*Chih-Chiang Chiu, MD, Department of Psychiatry, Taipei City Psychiatric Center, Taipei City Hospital, No. 309, Sungde Road, Taipei 110, Taiwan. Email: email@example.com
Aims: This 12-week open-label study was designed to investigate the pharmacokinetics and efficacy of a direct switch from a conventional depot to long-acting injectable risperidone in patients with schizophrenia and schizoaffective disorder.
Methods: Men or women from 18 to 65 years old with a diagnosis of schizophrenia or schizoaffective disorder were eligible for participation if they had been treated with conventional depot for at least 8 weeks before study entry. Intramuscular long-acting risperidone was administered starting from 25 mg, with the dose flexibly adjusted every two weeks for 12 weeks from week 4.
Results: Of the 25 patients enrolled in this study, 21 completed at least one post-baseline assessment and were thus included in the analysis. The mean serum concentration of risperidone plus 9-hydroxyrisperidone was 29.1 ng/mL at the 12th week after switching, with an average injection dose of 31.25 mg long-acting risperidone every two weeks. The levels of active moiety of risperidone seemed to be higher in Chinese patients compared to those in Caucasian patients. Positive and Negative Syndrome Scale total scores (from 67.5 to 56.4; P = 0.002), scores for negative symptoms (P = 0.006) and general symptoms (P = 0.001) were improved significantly 12 weeks after the switch. Mean Extrapyramidal Symptom Rating Scale scores were improved significantly from 20.1 to 5.5 (P < 0.001). Significantly decreased levels of cholesterol and triglyceride were found at the 12th week. The levels of fasting glucose, low-density lipoprotein, high-density lipoprotein and bodyweight remained unchanged.
Conclusions: These findings suggest that switching from conventional depot to long-acting risperidone is feasible with the advantage of symptom reduction and side-effect profile decrement.
PHARMACOLOGICAL TREATMENT IS a critical component in the long-term management of patients with schizophrenia.1 A previous study found that 82% of schizophrenic patients experience at least one relapse 5 years after the first break, and the relapse rate in unmedicated patients increases approximately 10% per month.2 Despite the demonstrated efficacy of antipsychotics, non-compliance is still a major reason for the high relapse rate among patients with schizophrenia. There are multiple factors that influence compliance, such as health belief models, comorbidities, social perceptions regarding psychiatric treatment, the doctor–patient relationship, and side-effect profiles.3
Long-acting conventional depot antipsychotics, developed in the 1960s, have been advocated as one of the ways to promote compliance in people with schizophrenia.4 Their reported merits include avoidance of problems with absorption and hepatic biotransformation, the predictability of stable plasma levels, and the lack of abrupt discontinuation.5 However, conventional depot antipsychotics have a number of limitations, especially adverse effects involving movement disorders.6 Atypical antipsychotics were developed to address the problems associated with typical antipsychotics, and their most notable improvement is a reduction of acute extrapyramidal system adverse effects (EPS).7 Despite the higher rate of compliance with atypical antipsychotics due to their superior side-effect profiles, interventions to improve compliance are still warranted.8
The introduction of long-acting risperidone meets the need for an agent combining the advantages of a long-acting formulation with those of an atypical antipsychotic. Results of several 12-week clinical trials and one large 12-month, open trial suggest the effectiveness and safety of switching to long-acting risperidone therapy from the oral form of risperidone.9–11 The pharmacokinetic profile of long-acting risperidone has been suggested to have 25–32% lower mean steady-state peak concentrations of the active moiety (risperidone plus 9-OH risperidone) and 32–42% lower fluctuations in plasma active-moiety levels than those of the oral form.12 Although previous studies suggested that patients could also be safely and effectively switched from a conventional depot antipsychotic to long-acting risperidone13,14 in Caucasian patients, data of pharmacokinetics and the efficacy of long-acting risperidone shifting from conventional depot in Chinese patients is still lacking.
It has been suggested that the oral form of risperidone may be associated with moderate risks of exacerbation of bodyweight gain, lipid profiles, and diabetes compared to other atypicals.15 However, there have been no reported studies of metabolic changes after shifting from conventional depot to risperidone long-acting injection. The objective of this open-label study in Chinese patients in Taiwan was to investigate the pharmacokinetics, efficacy, and tolerability of direct switching from conventional depot antipsychotics to long-acting injectable risperidone.
