Adding an atypical antipsychotic to treatment with an SSRI may –or may not –be effective in treatment‐ resistant depression, according to two studies reported from the USA (J Clin Psychiatry 2007;68:224‐36).

The paper reports two eight‐ week trials comparing olanzapine plus fluoxetine with both drugs as monotherapy in 605 patients for whom previous treatment (including fluoxetine) was unsuccessful. The primary endpoint was the change in Montgomery‐Asberg Depression Rating Scale (MADRS).

In one trial, there was no difference in MADRS scores between the groups, with reductions of 9.4‐11.0 from a mean baseline score of 30.0. In the second, the combination reduced MADRS score significantly more than either drug alone (14.5vs 8.6 for fluoxetine and 7.0 for olanzapine). The difference remained statistically significant in a pooled analysis.

The combination was associated with a greater increase in total cholesterol levels than either drug but the frequencies of other ADRs such as weight gain were similar for the combination and olanzapine.

Adding cognitive behavioural therapy (CBT) to treatment with imipramine reduces the frequency of adverse reactions (ADRs) and treatment discontinuation in patients with panic disorder, say US investigators (Am J Psychiatry 2007;164:273‐5).

During six months of treatment, ADRs typical of tricyclic anti‐ depressants were rated as no worse than mild. Nevertheless, CBT plus imipramine was associated with significantly less severe fatigue/weakness, dry mouth and sweating than imipramine alone. Thirteen per cent of patients taking imipramine alone discontinued treatment during the first three months due to ADRs; in those receiving combined treatment, 3 per cent discontinued during this phase and a further 8 per cent did so during treatment maintenance.

The authors acknowledge that CBT may have reduced symptoms associated with panic disorder that were perceived as ADRs but say their findings deserve further study.

Three new studies have provided reassuring data about the safety of certain drug treatments for women who are breastfeeding.

In the first of two Australian studies (Br J Clin Pharmacol 2007;63:371‐5), four women taking dexamphetamine 15‐45mg (median 18mg) daily for attention deficit hyperactivity disorder were nursing one infant each; the children were within acceptable weight limits and had developed normally.

Levels of dexamphetamine in milk were three times higher than in maternal plasma, giving an absolute dose to the infant of 21µg per kg per day, or 6 per cent of the adult dose. In three blood samples from the infants, dexamphetamine was undetectable in one and present at levels of 6 and 14 per cent of therapeutic concentrations in the other two. The authors say the findings support cautious use of dex‐ amphetamine during lactation but caution that longer‐term effects are unknown.

The second study (Br J Clin Pharmacol 2007;63:322‐7) involved eight women taking mirtazepine at a median dose of 38mg daily. The milk:plasma ratio was 1.1 for the parent drug and 0.6 for its active metabolite. Of four samples from the infants, mirtazepine was detected in only one at a level of 1.5µg per litre. No adverse effects were reported and the authors suggest that mirtazepine could be used safely in breastfeeding women.

USA specialists have reported on infant exposure to lithium (Am J Psychiatry 2007;164:342‐5). In 10 women taking an average lithium dose of 850mg daily and with an average trough serum level of 0.76mmol per litre, the concentration of lithium in breast milk was 0.35mmol per litre and infant serum concentrations averaged 0.16mmol per litre.

Overall, there was no evidence of abnormal renal or thyroid function in the infants, though levels of thyroid stimulating hormone, blood urea nitrogen and creatinine were slightly raised in several infants. The authors say their findings are reassuring for women who have to take lithium while breastfeeding.

Increased sympathetic activity occurs after acute stroke. It is associated with a poor neurological prog‐nosis and, experimentally, beta‐ blockers have been shown to reduce post‐stroke neurological deficits. In an observational study, US neurologists have now explored the impact of beta‐blockers on stroke outcomes (Neurology 2007;68:509‐14).

Of 111 patients admitted with acute ischaemic stroke, the 22 taking a beta‐blocker were broadly similar to those receiving other standard therapies except for a higher incidence of symptomatic ischaemic heart disease and a higher frequency of prior stroke. Use of a beta‐ blocker was associated with reduced stroke severity (according to the Canadian Neurologic Scale), even after adjustment for ethnicity, use of other drugs (statins, aspirin, warfarin), previous lacunar stroke and heart disease.

Patients taking beta‐blockers therefore have less severe stroke. In this study, beta‐blocker use was associated with lower cardiac sympathovagal tone but other contributory mechanisms may include reduction of acute inflammation, inhibition of thrombin formation or modulation of blood glucose.

Smoking cannabis reduces pain associated with neuropathy in patients with HIV to an extent comparable with that achieved by conventional analgesics, according to a study from California (Neurology 2007;68:515‐21).

Fifty‐five patients with HIV and confirmed painful neuropathy were randomised to use of cigarettes containing cannabis (3.6 per cent tetrahydrocannabinol, THC) or placebo (THC removed). Patients smoked one cigarette three times a day for five days, following a uniform puffing procedure to stan‐ dardise the dose.

Twenty five patients withdrew from the study, mostly because they could not tolerate the assessment procedure for acute pain. Among the remainder, more patients smoking cannabis had at least a 30 per cent reduction in pain severity (52 vs 24 per cent) compared with baseline. The median reduction in self‐assessed chronic neuropathic pain scores was twice as great with cannabis than with placebo (34 vs 17 per cent).

Adverse effects were more frequent with cannabis and included sedation, disorientation, confusion and dizziness, but no one stopped treatment as a result. Cannabis had no analgesic effect against acute pain.

