Richard Lehman’s journal review – 20 December 2010

Richard LehmanJAMA  8 Dec 2010  Vol 304
   When cardiac troponin measurements came into use about a decade ago, it was immediately clear that they would change clinical practice and redefine ischaemic heart disease. By providing a simple biochemical indication of the degree of myocardial injury and death, they also provided us with a new marker for prognosis, not just in acute ischaemia but also in heart failure. Small elevations of cardiac troponin indicate progressive apoptosis of cardiac myocytes, while large elevations occur during acute necrosis. All the while levels of B-type natriuretic peptide mark the progression of ventricular strain. By combining serial measurements of troponin and BNP you can predict the likely death trajectory of heart failure with at least the accuracy of any cancer biomarker, but nobody does this in clinical practice, even though this has been known for five years or more. This study adds to what we already knew about the prognostic value of high sensitivity cardiac troponin T (cTnT) in a general population of older adults. I suppose we don’t use these tests because we simply don’t know how to respond to this knowledge.

2503   Here the same team looks at a different, more general cohort of patients from Dallas County to seek correlations between small elevations of cTnT and structural abnormalities on cardiac MRI plus all-cause mortality over 6.4 years. This marker provided significant extra correlation after adjustment for eGFR, BNP and high sensitivity CRP. Again, I don’t know what we will be doing any time soon with this information.

2513   Low-dose aspirin use is set to shoot up following the Rothwell study showing that it may prevent several cancers, including of course bowel cancer. But won’t this completely negate the value of bowel cancer screening using immunochemical faecal occult blood testing, turning a bloody crap test into a completely useless one? Actually, this South German study suggests the opposite: using two immunochemical tests for occult faecal blood, the area under the ROC curve for non-aspirin users is 0.65 or 0.67 – in either case hair-raisingly bad for a population screening test – becoming significantly less bad at 0.73 or 0.79 if the participants were taking aspirin. The all-important sensitivity for population screening goes from the 30% bracket up to the 55-70% bracket in aspirin takers. But how did these bloody stool tests ever get adopted as a screening test in the first place? I once described them as just a smelly kind of randomisation to undergo sigmoidoscopy, and I haven’t seen anything to change my mind.

NEJM  9 Dec 2010  Vol 363
   Once patients require renal dialysis, they leave our humble purview and enter the temples of nephrology, where the priests have the inestimable gift of certain knowledge. If we need to know anything, we ask them. If patients need to know anything, they ask them. Such certainty is rare in medicine, and much to be prized, so do not disturb yourself by reading this randomized study of in-centre haemodialysis six times a week versus three times a week and its accompanying editorial, suggesting a degree of basic ignorance about relative strategies that is positively normal.

2301   Heart failure is the commonest cause of readmission and an unchanging major cause of death in most developed countries. It is a progressive and generally unpredictable condition, but great interest has focussed on interventions that might help to predict decompensation and so help patients to alter their treatment in time to prevent the need for readmission, or perhaps even pull themselves back from the brink of death. Telemonitoring for changes in symptoms and weight is one way to go about this: do it daily and you are bound to spot trouble in time, ran the logic and the thrust of several small trials recently meta-analysed in a Cochrane review, showing a 44% mortality benefit. But nobody had done a sufficiently large trial with rigorous methodology, until this Yale study which shows zero benefit for either end-point. Learning points: don’t pay any attention to meta-analyses of inadequate trials (should they even be published by Cochrane?); what may seem obvious may not be true; negative trials can be invaluable for preventing the widespread adoption of futile interventions; heart failure is even more unpredictable than we had thought.

2310  Some time in the mid-2000s, there was a near-universal switch from bare-metal to drug-eluting stents, and throughout the known universe all interventional cardiologists put on the battle livery of either paclitaxel or sirolimus (unless they were bribed to join the army of some competing olimus). Costs soared and everybody was happy. Stent Wars: an everyday tale of Pharma Folk became the media legend of its time: a unique mix of science fiction, soap opera and medieval jousting. But now, alas, comes an alliance of Calvinists and Catholics to spoil the fun. This Swiss/Dutch/Austrian/Italian study shows that for stenoses in large coronary arteries, bare metal stents give the same results as sirolimus- or everolimus-eluting stents. Though it has to be admitted that the presence of an olimus did reduce the need for revascularisation procedures. And although sirolimus and everolimus seemed to be equally good in this trial, are they really? Don’t miss Stent Wars: The Christmas Special featuring guest star Zotarolimus and lots of bare metal.

2320   Pregnant mothers who take methadone give birth to babies with neonatal opioid abstinence syndrome. This can cause hyper-irritability and autonomic nervous dysfunction and need lengthy stays in hospital. Buprenorphine is a partial mu-opioid receptor agonist and has some antagonist actions too: in this double-blinded, double-dummy trial (with a predictably high drop-out rate) in opioid-dependent pregnant women, it seemed to result in a significantly milder neonatal abstinence syndrome.

