Richard Lehman’s journal review – 27 February 2012

Richard LehmanJAMA  22 Feb 2012  Vol 307
813    When an Italian team of physicists reported that they had detected neutrinos travelling faster than light, the televisual physicist Jim Al-Khalili promised to eat his boxer shorts if it proved to be true. It turns out to have been a measurement error due to faulty wiring. Unbelievable results either shatter the laws of the known universe, or else they are wrong. So if a study tells us that 42% of women and 31% of men presenting with myocardial infarction do not have chest pain (or even pain in the arm or jaw), this either overturns clinical medicine as we know it – from experience and from several other large cohort studies – or else it is due to lousy recording. Guess which. This retrospective study is based on a single tick in a box completed by busy physicians looking after over a million patients coming into American hospitals with heart attacks between 1994 and 2006. It could be that they had better things to do than record the obvious. If these figures are true, then I will eat my elegant black Marks & Spencer long johns. These thermal underpants have proved very effective at protecting my lower parts from the ravages of winter on the eastern seaboard of America, and I just wish journal editors were as effective at protecting us against the ravages of bad data.

823   It’s a while since I was a proper GP, with a consulting room of my own and an obligation to endure baby clinics every so many weeks. As a result, I’ve completely lost track of what the standard infant immunization schedule now is in the UK. The poor mites seem to suffer an ever-increasing bombardment of antigens, but I no longer have to listen to them squawk somewhere down the corridor. So I will leave it to you to judge whether you need to know that “DTaP-IPV-Hib vaccination was associated with an increased risk of febrile seizures on the day of the first 2 vaccinations given at 3 and 5 months, although the absolute risk was small. Vaccination with DTaP-IPV-Hib was not associated with an increased risk of epilepsy.”

832    There was once an editorial in Gut which began, “The liver is a stupid organ. It can only grunt.” The point that this witty hepatologist was making is that estimating liver health from enzymes and ultrasound is like taking a history from a pig. Veterinary experts can no doubt pick up some diagnostic clues from the noises of their porcine clients, but it probably goes little beyond “oh dear, something’s the matter.” In this paper, the Rational Examination series revisits the question Does This Patient With Liver Disease Have Cirrhosis? In the best tradition of this generally wonderful series, 86 studies of patients referred to liver clinics are analysed in great detail and the findings are written up clearly and comprehensively. Astonishingly, alcohol history has no useful predictive value. The gold standard is liver biopsy, though even this can miss its target. All in all, this is a great contribution to the science of gruntology, but the end result is just a better kind of guessing.

NEJM  23 Feb 2012  Vol 366
687    This long-term report from the National Polyp Study made newspaper headlines the world over, which just shows that if you do lots of colonoscopies to remove colorectal adenomas for a period of up to 23 years, you may eventually get your 15 minutes of fame. More importantly, those you have treated will have half the rate of colorectal cancer of the general population. Since they were at greatly increased risk to start with, this is quite an achievement, and shows that a single colonoscopy per lifetime might reduce the population rate of colorectal cancer by up to 53%.

697    But the next paper brings us down to earth. A randomised non-inferiority trial of once-only colonoscopic screening compared with faecal immunochemical testing every two years has been going on in 8 regions of Spain since 2008. It seems that Spaniards are no different from the rest of us in their aversion to collecting poo (34.2% uptake) or having lengthy instruments introduced into their bottoms (24.6% uptake). The preliminary results seem to show a similar detection rate of cancers using the two methods, but a much higher rate of adenoma detection with colonoscopy. How to make screening more acceptable is the fundamental (sic) problem.

707    For patients with disseminated melanoma and their families, this is a tormenting period. Real advances are being made but they remain short of a cure; and they are also hideously expensive. Vemurafenib can extend the life of patient with metastatic melanoma by about a year, if it is one of the 60+% of tumours that has a mutation in the gene encoding the serine–threonine protein kinase B-RAF (BRAF). But the cost of a year’s vemurafenib is $113,000. And even if combination chemotherapy using this drug was developed which extended life beyond a year, there would be a high risk of developing other skin tumours. This phase 2 trial, paid for by Hoffmann-La Roche with company involvement at every stage, illustrates so much that is right and wrong with the current process of targeted drug development in cancer. If survival in advanced cancer is ever going to move beyond the prerogative of the wealthy, we need a whole new model of collaborative, global, not-for-profit cancer therapy research.

723   There is something deeply irritating about the names of skin conditions, which mix up Latin and Greek and seldom tell you anything useful. Lichen planus, for example. Moreover, we haven’t a clue what causes most skin diseases, and the treatment tends to be with local corticosteroids in the first instance, and oral corticosteroids if desperate. Lichen planus, for example. However, learned reviews of skin disease do tend to have elaborate accounts of molecular mechanisms and lots of sharp and lurid illustrations. Lichen planus for example; as reviewed in this article.

