Richard Lehman’s journal review – 16 May 2011

Richard LehmanJAMA  11 May 2011  Vol 305
1863    It’s hard to think of any physiological rationale for trying out a bolus dose of epoetin alfa following successful percutaneous intervention for myocardial infarction, but apparently it reduced infarct size in animal models, so this trial gave it to humans. It did not work; in fact it may have increased infarct size in some older patients. Another idea tested and destroyed, and another trial with the acronym REVEAL joins the list of those which were worth doing but can now be forgotten.

1873    The mapping of the human genome was a wonderful scientific triumph, achieved ahead of time. But it has proved remarkably feeble as a driver of therapeutic progress, except in the analysis of malignant cell lines and their response to established chemotherapy. I will spare you the detail, as you are most likely not someone with the responsibility of treating newly-diagnosed invasive breast cancer. It’s the principle that counts – “A genomic predictor combining ER status, predicted chemoresistance, predicted chemosensitivity, and predicted endocrine sensitivity identified patients with high probability of survival following taxane and anthracycline chemotherapy.” You may struggle with the syntax here but get the general meaning – what’s really difficult to predict is how this information will eventually be incorporated into the cost-effective management of breast cancer.

1882    The COURAGE trial, reported on 4 years ago, was a true landmark in the management of stable coronary artery disease, demonstrating that the outcomes of optimal medical management are as good as those of coronary artery reperfusion intervention. In an ideal world, this would have been followed by an immediate steep drop in reperfusion procedures. But as we saw in last week’s JAMA, no such drop occurred. Worse still, as this important study shows, doctors in the USA are still prescribing optimal medical therapy to fewer than half of patients who undergo PCI for stable angina. This is a classic example of the non-adoption of evidence-based medicine when it conflicts with an existing paradigm. It would be interesting to see some UK data: but the fact is that to perform elective PCI on someone who has not been offered optimal medical therapy is just bad practice, and is arguably unethical.

NEJM  12 May 2011  Vol 364
1795   Whenever the New England Journal publishes a paper about original non-clinical research, you can be pretty certain that it will be of fundamental importance. Past studies have demonstrated the existence of stem cells capable of organ repair in the heart and the brain: this one establishes the existence of such cells in the lung. They have already been discovered in the mouse, where they are capable of differentiating into new lung structures: these investigators have grown human lung stem cells for the first time, and watched them divide asymmetrically and generate bronchioles, alveoli, and pulmonary vessels. It’s the kind of thing that makes you want to buy a Carpathian castle, steal corpses, and harness lightning: but be warned from the experience of trying to repair other body organs that simply knowing that it might be possible is a very different thing from making it happen.

1817   And now for a very different kind of research: the sort that achieves little increments of benefit to patients dying of metastatic cancer. Few disseminated cancers are as unremitting as carcinoma of the pancreas, for which a common treatment is gemcitabine, which achieved a median survival of 6.8 months in this trial. The new chemotherapy regimen it was compared with here carries the acronym FOLFIRINOX – a combination consisting of oxaliplatin, irinotecan, fluorouracil, and leucovorin. Survival in this group reached a median of 11.1 months and quality of life at 6 months was much better than in the gemcitabine group, but adverse effects were nonetheless worse overall with the combination chemo.

1826    In today’s world of shared decision-making, patients who are asked for consent to surgery should be shown a video of the procedure plus the comparative outcome figures for this as opposed to any alternatives, including no treatment. Jack Wennberg pioneered this approach in the early 1990s but few have followed him, despite the dramatic reductions in surgery rates which followed. I think that most women would be put off if they watched videos of anterior colporrhaphy and transvaginal mesh insertion. The latter involves piercing the tenderest parts of their anatomy with a trocar resembling something one might use to suture a hippopotamus. Nonetheless, it has become a very popular day case procedure and probably has its place in the management of stress incontinence. This Swedish comparative study gives doctors and patients all the data they might need to judge between the two options – mesh is quicker and gives better short-term results, colporrhaphy takes longer but carries a lower risk of complications.

Lancet  14 May 2011  Vol 377
1655    Several of my patients have benefitted from the early adoption by an Oxford radiologist of the endovascular platinum coil for the treatment of intracranial aneurysms. As he writes in the editorial that accompanies this study, these work best in middle-sized aneurysms where complete occlusion is easiest to achieve. But the makers of a new hyrogel-coated coil, MicroVention Inc, would have liked this trial (HELPS) to prove that their self-expanding coil worked even better. In fact it did not, though they have managed to squeeze some evidence of superiority by using a composite end-point of death, disability and radiographic recurrence. The oddest feature of the trial was the higher incidence of late hydrocephalus with the new device. But for a change, I won’t lambast The Lancet for publishing this industry-sponsored study: it is rare enough for a device manufacturer to submit to rigorous comparison of any sort, as you’ll learn if you watch Channel 4 on Monday.

