Atypical Neuroleptics

Typical neuroleptics reduce the positive symptoms of schizophrenia at the expense of producing EPSs but the so-called atypical neuroleptics have less tendency to cause EPSs. With most of them, e.g. thioridizine, that is the extent of their atypicality but a few others, such as clozapine (and to a lesser extent risperidone and olanzapine) also reduce negative symptoms. Clozapine can even be effective in patients refractory to other neuroleptics. It is clearly a special drug, so special in fact that although it was once withdrawn because it causes agranulocytosis in some patients (2%), it has been reintroduced, alongside careful blood monitoring, for refractory cases. It will be given special consideration below.


There is certainly evidence that whereas typical neuroleptics are equally active in mesolimbic/cortical areas as well as the striatum, the atypical drugs are much less effective in the latter. This has been shown by (1) increased DA turnover through DOPAC and HVA production in vitro, (2) augmented DA and DOPAC release by microdialysis in vivo and (3) increased c-fos-like expression.

How the atypical neuroleptics achieve this differential effect is less clear but they could achieve some control of schizophrenia without producing EPSs by:

(1) Acting primarily on a particular subset of DA receptors

(2) Antagonising (or augmenting) some other NT(s) instead of, or in addition to, DA

(3) Having a particular but appropriate profile of DA and other NT (antagonistic) effects

These possibilities will be considered in turn. Significance of different DA receptors

So far we have generally just alluded to the neuroleptics as DA receptor antagonists. The reader will know that there are five such receptors (Chapter 7). Clearly, if the DA released at the terminals of one dopaminergic tract acted on a subset of DA receptors that were different from those found postsynaptically at other tracts then some specificity of antagonist action might be achieved. Unfortunately there is no evidence that different pathways innervate different DA receptor populations and as with the use of agonists in PD, the D2 receptor is dominant. Specific D1 antagonists have no anti-schizophrenic effect and antischizophrenic efficacy increases with neuroleptic affinity (potency) at D2 receptors — as unfortunately does the tendency to produce EPSs. Thus there is no great advantage in producing more potent D2 antagonists, other than that less drug needs to be incorporated into long-term release depot preparations.

PET studies show that at effective therapeutic plasma concentrations most neuro-leptics occupy some 80% of brain D2 receptors (in the striatum at least) and this is therefore considered to be a requirement for efficacy (Pilowsky, Costa and Eli 1992; Farde 1996). If that is so then clozapine, which occupies only 20-40% of the D2 receptors at a therapeutic concentration, must have some other action which accounts for its therapeutic effectiveness.

Its activity at D1 receptors has been put forward as a possibility and although it has a relatively higher affinity for D1 than D2 receptors, compared with typical neuroleptics, it is still a weak antagonist at both and in the absence of evidence for D1 (or D5) receptor involvement in schizophrenia the significance of any Di antagonism is unclear.

Ki (nM) values for clozapine at D2 and D1 receptors are 56 and 141 compared with 0.5 and 27 for haloperidol giving D1/D2 ratios of 2.5 and 54 for the two drugs. A relatively strong block of D1 compared with D2 receptors may not be the answer for schizophrenia but it could reduce the tendency to produce dyskinesias, if this depends on D1 receptor activation (see Fig. 17.2).

Among the D2 family of receptors (D2, D3 and D4) the D2 receptor itself seems to be the most important. At a therapeutic concentration, most neuroleptics, except clozapine (and risperidone), should, according to in vitro binding studies, be occupying 50-70% of brain D2 receptors. The picture is similar for D3 receptors but only clozapine (and risperidone and olanzapine) occupy more than 50% of D4 receptors at a therapeutic dose.

This relative selectivity of clozapine for D4 receptors with their restricted location, even if it is in small numbers, to the prefrontal cortex has stimulated much interest in their involvement in schizophrenia and the control of negative symptoms. There has been one report (Seeman, Guan and Van Tol 1993), refuted by others, of a sixfold increase in D4 receptors in schizophrenic brain. Unfortunately the measurements were made in striatum rather than cortex and depended on the difference in the binding of a D2, D3, D4 antagonist nemonopride compared with that of a D2 and D3 antagonist raclopride. D4 occupancy was thus inferred rather than established by a specific D4 antagonist. When such a selective D4 antagonist, L-745,870, became available and was tested in 38 schizophrenics it proved ineffective at what were considered to be doses sufficient to occupy 50% of the D4 receptors (Bristow et al. 1997). It has not been used apparently to assess D4 receptor number in schizophrenic brain.

