NEJM  3 Oct 2013  Vol 369
1295   The human birth canal, it has often been observed, is not particularly well designed for the passage of one baby, let alone two. So in an age of safety first, caesarean section has become almost routine for all twin pregnancies. Gut feeling and a number of cohort studies have supported the idea that this leads to better outcomes for the twins and perhaps the mother, but the Twin Birth Study claims to be the first trial in which mothers were randomly assigned to vaginal or operative delivery. Neonatal outcomes did not differ between groups. It’s a hard study to interpret though, because clinical judgement played a big role and only 44% of the women randomised to vaginal delivery escaped caesarean section.

1317   We definitely need better drugs to provide better outcomes in type 2 diabetes. But the pharmaceutical industry argues that so long as they can reduce HbA1c, their new products should be licensed for use on millions of people, and the hard outcomes can be collected later. And so it has been with dipeptidyl peptidase 4 (DPP-4) inhibitors saxagliptin and alogliptin. These drugs have been prescribed to hundreds of thousands of patients who were never told, “Actually, we have no idea what this stuff may do to you in the long term, but it will probably lower your blood sugar in the short term.” Well, now we know a bit more, because saxagliptin has been tested on 16 492 people with diabetes who are at high cardiovascular risk. In just over two years, it made no difference to their cardiovascular outcomes except admissions for heart failure, which were higher in the saxagliptin group. Moreover, the incidence of hypoglycaemia was higher than expected, in a group of drugs which is supposed to be relatively free of this risk. And the reduction in HbA1c was a measly 0.3%.

1327   At the same time, the rival agent alogliptin was undergoing a randomised double-blinded trial against placebo in diabetic patients who had recently had myocardial infarction. Good: the open-label methodology of the RECORD trial of rosiglitazone was not repeated here, even though the manufacturer had tight control over both the data collection and the write-up of this paper. A total of 5380 patients were followed up for 18 months and there was no difference in cardiovascular events. Nor was there an excess of pancreatic events in this and the previous trial, but it is too soon to relax vigilance: any increase in cancer would take much longer to show up.

1285   There is no ordinary editorial about these trials in the NEJM, but instead a Perspective at the beginning of the journal which is free for all to access, “The Cardiovascular Safety of Diabetes Drugs — Insights from the Rosiglitazone Experience.” Two of the three authors declare payments from NovoNordisk. They say that, “It is disappointing that neither intensive glycemic control nor the use of specific diabetes medications is associated with any suggestion of cardiovascular benefit. Thus the evidence does not support the use of glycated hemoglobin as a valid surrogate for assessing either the cardiovascular risks or the cardiovascular benefits of diabetes therapy.” Indeed; we’ve known that for five years. Nice to see the message is getting through. But earlier in the same piece they say that the new trials (and the RECORD trial!) show that the FDA is creating unnecessary work for itself and industry by insisting on trials to demonstrate cardiovascular safety before licensing drugs for diabetes. Pardon? The FDA “safety” margin indicates that diabetes drugs could be licensed if they show less than an 80% increase in CV events—is that too tight? For drugs that are meant to reduce CV risk in a high risk condition? Words fail me. What on earth are these drugs meant to achieve then? The answer comes in the saxagliptin paper: “Numerous studies, even those involving patients with advanced cardiovascular disease, have shown that improved glycemic control reduces microvascular complications.” But is this actually true? The evidence relating glycaemic control to real microvascular end-points, such as serious visual loss and end-stage renal failure, is very patchy, and involves numbers-needed-to-treat running into hundreds, with confidence intervals reaching infinity. The whole treat-the-sugar paradigm is wrong.

