NEJM  3 Nov 2016  Vol 375
Tolerating uncertainty
“At once it struck me what quality went to form a Man of Achievement . . . when a man is capable of being in uncertainties, mysteries, doubts, without any irritable reaching after fact and reason.” John Keats, December 1817

This quotation heads a wonderful short essay by Arabella Simpkin and Richard Schwartzstein about uncertainty in medicine. The year 1817 marked the point at which the young John Keats decided to leave medicine and take up poetry, and the “Man of Achievement” he was thinking of was almost certainly Shakespeare. The defining characteristic of Shakespeare (and of Keats in his finest works) is of being taken beyond the circumference of the individual mind to a place where feelings and words have no outer bound. Then Keats coughed up blood, and he knew he was going to die. A few years later, his illimitable gift was destroyed. But over the next 140 years, “irritable reaching after fact and reason” would result in a cure for tuberculosis. There would be no uncertainty left: a chest X-ray, a sputum smear, and then triple therapy. We love that kind of medicine: it gives us certainty and the power to save life. No wonder we want more of it. But most of what we do as doctors isn’t at all like that: I have often described my role as a GP over 35 years as being an uncertainty sump. This piece argues that this is the very essence of medical practice, and that we need to prepare doctors for it in new and better ways.

Horror stories of phase 1 trials
The phase 1 horror story that stays most in British minds is the cytokine storm, which hit six volunteers 10 years ago at Northwick Park when they were given an agent called TGN1412. But it is not a solitary example, and this week’s NEJM describes another. It was a dose-finding study of a potential new painkiller derived from the cannabinoid group. Previous single dose and 10 day dosing studies had been uneventful, but when another dose schedule was tried, an acute and rapidly progressive neurological syndrome developed in three of the four participants, starting on the fifth day of drug administration. The main clinical features were headache, a cerebellar syndrome, memory impairment, and altered consciousness. One of the group became brain dead, and the other two recovered although one had residual cerebellar symptoms and the other was left with memory impairment.

Beware new combinations too
For a few years now, we’ve been teetering on the brink of being able to induce permanent remission in metastatic melanoma by using the immune checkpoint inhibitors ipilimumab and nivolumab. Two patients in their 60s were given a combination of these drugs in the hope that they would be more effective in combination than singly. But both developed fulminant myocarditis and died.

Ribociclib for breast cancer
These tragic stories remind us how medical knowledge increases through the risks taken by others, and as Iain Chalmers reminds us, this means that there is a moral imperative to design trials which minimise risk and maximise clinical knowledge, and to disseminate the totality of their findings. In the cytokine storm case, for example, there was previous (hidden) knowledge of the risk and there was no possible reason why six volunteers should have been given it simultaneously. As I repeatedly complain, many new cancer drugs are tested in ways that make it very hard to determine their clinical utility, but are then marketed at phenomenal cost.
Here is a paper with the title “Ribociclib as First-Line Therapy for HR-Positive, Advanced Breast Cancer.” I know nothing about the treatment options for HR-positive advanced breast cancer, but this article tells me that “the duration of progression-free survival was significantly longer among those receiving ribociclib plus letrozole than among those receiving placebo plus letrozole, with a higher rate of myelosuppression in the ribociclib group.” Remember that, by and large, progression-free survival is a term without practical meaning for patients, whereas myelosuppression can lead to serious and distressing emergencies in the last months of life. The trial (median length six months) was not powered to determine overall survival, but there were four deaths (three [0.9%] in the ribociclib group and one (0.3%) in the placebo group) during treatment. I just don’t understand how this trial is supposed to inform clinical practice.

JAMA  1 Nov 2016  Vol 316
Hydrocortisone doesn’t prevent septic shock
Here’s a trial done in 34 German hospitals that took over four years, because the 380 participants were adults with severe sepsis but not septic shock. I really admire the effort that must have gone into identifying and getting informed consent from these very sick patients. And this trial falls into that all too uncommon but truly valuable category of one that addresses an important but difficult clinical question, which is of no possible commercial interest. Does giving intravenous hydrocortisone to septic patients prevent the onset of septic shock within 14 days? Rather to my surprise, the answer is no.

Released prisoners, psychotropic drugs, and reoffending
In Britain, the majority of people in prison are mentally ill, and the police are usually better at handling psychiatric emergencies than local mental health teams. Our press and our Home Office seem to accept these facts with indifference, while devoting much effort to the threat posed by refugee children. I don’t know if it’s the same in Sweden. A study of released prisoners there shows that rates of violent reoffending were lower during periods when individuals were dispensed antipsychotics, psychostimulants, and drugs for addictive disorders. As our overcrowded prisons become ever more violent and drug dominated, maybe we should just hand everyone a cocktail of quetiapine, methyphenidate, and buprenorphine, during confinement and after.

JAMA Intern Med  Nov 2016  Vol 175
Nutrifying your sponsors
“Is food industry sponsorship of nutrition studies associated with outcomes that favour the sponsor?” is the question addressed by this systematic review of 340 studies. The answer is not the resounding “yes” that you might have predicted. Instead it is a mild “perhaps,” because although there is a trend in that direction, it doesn’t reach statistical significance.

