The new generation of antipsychotic drugs : how atypical are they ?

R E V I E W A RT I C LE

International Journal of Neuropsychopharmacology (2000), 3, 339–349. Copyright # 2000 CINP

Jeffrey M. Goldstein Clinical Development and Medical Affairs Group, AstraZeneca Pharmaceuticals, Wilmington, Delaware, USA

Abstract The cardinal feature of traditional neuroleptic drugs is extrapyramidal symptoms (EPS), such as parkinsonism, akathisia, and dystonia. Irrespective of the autonomic nervous system side-effect profile of a specific neuroleptic, the entire class produces EPS. Therefore, EPS and antipsychotic activity were once thought to be inextricably linked. However, with the discovery of clozapine, this concept was no longer defensible. Clozapine produced antipsychotic actions without associated EPS or increases in serum prolactin levels, and the term ‘ atypical ’ was coined to differentiate its actions from those of the traditional agents. Later, the definition of atypical was expanded to encompass clozapine’s unique clinical spectrum of activity, including its effectiveness in treating some patients unresponsive to traditional neuroleptics. Clozapine thus became the archetype for a new generation of antipsychotic drugs, which now includes quetiapine, olanzapine, risperidone, sertindole, ziprasidone, zotepine and amisulpride. This paper will review the pharmacological actions that contribute to the unique features of clozapine, focusing on receptor profile and activity in animal models used for evaluations of antipsychotic activity and EPS. Similarities and differences amongst the new agents will also be discussed. Although conclusions regarding atypicality require controlled clinical trials in addition to preclinical and animal models, it is apparent from this review that not all agents match the profile of clozapine. Received 19 July 1999 ; Reviewed 14 September 1999 ; Revised 25 June 2000 ; Accepted 29 June 2000 Key words : Antipsychotic drugs, atypicality, EPS, prolactin.

Introduction Clozapine is the archetypal ‘ atypical ’ antipsychotic drug. It differs from traditional (i.e. haloperidol-like) antipsychotics in having a low propensity for producing extrapyramidal symptoms (EPS) after acute dosing, not producing tardive dyskinesia after chronic administration, and not elevating plasma prolactin levels. Moderate (rather then high) affinity for the striatal dopamine D # receptor, associated with low receptor occupancy, has been hypothesized to explain clozapine’s ability to produce antipsychotic, but not EPS, effects. In addition to its low EPS liability, clozapine also demonstrates superior antipsychotic efficacy compared with traditional agents, a property not easily explained by the drug’s pharmacology. Unfortunately, clozapine can cause agranulocytosis, a potentially fatal blood disorder, thus limiting its widespread therapeutic use. Address for correspondence : Dr J. M. Goldstein, Assistant Director, Clinical Development and Medical Affairs Group, AstraZeneca Pharmaceuticals, 1800 Concord Pike, Wilmington, Delaware 19850, USA. Tel. : j1 (302) 886 8071 Fax : j1 (302) 886 3078 E-mail : Jeffrey.Goldstein!AstraZeneca.com

The definition of ‘ atypicality ’ remains unclear, although it is generally accepted that, at a clinical level, diminished EPS and a reduced capacity to elevate plasma prolactin levels are central to the definition. Broader definitions include efficacy in treating negative symptoms and refractory patients. The pharmacological properties of clozapine that relate to the lack of EPS and prolactin elevation, e.g. neurotransmitter receptor affinity in the central nervous system (CNS), limbic selectivity and dystonic liability in haloperidol-sensitized monkeys, have provided insights into the potential mechanism for this drug’s unique effects (Goldstein 1995a, 1996). These properties have also provided a rationale for the discovery of several new and putatively atypical antipsychotic drugs (Goldstein, 1995b) including : risperidone (Risperdal2, Janssen) ; olanzapine (Zyprexa2, Lilly) ; quetiapine (Seroquel2, AstraZeneca) ; sertindole (Serdolect2, Lundbeck) ; ziprasidone (Zeldox2, Pfizer) ; zotepine (Zoleptil2, Orion) ; and amisulpride (Solian2, SanofiSynthelabo). How closely the properties of these drugs compare to those of clozapine, with respect particularly to EPS and prolactin elevation, as demonstrated both pharmacologically and clinically, will be the subject of this review.

340

J. M. Goldstein

Preclinical Receptor binding Among the various pharmacological effects of antipsychotics, their action on CNS neurotransmitter receptors has been an area of extensive investigation. Receptor profiles for antipsychotic drugs provide a convenient basis for making comparisons amongst agents, as well as providing a mechanistic rationale for predicting potency, efficacy and side-effects in humans. Figure 1 presents the receptor binding profiles for traditional (haloperidol) and atypical antipsychotics using relative binding affinity displayed as receptor pie charts. This method of presentation provides a convenient comparison of receptor properties across drugs as well as facilitating a comparison of receptor ratios. The pie charts were constructed by taking either published IC or Ki &! values and converting them to a percentage of total

activity (Goldstein, 1995b). Only the major endogenous CNS receptors hypothesized to be involved in the mechanism of action of antipsychotic drugs have been included in this analysis. However, it would be inappropriate to infer absolute receptor affinities based on a comparison of the slices across drugs. The receptor pie charts allow for the classification of the various antipsychotics, and these broadly fit into the following groups : (i) those interacting with multiple receptors (i.e. clozapine-like), a group that includes olanzapine, zotepine and quetiapine ; (ii) those with predominant selectivity for 5-HT receptors, e.g. # risperidone, sertindole, and ziprasidone ; and (iii) those predominantly selective for dopamine D receptors, e.g. # haloperidol and amisulpride. The receptor pie charts also provide an avenue to express and compare receptor ratios. These receptor ratios become more meaningful if the common denomi-

Figure 1. Comparative receptor binding profiles. Binding data for these charts taken from Goldstein (1995b). IC or Ki values &! were converted to percent of total activity by the following formula : [1\IC or Ki for D ]j[1\IC or Ki for D ]j[1\IC or Ki for 5-HT A]j[1\IC or Ki for 5-HT ]j[1\IC or Ki for M ] &! " &! # &! " &! # &! " j[1\IC or Ki for α ] j[1\IC or Ki for α ]j[1\IC or Ki for H ] l 100 %. &! " &! # &! " Receptor labels without respective pie slices indicate binding that was marginal. Only endogenous receptors that are targeted by antipsychotic agents were used for the analysis. No correlation to absolute affinity should be attempted between drugs based on the size of the pie slices as these represent relative binding.

