsewhere (Haefely, 1988b, 1989). A second direction for improvement may be the selective inhibitors of 5-HT uptake and the new generation of reversible MAO-A inhibitors in the treatment of panic, phobic and compulsive-obsessive disorders. A third promising area are selective ligands (agonists, partial agonists or antagonists) of 5-HT receptor subtypes. Initial clinical results are available on buspirone and compounds derived from it, while other compounds are still in the preclinical phase of evaluation and their potential therapeutic usefulness is purely hypothetical. Buspirone has also shown the problems that might be inherent for this class of drugs. These are the long latency in the onset of anxiolytic action, as well as the possible lack of efficacy in patients treated previously with BZR ligands. Efficacy in various forms of anxiety disorders, e.g. panic disorders, remains to be demonstrated. Treatment compliance appears to be a serious problem and long-tei:m side effects cannot be disregarded. The main a~lvantage of buspirone, namely lack of dependence and addiction liability, will have to be carefully weighed against the shortcomings just mentioned.
References Emmett-Oglesby, M.W., Spener, D.G., Elmesallamy, F. and Lai, H. (1983) The pentylenetetrazol model of anxiety detects withdrawal from diazepam in rats. Life Sci. 33, 161-168. Fontana, D.J., Carbary, T.J. and Commissaris, R.L. (1989) Effects of acute and chronic anti-panic drug administration on conflict behavior in the rat. Psyehopharmacology 98, 157-162. Gorman, J.M., Liebowitz, M.R., Fyer, A.J. and Stein, J. (1989) A neuroanatomical hypothesis for panic disorders. Am. J. Psychiat. 146, 148-161. Gray, J.A. (1982) The Neurobiology of Anxiety, Clarendon Press, Oxford. Haefely, W.E. (1987) Structure and function of the benzodiazepine receptor. Chimia 41,389-396. Haefely, W. (1988a) Endogenous ligands of the benzodiazepine receptor. Pharmacopsychiatry 21, 43-46. Haefely, W. (1988b) Partial agonists of the benzodiazepine receptor: from animal to results in patients. Adv. Biochem. Psychopharmacol. 45, 27--46. Haefely, W.E. (1989) Pharmacology of the allosteric modulation of GABAA receptors by benzodiazepine receptor ligands. In: Barnard, E.A. and Costa, E. (Eds.), Allosteric Modulation of Amino Acid Receptors: Therapeutic Implications, Raven Press, New York, pp. 47-69. Haefely, W., Cumin, R. and Martin, J.R. (1986) VerhaltensmodeUe beim Tier zur Erfassung einer anxiolytischen Wirkung. In: Keup, W. (Ed.), Biologische Psychiatric Forschungsergebnisse, Springer-Verlag, Berlin, pp. 321-327. Hartig, P.R. (1989) Molecular biology of 5-HT receptors. Trends Pharmacol. Sci. 10, 64--69. Kr~upl Taylor, F. (1989) The damnation of benzodiazepines. Br. J. Psychiat. 154, 697-704. Martin, J.R., Pieri, L., Bonetti, E.P., Sehaffner, R., Burkard, W.P., Cumin, R. and Haefely, W.E. (1988) Ro 16-6028: a novel anxiolytie acting as a partial agonist at the benzodiazepine receptor. Pharmacopsychiatry 21,360-362. Maurach, R. (1983) Neuropsychologie der Angst. In: Strian, F. (Ed.), Angst - Grundlagen und Klinik, Springer-Vv'erlag, Berlin, pp. 87-98. Murphy, D.L., Mueller, E.A., Hill, J.L., Tolliyer, T.J. and Jacobsen, F.M. (1989) Comparative anxiogenic, neuroendocrine, and other physiological effects of m-chlorophenylpiperazine given intravenously or orally to healthy volunteers. Psychopharmacology 98,275-282. Pieri, L., Hunkeler, W., Jauch, R., Merz, W.A., Roncari, G. and Tirnm, U. (1988) Ro 16-6028. Drugs Fut. 13,730--735. Pratt, J.A., Laurie, D.J. and McCulloch, J. (1988) The effects of FG 7142 upon local central glucose utilization suggest overlap between limbie structures important in anxiety and convulsions. Brain Res. 475, 218--231. Reiman, E.M., Fusselman, M.J., Fox, P.T. and Ralchle, M.E. (1989) Neuroanatomical correlates of anticipatory anxiety. Science 243, 1071-1074. Robins, L.N., Helzer, J.E., Weissman, M.M., Orvaschel, H., Gruenberg, E., Burke, J.D. and Regier, D.A. (1984) Lifetime prevalence of specific psychiatric disorders in three sites. Arch. Gen. Psyehiat. 41,949-958. Roth, M. and Harper, M. (1962) Temporal lobe epilepsy and the phobic anxiety-depersonalization syndrome. II. Practical and theoretical considerations. Comp. Psyehiat. 3,215-226.