The participants were men or women from 18 to 65 years old with a diagnosis of schizophrenia or schizoaffective disorder according to DSM-IV criteria.16 They were eligible for participation in the trial if they had been treated with depot formulations of flupenthixol, fluphenazine, haloperidol or zuclopenthixol during the 8-week period before study entry. Patients were determined to be clinically stable on their current treatment at screening and at baseline visits based on a total Positive and Negative Syndrome Scale (PANSS) score ≤80 and a score ≤4 on each of the following PANSS items: conceptual disorganization, hallucinatory behavior, suspiciousness and unusual thought content. Results of standard clinical laboratory tests were required to be within laboratory reference ranges or, if outside the ranges, judged to be not clinically significant by the investigator.
Exclusion criteria were as follows: a history of severe drug sensitivity/allergy or unresponsiveness to risperidone, a history of neuroleptic malignant syndrome or tardive dyskinesia, clinically significant electrocardiogram (ECG) abnormalities, documented organic disease of the central nervous system, unstable or significant untreated medical illness, and history of clozapine treatment within the previous 3 months. Also excluded were patients who had participated in an investigational drug trial in the 30 days prior to screening and women who were pregnant, likely to become pregnant, or lactating. This study was approved by the Institutional Review Board of Taipei City Hospital and written informed consent was obtained from each patient prior to participation.
Following a 1-week screening period, patients entered the 12-week treatment trial. The first injection of long-acting risperidone was introduced at the time when the previous conventional depot was due, with the latter being stopped immediately. Subjects received an i.m. injection of long-acting risperidone every 2 weeks, starting from 25 mg every two weeks, and the dosage was adjusted flexibly 4 weeks after initiation by in-charged physicians who were blind to the results of the rating scale.
Chronic antidepressant or benzodiazepine use, antiparkinsonian medication, mood stabilizers, and hypnotics were permitted if the subject was taking (had taken) the medication prior to the trial. However, use of CYP 2D6 inhibitors or inducers, such as fluoxetine, paroxetine, and cimetidine, were not permitted during the trial. Oral risperidone and all other antipsychotic supplementation were not permitted during the interim. Benzodiazepines were allowed as rescue medications during the trial. Any new conditions or a worsening of an existing condition which led to concomitant therapy being given as a treatment were documented on the Adverse Event Form.
Assessments were performed on week 0, 2, 4, 8, and 12 after the switch to long-acting risperidone. Efficacy was assessed with PANSS and the Clinical Global Impressions scale (CGI) at every study visit. Adverse events and the Extrapyramidal Symptom Rating Scale (ESRS) were assessed at every study visit. Bodyweight was measured at weeks 0, 4, 12, and a physical examination was performed at screening and at the study end-point. ECG was performed at week 0 and at the study end-point. Repolarization duration variables were corrected for heart rate using the Bazett formula (QTc = QT (ms)/RR0.5).
Serum concentrations of risperidone and 9-hydroxyrisperidone were measured at every visit. All overnight fasting blood samples were drawn between 7.00 hours and 8.00 hours, 12 h after the evening administration of the medication and just before the next long-acting risperidone injection. Venous blood was collected into an EDTA-tube and centrifuged at 3000 rpm for 15 min. The plasma samples were stored at −60°C until assayed. Determinations of risperidone and 9-hydroxyrisperidone were performed by high performance liquid chromatography (HPLC) using ultraviolet detection. A LC-18 PK/108 solid phase cartridge (Supelco, Bellefonte, PA, USA) with a vacuum manifold was used for plasma extraction. Prior to the application of the sample, the clean-up cartridge was conditioned by consecutive rinses with 1 mL methanol and 1 mL water. One milliliter of the plasma sample was mixed with 50 µL (3.2 µg/mL) methylrisperidone (the internal standard) and 1 mL 0.2 M potassium chloride (which was adjusted to pH 12.0 using 0.2 M sodium hydroxide), and then applied into the cartridge. After being washed with 50% methanol in water (v/v), the analytes were eluted with 1 mL 0.01 M acetic acid in methanol. The HPLC set was equipped with a Waters 600-MS system controller, a Waters 717 WISP autosampler, and a Spectra Series UV 150 ultraviolet detector. Separations were performed on a reverse phase Waters Nova-Pak Phenyl column (3.9 × 150 mm; 4 µm particle size; Milford, MA, USA). The mobile phase was composed of 0.05 M potassium phosphate and acetonitrile (70:30, v/v, pH = 6.5 with 1 M KOH). All water was Milli-Q grade. The isocratic separation was performed at a 0.9 mL/min flow-rate at ambient temperatures. The eluent was monitored by the ultraviolet detector at 277 nm. Linearity was observed in a standard curve of risperidone and 9-hydroxyrisperidone over a range of 2.5–200 ng/mL for plasma. Within-day and between-day standard curves in plasma with risperidone and 9-hydroxydrisperidone had a correlation coefficient greater than 0.99, with coefficient of variance less than 12% for all the concentrations. The recovery rate of risperidone and 9-hydroxyrisperidone from plasma were, respectively, 85 ± 10% and 72 ± 2%. The limit of quantitation was 2 ng/mL for both compounds in plasma.
