Treatment with Antipsychotics

As in other areas of clinical science, serendipity as well as careful investigation contributed to our knowledge about antipsychotics, including their

Discovery
Potential mechanisms of action
Range of complications

The efficacy of chlorpromazine (CPZ) was discovered primarily by chance in exploratory clinical trials after it had been initially synthesized as an antihistamine. Its discovery, however, was not entirely fortuitous, as it was chosen for human investigation because it was mildly sedating. The concept of an antipsychotic, however, was unknown. The sedative properties of CPZ then led the French anesthesiologist and surgeon Henri Laborit to use it in a lytic cocktail to reduce autonomic response with surgical stress (1). He also persuaded many clinicians to try it for the treatment of a wide variety of other disorders. In this context, he encouraged Delay and Deniker (2) who then administered CPZ to psychotic patients. The rest is history.

Even though CPZ did not produce a permanent cure for schizophrenia (or for any other psychotic disorder), it had a dramatic impact, benefiting many as no other treatment had before (3). News of its effectiveness spread rapidly, and within 2 years CPZ was used worldwide, altering the lives of millions of psychotic patients in a positive manner.

The relocation of treatment from chronic hospital settings to outpatient community mental health centers is, in great part, due to the efficacy of antipsychotics. Naturalistic studies before the era of psychotropics revealed that two of three psychotic patients (primarily schizophrenia patients) spent most of their lives in state asylums. Before the mid-1950s, there had been a steady increase in state hospital populations, which paralleled the general population growth, but after the introduction of antipsychotics, there was a marked reduction in those hospitalized for various psychoses. Presently, more than 95% of these patients live outside of the hospital, even though many continue to relapse or demonstrate residual symptoms. Thus, although the antipsychotics are not a panacea, they make community-based care a reality for many who would otherwise remain chronically institutionalized.

Reserpine, used as a folk medicine in India, was found to have antipsychotic properties at about the same time as CPZ. Both agents affected the dopaminergic system, albeit in different ways, but the functional results were similar [i.e., lowering dopamine (DA) activity]. This phenomenon continues to be an important factor in hypotheses about the mechanism of action of these drugs and for biological theories about the pathophysiology of psychotic disorders.

The revolutionary idea that drugs could exert a specific antipsychotic effect, coupled with a growing recognition of their significant drawbacks, led to the search for other substances with similar beneficial properties and fewer adverse effects. Thus, pharmacological animal screens, followed by systematic clinical investigations with chemically or structurally related compounds, led to a variety of effective, better tolerated drugs for schizophrenia and other major psychotic disorders.

The therapeutic efficacy of clozapine, the first truly different antipsychotic, spawned the development of a succession of newer agents (e.g., risperidone [RISP], olanzapine [OLZ], quetiapine [QTP], ziprasidone [ZPD], aripiprazole [ARIP], paliperidone [PALI], iloperidone [ILP], and asenapine [ASEN]). Other new compounds (e.g., bifeprunox, lurasidone, sertindole) may also receive US Food and Drug Administration (FDA) marketing approval in the near future (3a, 3b). Antipsychotics can be classified by their chemical class, potency, activities at receptors, or liability for causing neurological adverse effects. This text refers to these drugs as first-generation antipsychotics (FGAs) and second-generation antipsychotics (SGAs). FGAs include all of the agents marketed before clozapine. Currently marketed SGAs worldwide include clozapine, sulpiride, amisulpride, zotepine, RISP, OLZ, QTP, ZPD, ARIP, PALI, ILP, and ASEN. Although these agents have previously been referred to as atypical or novel, these terms seem dated, as SGAs are the most commonly prescribed antipsychotic agents in the United States, Canada, and most other countries.

Table 5-1 presents the classes, trade and generic names, and typical dose ranges for the antipsychotics available in the United States. Figures 5-1 and 5-2 present the basic chemical structure of antipsychotic agents.

Pharmacodynamics

All antipsychotics work in part through the central DA-2 (D₂) receptor (most cause blockade, but some [e.g., ARIP and bifeprunox] are partial agonists at the D₂ receptor). Extrapyramidal reactions, particularly parkinsonian symptoms, can be a major adverse effect of many of these drugs as well as an important clue to their mechanism of action. True Parkinson’s disease is caused by a DA deficiency in the nigrostriatal system. Further, crystallographic data have demonstrated that the molecular configuration of CPZ is similar to that of DA, which could explain its ability to block the receptors of this neurotransmitter. Drugs with similar structures that do not block DA receptors (e.g., promethazine, imipramine) do not have antipsychotic activity. For example, the isomer of flupenthixol that blocks DA receptors is an effective antipsychotic, whereas the isomer that does not is ineffective (4). Other DA receptors, D₁, D₃, D₄ and D₅, may also be implicated in psychosis.

TABLE 5-1 ANTIPSYCHOTIC AGENTS

Class/Trade Name			Generic Name	Dosage (Range, Orally, Per Day)
Phenothiazines
	Aliphatics
		Thorazine	Chlorpromazine	100-1,000 mg
		Sparine	Promazine	25-1,000 mg
		Vesprin	Triflupromazine	20-150 mg
	Piperidines
		Mellaril	Thioridazine	30-800 mg
	Serentil		Mesoridazine	20-200 mg
		Quide	Piperacetazine	20-160 mg
	Piperazines
		Stelazine	Trifluoperazine	2-60 mg
		Prolixin	Fluphenazine	5-40 mg
		Trilafon	Perphenazine	2-60 mg
		Tindal	Acetophenazine	40-80 mg
		Compazine	Prochlorperazine	15-125 mg
Thioxanthenes
	Navane		Thiothixene	6-60 mg
	Taractan		Chlorprothixene	10-600 mg
Dibenzoxazepines
	Loxitane		Loxapine	20-250 mg
Butyrophenones
	Haldol		Haloperidol	3-20 mg
	Inapsin		Droperidol	2.5-10 mg^a
Dihydroindolones
	Moban		Molindone	15-225 mg^b
Dibenzodiazepines
	Clozari		Clozapin	100-900 mg
Benzisoxazole
	Risperdal		Risperidone	2-8 mg
	Invega		Paliperidone	3-12 mg
	Fanapt		Iloperidone	12-24 mg
Thienobenzodiazepines
	Zyprexa		Olanzapine	5-20 mg
Dibenzothiazepines
	Seroquel		Quetiapine	75-800 mg
Dibenzo-oxepino pyrroles
	Saphris		Asenapine	10-20 mg
Benzisothiazolyls
	Geodon		Ziprasidone	40-160 mg
Quinolinones
	Abilify		Aripiprazole	5-30 mg
Diphenylbutylpiperidines
	Orap		Pimozide	1-10 mg
^aAdministered intramuscularly ^b“Recently taken off the market.”

Interestingly, some SGAs are nearly devoid of extrapyramidal adverse effects in their usual clinical dosing range. This is an important observation because it means that these two
effects can be dissociated. Initially, the term neuroleptic was applied to all drugs that produce both extrapyramidal and antipsychotic effects, but agents such as clozapine, RISP, OLZ, QTP, ZPD, ARIP, and others can be antipsychotics without being neuroleptics, making the interchangeable use of the terms neuroleptic and antipsychotic no longer appropriate.

Figure 5-1 Basic requirements for antipsychotic activity with first-generation antipsychotics (FGAs); distance between N₁₀ of radical and the base N atom must equa at least three carbons (n = 3), with a suitable substituent R₂ mainly determining the qualitative properties of the group; the A₂ substituent R₁ mainly determines the quantitative properties of the individual compounds.

The evidence that antidopaminergic agents ameliorate psychosis and that full DA agonists can produce psychosis or worsen preexisting disorders supports a central role for DA in any theory of pathophysiology. Further, there is a high correlation among directly measured D₂-type receptor binding, the clinical potency of these agents, animal behavioral models, and data about DA blockade.

Figure 5-2 Chemical structures of second-generation antipsychotics (SGAs).

Other drug effects that decrease DA activity also support this position. Thus, when DA synthesis is blocked by α-methyl-p-tyrosine (AMPT), the dose necessary for an antipsychotic effect is reduced (i.e., the dose-response curve is shifted to the left by the interaction between DA and AMPT). A drug such as reserpine that can deplete DA stores has relatively mild antipsychotic properties. Also, partial DA agonists (e.g., ARIP, bifeprunox) bind to DA receptors with higher affinity than endogenous DA but result in lower receptor activity. As a result, under conditions of high activity, they functionally decrease the effects of DA (5).

DOPAMINERGIC PATHWAYS OF THE CENTRAL NERVOUS SYSTEM

In 2000, Arvid Carlsson, Paul Greengard, and Eric Kandel received the Nobel Prize in Physiology or Medicine for their study of nerve cell communication (6). Their work included studies of the neurotransmitter DA and its role in Parkinson disease and schizophrenia. Because the existing evidence strongly implicates the DA system in
psychosis, it is pertinent to discuss the central pathways of this neurotransmitter. There are five tracts:

The striatal system (A-9)
The mesolimbic system (A-10)
The mesocortical system (A-10)
The retinal system
The neurohypophyseal system

The last system is of interest in regard to elevation in prolactin, an effect most closely associated with FGAs, RISP, and PALI.

Postmortem studies of patients with idiopathic Parkinson’s disease demonstrate cell loss in the striatal system (A-9), directly implicating this tract vis-a-vis the drug-induced pseudoparkinsonian adverse effects. The assumption that psychosis is related to the A-10 system is made by exclusion. Evidence also indicates that clozapine may differentially block DA pathways. Specifically, it seems to act on the mesolimbic dopaminergic system (A-10), while being relatively inactive in the striatal system (A-9); however, this remains controversial. Chronic administration of clozapine decreases the firing rate of A-10 mesocortical tract DA neurons while sparing A-9 neurons of the nigrostriatal pathway. By contrast, FGAs reduce the firing rate of both A-9 and A-10 neurons.

Because clozapine may block specific DA receptors, its antipsychotic activity could be consistent with an antidopaminergic mechanism of action. Conversely, clozapine does not typically induce extrapyramidal symptoms, which are presumably subserved by the A-9 system. Thus, although clozapine is known to block striatal DA receptors, in positron emission tomography (PET) studies, resolution is not sufficient to clarify effects on other tracts. Furthermore, low doses of metoclopramide, which significantly decrease the number of DA neurons spontaneously active in A-9, do not have antipsychotic effects (although high doses do) but can induce tardive dyskinesia (TD) as well as acute extrapyramidal side effects (EPSs).

NEUROPHYSIOLOGICAL STUDIES

Another way to measure striatal (A-9) or mesolimbic (A-10) activity is through neurophysiological studies involving the firing rate of these neurons. The acute administration of antipsychotics will cause DA neurons to initiate a burst pattern of firing, with the opposite occurring when a DA agonist is administered. Eventually, some degree of tolerance develops, and the enhanced DA synthesis decreases with time on drug, as demonstrated in both human cerebrospinal fluid (CSF) studies and rat brain slices. On a more chronic basis, repeated drug administration results in a dramatic decrease in the proportion of spontaneously active dopaminergic neurons (i.e., excitation-induced depolarization blockade of the DA neuron spike-generating region). Both the depolarization blockade and the antipsychotic effects of these drugs develop slowly. FGAs induce nigrostriatal and mesolimbic DA neurons into depolarization blockade, whereas clozapine induces depolarization only in the mesolimbic-mesocortical DA neurons.

The working assumption that the striatal system is involved only with extrapyramidal function (e.g., parkinsonian adverse effects, dystonias, TD) and the mesolimbic or mesocortical system is involved only with psychosis may be an oversimplification. Many of the neuroanatomical studies on the identified dopaminergic tracts are done with rats. In the monkey, by contrast, there are many more DA tracts that are either absent in the rat or at least markedly different. Human systems could be different from those of the rat or the monkey. Understanding the neuropharmacology of the antipsychotics is further complicated by the fact that neither the mesolimbic-mesocortical nor the striatal system is homogeneous but may include various subsystems.

DOPAMINE AUTORECEPTORS

There are also dopaminergic presynaptic receptors (or autoreceptors) that generally play a negative feedback role. These autoreceptors sense the level of neurotransmitter and produce a negative feedback influence on DA synthesis and release. Thus, high DA release results in high intrasynaptic concentrations, which stimulate the autoreceptors, ultimately inducing a slowing of DA synthesis and release. Conversely, blockade of the postsynaptic receptors may lead to positive feedback loops, which can

Activate tyrosine hydroxylase in the presynaptic dopaminergic neuron
Increase DA synthesis
Increase the firing rate
Increase the levels of DA metabolites such as homovanillic acid (HVA)

Further complicating the picture is the induction of autoreceptor supersensitivity with chronic antipsychotic administration, as well as the fact that at least one tract—the mesocortical, which projects to the prefrontal cortex—may lack such autoreceptors. Schaffer, et al conducted preliminary, single-dose studies with apomorphine at a dose that stimulated presynaptic DA autoreceptors reducing synthesis and neuroleptic threshold (NT) release. This produced a measurable acute antipsychotic effect (7).

DECREASING DOPAMINERGIC ACTIVITY

Given the typical time course of adaptation to the biochemical and electrophysiological effects induced by antipsychotics in most dopaminergic systems, it is important to consider the time course for their efficacy. Apparent clinical benefit can be seen within a few hours after treatment is initiated, improves at a roughly linear rate for the first 14 days, and then gradually levels off.

Time-dependent changes in plasma HVA (a major metabolite of DA) during antipsychotic treatment are consistent with the hypothesis that the mechanism of these agents involves a slowly developing decrease in presynaptic DA synthesis and release. Thus, there is an initial increase in HVA followed by a more sustained decrease. Although decreases in plasma HVA appear to parallel the time course of antipsychotic efficacy, peripheral DA systems may also contribute to plasma or urine HVA levels. Further, because different brain areas may or may not develop tolerance, it is quite possible that CSF HVA levels reflect those areas that produce most of the HVA but are not relevant to the psychotic process. With these caveats in mind, human CSF and plasma data are generally consistent with animal studies, both of which contribute considerable evidence implicating DA blockade and its aftermath. The relevance of these findings to the biological mechanisms subserving psychosis, however, is yet to be determined.