AAA in Australia
The abdominal aortas of older men in Western Australia have been the subject of study for more than 20 years, using electoral registry data to invite men aged between 64 and 83 for screening and elective repair in the context of a randomised controlled trial. This report shows that it has had no significant effect on AA aneurysm related mortality, in contrast with similar programmes in Denmark and the UK. The accompanying editorial provides a neat summary of the history of these trials and their context. Deaths from ruptured AAA have been tumbling in the developed world. In the US, there is an increasing tendency to move in and repair aneurysms below 5.5cm in diameter, despite any evidence of overall benefit, and everywhere endovascular repair is replacing open repair. Data from trials done in the 1990s may have little bearing on reality 20 years later. Evidence based medicine can only be as fresh as the evidence that it is based on.

Lancet  5 Nov 2016  Vol 388
A light to lighten our stenting
The stent wars are taking new directions and new surrogate outcomes are having to be invented. One of the latest is post percutaneous intervention minimum stent area. Don’t nod off just yet: there is lots more of this kind of thing to come before the end of this week’s reviews. It makes sense to believe that the better the lumen you create in the coronary artery, the better the end result will be. Like all sensible beliefs, this has a 50% chance of being true. Or false. Nobody will know until a randomised trial—in this case involving tens of thousands of participants—has been done. Meanwhile, the St Jude company has invented a stent guidance device, which uses optical coherence tomography (OTC) and has tested it against the alternatives of conventional angiography imaging and intravascular ultrasound. Apparently, it produces the same stent area as the latter and has some theoretical advantages. Maybe a randomised trial will follow. I knew you would be interested.

The stent that just biodegraded itself
“Very thin strut biodegradable polymer everolimus-eluting and sirolimus-eluting stents versus durable polymer zotarolimus-eluting stents in allcomers with coronary artery disease (BIO-RESORT): a three-arm, randomised, non-inferiority trial.” I know you can’t wait to hear my account of this article, which is currently lodged on the Lancet website. Sadly, I will have to disappoint you. Although the FDA approved the use of the first biodegradable coronary stents in July this year, it is unlikely to do so again in the near future. The three year data for the ABSORB II trial of those devices show that they are definitely not superior to existing stents and probably carry a higher risk for myocardial infarction. I’m not sure we will ever see these new ones in clinical action.

Whetting the appetite for CABG
Before the advent of widespread percutaneous coronary intervention, coronary artery bypass surgery had become the commonest kind of major surgery in the developed world. Nostalgic cardiac surgeons are still trying to find it a place. NOBLE was a prospective, randomised, open-label, non-inferiority trial, which compared PCI with CABG in the treatment of unprotected left main coronary artery stenosis. Despite being well powered, the trial showed no difference in mortality and few differences in other important outcomes. Although these tended to favour CABG, I’m not sure I would want my chest cutting open on the basis of these data.

The BMJ  5 Nov 2016  Vol 355
The fractional flow of premature adoption
The need for interventional cardiologists to keep intervening is not one that even The BMJ seems able to resist. Reading my paper copy this week, I was surprised to see an editorial stating that “It seems clear that initial angiographic assessment of patients with stable angina should be combined with pressure wire assessment when possible.” Why in the name of COURAGE do patients with stable angina even need angiographic assessment? And why does it seem clear to the authors that they also need this surrogate measurement, when they themselves point out that it is based on a single, unblinded, prematurely stopped trial?

Dual platelet drugs after stenting
I can’t help concluding that most non-acute stenting is done for no good reason other than having a catheter in the artery. And evidence for any useful difference between bare metal and drug eluting stents is scarce. We could do with a QALY based assessment of the cost to the NHS of the premature adoption of these expensive devices and the clopidogrel that had to be prescribed with them. At least we know from a new systematic review that the continuation of dual antiplatelet agents beyond a year after drug eluting stenting confers no extra benefit.

Vasectomy does not increase prostate cancer
In 1974 I was a student under the supervision of a very conservative urologist, who explained that vasectomy was an operation generally sought after by Guardian reading teachers or social workers. He expected his opinions to be regarded with vast deference, but I continued to read The Guardian and had my vas deferens* cut 10 years later. This roughly coincided with the first observational evidence of a link between vasectomy and an increased incidence in prostate cancer. I am glad that this caused me no concern whatever, because it is unlikely to exist. A new population based matched cohort study confirms this.
* Only referred to in the singular for the sake of a dreadful pun.

Plant of the Week: Vitis vinifera “Purpurea”
The rush to the bottom of the plant barrel is about to begin. November has arrived with freezing winds and soon everything will look dead.

Periods of sunshine still reveal wonders of turning foliage, and some of these are in plants which have foliage that is already purple throughout the growing season. As winter threatens, the leaves of Cotinus “Grace” and the purple leaved grape vine become translucent and turn bright red and orange, often for weeks before they are blown away. The vine will keep its shrivelling clusters of tiny dark grapes. Now is the time to bring them in and press them if you want to make verjuice, an intensely tannic and acid liquid that the Romans used in cookery. You can add it to fermented fish blood and bird droppings to embellish your dormouse based dishes.