The serum levels of glucose, triglyceride, cholesterol, low-density lipoprotein (LDL), and high-density lipoprotein (HDL) were measured at baseline and week 12 by fasting blood samples. Laboratory assays were performed at the chemistry laboratory of Taipei Institute of Pathology. The serum levels of glucose, triglyceride, cholesterol, LDL and HDL were measured using the Olympus AU400 chemistry analyzer. The plasma glucose level was measured with a hexokinase method. The plasma levels of triglyceride and cholesterol were measured with the enzymatic colormetric method. Plasma HDL level was determined by homogeneous liquid selective detergent. Plasma low-density lipoprotein-cholesterol was calculated from the Friedewald equation.17
All participants who completed at least one outcome assessment were included in the intent-to-treat analysis, which followed the rules of the last observation carried forward. Descriptive statistics are represented as mean ± SD. For outcome measures, including the PANSS, PANSS subscales, and ESRS, we analyzed the change from baseline using a repeated-measures ancova, controlling for baseline. The difference between the baseline and end-point of other measurements, such as CGI and metabolic parameters, was analyzed using the paired t-test for continuous variables and the χ2-test for categorical variables. Pearson correlation was used to assess the correlation between levels of active moiety and the scores of the PANSS, ESRS scores, or prolactin levels. A P-value of less than 0.05 was considered statistically significant. All analyses were carried out using the software package spss, version 12.0.
A total of 25 patients, 12 men and 13 women, were enrolled. Five of them failed to complete the 12-week trial. Four patients withdrew from the study at week 1 due to their unstable condition (one) or withdrawal of consent (three), none of whom were included in the statistical analysis. Another patient discontinued the study medication due to increased irritability at week 8. The number of patients at each visit was 25, 21, 21, 21, and 20 at weeks 0, 2, 4, 8, and 12, respectively. Twenty-one patients, four inpatients and 17 outpatients, with at least one post-test were analyzed. These patients were being treated with haloperidol decanoate (n = 15), flupenthixol decanoate (n = 3), fluphenazine decanoate (n = 1), and clopenthixol decanoate (n = 2) before this study. The average treatment duration with conventional depot was 34.2 ± 32.3 (3–108) months. The average time since the last injection of conventional depot was 3.2 ± 1.0 (2–4) weeks. The baseline dosage of conventional depot antipsychotics was equivalent to 231.8 mg/day of chlorpromazine.18 The baseline demographic and clinical characteristics of the patients are summarized in Table 1. At the end of the 12-week treatment, 10 patients were treated with 25 mg long-acting risperidone every 2 weeks and the other 10 patients were treated with 37.5 mg every two weeks. For the latter 10 patients, five began the 37.5 mg/2 weeks dose at week 4 and the other five began this dose at week 6.