2332   Tourette’s syndrome only hit my radar about 15 years ago: whether this was me being dim, or a change in diagnostic fashion, I simply don’t know. This good clinical review has little to say about the latter and I don’t want to examine the former. Suffice to say that nowadays Tourette’s is diagnosed in 1% of children, 50% of whom have relatively unobtrusive tics. In others, however, especially those with coexistent neurodevelopmental disorders, it can be very severe and hard to treat. Unfortunately the section on drug treatment bears out my worst fears about a lack of evidence: something I came to realise when trying to help a brain-damaged boy who was causing social havoc and self-harm through his severe compulsive behaviour. It’s a case of try this: then that: then perhaps another thing. “Experts” can be hired to do this in your place.

2339   The NEJM’s contribution to your festive season is the free full text of this article on genomics, type 2 diabetes and obesity. Personally, I’d suggest you try the Turkish Delight and that dubious liqueur your grateful patient gave you instead. Too much Oxford medicine these days seems to consist of unscientific hype and pie-in-the-sky: “The boundaries of personalized medicine will be much clearer in a few years, after large-scale genomewide resequencing efforts (now under way) provide a systematic, comprehensive description of the relations between genome sequence variation and major clinical phenotypes.” And then, dear doctor, you will be able to treat everyone as an individual! In fact you will be able to put their genome card into your computer and it will automatically prescribe all the 27 drugs they need. Unless they are ill, in which case they will need gene transfer therapy. As an antidote to all this, I earnestly recommend James Le Fanu’s Why Us: there is just time to get it before those dreaded days of festivity descend.

Lancet  11 Dec 2010  Vol 376
   Perhaps it’s not just Oxford medicine that is succumbing to its own unscientific oversell: here the MRC itself proclaims a general anticancer benefit from zoledronic acid in response to a marginal mortality benefit (95% CI for hazard ratio 0.74-0.96) in newly presenting multiple myeloma. For goodness’ sake, scientists, stick to what you have demonstrated! And for goodness’ sake, Lancet, do some editing! But I suspect I am pleading with the deaf. British Breakthrough Means End to Cancer: every reader is happy now.

2009   Here’s what proper reporting of a cancer study looks like: “As compared with idarubicin, mitoxantrone conferred a significant benefit in progression-free and overall survival in children with relapsed acute lymphobastic leukaemia, a potentially useful clinical finding that warrants further investigation.” I notice that the first five authors cited do not have medical degrees: perhaps that is why they talk like scientists.

2018   But just to prove that doctors can write well, here is a superb overview of sickle cell disease, a topic of global importance which can nonetheless seem deeply boring to those who see little of it and remember the haematology lectures they endured many years ago. This beautifully illustrated account points out that it is a potent source of brain damage in children as well as mortality in large parts of Africa; the old idea that it gives some protection against malaria may be partly true, but nonetheless people with sickle cells die fast when they get the disease, which is very frequently. Hydroxycarbamide is a therapeutic agent about which I’d like to know more: apparently it has cardiovascular protective effects.

2032   Risk assessment for recurrent venous thrombosis is a haematology topic closer to home for most of us, but it would be asking too much for it to be made into another really interesting read. I have given you some account of many of the studies mentioned here, most of them to do with thrombophilic markers which turn out to have no clinical utility. Maybe you will enjoy revisiting them here. My rule of thumb is that if there is an obvious cause for a clot, which has now gone (e.g. surgery), then the risk of recurrence is slight; if there is not such a cause, it is higher; and if there has been more than one clot, or there’s cancer around, then it’s very high. I’m not sure this review tells us all that much more.

BMJ  11 Dec 2010  Vol 341
   Here is an important analysis of data from the Multicentre study of Epilepsy and Single Seizures, which will be widely welcomed as showing that the risk of a recurrent seizure within one year is below 20%, whether or not treatment has been started. This means that all such patients can continue driving under current definitions of acceptable risk. Not legally, yet, in the UK: but the law should probably change, as this is the best (indeed the only) evidence we have on this issue.

1261  More good news for people with epilepsy, this time women who wish to conceive: the EUROCAT Antiepileptic Study Working Group has identified carbamazepine as an agent which is relatively free of risk for congenital malformations. There is a small increased risk of spina bifida when it is used in the first trimester, but such pregnancies should be detectable by enhanced screening procedures.

1263  Chronic scrotal pain is a condition which induces despair in both sufferers and doctors; and I suspect there was some answering discomfort in the lower regions of the urologist authors of this Clinical Review. They say that about 25% of it eludes diagnosis, though I would put it higher than that. Most men I have tried to help have been round several urologists until finding one who would detect some minor abnormality warranting surgery. They would then fail to improve and get to see a pain doctor who would open the usual pain doctor medicine chest. Some may even resort to the famous Chinese remedy for all ills of the male genitalia, Three Penis Wine. Early in my medical career I grew friendly with a family of Chinese restaurateurs, who once treated me to dinner, followed by a strange grey liqueur which occasioned much mirth. They never told me what it was, but thinking back, the taste was definitely right, and I have since fathered three children and been spared testicular pain.