89
ENP 00018
Review article
Psychopharmacology of anxiety Willy Haefely Pharmaceutical Research Department, F. Hoffmann-La Roche Ltd., CH-4002 Basel (Switzerland)
(Received 15 October, 1990) (Accepted 10 December, 1990)
Key words: Anxiety; Animal model; Anxiolytic; Anxiogenic; Benzodiazepine receptor; Serotonin receptor
Summary This overview deals first with the actual views on the neurobiology of anxiety, focusing on the limbic system and epileptiform neuronal activities. No single neurotransmitter system is likely to be involved exclusively in prevention or generation of anxiety and, hence, to be a target for drugs affecting anxiety specifically. Next, animal tests currently used to identify drugs effective in the treatment of anxiety, in particular specific forms of anxiety, are discussed. Drugs currently used in the treatment of anxiety as well as anxiogenic agents and their mechanisms of action are compared. The future development in the field of anxiety-reducing drugs is seen in partial agonists of benzodiazepine receptors. Specific ligands of the various 5-HT receptors may replace benzodiazepine receptor ligands in specific forms of anxiety disorders, but do not seem to represent general alternatives, in particular, when an immediate attenuation of anxiety is required.
Drugs capable of reducing anxiety in its various forms, besides being of extreme practical therapeutic importance, are also of considerable scientific interest as probes for the study of the biological basis of anxiety and related emotional disorders, as well as brain function in general. For the last three decades, pharmacotherapy of anxiety has been virtually synonymous with the use of benzodiazepines (BZs). The elucidation of their molecular mechanism of action has indicated ways to improve on the therapeutic profile of ligands of the benzodiazepine receptor (BZR), mainly by reducing some undesired effects encountered with short or long lasting use. Recently, novel drugs have emerged that might be useful in the
Correspondence to: Dr. W. Haefely, Pharmaceutical Research Department, F. Hoffmann-LaRocheLtd., CH-4002Basel, Switzerland.
treatment of at least certain forms of anxiety and which act through mechanisms unrelated to the GABAA receptor and its associated modulatory sites (e.g. the BZR). Although it is too early for a judgement of the positive and negative properties of these drugs, it is reasonable to expect that our anxiolytic armamentarium will continue to be expanded in the near future. In this Context, terminological and conceptual problems arise. BZs immediately reduce anxiety in man and in most animal models presently available. Such an immediate or direct 'anxiolytic' action is not seen with other drugs that ameliorate chronic anxiety disorders with a considerable latency, e.g. MAO inhibitors and imipramine-like agents, and which are not called anxiolytics. Increasingly refined diagnostic criteria (classification of anxiety disorders on the basis of mainly phenomenological characteristics) and rating scales
90 enabling quantification of the intensity of anxiety will be useful in the evaluation of drugs for the treatment of anxiety disorders. A negative aspect of the huge interest in the various aspects of anxiety disorders and their treatment is that it detracts attention from less sophisticated, but not less relevant, problems related to the frequent states of anxiety in the diverse fields of medicine which require treatment with immediately effective anxiolytic drugs. This brief overview only touches on a few selected aspects of the psychopharmacology of anxiety which have to be considered at a point in time when intensive research efforts are being made to explore potential new avenues.