Table 1. Baseline demographic and clinical characteristics of the patients
n = 21 [Mean (SD)]
Sex, n (%) female
Age at diagnosis
Age at first treatment
Age of first admission
Number of admissions
The changes of concentration/dose (C/D) ratio (the sum of active moiety/ the mean daily dose of long-acting risperidone between blood sampling) during the 12-week study period are shown in Figure 1 (n = 18). For example, if case A received 25 mg at week 4 and 37.5 mg at week 6 then the C/D ratio at week 8 will be (levels of risperidone + 9-OH risperidone)/(25 mg/28 day + 37.5 mg/14 day). For most of the participants, C/D ratios were quite low at week 2 and became more stable at week 8 and week 12. The mean serum concentration of risperidone plus 9-hydroxyrisperidone was 29.1 ng/mL at the 12th week after switching, with an average injection dose of 31.25 mg long-acting risperidone every two weeks. Participants whose last dose was 25 mg had lower levels of the active moiety than those with 37.5 mg (20.1 ± 13.8 vs 39.6 ± 26.7 ng/mL; P = 0.06) at week 12. No correlation was found between the plasma concentration of active moiety at week 12 and age, previous dose of conventional depot (in terms of chlorpromazine dosage), weight, or number of hospital admissions. There was no association between levels of the active moiety and the scores of PANSS (total scores or sub-scores-positive, negative, and general), ESRS scores or prolactin levels.
The changes in PANSS total scores and subscores, ESRS, and CGI scores during the study period are shown in Table 2. Symptom severity (PANSS scores) was generally low at baseline in these patients. However, 50% of the patients had a 20% reduction in the PANSS total score at the study end-point. Significant improvements in mean PANSS total scores, and scores for negative symptoms and general symptoms were achieved after 12 weeks. However, no significant changes were found in PANSS scores for positive symptoms although mild improvement was noted. PANSS factors measuring mood (anxiety/depression) were also significantly improved at the study end-point (P = 0.002). Improvement in the ESRS was prominent during this study. Although CGI severity results showed an increase in the number of patients rated as not ill, very mild, or mild from 66.67% to 76.19%, these changes were not significant at the end-point of the 12-week trial.
Table 2. Changes of PANSS-total and –subscale scores, ESRS, and CGI (Mean(SD)) during the study period
CGI, Clinical Global Impression; ESRS, Extrapyramidal Symptom Rating Scale; PANSS-Negative, scores for negative symptoms of Positive and Negative Syndrome Scale (PANSS); PANSS-total, total scores of PANSS; PANSS-Positive, scores for positive symptoms of PANSS.
Side-effects were reported by 55% of patients over the twelve-week period. The most common side-effects reported were menstrual irregularity, increased talkativeness, dizziness, and somnolence (Table 3). Most of the side-effects were tolerable, except for one participant who discontinued the study medication at week 8 due to increased talkativeness and excitement. The number of patients with concomitant medication at baseline was eight with propranolol, eight with anticholinergics, five with benzodiazepine, one with trazodone, and one with valproate. At week 12, the numbers using these medications had changed to 5, 5, 3, 1, and 1, respectively. The attenuation in the proportion of use of concomitant anticholinergic medications was also compatible with the improvement in movement disorders (from 38.1% at baseline to 23.8% at the study end-point). The proportion of benzodiazepine use also decreased during the 12-week study period (from 23.8% at baseline to 14.3% at the study end-point). Changes of metabolic parameters between baseline and the study end-point are summarized in Table 4. Significant decreases were found in levels of cholesterol and triglyceride while levels of glucose, LDL and HDL did not change statistically significantly. Mild bodyweight gain (1.2 ± 3.5 kg) was also noted although this increase did not reach a significant level. No definite QTc change (414.7 ± 21.0 ms at baseline to 420.6 ± 25.3 ms at end-point, P = 0.22) was found after the use of long-acting risperidone.
Table 3. Types and frequency of reported side-effects during trial period
n = 21
EPS, extrapyramidal symptoms.
Exacerbation of EPS
Prominent bodyweight gain
Table 4. Metabolic changes (Mean (SD)) after switching to long acting risperidone from conventional depot antipsychotic medication
Baseline (week 0)
End-point (Week 12)
High-density lipoprotein (mg/dL)
Low-density lipoprotein (mg/dL)
The results of this trial in Chinese patients suggest that switching from conventional depot antipsychotic medications to risperidone long-acting injection appears to be well-tolerated and efficacious as shown by symptom reduction, ESRS attenuation and decreased use of concomitant medications without an oral risperidone run-in period. Patients with higher injection dosages and men tended to have higher levels of the active moiety.