TOLERANCE

It is important to examine whether tolerance develops to the antipsychotic effect of these agents. For example, Sharma et al. (8) found that nontolerant psychotic patients had a significantly inferior clinical response to antipsychotics, in contrast to their tolerant counterparts. Further, they found an earlier age of illness onset and a more refractory course in the nontolerant group. Although biochemical and electrophysiological tolerances (i.e., return of HVA to more normal levels and depolarization inactivation, respectively) develop in most DA systems, they do not occur in all. Prefrontal and cingulate cortices, for example, may be spared, perhaps because of the absence of autoreceptors in these tracts. Evidence also indicates that many areas in nonhuman primate brains develop biochemical tolerance to antipsychotic-induced HVA rise but other areas do not. Thus, with chronic treatment, certain areas (cingulate, dorsofrontal, and orbitofrontal cortex) maintain their HVA increases. Patients who were examined postmortem after chronic antipsychotic therapy were found to have increased HVA levels in certain areas, such as the cingulate and perifalciform cortex. It is also interesting that stress and benzodiazepine (BZD) receptor agonists can selectively stimulate DA neurons in the mesoprefrontal area but not in other dopaminergic tracts.

By using D₂-blocking isotopes in a PET scanner, it is possible to image the striatum in psychotic patients. Evidence from an early PET scan study indicated that some drug-free schizophrenia patients manifested an increase in D₂ DA receptors, but another study failed to find such an alteration (9,10). These studies used different receptor ligands, different assumptions, and both had a small sample size, so definitive conclusions are not possible. Finally, when PET scans are done on patients receiving clinically effective doses of FGAs or SGAs, all of these agents (including haloperidol [HPDL], clozapine, RISP, OLZ, and ZPD) produced varying levels of D₂ receptor occupancy in the striatum (11,12 and 13). Furthermore, they do so at the typical doses used to treat schizophrenia. In addition, Kapur et al. (13) found that only transiently high D₂ receptor occupancy occurred with QTP, suggesting that this was sufficient for efficacy, while minimizing the possibility of EPS or elevated prolactin levels.

INCREASING DOPAMINERGIC ACTIVITY

If decreasing dopaminergic activity benefits psychosis, what is the effect of increasing dopaminergic activity? Potentiation of DA by a variety of mechanisms is a common denominator for
inducing paranoia, hallucinations, and other manifestations of psychosis. For example,

L-Dopa (3,4-dihydroxyphenylalanine), which is converted to DA in the body, can produce psychosis as a side effect
Amantadine, another dopaminergic drug, can also induce psychotic symptoms
Stimulants, such as amphetamine and methylphenidate, are potent releasers of DA, whereas cocaine also interferes with its reuptake, increasing DA concentrations at synaptic sites; all can produce paranoid reactions in some abusers
Bromocriptine, apomorphine, lisuride, and other direct-acting DA agonists benefit Parkinson’s disease and can also cause psychotic reactions at high doses

As noted, large doses of amphetamine, cocaine, and other sympathomimetics can cause acute paranoid reactions, either spontaneously in abusers or experimentally in normal volunteers. An injection of a large amphetamine dose in normal control subjects, for example, often produces a paranoid psychosis resembling schizophrenia within hours. Frequent smaller doses over several days can also produce a paranoid psychotic reaction. The duration of an episode usually parallels the length of time the drug remains in the body.

Additional evidence comes from studies of increasing dopaminergic activity in patients with active psychosis. Small intravenous doses of methylphenidate (e.g., 0.5 mg/kg) can result in a marked exacerbation of an acute schizophrenic episode (8). By contrast, such doses usually do not produce psychotic symptoms in normal control subjects or in remitted patients. Methylphenidate is more potent in this regard than dextroamphetamine, consistent with the hypothesis that it intensifies psychotic symptoms by releasing central intraneuronal stores of norepinephrine and DA from the reserpine-sensitive pool (14).

Finally, sensitivity to an amphetamine test dose in recovered patients may predict an impending relapse (15).

OTHER PATHWAYS

Interestingly, when physostigmine, a drug that increases brain acetylcholine (ACh), is administered before methylphenidate, it prevents the exacerbation, indicating that the worsening may be mediated by a dopaminergic-cholinergic imbalance (16). Physostigmine itself, however, does not reduce psychosis, suggesting that the underlying process is not amenable simply to altering cholinergic tone. Subsequently, nicotinic cholinergic receptors were reported to play an important role in the inhibition of the perception of extraneous environmental stimuli. This observation has implications for the deficits in sensory gating reported in schizophrenia (17).

Other systems (e.g., neurotransmitter, neuropeptide, neurohormonal) may also play important central or modulating roles in terms of the psychotic process or adverse effects. In addition to DA and ACh, the following have been implicated:

Serotonin (5-HT)
Norepinephrine (NE)
γ-Aminobutyric acid (GABA)
Glutamate (GLU)
Neurohormones
Neuropeptides

Table 5-2 provides a qualitative summary of the relative affinities (or binding profiles) of SGAs for various relevant neuroreceptors.

One leading hypothesis as to why newer agents produce fewer EPS and perhaps provide greater benefit for negative symptoms is that they have a more potent effect on the serotonin 5-HT_2A receptor subtype. Thus, the ratio of 5-HT_2A to D₂ blockade has become an important measure of the potential “atypicality” of an agent. In addition, these novel antipsychotics have a neuroreceptor profile that differs substantially from that of the FGAs. For example, clozapine is a dibenzothiazepine derivative with affinity for a variety of neurotransmitter receptors, including several serotonin receptors (5-HT_1C, 5-HT₃, 5-HT₆, and 5-HT₇) that may modulate DA activity and contribute to the atypical action of this agent. It is likely that clinically effective doses of clozapine also antagonize 5-HT_2A and 5-HT_2C receptors (18). The resulting increased 5-HT_2A/D₂ ratio may be one basis for the unique antipsychotic efficacy of clozapine (19). At clinical doses, clozapine also has a lower affinity for D₂ receptors, which may account for its decreased propensity to evoke EPS, as well as its negligible neuroendocrine effects (20,21). Other agents (e.g., ZPD,
ARIP) are partial agonists at the 5-HT_1A receptor, which may also modulate DA activity.

TABLE 5-2 NEURORECEPTOR BINDING PROFILE OF SECOND-GENERATION ANTIPSYCHOTICS^a

Receptor	Clozapine	Risperidone	Olanzapine	Quetiapine	Ziprasidone	Aripiprazole	Paliperidone	lloperidone	Asenapine
D₁	High	Low	High	Low/mod	Low	Low	Low	Low	High
D₂	Low	High	High	Low/mod	High	Very high	High	High	High
D₃		High	High		High	High	High	High	High
D₄	High	High	High	None	Mod	Low	High	Mod	High
D₆		Mod					Mod
D₇		High					High
5-HT_1A				Low/mod	High	High		Low	High^*
5-HT_2A	High	High	High	Low/mod	High	High	High	High	High
5-HT_2A/D₂ ratio	High	High	High	High	Very high	Low	High	High	High
5-HT_2C	High				High	High			High
5-HT3	High		High		Very low
5-HT₁₀					High
5-HT₆	High		High		High			Mod	High
5-HT₇		High			High		High	Mod	High
Alpha₁	Mod/high	High	Mod/high	Mod/high	Mod	Mod	High	Mod	High
Alpha₂	Mod	High	Low	Mod/high	Very low		High		High
Histamine₁	High	None	High	High	Mod	Low		Low	High
Muscarinic₁	High	Very low	High	None	Very low	Very low	Very low	Very low	Very low
^aBased on K₁ data ^*Also High affinity for 5-HT_1B; 5-HT_2B; 5-HT₅.
Mod, moderate.

Pharmacokinetics

Clinically relevant pharmacokinetic factors involving antipsychotics include

Good absorption from the gastrointestinal tract
An extensive “first-pass” hepatic effect
Subsequent high systemic clearance due to a large hepatic extraction ratio each time the plasma recirculates through the liver
Extensive distribution (V_D) due to highly lipophilic character
Plasma half-life (t_1/2) of typically about 20 hours
Primary route of elimination through hepatic metabolism (except PALI)
The presence of metabolites with varying pharmacological profiles (e.g., some may be more effective than their parent compound [e.g., mesoridazine], some may not reach the brain [e.g., sulfoxides], and some may have greater toxicity than the parent compound).

Table 5-3 lists the relevant pharmacokinetic properties of several SGAs.

THERAPEUTIC DRUG MONITORING

The use of drug plasma levels to effect optimal clinical response and to minimize adverse or toxic effects is standard practice in general medicine (e.g., phenytoin, digoxin) as well as in psychiatry (e.g., lithium, tricyclic antidepressants [TCAs], valproate [VPA]; see Chapter 2). The theoretical basis for plasma level monitoring rests on several factors, including

The existence of a long interval between drug administration and clinical response
Large interindividual differences in response to the same dose for the same diagnosis, in part reflecting differences in plasma levels achieved on a given dose
Possible usefulness in helping to establish the minimally effective dose
Possible usefulness in helping to estimate the average dose required to achieve a certain concentration when a positive correlation exists between a given steady state concentration (C_ss) and the dose required.

Unfortunately, for a number of methodological and clinical reasons, the success achieved with this strategy for some classes of drugs (e.g., certain mood stabilizers) has not been achieved with the monitoring of antipsychotic steady state plasma concentrations. Difficulties in study design have contributed to this uncertainty, including

Insufficient sample sizes, especially at the low and the high ranges
Inclusion of refractory, nonhomogeneous, nonadherent patients
Nonrandom adjustment of the dose based on response, adverse effects, or initial presentation
Concurrent treatments
Too brief an observation period for clinical effects to occur
Variable time of blood sampling
Inadequate evaluation of clinical response as well as the indefinite nature of the response
Presence of numerous active metabolites
Inadequate assay methods, such as radioreceptor assays

EXTRAPOLATION FROM PLASMA TO BRAIN LEVELS

It would be desirable to measure central nervous system (CNS) antipsychotic levels in human beings. Although this is possible with emerging imaging techniques, expense and technological barriers currently prohibit their routine clinical use. Direct measurements can also be made on postmortem tissue of patients who had recently discontinued the drug or were on antipsychotics at the time of death (22). HPDL is concentrated
in the brain at levels about 10 to 30 times higher than the plasma level. Animal studies find a brain-to-plasma ratio of about 20:1. Limited information from animal studies after drug withdrawal also shows a more rapid disappearance from plasma than from the brain. Although the half-life of HPDL is about 24 hours in plasma, even several days after cessation of oral dosing, significant CNS D₂ receptor activity is still evident. Indeed, PET studies indicate that after withdrawal of HPDL or fluphenazine decanoate, D₂ blockade can persist for several months. Thus, it is likely that human brain levels are approximately 10 to 20 times higher than plasma levels and that these agents stay in the brain longer than in plasma.

TABLE 5-3 PHARMACOKINETICS OF SECOND-GENERATION ANTIPSYCHOTICS

	Clozapine	Risperidone	Olanzapine	Quetiapine	Ziprasidone	Aripiprazole	Paliperidone	lloperidone	Asenapine
Absorption	Well absorbed	Rapid and complete	Well absorbed	Rapid	Rapid with food	Not affected by food	Well absorbed	Well absorbed	Well absorbed
Time to peak concentration	1-6 hours	1-3 hours	5 hours	About 1.5 hours	3.8-5.2 hours	3 hours	24 hours	2-4 hours	0.5-1.5 hours
Serum elimination half-life	12 hours	About 20 hours	20-70 hours	3-7 hours	5-10 hours	75 hours	23 hours	18 hours	24 hours
Time to C_ss	3-4 days	5-7 days	7 days	2-3 days	1-3 days	14 days	4-5 days	3-4 days	3 days
Protein binding	97%	89%; 77% for metabolite	93%	80%	>99%	98%	74%	95%	95%
Metabolism	CYP 1A2, 3A4	CYP 2D6	CYP 1A2, 2D6	Primarily by CYP 3A4	CYP 3A4	CYP 2D6, 3A4	Not clinically relevant	CYP 2D6, 3A4	UGT 1A4 CYP 1A2
Active metabolite	Norclozapine	9-OH-RISP (paliperidone)	None	None	None	Dehydro-ARIP	None	P 88	Activity related to parent drug
CYP, cytochrome P450; RISP, risperidone; ARIP, aripiprazole.

PLASMA LEVEL STUDY DESIGNS

Although large interindividual variability in steady state plasma concentrations among patients treated with similar doses of a given antipsychotic is well established, the existence of a critical range of plasma concentrations for therapeutic response or significant adverse effects remains controversial. A body of data from a number of fixed-dose studies, however, indicates a possible threshold for response or a linear or curvilinear relationship between plasma levels and clinical response for agents such as

CPZ (23)
Fluphenazine (24,25 and 26)
Trifluoperazine (27)
Thiothixene (28)
HPDL (29,30,31,32 and 33)
Clozapine (34,35)
RISP (36,37,38 and 39)
OLZ (40,41 and 42)
QTP (43)
ARIP (44)

Data comparing plasma concentrations with clinical response for ZPD and more recent SGAs are not currently available (45).

In addition, prospective studies in large numbers of acutely-ill patients targeting certain plasma levels to test a putative therapeutic threshold or range have been conducted with HPDL (46).

Chlorpromazine

Curry et al. (47,48) found a wide range of effective plasma drug levels in schizophrenia patients treated with comparable doses of CPZ, establishing that an upward or downward shift of 50% in dose usually produced adverse effects or a psychotic exacerbation, respectively. May et al. (49) found no relationship between plasma levels and response in 48 patients on fixed doses of CPZ (6.6 mg/kg/day). Effective plasma drug levels in this agent are particularly difficult to evaluate, however, because of its many potentially confounding metabolites, which are typically not measured.

Fluphenazine

Several studies investigated fluphenazine plasma levels with either the oral or the parenteral long-acting (LA) form. In one, clinical response as a function of the mean C_ss of oral fluphenazine suggested a therapeutic upper end based on three nonresponding patients who had a mean C_ss above 2.8 ng/mL (24). Further, a low end was suggested by two nonresponders and one partial responder, whose levels were below 0.2 ng/mL. Van Putten et al. (26) found that higher fluphenazine plasma levels (up to 4.23 ng/mL) were significantly associated with a higher rate of improvement; however, 90% (65 of 72 patients) experienced disabling adverse effects with levels greater than 2.7 ng/mL. In a 2-year, double-blind comparison of 5 or 25 mg of fluphenazine decanoate, Marder et al. (50) found a significant relationship between fluphenazine plasma levels and psychotic exacerbations after 6 to 9 months of maintenance therapy. Thus, those with levels less than 0.5 ng/mL did much worse than those with levels more than 1.0 ng/mL. These data suggest that the ideal plasma level range for many patients may be between 1.0 and 2.8 ng/mL. However, in a later study, Marder et al. (51) attempted to randomize patients into different plasma concentration ranges of fluphenazine. The results failed to support this strategy.

