Treatment of Refractory Schizophrenia

Original Author: Bruce Alexander, Pharm.D, BCPP
Latest Reviser: Bruce Alexander, Pharm.D, BCPP
Creation Date: 1996
Peer Review Status: Internally Peer Reviewed

The introduction of the first antipsychotic, chlorpromazine, into clinical practice in 1952 represented a major advance in the treatment of schizophrenia. Previous somatic treatment including continuous sleep therapy, insulin coma, electroconvulsive therapy (ECT), and psychosurgical treatments had been utilized but were found to be of limited benefit and/or associated with serious adverse effects. Reserpine was the first pharmacologic intervention in schizophrenia, and although it had antipsychotic activity, it was of limited benefit due to the adverse effects of depression, hypotension, peptic ulcers, exacerbation of asthma, and extrapyramidal side effects (EPSE). Thus prior to the availability of chlorpromazine, clinical treatment of schizophrenia was for the most part custodial with extensive use of seclusion, physical restraint, and various forms of hydrotherapy. With widespread use of chlorpromazine came reports of dramatic clinical improvement in schizophrenic patients and admission rates to psychiatric facilities began to decrease. Several years thereafter, trifluoperazine became available and added to the pharmacologic armamentum of antipsychotic agents. Over the ensuing 40 years several classes of compounds with antipsychotic activity have been developed, but until the recent development of clozapine and risperidone, none have been shown to possess superior antipsychotic activity.

Although antipsychotic medications remain the pharmacological foundation for the treatment of schizophrenic patients, it is obvious that not all patients benefit from these drugs. The goal of this chapter is to review the efficacy of various pharmacologic treatment strategies for patients with schizophrenia who respond poorly to treatment with conventional antipsychotic medications.

DEFINITION OF TREATMENT-RESISTANT SCHIZOPHRENIA

How do we define treatment-resistant schizophrenia? There have been numerous discussions as to what constitutes "treatment-resistance", yet there remains no consensus on this issue. This lack of consensus in definition of treatment-resistant schizophrenia is exemplified by the wide variation in terms used to describe poor response to antipsychotic medications (e.g. treatment-refractory, treatment-resistance, nonresponse, inadequate response, etc.). In addition, the criteria used for research exploring the efficacy of new treatments are likely to be much more rigid than those criteria used in clinical practice.

One definition for treatment-refractoriness that has received fairly wide acceptance is the definition used in the multicenter trial of clozapine (Kane 1989). This is a rigorous set of research criteria designed specifically to study the efficacy of a novel agent for treatment-refractory schizophrenia. The criteria are as follow:

1. Clinical History: Absence of periods of good functioning in the preceding five years during treatment with antipsychotics at doses equal to or greater than chlorpromazine 1000 mg/day. During this time the patient must have received at least two different chemical classes of antipsychotic for six-weeks or more with no episodes of significant relief;
2. Cross-sectional: The patient must have a Brief Psychiatric Rating Scale (BPRS) score " 45 (standard 18-item version , i.e. absent =1 through very severe =7) with item scores " 4 (moderate severity for two of the following four items: conceptual disorganization, suspiciousness, hallucinatory behavior, and unusual thought content) and a Clinical Global Impression (CGI) score " 4 (moderately ill);
3. Prospective Clinical Response: Failure to decrease BPRS score by 20% or below 35, or decrease CGI score to 3 (mildly ill) after a six-week prospective trial of haloperidol up to 60 mg/day.

In clinical practice, these very explicit criteria can be difficult to apply. Some patients who easily meet historical and cross-sectional criteria may have a substantial increase in their BPRS score when their antipsychotic medication is discontinued and may then have a greater than 20% reduction in BPRS score with reinstitution of treatment. After reinstitution of treatment they may be no more improved than prior to switching medications, but may appear to be responsive to medication by recovering from the exacerbation. Using this rigorous definition, such patients would not be considered treatment-refractory despite their unacceptable response to treatment.

In 1990, Brenner and members of an international study group published a position paper which represented ten years of work on defining treatment refractoriness in schizophrenia (Brenner et al 1990). The consensus of this group was that treatment refractory schizophrenia should be defined as follows: "the presence of ongoing psychotic symptoms with substantial functional disability and/or behavioral deviances that persist in well-diagnosed persons with schizophrenia despite reasonable and customary pharmacological and psychosocial treatment that has been provided for an adequate time period." In this paper, they recommended that a conservative diagnosis of schizophrenia be employed requiring continuous positive and characteristic symptoms of psychosis for at least two years. They further recognized that concomitant negative symptoms and disabling nonpsychotic symptoms are often present in the clinical picture of refractory schizophrenia (Brenner et al 1990).

The study group stated that in addition to persistent positive and negative symptoms of psychosis, associated functional disability in social, self-care, and occupational domains should be present for the patient to meet criteria for refractoriness. They noted that some patients with persistent psychotic symptoms may function with mild to moderate disabilities whereas others may have little in measurable psychotic symptoms while exhibiting bizarre and disabling behavioral deviances and acting out (e.g. incontinence, aggression, self injury, and fire setting) which may prohibit community placement (Brenner et al 1990).

This group also developed pharmacological criteria for defining refractoriness which was very similar to those used by Kane et al, 1989. However, they added that if blood levels had documented poor bioavailability of an antipsychotic drug, long-acting antipsychotic administration should have been shown to offer no significant improvement. They stated that it was important to exclude both inadequate dose and blood concentrations of antipsychotic drug as well as toxic concentrations of drug as the cause of poor response. They concluded that treatment-refractoriness is a multidimensional concept with an important component of psychosocial function and described seven levels of treatment response and resistance, including moderate and severe resistance, refractoriness and severe refractoriness based on severity of positive and negative symptoms, clinical global impression (CGI) and personal and social functioning (Brenner et al 1990).

In a commentary responding to this position paper, Meltzer agreed that the criteria for treatment-refractoriness for the use of clozapine need to be set high because of the risk of agranulocytosis, but argued that a more liberal definition may be used when considering other changes, such as the use of new medications without the risk of agranulocytosis (Meltzer 1995). He proposed that schizophrenic patients who do not return to their highest premorbid level of functioning should be considered treatment resistant to the extent their previous level of functioning is further impaired. Thus he conceptualized treatment-refractoriness as a relative term dependent on the previous level of premorbid functioning, taking into account the various social disruptions created by the illness such as interruptions in working, strain on personal and family relationships and negative impact on self esteem. This is a more clinically oriented approach in contrast to rigorous research criteria for treatment-refractoriness.

For the present purposes, we will define treatment-resistant schizophrenia as the presence of persistent positive psychotic symptoms such as hallucinations, delusions, bizarre behavior, or formal thought disorder and/or enduring negative symptoms such as affective flattening, alogia (poverty of speech, poverty of content of speech, blocking, or latency of response), avolition (lack of drive and physical anergia), anhedonia (lack of pleasure in normal interests), and asociality (impaired intimacy and few relationships with friends/peers) despite adequate treatment with antipsychotic medications. Most clinicians would agree that a patient is exhibiting a less than adequate response if he/she is receiving an antipsychotic at a dose equivalent to "1000 mg/d of chlorpromazine and continue to experience at least one positive psychotic symptom of marked severity, or two or more negative symptoms of moderate severity in addition to at least one positive psychotic symptom of moderate severity for the majority of the time.

Nevertheless, it is recognized that the definition of treatment-resistant schizophrenia used by the clinician is by necessity more flexible, and will be dictated by the treatment approach being considered. More lenient definitions may be used in deciding whether to adjust antipsychotic doses based on blood concentrations, switch to another antipsychotic other than clozapine, or add an adjunctive treatment (e.g. benzodiazepine or lithium). In contrast, more stringent criteria should be used when switching to a less benign treatment such as clozapine is considered.

INCIDENCE OF TREATMENT-RESISTANT SCHIZOPHRENIA

The precise incidence of poor treatment response in schizophrenia is not easy to determine. Data from the National Institute of Mental Health Psychopharmacology Research Branch multicenter, controlled trials of newly admitted patients who received antipsychotic treatment, found that about 10% of patients demonstrated "no change" or were "worse" and approximately 20% of patients were considered only "minimally improved" (Cole et al 1964; Davis et al 1980; Goldberg et al 1965). Kane et al concluded from these studies that 10 to 20% of patients "derive little benefit from typical antipsychotic medications" (Kane 1989).

Data from recent controlled trials have shown that 52-78% of schizophrenic patients respond favorably to 4-6 week trials of conventional antipsychotics (Lautin et al 1980; Mattes et al 1985; Pickar et al 1992; Siris et al 1987; Small et al 1975). This suggests that although not all of the nonresponders are necessarily treatment-resistant, 25 to 50 percent of patients with schizophrenia have a less than adequate response to conventional antipsychotics. In general clinical practice the impact of non-compliance on the incidence of poor treatment response is substantial and represents an entirely separate issue though we are not able to examine this important clinical problem in the scope of this chapter.