As in previous studies, the pharmacokinetic changes in our study showed wide inter-individual variation.12,19,20 The level of active moiety was quite low at week 2, an observation compatible with the pharmacokinetic study on single-dose long-acting risperidone, which showed a 3-week latent period prior to the gradual increase of active moiety. The delayed release may be due to the gradual hydrolysis of the matrix of glycolic acid-lactate copolymer of which drug was contained at the site of injection. The levels of active moiety became stable at week 8, and were similar to levels at week 12. There was a two-week delay in reaching a more steady concentration in our study compared to the previous report.21 This might be due to continued adjustment of the injection dosage in some patients at week 4 and week 6.21 There was a trend that plasma levels in subjects with a last dose of 37.5 mg were higher than those with 25 mg, a finding compatible with previous pharmacokinetics studies that suggested a dose-proportional increase in the administration of long-acting injectable risperidone.12 Paralleled, the correlation between dosage and plasma concentration of active moiety for oral risperidone administration has also been reported.22 Western studies found the average steady-state plasma level of active moiety of long-acting risperidone ranged from 13.1 to 18.7 ng/mL for the 25 mg dose group and from 23.2 to 27.5 ng/mL for the 37.5 dose group.19,23,24 Our results suggest that Chinese patients may have higher active moiety levels than Caucasian patients under the same treatment dose. However, the active moiety level was quite low (7.84 ± 3.52 ng/mL) in another Chinese population study that investigated a shift from oral risperidone to long-acting risperidone and the lower levels might have been attributable to lower dosage.20 In parallel, the concentration of risperidone and active moiety in Chinese patients treated with oral risperidone were higher than those of Caucasian patients.25 As the activity of both isozymes of cytochrome P450 (CYP) metabolizing risperidone, CYP2D6 and possibly CYP3A, are lower in East Asians than in Caucasian people,26 ethnic differences potentially influence the disposition of long-acting risperidone. In addition, the fact that the expression of the CYP2D6 enzyme is lacking in some subjects and abundant in others suggests that the activity of the CYP2D6 enzyme among subjects is highly variable, and is thus associated with different phenotypes, including ultrarapid, extensive, intermediate, and poor metabolizers.27 A previous study indicated that the CYP2D6 poor metabolizer (PM) phenotype, not CYP3A PM, plays an important role in individualizing dosage for oral risperidone. Further studies are needed to investigate the genetic effect of CYP2D6 and CYP3A on the pharmacokinetics in long-acting injection risperidone.28
In all of the symptom profiles, significant improvement was found in negative symptoms and general symptoms rather than positive symptoms. A significant improvement in PANSS total score after shifting to long-acting risperidone injection from conventional depot was found from week 4 in our study, an observation consistent with a similar previous study by Turner et al. in Caucasian patients with schizophrenia.13 Until now, despite the fact that there has been no positron emission tomography data focusing on the early effects of long-acting injectable risperidone in D2 occupancy,29,30 plasma concentration of active moiety has achieved stability at week 3–4 after the injection.12 The D2 receptor occupancy by conventional depots was reported to be persistent after discontinuation, notably, even 30% after 6 months.31 Therefore, the start of the stable release of active moiety, as well as continuing D2 receptor occupancy, may contribute to the early therapeutic effects of our subjects. Negative expectancy appraisals have been shown to be associated with negative symptoms, including low self-efficacy, low satisfaction, low acceptance, and low available resources, which affect patients' long-term functioning.32 Therefore, long-acting risperidone might reduce these negative appraisals and improve patients' functioning. One study reported that positive symptom domains also improved significantly after switching from conventional antipsychotics to risperidone,33 while another reported that positive symptom domains showed delayed improvement.34 The negative findings in positive domains of the PANSS in our study may have been due to the limitations of the study period or the small sample size.14,35 They may also imply that conventional depot antipsychotics have the same effects on positive domains as long-acting risperidone.
Although the mnemonic device SWITCH, which suggested gradual crossover from a conventional antipsychotic to an oral atypical antipsychotic, has been recommended as a strategy for shifting oral conventional to oral atypical agents,36 successful switching from one depot to another without oral supplementation or a washout has been reported for conventional antipsychotic depots. In previous experiments in treating with depot antipsychotics, the pertinent strategies included dose adjustments at designated intervals, flexible intervals between injections, and supplementary injections. In the present study, simplification of the procedure via a policy of direct switch from depot to long-acting risperidone seemed feasible.13,14 Because of a slow decline in drug concentration following withdrawal from conventional depots, the concentration can still be detected 24 weeks after the medication is discontinued.37,38 The prolonged release of a previous depot may partly explain the success of direct shift.