Trifluoperazine

We reported on a potential therapeutic window with the commonly used phenothiazine trifluoperazine (27). Specifically, there was evidence for a lower therapeutic threshold, around 1 ng/mL, and a suggestion of an upper end around 2.3 ng/mL (Fig. 5-3). Based on a review of dose-response studies, we concluded
that 9 to 15 mg of trifluoperazine would be comparable to 300 mg of CPZ and should fall near the lower part of the linear portion of the dose-response curve.

Figure 5-3 Brief Psychiatric Rating Scale (BPRS) change scores in relationship to trifluoperazine plasma levels. (From Janicak PG, Javaid JI, Sharma RP, et al. Trifluoperazine plasma levels and clinical response. J Clin Psychopharmacol. 1989;9:340-346, with permission.)

Thiothixene

Yesavage et al. (52) treated 48 acute schizophrenia patients with thiothixene (80 mg/day), measuring serum and red blood cell (RBC) concentrations 2 hours after the morning dose. Serum levels ranged from 3 to 45 ng/mL, with a linear relationship between clinical response during the first week of treatment and serum (r = 0.5) and RBC levels (r = 0.64). By contrast, Mavroidis et al. (28) found a curvilinear relationship between thiothixene plasma levels and clinical response. Thus, levels ranging from 2.0 to 15 ng/mL, measured 10 to 12 hours after the last dose, were associated with clinical improvement.

Haloperidol

HPDL, the most commonly prescribed FGA, has only one pharmacologically active metabolite (i.e., reduced HPDL). We reviewed the literature on HPDL plasma levels and summarized the outcome (53). Although there are several studies examining the relationship between HPDL steady state plasma levels and clinical response, they have used varying methodologies in terms of patient selection, symptom profile, diagnostic criteria, and assay techniques as well as variable or fixed-dose schedules. Hence, the results of these studies are difficult to interpret collectively. As with other agents, the results with the fixed doses of HPDL are conflicting, although at least six early studies demonstrated a curvilinear relationship (i.e., therapeutic window) between HPDL plasma levels and clinical response. Although optimal levels differed slightly among these studies, the mean low end was 4.2 ng/mL and the mean high end was 16.8 ng/mL (Table 5-4). By contrast, some studies did not find a correlation between plasma HPDL levels and clinical response (54).

Bleeker et al. (55) and Shostak et al. (56) also attempted to study the low end of a putative HPDL plasma level therapeutic window and found that none of three and two of five patients with low plasma levels, respectively, were responders. This contrasted with their outcome in the middle range, where 5 of 23 and 5 of 6, respectively, responded. Conversely, five other studies deliberately probed the upper end of the therapeutic window. Doddi et al. (57), Kirch et al. (58), and Bigelow et al. (59), using medium or high HPDL doses, failed to find any evidence for an upper end of the therapeutic window in newly admitted patients. Rimon et al. (60) used extremely high HPDL doses (120 mg/day) in very chronic patients but also failed to find an upper end of the therapeutic window. Coryell et al. (61) prospectively assigned patients to fixed HPDL doses for at least the first 2 weeks of
treatment to yield a distribution of plasma levels above and below a hypothesized therapeutic level of 18 ng/mL. They found a significant negative linear relationship (i.e., as dose increased, response decreased) between HPDL plasma levels and outcome at week 1, which did not persist into weeks 2, 3, and 4.

TABLE 5-4 EARLY FIXED-DOSE STUDIES FINDING A CURVILINEAR RELATIONSHIP BETWEEN HALOPERIDOL PLASMA LEVELS AND CLINICAL RESPONSE

Study	Assay Method	Therapeutic Range	Rating Scale	Study Duration
Garver (1984)/Mavroidis (1983)	GLC	4-11ng/mL	NHSI	14 days
Smith (1984)	GLC (RRA)	7-17 ng/mL	BPRS psychosis factor	24 days
Potkin (1985)	RIA	4-26 ng/mL	CGI	6 weeks
Van Putten (1985)	RIA	5-16 ng/mL	BPRS	7 days
Van Putten (1988)	RIA	2-12 ng/mL	BPRS psychosis factor	4 weeks
Santos (1989)	RIA	12-35.5 ng/mL (7.4-24.9 in subchronic group)	BPRS total score	21 days
BPRS, Brief Psychiatric Rating Scale; CGI, Clinical Global Impression; GLC, gas liquid chromatographic; NHSI, New Haven Schizophrenic Index; RIA, Radioimmunoassay; RRA, radioreceptor assay.
Adapted from Janicak PG, Javaid JI, Davis JM. Neuroleptic plasma levels: methodological issues, study design, and clinical applicability. In: Marder SR, Davis JM, Janicak PG, eds. Clinical Use of Neuroleptic Plasma Levels. Washington, DC: APPI Press; 1993:17-44.

TARGETED PLASMA LEVEL DESIGNS

One can pool data from several fixed-dose studies to achieve an adequate sample size that would define the optimal cutoff point for a lower end and possibly for an upper end as well. These data can then be used to design a targeted plasma level study.

Using such a design, Volavka et al. (62) failed to find evidence for a therapeutic window after randomly assigning 111 schizophrenia or schizoaffective patients to one of three HPDL plasma levels (i.e., 2 to 13, 13.1 to 24, or 24.1 to 35 ng/mL). However, too high a level for the low and perhaps middle ranges and a prolonged period to titrate to the middle and high HPDL plasma levels may have complicated interpretation of their results.

In a subsequent report, Volavka et al. (54) targeted acutely psychotic patients to a low (mean, 2.2 ng/mL) or a middle (mean, 10.5 ng/mL) HPDL plasma level group. Twenty-five patients with levels less than 3 ng/mL showed minimal response over 3 weeks, indicating that patients would benefit if titrated to higher plasma levels. There was no indication of diminished responding at higher levels.

Janicak et al. (63) attempted to address these issues by prospectively reassigning initial HPDL nonresponders to the putative therapeutic range (Fig. 5-4). During phase A, 25% to 30% of patients in the low, middle, and high HPDL plasma levels responded. Further, in the second phase of this study, initial partial or nonresponders to low or high plasma levels benefited from a dose adjustment (i.e., received an average dose of 25 mg/day) to achieve plasma levels of about 12 ng/mL (Fig. 5-5). When they analyzed only the phase A nonresponders in phase B, the sample size was smaller; but those in the middle range demonstrated significantly more improvement on both the total Brief Psychiatric Rating Scale (BPRS) and the positive symptom subscale scores in contrast to those who remained in the low or the high groups. Further, no benefit was achieved by raising plasma levels beyond the defined middle range.

In the two Case Examples, discussed, patients demonstrated clinically relevant and statistically significant decreases in psychotic symptoms after their plasma HPDL concentrations were targeted to within the 10 to 15 ng/mL range. Although it is possible that they improved during the second phase solely because of time on treatment, their scores on the BPRS dropped rapidly once switched into the middle range. In addition, they demonstrated minimal improvement during the first 3 weeks.

Further, scores on the Simpson-Angus side effect scale were in the mild range for both patients and did not change during either phase. Thus, this would not support decrease in toxicity
as the reason for improvement, especially when the dose and blood levels were reduced in the first patient.

Figure 5-4 Haloperidol (HPDL) plasma level ranges by targeted groups. (Adapted from Janicak PG, Javaid JI, Sharma RP, et al. A two-phase, double-blind randomized study of three haloperidol blood levels for acute psychosis with reassignment of initial non-responders. Acta Psychiatr Scand. 1997;95:343-350.)

As part of this study, a dose-prediction formula was developed to more rapidly achieve the desired HPDL C_ss. For this purpose, the first 28 patients, before receiving the initial assigned dose for achieving their targeted level, received a 15-mg “test” dose (orally) of HPDL, and blood samples were drawn 24 and 48 hours afterward. Data analysis indicated a strong linear relationship between the targeted log C_ss achieved and the dose required when the 24-hour log-transformed plasma level was included in a linear regression model (r = 0.933) (63). This formula was then prospectively tested and found to be valid in determining the dose required to achieve the desired HPDL C_ss (64).

The threshold data generated by McEvoy et al. (65) support the phase A findings of this study. This group determined what they called the NT dose (i.e., 3.7 ± 2.3 mg/day) for 106 schizophrenic and schizoaffective patients by starting with a low dose and increasing it until mild rigidity developed. The phase A results of the previous study were quite consistent with these data in that plasma levels as low as 2 ng/mL (mean dose, 3.3 mg/day) did not clearly fall below a therapeutic threshold for response. After a 2-week open trial at this dose, McEvoy et al. then randomly assigned the 95 remaining patients under double-blind conditions to remain at the NT dose or to be switched to a higher dose of HPDL (mean, 11.6 ± 4.7 mg/day) for an additional 2 weeks. Although there were similar rates of improvement on the total BPRS in phase A with McEvoy et al. at comparable
time points, the Janicak et al. (63) study found a greater improvement in phase B (8.1 BPRS points) compared with McEvoy et al.’s second phase (2.9 BPRS change score points). In the McEvoy study, although higher doses of HPDL did not produce greater improvement in measures of psychosis, they did lead to slightly more improvement in hostility.

Figure 5-5 Clinical response based on plasma leve l assignment in Phase B. (From Janicak PG, Javaid JI, Sharma RP, et al. A two-phase, double-blind randomized study of three haloperidol plasma levels for acute psychosis with reassignment of initial non-responders. Acta Psychiatr Scand. 1997;95:343-350, with permission.)

Case Example

A 29-year-old male schizophrenia patient was admitted to our research unit due to a worsening in his paranoid delusions and auditory hallucinations. After giving informed consent and undergoing a 12-day medication washout, he scored a 48 on the BPRS. He was randomly assigned initially to a low plasma level, received 2 mg/day of HPDL, and had an average plasma level of 1.29 ng/mL over 3 weeks. He demonstrated a slight improvement, scoring a 37 on the BPRS at that point (i.e., a 23% change). He was then randomly reassigned to a middle plasma level, and the dose of HPDL was increased to 18 mg/day. After 27 days, he had an average plasma level of 12.0 ng/mL and rated a 21 on the BPRS, or a 56% improvement from his baseline symptoms. Again, considering 18 is the lowest total BPRS score possible, this was a 90% drop in ratable symptoms.

These results also complement a series of PET studies by Nordstrom et al. (66) and Nyberg et al. (67), which support a potential optimal striatal D₂ occupancy level. Thus, below 60% D₂ occupancy response may be suboptimal and above 80% D₂ occupancy EPS may increase. In addition, Wolkin et al. (68) found that HPDL plasma levels of 5 to 15 ng/mL led to rapid increase in striatal D₂ occupancy, up to 80%. In a sample of seven acutely psychotic, first-episode patients, Kapur et al. (69), in a prospective controlled trial, found that 2 mg of HPDL for 2 weeks produced a 66% ± 7% D₂ receptor occupancy level. Patients achieved HPDL plasma concentrations ranging from 0.6 to 1.5 ng/mL (κ² = 1.1 ± 0.4). At these levels, five patients showed “much improvement” per the Clinical Global Impression (CGI), including a 45% improvement in positive symptoms and a 55% improvement in negative symptoms.

Although some studies using higher plasma levels find a modest decrement in clinical effect and others find a response plateau, none finds evidence that higher plasma levels lead to enhanced efficacy. In addition, though not a marked phenomenon, with unnecessarily high plasma levels (or doses), there is an increased risk of adverse events (AEs). For example, there is evidence that rapid dose escalation may predispose to neuroleptic malignant syndrome (NMS) (70). In this context, we emphasize the consistent agreement among studies that there is no
additional benefit to using higher doses or plasma levels.

Case Example

A 29-year-old woman was admitted to our research unit due to an exacerbation of her schizoaffective disorder. After giving informed consent, the patient participated in the targeted HPDL plasma level study. She was randomly assigned to the high plasma level and began treatment with 60 mg/day HPDL for 3 weeks. She scored a 40 on the BPRS at baseline (after an 11-day medication-free period). Her mean HPDL plasma level was 38.4 ng/mL for the first 3 weeks, and her BPRS score was 39 at the end of that period. She was then randomly reassigned to a middle plasma level, with a reduction of HPDL to 16 mg/day. At the end of this phase (24 days later), the patient’s BPRS score was 22, with an average plasma level of 14.2 ng/mL. After no improvement in the first phase, the patient experienced a 45% decrease from her total baseline BPRS score during the second phase. Keeping in mind the lowest score on the BPRS is 18, this was an 82% drop in ratable symptoms.

Clozapine

Clozapine Plasma Level-Drug Interactions. Clozapine is principally metabolized to N-desmethylclozapine (norclozapine). It is also metabolized to N-oxide, other hydroxyl metabolites, and a protein-reactive metabolite. The N-oxide can be converted back to clozapine. The enzyme responsible for the metabolism of clozapine to norclozapine is cytochrome P450 (CYP) 1A2 (71). This is consistent with a study showing that caffeine, a marker for 1A2, is cleared in relationship to the conversion of clozapine to norclozapine (72). Discontinuation of coffee intake can decrease clozapine plasma levels by more than 50%, and increasing caffeine intake can produce a reemergence of adverse effects (e.g., drowsiness, excess salivation). In addition, smoking, which induces 1A2, lowers clozapine plasma levels. Fluvoxamine, an inhibitor of 1A2, dramatically increases plasma levels, and on occasion, adverse effects are seen (73). This phenomenon can lead to clozapine intoxication in patients on high doses of fluvoxamine.

Phenytoin, an enzyme inducer, can reduce clozapine plasma levels. There have been two reported cases of RISP causing an increase in clozapine plasma levels; however, because this agent is not an inhibitor of 1A2, the mechanisms for this increase are unclear.

Clozapine Dose-Response Relationship. Data exits on dose-response curves for clozapine. In a blinded, controlled study, Simpson et al. (74) randomized patients to three doses of clozapine: 100, 300, and 600 mg. These researchers found that 600 mg was more effective than 300 mg and that doses up to 400 mg/day usually produce inadequate plasma levels. Thus, the results of this dose-response study are consistent with the plasma level studies. This provides some evidence that therapeutic drug monitoring (TDM) may help ensure that patients receive a clozapine dose high enough to reach an adequate plasma level.