The new generation of antipsychotic drugs

341

Table 1. Receptor affinity ratio Drug

5-HT A
D " #

5-HT
D # #

α D " #

α D # #

M D " #

H D " #

Clozapine Haloperidol Olanzapine Quetiapine Risperidone Sertindole Ziprasidone Zotepine Amisulpride

7 7 7 7 7 7 G 7 7

G 7 G G G G G G 7

G G 7 G G G 7 G 7

G 7 7 G 7 7 7 7 7

7 7 G 7 7 7 7 7 7

G 7 G G G 7 7 G 7

nator is the dopamine D receptor, since D receptor # # affinity is the one consistent feature of all clinically effective antipsychotic drugs, and correlates with clinical potency. The relative receptor ratios are summarized in Table 1, and the conclusions that can be drawn from the various receptor ratios are summarized as follows : (1) 5-HT A
D receptors. 5-HT A receptors are involved " # " with antidepressant activity and can modulate cortical glutamate transmission and contribute to antipsychotic effects (Reynolds and Czudek, 1995). Ziprasidone is the only listed drug that fits into this class and may produce antidepressant and antipsychotic effects via this mechanism. (2) 5-HT
D receptors. This has been hypothesized to # # be a distinguishing feature of all the atypicals (Meltzer, 1991, 1992), and predicts a separation of antipsychotic doses from those doses producing EPS. Clozapine, olanzapine, quetiapine, risperidone, sertindole, zotepine and ziprasidone all fit into this category. (3) α -adrenoceptor  D receptor. Predicts liability for " # cardiovascular effects, e.g. orthostatic hypotension (Reynolds, 1994 ; Goldstein 1995a). Clozapine, quetiapine, risperidone, zotepine and sertindole fit into this category. (4) α -adrenoceptor  D receptor. Predicted to be a mech# # anism associated with enhanced efficacy (Breier, 1994 ; Reynolds and Czudek, 1995). Clozapine and quetiapine fit into this category. (5) Muscarinic M  D receptors. Predicts anticholinergic " # activity (Goldstein, 1995a ; Reynolds, 1994). Olanzapine fits into this category. (6) Histamine H  D receptors. Predicts sedation " # (Goldstein 1995a ; Reynolds, 1994). Clozapine, quetiapine, olanzapine, zotepine and risperidone fit into this category. In an attempt to further understand the relationship between the commonality of D receptor binding for all #

clinically effective antipsychotic drugs and their obvious differences in EPS side-effect profiles, Seeman (1998) has provided recent evidence that the manner in which the drugs bind to the D receptor may be different. Using # antipsychotic drug dissociation constants (which are independent of the radioligand used), Seeman (1998) demonstrated that antipsychotic drugs that elicit EPS (e.g. trifluperazine, chlorpromazine, raclopride, thioridazine, haloperidol, fluphenazine and risperidone) bind more tightly than dopamine at the D receptor, whereas those # that elicit little or no EPS (e.g. sertindole, olanzapine, loxapine, molindone, remoxipride, clozapine, perlapine, quetiapine and melperone) bind more loosely than dopamine at the D receptor. # Seeman (1999) also provided direct evidence that loose or tight binding at the D receptor is associated with rapid # or slow release, respectively, of antipsychotic drugs from the receptor by physiological concentrations of dopamine (or the ligands used to label the D receptor in positron # emission tomography ; PET) studies, e.g. raclopride and iodobenzamide. Thus, clozapine and quetiapine, as examples of the most loosely bound atypical antipsychotic drugs, can be displaced within minutes from D # receptors, whereas haloperidol, chlorpromazine or olanzapine, as examples of the more tightly bound antipsychotic drugs, are displaced by dopamine 100 times more slowly. The relevance of the above findings to the receptor occupancy of these drugs is discussed below.

Receptor occupancy To help evaluate the mode of action of atypical antipsychotic agents in humans, PET studies have been undertaken to establish the level of receptor occupancy at clinically relevant doses. However, only limited data are currently available and only for striatal D and cortical 5# HT receptors. The D \5-HT receptor occupancy levels # # #