PHARMACOLOGICAL APPROACHES TO TREATING RESISTANT PATIENTS

The first step in approaching treatment-resistant patients is to optimize treatment with their present antipsychotic medication. This involves determining whether the individual has been on the medication for a long enough duration as well as whether the dose is adequate. The rate of improvement with antipsychotic medication may vary significantly among patients, with some showing improvement within several days of initiation of the antipsychotic while others may require several weeks (Karson et al 1982). Most literature suggests that the majority of the therapeutic gain occurs in the first four weeks of antipsychotic treatment, although 12-24 weeks may be needed (Davis et al 1977; Levi-Minzi et al 1991). However, it has been noted that many of these studies have methodological flaws (Breier et al 1987; Karson et al 1982). A recent controlled study by Kinon et al reported that only seven percent of patients who failed to respond after four weeks of treatment responded with an additional four weeks antipsychotic trial (Kinon et al 1993). 63 Kinon et al recommended that an adequate trial of antipsychotics in patients with first-episode psychosis or with acute exacerbation of chronic psychosis require at least four weeks at an adequate dose.

In spite of more than 40 years of use, establishing an adequate dose of a conventional antipsychotic for a given patient is generally done in a simple trial and error manner by adjusting the dose until a reduction of symptoms occurs or side effects prohibit further increase. Because the dosage of antipsychotic drugs only weakly predicts their therapeutic effects, and it is known that there is marked individual variability in the handling of antipsychotics, there has been a strong interest in ascertaining whether antipsychotic plasma concentrations may predict treatment outcome (Davis 1974).

There have been numerous studies comparing plasma concentrations and therapeutic response in schizophrenic patients receiving various standard antipsychotics. Many of the early studies relating antipsychotic plasma concentrations to clinical response suffered from poor study design (i.e. insufficient sample sizes, the inclusion of heterogeneous patient populations, non-random dose adjustments based on clinical response, concurrent treatments, nonstandardized timing of blood sampling, insufficient clinical evaluations, etc.). More recent studies have corrected many of these methodological problems. Nevertheless, present data do not allow a general recommendation for routine plasma concentration monitoring and subsequent dosage adjustment of any conventional antipsychotic, except possibly haloperidol (Dahl 1986; Davis 1974; Delva et al 1982).

Due to its simpler metabolic profile, haloperidol has been the most studied antipsychotic in terms of correlating serum levels and therapeutic response (Perry et al 1991; Silver et al 1992). However, the literature contains a number of discrepant findings. Of 29 studies, 15 report no relationship, five report a linear response, and 14 report a therapeutic window of approximately 5 to 15 ng/ml (Lingjaerde et al 1979; Marder et al 1994; Silver et al 1992). According to recent literature, maximal response may plateau at 15 ng/ml (Kinross-Wright 1955; Klien et al 1984; Miller et al 1994). Currently, levels exceeding 15 ng/ml are not recommended. More studies are needed to clarify the utility of maintaining levels at 5 ng/ml or less.

The wide variation in individual metabolism of antipsychotics creates a very loose relationship between dosage and plasma concentration. Therefore, Kane and Marder suggested that monitoring antipsychotic plasma concentrations measurements may be considered for the patient refractory to a four-week trial of therapeutic doses of typical antipsychotic medication to ensure that lack of response is not related to poor bioavailability (Donlon et al 1979). Although data on what should be considered a therapeutic plasma concentration for other antipsychotics are less conclusive, one can identify wholly noncompliant patients and get a rough idea as to whether the individual is a "rapid metabolizer" who may require a dose adjustment.

If antipsychotic blood levels are obtained, the recommend sampling time is just prior to the next dose. For most oral antipsychotics the patient should have received at least one week of consistent doses prior to sampling, and for decanoate preparations two to three months of consistent dosage is needed prior to sampling (Perry et al 1991). Following these guidelines will aid in the interpretation of subsequent antipsychotic levels. Dosing adjustments to achieve a desired level are made proportionately as antipsychotics follow linear kinetics with typical doses. Meaningful concentration information can not be determined if the patient is receiving more than one antipsychotic concurrently.

Management of the patient with an inadequate response after a four week treatment course includes five options: (1) continuing the same regimen for a longer duration; (2) increasing the dose of the current antipsychotic; (3) switching to a different class within the typical antipsychotic group (e.g., butyrophenone (haloperidol) to a phenothiazine(thioridazine)); (4) adding a second drug to the current antipsychotic (i.e., adjunctive treatment); or (5) switching to a newer agent such as risperidone or clozapine (Kane et al 1988; King et al 1980; Sramek et al 1988). The relative efficacy of these strategies has not been directly compared (Kane 1987; Karson et al 1982). A recent study directly evaluated the first three recommendations (King et al 1980). Patients who failed a four week trial of fluphenazine 20 mg/d were randomized to receive four additional weeks of fluphenazine 20 mg/d, fluphenazine 80 mg/d, or haloperidol 20 mg/d. Overall, there was no difference between the treatments as only 5%, 13% and 8% of patients, respectively, had responded after an additional four weeks of treatment. Though this study suggests that these strategies may not significantly increase the rate of response in patients, an extended trial on the same or increased dose may be considered for an individual patient who fails a four week trial. Often the presence of adverse effects may play a pivotal role in this decision making process. For example, if the patient is experiencing EPSEs, then maintaining the same dose for an additional four weeks might be attempted or the patient may be switched to another antipsychotic. However, if no EPSE or other side effects are present then the dose might be increased. No studies to date have directly compared the first three strategies to adjunctive treatments. The use of adjunctive treatments in combination with conventional antipsychotics will be addressed later in this chapter.

HIGH DOSE ANTIPSYCHOTIC TREATMENT

Although controversial, high-dose treatment with antipsychotics has been suggested as a possible management strategy for refractory patients. Uncontrolled trials using supranormal doses of antipsychotics have suggested that some patients respond better to high than to low doses of antipsychotics. Daily doses of chlorpromazine 5000 mg, fluphenazine 1500 mg, haloperidol 100 mg, as well as depot injections of fluphenazine enanthate starting at 250 mg have been used. However, a review of 11 double-blind studies failed to demonstrate a significant superiority of megadosing versus conventional dosing (Aubree et al 1980). Seven of 11 studies reported extrapyramidal side effects were more common in the high-dose group. A more recent review of 33 studies agreed with these earlier recommendations (Baldessarini et al 1988). After reviewing four studies that used high doses of fluphenazine or haloperidol, Kane concluded that high doses were of little benefit (Jimerson et al 1980). The author recommended that two to three antipsychotic trials with doses equivalent to chlorpromazine 400-600 mg/day be undertaken before considering a patient refractory to treatment. Similarly, there is no evidence to suggest that multiple antipsychotics administered simultaneously are better than one if equivalent doses are used (Davis et al 1977).

ALTERNATIVE TREATMENTS

Christison et al (1991) and Meltzer (1992) have extensively reviewed alternative pharmacological treatments for schizophrenia. In this section we will specifically review the data on alternative pharmacological treatments that have been used in the treatment of patients with schizophrenia who are resistant to conventional antipsychotics. This will include a review of medications that have been used as adjunctive treatments as well as new atypical antipsychotics. Unless stated otherwise, we will only discuss drugs that have been studied in systematic, controlled trials and will only review findings that relate to treatment-resistant schizophrenia.

BENZODIAZEPINES

Benzodiazepines are known to facilitate gamma-aminobutyric acid (GABA) transmission in the CNS. There are theoretical reasons to think that benzodiazepines may be useful in treating schizophrenia, as GABA is an inhibitory neurotransmitter that modulates the CNS release of serotonin, norepinephrine and dopamine in the nigrostriatal and possibly the mesolimbic brain regions via a negative feedback loop . It has been postulated that schizophrenic patients may have decreased GABA activity and the administration of benzodiazepines may reverse this biochemical deficit.

Based on this hypothesis, investigators have studied the use of benzodiazepines both in the treatment of acute exacerbations of psychosis and in the treatment of chronic schizophrenia. In reviewing several double-blind studies examining the effect of benzodiazepines in chronic schizophrenia, Christison et al found that approximately half reported significant improvement compared to placebo while half did not, though due to a number of methodological concerns, interpretation of these studies is inconclusive (Christison et al 1991). Nevertheless, there appears to be a small subgroup of patients with chronic schizophrenia who (receive) benefit from adding a benzodiazepine to conventional antipsychotics.

There are five studies specifically investigating the efficacy of benzodiazepines in patients resistant to treatment with conventional antipsychotics (Henderson et al 1995; Lieberman et al 1994; Lindstrom et al 1980; Potkin et al 1994; Tandon et al 1990). In these studies either a benzodiazepine or placebo was added to a stable dose of antipsychotic, and was subsequently crossed-over in a double-blind fashion for comparison. Three of these five studies report significant improvement associated with the benzodiazepine compared to placebo (Christison et al 1991; Lieberman et al 1994; Potkin et al 1994; Tandon et al 1990). Wolkowitz et al added alprazolam 3.5-4.0 mg/d in a double-blind design to fluphenazine 30 and 40 mg/d, respectively (Tandon et al 1990). The placebo-controlled OFF-ON-OFF design reported both patient groups responded with apparent improvement in positive and negative symptoms of schizophrenia. Conversely, both relapsed when the alprazolam was discontinued.

The optimal effective dosage of benzodiazepines for use as adjunctive therapy with antipsychotics in treatment-refractory schizophrenia remains unclear. In three studies reporting significant benefit, daily doses ranged from as low as 15 mg of diazepam (Lindstrom et al 1980) to 4 mg of alprazolam (i.e. the equivalent of 80 mg of diazepam) (Tandon et al 1990). In another study, diazepam 80 mg/day was no different from placebo (Potkin et al 1994). As there is no evidence that higher doses are superior to lower doses, and there is a greater risk of adverse effects, tolerance, and withdrawal with high doses, a trial of benzodiazepines as an adjunctive agent should maintain the lowest beneficial dose. Because of the dearth of information regarding the use of benzodiazepines in treatment-refractory schizophrenia, it is difficult to ascertain whether particular benzodiazepines may be more effective than others. The positive studies utilized diazepam, alprazolam, and estazolam whereas the negative studies used chlordiazepoxide and diazepam. In general the use of longer acting benzodiazepines to reduce the risk of withdrawal following abrupt discontinuation is recommended.

