Editorial

Regulatory authorities are an essential component of safe prescribing. Their agility – their ability to respond quickly to new findings – is crucial to their effectiveness. However, quick responses preclude, perhaps, a full assessment of emergent data and the new data’s consistency or otherwise with prior observations. Regulatory guidance might in some cases be rushed out and, as a consequence, be partly or wholly inappropriate. Guidance on venlafaxine is a good example of this. In 2004, the UK regulator (Medicines and Healthcare Products Regulatory Agency, MHRA) issued an urgent safety restriction for venlafaxine apparently because of fears about its safety in overdose. The message was “don’t use this stuff: it’s dangerous”. The message was received loud and clear and use of venlafaxine dropped substantially. In 2006, the MHRA to all intents and purposes reversed its safety restrictions, sending the message - “oh, hang-on a tick, venlafaxine is probably no worse than anything else, after all”. The effect of this second instruction was negligible – venlafaxine prescribing rates hardly changed: the damage was done. Worse still, the only other Serotonin Noradrenaline Reuptake Inhibitor available, duloxetine, was tainted by its pharmacological similarity to venlafaxine. In this issue I review data on SNRI cardiac safety with colleagues Andrew Bradley and Alan Lennox-Smith. We show that data available before and after the MHRA safety restriction suggest good cardiac safety for both venlafaxine and duloxetine both in everyday use and in overdose. Venlafaxine’s slightly higher reported toxicity in overdose may well reflect the type of patient for whom it is used. It is notable that the recent safety guidance on citalopram and escitalopram is much less Draconian than that issued for venlafaxine. This may be a sign that the MHRA realises that when a drug’s safety is questioned, the impression of toxicity both sticks and spreads.

Quetiapine is a widely used drug with numerous indications, perhaps as a consequence of active metabolites. The parent compound has a short half-life of only a few hours which is effectively prolonged somewhat by the use of extended release preparations. Determining a therapeutic plasma concentration for quetiapine has proved impossible because trough levels are roughly the same whatever the dose. Simon Handley and colleagues from King’s College Hospital show in this issue that there is only the faintest relationship between dose and plasma level. Each dose produced a huge range of plasma concentrations. Ironically this near absence of correlation between dose and plasma level makes the determination of a target plasma level all the more important. Not that plasma level monitoring is a complete waste of time – 9% of samples had undetectable levels of quetiapine, indicating total non-compliance.

Also in this issue, Hidenobu Suzuki and colleagues show that memantine helps reduce both antipsychotic dosage and behavioural symptoms in people with severe dementia. Jan Terock and co-workers report on the successful treatment of prolactinaemia-associated obesity with cabergoline, bupropion and methylphenidate and, in so doing, suggest very different effects for dopamine enhancers (and dopamine itself) in different parts of the brain. Lastly, Anusha Baskaran and colleagues show that the dopamine antagonist (and, as it happens, SNRI) ziprasidone substantially improves sleep architecture in bipolar depression. We know that most mental illness is associated with sleep disturbance but this paper adds to the growing weight of evidence that sleep disturbance may contribute to symptom severity and that correction of this disturbance brings about symptomatic improvement.