Two new analyses of the CATIE study provide more detailed comparisons of new and older anti‐ psychotics in patients with schizophrenia.

The first evaluated the efficacy of various atypicals in 114 patients who discontinued an older agent –in this case, perphenazine –due to lack of efficacy or adverse effects (Am J Psychiatry 2007;164:415‐27). After randomisation to olanzapine, quetiapine or risperidone, two‐ thirds of patients discontinued treatment prematurely, giving a median treatment duration of 5.8 months.

Time to treatment discontinuation was significantly longer for olanzapine and quetiapine than risperidone. The rates at which the atypicals were discontinued differed according to the reason for discontinuing perphenazine. Among patients who stopped due to lack of efficacy, about half then stopped quetiapine or risperidone but a quarter stopped olanzapine. Of those who stopped perphenazine due to poor tolerability, 40 per cent stopped quetiapine, 71 per cent stopped olanzapine and 85 per cent stopped risperidone.

The atypicals were similarly effective and, except for metabolic effects, similarly well tolerated. Olanzapine was associated with greatest increase in weight and changes in blood lipids. Hospitalisation for schizophrenia exacerbation was least frequent with quetiapine.

The second analysis showed that atypicals (this time also including ziprasidone and clozapine) and perphenazine were associated with similar but modest improvements in psychosocial functioning for up to 18 months (Am J Psychiatry 2007;164:428‐36). However, the authors noted that high rates of discontinuation meant that the potential benefits of sustained treatment were not available to many patients: the cohort in this study, who had been able to continue treatment for at least 12 months, represented only one‐third of the population originally randomised. Paradoxically, treatment gains were greatest for those with lowest quality of life scores at baseline –the very group with the highest discontinuation rate.

If they developed schizophrenia, most health professionals would choose to be treated with an atypical antipsychotic. A new study from London reveals which ones, and why (Psychiatric Bull 2007;31:94‐6).

Out of a total of 188 doctors, nurses and pharmacists working on mental health wards, 49 said they would opt for risperidone (26 per cent), 49 olanzapine (26 per cent), 35 aripiprazole (19 per cent) and 21 quetiapine (11 per cent). Eight (four per cent) would choose an older agent and five (3 per cent) said they would not want any treatment. Effectiveness was the principal criterion for 49 per cent and tolerability or safety for 45 per cent; nobody mentioned cost.

The authors note that their findings reflect current evidence and suggest that personal preference may be a sensitive indicator of treatment effectiveness.

Switching patients from an older atypical to aripiprazole is associated with improvements in lipid levels and loss of weight, according to a retrospective study by US psychiatrists (J Clin Psychiatry 2007; 68:406‐9).

They reviewed the medical records of 24 patients who had been switched from olanzapine, risperidone or quetiapine to monotherapy with aripiprazole. Ten patients were taking lipid‐ lowering medication and five were taking an oral hypoglycaemic drug.

After five to seven months' treatment with aripiprazole, mean total cholesterol levels decreased by 0.8mmol per litre and LDL‐cholesterol by 0.5mmol per litre; mean weight loss was 5.4kg. There were no significant changes in HDL‐ cholesterol, triglycerides or fasting blood glucose. Concurrent drug therapy did not affect these changes.

Predictably, these changes were most marked in the 15 patients switched from olanzapine –in fact, analysis of the remaining patients (switched from risperi‐ done or quetiapine) showed that only the change in LDL‐cholesterol remained statistically significant.

Encouraging as these figures are, the authors found relatively few patients who were treated with aripiprazole monotherapy at their Pharmacy during the study period. Of 133 who received a prescription for this drug, 29 were no longer taking it, four were non‐compliant and 56 were taking an additional antipsychotic. Aripiprazole may therefore offer benefits for some patients but the overall gains are less clear from this study.

The glutamate inhibitor riluzole, currently licensed for motor neurone disease, appears to improve treatment‐resistant depression (Biol Psychiatry 2007;61:822‐5).

Ten patients with major depression despite ongoing treatment with several antidepressants (mean duration of current episode, five years) received riluzole at an average dose of 95mg daily. After six weeks, scores of anxiety and depression were both significantly reduced; four patients met the Hamilton Depression Rating Scale (HDRS) criteria for response and three did so for remission. After a further six weeks, remission was maintained and one further patient achieved response criteria. The commonest adverse effect was fatigue.

