The article you requested is

A Randomized Controlled Trial of Risperidone and Olanzapine for Schizophrenic Patients With Neuroleptic-Induced Tardive Dyskinesia

J Clin Psychiatry 2010;71(9):1226-1233
10.4088/JCP.09m05155yel

Objective: To compare the efficacy of risperidone and olanzapine in schizophrenic patients with tardive dyskinesia on treatment with first-generation antipsychotics.

Method: We conducted a 24-week, rater-blinded, flexible-dose study. Sixty patients with DSM-IV schizophrenia (n=58) or schizoaffective disorder (n=2) met the DSM-IV research criteria for neuroleptic-induced tardive dyskinesia and were randomly assigned to a risperidone or olanzapine group. The primary outcome was a comparison of the change in the total scores on the Abnormal Involuntary Movement Scale (AIMS) from baseline to study end point between the groups. The study was conducted from July 2000 to June 2004.

Results: The mean±SD doses of risperidone and olanzapine from baseline to study end point were 1.9±0.7 to 4.1±1.4 mg/d and 8.1±2.0 to 12.6±5.4 mg/d, respectively. There were no statistically significant differences in demographic data, severity of tardive dyskinesia, or psychotic symptoms between risperidone and olanzapine groups at baseline assessment. Both groups showed significant improvement in mean±SD AIMS total scores (risperidone: −7.4±6.9, P<.0001; olanzapine: −6.2±8.0, P=.0002). However, there was a more statistically significant change in the slope of AIMS total scores in the risperidone group than in the olanzapine group (P=.0001).

Conclusions: Our findings demonstrated that olanzapine may not have better potential for tardive dyskinesia improvement than risperidone did. Double-blinded, fixed dose studies with a larger sample size on schizophrenic patients with tardive dyskinesia from different ethnic groups are needed to confirm the results of our study.

Trial Registration: clinicaltrials.gov Identifier NCT00621998

J Clin Psychiatry

Submitted: February 20, 2009; accepted April 21, 2009.

Online ahead of print: April 20, 2010 (doi:10.4088/JCP.09m05155yel).

Corresponding author: Mei-Shu Lai, MD, PhD, Institute of Preventive Medicine, College of Public Health, National Taiwan University, Room 518, No. 17, Hsu-Chow Rd, Taipei, Taiwan 100 (mslai@ntu.edu.tw).

A Randomized Controlled Trial of Risperidone and Olanzapine for Schizophrenic Patients With Neuroleptic-Induced Tardive Dyskinesia

Objective: To compare the efficacy of risperidone and olanzapine in schizophrenic patients with tardive dyskinesia on treatment with first-generation antipsychotics.

Method: We conducted a 24-week, rater-blinded, flexible-dose study. Sixty patients with DSM-IV schizophrenia (n=58) or schizoaffective disorder (n=2) met the DSM-IV research criteria for neuroleptic-induced tardive dyskinesia and were randomly assigned to a risperidone or olanzapine group. The primary outcome was a comparison of the change in the total scores on the Abnormal Involuntary Movement Scale (AIMS) from baseline to study end point between the groups. The study was conducted from July 2000 to June 2004.

Results: The mean±SD doses of risperidone and olanzapine from baseline to study end point were 1.9±0.7 to 4.1±1.4 mg/d and 8.1±2.0 to 12.6±5.4 mg/d, respectively. There were no statistically significant differences in demographic data, severity of tardive dyskinesia, or psychotic symptoms between risperidone and olanzapine groups at baseline assessment. Both groups showed significant improvement in mean±SD AIMS total scores (risperidone: −7.4±6.9, P<.0001; olanzapine: −6.2±8.0, P=.0002). However, there was a more statistically significant change in the slope of AIMS total scores in the risperidone group than in the olanzapine group (P=.0001).

Conclusions: Our findings demonstrated that olanzapine may not have better potential for tardive dyskinesia improvement than risperidone did. Double-blinded, fixed dose studies with a larger sample size on schizophrenic patients with tardive dyskinesia from different ethnic groups are needed to confirm the results of our study.

Trial Registration: clinicaltrials.gov Identifier NCT00621998

J Clin Psychiatry 2010;71(9):1226–1233

Submitted: February 20, 2009; accepted April 21, 2009.

Online ahead of print: April 20, 2010 (doi:10.4088/JCP.09m05155yel).

Corresponding author: Mei-Shu Lai, MD, PhD, Institute of Preventive Medicine, College of Public Health, National Taiwan University, Room 518, No. 17, Hsu-Chow Rd, Taipei, Taiwan 100 (mslai@ntu.edu.tw).

Many patients with schizophrenia experience tardive dyskinesia after the long-term treatment of first-generation antipsychotics (FGAs). The tardive dyskinesia incidence rate may be as high as 5% per year, and, in previous studies, the prevalence was about 20%.1,2 The risk factors for neuroleptic-induced tardive dyskinesia include old age, female gender, histories of neuroleptic-induced extrapyramidal syndrome (EPS), diabetes, mood disorder, and doses and duration of total neuroleptic exposure.3 Tardive dyskinesia might be irreversible and could cause personal embarrassment and emotional distress, and tardive dyskinesia could impact one’s physical, social, and occupational functions.4 Tardive dyskinesia may also decrease adherence to antipsychotics among schizophrenia patients, thus leading to relapse and rehospitalization.5,6

