Tardive dyskinesia is a common complication of conventional antipsychotic treatment in subjects with schizophrenia. This study investigates whether the addition of the omega-3 fatty acid, ethyl-eicosapentaenoic acid (EPA) to usual treatment improves movement disorder in 84 schizophrenia subjects with established tardive dyskinesia. The initial double-blinded, randomised trial duration is 12 weeks, followed by further 46 weeks of open-label treatment.

Background:

One of the major limitations of conventional antipsychotics is their high propensity to cause extrapyramidal symptoms (EPS). Tardive dyskinesia (TD) in particular causes problems, insofar as it is common, and resistant to treatment. TD is under-recognised in clinical settings. However, prevalence studies indicate that 5% of patients treated with conventional antipsychotics develop TD each year for the first eight years with an average reported prevalence rate in the region between 17 and 23%. Although in the majority of cases the disorder is mild and not distressing, TD contributes to social and vocational impairment, as well as to the further stigmatization of the illness. A minority develop severe symptoms, which are extremely distressing and disabling, and may even be life-threatening. Treatment of TD is difficult. Dose-reduction or discontinuation of antipsychotic medication may paradoxically result in exacerbation of the TD symptoms, and also run the risk of precipitating a psychotic relapse. With the exception of clozapine and possibly botulinum toxin (for tardive dystonia), vitamin E7 and vitamin B6, little evidence exists to indicate efficacy for any treatment modality for TD.

The novel antipsychotics have had a huge impact upon the treatment of schizophrenia. The major advantage of these agents over their predecessors is their reduced risk of inducing EPS (and probably TD). However, high acquisition costs have limited their availability worldwide and many patients will continue to be exposed to conventional antipsychotics for the foreseeable future. Development of an effective and affordable treatment for TD would therefore be of great benefit.

Rationale for this trial:

One possible candidate for an effective and affordable treatment for TD is eicosapentaenoic acid (EPA), an omega-3 polyunsaturated fatty acid obtained from marine and plant sources. In addition to the evidence suggesting that EPA may improve the symptoms of schizophrenia when combined with standard antipsychotic treatment, there is also reason to believe that EPA may have a role in the treatment of TD. An open study in 20 hospitalised subjects with chronic schizophrenia reported a significant inverse correlation between dietary EPA and severity of TD. Treatment of these subjects with a standard EPA-rich marine oil for 6 weeks resulted in significant improvement in TD. We recently completed a 12-week randomised, double-blind study with ethyl-EPA 3g/day versus placebo as add-on to standard antipsychotic treatment, in forty subjects with chronic, refractory schizophrenia (submitted for publication). While there were no differences between the groups for the changes in the Extrapyramidal Symptom Rating Scale (ESRS) for parkinsonism, dystonia or akathisia scores, the ethyl-EPA group showed a significantly greater reduction in ESRS dyskinesia scores at 12 weeks (p=0.008). This result is potentially of great importance. Given the chronic nature of illness in our sample, most of these dyskinetic symptoms are likely to have been due to TD. This same study also reported a significant advantage for the ethyl-EPA group in terms of overall symptom reduction (PANSS total) (p=0.03), and analysis of co-variance indicated an association between PANSS total score reduction and ESRS dyskinesia score reduction.

This study will also investigate whether EPA may have an antipsychotic effect that is associated with an anti-dyskinetic effect. It has been suggested that an intrinsic aspect of the disease process of schizophrenia predisposes individuals with severe forms of the illness to antipsychotic-induced TD. These authors hypothesise that pathologic over-activity of mesolimbic and mesocortical dopamine neural systems may mediate persistence of psychotic symptoms despite adequate treatment, while increased TD vulnerability may be caused by overactive nigrostriatal dopamine activity from this process that is enhanced by antipsychotic drug exposure. If this is so, then it is possible that EPA could, by a common mechanism, have both an anti-dyskinetic and antipsychotic effect. This would explain the significant association that we found between reduction in dyskinesia scores and reduction in overall symptoms in our preliminary study (unpublished).

Objectives:

Primary objective:

• To compare the efficacy of ethyl-eicosapentaenoic acid (ethyl-EPA) versus placebo as supplementary medication in reducing symptoms of tardive dyskinesia (TD).

Secondary objectives:

• To compare the efficacy of ethyl-EPA versus placebo as supplementary medication in reducing symptoms of psychosis in these subjects.
• To assess whether there is an association between the antidyskinetic and antipsychotic actions of ethyl-EPA.
• To investigate the safety and tolerability of ethyl-EPA.