J. M. Goldstein

342

Table 2. D \5-HT receptor occupancy # # Receptor Clozapine D # 5-HT

#

Risperidone

Moderate High Very high High\very high

Quetiapine

Olanzapine

Sertindole

Low\moderate Moderate\high

Moderate\high High

? High or low ? High High Very high

are summarized in Table 2 for clozapine (Nordstrom et al., 1995), risperidone (Busatto et al., 1995 ; Farde et al., 1995 ; Kapur et al., 1995 ; Kufferle et al., 1996), ziprasidone (Bench et al., 1996 ; Fischman et al., 1996), olanzapine (Kapur et al., 1998 ; Nyberg et al., 1997 ; Pilowsky et al., 1996), sertindole (Arnt and Skarsfeldt, 1998 ; Kasper et al., 1998), quetiapine (Gefvert et al., 1995, unpublished observations ; Kufferle et al., 1997), and amisulpride (Martinot et al., 1996 ; Trichard et al., 1998). Most of these studies were performed using small numbers of volunteers or patients, and larger studies are required to address fully the relationship between receptor occupancy and EPS. However, the results suggest that high D receptor occupancy can be tolerated in the # absence of EPS, but only with commensurately higher 5-HT receptor occupancy. High occupancy of 5-HT # # receptor appears to be a class effect and occurs even below clinically effective doses. Therefore, 5-HT receptor # blockade is likely to be insufficient for antipsychotic activity but may contribute to overall efficacy and to a side-effect profile characterized by fewer EPS. It is also interesting to note that quetiapine appears to have a unique profile for D receptor occupancy, showing # a modest peak with a rapid decline such that there is minimal D receptor occupancy 12–24 h after the last # dose (Kapur, 2000). This is in contrast to all other antipsychotics, with the exception of clozapine, which show sustained high (
60 %) D receptor occupancy # 12 h after the last dose. This transient occupancy that occurs with quetiapine may be related to its loose binding at the D receptor and the ability of dopamine to displace # quetiapine. It has been suggested that this transient and low D receptor occupancy may explain quetiapine’s # freedom from EPS (and prolactin elevation, discussed later) and hence its atypicality. It is further speculated that only transiently high D receptor occupancy may be # sufficient for antipsychotic response, with sustained high D receptor occupancy being associated with side-effects. # Limbic selectivity Another widely accepted hypothesis that explains the separation of antipsychotic activity from EPS liability is limbic selectivity. This hypothesis is based on the

Ziprasidone Amisulpride High Low

observation that antipsychotic drugs with minimal EPS selectively block dopamine receptors in the limbic or cortical brain regions, whereas those agents that produce EPS also block the effects of dopamine on nigrostriatal function. Although a variety of techniques have been employed to measure limbic selectivity, the current review will focus its discussion on the primary electrophysiological measure (i.e. depolarization inactivation of limbic dopamine cells), and the primary neurochemical measure (i.e. expression of early gene products).

Depolarization inactivation Repeated 21–28 d oral administration of typical antipsychotic drugs (e.g. haloperidol and chlorpromazine) causes a decrease in the number of spontaneously active dopamine cells in both the motor-related A9 and limbicrelated A10 cell regions, owing to the development of a state of tonic depolarization (depolarization inactivation ; Chiodo and Bunney, 1983 ; White and Wang, 1983). However, clozapine failed to cause depolarization inactivation of dopamine A9 cells after chronic administration. These findings suggest that the time-dependent inactivation of limbic-related dopamine A10 cells might be related to the efficacy of antipsychotic drugs, whereas the inactivation of motor-related dopamine A9 cells may be correlated with their side-effect liability. The delayed clinical effects observed with antipsychotic administration in humans and the necessity of repeated administration of drugs to produce the effect in this model add to its relevance. Utilizing this model, limbic (A10) selectivity has been established for olanzapine (Stockton and Rasmussen, 1996), quetiapine (Goldstein et al., 1993), and sertindole (Skarsfeldt, 1992), but not for risperidone and ziprasidone (Arnt and Skarsfeldt, 1998). Although one can conclude from these findings that EPS liability should be limited for olanzapine, quetiapine and risperidone, there is considerable variation in the results reported between laboratories, which questions the reliability and predictability of the model. Furthermore, a careful study of doseresponse relationships for A9 vs. A10 selectivity is required before meaningful predictions regarding EPS liability over the entire antipsychotic dose range (e.g. a

The new generation of antipsychotic drugs

343

Table 3. Dystonic reactions vs predicted antipsychotic doses in haloperidol-sensitized Cebus monkeys

Antipsychotic

Dose producing 100 % incidence of dyskinetic reactions (mg\kg p.o.)

Predicted antipsychotic dose range (mg\kg p.o.)

Quetiapine Clozapine Haloperidol Olanzapine Risperidone Sertindole Ziprasidone


40
60 0n25 0n5 0n125 2n5 0n62

3–9 6–18 0n1–1n0 0n1–0n4 0n08–0n16 0n24–0n48 0n8–3n2

clozapine-like profile) can be made. Thus far, this has only been performed for sertindole (Skarsfeldt, 1992). Early gene expression Acute administration of clozapine and haloperidol produce different induction patterns of c-fos expression in the forebrain. Haloperidol increases fos-like immunoreactivity (FLI) in both the motor-related (i.e. dorsolateral striatum) and limbic-related (i.e. nucleus accumbens, lateral septum and prefrontal cortex) areas, while clozapine produces such effects only in the limbic-related areas (Robertson et al., 1994). Accordingly, this approach can be used to characterize the EPS liability of antipsychotic agents. Recent findings using this model with the newer agents suggest a diversity of actions. Quetiapine and risperidone elevated FLI in the nucleus accumbens and medial striatum, and minimally in the dorsolateral striatum, consistent with predictions of reduced EPS liability. However, olanzapine increased FLI in the prefrontal cortex, nucleus accumbens, and the dorsolateral striatum, an effect that correlates with the ability to produce EPS (Robertson and Fibiger, 1996). There is also a discrepancy in the induction of FLI after acute treatment with antipsychotic drugs and the delayed clinical effects observed with these agents. Recent studies have attempted to resolve this problem by determining the chronic effects of antipsychotic drugs on an early gene marker that only appears after chronic dosing (deltafosB ; Vahid-Ansari et al., 1996). Quetiapine and clozapine demonstrate a limbic-related pattern of expression of deltafosB, whereas haloperidol exhibits a non-selective pattern of expression, i.e. limbic and motor-related areas equally affected. More recently, risperidone and olanzapine have been shown to induce deltafosB in the caudate–putamen (a region associated with EPS) but not in the nucleus accumbens or prefrontal cortex (Atkins et al., 1999).