Anxiety: general aspects Anxiety is one of the most frequent human emotions which every individual experiences on occasion. It is the perception or anticipation of a danger threatening the integrity, security, the well-being and self-esteem of the individual. The threat may be a real, or only a potential danger from the outside world. However, increasingly important are doubts about one's own ability to cope with demands and expectations imposed by other people or by the individual himself, or doubts concerning achievement of a chosen goal. The symptoms of anxiety are well known and affect all levels of biological activity (affective-cognitive, vegetative, endocrine, motor, behavioural). For the experimental biologist working on animals, behavioural inhibition is of prime interest for obvious reasons. While definition and symptomatology of anxiety do not pose major problems, the difficulties become apparent when dealing with the various forms of anxiety. Anxiety in patients is encountered very differently by general practitioners, anaesthetists, doctors in intensive care units, internists or psychiatrists. Normal (adaptive), and mostly transient, anxiety, (Table 1) i.e. anxiety as an appropriate, understandable reaction to situations of psychic or physical stress, may not only be extremely painful and distressing, but also very dangerous, and requires appropriate drug treatment. Anxiety disorders proper (Table 2), as classified, e.g., in DSM-III-R, are longlasting states of anxiety in the absence of actual danger or stress or outlasting them or are excessively intense emotions induced by normal stimuli. How relevant is anxiety as a medical problem? Anxiety disorders are the most common psychiatric
TABLE 1 FORMS OF ANXIETY normal (adaptive) anxiety situational short-term long-term personality type associated with physical illness pathological anxiety symptomatic physical illness (cardiovascular respiratory, gastrointestinal endocrine) psychopathological (affective disorders schizophrenia) personality disorders idiopathic (anxiety disorders) DSM III-R
illnesses (Robins et al., 1984). Epidemiological studies performed in the U.S.A. yield a prevalence rate for anxiety disorders of 4--8%, the latter figure being the more realistic. The life-time prevalence for an anxiety disorder is estimated to be 16% of the population. Over half of the patients seeking a doctor's help are thought to show symptoms of anxiety. Fortunately, in addition to being frequent, anxiety disorders are also the psychiatric illnesses most responsive to treatment. Against this background of information, the high prescription rate of antianxiety drugs cannot be considered as merely due to doctor's laziness and patients' lack of psychic resistance.
TABLE 2 CLASSIFICATION OF ANXIETY DISORDERS ACCORDING TO (DSM III-R) (IN PARENTHESES DISORDERS WITH ANXIETY CLASSIFIED ELSEWHERE IN DSM-III-R) Generalized anxiety disorder Panic disorder with Agoraphobia without Agoraphobia Phobic disorders Agoraphobia (without history of panic disorders) Social Phobia Simple Phobia Obsessive compulsive disorder Post-traumatic stress disorder Anxiety disorder not otherwise specified (Separation anxiety disorder) (Adjustment disorder with anxious mood)
91
Neurobiology of anxiety The continuum of anxiety as a characteristic state with emotional, behavioural, motor and vegetative symptoms between normal fear as a vital signal of danger and free-floating, incapacitating disorder suggests a priori a rather stereotyped activity pattern of interconnected neuronal networks as the basis of anxiety. Different sources of information on the possible nature and localization of CNS systems producing the symptoms of anxiety are listed in Table 3. An impressive number of studies deal with localized brain abnormalities that are frequently connected with states and bouts of anxiety. Epileptic activity, either focal in limbic structures or reaching them secondarily, has frequently been associated with anxiety (e.g., Maurach, 1983; Roth and Harper, 1962), pointing very convincingly to paroxysmal neuronal activity in these structures as the biological basis of anxiety and introducing the kindling phenomenon into psychobiology. Results obtained by local electrostimulation are fully congruent with the above evidence from 'experiments of nature'. Further support for this notion derives