Richard Lehman’s journal review—4 April 2016

richard_lehmanNEJM 31 Mar 2016 Vol 374
Going Dutch with Lyme
1209 When you open a journal with New England in its name and read about Lyme disease, you somehow don’t expect the study to have been carried out in the Netherlands. But never mind: it’s a good study and the investigators managed to collect enough cases of genuine Lyme disease from those densely inhabited flatlands. They gave a couple of weeks of intravenous ceftriaxone to 281 patients with persistent symptoms attributed to Lyme disease: musculoskeletal pain, arthritis, arthralgia, neuralgia, sensory disturbances, dysaesthaesia, neuropsychological disorders, or cognitive disorders, with or without persistent fatigue. They were then allocated randomly to receive 12 weeks of doxycycline, or a combination of clarithromycin and hydroxychloroquine, or placebo. There was no difference between the three groups in terms of symptom resolution. Whatever causes this disabling syndrome, it doesn’t seem amenable to longer antibiotic treatment.

Idolatry of the carotid surrogate
1221 Idols come in all shapes and sizes. Often they have an exciting range of appendages, like many-breasted Diana or many-armed Kali or super-endowed Priapus. Idols tend to be more interesting than ordinary life, where people can be boring and you just have to wait for stuff to happen. You may remember that one million women are being studied over decades in the UK to find out what happens to them if they do various things in real life. But by using idols—known in medicine as surrogate outcome measures—you can get to answers much more quickly. Mind you, it still takes time—five years was the median in this trial. A total of 643 healthy postmenopausal women were stratified according to time since menopause (<6 years [early postmenopause] or ≥10 years [late postmenopause]) and were randomly assigned to receive either oral 17β-estradiol (1 mg per day, plus progesterone [45 mg] vaginal gel administered sequentially [i.e. once daily for 10 days of each 30-day cycle] for women with a uterus) or placebo (plus sequential placebo vaginal gel for women with a uterus). The surrogate measured was carotid-artery intima–media thickness (CIMT), which I myself thought might be the answer to the time problem in cardiovascular outcome studies when it came into fashion 20 years ago. But no. It’s just another idol. “Oral estradiol therapy was associated with less progression of subclinical atherosclerosis (measured as CIMT) than was placebo when therapy was initiated within six years after menopause but not when it was initiated 10 or more years after menopause.” But as far as real women are concerned, that’s just hypothesis-generating. And we already had that hypothesis. Carry on.

Where does baricitinib sit among the nibs?
1243 Drugs whose names end in -inib have so far tended to be orally available TNF-α inhibitors, but it’s no longer so simple. The -inib suffix just stands for “inhibit”, innit? So it can denote all kinds of inhibitors, and baricitinib inhibits the Janus kinases JAK1 and JAK2, which are prominent in many inflammatory processes. Now let’s move on to rheumatoid arthritis, where there are lots of inibs already on the market. Where does this baricitinib fit in? I have absolutely no idea. All we now know is that “In patients with rheumatoid arthritis and an inadequate response to biologic DMARDs, baricitinib at a daily dose of 4 mg was associated with clinical improvement at 12 weeks. (Funded by Eli Lilly and Incyte).” It’s news to me that rheumatoid arthritis stops at 12 weeks.

When opioids are religion for the people
1253 The people I follow on Twitter, and the e-mail lists I get messages from, fret frequently about the problem of iatrogenic opioid dependency in North America. And it is a mighty problem: about 10 million US citizens take opioids on a long-term basis for non-cancer chronic pain. This review struggles to find answers. Anyone who has been in generalist medicine knows the options: psychological treatments, non-opioid analgesics, gabapentin, tricyclics, TENS machines, etc. The trouble is that the pain seldom goes away with any of them. Guérir quelquefois, soulager souvent, consoler toujours: if these are the three duties of the doctor, then the guy who invents new and effective ways to soulager will be among the world’s greatest benefactors.