Taken together, the above findings support a prediction of minimal EPS for quetiapine after both acute and chronic administration, but data for the other atypical agents are still needed before conclusions from this model can be drawn. Furthermore, the limited dose ranges explored again limit the extrapolation of results from this model to man.

Haloperidol-sensitized Cebus monkeys Cebus monkeys treated with haloperidol once a week will eventually become sensitized and develop dystonic reactions (Casey, 1995 ; Casey et al., 1980 ; Goldstein, 1996). The haloperidol-sensitized Cebus monkey remains the most predictive animal model for EPS in four respects : (i) the dystonic reactions in sensitized monkeys are identical to acute EPS in humans, i.e. dystonia and parkinsonism ; (ii) the reactions in sensitized monkeys are precipitated by agents that have EPS liability in humans, e.g. haloperidol-like drugs ; (iii) the dystonic reactions are controlled by the same medications that control EPS in humans, i.e. anticholinergic agents ; and (iv) clozapine is the only antipsychotic agent for which cases of dyskinesias in humans are absent, and clozapine is the only antipsychotic agent tested which produces no dystonic reactions in the Cebus monkey, even up to maximally tolerated doses. These similarities between antipsychotic-induced EPS in humans and the dystonic reactions in haloperidolsensitized Cebus monkeys suggest that this may be the most predictive model for antipsychotic-induced EPS in humans. A comparison of the potential dystonic effects of the new antipsychotics in the haloperidol-sensitized monkey is presented in Table 3 (Goldstein, 1995 ; Goldstein et al., 1994). Drugs are compared on the basis of the dose range over which dystonic reactions occurred and the predicted

344

J. M. Goldstein

antipsychotic dose range in monkeys, based on clinical data. As expected of a drug with EPS liability, haloperidol produced a 100 % incidence of dystonic reactions at the lower end of its predicted antipsychotic dose range. In contrast, clozapine failed to produce any dystonic reactions at doses 3n3-fold higher than its maximum predicted antipsychotic dose in man. Of the new antipsychotics, only quetiapine and sertindole showed a marked separation of dystonic vs. antipsychotic doses. However, quetiapine was closest to clozapine in overall incidence of dystonic reactions : quetiapine failed to produce more than a 15 % incidence of dystonic reactions at the highest dose tested, whereas sertindole eventually produced a 100 % incidence of dystonic reactions (but not until a dose 5-fold greater than the upper clinically effective antipsychotic dose was reached). Although risperidone and olanzapine produced 100 % incidence of dystonic reactions, they were considerably better than haloperidol and showed a separation of antipsychotic doses vs. dystonic doses. Ziprasidone showed no separation between the 100 % dystonic dose and the predicted minimal antipsychotic dose in monkeys. Inasmuch as the dystonic reactions in haloperidolsensitized Cebus monkeys resemble EPS observed in humans after treatment with standard agents like haloperidol (Casey, 1995), the following predictions can be drawn from these results. Quetiapine and sertindole should have a reduced potential to cause EPS across the entire antipsychotic dose range ; risperidone and olanzapine should have a reduced EPS liability, although this advantage should be lost at higher doses ; and ziprasidone should, like haloperidol, have virtually no separation of EPS vs. antipsychotic doses. Arnt and Skarsfeldt (1998) also reported similar findings on the complete lack of separation of EPS-inducing doses from antipsychotic doses for ziprasidone.

for haloperidol or risperidone. However, the pharmacology of this paradigm and its ability to predict antipsychotic atypicality require further study. The model is relatively new although the evidence suggests that it can be used as a specific model to identify antipsychotic agents with only atypical properties.

Prolactin The most important endocrinological side-effect of antipsychotic treatment observed in both laboratory animals and humans is a substantial increase in prolactin secretion (Meltzer, 1992). This increase is of relevance in humans as it may lead to neuroendocrine complications such as galactorrhoea, gynaecomastia, menstrual changes and impotence. Prolactin secretion is normally inhibited by dopamine (Clemens et al., 1980 ; Saller and Salama, 1986). Thus, it is not surprising that because of their dopamine D receptor antagonist action, all antipsychotics increase # prolactin after acute administration. However, the profile of clozapine in humans is different from that of other antipsychotics, in that it produces little or no prolactin stimulation (Meltzer, 1992). The profile of clozapine in laboratory animals also differs from the classical antipsychotics in that clozapine produces a transient rise in prolactin whereas haloperidol produces a more sustained increase (Gudelsky et al., 1987 ; Meltzer, 1975). Of the newer antipsychotics, only quetiapine, risperidone, zotepine and amisulpride have been studied for their effects on prolactin in rats (Kakigi et al., 1992 ; Uchida et al., 1979). Acute administration of quetiapine was associated with a clozapine-like transient rise in plasma prolactin levels as opposed to the sustained increase produced by haloperidol (Goldstein, 1999). This transient increase in prolactin in rats predicts that quetiapine may have little or no effect on prolactin in man. Risperidone has been shown to be 3–5 times more potent than haloperidol in stimulating rat prolactin levels in vivo (Bowden et al., 1992).

