Pii: s0006-3223(99)00059-

The pharmacologic treatment of schizophrenia remains a for a dopamine hypothesis of antipsychotic drug activity critical component in the short- and long-term manage- and, in turn, a dopamine hypothesis of schizophrenia.
ment of this disease. Considerable progress has been At the same time, further developments in establishing made in delineating different domains of this illness, the validity and reliability of psychiatric diagnosis resulted ranging from positive and negative symptoms to cognitive in a narrowing of the diagnosis of schizophrenia and to dysfunction and psychosocial vulnerabilities. Increas- some extent a reevaluation of drug efficacy in more ingly, treatments are being studied in relation to a variety selectively defined patient populations.
of different outcome measures with functional ability andquality of life achieving appropriate emphasis. It took many years after the introduction of antipsychot- The introduction of a new generation of antipsychotic ics to begin to establish the role of these medications in drugs has helped to raise optimism and expectations. long-term treatment, with the intention initially of reduc- Overall, second-generation drugs do provide clear advan- ing rates of relapse and rehospitalization. Even after tages in terms of reducing adverse effects (particularly relevant studies began to be conducted, very few lasted drug-induced Parkinsonism, akathesia, and, hopefully, more than 1 year, and they tended to involve multi-episode tardive dyskinesia). Advantages in alleviating refractory or chronically ill patients. At the same time that more and symptoms, negative symptoms, depression, and suicidal more data emerged supporting the value of maintenance behavior are found in some reports; however, much treatment (Davis 1975), increasing knowledge and con- remains to be done methodologically in establishing the cern was developing regarding the long-term risks associ- relative merits of specific drugs in the multiple domains of ated with antipsychotic drug treatment (Kane and Smith Biol Psychiatry 1999;46:1396 –1408 1999 Not surprisingly, outcome measures in clinical trials of Key Words: Schizophrenia, treatment, psychopharmacol-
antipsychotic drugs initially tended to focus more on positive symptoms than negative symptoms. Positivesymptoms were more likely to be associated with disrup-tive or troublesome behavior, more likely to result in Introduction
hospitalization (the focus for most initial clinical trials ofdrug efficacy), and to some extent were felt to be easier to The modern-age of pharmacologic treatment of schizo- measure.
phrenia began in the mid-1950s with the development With this background, for the first two decades of of chlorpromazine. Subsequent to that development, a widespread clinical use of antipsychotic medication, it was number of different chemical classes of antipsychotic generally accepted that among available agents there were drugs were introduced into clinical practice. Over the next no significant differences in clinical efficacy. In 100 three decades, considerable progress was made in several comparisons of different conventional antipsychotic drugs, areas which influenced subsequent development and uti- only one study reported a significant difference (Klein and lization of antipsychotic drugs as well the over all care of Davis 1969; Janicak et al 1993). This result did not patients suffering from schizophrenia.
necessarily mean that an individual patient was equally The observations that antipsychotic drugs were associ- likely to respond to any drug, but rather in group compar- ated with a variety of neurologic side effects, most isons similar proportions of patients responded regardless commonly drug-induced Parkinsonism, but also tardive of which drug was involved. The notion that if a particular dyskinesia and tardive dystonia, helped contribute support patient did not respond to one drug, then he or she mightrespond to another was widely assumed, however, as wewill discuss below, rarely tested.
From the Department of Psychiatry, Hillside Hospital, Division of Long Island The discovery that enormous variations in bioavailabil- Jewish Medical Center, Glen Oaks, New York and Department of Psychiatryand Neuroscience, Albert Einstein College of Medicine, Bronx, New York.
ity and metabolism occurred with these drugs led to Address reprint requests to John M. Kane, MD, Hillside Hospital, 75-59 263rd renewed hope that outcome could be improved substan- Received October 12, 1998; revised February 17, 1999; accepted March 5, 1999.
tially by measuring blood levels following the develop- ment of steady state and adjusting them up or down deficiency was a high incidence of acute and chronic accordingly (Kane et al 1976). Despite a fairly extensive neurologic effects. Second was frequent poor or only literature suggesting relationships between blood levels partial response of both positive and negative symptoms and clinical response, very few studies confirmed the with considerable residual disability (psychosocial and value of these measures by subsequently manipulating vocational). The third deficiency was high rates of non- blood levels into a putative therapeutic range, under compliance, which were felt to be at least partially due to controlled conditions and demonstrating a clinically and frequent adverse effects and subjective discomfort. Fourth statistically significant effect (Volavka et al 1992). The was the lingering uncertainty as to whether or not antipsy- reality is that blood levels have not come to be used chotic drugs were really affecting the long-term course of routinely as a reliable and valued guide to clinical man- schizophrenia or merely reducing the more florid mani- agement. (Whether this is the consequence of lack of knowledge transfer, impediments related to cost and fea- It is sobering to note that attempts to assess historical sibility, or an insufficiently compelling data base could be trends in the outcome of schizophrenia over the long term have not shown significant progress in the past several Another important development in the pharmacologic treatment of schizophrenia was the introduction of long- In the 1980s, a number of factors helped to change the acting injectable (depot) medications. Compliance in oral scope of both clinical investigation and expectation. First, medication-taking over long periods of time is an enor- enhanced methodology was developed for assessing neg- mous problem (Kane 1985). Initially, the mirror image and ative symptoms (problems differentiating primary from controlled trials with these medications led to considerable secondary negative symptoms on a short-term basis re- enthusiasm regarding their potential to improve long-term main evident) and a renewed emphasis was placed on their outcomes, by reducing rates of relapse and rehospitaliza- importance. Second, advances in neuropsychology helped tion. Interestingly, a number of trials comparing oral and to generate an extensive and more refined data base on depot medication were not successful in demonstrating as various aspects of cognitive dysfunction in schizophrenia.
significant a reduction in rates of relapse as expected These findings contributed to increased awareness of the (Glazer and Kane 1992). We have argued elsewhere (Kane relative independence of cognitive dysfunction from pos- and Borenstein 1985) that, to some extent, this is likely a itive psychotic symptoms (including the former preceding result of the methodology employed in these trials. Spe- the onset of the latter) and its important impact on cifically, relatively compliant patients were selected given functional outcome. Third, the renewed interest in and the nature of the study design (double-blind oral and depot ultimate marketing of clozapine reenergized the search for drug, frequent assessments, need for consent, etc.) Perhaps new drugs with different characteristics and challenged the most importantly, almost all of these trials lasted only 1 field to look for differential effects of new generation year. Given the time frame likely involved in becoming noncompliant and the average lengthy delay between Much remains to be done in establishing the relative actually discontinuing medication and subsequent relapse merits of new generation drugs. The one property that is in stable or remitted patients (i.e., 3–7 months) a 1-year clearly apparent in most studies is reduced propensity to trial is unlikely to find a difference. The only controlled produce neurologic adverse effects. Some researchers comparison between oral and depot preparations that have questioned the role of dosage of the comparative lasted more than 1 year (Hogarty et al 1979) found a drug (and the fact that it has usually been haloperidol) in striking difference in the second year, which did not reach this context. Interestingly, one study (involving sertindole) statistical significance because of the small sample size.
