Richard Lehman’s journal review – 10 September 2012

Richard LehmanJAMA  5 Sep 2012  Vol 308
869    Cancer, multiple sclerosis, stroke: do you want your patients to get the benefit of new drugs for these conditions as soon as possible? It’s pretty hard to say no to a question like that, but if you follow the flow of this rhetoric you can easily ignore poor evidence of benefit, and absence of evidence of safety. That’s what the authors of this short piece demonstrate in relation to vandetanib for medullary thyroid cancer, fingolimod for MS, and dabigatran for stroke prevention in AF. For all these drugs, the US Food and Drug Administration used its expedited approval program, as it did for 46% of the new drugs which came before it in 2011. All of them are expensive and in each of these three cases there are clear signals of harm: but they have been let loose on patients simply on condition that there will be post-marketing studies. The same happens in the UK, and we are about to lose what small protection NICE once offered, since now any manufacturer will be able to appeal against rejection and have a NICE decision overturned by an “independent” assessor picked by the Department of Health. This ridiculous travesty of proper regulation shows that nothing has been learnt from the lessons of Vioxx or Avandia. Pre-order your copy of Ben Goldacre’s Bad Pharma now.

871    The trials that drug and device companies run on their own products usually involve the recruitment of hundreds of subjects in dozens of centres around the world, and these subjects typically sign a lengthy consent form which points out the possibility of harm. People take part in clinical trials because they regard clinical trial data as a public good, and assume that if they suffer any ill effects as a result of their experiment, these will be carefully recorded and made public so that others can avoid them. But in fact clinical trials are typically reported in summary form in journal articles, where important harms may be under-reported and benefits exaggerated. For clinical trials to yield their full value, it is essential that independent analysts should be able to gain access to the full clinical study reports, and ideally to case report forms. This may seem like an issue far removed from clinical practice, but it is not. It is why we find ourselves time and again having to explain to patients why we are now withdrawing treatments that we previously thought were beneficial and now know may be useless or harmful.

875    Framingham is a town which is getting swallowed in the urban sprawl to the west of Boston, Massachusetts, and if you are very clever at negotiating its northern outskirts you will find yourself in wonderful woodland garden containing many gems of the spring season, especially trilliums. Oops, I digress. This is meant to be about the latest from the Framingham Heart Study, not Plant of the Week. Start again: if you want the best source of knowledge about the wild flowers of New England, or about the natural history of cardiovascular disease in a mainly white population since 1948, turn to Framingham. The Framingham Heart Study has generated about 2,400 papers since 1950, and if you piece them all together… well, you may be dead before you’re finished, like most of the original cohort. This latest paper explores the important question of aortic stiffness in the elderly: important because it is a predictor of total mortality and especially of heart failure without systolic dysfunction. The surprise for me here was that previous hypertension was not associated with late arterial stiffening. So it just seems to happen.

890    It’s one thing to recruit several thousand people in a Massachusetts town with a large Italian population, but just imagine if you had a few hundred thousand people inhabiting an island which few of them want to leave, descended mostly from the same few 1,000-year-old families. You could capture the entire population genome (this happened years ago) and do all the epidemiology you liked. ICELAND-MI does what it says on the side of the trawler: it looked for cardiac MRI (CMR) evidence of myocardial infarction in 936 participants aged 67 to 93 years, including 670 who were randomly selected and 266 with diabetes. Previously undetected MI showing up on CMR is about 60% commoner than recognized MI, and carries about the same prognosis. These sons and dottirs should all be on the standard post-MI drug cocktail.

898   Biologic response modifiers came into wide use for rheumatoid arthritis about a decade ago: abatacept, adalimumab, anakinra, certolizumab, etanercept, golimumab, infliximab, rituximab, and tocilizumab. See how long it takes you to memorize that list. This meta-analysis lumps them together for the purpose of determining the risk of malignancy from these “biologicals” and finds it to be the same as placebo or other disease-modifying drugs. But more time is needed for absolute certainty.

NEJM  6 Sep 2012  Vol 367
895    Castration is a question that has been written about ever since the invention of writing in Sumer 5,000 years ago. In ninth century Byzantium the castration of pre-pubertal boys assumed particular importance as a means of preserving and enhancing the singing voices of the finest choristers. This reached its height of popularity in the Roman Church between the sixteenth and nineteenth centuries. The singing of many of these castrati is reported to have been dazzling beyond imagination, but all we have to go by are the hoots and wails of the “last castrato,” Alessandro Moreschi, recorded in 1902-4. The timing of castration in localized prostate cancer is a much more difficult subject. The trial reported here shows that intermittent androgen suppression has functional benefits and no mortality difference compared with continuous androgen suppression over 7 years in men whose PSA had begun to rise above 3 following local treatment. But as the editorial on this study concludes: “Does early androgen-deprivation therapy in asymptomatic men with rising PSA levels provide more benefit than treatment in symptomatic men with metastases? This question bedevils our field, and we are no closer to an answer now than we were before.”