There are few specific drugs for D3 receptors but D3 knock-out mice show no behavioural defects. Thus the significance of any DA receptor other than the D2 still remains to be established (see Seeman and Van Tol 1994; Sokoloff and Schwartz 1995; Strange 1994).

Involvement of other NTs


Neuroleptic-induced Parkinsonism (but not tardive dyskinesias) can be reduced by antimuscarinic drugs. It is generally assumed that neuroleptic antagonism of DAmediated inhibition in the striatum leaves the excitatory muscarinic action of ACh unchecked (Fig. 15.9) so that blocking it will restore normality. The atypical neuro-leptics thioridizine and clozapine both have potent inherent antimuscarinic activity with PA2 values (7-8) similar to that for atropine and more than a hundredfold that of a typical neuroleptic-like haloperidol. Thus each compound has the ability to nullify its own antidopamine effect in the striatum and stop Parkinsonian symptoms developing (Fig. 17.7) without affecting DA antagonism, and possible antischizophrenic effects elsewhere. There is no evidence that antimuscarinic activity has any effect on schizophrenia and thioridizine has no more effect on negative symptoms than typical neuroleptics. Of course, since clozapine is also a weaker D2 antagonist than thioridizine this automatically reduces its ability to produce EPSs anyway.


Some neuroleptics, including clozapine, are potent 5-HT-receptor antagonists and the possible role of 5-HT in the action of neuroleptics and the development of schizophrenia has recently generated much interest (Busatto and Kerwin 1997). This has centred primarily on 5-HT2A receptors found in the limbic cortex, which are linked to neuronal excitation and believed to mediate the hallucinogenic effects of drugs such as lysergic acid diethylamide (LSD).

Generally most atypical neuroleptics have higher affinity for 5-HT2 than D2 receptors while typical ones retain a preference for the D2 receptor. There is, however, no infallible separation since chlorpromazine (typical neuroleptic) is more active at 5-HT2A

Neuroleptic Drugs And Function

Figure 17.7 Possible mechanism by which atypical neuroleptics with antimuscarinic activity produce few EPSs. Normally the inhibitory effects of DA released from nigrostriatal afferents on to striatal neuron D2 receptors is believed to balance the excitatory effect of ACh from intrinsic neurons acting on muscarinic (M2) receptors (a). Typical neuroleptics block the inhibitory effect of DA which leaves unopposed the excitatory effect of ACh (b) leading to the augmented activity of the striatal neurons and EPSs (see Fig. 15.2). An atypical neuroleptic with intrinsic antimuscarinic activity reduces this possibility by counteracting the excitatory effects of released ACh as well as the inhibitory effects of DA (c). Thus the control of striatal neurons remains balanced

Figure 17.7 Possible mechanism by which atypical neuroleptics with antimuscarinic activity produce few EPSs. Normally the inhibitory effects of DA released from nigrostriatal afferents on to striatal neuron D2 receptors is believed to balance the excitatory effect of ACh from intrinsic neurons acting on muscarinic (M2) receptors (a). Typical neuroleptics block the inhibitory effect of DA which leaves unopposed the excitatory effect of ACh (b) leading to the augmented activity of the striatal neurons and EPSs (see Fig. 15.2). An atypical neuroleptic with intrinsic antimuscarinic activity reduces this possibility by counteracting the excitatory effects of released ACh as well as the inhibitory effects of DA (c). Thus the control of striatal neurons remains balanced than D2 receptors and remoxidpride (atypical) more active at D2 than 5-HT2A. Also differences in the values for the dissociation constants between experimental studies (see later) make comparisons between D2 and 5-HT2A potency somewhat difficult.

No neuroleptic has purely 5-HT2A antagonist activity and a pure 5-HT2A antagonist drug may not have neuroleptic activity. Risperidone, an atypical neuroleptic with some benefit against negative symptoms, is the most potent of all neuroleptics at 5-HT2A receptors (K\: 0.2 nM). Some in vitro measurements show it to have up to 25 times more affinity for 5-HT2A than D2 receptors and PET studies indicate that at therapeutic doses it displaces a 5-HT2 ligand in preference to a D2 one. Clozapine is also claimed to occupy over 80% of 5-HT2 and less than half this number of D2 receptors at clinical doses. Neuroleptics with 5-HT2 antagonist activity not only produce fewer EPSs but 5-HT2 antagonists reduce neuroleptic-induced EPSs.