from brain lesions (spontaneous and in psychosurgery) eliminating limbic structures or separating them from other brain regions. Drugs as probes for exploring brain function are a further, extremely valuable source of information on the neurobiology of anxiety. On the one hand, all drugs with immediate anxiolytic activity have anticonvulsant properties typical of drugs active in petit mal type epilepsies. Increasing evidence is also accumulating for anxiolytic activity of drugs primarily used in the treatment of epilepsies (e.g., valproate). On the other hand, subconvulsive doses of convulsantagents or appropriate doses of proconvulsant agents have long been known to induce acute TABLE 3 APPROACHES TO THE NEUROBIOLOGICAL NATURE OF ANXIETY localized disorders of the CNS associated with anxiety (e.g. 'limbic' epilepsy) electrostimulation brain lesions (psychosurgery) reducing anxiety drugs anxiolytic anxiogenic PET studies (localized changes in brain metabolism and cerebral circulation)
anxiety in man and anxiety-related behaviours and somatic responses in animals. Limbic structures have been shown to be the most sensitive to many of these agents. Indirect measurements of regional brain activity using positron emission tomography have shown abnormal activity in the temporal pole in patients with panic disorders and in healthy volunteers during anxious anticipation of aversive stimuli (Reiman et al., 1989). Similar studies in animals, estimating regional glucose utilization by the 2-deoxy[14C]glucose autoradiography, localized increased activity in mamillary bodies, septal nuclei and hippocampi of rats in response to the partial BZR inverse agonist carboline FG 7142 (Pratt et al., 1988), known to be anxiogenic in man. Thus, all five sources of information point to the limbic structures as the core regions involved in the generation of anxiety and suggest neuronal activity patterns similar to or identical with those occurring in epileptic disorders. It is difficult to avoid associating ictal- and interictal-like activities with paroxysmal (phasic) and lasting (tonic) anxiety states, respectively. Behavioural inhibition is an outstanding symptom of anxiety and the one most accessible to experimentation in animal studies. It is not surprising, therefore, that a 'behavioural inhibition system' (Gray, 1982) has been proposed as a synthetic construct of neuronal systems generating inhibition of ongoing behaviour and other symptoms of anxiety (Fig. 1). Intrinsic neuronal circuits of the limbic system and input as well as output systems into and out of the limbic system are proposed to form the complex neuronal network that generates anxiety (Gray, 1982, Gorman et al., 1989). Specific neurotransmitter systems play an important part in the anxiety Systems according to the hypothesis of Gray (1982) and other authors (see Gorman et al., 1989): the ascending noradrenergic locus caeruleus pathway and the ascending serotoninergic raph6 systems. The crucial role of the noradrenergic system is not supported by more recent studies, and more attention is now being paid to the serotoninergic system, mainly because the identification of a diversity of 5-HT receptors (Hartig, 1989) is expected to resolve serious discrepancies in the pharmacological support of the 5-HT hypothesis. The two monoaminergic systems are well known to be involved in diverse functions of the CNS and they are, therefore, unlikely to provide the basis for 'anxioselective' drugs. Clearly, the notion of specific neurotransmitter or modulator systems determining specific CNS functions is wishful thinking and a primitive, isolated view on the complex interaction of various neuronal systems in the control and
92
Outputs
Inputs Signalsof Punishment Signalsof N.....ard ~ [
i
Behavioural Inhibition
System
N~_~IncreasedAttention I Impair Anti-Anxiety Drugs Fig. 1. Simplified diagram of the behavioural inhibition systems (from Gray, 1982).
generation of emotional responses. Excitatory (e.g. glutamic acid) and inhibitory transmitters (mainly GABA) are as indispensable as elements in the anxiety-generating systems as they are in any complex function of the CNS and, hence, cannot serve as targets for anxioselective drugs. How, for practical purposes, selectivity of action can be achieved with drugs affecting the ubiquitous GABAA receptor, will be discussed below.