Some studies7,8 showed that second-generation antipsychotics (SGAs) might have a lower liability for EPS, better drug adherence, and a lower relapse rate than FGAs. Therefore, SGAs are considered a good choice for patients who had FGA-induced tardive dyskinesia. SGAs have some advantages in treating schizophrenic patients, but many countries cannot use them as first-line medications due to higher cost.9 The National Health Insurance Bureau of Taiwan had restrictive regulations for the use of SGAs before 2003. In recent years, although the government has loosened regulations on the use of SGAs, many psychiatrists in Taiwan still need to prescribe FGAs as the first-line medications due to the limitations of global budget policy for hospitals.10

Different SGAs have different pharmacologic mechanisms and were classified as serotonin-dopamine antagonists, multiacting receptor targeted agents, selective D2 and D3 dopamine receptor antagonist, and dopamine receptor partial agonist11; however, little is known about which class of SGA is superior over the other classes of SGAs for those schizophrenic patients with tardive dyskinesia. Although several studies have demonstrated the effectiveness of clozapine in improving tardive dyskinesia, many severe clozapine-induced adverse effects have prevented it from being a suitable first-line antipsychotic.12,13 There were few randomized controlled studies of other SGAs focusing on the treatment of patients with neuroleptic-induced tardive dyskinesia. Chouinard14 analyzed the results of a multicenter study on fixed-dose risperidone in Canada and found that 6 mg/d of risperidone was more effective than haloperidol and placebo in treating tardive dyskinesia. A double-blind, placebo-controlled study15 of risperidone showed that the group receiving risperidone 6 mg/d compared to a placebo group had a significant improvement in total Abnormal Involuntary Movement Scale (AIMS) scores. A single-blind comparison study on the usages of quetiapine and haloperidol in schizophrenia patients with tardive dyskinesia showed that quetiapine 400 mg/d could more significantly improve tardive dyskinesia than haloperidol 8.5 mg/d.16 A single-arm, blinded, randomized dose-reduction study demonstrated that olanzapine improved preexisting tardive dyskinesia during 8-month treatment.17 However, there were limitations to this study. First, there was no control group. Second, like other observational studies on the effectiveness of olanzapine in treating tardive dyskinesia, the results may be influenced by selection or measurement bias.18,19 There are only observational studies or case reports on tardive dyskinesia patients treated with other SGAs.20–26 To our knowledge, there has been no published head-to-head studies comparing different SGAs in patients with tardive dyskinesia.

Risperidone and olanzapine might have different EPS severity, efficacy on psychotic symptoms, and other domains due to different receptor profiles.27 Positron emission tomography studies showed risperidone had tighter dopamine D2 receptor binding than olanzapine.28,29 Furthermore, olanzapine can selectively act on dopamine mesolimbic pathway and is less likely to act on nigrostriatal pathway.30 Besides, risperidone and olanzapine had different strength in 5-HT1A, 5-HT2A, 5-HT2C, α-adrenergic, histamine, and muscarinic receptors binding.31 Hence, it is expected that risperidone and olanzapine might have different effectiveness in schizophrenia with tardive dyskinesia populations. Our hypothesis is that olanzapine is more likely than risperidone to improve tardive dyskinesia in schizophrenic patients because of less dopamine D2 receptors blockade and less EPS potential of olanzapine than risperidone.29 We thus initiated a randomized, head-to-head, controlled study to compare the efficacy of risperidone and olanzapine in Han-Chinese schizophrenic patients with FGAs-induced tardive dyskinesia.

METHOD

The study protocol was approved by the Research Ethics Committee of a public mental hospital in North Taiwan before implementation. We conducted the study at this public mental hospital from July 2000 to June 2004 in accordance with the principles of the Helsinki Declaration and Good Clinical Practice. All patients or legal guardians gave their written informed consent after the study purpose and procedures had been fully explained to them, before their participation in the study.

Inclusion and Exclusion Criteria

Inclusion criteria included patients who were aged 18–70 years; were female and did not have plans to become pregnant and who agreed to use reliable contraception methods if at childbearing age; met the DSM-IV criteria for schizophrenia, schizophreniform or schizoaffective disorder; fulfilled DSM-IV neuroleptic-induced tardive dyskinesia research criteria, with a severity of tardive dyskinesia no less than moderate degree (≥4) as assessed by the global impression of Extrapyramidal Syndrome Rating Scale (ESRS) (item 42 of the ESRS)32; and agreed or had legal guardians who agreed to join the study and signed an informed consent. Exclusion criteria included patients with other Axis I diagnoses in the DSM-IV; major systemic diseases in unstable condition; a history of neurologic disorders that would influence the tardive dyskinesia assessment; and substance abuse or dependence other than coffee or tobacco in the last 6 months before the study. All eligible patients received a psychiatric evaluation to decide whether they met the inclusion or exclusion criteria at the screening visit.

Clinical Assessments

The Brief Psychiatric Rating Scale (BPRS [18 items: score, 1–7; range, 7–126]),33 and the Clinical Global Impression-Severity (CGI-S [range, 1–7])34 were used to assess the severity of psychotic symptoms, the global impression of the ESRS (4 items: dyskinesia, parkinsonism, dystonia, and akathisia; score, 0–8 for each item) for EPS, and the AIMS (10 items: score, 0–4; range, 0–40) for tardive dyskinesia.35 ESRS global impression can conveniently assess global impression in other EPS, such as parkinsonism, dystonia, and akathisia, and supplementary to AIMS rating. Two investigators (C.-H.C. and J.-J.C.) served as blinded raters. They received video training for using the AIMS and ESRS and performed interrater reliability through joint assessment of 10 patients with tardive dyskinesia and EPS before the study. The criterion of agreement is within±1 point of expert ratings as the manual for the ESRS.32 The coefficients of agreement for the AIMS and global impression of ESRS were 0.91 and 0.75 among the 10 patients, respectively.