1681   The Lancet aka R Horton has chosen some distinctly odd hero-figures in the past, but I would entirely agree in his journal’s praise of the principal author of this article about medical treatment in acute and long-term secondary prevention following TIA and ischaemic stoke. Peter Rothwell gets his Lancet encomium, and fully deserves it: in fact the piece omits to observe what an excellent and thorough clinician he is as well as a great researcher. And that is the source of his strength – observing what actually happens to people, and how small increments of better management can add up to big differences to outcomes. That’s one reason I like figure 1 of this paper, the other being that it reflects very well on primary care in my home locality. It’s worth taking some time reading this paper: you will learn how marginal the benefits of non-aspirin anti-platelet agents are, for example, and you may be surprised to learn that drugs which block beta-adrenergic activity actually increase blood pressure variability – and thus fail to prevent stroke. Also, remember that to be of best use, carotid endarterectomy has to be done within two weeks of a minor event.

1693   Alas, the same optimism and methodological rigour do not characterise the field of stroke rehabilitation. But the authors of this review make the best fist they can of their fuzzy evidence base, and interestingly observe that “several cohort studies suggest that recovery of body functions and activities is predictable in the first days after stroke.” There are lots of studies comparing various integrated care pathways with “usual care”, but none that indicate a convincing superiority over a well-functioning multidisciplinary team.

BMJ  14 May 2011  Vol 342
1065   We know so little about the causes of miscarriage that any new piece of information is worth mulling over. The immune system is an obvious place to start when looking at a rejection phenomenon, and we already know that antiocardiolipin antibodies are associated with an increased risk of miscarriage: this meta-analysis of the evidence around thyroid autoantibodies demonstrates a substantial (2-3x) increase in risk of miscarriage and doubled risk of preterm birth.

1068    From time to time it is necessary to point out to cardiologists and chest physicians that the organs they are interested in share a common body cavity and are capable of influencing one another. In fact if you read Heart you might be forgiven for thinking that cardiologists suffer from hemispatial neglect syndrome causing agnosia of half of the heart and its functions, and that most of them believe that the organ has two chambers. When’s the last time you read about the right ventricular ejection fraction? About the circulatory consequences of chronic obstructive pulmonary disease? I say all this because otherwise you may be surprised to learn that β-adrenergic blocking drugs increase survival in severe COPD by about the same factor (20-25%) as they increase survival in left heart failure. And yet the moment you hear a wheeze or squeak in these patients you stop any β-blocker they may be taking and give them a long-acting β-stimulant instead. In this retrospective cohort study as in others, the patients were allowed to remain on concurrent inhaler therapy, so some of the mortality benefit may have come from one drug blocking the action of the other. Oh what a tangled web… we clearly need some prospective studies done by people with full spatial awareness of the thoracic contents and their physiology.

Arch Intern Med  9 May 2011  Vol 171
831    Now speaking of the physiology of the thoracic contents, what are the lungs there to do? Oxygenate the blood – OK. Anything else? Clue: what happens when you get a right-to-left cardiac shunt? Yes, all sorts of tiny rubbish can get into the systemic circulation, causing cerebral or systemic emboli of clot, air or fat: normally these are trapped in the lungs, which therefore act as a sieve. For all we know, we may be having tiny pulmonary emboli most days of the week: the point being that they do nothing and don’t matter. “The Lung as a Sieve”: now there’s a good title for an interesting paper. Meanwhile, I commend to your attention this excellent contribution to the Less Is More series, demonstrating that since the widespread adoption of computed tomographic pulmonary angiography, the diagnosis of PE in the USA has gone up by 81%, the decline in mortality has not changed, case fatality of PE has diminished proportionately, and complications of anticoagulation for PE have increased. In other words, pulmonary embolism is being overdiagnosed. It’s not that PE isn’t happening – it’s just that we can now see lots of little clots that don’t matter.
Plant of the Week: Paeonia suffruticosa “Tria”
This plant is the culmination of about one and half millennia of peony breeding in China, and later in France and  the USA. The ultimate ideal – at least for the last 150 years – has been to mix the genetic material of herbaceous and shrubby peonies so as to create well-shaped flowers of pure clear yellow on stems capable of bearing them elegantly. This is a slow and fiendishly difficult process, and whenever previous attempts have neared the goal (e.g. “Souvenir de Madame Cornu” or “Argosy”), wealthy gardeners have paid prices of $1,000 or more to get hold of them. But they flower for about one week a year and the great flower-heads flop and are easily damaged by rain.
“Tria” has lovely cut leaves which emerge purple and turn clear green. The flowers tend to be borne in threes and emerge successively over about three weeks. They are of the most beautiful clear yellow – pale but not washy. They are held well. Pay up your £80 with a light heart – this plant will give you joy for life.

Greek To You

Last week a BMJ clinical review mentioned keratomileusis – any idea what that is?

In this week’s Lancet, there is casual mention of:

Do you understand?