(Zimbroff et al 1997) did use three different doses of Unfortunately, depot drugs were initially administered haloperidol in a fixed-dose design. The lowest dose (4 mg) in higher than necessary doses (Glazer and Kane 1992) was associated with significantly more extrapyramidal and were presumed to cause more side effects when in side effects (EPS) than placebo or any dose of sertindole, reality there is no evidence that this is the case. In addition, supporting the assumption that the superiority of the depot drugs have not been widely used in the United States second-generation drugs regarding EPS is not solely due for a variety of nonscientific reasons, despite their poten- tial value in an illness with very high rates of noncompli- As suggested previously, there are a number of critical ance. (The role of new generation drugs in potentially areas in which the new drugs need to be evaluated, reducing rates of non compliance will be discussed be- particularly against each other, but in some cases against conventional drugs as well. Clearly, the management of The first three decades of widespread antipsychotic use refractory or poorly responsive patients is a high priority, highlighted a number of major deficiencies. The first but so are the domains of negative symptoms, cognitive dysfunction, compliance, and relapse prevention. Data are will focus largely on populations of treatment-refractory beginning to emerge on some of these issues, but defini- tive conclusions in a number of areas are as yet difficult to Kane et al (1988) reported on 267 schizophrenia pa- make. In addition, the question as to whether or not these tients who met strict criteria for treatment resistance.
drugs can have more of an impact on various domains of Patients had to fail to respond to adequate trials of at least outcome if they are introduced earlier in treatment is three antipsychotic drugs and a prospective, single-blind beginning to receive appropriate attention. Another oppor- trial of haloperidol. Clozapine was associated with signif- tunity given better adverse-effects profiles of new drugs icantly greater improvement in both positive and negative might be to attempt to prevent the emergence of psychotic symptoms. Using a priori criteria for improvement, 30% symptoms in patients who are exhibiting prodromal signs, of clozapine-treated patients improved after 6 weeks in or eventually in those individuals with highly specific risk comparison with 4% of those receiving chlorpromazine.
factors even prior to becoming symptomatic.
Since that report, a number of controlled trials have As we learn more about the complexity of this disease confirmed clozapine’s superiority over conventional drugs (or diseases), we should also remain open to the possibility in improving positive symptoms (Pickar et al 1992; Breier that a combination of different pharmacologic agents as et al 1994). However, clozapine’s impact on primary well as nonsomatic treatment could be necessary to pro- negative symptoms or deficit state remains the subject of duce optimum results. We seem, at times, to harbor what ongoing debate (Carpenter et al 1995; Kane 1996a; Melt- might be an unrealistic expectation that a single agent will zer 1995). For example, in a 6-month, double-blind influence some basic pathophysiologic factor that is pri- comparison of clozapine and a modest dose of haloperidol mary to all of the diverse manifestations of this illness, (10 mg/day) conducted in poor or partially responsive rather than recognizing the more likely possibility that outpatients, we found clozapine to be superior in positive palliative treatments are necessary in diverse areas of brain symptoms, but not on negative symptoms.
function, while we await a deeper understanding of etiol- Two relatively long-term controlled trials have now been reported with clozapine. Essock et al (1996a) carriedout an effectiveness trial of clozapine in which 227treatment-refractory, state hospital inpatients were ran- Clozapine
domly assigned to nonblind, continued “usual care” ortreatment with clozapine. There was no significant differ- The introduction of clozapine into clinical practice helped ence between clozapine and the usual care condition in the to set the stage for a variety of new perspectives on rate of discharge over the 24-month follow-up period, nor antipsychotic drug treatment, drug development and out- were there any significant differences on measures of come assessment. Clozapine was the first drug to demon- psychopathology. However, clozapine subjects who were strate a qualitative difference in propensity to produce discharged were significantly less likely to be rehospital- neurologic side effects (Casey 1989) and the first drug to ized than those in the comparison group.
show clear superiority over other drugs in the treatment of Rosenheck et al (1997) conducted a randomized, 1-year, refractory patients (Kane et al 1988).
double-blind, comparative trial of clozapine and haloper- As a result, an enormous amount of research has been idol in 423 patients at 15 Veterans Affairs Medical conducted to determine what neuropharmacologic charac- Centers. Significantly, more (57%) clozapine-treated pa- teristics of clozapine are responsible for these novel tients continued on the assigned treatment for the whole effects. At present, there are an extraordinary array of year in comparison with only 28% in the haloperidol findings showing differences between clozapine and con- group. Also, significantly more haloperidol patients dis- ventional drugs on neuroimaging, and neurophysiologic continued drug treatment due to worsening of symptoms and neurochemical measures. However, at present it is or lack of efficacy (51%) in comparison with clozapine difficult to make definitive statements as to what charac- patients (15%). Among subjects taking clozapine, a 20% teristic(s) account for clozapine’s clinical profile.
reduction in the total score on the Positive and Negative As a result of clozapine’s adverse effect profile, specif- Syndrome Scale was seen in 24% of subjects after 6 weeks ically, what initially appeared to be a 1–2% risk of and 37% after 1 year. However, this difference was not potentially fatal agranulocytosis (incidence estimates have significant from the 13% and 32%, respectively, who been revised downward and now are below 0.4% in the showed the same level of improvement on haloperidol.
first year of treatment [Novartis, unpublished data]), the The data regarding relapse are intriguing. In general, we use of this drug has generally been confined to treatment- do not think of those patients who relapse despite contin- refractory, or in some cases, treatment-intolerant patients.
ued maintenance medications as “refractory,” but clearly Therefore, any discussion of clozapine’s clinical effects this subgroup is not responding optimally. Even with guaranteed medication delivery (i.e., depot drugs), approx- slightly less than 10% considered unchanged or worse). In imately 15–20% of patients will experience a relapse over more recent trials, the proportion of patients considered to a 1-year period (Kane 1996b). There are hardly any data to have an inadequate response appears to be somewhat inform clinical strategies for managing such patients. We higher. Representative trials that attempted to exclude (Pollack et al 1998) also have observed in a mirror image refractory patients published in the 1990s reported an analysis of hospitalization before and after clozapine in 81 average good response rate of 50% (Levinson et al 1990; patients that the rate on clozapine was significantly re- Van Putten et al 1990; Rifkin et al 1991; Volavka et al duced. Clearly, given the mirror image design there are other factors that may have contributed to this difference.