The primary outcome of this study was to compare the mean changes in the AIMS total scores from baseline to study end point between the risperidone and olanzapine groups. Secondary outcomes were to compare the differences in the changes of the ESRS, BPRS, and CGI-S scores between these 2 groups. On the basis of previous studies’ results, a 50% or more decrease in AIMS total scores from baseline to study end point was set a priori as treatment responder.13,16

Study Design

Eligible patients fulfilling inclusion and exclusion criteria entered a washout period of 3 to 7 days prior to randomization. The patients stopped using any antipsychotics, mood stabilizers, antidepressants, and dopamine agonists during the washout period and throughout the entire study period. Screening would take place after 1 injection cycle had passed since the last injection if the patients had received depot medications prior to the study. Anticholinergic drugs, propranolol and daytime benzodiazepines were allowed during the washout period but were discontinued when the study period started. If patients had acute dystonia, parkinsonism, or akathisia at later study period, anticholinergic drugs, propranolol and daytime benzodiazepines could be prescribed under clinical decision. Patients having at least moderate degree of tardive dyskinesia at baseline visit (item 42 of ESRS ≥4) were then randomly assigned to receive either olanzapine or risperidone with a 1-to-1 ratio by coin method with a 6-block design. The study period lasted 24 weeks. The BPRS, CGI-S, AIMS and global impression of ESRS were performed at baseline and at weeks 1, 2, 3, 4, 8, 12, 16, 20, and 24 or at end point visit by blinded-rater. Information on adverse events, body weight, vital signs, concomitant medications, and the dosage of study medications were recorded by the attending physicians. No more than 8 mg/d of intramuscular lorazepam was given to the patients who had agitated or exhibited violent behavior.

Dosing Strategy

This study was a flexible-dose design; the dose ranges for risperidone and olanzapine were 0.5–6.0 mg/d and 2.5–20.0 mg/d, respectively. The dose range was determined based on the suggested optimal dose for general schizophrenic patients in Taiwan (risperidone 2–6 mg/d, olanzapine 10–20 mg/d).36 Once patients appear to have tardive dyskinesia, American Psychiatric Association schizophrenia treatment guideline suggest tapering dose of FGAs to alleviate its severity.37 Hence, we hypothesized that patients with tardive dyskinesia may need lower doses and decided to use a dose range that was one-quarter lower than the dose range for general schizophrenic patients as the lower dose range of our study. The doses were adjusted according to the clinical decisions of the nonblinded investigators. Drug compliance was assessed by counting the medications left. Study protocol was violated if a patient’s medication uptake was below 80% or above 120%, and the patient was required to withdraw from the study early.

Statistical Methods

All patients who were randomly assigned and had at least 1 postbaseline assessment were included in the intent-to-treat (ITT) analysis. If the ITT subjects withdrew from the study earlier than scheduled, then the last observation carried forward method was employed to extend the end point scores. For between-group analysis, χ2 or Fisher exact test and independent t test were performed for categorical and continuous data, respectively. The linear mixed model was used for continuous variables that were repeatedly measured. All results were expressed as means and standard deviations. The level of significant difference was 1-sided, P<.05 for the AIMS total scores change comparison due to superiority hypothesis testing. The other statistics were 2-sided, P<.05. The data were analyzed using the Chinese version of SAS 9.1 (SAS Institute Inc, Cary, North Carolina).

Sample Size Calculation

Our hypothesis is that olanzapine is more effective in tardive dyskinesia improvement than risperidone, due to lower dopamine D2 receptors affinity and less hypersensitivity mechanism appeared. The sample size was calculated based on 2 previous risperidone studies,15,38 which showed that risperidone decreased on AIMS total scores in tardive dyskinesia patients by 5 in the ITT population, and we had a case report39 which showed that after 6 months of olanzapine treatment, AIMS total scores decreased by 12 in a tardive dyskinesia patient. We hypothesized that subjects who terminate early from this study would be followed up for 3 months on average, about half of the study period, and that one-third of subjects in the olanzapine group would drop out, resulting in AIMS total scores decreasing by 6, only half as much as those in patients who completed the study. Then the AIMS total scores in the ITT population of olanzapine group decreased by 10 on average after the combination of one-third dropouts and two-thirds patients who completed the study ([1/3×6]+[2/3×12]=10). Under the conditions of α=.05 1-sided, 80% power, and a standard deviation of 7.5, we needed at least a total of 58 ITT subjects.

RESULTS

Patient Disposition and Demographics

Figure 1

Click figure to enlarge

Table 1

Click figure to enlarge

Eighty-one patients were screened and entered a 3- to 7-day washout period. Twenty-one patients were excluded before randomization with the following reasons: misclassification as tardive dyskinesia (n=8), less than moderate tardive dyskinesia severity after 7-day washout period (n=5), other criteria not met (n=4), patient withdrew consent (n=2), and worsening of psychotic symptoms (n=2). The remaining 60 patients were allocated by randomization, with 30 patients in each group. All 60 patients received study medications and had at least 1 postbaseline assessment. Nine and 7 patients left the study early in the risperidone and olanzapine groups, respectively. The participant flowchart and the reasons for early termination are shown in Figure 1.