Fibres from 5-HT neurons in the raphe nucleus innervate and yet, despite the observed 5-HT2A receptor link with neuronal excitation, appear to inhibit DA neurons in the SN (A9). Thus antagonism of 5-HT released onto them would increase their firing and so reduce the likelihood of EPSs, although how 5-HT2A antagonists can overcome the established motor side-effects of another neuroleptic is less clear if that compound has already caused a depolarising block of the neurons.

The mechanism by which 5-HT2 antagonism could ameliorate schizophrenic symptoms and what effect 5-HT has on mesolimbic and mesocortical pathways through A10 neurons is even less certain. It is more likely that 5-HT's action occurs postsynaptically in the limbic system or PFC. The probability that neuroleptics benefit from a particular balance of DA and 5-HT2A antagonism is developed later.

The 5-HT3 receptor is found appropriately in mesocortical areas and while behavioural studies with their antagonists in rodents showed potential antipsychotic activity, they have proved ineffective in patients. 5-HT1A agonists may be more useful. They have been found to increase the extracellular concentration of DA in the frontal cortex of rats but diminish apomorphine-induced stereotypy (striatal effect). So they could be of some benefit, especially against negative symptoms, without causing EPSPs (see Chapter 9).


Many of the neuroleptics are a-adrenoceptor antagonists. Some, like chlorpromazine, block a1 postsynaptic receptors while clozapine (and risperidone) are as potent at a2 as D2 receptors. There is no evidence that either of these actions could influence striatal or mesolimbic function but NA is considered important for function of the prefrontal cortex and any increase in its release, achieved by blocking a2-mediated autoinhibition, might contribute to a reduction in negative symptoms and provide a further plus for clozapine (see Nutt et al. 1997). Centrally, however, most a2-receptors are found post-synaptically and their function, and the effect of blocking them, is uncertain.


Although there is no evidence that any of the neuroleptics have any significant effect on glutamate receptors, it will be of no surprise to learn that clozapine, but not pure D2 or 5-HT2 antagonists nor any typical neuroleptic, can overcome phencyclidine disruption of PPI in animals. Interestingly, the efficacy of clozapine (but not risperidone or olanzopine) is increased by the antiepileptic drug lamotrigine that has inhibition of glutamate release as one of its actions (see Chapter 16). Also glycine (and serine) have been shown to improve the negative symptoms by what is assumed (but not proven) to be a potentiation of NMDA receptor activity, but they can make positive symptoms worse.

Profile of NT antagonism in neuroleptic action

In deciphering the role of the different NTs, or more precisely their antagonists, in the antischizophrenic action of neuroleptic drugs it must be remembered that published binding data and calculated dissociation constants vary considerably, which, of course, affects correlation coefficients made with clinical activity. Factors to bear in mind are:

(1) In vitro binding studies use different cell lines or membrane preparations and generally only yield the apparent dissociation constants for a number of antagonists obtained by comparative displacement of one labelled ligand. Unfortunately few such ligands are specific for the receptor being analysed, i.e. they bind to other receptors to differing extents as do the displacing compounds. Reported values for clozapine's binding affinity vary from 84-388 nM depending on the D2 ligand being displaced. Real dissociation constants can be obtained from direct measurements of the binding of the neuroleptic alone in labelled form but because neuroleptics also bind to more than one receptor, the preparation must express only the receptor being studied.

(2) PET studies have almost always centred on measurements of binding and DA receptor number in the striatum rather than other DA-innervated areas of more significance in schizophrenia. Also in PM measurements of receptor number it is invariably the striatum which is used, because of its high density of DA receptors.

(3) Functional activity (clinical effect, catalepsy in animals, etc.) is invariably correlated with plasma concentrations whereas the brain levels of many neuroleptics, which are very lipophilic compounds, could be much higher. Some clinicians also believe that many newer compounds achieve atypical status compared with older ones because they are used at minimal dosage while older ones are prescribed at established levels which may be unnecessarily high.