This response rate raises the question as to whether or With clozapine’s evidence of novel properties, a con- not response to medication has declined over the past three trolled trial focusing on relapse prevention in patients who decades. There certainly are a number of factors that could have “broken through” other drugs would be of consider- have contributed to the appearance of such a decline. First, the nosology has changed considerably during that inter- There have been a series of reports suggesting superiority val, with the definition of schizophrenia becoming nar- for clozapine in a number of areas such as hostility and rower and that of affective illness becoming broader (at violence (Mallya et al 1992), substance abuse (Albanese et al least in the United States). Therefore, it is possible that 1994; Buckley et al 1994), and suicidal behavior (Meltzer more modern-day trials include a relatively more homog- and Okayli 1995). These observations have been derived enous population of patients with schizophrenia. Second, mainly from uncontrolled observations of primarily chronic the prevalence of substance abuse among patients with refractory or partially drug-responsive patients.
schizophrenia has increased enormously over the past 20 It is important now, given the availability of alternative years. Some studies (Regier et al 1990; Shaner et al 1993) second-generation drugs, to examine these effects in find a 50% lifetime prevalence of substance abuse among controlled trials. Despite the proven advantages of cloza- patients with schizophrenia. Although it is clear that drugs pine, this drug has been largely underused in clinical of abuse can exacerbate symptoms of psychosis, the extent practice. This was the case even before the introduction of to which chronic drug abuse alters the potential respon- other second-generation, or putative “atypical” com- siveness of the underlying condition to antipsychotic medication has not been determined adequately. Because,however, some of the same receptors are involved in theseeffects, there is reason for concern.
Treatment Refractoriness: General Issues
A third possible factor is the trend in recent years to The success of clozapine in refractory patients led to limit hospitalization whenever possible. This trend could renewed interest in and hope for developing better treat- have resulted in a change in the population of patients ment strategies for such patients. The first consideration in available for clinical trials. In addition, there may be an this context is the definition or criteria by which patients inherent bias in the way patients are selected for clinical are identified. In the original multicenter clozapine trial trials. Especially when placebo controls are utilized, pa- (Kane et al 1988), a very stringent definition was used; tients who have responded well to medication might be however, it is clear that many other patients (who would reluctant to risk placebo exposure or exposure to an not necessarily meet those criteria) derive only partial experimental drug. In contrast, those patients who have benefit from conventional drugs. As a rule, definitions of had only poor or partial response and/or those who have treatment resistance have relied more heavily on positive had significant adverse effects from earlier treatment symptom assessment, yet it is clear that many patients might be more likely to participate in such trials unless suffer from persistent negative symptoms such as apathy, explicit efforts are made to exclude them.
anhedonia, or affective impairment. In addition, many Recently, we (Kane and Borenstein unpublished data) patients suffer in the so-called third factor domain of reviewed response rates in more than 700 haloperidol- disorganization and cognitive dysfunction. (We have al- treated patients participating in placebo-controlled, new luded previously to the problem of those patients who drug development trials and found that the proportion of “break through” adequate maintenance or prophylactic patients improving 30% or more on the Brief Psychiatric treatment as another subgroup who might be considered Rating Scale (BPRS) total score ranged from 12% to 45%.
Overall, the demographic and treatment history character- By far the largest category of patients appears to be poor istics of patients participating in these trials reveals that or partial responders. In those trials conducted in the 1960s most are male, have an average age in the mid-to-late 30s, (Klein and Davis 1969), approximately 38% of patients have had six or more prior hospitalizations, and have been derived little if any benefit from medication (including the These demographics underscore the fact that patients for a study such as our original clozapine study; however, participating in these trials represent a particular subgroup given increasing pressure on length of stay clinicians are at a chronic phase of illness. It is likely that this population frequently inclined to try alternatives even within a rela- may have limitations or differences in terms of delineating tively brief time frame if patients are not improving. A the potential impact of novel pharmacologic agents on the sample of 156 acutely ill patients were enrolled initially.
course of schizophrenia in comparison with patients who At the end of 4 weeks of fluphenazine, 115 subjects had completed this portion of the trial and 68% of those stillhad a score of moderate or more on at least one psychoticitem. Fifty-eight patients entered a second 4-week phase in Alternative Treatment Strategies
which they were randomly assigned double-blind to con- When patients do not respond adequately, clinicians face a tinue on fluphenazine 20 mg/day; have the dose increased challenge as to when and if to abandon the current to 80 mg/day; or be switched to haloperidol 20 mg/day.
treatment, to increase the dose, to add adjunctive medica- Only 9% of subjects who completed this phase achieved tion, when to switch to another drug, and which drug to the a priori response criteria and there were no significant switch to. Despite the frequency of this situation, there are differences between the three alternatives.
still inadequate data by which to inform clinical practice.
It is possible, however, that further improvement would The dilemmas faced for years by clinicians with conven- be observed with specific alternatives administered over a tional drugs are now being repeated in many cases with the longer period of time than 4 – 8 weeks. Shalev et al (1993) evaluated the proportion of acutely exacerbated schizo- It is sobering to look at the literature with conventional phrenic patients who remained unimproved after consec- drugs to suggest what lessons might be learned for current utive administration of haloperidol, chlorpromazine, and and future research with newer medications. When drugs perphenazine in randomly determined order. Two criteria are developed initially, the pharmaceutical companies do were used concurrently to define therapeutic success: a not necessarily have an incentive for studying a broad decrease of 30% in a modified version of the BPRS; and range of doses or even establishing clearly what the clinical improvement sufficient to permit discharge back maximum tolerated dose will be. As a result, once a drug to the community. Sixty patients completed the trial (58% is marketed there is a natural tendency among clinicians to were woman, the mean age was 33 years, and the number try higher doses (assuming they are tolerated) in patients of previous admissions averaged four). Each phase of the who are not responding adequately. This practice became study averaged 28 –33 days. Patients who did not improve such a widespread tendency with the conventional drugs during each 4-week period were switched to the next drug that it was not at all uncommon to see many patients according to their predetermined schedule; however, if treated with extremely high doses of fluphenazine or there was a “subthreshold” improvement, the drug was haloperidol (or other mid-to-high potency drugs). Al- continued another week before determining ultimate re- though some studies suggested that high-dose or even sponse to that phase. The average daily doses were 27–30 “megadose” treatment might be beneficial in a proportion mg for haloperidol, 350 –380 mg for chlorpromazine, and of refractory patients, well-designed studies (Prien and Cole 1968; Quitkin et al 1975; Bjorndal et al 1980) found The total improvement rate was 95%; only 3 of 60 little consistent advantage. Despite this lack of data, many patients remained unimproved after the three phases of the clinicians continue to use high doses, especially in patients study. Of the patients entering each phase of the trial, 67% who are exhibiting hostile or aggressive behavior (Rem- improved after the first drug, 55% after the second, and 67% after the third. The differences in improvement rates There are remarkably few studies assessing the value of between the phases were not statistically significant, nor switching from one antipsychotic drug to another, either were differences found between men and women. The with conventional drugs or with the new generation drugs.
differences in improvement rates between the three drugs In addition, studies remain to be conducted of dosage over the entire trial were not statistically significant.
escalation (or reduction) in poor or partially responsive The disparity in results between these two trials might patients with the new generation drugs.
be due at least in part to differences in the criteria for good We (Kinon et al 1993) reported results from a study response. Whereas the Kinon et al (1993) study required designed to compare the relative efficacy of treatment ratings of no more than “mild” on any psychotic item, the alternatives using conventional drugs in patients who Shalev et al (1993) study required a 30% improvement on continued to exhibit clinically significant positive symp- the total BPRS scale. In addition, the Shalev study used a toms after a 4-week trial of fluphenazine (20 mg/day).