Behavioural animal tests used in anxiety research
The most common test procedures currently in use to identify anxiolytic agents (Haefely et al., 1986) are listed in Table 4. Inhibition of conditioned or spontaneous ongoing behaviour is the most frequently measured parameter. Behavioural inhibition is induced by response-contingent punishment, fear of response-contingent punishment, frustration, and aversive stimuli (one of which is thought to be novelty). The major debated aspects are whether behavioural inhibition is actually the result of conflict-induced anxiety and whether reversal of behavioural inhibition by drugs reflects an anxiolytic action proper or a more general behavioural disinhibition, e.g. inability to withhold responses, or even purely amnestic activity. The high degree of validity provided by the excellent correlation of antipunishment effect of drugs in these models and acute anxiolytic effect in man does not help to resolve these issues. Other critical arguments can be applied to aversive brain stimulation experiments and to the defensive burying test, which measures a coping response to an aversive object. While drug discrimination tests are subject to a great many possible interpretations, the precipitated withdrawal-induced convulsive cue
selection (Emmett-Oglesby et al., 1983) is highly interesting in view of the major problem of withdrawal anxiety in anxiolytic-dependent patients and the great need for drugs able to reduce this withdrawal symptom without the danger of replacing one type of dependency by another. Induction of anxiety-like behaviour by chemicals is possible using convulsive or proconvulsive agents as already mentioned above. The striking fact with these models is clearly that, if they involve anxiety at all, they represent acute situational, adaptive anxiety and that drugs are testTABLE 4 MAJOR ANIMAL BEHAVIOURAL TESTS CURRENTLY USED IN ANXIOLYTIC DRUG RESEARCH (1) 'Conflict-punishment tests' Punished instrumental behaviour response-contingent shock fear of punishment Punishment of spontaneous behaviour punished drinking (consumption) four-plate test (exploration) Conflict without punishment dark-light chamber elevated plus maze food consumption (novel environment, unfamiliar food) social interaction (2) 'Frustration' (non-reward) and aversive stimulation intermittent non-reward in a schedule of responding maintained by response-contingent reward defensive burying aversive PAG stimulation (3) Anxiety induced by anxiogenic chemicals (4) Withdrawal anxiety (5) Ultrasonic distress cries
9,3
ed for immediate anxiety-relieving activity. Animal models of chronic anxiety in general, as well as of panic disorders or phobic disorders in particular, if they exist, are not yet broadly usedin the drug discovery process. Interestingly, amine uptake inhibitors and an irreversible MAO inhibitor were found to be effective after chronic pretreatment in a conditioned drinking suppression test (Fontana et al., 1989).
Drugs used for the treatment of anxiety
Drugs used currently for the treatment of various forms of anxiety (Table 5) clearly fall into two very different categories: those with immediate anxiolytic effect in animal models and anxious patients (agonists of the B Z R and, in the past, barbiturates and propandiolcarbamates) and those effective in anxiety disorders after a rather long latency and lacking immediate anxiolytic activity (MAO inhibitors, imipramine-like drugs). Buspirone belongs to this last category based on clinical results; immediate effects in the common acute models of anxiety are either absent or inconsistent, weak, and restricted to a narrow dose range, except for an antipunishment effect in the pigeon. The anxiolytic effect of antipsychotics in low doses is controversial in man and absent in animal models. There is obviously an urgent need for novel animal tests of anxiety capable of picking up drugs that might ameliorate anxiety by yet unknown mechanisms differing from those of currently used 'classical' drugs. The mechanisms of action of drugs used in the treatment of anxiety are indicated in Table 6. Barbiturates and BZs act by enhancing the effectiveness of GABAergic synaptic transmission through activation of two different allosteric modulatory sites on the GABAA receptor/chloride channel. With barTABLE 5 DRUGS CURRENTLY USED IN THE TREATMENT OF ANXIETY (1) barbiturates meprobamate BZR agonists and partial agonists antipsychotics (2) antidepressants (MAO-I, amineuptake-I) buspirone (novel 5-HT receptor ligands) ' (1) with immediate efficacy (2) with delayed efficacy
TABLE 6 MECHANISMS OF ACTION OF MAJOR DRUGS EFFECTIVE IN VARIOUS ANXIETY STATES Barbiturates
modulation of GABAA-R inhibitionof various excitatory synapses inhibitionof various ion channels
Meprobamate Benzodiazepine receptor agonists
modulation of GABAA-R (increased sensitivity to GABA)
Buspirone
agonist of 5-HT1A-R other mechanisms?