Patients with a recent episode of self‐harm have lower levels of total cholesterol, LDL‐cholesterol and n‐3 and n‐6 essential fatty acids (EFAs) than controls with no psychiatric history, psychiatrists in Ireland have shown (Br J Psychiatry 2007;190:118‐22).

They randomised 49 patients presenting after an act of self harm to treatment with placebo or EFAs (eicosapentaenoic acid 1.2g daily plus decosahexaenoic acid 0.9g daily), in addition to usual psychiatric care (Br J Psychiatry 2007;190:118‐22).

After 12 weeks, treatment with EFAs was associated with significantly improved scores of depression compared with placebo and with more patients achieving a treatment response or remission (reductions of 50 or 70 per cent in rating scores).

Fewer patients taking EFAs reported suicidal ideation (though the difference fell slightly short of statistical significance) and per‐ceived stress. However, the frequency of self‐harm events was similar in the two groups and there were no differences in measures of impulsivity and aggression. A larger randomised trial is now warranted, the authors say.

London investigators have observed differences in the rate of deterioration in patients with probable Alzheimer's disease associated with drugs prescribed for chronic medical conditions (J Neurol Neurosurg Psychiatry 2007;78:233‐9).

Among 224 patients aged over 65 years who were living in the community, those treated with an antipsychotic or an hypnotic/anxiolytic were almost three times more likely to have deterioration of Alzheimer's disease (assessed by the Global Deterioration Scale) compared with those who were not taking these drugs. The combination of antipsychotic and benzodiazepine was associated with a nearly four‐fold increased risk.

By contrast, statins and ACE inhibitors were associated with substantially lower risks of deterioration of one‐third and one‐eighth respectively; this compared with a risk reduction of a half for drugs licensed specifically for the treatment of Alzheimer's disease.

Neurologists from Norway have flagged a possible complication of treatment with lamotrigine –sudden cardiac death (Acta Neurol Scand 2007;115:199‐203).

Noting that lamotrigine may affect cardiac potassium currents, they report four cases of sudden death in young women receiving monotherapy with the drug occurring at their outpatient clinic between 1995 and 2005. All were diagnosed with sudden death in epilepsy. There were other possible factors in each case –for example, low blood levels of lamotrigine, concurrent diabetes –and no deaths were observed. The authors call for a systematic review to clarify the possible risk.

In the first published trial of anti‐ epileptic drugs conforming to EU guidelines on non‐inferiority studies, specialists have compared levetiracetam and modified‐release carbamazepine as first‐line treatment for newly diagnosed adults with partial or generalised tonic‐clonic seizures (Neurology 2007;68:402‐8).

The proportions of patients remaining seizure‐free were 73 per cent in both groups after six months and about 50 per cent by intent‐to‐ treat analysis after one year. Of these, 80‐90 per cent achieved remission at starting doses (lamotrigine 500mg twice daily, carbamazepine 200mg twice daily).

More patients discontinued carbamazepine due to adverse effects (19vs 14 per cent) but this was not statistically significant. The authors say their study confirms uncontrolled observations that most newly‐diagnosed patients respond to low doses of their first anti‐ epileptic drug.

Melatonin could prove a useful hypnotic for patients with schizo‐phrenia, according to a study from India (J Clin Psychiatry 2007;68: 237‐41). In 40 patients reporting insomnia of at least two weeks duration, melatonin 3‐12mg at night improved sleep quality and duration, and reduced wakenings compared with placebo; it was not associated with hangover. Antipsychotic dosage (haloperidol 5‐15mg daily) was unchanged during the study.

The authors suggest that melatonin may offer a better tolerated alternative to a benzodiazepine.

Modafinil has been shown to reduce daytime sleepiness in patients with myotonic dystrophy so, researchers from The Netherlands asked, would it also improve activity (J Neurology 2007;254:26‐8)?

Their placebo‐controlled crossover trial in 13 patients evaluated 14 days treatment with modafinil 200‐400mg daily. The drug's effects on sleepiness were so obvious that most patients and their partners correctly guessed which treatment they were taking, so blinding was compromised. There was no apparent effect on spontaneous activity.

Valproate is associated with a higher risk of developing Parkinson's disease than other antiepileptic drugs –but the risk is lower than previously believed, according to a Canadian observational study (J Neurol Neurosurg Psychiatry 2007;78:147‐51).

Review of 201 patients attending a single clinic showed that postural tremor was the commonest movement disorder, with a prevalence of 45 per cent. The drugs most frequently associated with this disorder were valproate and carbamazepine.

Nine patients had parkinsonism, including six of the 59 patients taking valproate; this reflected a five‐fold increased odds of parkinsonism compared with other antiepileptic drugs. Symptoms were mild in all but one case, and this was the only patient in whom the possible diagnosis had been raised. The risk of parkinsonism was not related to duration of valproate use. Review of patients' records showed that concurrent therapies or comorbidities could account for parkinsonism in about half of cases.

SSRIs are associated with a higher risk of apathy in older patients than other antidepressants, Canadian investigators say (Ann Gen Psychiatry 2007;6:7).

Their case‐control study included 160 patients taking SSRIs and 224 taking other antidepressants discharged from a geriatric day care hospital. Using scales adapted from established depression inventories, they found that apathy scores were significantly higher among patients taking an SSRI and that these agents were associated with apathy at discharge but not at admission.

The authors admit they had no data on dose or duration of use, and their apathy assessment scales had markedly different thresholds for diagnosis. Overall, however, patients taking an SSRI were about twice as likely to be identified as having apathy. Other risk factors for apathy were age 70‐75 years and living alone. Copyright © 2007 Wiley Interface Ltd