Atomoxetine, currently licensed for attention‐deficit hyperactivity disorder (ADHD), is a selective inhibitor of noradrenaline reuptake that may cause anorexia and weight loss. Reason enough, US psychiatrists argued, to evaluate its potential in obese patients with binge eating disorders (J Clin Psychiatry 2007;68:390‐8).

Forty patients averaging four bingeing episodes per week, and with mean body mass index 37‐ 41kg per m², were randomised to placebo or atomoxetine 40‐120mg daily. Fifteen withdrew from the study (six atomoxetine, nine placebo). Binge frequency declined substantially in both groups, reaching a minimum after four weeks for placebo but continuing thereafter with atom‐ oxetine, becoming significantly lower after 10 weeks. Remission (cessation of bingeing) was achieved in 70 per cent of patients taking atomoxetine and 32 per cent with placebo. Placebo recipients did not lose weight whereas those taking atomoxetine lost about 3kg.

Adverse events, notably dry mouth, nausea, nervousness and insomnia, were common with atomoxetine and accounted for three withdrawals from the study, compared with one withdrawal in the placebo group.

The recommended way to begin treatment with quetiapine is to increase the dose to 300mg over four days then adjust up to a maximum of 750mg daily. US psychiatrists say an alternative rapid regimen may be useful for acutely ill patients (J Clin Psychiatry 2007;68:399‐405).

They compared the effectiveness and tolerability of the conventional regimen with rapid escalation to 800mg daily over four days in 40 patients admitted with acute schizophrenia. Of these, 12 were already taking an anti‐ psychotic.

About 90 per cent of patients in each group completed the two‐ week study, reaching final doses of 763mg daily with rapid initiation and 600mg daily with the conventional regimen. Adverse effects such as dizziness, sedation and dry mouth were more frequent with rapid dosing but the difference was not statistically significant. Symptom scores improved equally but more quickly with rapid dosing.

UK and Romanian investigators have reported another trial of the efficacy of cannabis in relieving the symptoms of multiple sclerosis (MS) (Eur J Neurol 2007;14:290‐6).

The placebo‐controlled, double‐ blind trial evaluated the cannabis‐ derived oromucosal spray Sativex (containing tetrahydrocannabinol and cannabidiol) in 189 patients with spasticity due to MS. Established treatments, other than nabilone and cannabis use, were continued. Spasticity was assessed using a numerical score on a rating scale from 0 to 10.

After six weeks, the spasticity score was reduced from a mean of 5.49 at baseline by 1.18 points with Sativex, and from 5.39 by 0.63 points with placebo –a statistically significant difference. The proportions of patients who experienced a ≥30 per cent reduction in spasticity were 40 and 22 per cent respectively, and 18 and 9 per cent respectively for a ≥50 per cent reduction.

Adverse events were reported by 82 per cent of patients using Sativex and 71 per cent with placebo; CNS effects (dizziness, impaired balance, attention distur‐bance, blurred vision) were more frequent with Sativex.

Recent concerns about the efficacy and safety of antidepressants in children have prompted a re‐evaluation of the evidence underlying their use. In a new systematic review (Acta Psychiatr Scand 2007;115:184‐95), US and Spanish investigators attempt to identify reasons for the different therapeutic response in children.

They note that neurotransmitter systems are not fully developed in children and adolescents; the adverse effect profiles of anti‐ depressants are dissimilar to those in adults; and pharmacokinetic characteristics are different. Notably, the placebo response is much greater among children than adults (50 vs 30 per cent) and childhood depression may be more heterogeneous. Finally, methodological shortcomings mean that clinical trials may be unable to detect small treatment differences.

Antidepressants improve outcomes after stroke by a mechanism independent of their effect on depression. One of these appears to be an improvement of executive function, say US investigators (Br J Psychiatry 2007;190:260‐5).

Their placebo‐controlled trial included 47 patients and showed that 12 weeks' treatment with fluoxetine or nortriptyline, started within six months of stroke, had no immediate effect on executive func‐tion by a variety of measures. However, 21 months later, patients assigned to placebo experienced a deterioration in executive function but there was significant improvement in those who had taken an antidepressant. There was no difference in outcomes between the two antidepressants.

The authors suggest that anti‐ depressants may influence monoaminergic modulation of cortical‐subcortical circuits or promote neurogenesis, and may be a useful adjunct in rehabilitation.