Clozapine Plasma Level-Response Relationship. Dose-response and plasma level-response information is conceptually very similar. In dose-response terms, it is important to know the average therapeutic threshold dose, the minimal dose to produce a full response. When plasma levels correlate with clinical response and there is a wide range of plasma levels for a given dose, then it is important to know the minimal therapeutic plasma levels to produce the optimal response for most patients. Several reports have indicated that plasma concentrations of clozapine above 350 to 450 ng/mL may be required to maximize response in treatment-refractory schizophrenic patients (34,35,75). Clozapine plasma levels, however, vary substantially from patient to patient. For example, Potkin et al. (35) found a greater than 20-fold (40 to 911 mg/mL) range in plasma levels after a fixed dose of 400 mg/day. Initially, Perry et al. (34) did a plasma level study assigning patients to a dose of 400 mg/day for 4 weeks and
identified 350 ng/mL as a threshold for response. At or above this plasma level, 64% of patients responded as compared with only 23% of patients below that level. This group followed up these original results with a larger sample, and in a follow-up of the existing samples, five of seven patients had an unsatisfactory response to a dose of 400 mg/day. When the dose was increased to achieve a plasma level above the threshold 350 ng/mL, however, most then responded (76).

Potkin et al. (35) also used a fixed dose of 400 mg/day in a study of 58 schizophrenic patients and found that very few (8%) patients with low clozapine plasma levels (<420 ng/mL) responded, in comparison with 60% of those with plasma levels greater than 420 ng/mL. This group used a targeted plasma level design so that patients with originally low plasma levels had these levels increased above 420 ng/mL in a double-blind, random-assignment procedure. Response rate increased to 73%, compared with 29% of those whose plasma levels remained low. This result is particularly impressive because it was a randomized prospective change in dose, which provides important confirmation of the relationship between clozapine plasma levels and clinical efficacy. Kronig et al. (75) also found a therapeutic level of 350 ng/mL. Although the chosen level varies from study to study, there is general agreement that a threshold therapeutic plasma level exists somewhere between 350 and 450 ng/mL.

It is important to consider whether parent compound levels only or the sum of clozapine plus norclozapine plasma levels are measured because in the latter case, the threshold would be higher. Note that a quality control study has been done comparing different laboratories on the assay reliability of clozapine plasma levels, with split samples finding good agreement. Clinically, it is important to start with a small dose and gradually increase to avoid adverse effects. Even so, it is desirable to move rapidly to the correct level in patients who may need a slightly higher than average dose. Perry (78) has provided a prediction formula for plasma concentration that requires information about the patient’s clozapine dose, smoking status, and gender.

Case Example

One patient ingested an overdose of clozapine on two different occasions. On the first, she took clozapine and recovered in 6 hours from a comatose state. Two months later, while on therapeutic doses of fluvoxamine, she again ingested an overdose and exhibited a state of fluctuating consciousness. Usually, she was unable to gesture a simple request, but sometimes she was able to react adequately. She was also bothered by vivid and frightening visual hallucinations of tigers under her bed and serpents on her pillow. Her speech was disturbed, and she could not sit up because of truncal ataxia. Twenty-four hours after the intoxication, the overdose symptomatology declined significantly. The patient was able to walk a few steps, and her speech was normal. Peak plasma levels were greater than 9,000 ng/mL, falling to 4,000 ng/mL by day 1, and less than 1,000 ng/mL by day 3. This excretion rate was slower than is normally observed in overdoses, presumably because of the pharmacokinetic interaction with fluvoxamine through CYP 1A2 (77).

Risperidone

Riedel et al. (38) measured RISP and 9-OHRISP levels in patients receiving open-label treatment. Those with higher levels were less likely to be clinical responders. Although plasma concentrations and EPS were not correlated, a high concentration at week 2 predicted the development of EPS. The lack of a clear relationship between levels and response could be related to clinicians adjusting the dose to manage EPS. The poor response of patients with higher plasma levels could be due to mild EPS, which went unrated but may have compromised the overall clinical outcome. Another trial (39) did find a relationship between RISP plasma levels and EPS but not clinical response.

Olanzapine

Perry et al. (42) measured OLZ concentrations in patients who participated in a study of three OLZ dose ranges. Receiver-operated curve (ROC) analyses indicated that a level greater than 23.2 ng/mL was a useful predictor of response. Thus, among patients with levels above 23.2 ng/mL, 52% met response criteria, whereas only 25% of patients with lower levels met these criteria.

SUMMARY

The value of routine plasma level monitoring with antipsychotics remains uncertain. However, TDM can be useful to

Determine patient adherence
Establish that adequate doses of pharmacotherapy were used in nonresponders
Avoid toxicity due to unnecessarily high plasma levels
Monitor patients with other medical disorders or on concurrent psychotropics or other medical drugs
Maximize the clinical response when the drug plasma level-response relationship has been elucidated
Help define the dose-response relationship
Safeguard clinicians in potential medicolegal situations

Antipsychotics for Acute Psychosis

ACUTE SCHIZOPHRENIA

Each class of antipsychotic was found superior to placebo for the treatment of acute schizophrenia. Some of the most informative studies regarding the effectiveness of these agents were carried out during the first years following their discovery, when there was considerable skepticism regarding their usefulness. For example, a classic, double-blind, random-assignment investigation by Cole et al. (79) compared the response of schizophrenia patients treated with antipsychotics with those treated with placebo. In a reexamination of the data (80), the authors evaluated the fate of the 10% to 20% patients who had the best outcome. A small percentage recovered completely without residual symptoms in 6 weeks, representing about 16% of those on an active drug but only about 1% of those on placebo. This is a remarkable difference, and although some did relatively well without medication, it is likely that they would have done even better with active drug therapy. Further, in the drug-treated group, only 2% deteriorated, in contrast to 33% in the placebo group. It is true that some patients showed little improvement with antipsychotics, but if they had been given only placebo, these patients most likely would have deteriorated even more. Thus, at both ends of the prognostic spectrum, there appears to be a substantial benefit from drugs. In some of the studies, the dose was too low to produce an effect, but when adequate these agents were consistently superior to placebo. The magnitude of improvement in the drug-treated group was considerable whether evaluated by the degree of change (e.g., worse, no change, slight improvement, marked improvement) or the degree of remission (e.g., full remission, only minimal symptoms, still mildly ill, moderately ill, or severely ill). Results from this National Institute of Mental Health (NIMH) Collaborative Study, combining both types of evaluation, are presented in Table 5-5 (79,80). These results can be summarized as follows:

TABLE 5-5 EFFICACY OF ANTIPSYCHOTICS

Degree of Remission	Degree of Change	Drug (%)	Placebo (%)
Remitted	Very much improved	16	1
Only borderline symptoms remain	Improved	29	11
Mild symptoms still present	Improved	16	10
Moderately ill	Slightly improved	31	31
Moderately ill	Not improved	6	15
Severely ill	Worse	2	33
From Cole JO, Davis JM. Antipsychotic drugs. In: Bellak L, Loeb L, eds. The Schizophrenic Syndrome. New York, NY: Grune & Stratton; 1969:478-568, with permission.

16% of the drug-treated versus only 1% of the placebo-treated group were in complete remission or very much improved
29% of the drug-treated and 11% of the placebo-treated group were evaluated as improved, with only borderline symptoms
16% of the drug-treated and 10% of the placebo-treated group were evaluated as improved, with mild symptoms remaining
8% of patients did poorly on active drug (i.e., rated as not improved or still ill), compared with 48% on placebo
Only 2% of those on an antipsychotic versus 33% of the placebo patients deteriorated during 6 weeks of treatment
The largest differences between the drug- and the placebo-treated patients are seen at both ends of the outcome spectrum

This study also found that new schizophrenia symptoms often emerged during the 6 weeks of placebo treatment, whereas worsening of symptoms was prevented by antipsychotics. Lehmann (81) coined the term psychostatic to describe the ability of phenothiazine antipsychotics to prevent the reemergence of symptoms.

The time course of improvement for most acutely psychotic patients on these agents demonstrates that most therapeutic gain occurs during the first 6 weeks, although further progress can be realized much later (e.g., clozapine). Thus, some patients improve rapidly, within a few days, whereas others show gradual changes over several months. There is no evidence for clinical tolerance, for if there were, the dose would have to be increased over time. Fixed-dose studies, however, show that the initial effective amount of drug does not lose efficacy. In addition, patients do not require higher doses after several weeks of treatment; instead, the reverse appears to be true. It is also clear that antipsychotics do not produce dependency of the barbiturate, stimulant, or narcotic type.

In the hope of developing either a more effective agent or one with fewer adverse effects, medicinal chemists have synthesized a number of new compounds. Preclinical animal model studies are used to select potentially useful drugs, using a profile of pharmacological properties reflecting differential blockade of DA as well as other neurotransmitter systems (e.g., 5-HT). This research has led to the development of several new classes, as noted earlier in this chapter.

The question immediately arises as to whether any of the drugs in these various classes are superior to CPZ for the typical patient, for particular symptoms, or for subgroups of patients. With the exception of promazine and mepazine, all antipsychotics are clearly superior to placebo or nonspecific sedatives. Comparisons with CPZ in controlled trials found mepazine and promazine inferior to CPZ, but all other agents were equal to CPZ in therapeutic efficacy. Other controlled trials using thioridazine or trifluoperazine as standards also found all other FGAs to be comparable with these agents. Although there were trends at times favoring one agent over another, they were usually not statistically significant, and an inspection of the data finds no agent to be consistently superior to any other. Furthermore, all these drugs produced consistent changes in the same symptoms. These similarities are quite striking and support the theory that they work through a common mechanism of action (e.g., DA blockade). Furthermore, relevant differences are primarily related to their adverse effect profiles. Pimozide is FDA labeled for Tourette syndrome and is particularly interesting in that it is a highly specific DA antagonist that may produce fewer adverse effects than HPDL. In open studies with adequate doses, this agent has demonstrated efficacy for acute schizophrenia. Several double-blind trials comparing pimozide with other FGAs also found it to be an equally effective maintenance therapy (82,83 and 84). We consider this agent to be as effective as the other standard agents, with the same, potential for adverse effects.

When the SGAs were first introduced in the early 1990s, there was hope that these agents would prove to be more effective than FGAs for the acute symptoms of schizophrenia. This hope was encouraged by comparisons of clozapine with FGAs, which found clozapine more effective (85,86). However, there was an important source of bias in these studies. Among the patients included were those who could not be adequately treated with FGAs because they failed to improve on them or because they could not tolerate them. These studies clearly identified clozapine as a valuable alternative for those patients who failed on FGAs, but they did not address whether it had an advantage for the broad group of patients who could be treated with FGAs.

Support for an advantage for SGAs came from early studies of RISP and OLZ (reviewed in Leucht et al. (87). These were Phase II and III
studies that compared these SGAs with both a placebo and an FGA, usually HPDL. All of these early studies found that the SGA was superior to placebo, and some found that the SGA was also more effective than HPDL for psychotic symptoms. However, a broader look at these and studies of other SGAs raised questions about their advantages. A meta-analysis by Geddes et al. (88) concluded that the advantages of SGAs were apparent only when patients received excessive doses of the comparator FGA. More specifically, studies that used doses of HPDL greater than 12 mg daily were more likely to find advantages for the SGAs. This suggests that the greater severity of extrapyramidal symptoms in the HPDL-treated groups may have given the SGAs an apparent advantage, though it was not related to their intrinsic antipsychotic properties. In addition, another meta-analysis (87) indicated that even when there were statistically significant advantages for the newer drug over HPDL, the effect sizes were very small. This finding is disputed in a meta-analysis by Davis et al. (89) who found substantial advantages for clozapine, amisulpride, RISP, and OLZ when compared with HPDL. These advantages were not present with other SGAs. Cochrane reviews of RISP (90) and OLZ (91), however, point out design flaws such as high dropout rates that make it difficult to determine whether these SGAs are superior to FGAs.

Results of the first phase of the Clinical Antipsychotic Trial of Intervention Effectiveness (CATIE) provide useful information regarding the relative effectiveness of various antipsychotics (92). This was a double-blind, randomized, NIMH-sponsored trial conducted at 57 US sites in which 1,493 non-treatment-resistant patients with schizophrenia were randomized to perphenazine (8 to 32 mg/day), OLZ (7.5 to 30 mg/day), QTP (200 to 800 mg/day), RISP (1.5 to 6 mg/day), or ZPD (40 to 160 mg/day). Concomitant psychotropics other than another antipsychotic were allowed. The primary outcome measure was the time that patients remained on their assigned drug. Seventy-four percent of subjects discontinued treatment during an 18-month period. Rates of discontinuation were 64% for OLZ, 75% for perphenazine, 82% for QTP, 74% for RISP, and 79% for ZPD. This study (like others) indicates that FGAs and SGAs are similar in efficacy. There was a statistically better outcome with OLZ in that

Time to discontinuation for any cause was longer for OLZ (but this was nonsignificant for perphenazine and ZPD after adjustment for multiple comparisons)
Time to discontinuation for lack of efficacy was longer for OLZ (but this was nonsignificant for ZPD after adjustment for multiple comparisons)
Duration of successful treatment was significantly longer for OLZ

The mean modal doses of some of the antipsychotics were problematic in that they were lower than those typically used in clinical practice (e.g., QTP = 543 mg/day; ZPD = 113 mg/day; RISP = 3.9 mg/day). Monitoring of important AEs found that

Time to discontinuation for intolerable side effects did not differ, but rates were highest in the OLZ group and lowest in the RISP group
Metabolic issues were highest with OLZ, while ZPD was the only agent associated with improvement in metabolic variables
EPSs were highest with perphenazine
Hyperprolactinemia only occurred with RISP
Clinically relevant changes in the QTc interval did not occur with any agent

The interpretation of the results from the first phase of CATIE, however, is complex. Although OLZ had an advantage in efficacy, it also caused greater increases in weight, glucose levels, and lipids. Patients who received perphenazine were more likely to discontinue their medication due to EPS. The study was too brief to evaluate the relative risk of TD among the agents, and for ethical reasons, those with a history of TD could not be randomized to perphenazine. In addition, the patients who entered the study were probably not representative of typical patients who receive treatment in clinical settings. On average, patients had been ill for more than 15 years, had received multiple trials with SGAs, but still had not found a completely satisfactory agent. This suggests that the study was focused on patients who were partial responders to antipsychotics— a group likely to have higher rates of discontinuation. Nevertheless, the large sample size of this study and the fact that it was supported by NIMH rather than industry give the results considerable weight. The results also indicate that all currently available antipsychotics have serious limitations for many patients.