biturates this G A B A potentiation is very marked and, at high doses, is accompanied by a direct GABA-independent opening of chloride channels'. Moreover, the effect of barbiturates on the GABAA receptor/chloride channel is only one of many effects on other ion channels and excitatory transmitter receptors, which readily accounts for the restricted anxiolytic usefulness of barbiturates. The BZR, the site at which BZs act with high specificity, is unique for various reasons. It is the first allosteric modulatory site on a transmitter receptor found to mediate a very important therapeutic .action. Even more surprising is the fact thati the B Z R mediates two opposite effects, facilitation and inhibition of GABA-mediated chloride channel gating, which can be competitively blocked by ligands of the B Z R that lack modulatory activity by themselves. The facilitatory effect of full and partial agonists and the inhibitory effect of full and partial inverse agonists of the B Z R on the effectiveness of G A B A in inducing a chloride conductance of the neuronal cell membrane, is shown in the diagram of Fig. 2. This diagram illustrates another important aspect of B Z R agonists, namely that the sensitization of neurones to the inhibitory effect of G A B A mediated through GABAA receptors is achieved.without increase of the maximal inhibitory efficacy of GABA. Alt.hough the B Z R mediates both anxiolytic and anxiogenic effects, it is no more an 'anxiety receptor' than adrenergic or 5-HT receptors. The potent and efficacious anxiolytic activity of B Z R agonists and the ease with which inverse B Z R agonists induce anxiety derives from the strategic location of the GABAA-BZR Complex in neuronal circuits. Abnormal neuronal activity of critical circuits in the generation of anxiety makes the involved neurones more responsive to the GABA enhancing action of B Z R agonists than circuits with normal activity. Conversely, tuning down
94 anxiety in healthy subjects (Murphy et al., 1989). The anxiogenic effect of peripheral sympathetic over-stimulation (e.g., injection of adrenaline) is likely to be due to excitatory stimuli originating in the periphery. The attack-inducing effect of lactate in patients with panic disorder is a special case and the mechanism is yet unknown. The suggested existence of endogenous anxiogenic agents, e.g. inverse agonists of the BZR, is still an intriguing possibility, although it only defers the problem to the mechanism and situation triggering the release of such a compound (Haefely, 1988a).
g Cl-
i '1
/
,ii
'
'
/\D
!
I
//I / / ,,'/ // //
/ / " " : ' ~" /"
///// I
I
Io
I
[GABA]
Fig. 2. Schematic diagram of the allosteric modulation of the GABAA receptor/chloride channel by BZR ligands. C: theoretical dose-response curve for the chloride conductance inducing effect of G A B A in a single neurone. A: leftward shift of the GABA dose-response curve induced by a maximally effective concentration of a full BZR agonist; B: shift produced by a submaximal concentration of a full agonist or by the maximally effective concentration of a partial BZR agonist. D: effect of a full inverse BZR agonist.
the G A B A A receptor function by inverse BZR agonists increases activity in anxiety-generating circuits more easily than in others. Drugs with a delayed effect on anxiety apparently do not affect directly the balance between excitation and inhibition in anxiety related circuits. Antidepressants and, possibly, buspirone-like compounds are more likely to induce slowly developing and longer lasting changes in neuronal sensitivities (plasticity changes), e.g. reduced responsiveness to excitatory inputs or increased responsiveness to inhibitory influences.