Prepulse inhibition Prepulse inhibition (PPI) of the startle reflex (the reduction of the startle reflex by a prepulse stimulus) is diminished in schizophrenic patients. In rats, PPI is reduced or eliminated by the psychotomimetic non-competitive glutamate antagonist phencyclidine (PCP). The effects of PCP are not reversed by antipsychotics such as haloperidol (Keith et al., 1991) but are reversed by atypical agents such as clozapine (Bakshi et al., 1994), olanzapine (Bakshi and Geyer, 1995), and quetiapine (Swerdlow et al., 1996). Risperidone does not restore PPI in PCP-treated rats (Swerdlow et al., 1996). The data predict atypical antipsychotic effects for clozapine and quetiapine but not

Clinical studies It is outside of the scope of this paper to review clinical efficacy on negative symptoms and refractory patients ; these data have recently been reviewed by Remington and Kapur (2000) and King (1998). This section will therefore provide an overview of the clinical incidence of EPS and prolactin elevation for the new antipsychotics. In the absence of any published meta-analyses, data presented in the US Label or UK Product Monograph have been used where available. Whilst this approach has limitations – differences in patient type and assessment

EPS (%)

16 26 Placebo Zotepine

methodology will occur between pivotal studies – in the absence of any meta-analyses these data provide the fairest comparisons. The antipsychotic dose ranges shown are also taken from these sources. For more details on individual studies readers are referred to the recent review by Remington and Kapur (2000).

31 42 45 55 100 mg 400 mg 800 mg 1200 mg 16 6 6 4 8 6 a From

Sertindole UK Product Monograph (1977) ; b from US Label ; c from Puech et al. (1998).

Placebo 75 mg 150 mg 300 mg 600 mg 750 mg 27 21 13 24 Placebo 12 mg 20 mg 24 mg 27 44 55 56 Placebo 4 mg 8 mg 16 mg

Placebo 2 mg 6 mg 10 mg 16 mg

13 13 16 20 31

Placebo 5 mg 10 mg 15 mg

16 15 25 32

EPS (%) EPS ( %) Olanzapineb EPS (%) Risperidoneb EPS (%) Haloperidola

Table 4. Percentage incidence of extrapyramidal symptoms (EPS)

345

EPS incidence

Sertindolea

Quetiapineb

EPS (%)

Amisulpridec

EPS (%)

Zotepinea

The new generation of antipsychotic drugs

Table 4 lists the EPS incidence for the new atypical antipsychotics compared with haloperidol. Haloperidol, at all antipsychotic doses listed, produced an incidence of EPS higher than placebo. Both risperidone and olanzapine demonstrated improvement in their EPS profile compared with haloperidol. Both drugs appear to be associated with levels of EPS similar to placebo at the lower end of the antipsychotic dose range. However, this separation is lost at the higher antipsychotic doses where the incidence of EPS becomes greater than placebo. Sertindole and quetiapine do not appear to produce EPS at greater than placebo levels across their entire antipsychotic dose range. For ziprasidone EPS labelling data were not available. Clinical reports have suggested a lower risk of EPS in comparison with haloperidol for zotepine (Fleischhacker et al., 1989 ; Petit et al., 1996), ziprasidone (Goff et al., 1998) and amisulpride (Freeman 1997 ; Puech et al., 1998), although for amisulpride the data suggest that the incidence of EPS may increase with dose.

Prolactin Clinical data for prolactin appear, for the most part, to follow the clinical data for EPS. Haloperidol, a drug clearly associated with EPS, elevates prolactin levels and the elevation persists during chronic administration [Physician’s Desk Reference (PDR), 1998]. Risperidone, although having fewer EPS than haloperidol at the lower doses, also elevates prolactin levels and the elevation persists during chronic administration (PDR, 1998). Olanzapine has a lower liability for EPS at lower doses and elevates prolactin levels, with a modest elevation persisting during chronic administration (PDR, 1998). Although there were significant differences in mean prolactin levels from baseline between sertindole and placebo, the values were within normal limits and there were no elevations recorded in long-term studies (Sertindole UK Product Monograph, 1997). Quetiapine also did not cause EPS at any antipsychotic doses (Arvanitis et al., 1997), and an elevation in prolactin levels was not demonstrated in any of the clinical trials (PDR, 1998). Transient elevations in prolactin have been noted for

346

J. M. Goldstein

ziprasidone (Goff et al., 1998). In a small endocrine study of schizophrenia patients, prolactin was noted to increase with zotepine (von Bardeleben et al., 1987) and significant prolactin elevations have been observed following amisulpride administration in healthy volunteers (Wetzel et al., 1994).

Weight gain Most atypical antipsychotics cause weight gain to some extent and the increases may be substantial. The mechanism is unclear : blockade of 5-HT receptors may lead to # appetite deregulation causing weight gain or, alternatively, weight gain may be due to other direct (noradrenergic blockade) and indirect side-effects such as inactivity. Allison et al. (1999) compared weight changes with several atypical antipsychotics and found clozapine and olanzapine caused the most weight gain, risperidone and sertindole intermediate, quetiapine produced modest weight gain and ziprasidone caused minimal weight gain. Similar results were reported in a comparative study by Wirshing et al. (1999).

Discussion The mechanism of action of atypical antipsychotic drugs has been the topic for several recent reviews (Kinon and Lieberman, 1996 ; Tandon et al., 1999). This review has focused on the pharmacological properties of clozapine that predict minimal EPS liability and transient effects on prolactin. Similarities and differences amongst the new antipsychotics have been highlighted and addressed in the light of clinical results. Only limited preclinical data are available for zotepine, amisulpride and ziprasidone. Each of the other new antipsychotics profiled as atypical agents with respect to having fewer EPS compared with haloperidol. However, there were notable differences in the EPS profiles of the drugs where full doseresponse data were available. For example, in the haloperidol-sensitized monkey, quetiapine and sertindole showed the greatest separation of dystonic doses vs. antipsychotic doses, whereas risperidone and olanzapine lost this advantage at high antipsychotic doses. With respect to prolactin, quetiapine demonstrated a transient clozapine-like increase in prolactin in rats as opposed to the sustained increase produced by haloperidol and risperidone, thereby predicting minimal prolactin effects in humans. The results of the clinical studies support those obtained from observations in animal tests. For EPS, quetiapine and sertindole have an incidence of EPS similar to placebo across the antipsychotic dose range, although