A study from the United Kingdom, Cost Utility of the Latest Antipsychotic Drugs in Schizophrenia Study (CUtLASS), also compared first- and second-generation antipsychotics (93). In this study, 227 patients with similar demographic characteristics to CATIE patients were randomized to a first- or secondgeneration antipsychotic. In contrast to CATIE, they were not assigned to a particular drug, but to a first- or second-generation agent selected by the subject’s psychiatrist. The study failed to find a difference between FGAs and SGAs. However, the findings are difficult to interpret since the most commonly chosen FGA in the study was sulpiride, an agent that is unfamiliar to US psychiatrists and one that is viewed as having less EPS than other FGAs.

Taken together, these large studies do not point to an efficacy advantage for the SGAs. On the other hand, many psychiatrists prefer the newer drugs because of their lower risk of EPS.

SPECTRUM OF ACTIVITY FOR ANTIPSYCHOTICS

Schizophrenia Versus Other Psychosis

Antipsychotics are effective for virtually every condition in which psychosis can be a symptom, including major depression, bipolar disorder, dementias, and psychoses secondary to medical conditions. In each of these illnesses, antipsychotics do not specifically treat all aspects of the underlying illness. Rather, these agents appear to target the psychotic manifestations of the illness.

Effects on Positive, Negative, and Neurocognitive Symptoms in Schizophrenia

Contemporary models for understanding schizophrenia have proposed that the illness includes a number of psychopathological dimensions. Different models accommodate different dimensions, but all include at least positive, negative, and neurocognitive symptoms (94,95 and 96). These dimensions are core features of the illness and are relatively independent of one another. Agitation/excitement and depression are often included as well (Table 5-6). If antipsychotics could be considered truly “antischizophrenic,” it would be expected that they would effectively improve each of these dimensions. However, treatment of an acute psychotic episode typically leads to improvement in positive symptoms such as hallucinations and delusions. As patients improve and enter remission, it is common for positive symptoms to be substantially reduced or eliminated and for negative (e.g., restricted affect, alogia, asociality, avolition, anhedonia) and neurocognitive symptoms to remain. This suggests that antipsychotics act primarily on the positive symptom component and that improvements in other dimensions are secondary.

TABLE 5-6 SCHIZOPHRENIA: SYMPTOM DIMENSIONS

Positive symptoms
	Hallucinations
	Delusions
Neurocognitive symptoms
	Dissociative thinking
	Disorganization of thought
	Attentional impairment
	Memory impairment
Negative symptoms (i.e., deficit syndrome)
	Restricted affect
	Alogia
	Asociality
	Apathy
	Anhedonia
Excitement/agitation
Mood symptoms
	Dysphoria
	Depression

Impairments in neurocognition, including memory, attention, executive functioning, and psychomotor performance (97), are a core feature of schizophrenia. Unlike positive symptoms, which tend to be episodic, cognitive impairments are usually present at the onset of the illness and remain relatively stable over the course of the patient’s life. These impairments tend to be an independent dimension of the illness with little relationship to positive symptoms. The importance of neurocognitive symptoms relates to the ability of patients with schizophrenia to function in the community. That is, the relationship between functional outcomes—which include the ability to function vocationally and socially—and neurocognitive impairments is stronger than the relationship between functional outcomes and positive symptoms.

Both FGAs and SGAs may improve cognitive symptoms in psychotic individuals (98),
although their effects are relatively weak compared with the severity of the impairments. Although a number of studies suggest that SGAs are superior to FGAs in improving cognition (99), patients treated with the newer agents usually still retain these impairments. For example, it is estimated that individuals with schizophrenia perform one and one half to two standard deviations below the mean on neuropsychological tests, and SGAs may improve performance by about one half of a standard deviation (99). Although a number of studies have compared SGAs to one another, it is unclear if any SGA demonstrates advantages over others for neurocognitive impairment (100). The NIMH CATIE study compared the effects of five different antipsychotics on a neurocognitive battery (101). Each of the medications led to relatively small improvements in a composite score (z = 0.12 to 0.26) after 2 months. Moreover, there were no differences among the antipsychotics which included perphenazine, an FGA. Interestingly, at 18 months, there was a suggestion that perphenazine had an advantage when compared with the SGAs. This provides further evidence that SGAs do not have specific effects on neurocognition.

The lack of a substantial effect of available antipsychotics for both cognition and negative symptoms has led to a search for medications that would target these domains. These agents might be ineffective as antipsychotics but would be added as comedications with antipsychotics. A number of experimental drugs targeting cholinergic, glutamatergic, dopaminergic, and other systems are in development and may be effective adjuncts to existing APs (102,103). Although most of these agents focus on improving cognition, there are suggestions that some may be effective for treating negative symptoms as well (104).

Some models of psychopathology also include a component of excitement or agitation. Early impressions following the introduction of the first FGAs focused on the calming they induced and led observers to propose that they were “major tranquilizers.” Antipsychotics can diminish excitation, but they also lead to increased activity in withdrawn psychotic individuals. Therefore, “tranquilizer” is a misnomer. Antipsychotics do not, in any real sense, produce a state of tranquility in normal or psychotic individuals; in fact, normal control subjects often find their effects unpleasant. An appropriate analogy may be aspirin, which reduces an elevated temperature but typically does not alter normal temperature. Another example is insulin, which replaces the absent endogenous supply and restores a diabetic to normal glucostasis.

Dysphoria and depression are also prominent symptoms in schizophrenia. They may overlap with certain negative symptoms, are worsened by some FGAs, and contribute to the high suicide rate in these patients.

SUMMARY

Assuming antipsychotics work (at least in part) by impacting DA receptors, it is not known why their effects lead to a lessening of psychosis—but clearly, these agents do more than control agitation. They are effective for reducing and often eliminating the reality distortions that can torment people with schizophrenia. Unfortunately, this improvement in psychosis has minimal effects on other symptom dimensions of schizophrenia (105).

Second-Generation Antipsychotics

There is no single mechanism of action that defines the SGAs. Rather, this group most clearly differs from FGAs in producing minimal or no extrapyramidal reactions at therapeutic doses. This may be explained by a higher 5-HT₂ to D₂ receptor affinity ratio; a lower, though not negligible, affinity for the D₂ receptor; the looseness with which the drug binds to the receptor; and/or partial D₂ and 5-HT_1A receptor agonism (106).

Clozapine

Receptor-Binding Studies. Clozapine (Clozaril) was the first clinically effective antipsychotic with a receptor profile that differentiated it from FGAs. Using PET scanning, Farde et al. (107) demonstrated that patients receiving antipsychotics other than clozapine exhibit striatal D₂ receptor occupancy in the 70% to 89% range. Further, those agents with greater propensity to induce EPS had higher D₂ receptor occupancy rates (82 ± 4% vs. 74 ± 4%), whereas no patients with occupancy rates below 75% had any EPS (51). Farde et al. then demonstrated that the D₂ occupancy rate of clozapine (125 to 600 mg/day) was between 20%
and 67% (108,109). Thus, the low incidence of EPS with clozapine may be due to a subthreshold level of D₂ occupancy at clinical doses.

Because clozapine is an effective D₄ receptor blocker, there has been speculation that its superior efficacy may be a consequence of this property. Several D₄ antagonists or D₄/5-HT₂ antagonists have been studied in early clinical investigations; however, they failed as effective treatments for schizophrenia in phase II trials.

There is a striking correlation between clinical efficacy and D₂ blockade with most antipsychotics. Clozapine is a partial exception because it is a weaker D₂ antagonist than other agents at usual clinical doses. Although the pharmacological reason for the greater efficacy of clozapine has not been elucidated, there is little evidence that it is explained by D₄ antagonism.

In summary, compared with other antipsychotics, clozapine

Has a lower affinity for D₂ receptors
Binds proportionately more to D₁ and D₄ receptors
Appears to act selectively on cortical and mesolimbic DA neuronal systems (in preference to the nigrostriatal and tuberoinfundibular areas)
Blocks 5-HT₂ receptors
Has few or no EPS effects across the recommended dosing range
Has effects on several other neurotransmitter systems (e.g., NE, ACh)
Has more than 25 years of experience establishing it as an effective treatment for
- Acute and chronic schizophrenia
- Psychoses that are responsive or nonresponsive to other antipsychotics

Therapeutic Efficacy in Schizophrenia. Early evidence supporting the efficacy of clozapine came from one double-blind, placebocontrolled study and six other FGA-controlled, random-assignment studies (110,111). When compared with FGAs such as CPZ and HPDL, clozapine was found to be more effective (i.e., three studies found clozapine significantly superior and two others found a trend favoring clozapine (p = 0.10). However, as mentioned earlier, the larger trials selected patients who did poorly on FGAs due to adverse effects or their inability to tolerate these agents. As a result, the studies may be biased in favor of clozapine.

Even though this agent may be superior to FGAs in a substantial proportion of nonresponders, its adverse effect profile has been a major obstacle to its widespread use. Specifically, agranulocytosis (approximately 0.5% to 1% incidence), seizures (usually with higher doses), and cardiovascular effects are the most serious. These safety concerns have led many clinicians to limit their use of clozapine to the most severely-ill patients. On the other hand, an 11-year followup study in Finland found that clozapine was associated with the lowest mortality among FGAs and SGAs (112). Other studies—discussed later in this Chapter—indicate that clozapine is also effective for reducing suicidal behaviors. These findings led to a reassessment of the overall safety of clozapine and encourage its use in a broad range of patients who have partial responses to other antipsychotics.

A number of studies compared other SGAs with clozapine. The overall results are difficult to interpret, with some studies suggesting that RISP (113) and OLZ (114,115) have efficacy similar to that of clozapine, whereas others suggest advantages for clozapine (116,117). A study by Volavka et al. (118) compared clozapine, RISP, OLZ, and HPDL in patients with suboptimal responses with FGAs. Effect sizes for improvement were largest for OLZ and clozapine, lowest for HPDL, and intermediate for RISP. The CATIE study randomized patients who discontinued their phase I medication for lack of efficacy to open-label clozapine, or double-blind OLZ, RISP, or QTP (119). For the primary measure of time until treatment discontinuation for any reason, clozapine was significantly superior to QTP or RISP; for discontinuations for lack of efficacy, clozapine was significantly superior to all three SGAs. A study in the United Kingdom randomly assigned patients who had poor responses to two or more antipsychotics to either clozapine or an SGA selected by the treatment psychiatrist (120). Clozapine patients showed greater improvement in their PANSS scores. Taken together, these studies suggest that patients who have an inadequate response to two antipsychotics should be considered for clozapine. Published guidelines from the American Psychiatric Association (APA) (121) and the Texas Medication Algorithm (122) suggest that patients receive trials of one or two other SGAs before receiving clozapine. Guidelines from an expert consensus meeting on the pharmacotherapy of schizophrenia held at
the Mount Sinai Hospital in 2001 (242) also state that patients should not be considered partial or nonresponders until they have received a trial of clozapine. Conley et al. (123) underlined the importance of trying clozapine in patients who are doing poorly on another SGA when they found that individuals who failed to respond to OLZ did substantially better with clozapine.

Clozapine and Suicide. Evidence from a number of sources suggests that clozapine reduces the risk of suicide in schizophrenia. The most obvious source emerged from the treatment of patients who had previously been severely psychotic and unresponsive. Improvement in these individuals appeared to be associated with a decreased risk of suicide (124). Other studies found that patients with schizophrenia who were treated with clozapine had much lower rates of completed suicide than other populations with schizophrenia. For example, Reid et al. (125) found that the annual rate of completed suicide among patients with schizophrenia treated by the Department of Mental Health in Texas was 63.1 per 100,000. In contrast, from an apparently similar group of patients who received clozapine, there was only one completed suicide (i.e., 12.7 per 100,000). The authors noted that this rate was similar to the rate in a clozapine national registry.

These observations led to the International Prevention of Suicide (InterSept) trial in which 980 individuals with schizophrenia or schizoaffective disorder and at risk for suicide were randomly assigned to clozapine or OLZ (126). The study found that clozapine-treated patients had a significant advantage over those treated with OLZ in decreasing suicidal behaviors. This included fewer suicide attempts and a lower likelihood of hospitalization or other rescue interventions. As a result of this study, the FDA approved clozapine for reduction of risk of suicide in patients with schizophrenia and schizoaffective disorder.

Clozapine in Other Disorders. Clozapine can also benefit schizoaffective and bipolar patients with treatment-refractory mania (127,128 and 129) (see also Chapter 10). The advantages of clozapine over other antipsychotics and mood stabilizers are most apparent in patients with the most severe refractory illnesses. For example, in one study that included 25 patients who were refractory to antipsychotics and mood stabilizers, 18 patients (72%) showed marked improvement with clozapine monotherapy (130).

Clozapine has been administered to children and adolescents for both schizophrenia and bipolar disorder (see also Chapter 14). Unfortunately, most trials were uncontrolled, providing relatively little guidance for clinicians. A study of children and adolescents with childhood-onset schizophrenia (131) randomized 21 patients (mean age, 14 years) to either HPDL or clozapine. Although clozapine was superior for both positive and negative symptoms, the results suggested that this group may be more sensitive to neutropenia and seizures.

The support for clozapine use in adolescent mania comes from uncontrolled studies (131,132) showing its effectiveness in patients refractory to other medications.

Clozapine for Comorbid Substance Abuse or Dependency. Comorbidity of alcohol/substance abuse or dependency with schizophrenia is common. In this context, Green et al. (133) made an interesting observation in two patients with schizophrenia and excessive alcohol abuse. These patients experience a dramatic decrease in alcohol use coincidental with beginning treatment with clozapine. Buckley et al. (134) also observed a similar phenomenon in patients with comorbid schizophrenia substance abuse treated with clozapine. Subsequently, Drake et al. (135) conducted several systematic observational studies. For example, 151 schizophrenia patients with comorbid alcohol abuse disorder were followed in a psychosocial treatment study. Of this group, 36 patients were treated with clozapine, which produced a significant reduction in the severity of alcohol abuse and also correlated with improvement of negative symptoms. The remission rate for alcohol abuse was 79% in patients taking clozapine, compared with only 34% in those not taking clozapine. These researchers also conducted a retrospective survey of 36 dual-diagnosis, schizophrenic patients with comorbid substance use disorder, and 85% demonstrated a decrease in symptoms. Green et al. (136) found that rates of abstinence in patients treated with either RISP or clozapine were significantly higher in the patients on clozapine.

Yovell and Opler (137) noted a decrease in craving for cocaine in clozapine-treated patients with schizophrenia. This was supported by
Farren et al. (138), who administered ascending doses of clozapine followed by intranasal cocaine in addicted patients. Clozapine was reported to diminish the subjective “high” following the cocaine.

