Richard Lehman’s journal review – 21 May 2012

Richard LehmanJAMA  16 May 2012  Vol 307
Do we all live on the same planet? I’m nearing the end of an amazing year at Yale, surrounded by superlatively intelligent people working on the outcomes of US healthcare. I myself occupy a space with brilliant newly qualified young doctors from India, Iran, and Brazil, putting together a book on patient-centred medicine. Most of the ideas we discuss about patient autonomy and shared decision making are quite new to them. All day long we work in cyberspace with people all over the world. When we leave the confines of our artificially lit space with temperamental air-conditioning, we blink in the sunshine and walk past black garbage-sifters begging for money, sallow drug addicts, and drunk disabled people shouting at each other. We share a street with these people, a town, a country, a world. But I cannot say that we really share anything with them, except perhaps for some coins we have in our pockets.

Which brings me to this JAMA issue devoted to global health.

2031   It begins, as it should, with an essay on Primary Health Care in Low-Income Countries: Building on Recent Achievements. This is a great piece: terse, comprehensive, and optimistic. It begins: “Small investments in improved health of the poor have a remarkable return in reduced morbidity and mortality. While the developed economies grapple with health systems that cost several thousand dollars per person per year…outlays of just a few dozen dollars per person per year in impoverished countries can add several years to life expectancy.” This is one of the reasons—together with the BMJ UK-India piece by Rao and Mant—that I put out this tweet:

“RCGP shd get trainees to spend extra year not in pampered UK but in xchanges to build 1ry care in India, China, Africa.”

2039   And to bring the issues really alive, read this narrative about the Lifeline Express by an Indian medical student: A Train of Hope, and a Chance to Train.

NEJM  17 May 2012  Vol 366
1859   Here’s a big international study to settle the question of whether warfarin or aspirin is better at preventing stroke and mortality in heart failure with sinus rhythm. Well, that’s what the title would have you believe. In fact this is a truly old-fashioned study which defines “heart failure” by an ejection fraction under 35 and so ends up with a cohort of patients of mean age 61 and 80% male. It tells you nothing at all about your average patients with clinical heart failure who have a mean age of 76 and are 50+% female, half with normal ejection fractions. Their chances of going into atrial fibrillation and throwing off clots are much higher than those of this cohort. Someone needs to do a trial comparing a fixed-dose new generation anticoagulant with aspirin in this “real world” population.

1870   Progressive multifocal leukoencephalopathy (PML) is a ghastly fatal disease caused by activation of the JC virus, most frequently seen after the use of natalizumab in the treatment of multiple sclerosis. Natalizumab is a humanised monoclonal antibody against the cellular adhesion molecule α4-integrin, and it seems that it has the property of converting this common and normally harmless virus into an aggressive CNS pathogen. It is also among the more effective of recent treatments for relapsing-remitting MS. In this study, the manufacturers of natalizumab combine different sources of observational data to conclude that the absence of anti-JC antibodies before the use of this drug may predict an extremely low risk for the later development of PML.

1881    This study about the increased risk of cardiovascular death while taking azithromycin has got a lot of publicity. The increase in CV death for the average adult population is put 47 per million taking a five day course of this antibiotic rather than amoxicillin: it is considerably higher in those at increased CV risk, of course. But there is an easy way to remember not to give this or any other macrolide antibiotic to those at highest risk: just heed your computer warning that all these drugs interact with statins.

1891   The paper that got the most publicity in this week’s New England Journal though was this one showing that coffee consumption is associated with lower all-cause mortality. This seems dose-related too, which is good news as I sit here all a-buzz from a mug brewed to the strength my wife likes. Legend has it that coffee reached the West via the Turkish armies who were defeated at the walls of Vienna in 1683, complete with bread rolls in the shape of the Islamic crescent. Which is still the best—and now perhaps the healthiest—breakfast; and the surest mark of advanced civilization.

Lancet  19 May 2012  Vol 379
1879    And still they come: industry-funded trials of amazingly expensive drugs for incurable cancers. This week there are two in The Lancet: this one looks at pazopanib, a tyrosine kinase inhibitor made by GlaxoSmithKline and compared with placebo in 72 institutions, across 13 countries on patients with angiogenesis inhibitor-naive, metastatic soft-tissue sarcoma. The drug costs about £2.3K per month, and life in the pazopanib group was extended by 1.8 months, within a mean survival period of just over a year. Diarrhoea and fatigue were very common side-effects. I see from the BMJ that the UK has just been ranked 12th in the European league of 34 health systems, and that the study leader, Arne Björnberg, marked us down because our access to new cancer treatments is “deplorable.” But then some new cancer treatments, and the prices charged for them, are themselves best described as deplorable.