Lancet  25 Feb 2012  Vol 379
713    The Lancet, you may have noticed, is a seriously weird journal. One of the things it likes to do is publish the results of cutting-edge human experiments before they have any clear outcomes. Two subjects – one nearly blind with Stargardt’s macular dystrophy and the other nearly blind from age-related macular degeneration – underwent subretinal transplantation of retinal pigment epithelium cells derived from human embryonic stem cells. Four months later, not much happened. A great breakthrough, the editorial announces.

721    A few minutes after the Big Bang, when the Universe had cooled to two trillion degrees C, the first protons and neutrons formed out of the gluon-quark plasma and formed the nuclei of hydrogen and helium, together with tiny numbers of lithium nuclei. Wind on 360 000 years, when the Universe had cooled sufficiently for electrons to form stable orbits, and lithium as we know it came about. A certain amount has since been made – or recycled – by stellar nucleosynthesis. So what happens if you put lithium into people? This meta-analysis of 385 studies shows that it can be an effective mood stabiliser but carries the risk of reduced urinary concentrating ability, hypothyroidism, hyperparathyroidism, and weight gain. The risk of renal failure is very small but the risk of hyperparathyroidism is 10% and of the rate of hypothyroidism is increased more than fivefold. The authors recommend that TSH and calcium should be checked before starting lithium therapy and annually thereafter. They also cast doubt on the need for 3-monthly monitoring of serum levels. “Because few patients spontaneously develop toxic effects without a precipitating illness, yearly monitoring plus monitoring of one-off lithium concentrations in high-risk circumstances might be more clinically relevant and cost effective.” But this means better education of patients and doctors.

BMJ  25 Feb 2012  Vol 344
So if protons are of such ancient cosmic pedigree, what happens if we put proton pump inhibitors into people? By and large, they are happy: their gastric symptoms disappear and they want to keep taking them. But as an increasing proportion of the population take them, hidden harms begin to emerge. One of these is increased risk of fracture. This analysis of data from the Nurses’ Health Study shows that regular use of PPIs is associated with a 30% increase in incident hip fracture – increasing to over 50% in women who smoke.

For the 14 years I have been writing weekly reviews, Scandinavians have been engaged in furious argument about breast cancer screening and its contribution to the rise and subsequent decline of invasive breast cancer in those hardy regions. Here a group of Norsemen from Oslo, Bergen, and Trondheim try to promote our understanding of breast cancer trends in Norway, and conclude that “Changes in incidence trends of invasive breast cancer since the early 1990s may be fully attributed to mammography screening and hormone treatment, with about similar contributions of each factor.” But in a formidable editorial from Harvard, Karin Michels points out that “the authors do not discuss artefacts that can arise in ecological data and age-period-cohort analyses when non-linearities are present.” No doubt she is right, but I’m afraid I gave up at that point.

On to the whole question of hormone replacement therapy. It’s now ten years since the Women’s Health Initiative Study published its findings and reversed our notions of HRT and cardiovascular risk. The subsequent mass cessation of HRT caused endless hot flushes in doctors and women alike; but has the dust now settled sufficiently for us to take a balanced view of the immediate benefits versus the longer-term risks? Perhaps: and the verdict of this review is that HRT taken for five years at the onset of menopause is reasonably safe. It also confirms my feeling that we know too little about the risk/benefit ratio of hormone combinations other than the conjugated equine oestrogen/medoxyprogesterone acetate combination used in the WHI study.

Ann Intern Med  21 Feb 2012  Vol 156
I can’t think of many studies from British primary care that have appeared in the Annals, and this one receives a rave review in the editorial: “That this study was done at all speaks to the better support for high-quality research in primary care in the United Kingdom; finding support for a study like this would be extraordinarily difficult in the United States.” It’s not that easy in the UK either, and the praise is deserved: a 4-university collaborative did a cluster-randomized trial of cardiovascular family history taking by patient-completed questionnaire. It had a 98% uptake and showed that this is useful and practicable and resulted in a 4.5% increase in patients categorized as high risk.

Plant of the Week: Prunus cerasifera “Pisardii Nigra”

The earth is beginning to stir and buds are swelling imperceptibly on shrubs and trees. Among the first to open – maybe even as I write, in the gentler regions of England – will be those of this flowering cherry. It was an Edwardian favourite and is often abundant in the suburbs of that era, usually as a thirty-foot tree with an untidy trunk and dark purple leaves throughout the season. I would not recommend it for small gardens where every subject has to earn its place throughout the year, but it is a lovely sight in flower for a couple of weeks at the very end of winter.

There are plenty of these trees in the town where I used to practice, and I would look forward to their abundant pink blossom as I drove on my rounds. Wonderful when underplanted with snowdrops, or better still with scillas or blue chionodoxas. To make the tree less gloomy later in the season, train a vigorous white clematis into it – montana “Wilsonii” for May, or viticella Alba for August.