Olanzapine in Chinese treatment-resistant patients with schizophrenia: An open-label, prospective trial


address: Dr Po See Chen, Department of Psychiatry, College of Medicine, National Cheng Kung University, 138 Sheng Li Road, Tainan 704, Taiwan.
Email: chenps@mail.ncku.edu.tw


The role of olanzapine in treatment-resistant schizophrenia has still not been clearly resolved. In addressing this issue, the current report presents an open-label, prospective, 13 week trial with olanzapine use in Chinese schizophrenic patients who were resistant to more than two different classes of antipsychotics during a minimal 4 week treatment period for each antipsychotic drug at adequate dosage. Fifty-one inpatients were recruited after a cross-titration period and given 10–25 mg of olanzapine daily, without any concomitant antipsychotic medication. Patients were evaluated with the Brief Psychotic Rating Scale (BPRS), the Positive and Negative Symptoms Scale, the Clinical Global Impression Scale (CGI), the Abnormal Involuntary Movement Scale, the Simpson–Angus scale, and the Barnes Akathisia Scale. The olanzapine-treated patients showed significant improvement in both the positive and negative symptoms of schizophrenia by the end of the study. Overall, 20 of 51 (39.2%) responded to 10–25 mg of olanzapine per day as measured by the BPRS and CGI scores. Five patients dropped out due to the worsening of their psychotic symptoms, two patients discontinued owing to poor drug compliance, and the remaining patient complained of a lack of efficacy. Extrapyramidal side-effects were mild, and anticholinergic medications required has decreased. The present open study suggests that olanzapine may be effective and well-tolerated in Chinese treatment-resistant schizophrenic patients. Further double-blinded trials are needed to confirm this result.


Despite the proven efficacy of antipsychotic drugs for schizophrenic patients, between 5% and 25% of schizophrenic patients demonstrate little or no improvement while taking classical antipsychotics. These patients, who had been given several antipsychotics without improvement, are known as treatment-resistant schizophrenic patients.1,2 Clozapine has shown an efficacy superior to conventional antipsychotics in treatment-resistant schizophrenic patients, and is currently the treatment of choice for these patients. However, it has been limited in use by the increasing fetal risks of agranulocytosis.3

Olanzapine has been shown to be effective in treating the positive and negative symptoms of schizophrenia, with minimal extrapyramidal side-effects.4 Although olanzapine and chlorpromazine have shown similar efficacy, olanzapine-treated patients had fewer side-effects than patients treated with chlorpromazine.5 Olanzapine is a potent serotonin (5-HT2A/C, 5-HT3, 5-HT6), dopamine (D4, D3, D1, D2), muscarinic (M1-5), adrenergic α1, and histamine H1 receptor antagonist. It is considered an atypical antipsychotic agent; it has a pharmacologic profile of activity similar to that of clozapine but without the serious side-effects.6 In a European study, clozapine and olanzapine were found to be of comparable efficacy.7

Conley et al. conducted an 8 week double blind trial to study efficacy with patients who were refractory to haloperidol.5 Some other preliminary reports have suggested that substantially higher doses of olanzapine may be beneficial for some severely ill patients.8–11 Although a number of clinical trials have proved that olanzapine is effective in treating refractory schizophrenic patients in Western countries, the literature remains scant with regard to Asian countries. The aims of the present study were to conduct an evaluation of the efficacy and safety of olanzapine treatment in Chinese treatment-resistant schizophrenic patients in Taiwan and to try to find possible predictors for respondent patients.



Fifty-one patients, with a mean age of 30.95 ± 10.75 years, and who met the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV) criteria for a clinical diagnosis of schizophrenia, were enrolled from the psychiatric clinics of Chang Hua Christian Hospital and National Cheng Kung University Hospital, two teaching general hospitals in Taiwan. All patients and their legal representatives gave their written informed consent to participate in this trial. The Ethics Committee for Human Research, National Cheng Kung University Medical Center approved the study protocol. All patients were required to have a documented treatment-resistant status, as defined by the absence of clinically significant improvement using at least two different classes of antipsychotics (including clozapine) at adequate dosage (equivalent to 750 mg/day of chlorpromazine or 15 mg/day of haloperidol), during a minimum treatment period of 4 weeks, and evidence of impaired social functioning for the past 2 years.