While there has been evidence to suggest benzodiazepines are helpful in refractory schizophrenia, adverse effects have also been noted. Investigators have reported social disinhibition, aggression, sedation and ataxia (Kane et al 1993; Lindstrom et al 1980). Due to the risk of abuse and dependence with benzodiazepines, particular care must be taken in patients with a history of drug abuse, although anecdotally this does not seem to be a significant problem in treatment-refractory persons with schizophrenia. In addition, patients may experience withdrawal symptoms including seizures, if benzodiazepines are stopped abruptly after chronic use. There is no evidence to suggest that treatment- refractory patients develop tolerance to the antipsychotic effect of the benzodiazepines.

Benzodiazepines appear to be potentially useful adjuncts to conventional antipsychotics in patients with treatment-resistant schizophrenia. It is difficult to determine if a particular agent is likely to be more beneficial than another, or which patients may receive benefit from adjunctive benzodiazepine use. There is some evidence that those patients with higher baseline anxiety and psychosis may show the greatest improvement (Keck et al 1989; Tandon et al 1990).

Although controlled trials have not been conducted, Henderson and Goff have reported significant improvement with buspirone 15-40 mg/day added to conventional antipsychotics (Goff et al 1991). In seven patients receiving oral haloperidol, they reported a 26% increase in mean haloperidol plasma concentrations while receiving buspirone. Thus the clinical improvement may have been due primarily to the increased availability of antipsychotic rather than any intrinsic benefit from buspirone.

CARBAMAZEPINE

Several small open studies have suggested that carbamazepine may be useful as an adjunct to antipsychotics in treatment-resistant schizophrenia (Ballenger et al 1984; Hakola et al 1982; Lingjaerde 1982). Carbamazepine-responsive signs and/or symptoms have included excitement, hostility, aggressive behavior, and not surprisingly, temporal lobe or nonspecific electroencephagraphic abnormalities. Two double-blind, placebo-controlled trials involved treatment-resistant schizophrenic patients who were selected for the studies because of manic-like symptoms, excitement and hyperactivity (Kirch et al 1988; Meltzer 1990). Both studies found that carbamazepine produced a statistically significant, but clinically insignificant improvement in excitement, manic symptoms, tension, and unusual thought content compared to placebo. Okuma et al reported that subjects with manifest symptoms of violence, aggression, or paranoia were more likely to show improvement (Meltzer 1990).

However, studies which did not select patients with manic-like symptoms have not had positive results. There have been two double-blind, placebo-controlled studies in which carbamazepine was added to a stable dose of antipsychotic in individuals with treatment-resistant schizophrenia (Dose et al 1987; Kellner et al 1975). In both studies, the authors reported no significant difference between carbamazepine and placebo.

In each of these studies, the dosage typically used was that required to produce therapeutic anticonvulsant plasma concentrations. It is important to note that carbamazepine may lower antipsychotic serum levels by > or = 50% through hepatic microsomal enzyme induction (Kellner et al 1975). Carbamazepine should not be added to clozapine because of its potential to suppress bone marrow function and increase the risk of clozapine-induced agranulocytosis.

Overall, carbamazepine appears to be a potentially useful adjunctive agent in the treatment-resistant schizophrenic with a history of EEG abnormalities, manic-like symptoms, excitement, and hyperactivity.

VALPROATE

Several anecdotal reports and uncontrolled studies, have suggested that valproate may be useful in treating resistant schizophrenic patients when added to conventional antipsychotics (Gundurewa et al 1980; Ko et al 1989; Kornig et al 1995). No controlled trials have been conducted to date and more study is needed before it can be recommended.

LITHIUM

The addition of lithium to conventional antipsychotics has resulted in improvement in some treatment-resistant patients. There have been three double-blind, controlled crossover trials examining the effect of combining lithium and antipsychotic medications in treatment-refractory schizophrenia and schizoaffective disorder (Carmen et al 1981; Growe et al 1979; Silver et al 1991). In each of the three studies patients received lithium + antipsychotic for four weeks, placebo + antipsychotic for four weeks, and then repeated the sequence. All three found lithium to be superior to placebo for hallucinations, delusions, negative symptoms, irritability and excitement. The results suggest that 33 -50 % of treatment-resistant patients will show some improvement within four weeks when lithium is added to their antipsychotic.

It has been suggested that treatment-resistant patients with an affective component to their clinical presentation may be most likely to respond to the addition of lithium. However, in the treatment-resistant patients in the studies noted above, improvement was also demonstrated in patients without affective symptoms. Doses necessary to achieve lithium plasma concentrations of 0.8-1.2 mEq/L were administered.

Notably, there have been reports of neurotoxicity when lithium has been combined with antipsychotics, so patients must be monitored closely for changes in cognition and neurological function (Braden et al 1982; Cohen et al 1974; Silver et al 1991). However, these reports consisted primarily of cases involving patients on high dose antipsychotics who may have been vulnerable to drug-induced delirium for a variety of reasons. This combination is widely used and with prudent clinical monitoring can be done safely. The combination appears to be potentially useful in some persons with treatment-resistant schizophrenia.

PROPRANOLOL

Numerous anecdotal reports have suggested that propranolol, a beta-adrenergic blocking agent, if given in doses as high as 4 g/d, was effective in treating refractory schizophrenia. However, these reports have been criticized because diagnostic criteria were often lacking. In addition, the onset of response varied from days to months (Donaldson et al 1986).

There have been a total of six double-blind, controlled trials examining the efficacy of propranolol in treatment-resistant schizophrenia. Two studies (Levinson et al 1992; Wolkowitz 1993) showed that propranolol significantly reduced psychotic symptoms, two found a modest benefit (Bigelow et al 1978; Peuskens 1995), and the other two (Kidron et al 1985; Meltzer et al 1986) found no difference between propranolol and placebo. Three of the studies used a crossover design (one positive study, one with modest results and one negative study) (Bigelow et al 1978; Kidron et al 1985; Levinson et al 1992), and three (Meltzer et al 1986; Peuskens 1995; Wolkowitz 1993) were head-to-head comparisons of propranolol and placebo. It is not possible to explain the difference in findings based on propranolol doses, duration of treatment, or other methodologic factors.

Several theories that attempt to explain why propranolol may produce some benefit in schizophrenic patients have been proposed. Peet et al reported that propranolol raises chlorpromazine plasma concentrations, suggesting that the clinical improvement may be related to the higher concentrations (Myers et al 1981). However, in several other studies patients who had not responded to higher doses of antipsychotics responded to propranolol (Donaldson et al 1986). Donaldson et al suggested that any benefit seen when propranolol is added to an antipsychotic may be related to its proven effectiveness for akathisia (Donaldson et al 1986). Another proposed mechanism for the improvement seen with propranolol may be related to temporal lobe function. It has been proposed that propranolol in high doses acts pharmacologically as a anticonvulsant, not an antipsychotic, and suppresses temporal lobe abnormalities. This may explain the beneficial effects of propranolol in decreasing aggression in patients with head injuries. Thus it has been hypothesized that in patients with schizophrenia what may actually be responsive to propranolol is the coarse/diffuse brain dysfunction as opposed to the psychosis per se.

In sum, the data on propranolol do not demonstrate any clear advantage of its use in treatment-resistant schizophrenic patients, and the potential for cardiovascular and/or pulmonary side effects (i.e. decreased inotropic effects, decreased blood pressure and heart rate, interference with antiasthmatics, etc.) (Donaldson et al 1986), severely limits its utility.

RESERPINE

Reserpine alone has been shown to exert antipsychotic action, but is less effective than typical antipsychotics in the treatment of schizophrenia (Sramek et al 1988). Various open trials have suggested that reserpine added to antipsychotic medication may be useful in the treatment-resistant patient. In l955 Kinross-Wright reported that 11/14 "unresponsive" patients with schizophrenia responded to a combination of reserpine and chlorpromazine after having failed to respond to either agent alone (Kinon et al 1993). In another uncontrolled report, Backer and Lewis in l978 reported a 50% response rate in "refractory schizophrenics" when reserpine was added to their antipsychotic regimen (Bacher et al 1978). Berlant, in a l986 retrospective chart review of DSM-III diagnosed schizophrenia, reported that 18/36 patients responded when reserpine 0.5-5.5 mg/d was added to the existing antipsychotic regimen (Berlant 1986). It was noted that most of these patients had failed treatment trials with lithium, carbamazepine, and/or tricyclic antidepressants. Symptoms reported to specifically improve in these open studies include psychosis, hyperactivity, agitation, and assaultiveness.

In the one double-blind, controlled trial examining the efficacy of reserpine in carefully selected individuals with treatment-resistant schizophrenia, Perenyi et al reported that reserpine had no added benefit compared to antipsychotics alone (Okuma et al 1989).

Reserpine has been associated with severe side effects including depression, hypotension, peptic ulcers, exacerbation of asthma, and EPSE (particularly when combined with antipsychotics). Due to the severity of these adverse reactions, and the equivocal nature of its antipsychotic effects, reserpine is no longer recommended for the treatment of schizophrenia whether adjunctive or in monotherapy.