Anxiogenic agents
Future developments in the psychopharmacology of
anxiety A question that arises immediately when considering possible future developments in the drug treatment of anxiety is whether ligands of the BZR will soon be replaced as anxiolytics by other drugs. I think that BZR ligands will be replaced as soon as true alternatives emerge, i.e. drugs with at least as good therapeutic efficacy and clearly less undesired effects. Such alternatives for a replacement of BZR ligands in their broad spectrum of indications do not
open field
• 300 > 50
horis.wit'e
tel
ataxia
ral
alexia
dol
sedation
rl/
/o
monkell
Iocom.acf,
re1
atasla
eel
max.(S
lal
conflict
ral
PTZ seizures conflict
Agents with generally accepted anxiogenic activity are all convulsants or proconvulsants (i.e., agents reducing GABAA receptor function by any of the many steps from GABA synthesis to chloride channel gating modulation, antagonists of adenosine receptor, such as caffeine). Similarly, convincing evidence for an anxiogenic action of agents affecting central noradrenergic and serotoninergic transmission (yohimbine) is lacking..To complicate the suggested role of 5-HT receptors in the anxiolytic activity of novel drugs, the 5-HT agonist mchlorophenylpiperazine has been found to induce
tel
cont,avotd, monkel
I
el.
•/,,,i>"
u/=
ral
monke , 0.01
, 01
, 1
, 1 mll/kg p,o.
, 100
, 1000
Fig. 3. Pharmacological profiles of a full B Z R agonist (diazepam) and partial agonist (Ro 16-6028). Plotted on the ordinate are effects with increasing inhibition of CNS activity. As can be seen, diazepam produces anticonflict (conflict), anticonvulsant (pentetrazol seizures, maximal electroshock seizures), muscle relaxation-ataxia, reduction of locomotor activity and severe sedation (horizontal wire test, continuous avoidance activity, open field behaviour) over a dose-range of approximately one log unit. Ro 166028 is more potent than diazepam in anticonflict and anticonvulsent tests, but is clearly less potent or even ineffective in producing severe sedation.
95
seem to be in sight. There will remain the need for anxiolytics with an immediate onset of action, for anxiolytics with sedative (even amnesic) effects; a muscle relaxant effect is not undesired in all situations. What will change - and the process is already under way - is a more correct use of the available drugs, which means, in particular, finding the balance between uncritical overuse and the hysterical damnation, underprescription, and withholding (Kratipl Taylor, 1989). Nevertheless, improvement on the available drugs is possible and has been initiated in three directions. One is to exploit the accumulated knowledge about the GABAA-BZR structure and function (Haefely, 1987). Partial agonists of the BZR, i.e. ligands with an intrinsic efficacy between that of full agonists and pure antagonists, have been shown to retain full anxiolytic and anticonvulsant efficacy with greatly reduced sedative and muscle relaxant-ataxic properties (Fig. 3). Even more important, the liability for development of tolerance and physical dependence has been shown to be greatly reduced or almost abolished with partial agonists (Martin et al., 1988; Pieri et al., 1988). The theoretical basis for this change in the pharmacological profiles has been discussed elsewhere (Haefely, 1988b, 1989). A second direction for improvement may be the selective inhibitors of 5-HT uptake and the new generation of reversible MAO-A inhibitors in the treatment of panic, phobic and compulsive-obsessive disorders. A third promising area are selective ligands (agonists, partial agonists or antagonists) of 5-HT receptor subtypes. Initial clinical results are available on buspirone and compounds derived from it, while other compounds are still in the preclinical phase of evaluation and their potential therapeutic usefulness is purely hypothetical. Buspirone has also shown the problems that might be inherent for this class of drugs. These are the long latency in the onset of anxiolytic action, as well as the possible lack of efficacy in patients treated previously with BZR ligands. Efficacy in various forms of anxiety disorders, e.g. panic disorders, remains to be demonstrated. Treatment compliance appears to be a serious problem and long-tei:m side effects cannot be disregarded. The main a~lvantage of buspirone, namely lack of dependence and addiction liability, will have to be carefully weighed against the shortcomings just mentioned.