There were 58 patients with schizophrenia and 2 patients with schizoaffective disorder. The study population comprised chronic schizophrenic patients whose mean age was about 40 years and exhibited moderate to severe psychotic symptoms. The mean duration of neuroleptic exposure was about 10 years. After the washout period, AIMS total scores from screening visit to baseline visit was 14.9±2.2 to 17.2±4.6 in the risperidone group and from 17.7±6.0 to 19.4±6.1 in the olanzapine group. All of the subjects received FGAs prior to participation in this study. The FGAs (the risperidone group, the olanzapine group) used prior to the study were as follows: sulpiride (n=10, n=11), haloperidol (n=5, n=7), zotepine (n=3, n=4), chlorpromazine (n=3, n=2), thioridazine (n=3, n=1), flupenthixol (n=2, n=2), pipotiazine (n=2, n=0), clopenthixol (n=1, n=3) and loxapine (n=1, n=0). There were no significant differences between the 2 treatment groups in the distribution of prior FGAs (χ2 or Fisher exact P values ranging from 0.492 to 1). Table 1 illustrates the baseline characteristics of the study population. There were no statistically significant between-group differences in age, sex, duration of illness of schizophrenia, duration of antipsychotic treatment, duration of tardive dyskinesia, AIMS total scores, global impression of dyskinesia, parkinsonism, dystonia, and akathisia on ESRS, BPRS total scores, CGI-S, and body weight at the baseline visit. There were no between-group differences in prescreening chlorpromazine equivalent dose and the lengths of washout period.

Dose of Study Medications

The dosages used in the risperidone and the olanzapine groups were gradually increased from baseline to end point. For the risperidone group, the mean doses were 1.9±0.7 (baseline), 2.5±1.3 (first week), 3.1±1.3 (second week), 3.4±1.5 (third week), 3.8±1.6 (fourth week), 4.2±1.4 (eighth week), 4.2±1.4 (12th week), 4.3±1.4 (16th week), 4.1±1.5 (20th week), and 4.1±1.4 (24th week) mg/d. For the olanzapine group, the mean doses were 8.1±2.0 (baseline), 10.0±3.9 (first week), 11.9±4.5 (second week), 12.9±5.2 (third week), 12.5±5.3 (fourth week), 13.1±4.3 (eighth week), 13.1±4.4 (12th week), 12.4±5.2 (16th week), 12.6±4.7 (20th week) and 12.6±5.4 (24th week) mg/d.

Concomitant Medications

The overall use of concomitant anticholinergic drugs during the study was significantly higher in the risperidone group than in the olanzapine group (n/n=8/30 vs n/n=2/30; OR=5.1; 95% CI, 1.0–26.4; P=.04). Although there was a higher percentage of concomitant propranolol usage to treat akathisia in the risperidone group during the study, the between-group difference was not statistically significant (n/n=3/30 vs n/n=1/30, P=.6). The incidences of concomitant daytime benzodiazepines to treat akathisia in these 2 groups were the same (n/n=2/30 vs n/n=2/30). One case in the risperidone group and 2 cases in the olanzapine group needed to use intramuscular lorazepam to manage agitated symptoms.

EPS and Psychotic Symptoms Assessments

Paired t test revealed significant reduction in AIMS total scores (risperidone, P<.0001; olanzapine, P=.0002) and global impression of ESRS dyskinesia (risperidone, P=.002; olanzapine, P<.001) from baseline to study end point in both groups; yet, significant reduction in the global impression of parkinsonism (P=.01) and akathisia (P<.05) was noted only in the olanzapine group. There was significant reduction in CGI-S in the risperidone group only (P=.02). There were no significant reduction in both groups in terms of the BPRS total scores and global impression of dystonia.

Independent t test showed that there was no significant difference in AIMS total score changes from baseline to study end points between the 2 treatment groups. The proportions of patients meeting response criteria were also not significantly different (OR=1.75; 95% CI, 0.62–4.97; P=.29) between risperidone (14/30, 47.7%) and olanzapine (n/n=10/30, 33.3%). Except global impression of parkinsonism and dystonia of ESRS, the changes in all other rating scale assessments (global impression of dyskinesia and akathisia of ESRS, CGI-S, and BPRS) showed no significant differences between groups (Table 2).

Table 2

Click figure to enlarge

Figure 2

Click figure to enlarge

Table 3

Click figure to enlarge

Figure 2 illustrates the AIMS total scores of each visit. An unstructured, covariance-matrix mixed model was used for repeated measurements, which tested the changes in the slopes of AIMS total scores and global impression of ESRS dyskinesia. The risperidone group showed more significant improvement in AIMS total scores (F1,538=14.74, P=.0001) and global impression of ESRS dyskinesia (F1,538=13.47, P=.0003) than the olanzapine group. However, the olanzapine group showed more statistically significant decrease than the risperidone groups in global impression of parkinsonism of ESRS (F1,538=6.4, P=.01). There were no significant slope differences between groups in the other assessments.

Safety Results

Table 3 describes adverse events among patients. There were no significant between-group differences in adverse events. The common adverse events (≥10%) in the risperidone group were headache, nausea, dizziness, drowsiness, weakness, palpitation, and constipation. As for the olanzapine group, the common adverse events included headache, dizziness, thirst, drowsiness, weakness, and muscle ache. For those who completed the entire study, the mean±SD gains in body weight were 4.9±7.1 kg and 4.6±6.0 kg for risperidone and olanzapine groups, respectively. There were 42.9% (9/21) and 47.8% (11/23) of patients in risperidone and olanzapine groups that had greater than 7% of body weight gain, respectively. The differences seen above were not statistically significant. No severe adverse events appeared during the study period.

DISCUSSION

To the best of our knowledge, this study is the first randomized controlled study to compare risperidone and olanzapine in schizophrenic patients who have had neuroleptic-induced tardive dyskinesia. This study may provide more clear guidance for drug choices for patients with tardive dyskinesia.