Trial design:

This is a double-blind, randomised, parallel-group comparison of ethyl-EPA and placebo in the treatment of established TD in patients with schizophrenia or schizo-affective disorder. There will be a pre-trial screen, following which subjects will enter the randomised treatment phase for 12 weeks. Responders will then be offered the option of open add-on treatment with ethyl-EPA for a further 40 weeks.

Patient selection:

Source of patients: In- and out-patients from state hospital and community psychiatric services in the greater Cape Town area.

Number of patients: Recruitment will stop when 84 eligible patients have been randomised to trial treatment. A recruitment period of 12 to 18 months is anticipated.

Pre-trial screenings:

Written, informed consent will be obtained prior to screening procedures. All patients will be screened to assess their eligibility for the trial. The screening visit will include the following:

• Informed consent
• Psychiatric history
• Medical history
• Physical examination
• Diagnosis
• Demography
• Vital signs
• Laboratory tests

Trial treatment:

Subjects will be randomly assigned to receive either an encapsulated ethyl-EPA supplement 2 g/day (2X500 mg capsules twice daily) (Laxdale Ltd.), or placebo (medicinal liquid paraffin BP 2 g/day), in addition to the medication that they had been receiving, for the duration of the study. Ethyl EPA is a highly purified derivative of fish-oil. The Food and Drug Administration has affirmed the status of fish-oil as generally recognised as safe with EPA doses up to 3 g/day. Trial supplies will be packed by an independent contract clinical trials supplies company (DHP), who will prepare the placebo and active packs for the entire trial and assign the randomisation numbers to the packs. The randomisation code will be broken after completion of the trial.

Concomitant treatment:

Only subjects who have been on stable medication for the preceding 6 weeks will be considered. Psychotropic medication will be fixed for the duration of the trial.

Anticholinergic medication:

Anticholinergic medication may be prescribed following randomisation, for treatment-emergent extrapyramidal symptoms (EPS).

Other psychotropic medication:

Patients who were stabilised on other psychotropic medications (anxiolytic, hypnotic, antidepressant, mood stabilising) before entry to the trial may continue on these medications. Anxiolytic or hypnotic medication may be prescribed for treatment-emergent conditions, eg. insomnia or acute anxiety.

Other concomitant medication:

Any medication for physical conditions that was taken prior to the commencement of the trial may be continued.

Medication for other conditions that arise during the course of the trial will be permitted, at the investigator's discretion.

Other omega-3 fatty acid preparations are not permitted.

Trial methods

Assessments:

Primary outcome measure:

• Change in Extrapyramidal Symptom Rating Scale (ESRS) dyskinesia score from baseline to week 12.

Secondary outcome measures for effect on TD:

• Change in ESRS for parkinsonism, dystonia, akathisia, and total scores from baseline to week 12
• The proportion of subjects in each group who achieve a 30% reduction in ESRS total scores at week 12.
• Time to remission (defined as a 30% reduction in ESRS total scores).
• The proportion of patients achieving a CGI Severity of TD score of < 3 at 12 weeks.

Secondary outcome measures for effect on psychosis:

• Change in Positive and Negative Syndrome Scale (PANSS) total, positive, negative and general psychopathology scores from baseline to week 12

Assessment of relationship between change in TD scores and change in symptoms of psychosis:

• Correlations between the above measures of TD and psychosis symptoms will be sought.

Secondary outcome measures for safety and tolerability:

1. The incidence of treatment-emergent EPS will be assessed at each visit by the proportion of subjects requiring anticholinergic medication.
2. Other tolerability and safety measures will include the reporting of adverse events, vital signs and weight.
3. All concurrent medication will be recorded.
4. A physical examination will be carried out at visit 1.
5. Laboratory tests:

In addition to liver functions, haematology and prolactin, the following tests will be done: bleeding time, fasting blood sugar and fasting lipogram. Although EPA is classified by the FDA as Generally Recognised as Safe in doses up to 3g/day, there have been reports of adverse effects on bleeding time, glycaemic control and low-density lipoprotein cholesterol.

Venepuncture will be performed at the pre-trial screen (visit 1), visit 4 (week 4) and at endpoint (visit 6).

The following tests will be performed:

• alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase
• haemoglobin, platelet count, total white cell count, differential white cell count
• prolactin
• bleeding time
• lipogram (fasting)
• blood-sugar (fasting)

Adverse events and patient withdrawals:

Any new medical condition, or the deterioration of a pre-existing medical condition occurring during the wash-out and randomised phase will be recorded on the case record forms (CRF's).