CALCIUM CHANNEL BLOCKERS

As the diphenylbutylpiperidine class of antipsychotics (i.e. pimozide) share structural similarities with calcium channel blockers, investigators have sought to determine if calcium channel blockers may play a role in the treatment of schizophrenia. Kramer and coworkers studied five DSM III-diagnosed persons with schizophrenia who had failed to respond to three different antipsychotics (Ko et al 1985). All had been ill for at least ten years and had received high dose antipsychotic treatment. After discontinuing the antipsychotic, placebo was prescribed for two weeks. Nifepidine 60 mg/d for four weeks was followed by two weeks of placebo in a double-blind design. Weekly BPRS and CGI did not reflect a significant change in positive or negative symptoms. One subject experienced marked clinical worsening as the study progressed which was reversed when antipsychotics were reinstituted. Nifedipine treatment was associated with mild tremor and lightheadedness. It is noted that nifedipine 60 mg/d is a low dose.

Other calcium channel blockers have also been investigated. Pickar et al administered verapamil in doses of 480 mg/day to seven "chronically ill" schizophrenic patients for five weeks in double-blind, placebo-controlled conditions (Perenyi et al 1988). Verapamil had no beneficial effects for either positive psychotic symptoms or emotional blunting. At present, there is no support for the use of calcium channel blockers in the treatment of refractory schizophrenia, although no studies have examined the possible benefit of adding a calcium channel blocker to antipsychotics.

OTHER PHARMACOLOGICAL APPROACHES

Numerous other pharmacological agents have been used alone and in combination with antipsychotic in treatment-resistant schizophrenic patients (Table 1). Unfortunately the majority of these agents have been found to have modest if any therapeutic effects and very few of the positive studies have been replicated.

TREATMENT OF REFRACTORY PATIENTS WITH PREDOMINANTLY NEGATIVE SYMPTOMS

The term "treatment-refractory schizophrenia" has traditionally referred to patients whose positive psychotic symptoms do not respond adequately to antipsychotic medications. However, it is now recognized that patients who experience prominent negative symptoms such as decreased emotional expressiveness, volition, social drive, and thought/speech are also severely impaired. Various investigators have shown improvement in negative symptoms during conventional antipsychotic trials in acutely psychotic individuals (Breier et al 1987; Coryell et al 1990; Goldberg 1985; Meltzer 1992; Simpson et al 1967). However, since improvements in negative symptoms can co-occur with amelioration of positive psychotic symptoms, it is believed that these studies are probably measuring improvement in "secondary" negative symptoms (Bustillo et al 1995). "Secondary" negative symptoms are those which occur in response to another facet of the illness, for example avolition due to overwhelming positive psychotic symptoms, drug side effects (i.e. EPSE), and/or depression. Existing literature fails to adequately distinguish improvement of primary negative symptoms from secondary negative symptoms. Nevertheless, there is evidence to suggest that several alternative agents may be useful in treating patients who continue to be disabled by negative symptoms in spite of adequate remission of positive symptoms. We provide a brief overview of agents that appear to have some benefit.

L-DOPA

As a dopamine agonist, L-dopa is reported to worsen positive psychotic symptoms, agitation and hostility in some patients with schizophrenia (Angrist et al 1973; Calil et al 1977; Volavka et al 1992). However, in schizophrenics selected for a predominance of negative symptoms such as: affective flattening, amotivation, anhedonia, asociality and alogia and minimal positive psychotic symptoms, several double-blind, placebo controlled studies have found L-dopa to be superior to placebo (Brambilla et al 1979; Gerlach et al 1975; Hommer et al 1984). In the largest of these studies, Inanaga et al, studied 104 inpatients with a diagnosis of schizophrenia who were experiencing negative symptoms but no positive psychotic symptoms (Davis et al 1980). Patients were randomly assigned to have either L-dopa 300-600 mg/day or placebo added to neuroleptic for eight weeks. They found a significantly greater improvement in motivation and sociality with L-dopa than with placebo. The improvement with L-dopa was seen most often in patients with a duration of illness of less than five years.

Brambilla et al compared L-dopa combined with carpidopa to placebo in a double-blind cross-over study of patients with schizophrenia with predominantly negative symptoms who were not receiving antipsychotics (Brambilla et al 1979). Two out of the six patients studied were reported to experience significant improvement in negative symptoms on L-dopa/carpidopa.

It appears that L-dopa may have a role in patients with schizophrenia who have prominent negative symptoms with minimal positive symptoms. Christison et al have suggested combining L-dopa with a neuroleptic in these patients and potentially gradually withdrawing the neuroleptic in patients who show improvement (Christison et al 1991).

Other agents that have been studied in the treatment of patients with schizophrenia who have persistent negative symptoms are listed in Table 2. There is data suggesting that amphetamines (Cesarec et al 1985), the tricyclic antidepressants (Schooler et al 1995), cyproheptadine (Pugh et al 1983), fluoxetine (Goff et al 1990), and fluvoxamine (Rifkin et al 1991), produce statistically significant improvement in negative symptoms when added to conventional antipsychotics. However, it is unclear whether these improvements noted in the research setting will translate into clinically significant when utilized in patient care settings. Additionally, in the studies completed to date, there is little or no distinction between improvement in primary and secondary negative symptoms.

RISPERIDONE

Risperidone is a newer antipsychotic that combines the antagonism of both dopamine D2 and serotonin 5HT2 receptors. In the North America Multicenter trial involving 523 patients with schizophrenia, risperidone (6 mg/day) was found to be more effective for total psychopathology, positive psychotic symptoms and negative symptoms than either placebo or haloperidol (20 mg/day) (Luchins 1983). In addition, risperidone was associated with fewer extrapyramidal side effects than haloperidol. In another large, multinational, parallel-group, double blind study, risperidone in doses of 1, 4, 8, 12, or 16 mg daily was compared to haloperidol 10 mg daily in 1362 patients with chronic schizophrenia (Palao et al 1994). Risperidone in doses of 4 to 16 mg/day showed superior efficacy compared to risperidone 1 mg/day and at doses of 4 and 8 mg had a lower incidence of side effects than haloperidol.

Although risperidone has not been studied in rigorously defined treatment-refractory schizophrenic patients, there was some suggestions that it may have a therapeutic role in this population. In a 12-week, double-blind comparison study of 44 patients in whom previous treatments had been unsuccessful "despite optimization of conventional neuroleptic treatment" risperidone (mean dose 12 mg/day) proved to be superior to haloperidol (mean dose 10 mg/day) for both positive and negative symptoms, with fewer extrapyramidal side effects (Claus et al 1992). The fact that the haloperidol treated patients had no significant improvement supports the likelihood that these patients were relatively treatment-resistant.

On the assumption that prolonged hospitalization denotes treatment resistance, a separate, post-hoc analysis of the North America Multicenter trial was performed on 48 of the above patients who had been hospitalized for longer than six months at the time of entry into the study (Luchins 1983). Those patients treated with risperidone had significantly greater improvement than those treated with haloperidol (p<0.01) or placebo (p< 0.01). The overall improvement in these patients treated with risperidone tended to be greater than the improvement seen in the total group.

Using a more clearly defined assessment of "treatment-refractory," Kronig et al examined the efficacy of risperidone in a small open study (Kramer et al 1987). Fifteen patients who had been refractory to two trials of antipsychotic therapy (at dose equivalent of > 1000 mg/day chlorpromazine for 4 weeks). In addition, patients could be included if they were intolerant of two or more antipsychotics despite efforts to alleviate side effects were studied. All patients had a BPRS total score >43 with a score of moderate or worse on one or more BPRS items of unusual thought content, conceptual disorganization, hallucinations, or suspiciousness. In this group, eight out of the 15 patients responded to risperidone (response defined as > 20% reduction in BPRS).

To date, there have been no studies directly comparing risperidone to clozapine in carefully selected treatment-resistant patients. Therefore, in terms of clinical recommendations, risperidone appears to be an intermediate step before the use of clozapine in treatment-refractory schizophrenia, particularly in light of the fact that risperidone does not cause agranulocytosis. Additionally, the overall cost of risperidone is less than clozapine, and the patient is spared the frequent blood tests, making risperidone more "user-friendly".

It is recommended that risperidone be started at doses of 1.0 mg twice daily and gradually titrated up to a dose of 6.0 mg/day over a period of three to seven days (Pickar et al 1987). In the patient with treatment-resistant schizophrenia who does not respond to a dose of 4-6 mg/day for three to four weeks and who is not experiencing troublesome side effects, the clinician may want to slowly increase the dose, with a maximum of 16 mg/day. In switching from another antipsychotic, we recommend a gradual transition from the previous antipsychotic to risperidone. Exactly how this is accomplished depends on various patient characteristics, clinician preferences, etc.