4-point version of the BPRS so the rating scale data cannot These were by no means patients who would meet criteria be compared easily. It is also not possible to determine whether or not the improvement occurring in the second mean dose 7.75 mg) or clozapine (n ϭ 57, mean dose 420 and third phases was in fact due to the change in drug or mg/day) over an average of 12 weeks. Utilizing a response the passage of additional time. In addressing this issue, criterion of a 20% reduction in total Positive and Negative Shalev et al (1993) rightly point out that it is difficult to Syndrome Scale (PANSS) score, 44% of the clozapine- maintain poor responders on the same drug for 12 weeks.
treated and 28% of the risperidone-treated patients re- It is notable that none of the drugs appeared superior in sponded. Bondolfi et al (1998) have reported on 86 this context, despite some differences in receptor binding inpatients with treatment refractory schizophrenia as- signed randomly to an 8-week double-blind trial of ris- We have already discussed the role of clozapine in peridone or clozapine. The mean daily dose for clozapine treatment-refractory schizophrenia. In general, clozapine was 291 mg and for risperidone 6.4 mg. The criterion for has the most data supporting its efficacy in this population.
categorizing patients as improved was a 20% reduction in A recent meta-analysis conducted by Wahlbeck (1998) total PANSS score. At the end of the study, 67% of involved a review of 2,560 randomized participants from risperidone-treated patients and 65% of clozapine-treated 31 mainly short-term trials. Wahlbeck concluded that patients met response criteria. This study has been criti- clozapine was clearly more effective than conventional cized on the grounds that the dose of clozapine was low by antipsychotics in reducing symptoms of schizophrenia, some standards and the response rate was higher than both in treatment-resistant and nonresistant patients. How- generally seen in 8-week trials of clozapine in refractory ever, Wahlbeck found no convincing evidence that the patients, suggesting this population may have been some- superior clinical effect of clozapine results in improved levels of functioning. It is not possible to determine to Additional data will be necessary to clarify the relative what extent this is a result of inadequate methodology or value of risperidone in treatment-refractory patients and a number of relevant trials are currently underway.
Other treatment strategies for refractory patients have included adjunctive treatments such as lithium, benzodi-azapines, anticonvulsants ,and electroconvulsive therapy.
Though some patients may benefit from these strategies, Olanzapine was marketed in the United States beginning the proportion of such patients is low (Christison et al in 1996. Several large-scale trials have been conducted 1991). We continue to struggle with the question of which demonstrating olanzapine’s efficacy in the treatment of alternatives should be pursued when the one or two most schizophrenia. Most studies were controlled with placebo likely to succeed do not. Clinicians must balance the and/or haloperidol, but more recently studies are being appropriateness of trying numerous unproven treatments reported comparing olanzapine with risperidone or cloza- or accepting the possibility that with our current state of pine. As is the case with all new drugs, in the initial studies knowledge we are not always successful.
conducted for drug regulatory purposes the patient demo-graphics were such that the efficacy results might have Second-Generation Antipsychotics
underestimated the impact in younger patients, and/orpatients closer to the onset of the illness. In general, the studies showed at least equivalent antipsychotic efficacy Risperidone was marketed beginning in 1994 in the United to haloperidol, with some evidence of superiority on States and has been used widely since then. Risperidone’s negative symptoms and depression (Beasley et al 1996; efficacy was established in seven clinical studies of Tollefson et al 1997b, 1998). Clear superiority was evident acutely psychotic patients (Borison et al 1992; Claus et al concerning neurologic side effects and several other ad- 1992; Muller-Spahn 1992; Ceskova and Svestka 1993; Chouinard et al 1993; Heinrich et al 1994; Marder and In addition, analysis of double-blind, continuation data Meibach 1994). In the United States-Canadian multicenter from patients receiving olanzapine or haloperidol revealed trial, 6 or 16 mg of risperidone per day were superior to a significantly lower incidence of treatment-emergent haloperidol (20 mg/day) on reduction in total positive and dyskinesia among olanzapine-treated patients (Tollefson negative syndrome scale scores. At present, there is no et al 1997a). The incidence observed with haloperidol was evidence that doses higher than 6 mg/day offer any similar to that observed in incidence studies involving advantage and are associated with a higher incidence of conventional antipsychotics (Kane 1995), a result that is EPS. In initial trials (Marder and Meibach 1994), risperi- helpful in supporting the validity of the findings.
done appeared to be helpful for some refractory patients.
In a more recent study (Tran et al 1997), olanzapine was Flynn et al (1998) described a series of cases of refractory compared with risperidone in 339 schizophreniform, patients treated openly with either risperidone (n ϭ 29, schizophrenia, and schizoaffective patients over a 28-week treatment period. The mean dosage of olanzapine was 16.9 of double-blind trials (Arvanitis et al 1997; Peuskens and mg/day and that of risperidone 7.3 mg/day. Side effects, Link 1997). Superiority was demonstrated over placebo on particularly extrapyramidal, were more common in the measures of positive and negative symptoms. No signifi- risperidone group at week 8 and week 28. The olanzapine cant superiority in efficacy was demonstrated in compar- treatment group showed significantly greater improvement ison with conventional drugs, but quetiapine was shown to on measures of negative symptoms, and a significantly produce very low levels of EPS and to be generally well lower rate of exacerbation during the 28-week trial. The tolerated. To date, there are no published data available in average dose of risperidone used in this study was some- treatment-refractory patients for quetiapine. Concern what greater than is generally recommended today for the about possible ophthalmologic effects has led to recom- treatment of psychosis and this difference might have mendations of eye examinations at baseline and during influenced the higher rate of adverse effects seen with risperidone. However, half the patients on risperidone In summary, as a class, the second-generation antipsy- received dosages less than 6 mg/day.
chotics clearly have advantages over conventional drugs Studies are beginning to be reported examining olanza- particularly in the area of adverse effects (with the pine’s potential for improving outcome in treatment- exception of weight gain). Data with regard to positive and refractory patients. In one open trial of olanzapine, 36% of negative symptom superiority are still inconsistent and patients achieved response criteria (Martin et al 1997) and limited by methodologic considerations.
in another, 48% of treatment refractory patients were When differences have been found between new com- reported to have responded to olanzapine (Baldacchino et pounds and standard agents (i.e., haloperidol), the effect sizes have generally been small to medium. Schooler Conley et al (1998) reported on an 8-week, double-blind (1998) reviewed data on representative studies and found comparison of olanzapine and chlorpromazine in 84 treat- a mean effect size of .20 in comparison with haloperidol ment-refractory subjects. No difference in efficacy was on negative symptoms (with the highest being .53).
observed on measures of psychopathology. Seven percent Clearly, diminution of parkinsonian side effects can of the olanzapine-treated patients responded according to a have important effects on appearance, subjective well- priori criteria and no chlorpromazine-treated patients re- being, and functioning, so these gains are considerable. In sponded; however, this difference was nonsignificant. The addition, reducing the need for the adjunctive antiparkin- mean dosage of chlorpromazine was 1,173 mg/day. All sonism medications has other potential benefits.
olanzapine-treated patients received 25 mg/day. Overall, Other domains of outcome are now being explored as neither drug group experienced substantial change in the well. Increasing attention has been focused on cognitive severity of psychosis from their baseline state.
function as an area of considerable impairment in schizo- Breier and Hamilton (1999) reported on a subpopulation phrenia. An estimated 75% of patients with schizophrenia of patients meeting retrospectively applied criteria for exhibit some neuropsychological dysfunction (Gold and treatment-refractory schizophrenia, selected from a large Harvey 1993). It has been shown that cognitive dysfunc- prospective, double-blind, 6-week study of olanzapine and tion can be an important determinant of poor social and haloperidol (Tollefson et al 1997b). Olanzapine showed vocational outcome and can play more of a role in these superiority over haloperidol for those who completed the domains than psychotic signs or symptoms (Green 1996).
study on most of the major measures of psychopathology.