the range of EPS-free dosing examined in clinical studies is considerably narrower for sertindole (12–24 mg\d) compared with quetiapine (75–750 mg\d). This separation of EPS liability from antipsychotic doses was predicted on the basis of the haloperidol-sensitized monkey test. Although risperidone and olanzapine also have an EPS incidence similar to placebo at lower doses, this advantage is lost at the higher antipsychotic doses ; at these doses the incidence of EPS is greater than placebo. Again, this was predicted on the basis of the haloperidolsensitized monkey. The EPS liability for ziprasidone and zotepine appears to be less than that for haloperidol, and for amisulpride it appears to be dose related. Inasmuch as the haloperidol-sensitized monkey test resembles acute EPS in humans, predictions based on this model so far closely parallel the clinical effects of the new antipsychotics. The results obtained for prolactin liability appear to follow closely those for EPS liability. Drugs capable of causing EPS, albeit only at high doses (risperidone, olanzapine, amisulpride), caused increases in prolactin. On the other hand, agents not associated with increases in EPS across the entire antipsychotic dose range (quetiapine, sertindole) either did not affect prolactin levels (quetiapine), or produced an increase in prolactin that was within normal clinical limits (sertindole). Finally, it is tempting to speculate that loose binding at D receptors, coupled with transient and low D receptor # # occupancy, as demonstrated by both clozapine and quetiapine, may be the mechanism responsible for these drugs’ lack of EPS and prolactin effects.

References Allison DB, Mentore JL, Heo M, Chandler LP, Capelleri JC, Infante MC, Weiden PJ (1999). Antipsychotic-induced weight gain: a comprehensive research analysis. American Jornal of Psychiatry 156, 1686–1696. Arnt J, Skarsfeldt T (1998). Do novel antipsychotics have similar pharmacological characteristics? A review of the evidence. Neuropsychopharmacology 18, 63–101. Arvanitis LA, Miller BG, Seroquel Trial 13 Study Group (1997). Multiple fixed doses of ‘ Seroquel ’ (quetiapine) in patients with acute exacerbation of schizophrenia, a comparison with haloperidol and placebo. Biological Psychiatry 42, 233–246. Atkins JB, Chlan-Fourney J, Nye HE, Hiroi N, Carlezon Jr. WA, Nestler EJ (1999). Region specific induction of deltaFosB by repeated administration of typical versus atypical antipsychotic drugs. Synapse 33, 118–128. Bakshi VP, Geyer MA (1995). Antagonism of phencyclidineinduced deficits in prepulse inhibition by the putative

The new generation of antipsychotic drugs atypical antipsychotic olanzapine. Psychopharmacology 122, 198–201. Bakshi VP, Swerdlow NR, Geyer MA (1994). Clozapine antagonizes phencyclidine-induced deficits in sensorimotor gating of the startle response. Journal of Pharmacology and Experimental Therapeutics 271, 787–794. Bench CJ, Lammertsma AA, Grasby PM, Dolan RJ, Warrington S, Boyce M, Gunn KP, Brannick LY, Frackowiak RS (1996). The time course of binding to striatal dopamine D2 receptors by the neuroleptic ziprasidone (CP-88,059-01) determined by positron emission tomography. Psychopharmacology 124, 141–147. Bowden CR, Voina SJ, Woestenborghs R, De Coster R, Heykants J (1992). Stimulation by risperidone of rat prolactin secretion in vivo and in cultured pituitary cells in vitro. Journal of Pharmacology and Experimental Therapeutics 262, 699–706. Breier A (1994). Clozapine and noradrenergic function, support for a novel hypothesis for superior efficacy. Journal of Clinical Psychiatry 55, 122–125. Busatto GF, Pilowsky LS, Costa DC, Ell PJ, Verhoeff NP, Kerwin RW (1995). Dopamine D2 receptor blockade in vivo with the novel antipsychotics risperidone and remoxipride – an 1231-IBZM single photon emission tomography (SPET) study. Psychopharmacology 117, 55–61. Casey DE (1995). Neuroleptic-induced acute extrapyramidal syndromes and tardive dyskinesia. In: Hirsch SR, Weinberger DR (Eds), Schizophrenia UK (pp. 546–565). London : Blackwell Science. Casey DE, Gerlach J, Christensson E (1980). Dopamine, acetylcholine, and GABA effects in acute dystonia in primates. Psychopharmacology 70, 83–87. Chiodo LA, Bunney BS (1983). Typical and atypical neuroleptics, differential effects of chronic administration on the activity of A9 and A10 midbrain dopaminergic neurons. Journal of Neurological Sciences 3, 1607–1619. Clemens JA, Shaar CJ, Smalstig EB (1980). Dopamine, PIF, and other regulators of prolactin secretion. Federation Proceedings 39, 2907–2911. Farde L, Nyberg S, Oxenstierna G, Nakashima Y, Halldin C, Ericsson B (1995). Positron emission tomography studies on D2 and 5-HT2 receptor binding in risperidone-treated schizophrenic patients. Journal of Clinical Psychopharmacology 15 (Suppl. 1), 19–23. Fischman AJ, Bonab AA, Babich JW, Alpert NM, Rauch SL, Elmaleh DR, Shoup TM, Williams SA, Rubin RH (1996). Positron emission tomographic analysis of central 5hydroxytryptamine2 receptor occupancy in healthy volunteers treated with the novel antipsychotic agent, ziprasidone. Journal of Pharmacology and Experimental Therapeutics 279, 939–947. Fleischhacker WW, Barnas C, Stuppack CH, Unterweger B, Miller C, Hinterhuber H (1989). Zotepine vs. haloperidol in paranoid schizophrenia, a double blind trial. Psychopharmacology Bulletin 25, 97–100. Freeman H. (1997). Amisulpride compared with standard neuroleptics in acute exacerbations of schizophrenia, three