Several groups also reported decreased nicotine use in schizophrenic patients treated with clozapine (139,140). A prospective trial, however, failed to find that clozapine led to decreased rates of smoking (141). In this light, we note that at this time the effectiveness of clozapine in schizophrenia patients addicted to cocaine, alcohol, or other substances has not been adequately demonstrated in controlled, prospective studies.

Starting Clozapine. Starting doses for clozapine should be in the range of 12.5 to 50 mg/day and titrated up slowly to minimize hypotension. There is evidence that some patients develop hypotension or apnea when started on as little as 25 mg. This has led to the suggestion that patients be started on as little as 12.5 mg. The usual therapeutic range is 300 to 500 mg/day (the recommended maximum dose is 900 mg/day). In this context, Perry et al. (34) found a therapeutic threshold clozapine plasma level of 350 ng/mL (i.e., 64% with levels above this responded vs. only 22% with levels below this point), which required an average dose of 380 mg/day. The usual therapeutic dose is typically achieved in 3 to 6 weeks. As hypotension or seizures may occur with a more rapid early titration, we recommend increasing the dose by only 25 to 50 mg every 3 to 7 days depending on the patient’s tolerance.

Risperidone

RISP is available in a number of oral formulations as well as a long-acting injectable formulation, RISP microspheres (RISP LA).

Receptor-Binding Studies. RISP is a benzisoxazole derivative with a relatively high 5-HT_2A/D₂ ratio and relatively fewer and milder EPS than FGAs at its clinically effective doses. Receptor-binding studies of RISP and its active metabolite, 9-OH-RISP, in comparison with other novel antipsychotics revealed that In vitro:

RISP, 9-OH-RISP, ZPD, and ARIP had the highest affinity for human D₂ receptors
OLZ was slightly less potent at D₂ receptor sites
Clozapine and QTP showed two orders of magnitude lower D₂ affinity (142)

In vivo:

RISP, 9-OH-RISP, OLZ, and clozapine maintained a higher potency for occupying 5-HT_2A than D₂ receptors
RISP had a 5-HT_2A versus D₂ potency ratio of about 20
High potency for 5-HT_2A occupancy was observed for RISP and OLZ

All the various compounds also displayed high-to-moderate occupancy of adrenergic α₁ receptors, with clozapine occupying even more α₁ receptors than D₂ receptors. Clozapine also showed predominant occupancy of H₁ receptors and occupied cholinergic receptors with equivalent potency to D₂ receptors.

A stronger predominance of 5-HT_2A versus D₂ receptor occupancy, combined with a more gradual occupancy of D₂ receptors, differentiated RISP and 9-OH-RISP from the other SGAs. Combined 5-HT_2A and D₂ occupancy and the avoidance of excessive D₂ receptor blockade are believed to reduce the risk for EPS (142,143). Although RISP also has a high affinity for the 5-HT₇ but not the 5-HT₆ receptor, the importance of this differential activity is not clear. Nyberg et al. (144) were the first to demonstrate that RISP induces marked occupancy of central 5-HT₂ and D₂ receptors in vivo. About 60% (range, 45% to 68%) of the 5-HT₂ receptors in the frontal cortex and about 50% (range, 40% to 64%) of the D₂ receptors in the striatum were occupied 4 and 7 hours, respectively, after a single oral dose of 1 mg. These results indicate that 5-HT₂ receptor occupancy should be very high at doses of 4 to 10 mg daily.

Because antipsychotic-induced EPSs are related to the level of D₂ receptor occupancy, Kapur et al. (145) used [11C]raclopride PET scans to determine in vivo D₂ receptor-binding characteristics in nine patients receiving 2 to 6 mg/day of RISP. These authors found that the mean level of receptor occupancy was 66% at 2 mg, 73% at 4 mg, and 79% at 6 mg. Further, the three patients who had the highest receptor occupancies experienced mild EPS but did not require antiparkinsonian medication. These results suggest that RISP 4 to 6 mg daily
achieves a D₂ receptor occupancy similar to that of FGAs and higher than that of clozapine. The occurrence of EPS at higher levels of D₂ receptor occupancy in this study and in previous clinical trials indicates that the high 5-HT₂ affinity of RISP provides some protection against EPS. Clinical and laboratory data, however, also suggest that when the D₂ receptor occupancy exceeds a certain threshold, any 5-HT₂-mediated protection against EPS may be lost. The lower propensity of RISP to evoke EPS is progressively diminished as its dose increases and its occupancy of D₂ receptors rises. Thus, there is a dosage range between inducing and not inducing EPS.

Remington and coworkers (2006) evaluated D₂ receptor occupancy in patients administered 25, 50, or 75 mg of injectable RISP LA after they had already been stabilized on their dose (146). The three doses were associated with occupancies greater than 65% suggesting that they were effective. However, the 75 mg dose led to occupancy near 80% indicating that this may be associated with EPS. Another study (147) found D2 receptor occupancies of 25% to 48% for 25 mg, 59% to 83% for 50 mg, and 62% to 72% for 75 mg.

Therapeutic Efficacy. In the United States, oral RISP is indicated for the following conditions:

Schizophrenia in adults and adolescents
Bipolar mania in adults, children, and adolescents
Irritability associated with autistic disorder

Long-acting RISP is indicated for

Schizophrenia
As monotherapy or an adjunctive therapy to VPA or lithium for bipolar I disorder

Schizophrenia. The therapeutic efficacy of RISP for schizophrenia was well established in several controlled trials conducted worldwide (148,149). It is appropriate to combine these data using meta-analytic techniques to explore the efficacy of RISP compared with FGAs. For most drugs, the relationship of dose and response is defined by the classic sigmoidal curve. Thus, as the dose (or plasma level) increases beyond a threshold and reaches the linear portion of the curve, response increases. Once the dose is high enough to produce maximal clinical response, the dose-response curve then levels off.

Because a therapeutic window has been hypothesized for antipsychotics in general, its existence for RISP is a reasonable supposition. Indeed, evidence to support such a window was established in the North American Clinical Trial and the International Collaborative Study, both of which found doses of 4, 6, and 8 mg superior to higher or lower doses of RISP.

The efficacy of RISP for adolescents with schizophrenia was demonstrated in two acute trials of patients ages 13 to 17 years (150,151). One study compared RISP 1.5 to 6.0 mg/day with 0.15 to 0.6 mg/day. The higher dose was more effective. The other study compared 1 to 3 and 4 to 6 mg/day with placebo. For both studies, doses above 1 mg were effective. Moreover, there was no indication that doses above 3 mg were superior to lower doses (see also Chapter 14).

Bipolar Mania. A number of trials have demonstrated that RISP is effective as monotherapy for treating the symptoms of bipolar mania. A 3-week trial found that RISP at an average dose of 4.1 mg/day led to greater decreases in scores on the Young Mania Rating Scale (YMRS) when compared with placebo (152). Differences between RISP and placebo were evident at just 3 days. Another trial (153) found that RISP was superior to placebo and showed similar efficacy to HPDL for mania. In a trial of children and adolescents between 10 and 17 years, RISP at dose ranges of 0.5 to 2.5 or 3 to 6 mg/day was superior to placebo (154).

The use of antipsychotics in youth with mania is reviewed in more detail in Chapter 14.

Autistic Disorder. Three clinical studies compared RISP with placebo for symptoms associated with autistic disorder (AD) (155). The largest was a double-blind, placebo-controlled multicenter study (156,157) in 101 children and adolescents with autism and found that RISP led to greater improvements in core symptoms of autism such as repetitive and stereotyped behaviors as well as tantrums and self-injurious behaviors. RISP-treated patients, however, experienced greater weight gain as well as fatigue and dizziness (see also Chapter 14).

RISP is approved in the United States for irritability associate with autistic disorder in children and adolescents between 5 and 16 years. Symptoms that improved with RISP include aggression towards others, self-injurious behaviors, tantrums,
and rapid mood change. Patients should be started on 0.25 mg/day for children weighing less than 20 kg and 0.5 mg/day for heavier children. Most patients respond between 0.5 and 2.5 mg daily.

Nonapproved Uses

Schizoaffective Disorder. Janicak et al. (158) studied the relative efficacy and safety of RISP versus HPDL in the treatment of schizoaffective disorder. Sixty-two patients (29 depressed type, 33 bipolar type) entered a randomized, doubleblind, 6-week trial of RISP (up to 10 mg/day) or HPDL (up to 20 mg/day). They found no difference between RISP and HPDL in the amelioration of psychotic and manic symptoms nor any significant worsening of mania with either agent. For the total PANSS, RISP produced a mean decrease of 16 points from baseline, compared with a 14-point decrease with HPDL. For the total Clinician-Administered Rating Scale for Mania (CARS-M) Scale, RISP and HPDL produced mean change scores of 5 and 8 points, respectively; and for the CARS-M mania factor, 3 and 7 points, respectively. In addition, RISP produced a mean decrease of 13 points from the baseline 24-item Hamilton Rating Scale for Depression (HAM-D) compared with an 8-point decrease with HPDL. In those patients who had more severe depressive symptoms (HAM-D baseline score > 20), RISP produced at least a 50% mean improvement in 12 of 16 (75%) patients in comparison with 8 of 21 (38%) patients receiving HPDL. HPDL produced significantly more EPS and resulted in more dropouts due to any adverse effect.

Other Disorders. RISP has been tried in several other conditions with mixed results, including

Tic disorders
Mental retardation or pervasive developmental disorders
Autistic disorder
Psychiatric disorders caused by a medical condition (e.g., acquired immunodeficiency syndrome [AIDS]; organic delusional disorders)
Agitated or aggressive states, including those associated with dementia (NB SGAs, including RISP, now carry a recently mandated box warning for increased risk of cerebrovascular AEs and a box warning for increased mortality risk in patients with dementia and psychosis; see also Chapter 15) (159,160); further, the risk of death in elderly patients (>65 years) may be even higher on FGAs versus SGAs (161)
Anorexia nervosa
Personality disorders (e.g., borderline and schizotypal)
Obsessive-compulsive disorder
Posttraumatic stress disorder
Various other psychotic disorders, such as brief psychotic disorder, delusional disorder, shared psychotic disorder, substance-induced psychotic disorder, and psychotic disorder not otherwise specified (162)

RISP is also reported to benefit major depressive disorder (with or without psychosis), bipolar mania, and schizoaffective disorder (see also Chapters 7 and 10) (152,153,163,164,165,166 and 167).

Risperidone for Conduct Problems. The effect of RISP in 63 adult, developmentally disabled men with disruptive, agitated, aggressive, or self-injurious behaviors was compared with their previous behavior on FGAs, usually HPDL (168). Although not blinded, the clinical staff observed behavior without knowledge of medication. Targeted behaviors were measured for 6 months before treatment with RISP began and then for each month after treatment. The mean dose of RISP was 5.6 mg/day, and the mean duration was 643 days. RISP lowered ratings of selfinjurious behavior from 55 to 18 and aggressive behavior from 23 to 9 while also producing a modest increase in appropriate social behaviors. The most striking finding of this mirror-image study was the reduction of aggression, which was not due to sedation because social interactions also increased.

In another study, 118 children with IQs ranging from 35 to 84 who demonstrated conduct disorder were randomized to RISP (at a mean dose of 1.23 mg/day) versus placebo in a double-blind design (169). In comparison with placebo, RISP produced a statistically significant reduction in insecure/anxious behavior, hyperactivity, self-injurious/stereotyped behavior, irritability, and aggressive/destructive behavior as well as an increase in adaptive social behavior. The latter is important because it shows that the changes are not due to sedation, although more somnolence was apparent with RISP. A study of 79 children aged 5 to 12 years with pervasive developmental disorder (170)
found an advantage for RISP over placebo for irritability, insecurity/anxiety, and hyperactivity. RISP-treated patients, however, experienced greater sedation and weight gain. Snyder et al. (171) compared RISP and placebo in 110 children with IQs between 36 and 84 who exhibited conduct problems. The RISP-treated children demonstrated greater improvements than children on placebo on scales measuring the severity of the conduct problems.

The response of aggressive behavior in 20 patients with conduct disorder was also evaluated in a double-blind, random-assignment, 10-week study comparing placebo with RISP (dose range, 1.5 to 3 mg/day) (172). Children aged 5 to 15 years without mental retardation (i.e., IQ > 80) and without attention deficit hyperactivity disorder (ADHD) were studied. The primary outcome variable was the rating of aggression against people and/or property scale. A score of 5 on this scale indicates that the person often attacks others and destroys property; 4, the person occasionally attacks people and destroys property; 3, the person either fights or is destructive; 2, the person is verbally abusive; 1, no aggression reported. At baseline, the children’s mean score was about 3.7. At 10 weeks, those on placebo had scores of about 3.5, but those on RISP had dropped to about 2.0.

The effectiveness and use of RISP microspheres is discussed in the section on “Long-Acting Antipsychotics.”

Switching from Clozapine to Risperidone

Patients may switch from clozapine to other SGAs, including RISP, because of

The adverse effects of clozapine
The frequent venipunctures required for hematological monitoring
The cost of clozapine and laboratory testing
The hope that RISP will be safer and at least as effective as clozapine

Early after the introduction of RISP, many patients who were doing reasonably well on clozapine risked deterioration by abruptly switching to RISP. Indeed, a number of case reports were published of patients who switched from clozapine to another SGA such as RISP and then relapsed. Clinical evidence exists for a clozapine withdrawal syndrome manifested by worsening psychosis. Because it is possible that clozapine is uniquely effective for some patients, before switching a previously antipsychoticrefractory patient benefiting from clozapine, the clinician should very carefully weigh the risk of clozapine discontinuation with the possibility of a safer, as well as comparable, therapeutic response to an agent such as RISP. This discontinuation is done best by cross-tapering or overlapping clozapine with the alternate agent (e.g., starting with a low RISP dose [0.5 mg/day] and gradually increasing it while reducing the clozapine over several weeks (173)).

Clozapine plus Risperidone

There is anecdotal clinical data indicating that RISP and clozapine can be overlapped or used concomitantly with beneficial results and no serious adverse reactions (173,174 and 175). Controlled trials have yielded mixed results. In one study (176), 40 patients who were nonresponsive or partially responsive to clozapine were randomly assigned to receive double-blind supplementation with either 6 mg of RISP or placebo. The patients who received RISP demonstrated greater improvements in positive symptoms without an increase in adverse effects. Two other controlled studies (177,178), however, found no advantage for adding RISP over placebo. A review of available studies found that a lower dose of RISP and a longer trial duration were associated with improvement on the combination (179).