It also appears that some type of cognitive dysfunction This study was limited by the retrospective nature of the precedes the overt onset of a schizophrenic illness (Erlen- determination of nonresponse, which relied on failure to respond to at least one 8-week trial of a neuroleptic during The new generation antipsychotics appear to hold con- the previous 2 years and presence of a specified level of siderable promise in either improving or not worsening severity at baseline when subjects were still being treated specific aspects of cognitive functioning. So far, there with a prior neuroleptic. These criteria are not as clear or remain only a relatively small number of studies and very selective as those used in Conley’s study.
few controlled trials focusing on the impact of second Further research is necessary to clarify the role of generation drugs on cognitive function.
olanzapine in treatment-refractory patients.
Reports on clozapine are mixed and somewhat contra- dictory (Hagger et al 1993; Goldman et al 1996). Reportson risperidone are beginning to emerge (Green et al 1997) as are data on olanzapine (Canadian Cognition and Out- The most recently marketed antipsychotic in the United come Study Group 1998), but it is too soon to draw firm States is quetiapine. This compound was also compared with placebo and conventional antipsychotics in a number A critical area for assessing the ultimate impact of Figure 1. Cumulative rates of relapse amongpatients with schizophrenia after 1 year of main-tenance therapy with various doses of antipsy-chotic drugs, in six studies. Most patients hadhad more than one prior psychotic episode.
Fluphenazine decanoate was given biweekly,and haloperidol decanoate was given monthly.
Doses of fluphenthixol decanoate, which wasused in some of the studies, are expressed interms of estimated equivalent doses of fluphen-azine decanoate. The number of patients in eachgroup in these studies ranged from 29 to 63.
intervention strategies involves measures of functioning.
effects in relatively short-term trials, there has been far Here too, results from controlled trials (even with cloza- less activity in the maintenance treatment arena. In addi- pine) are not as clear-cut as one would like, but this will tion, such trials are difficult to conduct, take long periods become an increasingly important domain in comparing of time, and can be very expensive. At the same time, if we the impact of different treatment strategies.
really are interested in altering the course and improving Overall, there are still too few direct comparisons of the functional outcome of this disease, it is producing and new drugs against each other. There are also no data sustaining gains over the long-term that is critical.
providing predictors of response that might help clinicians The importance of maintenance medication has been to choose the ideal drug for a specific patient. There is well established in numerous double-blind, placebo-con- hope that pharmacogenomics may provide important guid- trolled trials (Davis 1975; Kane and Lieberman 1987; ance; however, we have not yet seen consistent results in Gilbert et al 1995). Even if the focus is just on patients this regard (Arranz et al 1995; Malhotra et al 1996).
who had been in remission for a year or more, availablestudies indicate that approximately 75% will relapsewithin 12–18 months after the discontinuation of antipsy- Maintenance Treatment
chotic medication (Kissling 1991; Davis et al 1994). These Given the nature of schizophrenia and its long-term data are particularly important for clinicians to appreciate course, the role of maintenance treatment is absolutely because it is not uncommon for both patients and clini- critical. Because drug development relies heavily on acute cians to develop a false sense of security regarding Figure 2. Rates of relapse among patientswith schizophrenia after 1 year of contin-uous or intermittent maintenance therapy,in five studies. Black bars represent con-tinuous treatment, and shaded bars inter-mittent treatment. Most patients had hadmore than one prior psychotic episode.
The number of patients in each group inthese studies ranged from 27 to 121.
diminished need for medication when the patient has been treatment is feasible, but that the risk of relapse during in remission for a year or more. Current American therapy is somewhat greater than if the maintenance dose Psychiatric Association (APA) guidelines (Herz et al is higher (Figure 1). Although lower doses were associated 1997) recommend that individuals recovering from their with fewer adverse effects, in general, this relationship has first episode of illness should probably be treated for 1–2 not been demonstrated clearly with respect to tardive years; however, 75% of patients will have relapses after therapy is discontinued. These guidelines should be re- Intermittent treatment is a strategy intended to take vised as fears of long-term drug toxicity decline. Patients advantage of the observation that many individuals do not who have had two or more episodes should receive experience a relapse until several months after the discon- maintenance treatment for at least 5 years (according to tinuation of antipsychotic medication. This strategy in- current APA guidelines) and many experts believe they cludes the discontinuation of antipsychotic medication in should be treated indefinitely (Davis et al 1994). We have stable outpatients who are in remission and the reinstitu- been hard pressed to identify subgroups of patients to tion of medication if and when prodromal signs of relapse whom these recommendations do not apply. It is also occur. The hope of this “early intervention” is to prevent clear, as previously discussed, that 15–20% of patients a full-blown psychotic relapse. The overall goal is to will relapse within a year despite guaranteed medication reduce the cumulative exposure to antipsychotic drugs and delivery (i.e., depot drugs). When rates of noncompliance limit adverse effects. Five major studies have examined are also taken into consideration, the frequency and this strategy (Jolly et al 1989, 1990; Carpenter et al 1991; consequences of relapse take on enormous public health Herz et al 1991; Pietzcker et al 1993; Schooler et al 1997).
proportions (Weiden and Olfson 1995).
All of the patients in these trials were treated for 2– 6 An important focus of recent research with conventional months before therapy was discontinued in those individ- antipsychotic drugs has been to improve the overall uals who were assigned to the intermittent treatment benefit-to-risk ratio of long-term antipsychotic drug treat- groups. The cumulative rates of relapse after 1 year in the ment by establishing optimal strategies. Studies comparing groups that received continuous treatment were similar to different fixed doses and comparing continuous with those reported for the conventional doses of antipsychotic intermittent therapy have been conducted. Six studies drugs in the trials involving low-dose therapy. The results involving different doses given for at least 1 year have with intermittent treatment, however, were in general been reported (Kane et al 1983, 1986, 1993a; Marder et al discouraging (Figure 2). The rates of relapse during the 1984, 1987; Johnson et al 1987; Hogarty et al 1988; 1-year study follow-up interval were high and in addition Schooler et al 1997). All involved long-acting, injectible there was little benefit in terms of psychosocial adjust- medication given at intervals of 2– 4 weeks. The results ment, enhanced subjective well-being, or diminished inci- were similar in suggesting that low-dose maintenance dence of tardive dyskinesia. In light of the overall twofold increase in the risk of relapse, this approach is difficult to oral medication), but in addition providing some further justify. On the other hand, the intermittent approach may protection against relapse for several weeks, allowing all be useful to those patients who adamantly refuse to take interested parties to respond to this “crisis.” It is noteworthy that all of these studies involving Conclusion
alternative maintenance treatment strategies used depotantipsychotic drugs. The reason for this method is that Considerable progress has been made in multiple areas of investigators attempting to make comparisons between schizophrenia research. Undoubtedly, these advances will different dosage levels or different strategies for adminis- help to set the stage for further gains. At present, there are tration need to be absolutely certain that the patient is more new antipsychotic drugs and more activity in drug taking the medication as prescribed. The only way that it development than at any time in the past three decades.
