If you haven’t managed to open up an occluded coronary artery within 3 days of a myocardial infarction, should you go ahead and stent it anyway? Don’t bother, is the message of this multinational study: you may even harm the patient by trying. Or, to look at it a different way, just make sure you do your PCI as soon as possible after MI, as there is no rescue option later.

Famed for its orchestra and a Cream Cheese Spread, the City of Brotherly Love also gives its name to a chromosome which is involved in the pathogenesis of chronic myeloid leukaemia. The Philadelphia chromosome encodes the production of BRC-ABL tyrosine kinase, which is specifically blocked by imatinib. I mention these interesting facts because they may have led to a cure for this nasty condition. Nearly 90% of patients from the initial imatinib study are still alive after five years. And a similar genomic approach may bring cures for other myeloproliferative disorders – see the review on p.2452.

Antibiotic-associated haemorrhagic colitis is of course usually due to Clostridium difficile, however you choose to pronounce the beast. But now it is being joined by a fellow-beast, Klebsiella oxytoca. At least we can agree how to pronounce that. It was detected in 22 patients with post-antibiotic colitis who were negative for C-diff: the klebsiella was then given to rats and duly gave them haemorrhagic colitis, but only if they had been pre-treated with co-amoxiclav. So Koch’s postulates were handsomely satisfied.

We now have three oral drug classes for the routine management of type 2 diabetes, but which one should we ADOPT as our first choice for initial monotherapy? Metformin seems the obvious answer, since it is the only one which has been proved to improve outcomes by helping patients make better use of their own insulin. But it does tend to cause bloating and diarrhoea, so we often end up substituting sulfonylureas like glibenclamide (glyburide) which thrash the islet cells and in all likelihood burn them out faster. Then there are the thiazolidinediones like rosiglitazone which work on insulin sensitivity but can cause weight gain, oedema and even heart failure in the susceptible, and take a long time to kick in. The ADOPT trial, paid for by GlaxoSmithKline, shows that their drug keeps people going on monotherapy a little longer than metformin and a lot longer than glibenclamide. The confidence intervals just manage to part company at five years, ensuring a rosy future for GSK, which in any case hedges its bets by producing combinations of rosiglitazone with metformin or glimepiride.