JAMA 29 Mar 2016 Vol 315
Duelling over dual antiplatelets
OL There is no print JAMA this week, but if you fancy a paper that will take you as long to chew through as any ordinary issue of the journal, you can find this free one on the website. Papers like this are an acquired taste. I have just eaten my way through a large bag of whelks. They are meaty but hard to extract, they taste of rubber and estuarine mud, and they contain particles of grit and sand to wear down your teeth. I love ’em. And I rather love this article for all the careful work that went into it and the text and graphics that have resulted. To a degree it encapsulates how the future of medicine might look, but I can’t say it’s easy going. It sets out to develop and validate a prediction rule for the benefit and harm of dual antiplatelet therapy beyond one year after percutaneous coronary intervention. It is very modest in its conclusion: “Among patients not sustaining major bleeding or ischemic events one year after PCI, a prediction rule assessing late ischemic and bleeding risks to inform dual antiplatelet therapy duration showed modest accuracy in derivation and validation cohorts. This rule requires further prospective evaluation to assess potential effects on patient care, as well as validation in other cohorts.” It’s by this kind of self-effacing, contingent effort that we can start to draw conclusions which inform decision-making with individual patients. The future lies in doing it more efficiently, using smarter analytic software and better data. And the end-product will be interactive graphics that make perfect sense to both clinicians and patients. We are certainly not there yet, but studies like this edge us closer to that great goal.

JAMA Intern Med April 2016
Hip diet
OL I suppose the “Mediterranean” diet has been around for too long to be considered hip any more. It’s a nice middle-class array of foods, some of which are often eaten around the Mediterranean basin. There have been lots of observational studies of this diet and even the odd randomized trial of uncertain rigour. “Healthy Eating” Indices have been derived from these data, though naturally they tend to have a short lifespan. Here’s a study based on food diaries from the Women’s Health and it concludes, “Higher adherence to a Mediterranean diet is associated with a lower risk for hip fractures. These results support that a healthy dietary pattern may play a role in maintaining bone health in postmenopausal women.” Ah yes, and the number needed to treat is 342. So is diet an essential part of public health strategies to reduce hip fractures in older women? I guess public health physicians will continue to say so, on the basis that if 342,000 women changed to a Mediterranean diet, the NHS might have 1,000 fewer hip fractures to fix. Bless them. It’s the way they’ve been trained.

Unnecessary test, unnecessary treatment, lotsa $
OL I’ll just give you the summary of this study from the USA, where tests and procedures mean dollars. Couldn’t happen here:
“In the largest contemporary US cohort reported to date, most patients undergoing diagnostic catheterization before noncardiac surgery are asymptomatic. The discovery of obstructive coronary artery disease is common, and although randomized clinical trials have found no benefit in outcomes, revascularization is recommended in nearly half of these patients. The overall findings highlight management patterns in this population and the need for greater evidence-based guidelines and practices.”
Shouldn’t that last bit read “disciplinary action and punitive fines”?

Lancet 2 April 2016 Vol 387
The fat of the lands
1377 “Over the past four decades, we have transitioned from a world in which underweight prevalence was more than double that of obesity, to one in which more people are obese than underweight, both globally and in all regions except parts of sub-Saharan Africa and Asia.” So this is a wonderful achievement, isn’t it? Because humans can cope much better with overnutrition than undernutrition, as the epidemiology of cardiovascular disease clearly shows. It’s the places where mean BMI is still less than 25 that we need to worry about.

Bevacizumab for mesothelioma
1405 There are few patient journeys sadder than those of people with mesothelioma. In this study, the group receiving standard treatment after first diagnosis survived for a mean of 16.1 months, while those who received additional (open-label) bevacizumab stayed alive about 11 weeks longer. This is not going to be an easy choice for people who have just been told they are going to be dead within two years. Theoretically, this trial may represent the sort of incremental progress which has led to eventual breakthroughs in cancer treatment, but for an individual it means deciding whether those extra weeks are worth a decrement in quality of life. I wonder how often these issues are actually discussed honestly with patients.