Significant reduction of ESRS scores and usage of anticholinergics has been demonstrated to be beneficial for movement disorders in patients with schizophrenia. Previous studies conducted over 6-month treatment periods also showed similar results.33,39 A previous report also indicated that long-acting injectable risperidone was tolerable in elderly patients due to lower EPS.33 In addition, the use of long-acting risperidone was also reported to be efficacious in recent-onset psychosis40 and schizophrenic patients with comorbid substance abuse.41 Thus, long-acting risperidone appears beneficial in treating a variety of psychotic patients with satisfactory efficacy and safety.
The lipid profiles in patients taking conventional antipsychotics exhibited diverse changes. Chlorpromazine and other phenothiazines could elevate serum triglycerides and total cholesterol whereas butyrophenone derivates such as haloperidol did not show these dyslipidemic effects despite an increase of LDL that was noted in a butyrophenone cohort.42 Furthermore, relatively mild or neutral effects of risperidone on serum lipids has been demonstrated.42 There is also no difference concerning the changes over TG or cholesterol in first-episode schizophrenic patients receiving risperidone or conventional antipsychotics (sulpiride).43 There remains a paucity of data to compare the effects on lipid profiles between oral risperidone and conventional antipsychotics, or between long-acting injectable risperidone and conventional depot. Our study showed that cholesterol and triglyceride levels were significantly decreased after 12-weeks of long-acting risperidone treatment. A trend toward lower LDL was also found. These results seem to be compatible with those of a recent large Clinical Antipsychotic Trial of Intervention Effectiveness (CATIE),44 which showed that changes in the levels of cholesterol and triglyceride in patients treated with risperidone were lower than those treated with conventional antipsychotics. However, the cholesterol and triglyceride levels, measured only at baseline and week 12, were all within normal range. In addition, the mean bodyweight change after 12 weeks of the depot shift was an increment of 1.2 kg. The slight increase in bodyweight may be related to insulin resistance and would potentially further impair lipid metabolism in a longer follow up.42 Pertinent investigations are warranted clarifying the dyslipidemic issue. Previously, we have reported a significantly increased prolactin level from 27.7 ± 28.5 to 55.9 ± 43.0 ng/mL after shifting from conventional depot to long-acting injectable risperidone.45 As hyperprolactinemia may lead to a variety of complications, such as the menstrual irregularity in our subjects, it should be paid more attention on the prolactin changes in practice.
Interpretation of the study results is limited by the open-label uncontrolled design and the fact that it was funded by a pharmaceutical company, both of which may potentially influence clinical evaluation such as reducing dosage of benzodiazepines and anticholinergics. In addition, lack of direct comparison with conventional depot antipsychotics, the small sample size, prolonged effects of previous conventional depot antipsychotics, the relatively short-term duration of treatment, and the fact that the pharmacokinetics did not reach a steady-state also should be taken into consideration. Furthermore, there were no pharmacogenetic tests performed in the patients and the activity of CYP2D6 and CYP 3A4 would influence the results of plasma levels.
In conclusion, our study demonstrated that Chinese patients with schizophrenia or schizoaffective disorders, who are being treated with a conventional depot antipsychotic, can be switched to long-acting risperidone safely and efficaciously without a run-in period with oral risperidone. The levels of active moiety of risperidone may be higher in Chinese patients compared to those in Caucasian patients. Further studies to investigate the pharmacokinetics in the steady-state of long-acting injectable risperidone with the effects of pharmacogenetics and activity of CYP2D6 and CYP 3A4, and its long-term effects on metabolic parameters are warranted.
The authors would like to thank Ms Su-Cheng Chang and Ms Shaw-Tein Wu for data collection and laboratory analysis.
This study was supported by grants from Janssen-Cilag Taiwan, Taipei, and the Johnson & Johnson Corporation, Taipei. They were not involved in any of the processes of case enrollment or data analysis.