CLOZAPINE

Currently, the pharmacologic treatment with the most evidence of efficacy in treatment-refractory schizophrenia, is the atypical antipsychotic, clozapine. When clozapine was compared in a double-blind fashion to chlorpromazine in rigorously defined treatment-resistant schizophrenic patients, 30% of the patients treated with clozapine improved (>20% reduction in BPRS) compared to 3.5% of those treated with chlorpromazine over a six-week trial (Kane 1989). Clozapine was associated with superior efficacy for both positive and negative symptoms. Two subsequent double-blind comparisons of clozapine in treatment-resistant schizophrenic patients have shown superior efficacy when compared to haloperidol (Breier et al 1994; Preskorn et al 1993). Breier et al compared clozapine to haloperidol (25 mg/day) in 39 "partially responsive" outpatients with schizophrenia in a ten-week trial. They reported that 44% of the patients treated with clozapine improved compared to only 6% of those treated with haloperidol (Breier et al 1994). In order to extend the understanding of clozapine's role in longer term treatment of outpatients who experience psychotic symptoms despite treatment with antipsychotics, Schooler et al compared clozapine to haloperidol (10 mg/day) in a 29-week multicenter trial. Preliminary results from this study demonstrated that 54% of the patients treated with clozapine improved compared to 8% of those treated with haloperidol (Preskorn et al 1993) (See Table 3). These data suggest that the clozapine response rate may reach 50-60% with longer trials. Although there is an ongoing debate as to the appropriate length of a clozapine trial, it seems that a six-month trial is justified (Carpender et al 1995; McEvoy et al 1991).

Although several trials have shown that treatment-refractory schizophrenic patients with a diagnosis of paranoid subtype, a later age of onset of illness, higher EPS scores on typical antipsychotics, lower plasma HVA levels on clozapine, and lower HVA/5-HIAA ratios may be more likely to respond to clozapine, it is not possible to predict which patients will respond to clozapine with any degree of certainty (Levinson et al 1990; Peet et al 1981). Therefore, any treatment-resistant patient who has not responded adequately to at least two trials of an antipsychotic at reasonable doses (e.g.. >600-800 mg chlorpromazine equivalents) for at least 6 weeks should be given a trial of clozapine.

It is recommended that clozapine be started at doses of 12.5 mg once or twice daily and gradually titrated up to a dose of 300-600 mg/day over a period of several weeks (Clozaril 1994). As it takes at least several weeks to appreciate clozapine's therapeutic effects, we recommend continuing the previous antipsychotic initially and gradually tapering the dose as the clozapine dose is being increased. The switch over from the previous antipsychotic to clozapine may take several weeks to several months depending upon the dose of the previous antipsychotic, the patients clinical status side effects, etc. Most treatment-refractory patients will respond to doses of 400-500 mg/day.

There is growing evidence that clozapine plasma concentrations may be a useful predictor of response in treatment-refractory schizophrenics. Although early studies did not demonstrate a relationship between clozapine plasma concentrations and therapeutic response, they were not specifically designed to answer this question. Five out of six recent studies have reported that clozapine response rates are significantly higher in patients who have plasma concentrations above a threshold concentration of 350 ng/ml, 370 ng/ml or 420 ng/ml (Ballenger et al 1984; Braden et al 1982; Carmen et al 1981; Csernansky et al 1984; Csernansky et al 1988) (see Table 4).

Three of the studies reported a threshold clozapine plasma concentration of 350 ng/ml (Braden et al 1982; Carmen et al 1981; Csernansky et al 1988); one found a similar optimal cutoff of 370 ng/ml (Ballenger et al 1984); and Potkin et al (1994) reported that a threshold of 420 ng/ml differentiated responders from nonresponders. Further research, particularly with a greater number of subjects with plasma concentrations between 350 and 420 ng/ml will be needed to determine precisely what should be considered the optimal cutoff. Currently a clozapine plasma concentration above 350 ng/ml appears to have the greatest potential benefit. If a patient is unresponsive with a plasma concentration above 350 ng/ml but below 420 ng/ml, increasing the dose further to obtain concentration of at least 420 ng/ml may be of benefit.

Given our current knowledge, clozapine plasma concentrations may provide useful information in the following circumstances: a) when patients fail to respond to a typically adequate dose of clozapine; b) when the clinician must discriminate clozapine side effects from symptoms of schizophrenia; c) when patients are responding to clozapine but are experiencing drug side effects; d) when clozapine is combined with other drugs that may affect its pharmacokinetics, such as phenytoin, carbamazepine, valproate, fluoxetine, and cimetidine; e) when the pharmacokinetics of clozapine may be significantly altered as in the very young, the elderly, and the medically compromised; and f) when noncompliance or poor compliance is suspected.

Clozapine plasma concentrations should be obtained approximately 12 hours after the last dose to avoid falsely elevated concentrations that occur during the period of distribution into various tissues. This is generally achieved by taking the blood sample in the morning prior to the next clozapine dose. The clinician should be certain that the patient has been on a stable dose of clozapine for at least 5-7 days to assure that the plasma concentration has reached steady-state.

In patients who does not respond to doses of 900 mg per day and has a clozapine concentration below 350 ng/ml, the clinician is left with the question of whether the dose should be increased to greater than manufacturers recommended maximal dose (Clozaril 1994). Clozapine doses of greater than 600 mg per day, particularly greater than 900 mg per day, have been associated with a greater risk of seizures (Clozaril 1994). Although this has not been studied in a systematic manner, several cases of clozapine-induced seizures have been associated with extremely high plasma concentrations (Kinon et al 1993). Theoretically, if an individual metabolizes clozapine very rapidly and has a low plasma concentration, the risk of side effects including seizures, should also be low. In this situation the clinician may consider other alternative medications and should discuss the plan with the patient and caregivers before increasing the doses above 900 mg per day. If it is decided to increase the dose further, this should be done very cautiously with close monitoring. Following-up with subsequent clozapine plasma concentrations will provide the clinician with added information regarding further need for dose adjustment.

Based on the above data, it appears that it may be premature to classify a patient as refractory to clozapine without first determining a clozapine plasma concentration and adjusting the dose based on the results. However, it is recognized that some patients will not be able to tolerate doses needed to achieve concentrations above the purported threshold, and other patients with concentrations above the threshold may not show a therapeutic response to clozapine. Clozapine has already shown tremendous promise in symptom relief for the refractory patient with schizophrenia. It appears that additional benefits may be obtained by careful clinical follow-up possibly involving longer trials at targeted serum concentrations.

Little data from systematic trials is available to guide treatment decisions in the 40-60% of treatment-refractory patients with schizophrenia who do not respond to adequate doses of clozapine. Henderson and Goff have completed an open trial in which risperidone (2-6 mg/day) was added to clozapine (250-700 mg/day) in twelve patients who were refractory to clozapine in spite of "optimal doses" (Hasegawa et al 1993). They found that the combination was well-tolerated and produced significant reduction of positive symptoms, negative symptoms, and depressive symptoms. The addition of risperidone did not alter clozapine plasma concentrations significantly. They conclude that the addition of risperidone may be an appropriate next step for patients not responding to optimal doses of clozapine. Others have suggested adding a high potency typical antipsychotic, lithium, a selective serotonin reuptake inhibitor or buspirone to clozapine may be of benefit in such patients (Navroidis et al 1984). However, at the present time there is no published data supporting this recommendation.

CONCLUSION

Treatment of the refractory patient with schizophrenia represents a significant challenge for the clinician, and unfortunately there are no definite guidelines guaranteed to elicit a successful response in this complicated population of patients. Each case presents with a unique symptom profile and individual vulnerabilities to adverse effects. The pitfalls of aggressive treatment often include polypharmacy with multiple medication side effects and numerous adjunctive medications complicating the clinical picture. Thus the clinical approach should be one of determined simplicity with very careful and deliberate changes made in a non-reactive manner. The clinician should be reminded that antipsychotic medications may require weeks to months before full benefits are apparent, making patience a key feature in managing refractory symptoms. Kane and Marder, emphasize the importance of a well thought-out plan in terms of target symptoms, dosages, and duration of treatment in order to accurately assess the impact of a particular trial (Kane et al 1988). This includes the recommendation that only one change in a patient's pharmacological regimen should be made at a time, with adequate amount of time allowed to respond to an intervention. In addition, we recommend that if a particular intervention has little benefit, the medication should be tapered and discontinued over a reasonable period of time before a new intervention is implemented. As Kane and Marder emphasize, it is important that the physician fully explain the rationale as well as the potential benefits and risks of the alternative treatment to the patient and significant others, since many of these treatments are not approved for this indication.

Even under the best circumstances of patient compliance and minimization of side effects, all of the antipsychotic medications today continue to have adverse effects in both short and long term treatment. However, there is growing evidence to suggest that long term consistent treatment of psychotic symptoms may have a beneficial effect on symptom severity and cognitive function in later life. This notion fuels the incentive to continue to search for better tolerated more efficacious agents for the treatment of schizophrenia. A number of new agents including olanzapine, sertindole, seroquel, ziprasidone, etc. are currently undergoing trials for future antipsychotic use, and hold promise for improved outcome in this otherwise devastating illness. To date, none of these experimental agents have been studied in rigorously defined treatment-refractory patients with schizophrenia, so we can only speculate what role they may play in this population.