Both risperidone and olanzapine significantly reduced dyskinesia, which was demonstrated in the changes in the AIMS total scores and global impression of ESRS dyskinesia after 24-week treatment. Contrary to our hypothesis, olanzapine did not have better efficacy in decreasing tardive dyskinesia severity than risperidone did. Instead, risperidone might be more effective than olanzapine in alleviating tardive dyskinesia. In our study, the high EPS liability of risperidone reflected a higher incidence of anticholinergic drug use than the olanzapine group and a slight increase in global impression of parkinsonism and dystonia from baseline to study end point. Fewer patients (6.7%) in the olanzapine group needed anticholinergic drugs, and they had slightly decreased scores on global impression of parkinsonism, dystonia, and akathisia. The differential effects of risperidone and olanzapine on EPS in this susceptible population seemed comparable to those general schizophrenic patients. The above observations were consistent with the fact that risperidone has stronger D2 receptors antagonist effects and higher EPS side effects than olanzapine.40,41

There are significant between-group differences in the use of anticholinergic drugs to manage EPS. The effects of anticholinergic medications and cholinergic medications on tardive dyskinesia are still controversial. Some studies42,43 have shown that anticholinergic medications might worsen tardive dyskinesia and cholinergic medications might improve tardive dyskinesia, but Cochrane systematic reviews44,45 did not draw a definite conclusion and suggested further studies with sound study design to disclose this issue. Another point is that olanzapine had higher anticholinergic action than risperidone. Both are potential confounders of the treatment outcome. The influence of concomitant anticholinergic drugs in patients with tardive dyskinesia in the context of treatment with risperidone and olanzapine warrants further investigation.

In our study, however, there were no between-group differences in the change of tardive dyskinesia severity from baseline to study end point. Two possible explanations are that risperidone also had high 5-HT2A receptors antagonist effects, which may increase endogenous dopamine release,46 to prevent the long-term complication of D2 receptors blockade and that tardive dyskinesia is attributable to many etiologies so more than 1 mechanism can explain the results, such as dopamine receptors hypersensitivity.47

Some studies48–50 have revealed that early parkinsonism or akathisia side effects of antipsychotics may predict future occurrences of tardive dyskinesia. The risperidone group did not show exacerbation of tardive dyskinesia at a later period in our study. This result is consistent with some previous studies.38,51,52 But we still need to consider that the improvement of tardive dyskinesia might be due to the masked effect of risperidone and the deterioration of tardive dyskinesia after long-term treatment. Meanwhile, the AIMS total scores in the olanzapine group gradually increased from the 16th week to 24th week. There were also some case reports of olanzapine-induced or worsening tardive dyskinesia.53,54 Previous studies55 of tardive dyskinesia have shown that FGAs can decrease scores of dyskinesia. The symptomatic improvement may be due to masking effects of FGAs. Long-term effects of risperidone and olanzapine in a population with tardive dyskinesia need further investigation.

The washout period in our study was between 3 and 7 days, with a mean of about 4 days. On the contrary, previous studies’ washout periods were about 2 to 4 weeks long.15,16 Prolonged washout periods had the advantages of less carryover effect from previous antipsychotics and provided more reliable diagnoses and assessments for tardive dyskinesia. However, a prolonged washout period could worsen psychotic symptoms in schizophrenia patients and discourage patients from participation the study. Two patients in our study experienced exacerbation of psychotic symptoms during 7 days of washout and withdrew from the study. We can expect that longer washout periods may put more patients at risk for exacerbation of psychotic symptoms. Shorter washout periods also have the advantage of close to real clinical practice and increase external validity of this study.

The mean dose of risperidone with flexible-dose design was lower than that in the fixed-dose design study of 6 mg/d.15 At the same time, our mean dose was slightly higher than the open-level study of Bai et al,38 which had a mean dose of 3.6 mg/d. The mean dose of olanzapine in the Kinon et al17 study was 12.1 mg/d, which was similar to our result. Further multidose comparison studies are needed to provide the evidence of optimal doses of risperidone and olanzapine for schizophrenia patients with tardive dyskinesia.

The degree of tardive dyskinesia improvement in the risperidone group was similar to but slightly higher than that in previous studies. AIMS total scores decreased by 5.5 in a 12-week study and by 6.1 in a 48-week study.15,16 Tardive dyskinesia improvement in the olanzapine group was also higher than Kinon and colleagues’17 study, whose AIMS total scores decreased by 4.5 in an 8-month period. Compared to previous studies, there were more decreases in AIMS total scores in both risperidone and olanzapine groups. This may be due to higher baseline AIMS total scores in our study population. Linear regression analysis showed that patients with higher baseline AIMS total scores had greater AIMS total score reduction (regression coefficient: −0.51, P=.003). This result is consistent with that of the Kinon et al17 study.

Tardive dyskinesia improvement in both groups was most evident in the first 4 weeks, and then stabilized during the remaining study periods. This result was compatible with some of the previous studies on risperidone showing the most significant improvement was in the first 2 to 3 months of treatment.38 Previous studies on olanzapine also showed that tardive dyskinesia improvement was most evident in the first few weeks, reaching maximum improvement in the 20th week.17 Nevertheless, the results from some studies on quetiapine and clozapine revealed that tardive dyskinesia improvement could persist up to 9 months or even 1 year.12,16 Clozapine and quetiapine had the mechanism of fast dissociation and might exhibit a different improvement profile.56 If we use more than 50% improvement in severity of tardive dyskinesia from baseline to study end point as the definition of responder, the proportion of responders in our study was similar to that in a clozapine study13 (43%) but lower than that in a quetiapine study (64% and 55% at 6 months and 1 year, respectively).16 Nevertheless, due to different tardive dyskinesia populations and study designs, it would be difficult to directly compare results from this and previous studies. Head-to-head comparison studies of different SGAs are still warranted to make a straightforward comparison and interpretation.