Among the reports on overlapping or combining clozapine and RISP, some pharmacokinetic interactions were reported (174).

Because coadministered clozapine and RISP may result in elevated plasma levels of clozapine and norclozapine, plasma levels should be monitored when these drugs are prescribed concurrently. Not all patients receiving RISP as an adjunct to clozapine therapy, however, experience significant increases in blood levels. For example, in an open clinical trial, 12 patients were treated with clozapine (mean daily dose, 479.2 ± 121.5 mg; range 250 to 700 mg/day) and RISP (mean daily dose, 3.8 ± 1.4 mg; range 2 to 6 mg) (34). In seven patients, clozapine serum concentrations assayed before and after 4 weeks of RISP treatment increased nonsignificantly by 2.2% (i.e., from 374.6 ± 143.9 to 382.9 ± 218.3 ng/mL). The
addition of RISP to clozapine was well tolerated and produced significant reduction of symptoms on the total BPRS and the psychotic, negative symptom, and depression subscales. No patients complained of dizziness or demonstrated changes in their blood pressure or pulse.

Case Example

Because of a patient’s partial response to 5 months of clozapine therapy at 600 mg/day, RISP was added for augmentation (started with 0.5 mg twice daily and increased to 1 mg twice daily after 1 week). Before this addition, the clozapine plasma level was 344 ng/mL, but after 2 weeks of RISP augmentation, the level was elevated to 598 ng/mL with no adverse effects and substantial clinical benefit. In another report, there was an increase in the steady state plasma levels of clozapine (675 mg/day) and its active metabolite norclozapine after the addition of RISP 2 mg/day in a patient treated for 2 years. Before the addition of RISP, her clozapine and norclozapine levels were 829 and 1,384 ng/mL, respectively. Two days after RISP was added, these levels rose to 980 and 1,800 ng/mL. Clozapine dosage was reduced to 500 mg/day, and after 5 days of combined treatment with 4 mg/day of RISP, the clozapine and norclozapine levels were 110 and 760 ng/mL, respectively. Aside from some mild oculogyric crises, she had no symptoms of clozapine toxicity or clinical changes during the period of cross-tapering. In another case, RISP was added to clozapine because the patient had relapsed after discontinuation of fluphenazine and had not responded to clozapine alone. The addition of RISP resulted in an acute remission of psychosis (180).

Olanzapine

OLZ is available in a number of oral preparations, as well as a short-acting injectable and a long-acting injectable formulation.

Receptor-Binding Studies. OLZ is a thienobenzodiazepine derivative with a receptor-binding profile similar to that of clozapine (Table 5-2) (181). The effects of OLZ in several animal behavioral tests (e.g., conditioned avoidance response, apomorphine-induced climbing, 5-HTP-induced head twitch, punished responding, drug discrimination studies), neuroendocrine assays, and electrophysiological studies suggest that it has a pharmacological profile similar to that of clozapine and therefore may have similar antipsychotic effects in humans (182,183 and 184). Furthermore, as with clozapine, chronic OLZ administration selectively decreases the number of spontaneously active A-10, but not A-9, DA neurons (184). In contrast to RISP, which has a high affinity for the newly cloned 5-HT₇ receptor, OLZ has a greater affinity for the 5-HT₆ receptor (185).

Although early studies (186,187) suggested that patients treated with OLZ had D₂ receptor occupancies that were similar to those for clozapine, more recent studies indicate that they more closely resemble to those patients treated with FGAs and RISP. Kapur et al. (188) found that patients treated in the normal therapeutic range of 10 to 20 mg had D₂ occupancies of 71% to 80%. This suggests that the low propensity of OLZ for causing EPS in patients is not related to low D₂ occupancy but rather to 5-HT₂ receptors that were nearly saturated at the 5-mg dose.

Therapeutic Efficacy. In the United States, oral OLZ is approved for the following illnesses:

Schizophrenia (acute and maintenance) in adults and adolescents (13,14,15,16 and 17)
Acute treatment for manic or mixed episodes in bipolar I disorder in adults and adolescents (also as an adjunct to VPA or lithium)
Maintenance treatment of bipolar I disorder

Long-acting OLZ is indicated for schizophrenia. Acute intramuscular OLZ is approved for the treatment of acute agitation associated with schizophrenia and bipolar I mania.

OLZ combined with fluoxetine is approved for depressive episodes in bipolar I disorder and treatment-resistant depression.

The use of OLZ in bipolar disorder and major depressive disorder is discussed in Chapters 7 and 10.

Efficacy in Schizophrenia. Data from an international, fixed-dosing range (5 ± 2.5, 10 ± 2.5, and 15 ± 2.5 mg/day), controlled study demonstrated that OLZ significantly reduced symptoms in schizophrenic patients and showed a trend toward superiority in symptom reduction compared with HPDL (15 ± 5 mg/day) (189). The results indicate that the highest dose range, 15 ± 2.5 mg daily, was the most effective. Each of the OLZ dose groups experienced reduced acute EPS and negative symptoms compared with patients assigned to HPDL.

An international, multicenter, double-blind trial (190) addressed the acute efficacy and safety of a single-dose range of OLZ (5 to 20 mg/day) compared with a single-dose range of HPDL (5 to 20 mg/day). A total of 1,996 patients with a Diagnostic and Statistical Manual of Mental Disorders, 3rd ed., Revised (DSM-III-R) diagnosis of schizophrenia (83.1%), schizophreniform disorder (1.9%), or schizoaffective disorder (15%) participated in this study. The primary overall efficacy analysis (i.e., the difference in baseline to endpoint [last observation carried forward (LOCF) mean change on the BPRS]) found OLZ to be statistically superior to HPDL (i.e., —10.98; -7.93; p < 0.015).

In addition, in the international trial, the 300 patients with schizoaffective disorder (177 bipolar types and 123 depressive types) were the focus of a separate analysis. At the end of the maintenance phase, the bipolar subgroup had significantly greater improvement in mean change scores on the BPRS total score, the PANSS total score, the PANSS negative symptom subscale score, the CGI severity, and the Montgomery-Asberg Depression Rating Scale (MADRS). For the depressed subgroup, the magnitude of the improvement in BPRS total, PANSS total, and PANSS negative symptom scores was even greater than that observed in the bipolar subgroup. Overall, these results indicate that OLZ may be significantly superior to HPDL in the treatment of patients with schizoaffective disorder.

OLZ is available in a short-acting parenteral preparation. Several controlled and open-label trials assessed its efficacy, safety, and tolerability in acute agitation associated with schizophrenia, bipolar disorder, and dementia (191). In schizophrenia subjects, this formulation (5 to 10 mg) was, in general, superior to placebo and comparable to HPDL (7.5 mg). Sedation was more common with OLZ than with placebo.

Olanzapine Versus Clozapine. Tollefson et al. (114) reported the results of a large, collaborative (41 centers), random-assignment, parallelgroup, double-blind, placebo-controlled study of OLZ versus clozapine in patients who were nonresponsive to previous antipsychotics. The two drugs demonstrated similar effectiveness, with OLZ having advantages in its tolerability. Another study of refractory patients (118) yielded similar results. An interesting study by Conley et al. (192) gives a different perspective on the relative effectiveness of OLZ and clozapine for refractory patients. In a double-blind comparison of OLZ and CPZ (with benztropine), improvement was similar and modest in each group. The patients in this study were probably more severely ill than in the Tollefson study. Moreover, all of the patients failed to respond to a prospective trial of HPDL before entering the double-blind phase. In a later study (123), this same group also found that 41% of the patients who failed to respond to OLZ met criteria for improvement during a trial of clozapine. Phase II of CATIE compared OLZ and other antipsychotics with clozapine (119). As noted in the section on clozapine, there was an advantage for clozapine in this study of partially treatmentresistant patients.

Quetiapine

Receptor-Binding Studies. QTP is also a dibenzothiazepine derivative produced by altering the structure of clozapine. This agent has an affinity for multiple brain receptors, a low propensity to evoke EPS in preclinical tests considered predictive of antipsychotic activity, and does not produce sustained plasma prolactin levels.

In vitro binding studies demonstrate that QTP has low-to-moderate affinity for D₁, D₂, 5-HT_1A, and 5-HT_2A receptors and moderate-to-high affinity for α₁– and α₂-adrenergic receptors. Like other SGAs, QTP also has greater relative affinity for 5-HT_2A receptors than for D₂ receptors, only weakly induces catalepsy in rats, and causes few EPS in humans (193,194). QTP lacks appreciable affinity for muscarin ic cholinergic and BZD
binding sites and causes selective activation of the A-10 DA neurons when given acutely and depolarization inactivation of A-10 DA neurons when given chronically (195). Early PET studies of QTP receptor occupancy indicated substantially lower D₂ occupancy than with other SGAs, with the exception of clozapine, which is also associated with lower occupancy (196). Kapur et al. (13) found that QTP led to transiently high D₂ occupancy of 58% to 64% during the first 2 to 3 hours, which fell to minimal levels by 12 hours. The extended and immediate release formulations of QTP result in similar plasma levels and D₂ occupancy (197). These studies suggest that transient occupancy may be sufficient for an antipsychotic effect. These observations provide strong support for Seeman’s proposal (106) that an important property of SGAs is their ability to occupy D₂ receptors briefly and then rapidly dissociate from them to allow normal DA neurotransmission.

Therapeutic Efficacy. In the United States, QTP is approved for the following conditions:

Schizophrenia (acute and maintenance) in adults and adolescents (13,14,15,16 and 17)
Acute treatment for manic or mixed episodes in bipolar I disorder in adults and adolescents (also as an adjunct to VPA or lithium)
Depressed episodes in bipolar I and II disorder
Maintenance treatment of bipolar I disorder
Adjunctive treatment in major depressive disorder

The use of QTP in bipolar disorder and major depressive disorder is discussed in Chapters 7 and 10.

Efficacy in Schizophrenia. The efficacy and safety of QTP were tested in 109 schizophrenic patients in a multicenter, randomized, doubleblind, placebo-controlled, parallel-group trial (198). Subjects randomized to QTP initially received 25 mg three times per day for 1 to 2 days. Thereafter, the dose was titrated upward to a maximum daily dose of 750 mg. Analysis of BPRS factor scores showed that treatment with QTP resulted in selective improvement in the psychosis-related factors of thought disturbance and hostility/suspiciousness as well as improvement in the mean BPRS positive symptoms cluster score. By day 21, QTP was also clinically and statistically superior to placebo in moderating negative symptoms.

A study in 286 subjects (199) evaluated QTP at high doses (>250 mg, but <750 mg), low doses (<250 mg), or placebo. Only the higher dose was related to significantly greater improvement when compared with placebo, suggesting that the effective dose was greater than 250 mg. Arvanitis and Miller (200) reported a multiple fixed-dose, placebo-controlled, double-blind study of QTP in comparison with HPDL and placebo in acutely exacerbated patients with chronic schizophrenia. QTP was administered in five doses: 75, 150, 300, 600, and 750 mg/day; HPDL was given at 12 mg/day. The study design had slightly more than 50 patients in each group. The 75-mg dose of QTP was clearly less efficacious than the higher doses. Doses of 150 to 750 mg/day were superior to placebo and comparable with HPDL in reducing positive symptoms, and 300 mg/day was superior to placebo and comparable with HPDL for negative symptoms.

A comparison of QTP (mean dose, 455 mg/day) and HPDL (mean dose, 8 mg/day) in 448 acutely psychotic patients found similar efficacy for the two agents (201). Another study compared 600 mg/day of QTP with 20 mg/day of HPDL in patients only partially responsive or nonresponsive to a trial of fluphenazine (20 mg/day) (202). There was a nonsignificant trend toward an advantage for QTP in total PANSS score. Another study compared 200 mg bid of immediate release QTP with 400, 600, or 800 mg of extended release (ER) QTP administered once daily in acute schizophrenia. All four drug groups were superior to placebo and showed similar effectiveness (203).

The data from these controlled trials provide relatively sparse information regarding the optimal dose of QTP for patients with acute schizophrenia. The data from the Arvanitis and Miller paper suggested that 300 mg/day may be an appropriate dose. It is the authors’ experience and that of others (204), however, that substantially higher doses in the range of 600 to 750 mg/day may be more reasonable for most patients and that some may do well with even higher doses (e.g., 1,000 mg/day). Unfortunately, these higher doses were seldom used in controlled trials. There is evidence from one study of only 21 patients indicating that QTP is effective when administered as a single daily dose (205). Once-daily administration of an ER formulation
of QTP (Seroquel XR) demonstrates similar efficacy when compared with twice-daily administration of the immediate release formulation (203,206).

Ziprasidone

Receptor-Binding Studies. Among the SGAs, ZPD has a distinctive pharmacological profile. Binding studies indicate that ZPD has a high affinity for D₂ receptors, approximately equal to that of RISP (Table 5-2). In vitro, ZPD

Is a moderately potent D₂ receptor antagonist (K; = 8.32)
Has a high affinity for the D₃ receptor (pK. = 8.14)
Shows a 100-fold lower affinity for the D₁ receptor (approximately equivalent to that of RISP and HPDL) than for the other DA receptor subtypes

ZPD also has a significant affinity for several 5-HT receptor sites, including very potent antagonistic activity at 5-HT_2A receptors (pKi = 9.38) (207,208). ZPD also has potent antagonistic action at the 5-HT_2C receptor, which may contribute to its capacity to alleviate positive symptoms. Furthermore, its 5-HT_1A partial agonism may improve negative and cognitive symptoms (209).

ZPD has a high 5-HT_2A/D₂ receptor-affinity ratio. It is also an antagonist at the α₁-adrenoreceptor, but its affinity there is half that of its D₂ affinity. In addition, compared with its affinity for D₂ and 5-HT_2A receptors, ZPD has a relatively low affinity for H₁ receptors (pKi = 7.33). In vitro, ZPD is a moderately potent inhibitor of neuronal reuptake of norepinephrine; serotonin; and, to a lesser extent, DA. This property is shared with some antidepressants and contrasts with other SGAs, which are inactive at all three monoamine uptake sites.