SUMMARY

1. First and foremost management of the treatment-refractory patient requires the assessment of compliance. Non-compliance is a notoriously common cause of a false appearance of treatment-resistance when in fact the patient may be quite responsive to medication. Plasma concentrations may be quite helpful both in identifying non-compliance and in optimizing levels in the compliant patient. Often upward titration of the dose (with careful attention to side effects and prompt intervention) will assist the patient in achieving an adequate response to typical antipsychotic medications. Often simply switching a dosing schedule or selecting an alternate typical agent with a different side effect profile will be sufficient to maintain improvement.
2. When non-compliance is excluded, and the patient continues to suffer significant symptoms in spite of an adequate trial of haloperidol, for example, (in the therapeutic range of 5-15 ng/ml) a novel agent may be considered. Risperidone is likely the next most reasonable choice in terms of least cost and effort for the patient, though there is comparatively much less data to support its efficacy in treatment-resistant schizophrenia. A dose of 4-8 mg/day of risperidone in divided doses is considered optimal, though higher doses may be necessary in treatment-resistant patients.
3. If the patient continues to remain treatment-resistant after the above antipsychotic trials, clozapine is the next logical choice. We recommend gradually titrating up the dose of clozapine as the dose of the previous antipsychotic is being gradually tapered. Most treatment-refractory patients will respond to doses of 400 - 600 mg/day. If the patient does not achieve an adequate response at these doses, we suggest obtaining a clozapine plasma concentration and adjusting the dose to reach a plasma concentration above 350 ng/ml (maximum dose of 900 mg/day).
4. If there is no response or inadequate response to "therapeutic doses" of clozapine (dose required to obtain a plasma concentration above 350 ng/ml) for 6 months, we would suggest adding risperidone in doses of 2 to 6 mg/day. If there is no improvement with this combination and the patient presents with affective symptoms (i.e. manic-like symptoms, excitement, hyperactivity, etc.), gradually tapering risperidone and adding lithium to clozapine is a reasonable next step, although there are currently no controlled studies demonstrating definite efficacy. If there is still inadequate response, other considerations for adjunctive therapy include a high-potency antipsychotic, valproate, a selective serotonin reuptake inhibitor or buspirone. We suggest gradually tapering and discontinuing an adjunctive agent that does not improve response, before adding another agent to clozapine.
5. If a patients does not respond to "therapeutic doses" of clozapine alone or in combination with the above agents, one must consider the feasibility of continuing clozapine given its cost and side effects. Although, because of clozapine's lower liability for tardive dyskinesia, one could argue that it should be continued. In a clozapine-refractory or clozapine-intolerant patient or in patients for whom clozapine is unavailable because of financial considerations, combining a conventional antipsychotic or risperidone with lithium, carbamazepine, or a benzodiazepine (depending on the presenting symptom profile) are rational alternatives.
6. As new antipsychotic agents are developed, patients with treatment-refractory schizophrenia should most certainly have the opportunity to receive further trials, as a history of non-response to multiple antipsychotics does not preclude a future response to a novel antipsychotic agent. Clearly the clinical benefit observed with clozapine (and possibly risperidone) has demonstrated the rewards of perseverance and willingness to try new agents in treating the symptoms of schizophrenia.
7. In patients with persistent, disabling negative symptoms, trials involving the adjunctive use of L-dopa, amphetamines, tricyclic antidepressants, fluoxetine, and fluvoxamine are probably warranted. The clinician must keep in mind that these medications may interact with antipsychotic medication, and patients need to be monitored closely.

Table 1. Other Pharmacological Treatments Studied In Treatment-Resistant Schizophrenia

• Baclofen (Simpson et al 1976; Ruskin et al 1979)
• Beta-endorphin (Berger et al 1983)
• Ceruletide (Hommer et al 1984; Holden et al 1968)
• Cholecystokinin (Mattes et al 1985; Marder et al 1989)
• Clonidine (Simposn et al 1967; Jimerson et al 1980; Risperdal 1994)
• Gamma-acetylenic acid (Casey et al 1976)
• Methadone (Brizer et al 1985)
• Prazosin (Hommer et al 1984)
• Thyrotropin-releasing hormone (Wilson et al 1973)

Table 2. Pharmacological Treatments Studied in Patients With Schizophrenia With Persistent Negative Symptoms

Alprazolam (Csernansky et al 1984; Csernansky et al 1988; Douyon et al 1989; Tandon et al 1990; Wolkowitz et al 1988)

Amphetamine (Cesarec et al 1985)

Biperiden (Simpson et al 1976; Tandon et al 1990)

Bromocriptine (Levi-Minzi et al 1991; Kronig et al 1995)

Cyproheptadine (Silver et al 1991; Pugh et al 1983)

Fenfluramine (Alphs et al 1989)

Fluoxetine (Goff et al 1990)

Fluvoxamine (Silver et al 1992; Rifkin et al 1991)

L-Dopa (Gerlach et al 1975; Inanaga et al 1972; Brambilla et al 1979)

Maprotiline (Yamagami and Soejima 1989; Van Putten et al 1990)

Tranylcypromine (Bucci 1987)

Tricyclic Antidepressants (Siris et al 1987; Schooler et al 1995)

Vasopressin (Brambilla et al 1988)

Table 3. Response Rates To Clozapine In Double-Blind Controlled Studies of Treatment-Refractory Schizophrenia
STUDY	(N)	COMPARATIVE DRUG (DOSE)	PATIENT STATUS	DURATION OF TRIAL	RESPONSE RATE (%)
Kane et al 1988	267	Chlorpromazine (1200 mg/day)	Inpatient, Refractory	6 weeks	30 cloz 4 cpz
Breier et al 1994	39	Haloperidol (25 mg/day)	Outpatient, Partially responsive	10 weeks	44 cloz 6 hlp
Schooler et al 1995	52	Haloperidol (10 mg/day)	Outpatient, Refractory	29 weeks	54 cloz 8 hlp

cloz = clozapine
cpz = chlorpromazine
hlp = haloperidol

Table 4. Studies Reporting Relationship Between Clozapine Plasma Concentrations and Clinical Response in Schizophrenia
STUDY, YEAR	(N)		DOSING SCHEDULE	MEAN DOSE (mg/day)	MEAN PLASMA CONC.	DURATION OF TRIAL	RESULTS
Perry et al 1991	29	TR	Fixed dose	384.5 ± 42.5	356 ± 255	6 weeks	Plasma concentration above 350 ng/ml were associated with a 64% response rate compared to a 22% response rate below 350 ng/ml
Pickar et al 1992	21	TR & EPS	Variable dose	373.8 ± 110.8 542.9 ± 207.4	439 ± 431	3-7 weeks 7 weeks	No correlation between plasma concentrations and clinical response
Hasegawa et al 1993	59	TR	Variable dose	443.8 ± 270.7	371 ± 257	6 months	Plasma concentration above 370 ng/ml were associated with a 73% response rate compared to a 38% response rate below 370 ng/ml
Potkin et al 1994	58	TR	Variable dose	400 400 or 800	419 525	4 weeks 8 weeks	Plasma concentration above 420 ng/ml were associated with a 73% response rate compared to a 29% response rate below 420 ng/ml
Miller et al 1994	24	TR	Variable dose	568.7 ± 151.8	394 ± 229	2.5 years	Plasma concentration above 350 ng/ml were associated with a 86% response rate compared to a 40% response rate below 350 ng/ml
Koning et al 1995	45	TR	Variable dose	623.2 ± 203.5	370 ± 336	6 weeks	Plasma concentration above 350 ng/ml were associated with a 55% response rate compared to a 20% response rate below 350 ng/ml

a = Treatment status of patients not reported
TR = Treatment-refractory schizophrenic patients
EPS = Schizophrenic patients intolerant to typical antipsychotics due to extrapyramidal side effects

REFERENCES

Alphs L, Lafferman J, Ross L, et al: Fenfluramine treatment of negative symptoms in older schizophrenic inpatients. Psychopharmacol Bull 25:149-153, 1989

Angrist B, Sathananthan G, Gershon S: Behavioral effects of L-dopa in schizophrenic patients. Psychopharmacalogia 31:1-12, 1973

Aubree J, Lader M: High and very high dosage anti-psychotics: A critical review. J Clin Psychiatry 41:341-350, 1980

Bacher NM, Lewis HA: Addition of reserpine to antipsychotic medication in refractory chronic schizophrenic outpatients. Am J Psychiatry 135:488-489, 1978

Baldessarini RJ, Cohen BM, Teicher MH: Significance of neuroleptic dose and plasma level in the pharmacological treatment of psychoses. Arch Gen Psychiatry 45:79-91, 1988

Ballenger JC, Post RM: Carbamazepine and neuroleptics in the treatment on schizophrenia. Psychopharmacology Bull 20:572-584, 1984

Berger P, Barchas J: Pharmacology studies of beta-endorphi in psychopharmacology. Psychiatric Clinics of North America 6:377-391, 1983

Berlant JL: Neuroleptics and reserpine in refractory psychosis. J Clin Psychopharmacol 6:180-184, 1986

Bigelow LB, Zalcman JE, Klienman JE: Propranolol treatment of chronic schizophrenia: Clinical response, catecholamine metabolism and lymphocyte beta-receptors. Fourth International Catecholamine Symposium 4:1851-1853, 1978

Braden W, Fink EB, Qualls B, et al: Lithium and chlorpromazine in psychotic inpatients. Psychiatry Research 7:69-81, 1982

Brambilla F, Boniolotti G, Maggioni M, et al: Vasopressin (DDAVP) therapy in chronic schizophrenia: Effects on negative symptoms and memory. neuropsychobiology 20:113-119, 1988

Brambilla F, Scarone S, Ponzano M, et al: Catacholaminergic drugs in chronic schizophrenia. Neuropsychobiology 5:185-200, 1979

Breier A, Buchanan R, Kirkpatrick B, et al: Effects of clozapine on positive and negative symptoms in outpatients with schizophrenia. Am J Psychiatry 151:20-26, 1994

Breier A, Wolkowitz O, Doran A, et al: Neuroleptic responsivity of negative and positive symptoms in schizophrenia. Am J Psychiatry 144:1549-1555, 1987

Brenner HD, Dencker SJ, Goldstein MJ, et al: Defining treatment refractoriness in schizophrenia. Schizophrenia Bull 16:551-562, 1990