Some evidence suggests that cholinesterase inhibitors may reduce the severity of tardive dyskinesia (TD) in patients taking an antipsychotic. Galantamine, a cholinesterase inhibitor that also modulates nicotinic receptors, has not previously been studied for this indication (J Clin Psychiatry 2007;68:410‐5).

Thirty‐five patients with TD due to antipsychotic treatment for schizophrenia (696mg daily chlorpromazine equivalents) were randomised to placebo or galantamine 4‐12mg twice daily for 12 weeks; after a two‐week washout period they were then switched to the alternative treatment. TD was assessed by the Abnormal Involuntary Movement Scale (AIMS).

Mean total AIMS score and TD severity both improved more during galantamine treatment but the difference from placebo was not statistically significant. No prognostic factors, including use of first‐ or second‐generation antipsychotics, were found to influence the outcome.

However, patients who crossed over from galantamine to placebo experienced a small but significantly delayed increase in AIMS score. Furthermore, a small increase in parkinsonism occurred during galantamine use. The authors say the importance of these findings is uncertain but recommend surveillance of patients, particularly the elderly, who are taking both an antipsychotic and a cholinesterase inhibitor.

A small study has shown that the noradrenaline reuptake inhibitor reboxetine may improve motor function following stroke (J Neurol 2007;254:197‐201).

Ten patients with impaired fine finger movements and hemiparesis of three to four years' duration following a stroke received a single dose of reboxetine 6mg or placebo; two weeks later, they were switched to the alternative treatment. Motor function was assessed by grip strength and finger tapping tests. Reboxetine was associated with significant improvement in motor function of the paretic hand but not of the unaffected hand. Physiotherapy did not further improve outcomes.

Several small studies suggest that the response to an SSRI antidepressant may be influenced by variation in the SLC6A4 gene. SLC6A4 encodes for the serotonin transporter, a protein that terminates serotonin transmission and the molecular target for SSRIs.

According to new data from US investigators (Biol Psychiatry 2007;61:734‐42), based on analysis of 1914 participants in the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial, there is no association between the clinical response to citalopram and variation in SLC6A4. They could not exclude interactions between this gene and the environment but speculate that other genes in the serotonin pathway, or events dependent on serotonin function, may influence the response to treatment.

A meta‐analysis suggests that lithium greatly reduces the risk of suicide in patients with major depressive disorder (J Clin Psychiatry 2007;68:380‐3).

Analysis of eight trials involving a total of 329 patients showed that lithium was associated with an 89 per cent lower risk of suicide overall and an 85 per cent lower risk of completed suicides. The authors acknowledge that suicide may be under‐reported in studies when it is not a primary outcome, and that lithium is neurotoxic in overdose, but they suggest further studies are warranted.

The NMDA antagonist memantine appears to reduce craving in people with alcohol dependence, US and Russian specialists have found (Am J Psychiatry 2007; 164:519‐23).

Thirty‐eight hospital patients with a history of alcohol dependence were randomised in this placebo‐controlled trial. They had been sober for an average of three weeks and were not taking psychotropic medication. Four hours after taking a single 20mh or 40mg dose of memantine, they were told to smell and handle a glass of their drink of choice (but not drink it). Their craving was then assessed and after 6.5 hours they received relaxation therapy and a clinical briefing.

Memantine produced modest alcohol‐like effects but did not stimulate craving. It significantly reduced cue‐induced craving, the higher dose being more effective, and no adverse effects were reported.

Not all patients with Parkinson's disease are at equal risk of developing compulsive behaviour during treatment with a dopamine agonist. A North American study has now compared the characteristics of 21 patients who developed pathological gambling with those of 42 patients with no compulsive behaviours to identify possible risk factors (Arch Neurol 2007; 64:212‐16).

Those with pathological gambling were younger (mean age 60 yearsvs 66 years) and developed Parkinson's disease at a younger age (mean 51 years vs 58 years). They had a personal or family history of alcohol misuse, they were more likely to be novelty seekers, and to have developed mania or hypomania after starting treatment with a dopamine agonist.

Disease severity and levodopa dose were not predictive. Pathological gambling was associated with adjunctive use of dopamine agonists but not with monotherapy. Copyright © 2007 Wiley Interface Ltd