PET studies of ZPD’s displacement of [11C]raclopride at D₂ receptors have revealed decreased raclopride binding in a dose-dependent manner up to single doses of 60 mg. ZPD receptor occupancy was >65% after single doses of 20 to 40 mg, with occupancy reaching a maximum of approximately 80% following a single dose of 60 mg (208). The time course and proportion of 5-HT₂ receptor occupancy by single doses of ZPD (40 mg) was determined in normal controls using the PET ligand [¹⁸F]setoperone (210). Serum ZPD concentrations ranging from approximately 4 to 125 ng/mL at time points 4 to 18 hours postdose were associated with 5-HT₂ occupancy rates from about 50% to nearly 100%. Receptor occupancy was related to plasma concentration, suggesting that twice-daily administration of 20 mg or more to steady state levels would be expected to provide greater than 80% occupancy of 5-HT₂ receptors.

Another PET study in patients indicated a high 5-HT₂/D₂ occupancy ratio and that the optimal effective dose of ZPD may be closer to 120 mg/day (207).

Overall, the results of PET, pharmacokinetic, and pharmacodynamic studies indicate that a twice-daily dosage regimen is appropriate.

Therapeutic Efficacy. In the United States, ZPD is approved for the following conditions:

Schizophrenia
Acute treatment as monotherapy for bipolar I manic or mixed episodes
Maintenance treatment of bipolar I disorder as an adjunct to lithium or VPA
As an intramuscular injection for the acute treatment of agitation in schizophrenia patients

The use of ZPD in bipolar disorder and major depressive disorder is discussed in Chapters 7 and 10.

Efficacy in Schizophrenia. ZPD (daily dose range, 40 to 160 mg given on a twice-daily schedule with food) has been evaluated in acutely disturbed patients with schizophrenia or schizoaffective disorder. A study by Goff et al. (211) compared 15 mg of HPDL with 4, 10, 40, or 160 mg of ZPD daily. Both HPDL and 160 mg of ZPD were superior to the 4 mg of ZPD. Another study compared flexible doses of 80 to 160 mg of ZPD with 5 to 15 mg of HPDL (212). Improvements in both groups were similar, but ZPD had a significant advantage for negative symptoms. A summary of the clinical literature on ZPD is published electronically by the FDA. Based on this data set, we combined four studies (213,214) showing that ZPD in doses of 120 to 200 mg (Table 5-7) is both clinically and statistically superior to placebo and as effective as HPDL in acutely ill patients, with beneficial effects on both positive and negative symptoms. Doses below 100 mg were less effective than the higher doses.

TABLE 5-7 ZIPRASIDONE VERSUS PLACEBO OR HALOPERIDOL FOR ACUTE TREATMENT OF SCHIZOPHRENIA OR SCHIZOAFFECTIVE DISORDER

Ziprasidone/Placebo Difference				Ziprasidone/Haloperidol Difference
No. of Studies	n	Effect Size Units	p Value	No. of Studies	n	Effect Size Units	p Value
4	608	+0.46	0.00000002	1	165	-0.29	0.04

In one double-blind, randomized, placebocontrolled study, 132 acute patients with schizophrenia or schizoaffective disorder were given a daily dose of 40 or 120 mg ZPD or placebo for 28 days (215). There was a statistically significant improvement in psychotic symptoms versus placebo in the 120 mg/day ZPD group as measured by the total BPRS and the CGI scores. Evaluations for parkinsonian symptoms, akathisia, abnormal movements, and sedation did not reveal any notable treatment effects. No significant differences existed between drug and placebo in the total number of AEs, laboratory test abnormalities, or more serious AEs. Thus, this study documented that 60 mg of ZPD twice daily was an effective strategy with minimal risks.

Another report involved a 6-week, multicenter, double-blind, placebo-controlled study that compared the efficacy, safety, and tolerability of two fixed-dose regimens of ZPD in patients with an acute exacerbation of schizophrenia or schizoaffective disorder (216). Both the 80- and the 160-mg dose groups demonstrated statistically significant changes from the baseline BPRS, CGI severity, and PANSS total scores. Negative symptoms on the PANSS subscale also showed statistically significant differences favoring the ZPD groups over placebo. This trial also demonstrated efficacy for affective symptoms, as measured by the MADRS. In this context, two placebo-controlled trials for schizoaffective disorder also suggested that ZPD may be useful both for affective and psychotic symptoms.

Efficacy in Acute Agitation. ZPD is also available in a short-acting, parenteral preparation. Studies (217,218) indicate that short-acting intramuscular formulations of ZPD and HPDL demonstrate similar efficacy for agitated patients, with HPDL resulting in more EPS. A 6-week trial found similar advantages to starting patients on intramuscular ZPD and changing to the oral formulation compared with the same sequence using HPDL (219).

These studies indicate that the optimal dose of intramuscular ZPD is between 10 and 20 mg.

Aripiprazole

ARIP is available in oral formulations and a short-acting intramuscular formulation. Studies are also presently ongoing with a long-acting injectable formulation.

Receptor-Binding Studies. ARIP is the first approved antipsychotic that is not a pure D₂ antagonist. Rather, it is a partial agonist at D₂ (as well as 5-HT_1A) receptors. Like other SGAs, it is also an antagonist at the 5-HT₂ receptor. As an effective partial agonist (5), ARIP binds with high affinity to D₂ receptors but with a lower level of intrinsic activity. As a result, if there are high levels of DA in a particular area of the brain, ARIP will replace DA, leading to a reduced DA effect due to its lower intrinsic activity. If an area is low in DA, ARIP can lead to a net increase in DA activity. These differing actions at DA receptor subtypes in distinct cellular locations may produce different functional results in comparison with drugs that are D₂ postsynaptic receptor antagonists only. Theoretically, this could lead to improving psychosis by decreasing DA activity in limbic areas and possibly improving negative symptoms and cognitive impairments by indirectly (e.g., via 5-HT effects) increasing DA activity in frontal areas. PET studies (220) confirm that doses in the therapeutic range for ARIP result in D₂ receptor occupancy of up to 90%. However, even at these higher occupancies, EPSs are minimal. In addition, ARIP has moderate affinity for α₁-NE and H_l receptors, indicating that cardiovascular, sedative, and weight gain problems should be lessened compared with other SGAs.

Therapeutic Efficacy. ARIP is approved in the United States for the following conditions:

Schizophrenia in adults and adolescents (ages 13 to 17 years)
Acute treatment of manic or mixed episodes associated with bipolar I disorder as monotherapy and as an adjunct to lithium or VPA
Maintenance treatment of bipolar I disorder
Treatment of irritability associated with autistic disorder

As an injection for the acute treatment of agitation associated with schizophrenia or bipolar I disorder.

The efficacy of ARIP for acute schizophrenia has been demonstrated in a series of doubleblind controlled trials. In one study, Kane et al. (221) compared 15 and 30 mg daily of ARIP to 10 mg daily of HPDL or placebo over 4 weeks in 414 patients with schizophrenia or schizoaffective disorder. Both ARIP doses and HPDL were significantly superior to placebo for PANSS total, PANSS positive, and CGI. ARIP (15 mg) and HPDL were also superior to placebo for PANSS negative item symptoms. Another 4-week study (222) compared 20 and 30 mg daily of ARIP with 6 mg of RISP and placebo. Both doses of ARIP and RISP were superior to placebo for PANSS total, positive, and negative symptoms as well as the CGI.

Reviews of controlled trials in acute schizophrenia (223) concluded that once-daily doses of 10 to 15 mg were consistently effective (223,224). There is no indication from these trials that higher doses are associated with increases in efficacy. A Cochrane review (225) concluded that ARIP was effective, but there was inadequate information to conclude that it was superior to other FGAs or SGAs. A recent post hoc analysis of five short-term trials also indicated ARIP was superior to placebo across several symptom factors (e.g., positive, negative, disorganized thoughts, hostility, mood) (225a).

Acute parenteral ARIP was studied in 357 acutely agitated patients with schizophrenia, schizoaffective disorder, or schizophreniform disorder (226). Patients were randomized to one of four doses of intramuscular ARIP (1, 5, 10, or 15 mg), intramuscular HPDL (7.5 mg), or placebo. Based on improvement in the PANSS-Excited Components (PEC), ARIP and HPDL were superior to placebo over a 2-hour period, but ARIP did not produce excessive sedation and was associated with only minimal pain at the injection site.

Paliperidone

Paliperidone is available as an oral ER formulation and as a long-acting injectable formulation (i.e., paliperidone palmitate). Paliperidone ER is administered in a capsule that contains the drug itself as well as an osmotically active component. Water in the gastrointestinal tract enters the capsule and osmotic pressure pushes the drug out of the capsule through very small holes. This leads to a controlled release of the drug and permits once-a-day dosing.

Receptor-Binding Studies.Paliperidone (PALI) is 9-hydroxy RISP, the major active metabolite of RISP. As a result, it shares similar binding properties to RISP. That is, PALI has substantial antagonist activity at both D₂ and 5HT_2A receptors as well as activity at α_1;, a₂, and H₁ receptors. PET studies found that doses of paliperidone (ER) between 6 and 9 mg resulted in D₂ occupancy of 70% to 80% (227).

Therapeutic Efficacy

Paliperidone ER (oral) is approved in the United States for the following conditions:

Acute and maintenance treatment of schizophrenia
Acute treatment of schizoaffective disorder as monotherapy
Acute treatment of schizoaffective disorder as an adjunct to mood stabilizers and/or antidepressants

Paliperidone palmitate (long-acting injectable) is approved in the United States for the acute and maintenance treatment of schizophrenia in adults.

Efficacy in Schizophrenia. Five studies compared PALI with placebo in acute schizophrenia (228). Overall, PALI was more effective than placebo in managing psychosis in this population. Three trials included OLZ and found no advantage to either agent for reducing psychotic symptoms. A pooled analysis of three 6-week studies (229) is helpful in comparing doses of PALI. Doses of PALI between 3 and 15 mg were all superior to placebo. PALI was more effective than placebo for preventing relapse in stabilized patients (230). Unfortunately, there are no studies at the time of this publication comparing RISP with PALI (230a). As a result, there is no data to support selecting
PALI which is substantially more expensive when compared with generic RISP.

One controlled trial, however, found PALI superior to QTP in 309 patients with a recent exacerbation of their schizophrenia (230b).

Asenapine

Receptor-Binding Studies. Asenapine (ASEN) has a relatively complex pharmacology. It is an antagonist at 5HT_2A, 5HT_2B, 5HT_2C, 5HT₆, and 5HT_7α (1A); α_2A, α_2B, and α_2C; and D₂, D₃, and D₄ dopaminergic receptors. As with other new antipsychotics, it has a relatively high 5HT_2A to D₂ affinity (231). Studies in rodents suggest that ASEN increases cortical DA (232). Another study found that ASEN led to increases in DA, norepinephrine, and ACh efflux in the rat medial prefrontal cortex and hippocampus (233). If these effects are meaningful in humans, it could lead to improvements in negative symptoms and impaired cognition in schizophrenia.

Therapeutic Efficacy. Asenapine is approved in the United States for the following conditions:

Acute treatment of schizophrenia in adults
Acute treatment of manic or mixed episodes in bipolar I disorder in adults

Efficacy in Schizophrenia. In three, 6-week trials, asenapine 5 mg bid was compared with placebo and an active control (233, 233a, 233b). Asenapine was superior to placebo for both positive and negative symptoms and there was evidence in one trial of an advantage compared with RISP for negative symptoms. This advantage, however, was not seen in other trials in schizophrenia.

Iloperidone

Receptor-Binding Studies. Iloperidone (ILP), RISP, and PALI are benzisoxazoles. As a result, they share similar properties at various neuroreceptors. ILP shows high affinity for α₁, D₃, and 5-HT_2A receptors; intermediate affinity for α_2C, D₂, and D₄, and 5-HT_1A, 5-HT_1B, 5-HT_2C, and 5-HT₆ receptors (234). The relatively high 5-HT_2A to D₂ ratio may explain its relatively low EPS risk and the high affinity for α₁ explains its ILP risk for orthostatic hypotension (234a).

Therapeutic Efficacy. Iloperidone is approved in the United States for the acute treatment of schizophrenia in adults. It comes with two warnings:

Because it may prolong the QT interval, other antipsychotics should be considered first.
There is a need to titrate the dose due to orthostatic hypotension. This may delay the onset of effectiveness.

Early phase II trials of ILP showed that it was effective at doses of 8 mg or greater and tolerated at doses up to 32 mg/day (235). More recent studies evaluated ILP in acutely-ill schizophrenia and schizoaffective patients. These studies showed that ILP separated from placebo at doses between 8 and 24 mg (236). Another acute trial (237) compared 24 mg of ILP with 160 mg of ZIP. Both active drugs separated from placebo. A careful look at these studies shows that the response to ILP is commonly delayed by the need for titration. A 46-week follow-up of patients receiving ILP and HPL found a similar risk for relapse (238), suggesting that ILP is effective for maintenance treatment.

The value of genotyping to predict which patients are more likely to respond to ILP was evaluated in a large clinical trial (239). Using a six-marker genotype combination, four groups of patients were evaluated. For the “optimal” genotype, 75% of iloperidone-treated patients showed a 20% or greater improvement, whereas 37% for patients with other genotypes met this criterion. This is an interesting approach to predicting drug response and/or adverse effects, which may persuade clinicians to select ILP for some patients with schizophrenia.

DOSING STRATEGIES

The goal of pharmacotherapy for acute schizophrenia is to minimize psychotic symptoms that impair the patient’s ability to function and cause substantial distress. For some patients, drug treatment results in a complete remission of psychotic symptoms. Others continue to experience symptoms that are attenuated. Premature attempts to treat the lingering psychosis by increasing the dose, changing antipsychotics, or adding a second agent may not lead to further improvement and may worsen adverse effects. As with most treatments in medicine, antipsychotics diminish symptoms but may not alter the
underlying disease process. The result is that patients may enjoy a robust antipsychotic effect for positive symptoms but still retain psychopathology in other domains such as negative symptoms or cognitive impairments. Although goals may include eliminating every trace of hallucinations or delusions, even a substantial improvement in these positive symptoms may be associated with reduced suffering, improved functioning, and a better quality of life.

Only gold members can continue reading. Log In or Register to continue

Stay updated, free articles. Join our Telegram channel

Tags: Principles and Practice of Psychopharmacotherapy

Aug 27, 2016 | Posted by drzezo in PHARMACY | Comments Off

Basicmedical Key

Fastest Basicmedical Insight Engine

Treatment with Antipsychotics

Like this:

Related

Stay updated, free articles. Join our Telegram channel

Full access? Get Clinical Tree

Basicmedical Key

Fastest Basicmedical Insight Engine

Treatment with Antipsychotics

Share this:

Like this:

Related

Related posts:

Stay updated, free articles. Join our Telegram channel

Full access? Get Clinical Tree