904    Now that castration has fallen out of favour, one of the most popular interventions we impose on pre-pubertal children is the prescribing of continuous inhaled corticosteroids for asthma. We know that inhaled budesonide has growth-retarding effects but these are supposed to even out over time. This report from a randomized controlled trial (CAMP) of inhaled budesonide versus nedocromil or placebo shows that this is not the case: a mean reduction in adult height of 1.2cm results from continuous inhaled steroid use in childhood. Make very sure that you are prescribing these drugs for good reason.

Lancet  8 Sep 2012  Vol 380
908   The art of dying is another subject with a very long history. Personally I would wish to be spared any necessity to maintain fortitude, dignity, repentance or any other kind of art while dying: if these began to run short, I would just like an easy way out. Most people who set up the means for assisted dying never use them. But I would like to have that bottle in the bathroom just in case, and I would not like to have my doctor face a sentence of 14 years for assisting suicide if I used it. Those who object to me and others having this choice often claim that no legislation which might permit this can protect the vulnerable; that assisted dying would not be needed if good palliative care were available; and that any societal move in this direction is going down a slippery slope. That none of this is true is easy to prove simply by looking at what has happened in countries where assisted dying laws have been introduced. Objectors should read this report from the Netherlands with care and an open mind and see if their doubts are in any way justified. It seems to me just the way any civilised secular society should operate.

BMJ  8 Sep 2012  Vol 345
Attempted suicide is 10-40 times commoner than completed suicide. Perhaps the most important purpose of any “medical” intervention after self-harm is to prevent one being followed by the other. Unfortunately, we have yet to discover any generic intervention that works. Here is a well-conducted trial of assertive outreach carried out in a wealthy, economically stable country (Denmark): of 243 people recruited after a first suicide attempt, 20 from the intervention group went on to make another attempt within a year, compared with 13 from the “usual care” group.

The paper which follows is from the UK and shows that completed suicide rates have been highest in the areas worst hit by our economic recession.

Some centres apparently try to reduce vomiting after tonsillectomy by the use of systemic corticosteroids. They shouldn’t: this meta-analysis shows that there may not be any detectable evidence of increased bleeding due to this practice, but there is an increased rate of hospital readmission.

Good negative news on two fronts: bariatric surgery is not associated with increased fracture risk, at least in the medium term; and migraine in women is not associated with cognitive decline.

Clinical reviews of irritable bowel syndrome are an unlikely source of reading pleasure, but this one is so nicely done that it should be perused by everyone who encounters this condition—i.e. pretty well every health professional, and a large swathe of the public. It deals very well with the uncertainties and limitations of current treatments, and to repeat my usual moan, it’s a shame it can’t go straight into Wikipedia.

Ann Intern Med  4 Sep 2012  Vol 157
336    Artificial hearts and kidneys work far less well than real ones, and the same applies to the artificial endocrine pancreas. At the moment, this takes the form of a continuous subcutaneous insulin infusion pump linked up to a blood glucose sensor which allows real-time feedback.  Here’s a systematic review which looks at various complete or half-way systems and how they compare with multiple daily injections for type 1 diabetes. As you’d expect, it’s just the devices which complete the feedback loop in real time which provide the best glucose control and the least hypoglycaemia.

Fungus of the Week: Cortinarius speciosissimus

This is the handsome web fungus which decorates the front of this week’s BMJ, and is the villain of the Patient’s Journey piece. Unless you go mushroom hunting in northern Scotland, you are extremely unlikely to see this Cortinarius, and if you do you should be able to distinguish it with ease from any edible fungus species. I can’t understand how the author of The Horse Whisperer can have mistaken it for a cep or a chanterelle, which it does not resemble in any way at all. It is fabulously toxic, especially to the kidneys, and the author and his wife were lucky to survive: even luckier that their children refused to eat any.

In good fungus seasons, all sorts of Cortinarius species make a rare appearance and then disappear for years. If you have a stamp collector’s interest in fungi, they are splendidly elusive prey; but if you are collecting for the table, they are all worthless and some are dangerous. Only the most beautiful—speciosissimus—is truly deadly. The safest plan is to avoid the lot, doing no more than take photographs of the ones which interest you.

  • ” it’s a shame it can’t go straight into Wikipedia.” — of course, it can, if someone with full-text access to the BMJ has the time to edit the appropriate Wikipedia article. While I’m reasonably proficient at editing Wikipedia, I don’t have full text access to the BMJ…

  • Richard Lehman

    It’s not quite as simple as that, unfortunately. You could trawl the piece for information, certainly, but you would have to be wary of copyright if you started cutting and pasting into Wikipedia. And often the most valuable features of these Clinical Reviews are their tables (certainly in this one) and illustrations, which are under copyright. Personally I would like to see these articles come out from behind the paywall, and for authors to be asked in advance if they would agree to their contents being used in Wikipedia – which is by far the most consulted source of medical information by patients and health professionals.

  • rpg

    I’m with you now. It’s an open access issue you’re flagging up.