has been possible to ensure compliance up until now has New generation medications are demonstrating important been by giving long-acting injectible medications. It is advantages over conventional drugs, particularly in the ironic, however, that in clinical practice these drugs are area of adverse neurologic effects. Further research will used far less frequently (Glazer and Kane 1992). Now, clarify their impact on negative symptoms, depression, with the introduction of a new generation of antipsychotic cognitive dysfunction, suicidal ideation or behavior, and, drugs, which appear to be associated with a lower risk of ultimately, the most critical domain of functional outcome.
neurologic adverse effects, the clinician faces a dilemma The improved benefit-to-risk ratio of new drugs should in terms of patients who are potentially noncompliant.
reduce ambivalence about long-term treatment and en- None of the new drugs are currently available in a hance the possibility of early intervention in prodromal long-acting injectible form. It is assumed that the reduc- tion in adverse effects expected over long-term treatment It is critical that methodologic rigor and replication of with the new antipsychotic drugs will lead to a reduction findings as well as appropriate knowledge transfer be in rates of noncompliance. However, this hypothesis facilitated to solidify and implement the gains that have remains untested so far and it is also clear to clinicians and investigators that adverse effects of medication do notexplain all of the variance in the development of noncom-pliance. We continue, therefore, to be faced with the This work was supported by the following NIMH grants: MH 31776; MH dilemma of increasing our certainty of preventing relapse, 32369; MH 3992; MH 41960; MH 42929; MH 46633.
This work was presented at the conference, “Schizophrenia: From but also increasing the potential risk of adverse effects Molecule to Public Policy,” held in Santa Fe, New Mexico, in October such as tardive dyskinesia. It is hoped that new medication 1998. The conference was sponsored by the Society of Biological delivery methods will become available for the new Psychiatry through an unrestricted educational grant provided by Eli Lily generation antipsychotic drugs to resolve this dilemma.
Another interesting aspect of long-acting injectible medication is the suggestion that even following the References
discontinuation of such drugs, relapse rates are reducedbecause of the very gradual withdrawal resulting from the Albanese MJ, Khantzian EJ, Mercy FL, Green AI (1994): pharmacokinetic properties of these agents. Viguera et al Decreased substance use in chronically psychotic patientstreated with clozapine. Am J Psychiatry 151:780 –781.
(1997) reported that with abrupt cessation of oral antipsy- Arranz M, Collier D, Sodhi M, et al (1995): Association between chotic drugs the relapse rate reached 50% within 30 clozapine response and allelic variation in 5-HT 2A receptor weeks, with little additional risk thereafter up to 3.7 years.
The risk was lower, however, after gradual discontinuation Arvanitis LA, Miller DG and the Seroquel Trial Thirteen Study of oral treatment. A similar finding was reported by Group (1997): Multiple fixed doses of “SEROQOUEL” Gilbert et al (1995) in their review of the literature on (quetiapine) in patients with acute exacerbation of schizo- discontinuation of maintenance medication. The conclu- phrenia: a comparison with haloperidol and placebo. Biol sions that can be drawn from these reviews are tentative because of methodologic concerns (lack of random assign- Baldacchino AM, Stubbs JH, Nevison-Andrews D (1998): The use of olanzapine in non-compliant or treatment-resistant ment in comparison groups, etc.); however, this finding clozapine populations in hospital. Pharm J 260:207–209.
should be considered in planning future studies.
Beasley CM, Tollefson G, Tran P, et al (1996): Olanzapine Depot medications, therefore, have the advantage of not versus placebo and haloperidol: acute phase results of the only making it clear to family members and clinicians North American double-blind olanzapine trial. Neuropsycho- when medication has been discontinued (which may not necessarily be the case when an individual is receiving Bjorndal N, Bjerre M, Gerlach J, et al (1980): High dosage haloperidol therapy in chronic schizophrenic patients: a dou- Essock SM, Hargreaves WA, Covell NH, Goethe J (1996a): ble-blind study of clinical response, side effects, serum Clozapine’s effectiveness for patients in state hospitals: re- haloperidol, and serum prolactin. Psychopharmacology 67: sults from a randomized trial. Psychopharmacol Bull 32:683– Bondolfi G, Dufour H, Patris M, et al (1998): Risperidone versus Essock SM, Hargreaves WA, Dohm F-A, Goethe J, Carver L, clozapine in treatment-resistant chronic schizophrenia: a ran- Hipshman L (1996b): Clozapine eligibility among state hos- domized double-blind study. Am J Psychiatry 155:499 –504.
pital patients. Schizophr Bull 22:15–25.
Borison RL, Pathiraja AP, Diamond BI, Meibach RC (1992): Flynn SW, MacEwan GW, Altman S, Kopala LC (1998): An Risperidone; clinical safety and efficacy in schizophrenia.
open comparison of clozapine and risperidone. Pharmacopsy- Psychopharmacol Bull 28:213–218.
Breier A, Buchanan RW, Kirkpatrick B, et al (1994): Effects of Gilbert PL, Harris NJ, McAdams LA, Jeste DV (1995): Neuro- clozapine on positive and negative symptoms in outpatients leptic withdrawal in schizophrenic patients: a review of the with schizophrenia. Am J Psychiatry 151:20 –26.
literature. Arch Gen Psychiatry 52:173–188.
Breier A, Hamilton SH (1999): Comparative efficacy of olanza- Glazer WM, Kane JM (1992): Depot neuroleptic therapy: an pine and haloperidol for patients with treatment-resistant underutilized treatment option. J Clin Psychiatry 53:426 – schizophrenia. Biol Psychiatry 45:403– 411.
Buckley P, Thompson P, Way L, Melter HY (1994): Substance Gold JM, Harvey PD (1993): Cognitive deficits in schizophrenia.
abuse among patients with treatment-resistant schizophrenia: Psychiatr Clin North Am 16:295–312.
characteristics and implications for clozapine in therapy. Am J Goldman RS, Bates JA, Bilder RM, Kane J, Schooler N, Marter S (1996): The effects of clozapine and haloperidol on mem- Canadian Cognition and Outcome Study Group (1998): Neuro- psychological change in early phase schizophrenia over twelve months of treatment with olanzapine, risperidone orhaloperidol. Schizophr Res 29:152–153.
Green MF (1996): What are the functional consequences of neurocognitive deficits in schizophrenia? Am J Psychiatry Carpenter WT Jr, Conley RR, Buchanan RW, Breier A, Tam- minga CA (1995): Patient response and resource manage-ment: another view of clozapine treatment of schizophrenia.