Withdrawal from trial:

This may occur for any of the following reasons:

• Protocol violation
• Withdrawal of consent
• Deterioration in the patient's condition, or lack of efficacy
• The occurrence of an adverse event
• Patient lost to follow-up

Data management:

All the data will be recorded in the case record forms. Two trained persons will independently perform a double data entry in a database. Verifications will be conducted during each of the two data entry processes and after comparison of the double data entry.

Statistical plan:

Statistical analyses will be performed by a biostatistician. The primary intent of this study is to evaluate the ability of ethyl-EPA versus placebo to reduce TD symptoms in 12 weeks of treatment. The principal null hypothesis is therefore that ethyl-EPA will not differ from placebo on the primary efficacy measure.

Sample size estimation:

We obtained an estimate of the variability of the change in ESRS dyskinesia scores after 12 weeks of treatment with ethyl-EPA and placebo as add-on to their previous antipsychotic medication from a previous trial conducted at our centre, which gave 2.13 as the standard deviation (unpublished). Using this estimate and with a significance level of 5% and 90% power, 32 patients per randomised group would be sufficient to detect a 1.75 point difference in change in ESRS dyskinesia scores (the difference obtained in our preliminary study) from baseline to endpoint. Allowing for an estimated withdrawal rate of 30%, we will need to recruit 42 patients per treatment group into this study.

Statistical Analysis:

A single statistical analysis will be performed at the end of the study. All of the tests will be interpreted at 5% significance level (2-tailed). Data analyses will be performed with Statistica software (StatSoft, Inc.).

Analysis of efficacy

Primary efficacy endpoint:

The primary efficacy measure will be the change in the ESRS dyskinesia score from baseline to endpoint. This will be analysed using analysis of covariance, including baseline score, treatment received, the centre and the centre-by-treatment interaction as factors. Comparisons will be performed by intention-to-treat, with last observation carried forward (LOCF). An additional analysis will be performed on the per-protocol population, and an analysis will be performed on the per-protocol population without the LOCF to assess the effects of any withdrawals from this trial.

Secondary efficacy endpoints:

The ESRS scores for parkinsonism, akathisia and dystonia, as well as the PANSS total and subscale scores will be analysed in the same way as the ESRS dyskinesia score.

Both p-values and 95% confidence intervals will be presented to evaluate the main comparison of interest.

Student's t test and Pearson's Product Moment Correlation Coefficient will be used for univariate differences and correlations respectively between numeric variables. To determine the contributions of individual variables toward TD scores, significant univariate results will be followed with regression analysis with simultaneous entry, using the method of least squares.

The significance of the contributions of the predictor variables to the proportion of patients classed as responders will be analysed using logistic regression.

Time to remission will be calculated. Comparison of the time to remission for the two groups will be in terms of Kaplan-Meier estimates for their survival curves. Controlling for covariates will be done by Cox proportional hazard regression.

All patients receiving at least one dose of trial medication will be included in the assessments of safety/tolerability.

Adverse events:

Adverse event frequencies will be summarised by body system. All adverse events will be presented in a table sorted by body system classification, along with degree of severity and relationship to treatment. Adverse events will be listed in full in an Appendix.

Laboratory tests:

Laboratory findings (haematology and biochemistry) will be summarized and evaluated using the laboratory's normal ranges. The data will be tabulated in normal/abnormal categories. Shift tables will be constructed where appropriate, to summarize changes in the data over the course of testing.

Vital signs and physical examination:

Systolic blood pressure, diastolic blood pressure, double pressure product and heart rate will be summarized descriptively by treatment group. Frequency tables will be calculated for the respective physical examination variables.

Ethical considerations:

The study will comply with The Declaration of Helsinki (Republic of South Africa Revision, 1996) and ICH Guidelines for Good Clinical Practice (1997).

The trial will be submitted for approval to the Institutional Review Board of the University of Stellenbosch and the Medicines Control Council of South Africa (regulatory authority).

Inclusion Criteria:

• Male or female aged 18 to 60 yrs
• Meeting Diagnostic and Statistical Manual of Mental Diseases, Fourth Edition (DSM-IV) criteria for TD.
• Meeting DSM-IV criteria for schizophrenia or schizo-affective disorder.
• CGI severity of TD score >3.
• Patients from whom informed, written consent is obtained.
• Patients who have been on a fixed dose of antipsychotic medication for at least 6 weeks prior to trial entry.

Exclusion Criteria:

• Significant neurological disorder other than TD
• Substance abuse
• Significant other medical illness
• Psychiatric disorder not stabilised
• Patients currently receiving clozapine
• Pregnancy or lactation