There were some limitations to our study. First, we did not have an FGA or placebo control group due to ethical concerns. Hence, the study result might be confounded by placebo effect and thus overestimated the effects of SGA on improving FGA-induced tardive dyskinesia. Second, our study was a rater-blinded study in which primary care physicians and patients were not blinded. Thus, the result may be influenced by measurement bias. Third, the results of our studies were limited by small sample size and inadequate power to detect the between-group differences. In addition, small sample size also weakens the effects of randomization to balance the potential confounder distribution. Fourth, our study was a 24-week study, so our results cannot be applied to patients receiving long-term treatments. Fifth, our study was a flexible-dose study, so we cannot directly infer our results to those of fixed-dose studies on risperidone and olanzapine. In addition, there may be a possibility of inadequate dose titration in some patients. Finally, our study did not perform laboratory tests, so we could not assess the patients’ blood sugar levels, triglyceride, prolactin, and other biochemical data.

In summary, although our study had above limitations, the results have some clinical implications for clinicians: (1) Although olanzapine had fewer EPS side effects than risperidone in general, olanzapine did not show superior efficacy to risperidone in a tardive dyskinesia population. (2) Tardive dyskinesia may be a heterogeneous condition and cannot be explained by a single hypothesis, such as dopamine receptor hypersensitivity. More studies focusing on tardive dyskinesia populations are recommended to help clinicians make appropriate clinical decisions.

Drug names: clozapine (FazaClo, Clozaril, and others), haloperidol (Haldol and others), lorazepam (Ativan and others), olanzapine (Zyprexa), quetiapine (Seroquel), risperidone (Risperdal and others).

Author affiliations: Taoyuan Mental Hospital, Taoyuan County (Drs Chan, Chiang, J.-J. Chen, and C.-H. Chen); National Taiwan University Hospital, Taipei (Drs Chan, Gau, and Hwu); Institute of Preventive Medicine, College of Public Health, National Taiwan University, Taipei (Drs Chan and Lai); Department of Psychology, Chung Yuan Christian University, Chung-Li (Dr Chan); Institute of Preventive Medicine, College of Public Health, National Yang-Ming University, Taipei (Dr Chiang); Cathay General Hospital, Taipei (Dr Chang); School of Medicine, Fu Jen Catholic University, Hsin-Chuang (Dr Chang), Taiwan.

Potential conflicts of interest: Drs Chan and Chang have received speech honoraria from Astellas, AstraZeneca, Eli Lilly, Janssen, Otsuka, Pfizer, and SanofiAventis. Drs Chiang, Gau, J.-J. Chen, C.-H. Chen, Hwu, and Lai report no potential conflicts of interest relevant to the subject of this article.

Funding/support: This study was supported by research grant TMH-91-02 from the Taoyuan Mental Hospital and DOH-890010 from the Department of Health, Executive Yuan, Taiwan.

Previous presentation: Part of the data in this article was presented at the 24th congress of the Collegium Internationale Neuro-psychopharmacologicum, June 20–24, 2004, Paris, France.

REFERENCES

1. Glazer WM, Morgenstern H, Doucette JT. The prediction of chronic persistent versus intermittent tardive dyskinesia: a retrospective follow-up study. Br J Psychiatry. 1991;158(6):822–828. doi:10.1192/bjp.158.6.822 PubMed

2. Chakos MH, Alvir JM, Woerner MG, et al. Incidence and correlates of tardive dyskinesia in first episode of schizophrenia. Arch Gen Psychiatry. 1996;53(4):313–319. PubMed

3. Casey DE. Tardive dyskinesia and atypical antipsychotic drugs. Schizophr Res. 1999;35(suppl):S61–S66. doi:10.1016/S0920-9964(98)00160-1 PubMed

4. Krausz M, Moritz SH, Naber D, et al. Neuroleptic-induced extrapyramidal symptoms are accompanied by cognitive dysfunction in schizophrenia. Eur Psychiatry. 1999;14(2):84–88. doi:10.1016/S0924-9338(99)80722-6 PubMed

5. Barnes TR, McPhillips MA. Novel antipsychotics, extrapyramidal side effects and tardive dyskinesia. Int Clin Psychopharmacol. 1998;13(suppl 3):S49–S57. doi:10.1097/00004850-199803003-00009 PubMed

6. Collaborative Working Group on Clinical Trial Evaluations. Adverse effects of the atypical antipsychotics. Collaborative Working Group on Clinical Trial Evaluations. J Clin Psychiatry. 1998;59(suppl 12):17–22. PubMed

7. Correll CU, Leucht S, Kane JM. Lower risk for tardive dyskinesia associated with second-generation antipsychotics: a systematic review of 1-year studies. Am J Psychiatry. 2004;161(3):414–425. doi:10.1176/appi.ajp.161.3.414 PubMed

8. Correll CU, Schenk EM. Tardive dyskinesia and new antipsychotics. Curr Opin Psychiatry. 2008;21(2):151–156. doi:10.1097/YCO.0b013e3282f53132 PubMed

9. Emsley RA, Oosthuizen PP, Joubert AF, et al. Treatment of schizophrenia in low-income countries. Int J Neuropsychopharmacol. 1999;2(4):321–325. doi:10.1017/S1461145799001613 PubMed