Brizer DA, Hartman N, Sweeney J, et al: Effects of methadone plus neuroleptics on treatment-resistant chronic paranoid schizophrenia. Am J Psychiatry 142:1106-1107, 1985

Bucci L: The negative symptoms of schizophrenia and the monoamine oxidase inhibitors. Psychopharmacology 91:104-108, 1987

Bustillo J, Kirkpatrick B, Buchanan R: Neuroleptic treatment and negative symptoms in deficit and nondeficit schizophrenia. Biol Psychiatry 38:64-67, 1995

Calil HM, Yesavage JA, Hollister LF: Low dose levodopa in schizophrenia. Communication in Psychopharmacology 1:593-596, 1977

Carmen JS, Bigelow LB, Wyatt RJ: Lithium combined with neuroleptics in chronic schizophrenia and schizoaffective patients. J Clin Psychiatry 42:124-128, 1981

Carpender W, Conley R, Buchanan R, et al: Patient response and resource management: Another view of clozapine treatment of schizophrenia. Am J Psychiatry 152:827-832, 1995

Cesarec Z, Nyman AK: Differential response to amphetamine in schizophrenia. Acta Psychiatrica Scand 71:523-528, 1985

Christison GW, Kirch DG, Wyatt RJ: When Symptoms persist: Choosing among alternative somatic treatment in schizophrenia. Schizophrenia Bull 17:217-246, 1991

Claus A, Bollen J, De Cuyper H, et al: Risperidone versus haloperidol in the treatment of chronic schizophrenic inpatients: a multicenter double-blind comparative study. Acta Psychiatrica Scand 85:295-305, 1992

Clozaril: Package Insert. Sandoz Pharmaceutical Corporation, East Hanover, NJ, 1994

Cohen W, Cohen N: Lithium carbonate, haloperidol and irreversible brain damage. JAMA 230:1283-1287, 1974

Cole J, Goldberg S ,Klerman G: Phenothiazine treatment in acute schizophrenia. Arch Gen Psychiatry 10:246-261, 1964

Coryell W, Kelly M, Perry P, et al: Haloperidol plasma levels and acute clinical change in schizophrenia. J Clin Psychopharmacol 10: 397-402, 1990

Csernansky JG, L. L, Gulevich G, et al: Treatment of negative schizophrenic symptoms with alprazolam: A preliminary open-label study. J Clin Psychopharmacol 4:349-352, 1984

Csernansky JG, Riney SJ, et al: Double-blind comparison of alprazolam, diazepam, and placebo for the treatment of negative schizophrenic symptoms. Arch Gen Psychiatry 52:655-659, 1988

Dahl SG: Plasma level monitoring of antipsychotic drugs: Clinical utility. Clin Pharmacokinetics 11:36-61, 1986

Davis J, Casper R: Antipsychotic drugs: clinical pharmacology and therapeutic use. pharmacology and therapeutic use. Drugs 14:260-282, 1977

Davis J, Schaffer C, Killian G, et al: Important issues in the drug treatment of schizophrenia. Schizophrenia Bull 6:70-87, 1980

Davis JM: Dose-equivalence of the antipsychotic drugs. J Psychiatric Research 11:65-69, 1974

Delva NJ, Letemendia FJJ, Prowse AW: Lithium withdrawal trial in chronic schizophrenia. Br J Psychiatry 141:401-406, 1982

Donaldson S, Gelenberg A, Baldessarini R: Alternative treatments for schizophrenic psychoses. In Arieti, S. (ed): American Handbook of Psychiatry. New York, Basic Books, 1986

Donlon PT, Hopkin J, Tupin JP: Overview: Efficacy and safety of the rapid neuroleptization method with injectable haloperidol. Psychiatry Research 22:303-310, 1979

Dose M, Apelt D, Emrich HM: Carbamazepine as an adjunct of antipsychotic therapy. Psychiatry Research 22:303-310, 1987

Douyon R, Angrist B, Peselow E, et al: Neuroleptic augmentation with alprazolam: Clinical effects and pharmacokinetic correlates. Am J Psychiatry 146:231-234, 1989

Gerlach J, Luhdorf K: The effect of L-dopa on young patients with simple schizophrenia treated with neuroleptic drugs. Psychopharmacologia 44:105-110, 1975

Goff D, Midha K, Brotman A, et al: An open trial of buspirone added to neuroleptics in schizophrenic patients. J Clin Psychopharmacol 11:193-197, 1991

Goff DC, Brotman AW, Waites M, et al: Trial of fluoxetine added to neuroleptics for treatment-resistant schizophrenic patients. Am J Psychiatry 147:492-494, 1990

Goldberg S: Negative and deficit symptoms in schizophrenia do respond to neuroleptics. Schizophrenia Bull 11:453-456, 1985

Goldberg S, Klerman G, Cole J: Changes in schizophrenic psychopathology and ward behavior as a function of phenothiazine treatment. Br J Psychiatry 111:120-123, 1965

Growe G, Crayton J, Klass D, et al: Lithium in chronic schizophrenia. Am J Psychiatry 136:454-455, 1979

Gundurewa V, Beckman H, Zimmer R, et al: Effect of valproic acid on schizophrenic syndromes. Arzneimittle-Forschung 30:1212-1213, 1980

Hakola HPA, Laulumma VA: Carbamazepine in treatment of violent schizophrenics. Lancet II:1358, 1982

Hasegawa M, Gutierrez-Esteinou R, Way L, et al: Relationship between clinical efficacy and clozapine concentrations in plasma in schizophrenia: Effect of smoking. J Clin Psychopharmacol 13:383-390, 1993

Henderson D, Goff D(: Risperidone adjunct to clozapine therapy in chronic schizophrenics. New Clinical Drug Evaluation Unit (NCDEU) Program 35th Annual Meeting Orlando, Florida:60, 1995

Holden JM, Itil TM, Keskiner A, et al: Thioridazine and chlordiazepoxide, alone and combined, in the treatment of chronic schizophrenia. Comprehensive Psychiatry 9:633-643, 1968

Hommer DW, Pickar D, Roy A, et al: The efficacy of ceruletide in schizophrenia. Arch Gen Psychiatry 41:617-619, 1984

Hommer DW, Zahn TP, Pickar D, et al: Prazosin, a specific alpha1-noradrenergic receptor antagonist, has no effect on symptoms but increases autonomic arousal in schizophrenic patients. Psychiatry Research 11:193-204, 1984

Inanaga K, Inoue K, Tachibana H, et al: Effect of L-dopa in schizophrenia. Psychiatrica et Neurologica Japonica 26:145-157, 1972

Jimerson D, Post R, Stoddard F, et al: Preliminary trial of the noradrenergic agonist clonidine in psychiatric patients. Biol Psychiatry 15:45-58, 1980

Kane J: Treatment of schizophrenia. Schizophrenia Bull 13:133-156, 1987

Kane J: The current status of neuroleptic therapy. J Clin Psychiatry 50:322-328, 1989

Kane J, Honigfeld G, Singer J, et al: Clozapine for the treatment-resistant schizophrenic: A double-blind comparison with chlorpromazine. Arch Gen Psychiatry 45:789-796, 1988

Kane J, Marder S: Psychopharmacologic Treatment of Schizophrenia. Schizophrenia Bull 19(2):287-302, 1993

Karson CN, Weinberger DR, Bigelow L, et al: Clonazepam treatment of chronic schizophrenia: Negative results in a double-blind, placebo-controlled trial. Am J Psychiatry 139:1627-1628, 1982

Keck P, Cohen B, Baldessarini R, et al: Time course of antipsychotic effects of neuroleptic drugs. Am J Psychiatry 146:1289-1292, 1989

Kellner R, Wilson RM, M.D. M, et al: Anxiety in schizophrenia: The responses to chlordiazepoxide in an intensive design study. Arch Gen Psychiatry 32:1246-1254, 1975

Kidron R, Averbuch I, Klein E, et al: Carbamazepine-induced reduction of blood levels of haloperidol in chronic schizophrenia. Biol Psychiatry 20:199-228, 1985

King D, Turkson S, Liddle J, et al: Some clinical and metabolic aspects of propranolol in chronic schizophrenia. Br J Psychiatry 137:458-468, 1980

Kinon BJ, Kane JM, Johns C, et al: Treatment of neuroleptic resistant schizophrenic relapse. Psychopharmacol Bull 29:309-314, 1993

Kinross-Wright V: Chlorpromazine and reserpine in the treatment of psychoses. Annals of the New York Academy of Sciences 61:174-182, 1955

Kirch D, Bigelow L, Korpi E, et al: Serum haloperidol concentration and clinical response in schizophrenia. Schizophrenia Bull 14:283-289, 1988

Klien E, Bental E, Lerer B, et al: Carbamazepine and haloperidol vs. placebo and haloperidol in excited psychotics: A controlled study. Arch Gen Psychiatry 41:165-170, 1984

Ko G, Korpi E, Kirch D: Haloperidol and reduced haloperidol concentrations in plasma and red blood cells from chronic schizophrenic patients. J Clin Psychopharmacol 9:186-190, 1989

Ko GN, Korpi ER, Freed WJ, et al: Effect of valproic acid on behavior and plasma amino acid concentrations in chronic schizophrenic patients. Biol Psychiatry 20:199-228, 1985

Kramer M, DeMaria P, Morrow R, et al: Nifedipine in neuroleptic-refractory schizophrenia. J Clin Psychopharmacol 7:195-196, 1987