347

efficacy studies. International Clinical Psychopharmacology 12 (Suppl. 2), 11–17. Gerfvert O, Lindstom LH, Langstrom J, Bergstrom M, Lundberg T, Yates RA, Larron SD, Tuersley MD (1995). Time course for dopamine and serotonin receptor occupancy in the brain of schizophrenic patients following dosing with 150 mg ‘ Seroquel ’ tid [Abstract]. European Journal of Neuropsychopharmacology (Special Issue), 347. Goff DC, Posever T, Herz L, Simmons J, Kletti N, Lapierre K, Wilner KD, Law CG, Ko GN (1998). An exploratory haloperidol-controlled dose-finding study of ziprasidone in hosptialized patients with schizophrenia or schizoaffective disorder. Journal of Clinical Psychopharmacology 18, 296–304. Goldstein JM (1995a). Preclinical tests that predict clozapinelike atypical antipsychotic actions. In : Brunello N, Racagni G, Langer SZ, Mendlewicz J (Eds.), Critical Issues in the Treatment of Schizophrenia (pp. 95–101). Basel : Karger. Goldstein JM (1995b). Pre-clinical pharmacology of new atypical antipsychotics in late stage development. Expert Opinion on Investigational Drugs 4, 291–298. Goldstein JM (1996). Preclinical tests that further distinguish the pharmacological properties of a clozapine-like atypical antipsychotic. In : Beninger RJ, Palomo T, Archer T (Eds.), Dopamine Disease States (pp. 299–306). Madrid: Editorial CYM. Goldstein JM (1999). Seroquel (quetiapine fumarate): a new atypical antipsychotic. Drugs Today 35, 193–210. Goldstein JM, Litwin LC, Sutton EJ, Malick BM (1993). Seroquel, electrophysiological profile of a potential atypical antipsychotic. Psychopharmacology 112, 293–298. Goldstein JM, Snyder DH (1994). Effects of Seroquel, clozapine and other atypical agents in primate models of EPS. Schizophrenia Bulletin 15, 152. Gudelsky GA, Koenig JI, Simonovic M, Koyama T, Ohmori T, Meltzer HY (1987). Differential effects of haloperidol, clozapine, and fluperlapine on tuberinfundibular dopamine neurons and prolactin secretion in the rat. Journal of Neural Transmission 68, 227–240. Kakigi T, Maeda K, Tanimoto K, Kaneda H, Shintani T (1992). Effect of substituted benzamides on prolactin secretion in the rat. Biological Psychiatry 31, 827–831. Kapur S, Remington G, Zipursky RB, Wilson AA, Houle S (1995). The D2 dopamine receptor occupancy of risperidone and its relationship to extrapyramidal symptoms: a PET study. Life Sciences 57, 103–107. Kapur S, Zipursky R, Jones C, Shammi CS, Remington G, Seeman P (In Press). A PET study of quetiapine in schizophrenia, a preliminary finding of an antipsychotic effect with transiently high dopamine D2 receptor occupancy. Archives of General Psychiatry. Kapur S, Zipursky RB, Remington G, Jones C, DaSilva J, Wilson AA, Houle S (1998). 5-HT2 and D2 receptor occupancy of olanzapine in schizophrenia : a PET investigation. American Journal of Psychiatry 155, 921–928. Kasper S, Tauscher J, Kufferle B, Barnas C, Hesselmann B, Asenbaum S, Podreka I, Brucke T (1998). Sertindole and

348

J. M. Goldstein

dopamine D2 receptor occupancy in comparison to risperidone, clozapine and haloperidol – a 1231-IBZM SPECT study. Psychopharmacology 136, 367–373. Keith VA, Mansbach RS, Geyer MA (1991). Failure of haloperidol to block the effects of phencyclidine and dizocilpine on prepulse inhibition of startle. Biological Psychiatry 30, 557–566. King DJ (1998). Drug treatment of the negative symptoms of schizophrenia. European Journal of Neuropsychopharmacology 8, 33–42. Kinon BJ, Lieberman JA (1996). Mechanisms of action of atypical antipsychotic drugs : a critical analysis. Psychopharmacology 124, 1–34. Kufferle B, Brucke T, Topitz-Schratzberger A, Tauscher J, Gossler R, Vesely C, Asenbaum S, Podreka I, Kasper S (1996). Striatal dopamine-2 receptor occupancy in psychotic patients with risperidone. Psychiatric Research 68, 23–30. Kufferle B, Tauscher J, Asenbaum S, Vesely C, Podreka I, Brucke T, Kasper S (1997). IBZM SPECT imaging of striatal dopamine-2 receptors in psychotic patients treated with the novel antipsychotic substance quetiapine in comparison to clozapine and haloperidol. Psychopharmacology 133, 323–328. Martinot JL, Paillere-Martinot ML, Poirier MF, Dao-Castellana MH, Loc’h C, Maziere B (1996). In vivo characteristics of dopamine D2 receptor occupancy by amisulpride in schizophrenia. Psychopharmacology 124, 154–158. Meltzer HY (1991). The mechanism of action of novel antipsychotic drugs. Schizophrenia Bulletin 17, 263–287. Meltzer HY (1992). The mechanism of action of clozapine in relation to its clinical advantages. In : Meltzer HY (Ed.), Novel Antipsychotic Drugs (pp. 1–13). New York : Raven Press. Meltzer HY, Daniels S, Fang VS (1975). Clozapine increases rat serum prolactin levels. Life Sciences 17, 339–342. Nordstrom AL, Farde L, Nyberg S, Karlsson P, Halldin C, Sedvall C (1995). D1, D2 and 5-HT2 receptor occupancy in relation to clozapine serum concentration : a PET study of schizophrenia patients. American Journal of Psychiatry 152, 1444–1449. Nyberg S, Farde L, Halldin C (1997). A PET study of 5-HT2 and D2 dopamine receptor occupancy induced by olanzapine in healthy subjects. Neuropsychopharmacology 16, 1–7. Petit M, Raniwalla J, Tweed J, Leutenegger E, Dollfus S, Kelly F (1996). A comparison of an atypical and typical antipsychotic, zotepine versus haloperidol in patients with acute exacerbation of schizophrenia, a parallel-group double-blind trial. Psychopharmacological Bulletin 32, 81–87. Physician’s Desk Reference (1998). 51st edn. New Jersey, Medical Economics Company. Pilowsky LS, Busatto GF, Taylor M, Costa DC, Sharma T, Sigmundsson T, Ell PJ, Nohria V, Kerwin RW (1996). Dopamine D2 receptor occupancy in vivo by the novel antipsychotic olanzapine – a 1231 IBZM single photon