In an ethnic comparison of haloperidol plasma levels between Taiwan Chinese schizophrenic patients and American non-Chinese patients, Taiwan Chinese schizophrenic patients had higher mean plasma levels compared to non-Chinese patients.12 The ratio of the haloperidol plasma level between Taiwan Chinese and American non-Chinese patients is approximately 0.75. In accordance with this, we defined a treatment-resistant status as equivalent to 750 mg/day chlorpromazine in the present study, just 0.75 times Kane's criteria.2,13 The antipsychotic and anticholinergic medications used before and after olanzapine therapy are shown in Table 1.

Table 1. Antipsychotic and anticholinergic medications used before and after olanzapine therapy (n = 43)
No.Antipsychotics before olanzapine therapyAnticholinergic medications
Before therapy (mg/day)After therapy (mg/day)
  • Trihexyphenidyl;

  • biperiden.

 1Risperidone10 4
 3Sulpiride 5Nil
 4Risperidone 2.5Nil
 5Risperidone 5Nil
 6Pipothiazine 5Nil
 7Risperidone 6Nil
 8Haloperidol 6Nil
10Sulpiride 8 4
11Sulpiride15 5
12Haloperidol 5Nil
13Risperidone 3Nil
16RisperidoneNil 5
17Haloperidol10 5
18Flupenthixol15 4
20Risperidone 515
21Sulpiride 5 2
22Haloperidol 7.5Nil
23Risperidone 4 4
24Risperidone15 4
25Risperidone 4Nil
27Sulpiride 6 2
28Risperidone 4 4
29Flupenthixol 5Nil
30Risperidone 7.5Nil
31Risperidone 4 4
32Risperidone 7.5Nil
33Trifluoperazine10 5
34Haloperidol15 2
35Haloperidol 5Nil
39Risperidone 7.5Nil
41Risperidone 4Nil
43Risperidone 4 2

Additionally, the patient's severity of psychotic symptoms were required to be defined by a minimum normalized Brief Psychotic Rating Scale (BPRS)14 score of 12 of 24 points and at least 4 on the Clinical Global Impression Scale-Severity (CGI-Severity scale)15 plus either at least 4 points on any two of the following items on the Positive and Negative Symptoms Scale (PANSS):16 hallucinatory behavior, suspiciousness and unusual thought content, grandiosity, and conceptual disorganization, or 4 points on one of the PANSS items aforementioned plus a PANSS negative subscore of at least 21. Patients with serious medical conditions, substance-related disorders, pregnancy, or lactation were excluded.


This was a 13 week prospective, open-label clinical trial. Patients had discontinued their previous antipsychotics 2–9 days before entering the present study, then olanzapine treatment was initiated at 5 mg/day. Dosing was adjusted afterward to within the range of 10–25 mg per day, based on the clinical response. Patients were evaluated weekly during the initial 4 weeks, and biweekly for the subsequent 9 weeks using the efficacy rating scales; these included the PANSS, BPRS, and CGI. Extrapyramidal adverse effects were assessed thorough the Abnormal Involuntary Movement Scale (AIMS),17 Simpson–Angus Scale (SAS)18 and Barnes Akathisia Scale (BAS).19 Treatment responders were defined as those having at least a 20% reduction in the 16-item normalized BPRS total score from baseline to end-point, plus an end-point CGI-Severity scale score of ≤3.3 Clinical laboratory testing, including hematology, clinical chemistry, and electrolytes, was also performed before and after the study.

Statistical analysis

Data management and analysis were carried out using spss software. Improvement in treatment was defined by a decrease in scores for all rating scales of the end-point, with respect to baseline scores. Comparisons of efficacy, extrapyramidal adverse effects (EPSE), and bodyweight change were done with the t-test. A logistic regression model was used with the patients’ demographic characteristics (years of education, marital status), symptom severity (PANSS scores), global function (CGI scores), and side-effects of medication (AIMS, SAS, and BARS scores) as independent variables, and treatment response as a dependent variable. The mean modal dose was calculated by deriving the cumulative dosage received during the study divided by the number of days of therapy. The maximum daily dose was defined as the highest dose given to a patient.


