On the Pulse

On the Pulse - 6th November 2009

Migraine and stroke 
A systematic review in the BMJ pools nine observational studies to report a doubling of risk of stroke in people who suffer from migraine. Further analysis seems to show that the risk is confined to women whose migraine was accompanied by an aura – especially if they were young, smoked, and were taking oral contraception. An editorial urges active management of cardiovascular risk factors in this group, but Dr Lydgate is sceptical. The findings are based on post hoc subgroup analyses of observational data, none of the relative risks is much greater than 2, and the absolute risk of stroke in young women, even those who smoke and take oestrogens, is tiny. Even if the findings are true and not the result of bias or confounding, vast numbers of women would need to be treated to prevent one stroke.

Insulin needs a benefactor 
A trial in the NEJM assesses insulin treatments for type 2 diabetes, comparing basal with prandial and biphasic regimens over 3 years. The results tend to favour the first two over the last, although, as an accompanying Editorial says, most of the patients needed a complex combination of regimens by the end of the study. But more problematically, the treatments were all insulin analogues manufactured by the trial’s pharmaceutical sponsor, not the human insulin recommended by guidelines, so the results are more relevant to its marketing department than to clinical practice. Who will pay for the study we really need?

Treatment of a single episode of demyelination
Almost by definition, a definite diagnosis of multiple sclerosis can’t be made until a second attack. Since early treatment is likely to be more effective than delayed treatment, should a patient with a single episode of demyelination be advised to start treatment immediately, or to wait until the diagnosis is confirmed? The results of an RCT of glatiramer acetate in the Lancet suggest the former. Glatiramer reduced the risk of developing clinically definite multiple sclerosis by 45% compared with placebo (hazard ratio 0•55, 95% CI 0•40—0•77) and lessened the burden of disease as monitored by MRI.

Cerebellar haemorrhage 
A 68-year-old man who had hypertension, diabetes, and atrial fibrillation, and was being treated with warfarin, was discovered snoring and unarousable at home. His score on the Glasgow Coma Scale was 3, his pupils were dilated and unreactive to light, and he did not respond to external stimulation. His INR was only 2.2 but a computed tomographic image of the head, available in the NEJM, showed a cerebellar haemorrhage extending into the midbrain and ventricles. Of course, one can’t be sure that warfarin was responsible, but the story is a reminder that prophylactic treatment may have serious harms as well as benefits.

Childhood UTIs and prophylactic antibiotics 
Recurrent urinary tract infections are common in children, particularly when vesicoureteral reflux is present. But do prophylactic antibiotics help? Several recent studies say not, but a paper in the NEJM describes a modest but significant preventative effect, particularly in those with more severe reflux. An accompanying Editorial thinks that some previous studies may have been underpowered or suffered from blinding problems. However, it’s necessary to treat 14 children to prevent one infection, and there are no data on the true outcome of interest, long-term renal damage.

Asthma update
Asthma is probably the commonest chronic medical condition affecting both children and adults. Many surveys have shown a striking increase in prevalence, at least in developed countries, over the past 40 years, yet we understand remarkably little either about its environmental causes or about the genes that determine individual susceptibility. A review in CMAJ reckons that detailed studies of gene-by-environment interactions and epigenetics will sort it all out, which seems optimistic in view of the slow progress so far. Setting that aside, the article is well worth reading for a concise update on what we do know about the condition.

"Second generation" antipsychotics in the dock again 
Regular readers of the column will recall that claims of greater efficacy and safety for the ‘atypical’ or ‘second-generation’ antipsychotics are rather doubtful. A cohort study in JAMA follows 505 patients aged 4–19 years treated with these drugs from 2001 to 2007. Though the exact metabolic changes differed somewhat by agent, all four drugs studied were associated with significant weight gains, from a mean of 4.4 kg with aripiprazole to 8.8 kg with olanzapine. An accompanying Editorial calls it a challenge to the widespread use of atypical antipsychotic medications in youth, and advocates more independent investigations of their safety and benefits.

Risks of splenectomy
A large Danish study in Annals of Internal Medicine confirms existing evidence that splenectomy is associated with increased risk of infection, even when the splenectomy is not carried out for a haematological malignancy. Among 3812 patients who underwent splenectomy from 1996 to 2005, the rate of bacteraemia in the first 90 days was 10.2%, vs 0.6% in general population controls, and there was a similar increase in risk for pneumonia. The authors call for more intensive preventive measures, including pneumococcal vaccination.