Despite these problems it remains necessary to attempt some explanation in terms of differential NT antagonism, of why clozapine is so effective (see Reynolds 1997) in that it causes fewer EPSs, reduces negative symptoms and is effective in some patients refractory to other drugs. Considering these benefits in turn:

(1) Reduced EPSs. This may be achieved with clozapine because it is a:

(a) Relatively weak D2 antagonist. The one thing that is reasonably certain about the neuroleptics is that irrespective of the role of D2 antagonism in controlling schizophrenia the more potent the D2 antagonist, the more likely are EPSs. Just as Parkinsonian symptoms only occur in PD patients when 50-80% of the DA innervation to the striatum is lost (Chapter 15) so neuroleptic-induced Parkinsonism only follows blockage of some 80% of D2 receptors. Clozapine only achieves about half of this at therapeutic doses and its weak binding may allow DA to override its antagonism at appropriate times in the striatum. Thus clozapine has little potential for inducing EPSs and what it has could be reduced by its other activities.

(b) Potent antimuscarinic. ACh excitation counteracts DA inhibition in the striatum.

(c) Strong 5-HT2 antagonist. Compounds with this property appear to reduce EPSs.

(d) Relatively strong Dj antagonist. This may not stop PD symptoms but could reduce dyskinesias (Fig. 15.8).

As a result of these features clozapine is likely to have little effect on A9 (SN) neurons and does not cause their depolarisation in chronic dosing.

(2) Negative symptoms. These may be reduced because either clozapine antagonises appropriate receptors in the prefrontal cortex or it does not act as an antagonist there. This apparently stupid statement is prompted by the lack of knowledge of what is required to reduce negative symptoms. D4 and Dj receptors are found in the prefrontal cortex and only clozapine among current neuroleptics is more active at both of these than the D2 receptor. Thus on this basis it is well placed to block DA's influence but if negative symptoms follow an impairment of DA input (see above) further blockage is undesirable. In fact clozapine would have to augment DA function and based on the knowledge that D1 receptor activation appears to be required for optimal cognitive performance it has been suggested that neuroleptics should optimise activation of D1 receptors in addition to blocking D2 receptors (Lidow, Williams and Goldman-Rakic 1998). Little is known of the effect of DA or its agonists on cortical neurons, although most studies show it to be inhibitory. Even less is known about clozapine's action on neuronal firing but in one study on the prefrontal cortex of anaesthetised rats it was found to mimic the action of the DA agonist apomorphine, an effect blocked by haloperidol (Dalley and Webster 1993). A number of microdialysis studies have also shown that it is the only neuroleptic to increase DA efflux in the prefrontal cortex although most of them have that effect in the striatum. So perhaps clozapine can in some way increase DA transmission in PFC, even if that is achieved through initially antagonising an effect of DA or another NT. Recently risperidone has also been shown to increase both 5-HT and DA release in the rat prefrontal cortex (Knoble and Weinberger 1997) but possibly through a2 and 5-HT receptor antagonism. In view of the strong antimuscarinic activity of clozapine it is interesting that cholinergic overactivity has been reported to induce behaviour in animals that was thought to reflect negative symptoms.

(3) Refractory cases respond to clozapine. If D2 antagonism is considered necessary, or at least desirable, for counteracting positive symptoms it is surprising that a relatively weak D2 antagonist like clozapine should not only be so effective but also prove successful in patients who have not responded to other neuroleptics more potent at D2 receptors.

Certainly clozapine can avoid EPSs by only blocking a fraction of D2 receptors but that seems insufficient on its own to make clozapine so effective in schizophrenia. That is probably achieved by a unique combination of other blocking actions, at D1, D4, 5-HT2, a2 and possibly other receptors (see Fig. 17.8). It may simply be that clozapine is so effective because it is so 'dirty', a view held for many years about the first neuroleptic chlorpromazine. Indeed it is unlikely that the varied symptoms of such a complicated disorder could be rectified by manipulating just one NT.

Unfortunately although much is known about the pathways and receptors involved in extrapyramidal activity and the mechanism of the EPSs that follow neuroleptic therapy and even the possible origin of negative symptoms in the prefrontal cortex, the precise site of origin and NT involvement in the overriding positive symptoms is less clear. Until that is corrected, permutations of NT antagonisms are likely to multiply with the neuroleptics.

Continue reading here: Perspective

Was this article helpful?

0 0