Green MF, Marshal VD, Wirshing WC, et al (1997): Risperidone Am J Psychiatry 152:827– 832.
effects on verbal working memory. Schizophr Res 24:214.
Carpenter WT Jr, Hanlon TE, Heinrichs DW, et al (1990): Hagger C, Buckley P, Kenny JT, Freedman L, Ubogy D, Meltzer Continuous versus targeted medication in schizophrenic out- HY (1993): Improvement in cognitive functions and psychi- patients: outcome results. Am J Psychiatry 147:1138 –1148 atric symptoms in treatment-refractory schizophrenic patients [Erratum, Am J Psychiatry 1991;148:819].
receiving clozapine. Biol Psychiatry 34:702–712.
Casey DE (1989): Clozapine: neuroleptic-induced EPS and Hegarty JD, Baldessarini RJ, Tohen M, Waternaux C, Open G tardive dyskinesia. Psychopharmacology 99(suppl):S47–S53.
(1994): One hundred years of schizophrenia: a meta-analysisof the outcome literature. Am J Psychiatry 151:1409 –1416.
Ceskova E, Svestka J (1993): Double-blind comparison of risperidone and haloperidol in schizophrenia and schizoaffec- Herz MI, Glazer WM, Mostert MA, et al (1991): Intermittent vs tive psychoses. Pharmacopsychiatry 26:121–124.
maintenance medication in schizophrenia: two-year results.
Chouinard G, Jones B, Remington G, et al (1993): A Canadian Arch Gen Psychiatry 48:333–339.
multicenter placebo-controlled study of fixed doses of risperi- Herz MI, Liberman RP, Lieberman Ja, et al (1997): American done and haloperidol in the treatment of chronic schizo- Psychiatric Association practice guideline for the treatment of phrenic patients. J Clin Psychopharmacol 13:25– 40 [Erra- patients with schizophrenia. Am J Psychiatry 154(suppl):1– tum, J Clin Psychopharmacol 1993;13:149].
Christison GW, Kirch DG, Wyatt RJ (1991): When symptoms Hogarty GE, McEvoy JP, Munetz M, et al (1988): Dose of persist choosing among alternative somatic treatments for fluphenazine, familial expressed emotion, and outcome in schizophrenia. Schizophr Bull 17:217–245.
schizophrenia results in a two-year controlled study. Arch Claus A, Bollen J, De Cuyper H, et al (1992): Risperidone versus Gen Psychiatry 45:797– 805.
haloperidol in the treatment of chronic schizophrenic inpa- Hogarty GE, Schooler N, Ulorich RF, Mussare F, Ferro P, tients: a multicentre double-blind comparative study. Acta Horron E (1979): Fluphenazine and social therapy in the Psychiatr Scand 85:295–305.
aftercare of schizophrenic patients: relapse analysis of two- Conley RR, Taminga CA, Bartko JJ, et al (1998): Olanzapine year controlled study of fluphenazine decanoate and fluphen- compared with clopromazine in treatment-resistant schizo- azine hydrochloride. Arch Gen Psychiatry 36:1283–1294.
phrenia. Am J Psychiatry 155:914 –920.
Janicak PG, Davis JM, Preskorn SH, Ayd FJ Jr (1993): Princi- Davis JM (1975): Overview: maintenance therapy in psychiatry.
ples and Practice of Psychopharmacotherapy. Baltimore: I. Schizophrenia. Am J Psychiatry 132:1237–1245.
Williams & Wilkins, pp 104 –106.
Davis JM, Matalon L, Watanabe MD, Blake L (1994): Depot Johnson DA, Ludlow JM, Street K, Taylor RD (1987): Double- antipsychotic drugs: place in therapy. Drugs 47:741–773 blind comparison of half-dose and standard-dose flupenthixol [Erratum, Drugs 1994;48:616].
decanoate in the maintenance treatment of stabilized out- Erlenmeyer-Kimling L, Cornblatt BA, Rock D, et al (1993): The patients with schizophrenia. Br J Psychiatry 151:634 – 638.
New York High-Risk Project: Anhedonia, attentional devi- Jolley AG, Hirsch SR, McRink A, Manchanda R (1989): Trial of ants, and psychopathology. Schizophr Bull 19:141–153.
brief intermittent neuroleptic prophylaxis for selected schizo- phrenic outpatients: clinical outcome at one year. Br Med J Malhotra AK, Goldman D, Ozaki N, Breier A, Buchanan R, Pickar D (1996): Lack of association between poly-mor- Jolley AG, Hirsch SR, Morrison E, McRink A, Wilson L (1990): phisms in the 5-HT 2A receptor gene and the antipsychotic Trial of brief intermittent neuroleptic prophylaxis for selected response to clozapine. Am J Psychiatry 153:1092–1094.
schizophrenic outpatients: clinical and social outcome at two Mallya AR, Roos PD, Roebuck-Colgan K (1992): Restraint, seclusion and clozapine. J Clin Psychiatry 53:395–397.
Kane JM (1985): Compliance issues in outpatient treatment.
Marder SR, Meibach RC (1994): Risperidone in the treatment of J Clin Psychopharmacol 5:22S–27S.
schizophrenia. Am J Psychiatry 151:825– 835.
Kane JM (1995): Tardive dyskinesia: epidemiological and clin- Marder SR, Van Putten T, Mintz J, et al (1984): Costs and ical presentation. In: Bloom FE, Kupfer DJ, editors. Psycho- benefits of two doses of fluphenazine. Arch Gen Psychiatry pharmacology: The Fourth Generation of Progress. New Marder SR, Van Putten T, Mintz J, Lebell M, McKenzie J, May Kane JM (1996a): Who shall decide when doctors disagree? PR (1987): Low-and conventional-dose maintenance therapy Am J Psychiatry 153:1507–1508.
with fluphenazine decanoate: two-year outcome. Arch Gen Kane JM (1996b): Schizophrenia. N Engl J Med 334:34 – 41.
Kane JM, Borenstein M (1985): Compliance in the long-term Martin J, Gomez J-C, Garcia-Bernardo E, Cuesta M, Alvarez E, treatment of schizophrenia. Psychopharmacol Bull 21:23–27.
Gurpegui M (1997): Olanzapine and treatment-refractoryschizophrenia: results of an open-label study. J Clin Psychi- Kane JM, Davis JM, Schooler NR, Marder SR, Brauzer B, Casey DE (1993a): A one-year comparison of four dosages ofhaloperidol decanoate [abstract]. Schizophr Res 9:239 –240.
Meltzer HY (1995): Clozapine: is another view valid? Am J Kane JM, Honigfeld G, Singer J, Meltzer H and the Clozaril Collaborative Study Group (1988): Clozapine for the treat- Meltzer HY, Okayli O (1995): Reduction of suicidality during ment-resistant schizophrenic: a double blind comparison with clozapine treatment of neuroleptic-resistant schizophrenia: clopromazine. Arch Gen Psychiatry 45:789 –796.
impact on risk– benefit assessment. Am J Psychiatry 152:183–190.