1887   An Oxford-based group called IDEAL was recently set up to improve the methodology of surgical trials: this RCT of minimally invasive versus open oesophagectomy for patients with oesophageal cancer nicely illustrates the problems. Learning curves; case selection; end-points; inter-centre differences; duration; power—to name but a few. Here the groups totalled 57 and 59 each: there were two in-hospital deaths in the minimally invasive group and one in the open group, but far more of the latter got post-operative pneumonia. We will have to wait for longer-term data.

1893   Here is the second GSK cancer drug trial—a phase 1 safety and dosing study of dabrafenib, an inhibitor of BRAF kinase that is selective for mutant BRAF. It is always a mistake to get excited by phase 1 trials, but this is a drug to watch: it shrinks away a whole range of solid metastatic tumours, even including metastases of melanoma in the brain, which should be beyond its reach. There were no discontinuations due to adverse events.

After flirting briefly with items in the online first sections of the journals, I’ve largely gone back to the weekly printed items, for the sake of simplicity. But two papers on The Lancet’s website this week cry out for comment, as they put the final nails in the “treat-to-target” lipid management coffin. I hope.

The first one is old news, but repackaged as a huge meta-analysis of individual patient data from 27 randomised trials of statins which included subjects at low risk of cardiovascular events. Guess what? Everybody’s risk came down, whatever it was to start with. And the mortality benefit of the statin therapy was directly in proportion to the fall in LDL-cholesterol, and it far outweighed any measurable harms. So should everybody take a statin, so as to reduce the population rate of CV disease? No, I object to this kind of public-health-speak: everyone who wants to should be able to: it’s a personal choice. And does this prove that statins work by LDL-C lowering? Again no. They just work, and people should take them or not, as they wish to adjust their life chances.

And now onto the question of “good” cholesterol—HDL-C. There is a linear relationship between this lipid fraction and a decrease in cardiovascular risk. So far, 200 or more trials have been done with HDL-C raising agents; and not one of them has succeeded. Roche has just terminated its trial of dalcetrapib for futility, leaving the Oxford CTSU trial of anacetrapib (REVEAL) about the only one left standing. I wonder if it will ever recruit its 30,000 subjects. Here a massive Mendelian randomisation study shows that the HDL-C/CV protection association is unlikely to be causal, and plasma measurements of HDL-C may actually tell us very little.

BMJ  19 May 2012  Vol 344
A study from the Highlands and islands of Scotland looks at mortality in men screened for abdominal aortic aneurysm. I was somewhat fazed to see this defined by an aortic diameter of 30mm or more: in the MASS study and in my clinical practice, the threshold was 55mm. But this is only a fairly crude associational study which unsurprisingly shows that the diameter of the aorta in old men is a marker for high cardiovascular risk and cancer risk, almost all of it attributable to smoking.

Another observational study from the Hibernian regions looks at the outcomes of 1 271 549 bonnie lassies who produced wee bairns between 1981 and 2007. “At each gestation between 37 and 41 completed weeks, elective induction of labour was associated with a decreased odds of perinatal mortality compared with expectant management.” Nature is not a wise Mother. Human parturition is a chancy business, and it’s no wonder that by some estimates, the total population of Homo sapiens fell to 10,000 at one point before we learned to conquer stupid cruel Nature and take over the world.

Ann Intern Med  15 May 2012  Vol 156
673    Here’s how evidence-based medicine is supposed to work. We categorize a certain group of patients as having a condition in common: in this case it is chronic obstructive pulmonary disease. We conduct randomized controlled trials of interventions in a sufficiently large group of patients with the defining features of the condition, to determine what effect these have on important end-points which usually include death and hospitalization. Finally—and by no means always—we conduct further trials on methods of ensuring that these patients get the interventions which we have shown to be beneficial. A Comprehensive Care Management Program to Prevent Chronic Obstructive Pulmonary Disease Hospitalizations is an example of the final stage of EBM: implementation research. Comprehensive care management seems such a self-evidently good thing that it comes as little surprise that the study was terminated early. But the reason was that after a mean follow-up of 250 days, there had been three times as many deaths in the intervention group, and the same rate of hospitalizations. It seems to me that there could be two factors at work here. First, the interventions we encourage for COPD may be doing more harm than good, especially in the large number of patients with coexisting cardiovascular disease. Secondly, the effect of constantly reminding people of their disease status may be harmful in its own right—as demonstrated by the study which showed a fourfold mortality in sick elderly patients kept under close telemonitoring.