emission tomography (SPET) study. Psychopharmacology 124, 148–153. Puech A, Fleurot O, Rein W (1998). Amisulpride, an atypical antipsychotic, in the treatment of acute episodes of schizophrenia, a dose-ranging study vs. haloperidol. The Amisulpride Study Group. Acta Psychiatrica Scandinavica 98, 65–72. Remington G, Kapur S (2000). Atypical antipsychotics, are some more atypical than others? Psychopharmacology 148, 3–15. Reynolds GP (1994). Antipsychotic drug mechanisms and neurotransmitter systems in schizophrenia. Acta Psychiatrica Scandinavica 89 (Suppl. 380), 36–40. Reynolds GP, Czudek C (1995). New approaches to the drug treatment of schizophrenia. Advances in Pharmacology 13, 461–503. Robertson GS, Fibiger HC (1996). Effects of olanzapine on regional C-Fos expression in rat forebrain. Neuropsychopharmacology 14, 105–110. Robertson GS, Matsumura H, Fibiger, HC (1994). Induction patterns of Fos-like immunoreactivity in the forebrain as predictors of atypical antipsychotic activity. Journal of Pharmacology and Experimental Therapeutics 271, 1058–1066. Saller CF, Salama AI (1986). D-1 dopamine receptor stimulation elevates plasma prolactin levels. European Journal of Pharmacology 122, 139–142. Seeman P, Tallerico T (1998). Antipsychotic drugs which elicit little or no parkinsonism bind more loosely than dopamine to brain D2 receptors, yet occupy high levels of these receptors. Molecular Psychiatry 3, 123–134. Seeman P, Tallerico T (1999). Rapid release of antipsychotic drugs from dopamine D2 receptors, an explanation for low receptor occupancy and early clinical relapse upon withdrawal of clozapine and quetiapine. American Journal of Psychiatry 156, 876–884. Sertindole UK Product Monograph (1997). Lundbeck Ltd. Skarsfeldt T (1992). Electrophysiological profile of the new atypical neuroleptic, sertindole, on midbrain dopamine neurones in rats, acute and repeated treatment. Synapse 10, 25–33. Stockton ME, Rasmussen K (1996). Electrophysiological effects of olanzapine, a novel atypical antipsychotic, on A9 and A10 dopamine neurons. Neuropsychopharmacology 14, 97–104. Swerdlow NR, Bakshi V, Geyer MA (1996). Seroquel restores sensorimotor gating in phencyclidine-treated rats. Journal of Pharmacology and Experimental Therapeutics 279, 1290–1299. Tandon R, Milner K, Jibson MD (1999). Antipsychotics from theory to practice, integrating clinical and basic data. Journal of Clinical Psychiatry 60 (Suppl. 8), 21–28. Trichard C, Paillere-Martinot ML, Attar-Levy D, Recassens C, Monnet F, Martinot JL (1998). Binding of antipsychotic drugs to cortical 5-HT2A receptors : a PET study of chlorpromazine, clozapine and amisulpride in schizophrenia patients. American Journal of Psychiatry 155, 505–508. Uchida S, Honda F, Otsuka M, Satoh Y, Mori J, Ono T,

The new generation of antipsychotic drugs Hitomi M (1979). Pharmacological study of [2-chloro-11-(2dimethylaminoethoxy)dibenzo[b,f]thiepine] (zotepine), a new neuroleptic drug. Arzneimittelforschung 29, 1588–1594. Vahid-Ansari F, Nakabeppu Y, Robertson GS (1996). Contrasting effects of chronic clozapine, Seroquel (ICI 204,636) and haloperidol administration on deltaFosBklike immunoreactivity in the rodent forebrain. European Journal of Neuroscience 8, 927–936. von Bardeleben U, Benkert O, Holsboer F (1987). Clinical and neuroendocrine effects of zotepine – a new neuroleptic drug. Pharmacopsychiatry 20, 28–34.

349

Wetzel H, Wiesner J, Hiemke C, Benkert O (1994). Acute antagonism of dopamine D2-like receptors by amisulpride, effects on hormone secretion in healthy volunteers. Journal of Psychiatric Research 28, 461–473. White FJ, Wang RY (1983). Differential effects of classical and atypical antipsychotic drugs on A9 and A10 dopamine neurons. Science 221, 1054–1057. Wirshing DA, Wirshing WC, Kysar L, Berisford MA, Goldstein D, Pashdag J, Mintz J, Marder SR (1999). Novel antipsychotics, comparison of weight gain liabilities. Journal of Clinical Psychiatry 60, 358–363.