Kane JM, Lieberman JA (1987): Maintenance pharmacotherapy in schizophrenia. In: Meltzer HY, editor. Psychopharmacol- Muller-Spahn F (1992): Risperidone in the treatment of chronic ogy: The Third Generation of Progress. New York: Raven schizophrenic patients: an international double-blind parallel- group study versus haloperidol. Clin Neuropharmacol15(suppl 1):90A–91A.
Kane JM, Rifkin A, Quitkin F, Klein DF (1976): Antipsychotic drug blood levels and clinical outcome. In: Klein DF, Gittel- Peuskens J, Link CGG (1997): A comparison of quetiapine and man-Klein R, editors. Progress in Psychiatric Drug Treat- clopromazine in the treatment of schizophrenia. Acta Psychi- ment. New York: Brunner Mazel, pp 399 – 408.
Kane JM, Rifkin A, Woerner MG, et al (1983): Low-dose Pickar D, Owen RR, Litman RE, Konicki E, Gutierrez R, neuroleptic treatment of outpatient schizophrenia. I. Prelimi- Rappoport MH (1992): Clinical and biologic response to nary results for relapse rates. Arch Gen Psychiatry 40:893– clozapine in patients with schizophrenia: crossover compari- son with fluphenazine. Arch Gen Psychiatry 49:345–353.
Kane JM, Smith J (1982): Tardive dyskinesia: prevalence and Pietzcker A, Gaebel W, Kopcke W, et al (1993): Intermittent risk factors 1959 –1979. Arch Gen Psychiatry 39:473– 481.
versus maintenance neuroleptic long-term treatment in Kane JM, Woerner MG, Pollack S, Safferman AZ, Lieberman JA schizophrenia—2-year results of German multicenter study.
(1993b): Does clozapine cause tardive dyskinesia? J Clin J Psychiatr Res 27:321–339.
Pollack S, Woerner MG, Howard A, Fireworker RB, Kane JM Kane JM, Woerner MG, Sarantakos S (1986): Depot neurolep- (1998): Clozapine reduces rehospitalization among schizo- tics: a comparative review of standard, intermediate, and phrenia patients. Psychopharmacol Bull 34:89 –92.
low-dose regimens. J Clin Psychiatry 47(suppl):30 –33.
Prien RF, Cole JO (1968): High dose of chlorpromazine in Kinon BJ, Kane JM, Johns C, Perovich R, Ismi M, Koreen A chronic schizophrenia. Arch Gen Psychiatry 18:482– 495.
(1993): Treatment of neuroleptic resistant schizophrenic re- Quitkin F, Rifkin A, Klein D (1975): Very high dosage vs.
lapse. Psychopharmacol Bull 29:309 –314.
standard dosage fluphenazine in schizophrenia. A double- Kissling W, editor (1991): Guidelines for Neuroleptic Relapse blind study of non-chronic treatment-refractory patients. Arch Prevention in Schizophrenia. Berlin, Germany: Springer- Gen Psychiatry 32:1276 –1281.
Regier DA, Farmer ME, Rae DS, et al (1990): Co-morbidity of Klein DF, Davis JM (1969): Diagnosis and Drug Treatment of mental disorders with alcohol and other drug abuse. JAMA Psychiatric Disorders. Baltimore: Williams & Wilkins.
Klieser E, Lehman E, Kinzler E, et al (1995): Randomized, Remington G, Pollock B, Voineskos G, Reed K, Coulter K double-blind controlled trial of risperidone versus clozapine (1993): Acutely psychotic patients receiving high-dose halo- in patients with chronic schizophrenia. J Clin Psychophar- peridol therapy. J Clin Psychopharmacol 13:41– 45.
Rifkin A, Doddi S, Karajgi B, Borenstein M, Wachspress M Levinson DF, Simpson GM, Singh H, et al (1990): Fluphenazine (1991): Dosage of haloperidol for schizophrenia. Arch Gen dose, clinical response, and extrapyramidal symptoms during acute treatment. Arch Gen Psychiatry 47:761–768.
Rosenheck R, Cramer J, Xu W, et al (1997): A comparison of clozapine and haloperidol in hospitalized patients with refrac- Tollefson GD, Sanger TM, Lou Y, Thieme ME (1998): Depres- tory schizophrenia. N Engl J Med 337:809 – 815.
sive signs and symptoms in schizophrenia: a trial of olanza- Schooler N (1998, May): Comparing new anti-psychotic medi- pine and haloperidol. Arch Gen Pyschiatry 55:250 –258.
cations: what do the data say. Presented at the Society of Tran PV, Hamilton SH, Kuntz AJ, et al (1997): Double-blind Biological Psychiatry Annual Meeting, Toronto Canada.
comparison of olanzapine versus risperidone in the treatment Schooler NR, Keith SJ, Severe JB, et al (1997): Relapse and of schizophrenia and other psychotic disorders. J Clin Psy- reshospitalization during maintenance treatment of schizo- phrenia. Arch Gen Psychiatry 54:453– 463.
Van Putten T, Marder SR, Mintz J (1990): A controlled dose Shalev A, Hermash H, Rothberg J, Munitz H (1993): Poor comparison of haloperidol in newly admitted schizophrenic neuroleptic response in acutely exacerbated schizophrenic patients. Arch Gen Psychiatry 47:754 –758.
patients. Acta Psychiatr Scand 87:86 –91.
Viguera AC, Baldessarini RJ, Hegarty JD, et al (1997): Clinical risk following abrupt and gradual withdrawal of maintenance Shaner A, Khalsa MA, Roberts L, Wilkins J, Anglin D, Hsieh SC neuroleptic treatment. Arch Gen Psychiatry 54:49 –55.
(1993): Unrecognized cocaine use among schizophrenic pa-tients. Am J Psychiatry 150:758 –762.
Volavka J, Cooper T, Czobor P, et al (1992): Haloperidol blood levels and clinical effects. Arch Gen Psychiatry 49:354 –361.
Tollefson GD, Beasley CM Jr, Tamura RN, Tran PV, Potvin JH (1997a): Blind, controlled, long-term study of the compara- Wahlbeck K (1998, February 7–13): Evidence of clozapine tive incidence of treatment-emergent tardive dyskinesia with effectiveness in schizophrenia. Presented at the Ninth Bien- olanzapine or haloperidol. Am J Psychiatry 154:1248 –1254.
nial Winter Workshop on Schizophrenia, Davos Switzerland.
Tollefson GD, Beasley CM, Tran P, Street JS, Kruger JA, Weiden PJ, Olfson M (1995): Cost of relapse in schizophrenia.
Tamura RN (1997b): Olanzapine versus haloperidol and the Schizophr Bull 21:419 – 429.
treatment of schizophrenia and schizoaffective and schizo- Zimbroff DL, Kane JM, Tamminga CA, et al (1997): A con- phreniform disorders: results of an international collaborative trolled, dose–response study of sertindole and haloperidol in trial. Am J Psychiatry 154:457– 465.
schizophrenia. Am J Psychiatry 154:782–791.

Source: http://www.addiction.umd.edu/Psyc434/Kane%201999.pdf