Arch Intern Med  16 May 2012
686     I am an avid eater of all kinds of fish, and I think my absolute favourite is a large fresh herring fried in butter. Unfortunately you probably have to live on the West Coast of Scotland to obtain such an article: away from fishing ports you are safest with kippers, as there is nothing worse than a stale herring. Eating is meant for sustenance and pleasure. It has also been found that eating a diet rich in fatty fish is associated with better cardiovascular health—and if there is any evidence against butter, I have yet to read it. Unfortunately a large proportion of mankind seems to be averse to eating the oilier species of fish, and instead, many seek to obtain CV benefits from omega-3 fatty acid supplements. But as this systematic review reveals, there is no evidence worth the name that these achieve anything for secondary prevention following cardiovascular events.

715    Several years ago I was intrigued to learn about the possibility of distinguishing between bacterial and non-bacterial infections using measurement of procalcitonin. But this is far more expensive in time and money that prescribing a course of antibiotics, and it has not caught on very widely. This observational study from primary care in Switzerland, France, and the USA shows that where it is used, antibiotic prescribing can fall markedly without affecting patient outcomes.

724   Finally, an insight into the bizarre world of American lipid prescribing: target-driven, irrational, and profoundly distorted by big pharma. Fenofibrate is a drug which has repeatedly been found to have no beneficial effects whatsoever, though it lowers “bad” lipids, including triglycerides. As evidence for its uselessness mounted, fenofibrate prescribing in the USA soared, driven by advertising from Abbott Laboratories. But this $1bn-a-year triumph of marketing over evidence was due to come to an end when Abbott’s patent expired. This paper describes how Abbott fought off the threat of generic competition and continues to sell modified fenofibrate products with rights to exclusivity. Its lead author is Nick Downing, a medical student with a unique accent forged in London and Harvard, whom I met when he first began work on this at the end of his first year as a medical student at Yale last year. Well done Nick! But hang on—what is this I see on the NEJM website? A full special report on regulatory agencies in the US, Europe, and Canada, with first author Nicholas Downing. For a second-year medical student to publish one paper in a leading medical journal might be called good fortune; to publish two begins to look like brilliance.

And do read this inspiring advice to Be Brave, from Nick’s mentor Harlan Krumholz.

Plant of the Week: Paulownia tomentosa

This delightful tree has vast leaves and wonderful panicles of scented blue foxglove flowers at this time of the year. It comes from China where it was discovered in the mid-nineteenth century and named for a Russian princess Pavlovna, who had moved to the Netherlands and was known there as Anna Paulowna. She would be forgotten now, but for this botanical tribute, being a mis-spelt version of her name with the Latin word for “hairy” added on. Perhaps the poor lady was not very popular in her new homeland.

Certainly the tree itself does not thrive in damp northern maritime climates. But I was astonished to see numerous paulownias in full flower on a recent rail journey into New York City. Wikipedia tells the story of how they came there:

In China, an old custom is to plant an Empress Tree when a baby girl is born. The fast-growing tree matures when she does. When she is eligible for marriage the tree is cut down and carved into wooden articles for her dowry. Carving the wood of Paulownia is an art form in Japan and China. In legend, it is said that the Phoenix will only land on the Empress Tree and only when a good ruler is in power. Several Asian string instruments are made from P. tomentosa, including the Japanese koto and Korean gayageum zithers.

The soft, lightweight seeds were commonly used as a packing material by Chinese porcelain exporters in the 19th century, before the development of polystyrene packaging. Packing cases would often leak or burst open in transit and scatter the seeds along rail tracks. This, together with seeds released by specimens deliberately planted for ornament, has allowed the species to become an invasive weed tree in areas where the climate is suitable for its growth, notably Japan and the eastern United States.