Forty-three (84%) of the 51 enrolled patients completed the 13 week study. Five patients dropped out due to the worsening of their psychotic symptoms, two patients discontinued owing to their poor drug compliance, and the remaining patient complained of a lack of efficacy. The comparison of the demographic variables between the completed subgroup and the non-completed subgroup showed no statistical difference. The mean modal drug dose was 20.4 ± 4.0 mg per day, and the mean maximal dose was 23.5 ± 3.4 mg per day. At the end of the study, 20 patients (39.2%) were categorized as responders. Olanzapine produced significantly greater improvement (P < 0.05) on the BPRS, CGI, and PANSS for patient evaluation; this improvement included negative as well as positive symptom areas by the end of 13 weeks of therapy (Table 2). Results from logistic regression analysis indicated that the AIMS score prior to receiving olanzapine treatment was the only marginally significant predictor for the respondent subgroup (β = −1.13, P = 0. 056).

Table 2. Mean changes of the PANSS, BPRS, and CGI scores from week 1 to week 13 (n = 43)
Positive scoreNegative scoreGeneral scoreTotal scoreScore
Mean ± SD% changePMean ± SD% changePMean ± SD% changePMean ± SD% changePMean ± SD% changeP
  1. % change, compared with week 0; PANSS. Positive and Negative Syndrome Scale; BPRS, Brief Psychotic Rating Scale; CGI, Clinical Global Impression Scale.

 0 27.3 ± 7.68  28.55 ± 7.90  49.57 ± 13.56  58.77 ± 16.31  5.25 ± 0.75  
 324.23 ± 6.9311.250.0007224.86 ± 7.9212.960.000344.34 ± 11.2810.550.0007651.93 ± 13.1811.640.000454.66 ± 0.9111.240.00032
1022.25 ± 6.7518.500.0001923.27 ± 7.8218.490.0003341.93 ± 10.6415.410.0005648.57 ± 13.6017.360.000744.11 ± 0.8121.710.00071
1321.91 ± 6.5419.740.0005122.70 ± 7.5320.490.0001641.02 ± 9.9817.250.0002547.43 ± 13.2019.300.00023.98 ± 0.8524.100.00020

As to safety, olanzapine was well tolerated. The common adverse events experienced were nervousness (20.6%), insomnia (11.8%), and weight gain (5.9%). The mean bodyweight at the start and the end of the study was 63.9 ± 13.9 kg and 66.0 ± 12.7 kg, respectively. The t-test showed that the weight gain during the present 13 week study was statistically significant (P < 0.05; Table 3). Eighty-eight percent of the patients required anticholinergic medication at the beginning of the study and 42% of the patients required anticholinergic medication in the end of the study (Table 1). The changes in the SAS scores from baseline to end-point were statistically significant (Table 3). There was no significant change in the results of the clinical laboratory tests, except in the case of one female subject who developed hyperglycemia after olanzapine treatment.

Table 3. Extrapyramidal treatment-emergent advance events and weight change (n = 43)
WeekAIMSSASBARSWeight (kg)
Mean ± SD% changePMean ± SD% changePMean ± SD% changePMean ± SD% changeP
  1. % change, compared with week 0; AIMS, Abnormal Involuntary Movement Scale; SAS, Simpson–Angus Scale; BAS, Barnes Akathisia Scale.

 01.18 ± 2.27  1.89 ± 3.38  1.11 ± 1.96   63.1 ± 13.94  
 31.18 ± 2.27 01.001.73 ± 3.26 8.470.5751.07 ±   
10 0.8 ± 1.8432.200.0251.43 ± 2.1824.340.1531.00 ± 1.63 9.910.672   
130.82 ± 1.6530.510.0581.18 ± 2.0137.570.0410.82 ± 1.426.130.30565.98 ± 12.864.60.001