Molecular profiling in lung cancer
1415 Still, everything will come right when we can tailor treatment to the molecular characteristics, won’t it? Maybe; but don’t hold your breath. Here’s the report of a nationwide programme for the systematic routine analysis of EGFR mutations and ALK rearrangements as well as HER2, KRAS, BRAF, and PIK3CA mutations in patients with advanced stage, non-squamous, non-small-cell lung cancer (NSCLC) in 28 certified molecular genetics centres covering the whole of France, including overseas entities. The title of the paper describes this as “Routine molecular profiling of patients with advanced non-small-cell lung cancer”, but what it means is routine molecular profiling of tumour samples from patients with advanced non-small-cell lung cancer, which is a very different thing. And the trial proved it can be done: “Routine nationwide molecular profiling of patients [sic] with advanced NSCLC is feasible. The frequency of genetic alterations, acceptable turnaround times in obtaining analysis results, and the clinical advantage provided by detection of a genetic alteration suggest that this policy provides a clinical benefit.” But that too is a very different thing, requiring separate proof.

BMJ 2 April 2016 Vol 352
Pi-ous fears
Here’s a record linkage study from the UK Clinical Research Database which suggests that pioglitazone may increase the risk of bladder cancer, whereas rosiglitazone doesn’t. The absolute risk is tiny, at 121.0 v 88.9 per 100 000 person years, but this is the sort of thing that gets drugs taken off the market. And it reinforces the point I made last week about lumping together whole drug classes: individual drug effects can get diluted out, but they are all-important for patients.

All these confounded drugs for diabetes
Whereas the pioglitazone-bladder cancer association is very unlikely to arise by confounding, and is consistent with randomized trial evidence, broad associations from observing whole drug classes are a different ball-game altogether. You could argue that they are always lethally prone to “confounding by indication”—in other words, you can’t get round the fact that different doctors prescribe differently for different patients according to disease stage and a whole bundle of indeterminable reasons. Using the ball-game metaphor, the pitch is full of random bumps and holes and the goal-posts keep shifting, so the scores of each team end up being almost random. It’s the “almost” that is then the question. Failing better evidence, should we ever use observational evidence to give us an approximate idea of how to treat patients on the muddy playing field of real life? Julia Hippisley-Cox has worked tirelessly to try and disentangle these issues, setting up her own observational QResearch database in England many years ago, and analyzing the data into useful clinical categories. Hidden here are some damning figures about sulfonylureas, but this paper accentuates the positive. I’ll leave you all to pick through the detail, and to decide for yourselves what to believe.

These papers inspire a stunningly good editorial from Victor Montori on our ignorance about drugs for type 2 diabetes. “We need broad collaboration to conduct randomised trials embedded in the flow of practice; we need to turn uninformed choice into random allocation. When participation in a randomised trial is not an option, shared decision making will have to compensate for imperfect and sometimes corrupt information, with a humble and generous commitment to understand fully what is best for each patient.”

Plant of the Week: Fritillaria uva-vulpis

Fritillaries bring joy to the April garden as their grey-green leaves appear straight out of the soil, soon to be topped by nodding flowers of unique shape and grace. Britain is lucky to be host to the most beautiful of all, the damp-loving snake’s-head. But in the Middle East there are many others which grow on dry stony hills and come in a variety of intriguing shades and sizes.

At the local farmers’ market yesterday my eye was caught by some small dark flowered fritillaries which were selling at three pots for £2. Over the years, I must have killed about £200-worth of rare fritillaries. Why should these modestly priced beauties be spared a similar fate? They would have been a rarity about 20 years ago, but it’s now possible to get F uva-vulpis or the closely similar F michailovskyi very readily in packets of bulbs or pots of plants. And I think in suitable spots they will thrive and multiply readily. I’m hoping so, or I will have paid out a whole £2 for but a single year’s pleasure.

So there they are, still in their pots and nodding prettily. The bell-like flowers are of darkest purple, improbably fringed with a brocade of bright yellow. Delightful. May they live as happily in muddy Oxfordshire as they have for millions of years in the brooding mountains of Western Iran.