References Emmett-Oglesby, M.W., Spener, D.G., Elmesallamy, F. and Lai, H. (1983) The pentylenetetrazol model of anxiety detects withdrawal from diazepam in rats. Life Sci. 33, 161-168. Fontana, D.J., Carbary, T.J. and Commissaris, R.L. (1989) Effects of acute and chronic anti-panic drug administration on conflict behavior in the rat. Psyehopharmacology 98, 157-162. Gorman, J.M., Liebowitz, M.R., Fyer, A.J. and Stein, J. (1989) A neuroanatomical hypothesis for panic disorders. Am. J. Psychiat. 146, 148-161. Gray, J.A. (1982) The Neurobiology of Anxiety, Clarendon Press, Oxford. Haefely, W.E. (1987) Structure and function of the benzodiazepine receptor. Chimia 41,389-396. Haefely, W. (1988a) Endogenous ligands of the benzodiazepine receptor. Pharmacopsychiatry 21, 43-46. Haefely, W. (1988b) Partial agonists of the benzodiazepine receptor: from animal to results in patients. Adv. Biochem. Psychopharmacol. 45, 27--46. Haefely, W.E. (1989) Pharmacology of the allosteric modulation of GABAA receptors by benzodiazepine receptor ligands. In: Barnard, E.A. and Costa, E. (Eds.), Allosteric Modulation of Amino Acid Receptors: Therapeutic Implications, Raven Press, New York, pp. 47-69. Haefely, W., Cumin, R. and Martin, J.R. (1986) VerhaltensmodeUe beim Tier zur Erfassung einer anxiolytischen Wirkung. In: Keup, W. (Ed.), Biologische Psychiatric Forschungsergebnisse, Springer-Verlag, Berlin, pp. 321-327. Hartig, P.R. (1989) Molecular biology of 5-HT receptors. Trends Pharmacol. Sci. 10, 64--69. Kr~upl Taylor, F. (1989) The damnation of benzodiazepines. Br. J. Psychiat. 154, 697-704. Martin, J.R., Pieri, L., Bonetti, E.P., Sehaffner, R., Burkard, W.P., Cumin, R. and Haefely, W.E. (1988) Ro 16-6028: a novel anxiolytie acting as a partial agonist at the benzodiazepine receptor. Pharmacopsychiatry 21,360-362. Maurach, R. (1983) Neuropsychologie der Angst. In: Strian, F. (Ed.), Angst - Grundlagen und Klinik, Springer-Vv'erlag, Berlin, pp. 87-98. Murphy, D.L., Mueller, E.A., Hill, J.L., Tolliyer, T.J. and Jacobsen, F.M. (1989) Comparative anxiogenic, neuroendocrine, and other physiological effects of m-chlorophenylpiperazine given intravenously or orally to healthy volunteers. Psychopharmacology 98,275-282. Pieri, L., Hunkeler, W., Jauch, R., Merz, W.A., Roncari, G. and Tirnm, U. (1988) Ro 16-6028. Drugs Fut. 13,730--735. Pratt, J.A., Laurie, D.J. and McCulloch, J. (1988) The effects of FG 7142 upon local central glucose utilization suggest overlap between limbie structures important in anxiety and convulsions. Brain Res. 475, 218--231. Reiman, E.M., Fusselman, M.J., Fox, P.T. and Ralchle, M.E. (1989) Neuroanatomical correlates of anticipatory anxiety. Science 243, 1071-1074. Robins, L.N., Helzer, J.E., Weissman, M.M., Orvaschel, H., Gruenberg, E., Burke, J.D. and Regier, D.A. (1984) Lifetime prevalence of specific psychiatric disorders in three sites. Arch. Gen. Psyehiat. 41,949-958. Roth, M. and Harper, M. (1962) Temporal lobe epilepsy and the phobic anxiety-depersonalization syndrome. II. Practical and theoretical considerations. Comp. Psyehiat. 3,215-226.