The results of the present clinical trial demonstrate that a portion of Chinese schizophrenic patients who have not responded to classical neuroleptics might respond to olanzapine. Although the open-label nature of the study limited its interpretation, the result is similar to that of other open-label naturalistic studies among Caucasian subjects.7,8,10 The current guideline for olanzapine use in schizophrenic patients suggests a maximum dose of 20 mg per day; the literature regarding the efficacy of olanzapine in doses above this level is still scant. The maintenance dosage for olanzapine, however, was not a significant predictive factor for the respondent subgroup in the present study. The mean dosage for the olanzapine-treated patients in the current respondent subgroup was 21.0 ± 3.8 mg per day, which is higher than the dosage (10.31 ± 3.97 mg/day) used for treating non-resistant patients in Japan.20–22 Taylor et al. reported that the mean dose of olanzapine prescribed longer than 6 weeks for schizophrenic patients was 15.8 mg per day.23 It seems that Chinese treatment-resistant schizophrenic patients need a higher dose than the treatment-responders.4,24 Several reports of Caucasian patients have also demonstrated that a higher dose is needed in olanzapine-treated patients with treatment resistance.8,10,11

Patients in the present study had received high doses of antipsychotic medications long before the present olanzapine treatment because of their poor treatment response. Hence, more of our patients took anticholinergic medication during the period of olanzapine treatment than the patients that Taylor et al. reported (15%).23 However, both the SAS scores and the necessity of anticholinergic medication decreased significantly during the study. Ethnic comparison reports suggest that the metabolism of antipsychotics differs among ethnic populations, and that ethnicity is an important variable in predicting extrapyramidal symptoms.12,25,26 The difference in plasma drug concentrations may explain why Asian patients are reported to require smaller dosages of neuroleptic drugs than non-Asian patients, and why they may be more sensitive to neuroleptic-induced side-effects.27,28

Regarding the degree of correlation of EPSE to the efficacy of antipsychotic treatment, researchers have demonstrated that the emergence of EPSE can significantly compromise patient compliance with treatment, and can have profound effects on long-term outcomes. However, the relation between the efficacy of short-term treatment and the emergence of EPSE is still inconclusive. This can be explained by the fact that EPSE during classical neuroleptic therapy would make the schizophrenic symptoms worse, and that the improvement in EPSE after switching to olanzapine therapy would make the schizophrenic symptoms better. The pathophysiological basis of patients who are exceptionally sensitive to the adverse effects associated with antipsychotics, particularly the EPSE, should be further explored. Compared with the baseline stage, the stage of previous antipsychotics use, and the low incidence of treatment-emergent akathisia, dystonia, and parkinsonism, suggest that olanzapine has less induced EPSE, and that even high doses of olanzapine were actually free from extrapyramidal adverse effects in some cases in the present study.

One study result suggests that clozapine and olanzapine have similar general antipsychotic efficacy in the treatment of patients with chronic schizophrenia.29 In addition, a successful conversion from clozapine to olanzapine for clozapine responders has been reported.30 However, not all patients who received clozapine before olanzapine treatment in the present study had a successful conversion. The five patients who dropped out in the present study all received clozapine before olanzapine treatment. and the one who completed the study was an olanzapine non-responder. The clozapine non-respondence of the subjects enrolled may be the reason for the conversion failure. Additionally, the switch from clozapine to olanzapine could also have led to severe withdrawal symptoms.31

Although nervousness, insomnia, and other adverse effects were reported, the rate of severe adverse effects was low. Weight gain is an issue that concerns many patients and physicians. One previous study revealed that patients receiving clozapine or olanzapine showed significant increases in weight, body mass index, and leptin level.32 The weight change in the present study was significant. In addition, hyperglycemia in patients with schizophrenia treated with olanzapine is also a critical issue.33,34 Case reports about hyperglycemia and diabetic ketoacidosis associated with olanzapine treatment are not limited.35–39 One 29-year-old female subject in the present study developed hyperglycemia after 13 weeks of olanzapine treatment. Her medical records had been reviewed. There was no history of elevated serum glucose levels or glucose in the urine. There was no family history of diabetes. She was in good general health and did not have a past history of active substance abuse. Clinicians should now be alert when blood glucose deteriorates in psychotic patients, and the glucose level should perhaps be monitored when a high dose of olanzapine is used in treatment-resistant patients. Additional well-controlled and double-blinded prospective clinical trials would be of interest to further determine the effectiveness of high-dose olanzapine in treatment-resistant schizophrenic patients.


This work was supported partially by Eli Lilly and Company (Taiwan). The authors wish to thank Mr Mitchell Chen, Miss Karen Hsu, and Dr Chwen-Cheng Chen of the Department of Psychiatry, National Cheng Kung University, for statistical, administrative, and editorial assistance with the manuscript.