JAMA  2 Oct 2013  Vol 310
1353    The Women’s Health Initiative is often cited as the prime example of a big randomised trial which demonstrated harm after the observational evidence had suggested benefit. Women randomised to receive conjugated equine oestrogens (CEE) had worse cardiovascular and cancer outcomes than those randomised to placebo. Red-faced doctors (myself included) had to eat our encouraging words and advise droves of patients to come off hormone replacement therapy. A lot of the ladies then had red faces as their flushes returned, and nobody was very happy. And half of them were not taking CEE anyway, but oestradiol-based preparations, which for all we know may have a quite different balance of harms and benefits. Here, free to the world, is a follow-up report from the WHI. “Menopausal hormone therapy has a complex pattern of risks and benefits. Findings from the intervention and extended postintervention follow-up of the 2 WHI hormone therapy trials do not support use of this therapy for chronic disease prevention, although it is appropriate for symptom management in some women.” Methinks they generalise too much. How about a rerun of WHI using different a different oestrogen? Or a risk/benefit chart so women can decide for themselves?

Lancet  5 Oct 2013  Vol 382
1175    Speaking of generalising too much, how about this: “High volume prescribing of antibiotics in primary care is a major driver of antibiotic resistance.” It’s the opening sentence in the abstract of report of a trial which used internet training and CRP to reduce rates of antibiotic prescribing for respiratory infections in six European countries. What we really need to know is how high a volume of which antibiotics is needed to drive an increase in which kinds of resistance, and the best data on this come from a Europe wide study published in The Lancet in 2005. It clearly shows that antibiotic prescribing for RTIs in the UK at that time was associated with a 3% level of penicillin resistance in S pneumoniae, and we know that primary care antibiotic prescribing here has dropped since then. In France the situation was completely different: there was a resistance level of about 47%, and Spain was next worst at 37%. Belgium and Poland had about 12%. The trial here selected primary centres in countries with low, medium, and high levels of penicillin resistance and succeeded in lowering rates of antibiotic usage in RTI by the use of either rapid CRP testing or online doctor education, or both. This may be useful in countries with high levels of prescribing and resistance. The evidence does not suggest that GPs in the UK come into this category; but perhaps Sally Davies may be aware of something that I am not.

BMJ  5 Oct 2013  Vol 247
Clopidogrel was a massive earner for its manufacturer while its patent lasted, but like so many drugs which have cost health systems billions in long-term treatment, it may not actually do that much for most people who take it. A meta-analysis of nine trials in patients with established cardiovascular disease, the effect of clopidogrel was highly dependent on smoking status. Those who smoked showed a 25% reduction in a composite outcome of cardiovascular death, myocardial infarction or stroke, compared with a reduction of 8% in non-smokers. This may also apply to the newer agents ticagrelor and prasugrel, though with less certainty.

It has been said that the recommended limits of body mass index should be set lower for people of South Asian origin, because they store fat differently. But this study which pooled data from 20 prospective cohort studies was able to look at 3893 deaths in people from India and Bangladesh and was only able to identify a significant rise in cardiovascular deaths at a BMI level of 35 and above. This contrasts with people from further east (China, Taiwan, Singapore, Japan, and Korea) where the rise begins at a BMI of 25. In the “Far East” people really do seem to carry the “thrifty genotype.”

Plant of the Week: Fallopia baldschauanica

One reason that the reviews are a bit shorter than usual this week is that I am writing from Dresden, the Florence of the Elbe, which has many distractions to offer. The other reason is that the journals offer thin pickings. This city grows ever more beautiful as it continues to be reconstructed from the ruins of 1945 and the grey monotony of the DDR years. But many fine buildings remain neglected and unclaimed, and provide excellent growing space for the Russian vine, formerly classed as a polygonum.

It is probably best not to grow this plant unless you do indeed have a fine Baroque town house partially destroyed by the RAF to provide it with sufficient wall space and a suitably impressive backdrop. It fully deserves such a setting. At this time of year it is covered with a delightful froth of scented cream-white flowers with hints of pink, and very soon its abundant foliage will turn orange and crimson. Unwise people who plant it to hide a garden shed soon find themselves unable to find the shed, or to escape the wrath of neighbours who now find themselves invaded by the Russian vine.

Though in fact this plant comes from China. To hear real stories about Russian invaders, it is necessary to come to Dresden.

Next week I shall be in Canada and the USA, and the reviews will appear late.