10. Gau SS, Chung CH, Gau CS. A pharmacoeconomic analysis of atypical antipsychotics and haloperidol in first-episode schizophrenic patients in Taiwan. J Clin Psychopharmacol. 2008;28(3):271–278. doi:10.1097/JCP.0b013e3181723713 PubMed

11. Horacek J, Bubenikova-Valesova V, Kopecek M, et al. Mechanism of action of atypical antipsychotic drugs and the neurobiology of schizophrenia. CNS Drugs. 2006;20(5):389–409. doi:10.2165/00023210-200620050-00004 PubMed

12. Tamminga CA, Thaker GK, Moran M, et al. Clozapine in tardive dyskinesia: observations from human and animal model studies. J Clin Psychiatry. 1994;55(suppl B):102–106. PubMed

13. Lieberman JA, Saltz BL, Johns CA, et al. The effects of clozapine on tardive dyskinesia. Br J Psychiatry. 1991;158(4):503–510. doi:10.1192/bjp.158.4.503 PubMed

14. Chouinard G. Effects of risperidone in tardive dyskinesia: an analysis of the Canadian multicenter risperidone study. J Clin Psychopharmacol. 1995;15(suppl 1):228–230. doi:10.1097/00004714-199506000-00017

15. Bai YM, Yu SC, Lin CC. Risperidone for severe tardive dyskinesia: a 12-week randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2003;64(11):1342–1348. PubMed

16. Emsley R, Turner HJ, Schronen J, et al. A single-blind, randomized trial comparing quetiapine and haloperidol in the treatment of tardive dyskinesia. J Clin Psychiatry. 2004;65(5):696–701. PubMed

17. Kinon BJ, Jeste DV, Kollack-Walker S, et al. Olanzapine treatment for tardive dyskinesia in schizophrenia patients: a prospective clinical trial with patients randomized to blinded dose reduction periods. Prog Neuropsychopharmacol Biol Psychiatry. 2004;28(6):985–996. doi:10.1016/j.pnpbp.2004.05.016 PubMed

18. Tenback DE, van Harten PN, Slooff CJ, et al; SOHO Study Group. Effects of antipsychotic treatment on tardive dyskinesia: a 6-month evaluation of patients from the European Schizophrenia Outpatient Health Outcomes (SOHO) Study. J Clin Psychiatry. 2005;66(9):1130–1133. PubMed

19. Brar JS, Parepally H, Chalasani L, et al. The impact of olanzapine on tardive dyskinetic symptoms in a state hospital population. Ann Clin Psychiatry. 2008;20(3):139–144. PubMed

20. Lykouras L, Agelopoulos E, Tzavellas E. Improvement of tardive dyskinesia following switch from neuroleptics to olanzapine. Prog Neuropsychopharmacol Biol Psychiatry. 2002;26(4):815–817. doi:10.1016/S0278-5846(01)00303-7 PubMed

21. O’Brien J, Barber R. Marked improvement in tardive dyskinesia following treatment with olanzapine in an elderly subject. Br J Psychiatry. 1998;172(2):186. doi:10.1192/bjp.172.2.186a PubMed

22. Littrell KH, Johnson CG, Littrell S, et al. Marked reduction of tardive dyskinesia with olanzapine. Arch Gen Psychiatry. 1998;55(3):279–280. doi:10.1001/archpsyc.55.3.279 PubMed

23. Almeida OP. Olanzapine for the treatment of tardive dyskinesia. J Clin Psychiatry. 1998;59(7):380–381. PubMed

24. Sacchetti E, Valsecchi P. Quetiapine, clozapine, and olanzapine in the treatment of tardive dyskinesia induced by first-generation antipsychotics: a 124-week case report. Int Clin Psychopharmacol. 2003;18(6):357–359. doi:10.1097/00004850-200311000-00008 PubMed

25. Kantrowitz JT, Srihari VH, Tek C. Resolution of tardive dyskinesia after addition of aripiprazole to haloperidol depot. J Clin Psychopharmacol. 2007;27(5):525–526. doi:10.1097/JCP.0b013e31814f3002 PubMed

26. Peritogiannis V, Tsouli S, Zafiris S, et al. Improvement of tardive dyskinesia following amisulpride treatment. Prog Neuropsychopharmacol Biol Psychiatry. 2006;30(4):748–750. doi:10.1016/j.pnpbp.2006.01.022 PubMed

27. Seeman P. Atypical antipsychotics: mechanism of action. Can J Psychiatry. 2002;47(1):27–38. PubMed

28. Seeman P, Corbett R, Van Tol HH. Atypical neuroleptics have low affinity for dopamine D2 receptors or are selective for D4 receptors. Neuropsychopharmacology. 1997;16(2):93–110, discussion 111–135. doi:10.1016/S0893-133X(96)00187-X PubMed

29. Seeman P, Tallerico T. Antipsychotic drugs which elicit little or no parkinsonism bind more loosely than dopamine to brain D2 receptors, yet occupy high levels of these receptors. Mol Psychiatry. 1998;3(2):123–134. doi:10.1038/sj.mp.4000336 PubMed

30. Bymaster FP, Rasmussen K, Calligaro DO, et al. In vitro and in vivo biochemistry of olanzapine: a novel, atypical antipsychotic drug. J Clin Psychiatry. 1997;58(suppl 10):28–36. PubMed

31. Schotte A, Janssen PF, Gommeren W, et al. Risperidone compared with new and reference antipsychotic drugs: in vitro and in vivo receptor binding. Psychopharmacology (Berl). 1996;124(1–2):57–73. doi:10.1007/BF02245606 PubMed

32. Chouinard G, Margolese HC. Manual for the Extrapyramidal Symptom Rating Scale (ESRS). Schizophr Res. 2005;76(2–3):247–265. doi:10.1016/j.schres.2005.02.013 PubMed

33. Bech P, Larsen JK, Andersen J. The BPRS: psychometric developments. Psychopharmacol Bull. 1988;24(1):118–121. PubMed

34. Clinical Global Impression (CGI). In: Guy W, ed. ECDEU Assessment Manual for Psychopharmacology US Department of Health, Education and Welfare publication (ADM) 76-338. Rockville, MD: National Institute of Mental Health; 1976;218–222.