Kronig M, Loebel A, Mendelowitz A, et al: Risperidone in treatment-resistant schizophrenic and schizoaffective patients. American Psychiatric Association Annual Meeting: New Research And Abstracts Miami, FL, 1995

Kronig M, Munne R, Szymanski S, et al: Plasma clozapine levels and clinical response for treatment-refractory schizophrenic patients. Am J Psychiatry 152:179-182, 1995

Lautin A, Angrist B, Gershon S, et al: Sodium valproate in schizophrenia: Some biochemical correlates. Br J Psychiatry 137:240-244, 1980

Levi-Minzi S, Bermanzohn P, Siris S: Bromocriptine for "negative" schizophrenia. Comprehensive Psychiatry 32:210-216, 1991

Levinson D, Simpson G, Singh H, et al: Fluphenazine dose, clinical response, and extrapyramidal symptoms during acute treatment. Arch Gen Psychiatry 47:761- 768, 1990

Levinson D, Singh H, Simpson G: Timing of acute clinical response to fluphenazine. Br J Psychiatry 160:365-71, 1992

Lieberman J, Kane J, Safferman A, et al: Predictors of response to clozapine. J Clin Psychiatry 55 (Suppl B):126-128, 1994

Lindstrom LH, Persson E: Propranolol in chronic schizophrenia: A controlled study in neuroleptic-treated patients. Br J Psychiatry 137:126-130, 1980

Lingjaerde O: Effect of the benzodiazepine derivative estazolam in patients with auditory hallucinations: A multicentre double-blind, crossover study. Acta Psychiatrica Scand 65:339-354, 1982

Lingjaerde O, Engstraand E, Ellingson P, et al: Antipsychotic effect of diazepam when given in addition to neuroleptics in chronic psychotic patients: A double-blind clinical trial. Current Therapeutic Research 26:505-514, 1979

Luchins D: Carbamazepine for the violent psychiatric patient. Lancet I:766, 1983

Marder S, Hubbard J, Van Putten T, et al: Pharmacokinetics of long-acting injectable neuroleptic drugs: Clinical implications. Psychopharmacology 98:433-439, 1989

Marder S, Meibach R: Risperidone in the treatment of schizophrenia. Am J Psychiatry 151:825-835, 1994

Mattes JA, Hom W, Rockford JM, et al: Ceruletide for schizophrenia: A double-blind study. Biol Psychiatry 20:533-538, 1985

Mavroidis M, Kanter D, Hirschowitz J: Therapeutic blood levels of fluphenazine: plasma or RBC determinations? Psychopharmacol Bull 20:168-70, 1984

McEvoy J, Hogarty G, Steingard S: Optimal dose of neuroleptic in acute schizophrenia: a controlled study of the neuroleptic threshold and higher haloperidol dose. Arch Gen Psychiatry 48:739-745, 1991

Meltzer H: Treatment of the neuroleptic-nonresponsive schizophrenic patient. Schizophrenia Bull 18:515-542, 1992

Meltzer H: Clozapine: Is another view valid? Am J Psychiatry 152:821-825, 1995

Meltzer H, Sommers A, Luchins D: The effect of neuroleptics and other psychotropic drugs on negative symptoms in schizophrenia. J Clin Psychopharmacol 6:329-338, 1986

Meltzer HY: Commentary: Defining treatment refractoriness in schizophrenia. Schizophrenia Bull 16:563-565, 1990

Miller D, Fleming F, Holman T, et al: Clozapine plasma concentrations as a predictor of clinical response: A follow-up study. J Clin Psychiatry 55(9, suppl B):117-121, 1994

Myers D, Campell P, Cocks N, et al: A trial of propranolol in chronic schizophrenia. Br J Psychiatry 139:118-121, 1981

Okuma T, Yamashita I, Takahashi R, et al: A double-blind study of adjunctive carbamazepine versus placebo on excited states of schizophrenia and schizoaffective disorders. Acta Psychiatrica Scand 80:250-259, 1989

Palao D, Arauxo A, Brunet M, et al: Haloperidol: Therapeutic window in schizophrenia. J Clin Psychopharmacol 14:303-310, 1994

Peet M, Bethell S, Coates A, et al: Propranolol in schizophrenia: I. Comparison of propranolol, chlorpromazine, and placebo. Br J Psychiatry 139:105-111, 1981

Perenyi A, Goswani U, Bela A, et al: Addition of reserpine in ongoing neuroleptic exposure in chronic schizophrenia. Schizophrenia Res 1:307-308, 1988

Perry P, Miller D, Arndt S, et al: Clozapine and norclozapine plasma concentrations and clinical response in treatment-refractory schizophrenics. Am J Psychiatry 148:231-235, 1991

Peuskens J: Risperidone in the treatment of patients with chronic schizophrenia: a multi-national, multi-centre, double-blind, parallel-group study versus haloperidol. Br J Psychiatry 166:712-726, 1995

Pickar D, Owen R, Litman R, et al: Clinical and biologic response to clozapine in patients with schizophrenia; Crossover comparison with fluphenazine. Arch Gen Psychiatry 49:345-353, 1992

Pickar D, Wolkowitz OM, Doran AR, et al: Clinical and biochemical effects of verapamil administration to schizophrenic patients. Arch Gen Psychiatry 44:113-118, 1987

Potkin S, Bera R, Gulasekaram B, et al: Plasma clozapine concentrations predict clinical response in treatment-resistant schizophrenia. J Clin Psychiatry 55(9, suppl B):133-136, 1994

Preskorn S, Burke M, Fast G: Therapeutic drug monitoring. Psychiatric Clin N Am 16:611-645, 1993

Pugh CR, Steinert J, Priest RG: Propranolol in schizophrenia: A double-blind, placebo-controlled trial of propranolol as an adjunct to neuroleptic medication. Br J Psychiatry 143:151-155, 1983

Rifkin A, Doddi S, Karajgi B, et al: Dosage of haloperidol for schizophrenia. Arch Gen Psychiatry 48:166- 170, 1991

Risperdal: Package Insert. Janssen Pharmaceutica, Titusville, NJ, 1994

Ruskin P, Averbukh I, Belmaker RH, et al: Benzodiazepines in chronic schizophrenia. Biol Psychiatry 14:557-558, 1979

Schooler N, Kane J, Marder S, et al: Efficacy of clozapine versus haloperidol in a long term clinical trial: Preliminary findings. Schizophrenia Res 15(1,2):165, 1995

Silver H, Blacker M, Weller M, et al: Treatment of chronic schizophrenia with cyproheptadine. A double-blind controlled study. Biol Psychiatry 30:523-525, 1991

Silver H, Nassar A: Fluvoxamine improves negative symptoms in treated chronic schizophrenia: An add-on double-blind, placebo controlled study. Biol Psychiatry 31:698-704, 1992

Simpson G, Duntz-Bartholini E, Watts T: A preliminary evaluation of the sedative effects of catapres, a new antihypertensive agent, in chronic schizophrenic patients. J Clin Pharmacology 7:221-225, 1967

Simpson GM, Branchey MH, Shrivastava RM: Baclofen in schizophrenia. Lancet I:967-968, 1976

Siris SG, Moorage V, Fagerstrom R, et al: Adjunctive imipramine in the treatment of postpsychotic depression. Arch Gen Psychiatry 42:533-539, 1987

Small JG, Kellams JJ, Milstein V, et al: A placebo-controlled study of lithium combined with neuroleptics in chronic schizophrenic patients. Am J Psychiatry 132:1315-1317, 1975

Sramek J, Potkin S, Hahn R: Neuroleptic plasma concentrations and clinical response: in search of a therapeutic window. Drug Intell Clin Pharm 22:373-380, 1988

Tandon R, Goldman R, Goodson J, et al: Mutability and relationship between positive and negative symptoms during neuroleptic treatment in schizophrenia. Biol Psychiatry 27:1323-1326, 1990

Tandon R, Mann N, Eisner W, et al: Effect of anticholinergic medication on positive and negative symptoms in medication-free schizophrenic patients. Psychiatry Res 31:235-241, 1990

Van Putten T, Marder S, Mintz J: Controlled dose comparison of haloperidol in newly admitted schizophrenic patients. Arch Gen Psychiatry 47:754-758, 1990

Volavka J, Cooper T, Czobor P, et al: Haloperidol blood levels and clinical effects. Arch Gen Psychiatry 49:354-361, 1992

Wolkowitz O: Rational polypharmacy in schizophrenia. Ann Clin Psychiatry 5:79-90, 1993

Wolkowitz OM, Breier A, Doran AR, et al: Alprazolam augmentation of the antipsychotic effects of fluphenazine in schizophrenic patients. Arch Gen Psychiatry 45:664-671, 1988

Wolkowitz OM, Pickar D, Doran AR, et al: Combination alprazolam-neuroleptic treatment of the positive and negative symptoms of schizophrenia. Am J Psychiatry 143:85-87, 1986

Yamagami S, Soejima K: Effect of maprotiline combined with conventional neuroleptics against negative symptoms of chronic schizophrenia. Drugs Exo Clin Res 15:171-176, 1989

Yaryura-Tobias JE, Diamond B, Merlis S: The action of L-dopa on schizophrenic patients. Current Therapeutic Research 12:528-533, 1970

Yorkston N, Zaki S, Malik M, et al: Propranolol in the control of schizophrenic symptoms. Lancet II:575-578, 1977

Title Page