35. Munetz MR, Benjamin S. How to examine patients using the Abnormal Involuntary Movement Scale. Hosp Community Psychiatry. 1988;39(11):1172–1177. PubMed

36. Liu CY, Chiu NY, Wu CK, et al. Optimal dose of risperidone and olanzapine for patients with schizophrenia in Taiwan. Int Clin Psychopharmacol. 2003;18(1):49–51. doi:10.1097/00004850-200301000-00008 PubMed

37. Lehman AF, Lieberman JA, Dixon LB, et al. Practice guideline for the treatment of patients with schizophrenia, second edition. Am J Psychiatry. 2004;161(suppl 2):1–56. PubMed

38. Bai YM, Yu SC, Chen JY, et al. Risperidone for pre-existing severe tardive dyskinesia: a 48-week prospective follow-up study. Int Clin Psychopharmacol. 2005;20(2):79–85. doi:10.1097/00004850-200503000-00003 PubMed

39. Chen JJ, Chan HY, Sun HJ, et al. Response to olanzapine for a patient with tardive dyskinesia. Taiwanese J Psychiatry. 2002;16:163–168.

40. Chan HY, Chang CJ, Chiang SC, et al. A randomised controlled study of risperidone and olanzapine for schizophrenic patients with neuroleptic-induced acute dystonia or parkinsonism. J Psychopharmacol. 2010;24(1):91–98. doi:10.1177/0269881108096070 PubMed

41. Haro JM, Salvador-Carulla L. The SOHO (Schizophrenia Outpatient Health Outcome) study: implications for the treatment of schizophrenia. CNS Drugs. 2006;20(4):293–301. doi:10.2165/00023210-200620040-00003 PubMed

42. Greil W, Haag H, Rossnagl G, et al. Effect of anticholinergics on tardive dyskinesia: a controlled discontinuation study. Br J Psychiatry. 1984;145(3):304–310. doi:10.1192/bjp.145.3.304 PubMed

43. Bergman J, Dwolatzky T, Brettholz I, et al. Beneficial effect of donepezil in the treatment of elderly patients with tardive movement disorders. J Clin Psychiatry. 2005;66(1):107–110. PubMed

44. Soares KV, McGrath JJ. Anticholinergic medication for neuroleptic-induced tardive dyskinesia. Cochrane Database Syst Rev. 2000;(2):CD000204. PubMed

45. Tammenmaa IA, McGrath JJ, Sailas E, et al. Cholinergic medication for neuroleptic-induced tardive dyskinesia. Cochrane Database Syst Rev. 2002;(3):CD000207. PubMed

46. Kapur S, Zipursky RB, Remington G. Clinical and theoretical implications of 5-HT2 and D2 receptor occupancy of clozapine, risperidone, and olanzapine in schizophrenia. Am J Psychiatry. 1999;156(2):286–293. PubMed

47. Margolese HC, Chouinard G, Kolivakis TT, et al. Tardive dyskinesia in the era of typical and atypical antipsychotics, Pt 1: pathophysiology and mechanisms of induction. Can J Psychiatry. 2005;50(9):541–547. PubMed

48. Keepers GA, Casey DE. Use of neuroleptic-induced extrapyramidal symptoms to predict future vulnerability to side effects. Am J Psychiatry. 1991;148(1):85–89. PubMed

49. Sachdev P. Early extrapyramidal side-effects as risk factors for later tardive dyskinesia: a prospective study. Aust N Z J Psychiatry. 2004;38(6):445–449. PubMed

50. Sundram S, Lambert T, Piskulic D. Acculturation is associated with the prevalence of tardive dyskinesia and akathisia in community-treated patients with schizophrenia. Acta Psychiatr Scand. 2008;117(6):474–478. doi:10.1111/j.1600-0447.2008.01183.x PubMed

51. Jeste DV, Okamoto A, Napolitano J, et al. Low incidence of persistent tardive dyskinesia in elderly patients with dementia treated with risperidone. Am J Psychiatry. 2000;157(7):1150–1155. doi:10.1176/appi.ajp.157.7.1150 PubMed

52. Chen JY, Bai YM, Pyng LY, et al. Risperidone for tardive dyskinesia. Am J Psychiatry. 2001;158(11):1931–1932. doi:10.1176/appi.ajp.158.11.1931 PubMed

53. Bhanji NH, Margolese HC. Tardive dyskinesia associated with olanzapine in a neuroleptic-naive patient with schizophrenia. Can J Psychiatry. 2004;49(5):343. PubMed

54. Ananth J, Kenan J. Tardive dyskinesia associated with olanzapine monotherapy. J Clin Psychiatry. 1999;60(12):870. PubMed

55. Glazer WM, Hafez HM, Benarroche CL. Molindone and haloperidol in tardive dyskinesia. J Clin Psychiatry. 1985;46(8, Pt 2):4–7. PubMed

56. Kapur S, Seeman P. Does fast dissociation from the dopamine d(2) receptor explain the action of atypical antipsychotics? a new hypothesis. Am J Psychiatry. 2001;158(3):360–369. doi:10.1176/appi.ajp.158.3.360 PubMed