Richard Lehman’s journal review—4 September 2017


NEJM  31 Aug 2017  Vol 377

Liraglutide and the egg-box

Imagine that you are an egg-seller who is short of real eggs. But you happen to have lots of egg-white, and lots of egg-shells to put it in. You fill up the shells with egg-white and seal them so they look just like real eggs. Off you go to market and sell them in egg-boxes, mixed up together with a few real eggs. Before you have had time to get away, an enraged customer returns and complains that you have sold him a load of dud eggs with no yolk in them. Ah, you reply, the boxes just said : contains 100% egg. The customer proceeds to throw one at you and you end up with albumin all over your face.

Now here we have a paper called “Liraglutide and Renal Outcomes in Type 2 Diabetes.”

The conclusion of the abstract states that “This prespecified secondary analysis shows that, when added to usual care, liraglutide resulted in lower rates of the development and progression of diabetic kidney disease than placebo.” But then you read in the results section that “The rates of renal adverse events were similar in the liraglutide group and the placebo group (15.1 events and 16.5 events per 1000 patient-years).” So what is in the eggs that they are selling you here? Mere albumin, I’m afraid. “This result was driven primarily by the new onset of persistent macroalbuminuria, which occurred in fewer participants in the liraglutide group than in the placebo group (161 vs. 215 patients; hazard ratio, 0.74; 95% CI, 0.60 to 0.91; P=0.004).”

Read back and you’ll find that the prespecified renal outcome was “a composite of new-onset persistent macroalbuminuria, persistent doubling of the serum creatinine level, end-stage renal disease, or death due to renal disease.” Moral: never take composite outcomes (egg boxes) at face value. Always look at the contents separately. Having dipstick proteinuria is not the same as being dead.

Rivaroxaban plus aspirin: your choice

When the Bayer company started marketing acetylsalicylic acid as Aspirin in 1899, its fortune was assured. In fact this is a very rare example of a drug whose proprietary name has officially replaced its generic. Also one whose original indication (for pain) has been eclipsed by an unrelated indication (cardiovascular protection). Bayer now also makes billions from rivaroxaban, a direct factor X inhibitor whose patent has been extended (in the US) until 2024. So which is best for preventing events in people with established cardiovascular disease: aspirin, which is virtually cost-free, or rivaroxaban which costs about £60 a month, or a mixture of the two? The Bayer-run COMPASS trial recruited 27,395 participants and found that rivaroxaban 5mg bd caused more bleeds than aspirin 100mg daily with no greater benefit, while combining rivaroxaban 2.5mg with aspirin also caused more bleeds but conferred greater protection. But the differences were minute, despite being statistically significant. Figure 1 of this paper is a great teaching slide for illustrating this. It needs making into a decision aid for patients (and providers) right away.

Getting inflamed about canakinumab for CVD

In the good old days, human beings went around in a state of semi-permanent inflammation from parasites and infections, ate when they could, starved when they couldn’t, and rarely sat down. Few lived long enough to develop atherosclerosis. But in sedentary, well-fed modern people who live beyond 60, inflammation has come to be feared as a cardiovascular risk factor. In a much hyped study presented at the European Society of Cardiology last week, a monoclonal antibody targeting interleukin-1β called canakinumab was shown to reduce C-reactive protein—a crude marker for inflammation first used in the 1930s—in over 10,000 people with previous myocardial infarction. In one dose group of three, this was associated with a statistically significant reduction in a composite outcome of nonfatal myocardial infarction, nonfatal stroke, or cardiovascular death over four years. There were no differences in overall mortality. Canakinumab currently costs about $200K per person per year in the USA and is associated with an increase in fatal infections.

JAMA Intern Med  Sep 2017

Monitoring lipids on statin therapy

Next week I’ll be going to an international get-together of the various national Choosing Wisely campaigns. Getting rid of useless medical practices needs the help of the public and of individual patients. We need to be humble and acknowledge that much of it is to do with saying that we got stuff wrong and now need to put it right. The world still contains hundreds of guidelines that recommend regular testing of blood lipids in people taking statins. Fortunately this practice is already dying out in the UK, but is very much alive and well in the USA. Here’s a study showing that in 90% of cases, it makes no difference to clinical decision-making. Even when it does, it probably shouldn’t.

Ann Intern Med  29 Aug 2017

Screening for cancer in unprovoked VTE

Now for another Choosing Wisely topic: should we screen for cancer in everybody who has an unprovoked episode of venous thromboembolism? The overall risk for having cancer detected in the year after unprovoked VTE is one in 20. But it is not evenly distributed: the risk increases markedly over the age of 50. So there is probably a good case for screening older people after VTE: but just how far should you go with testing (whole body CT?) and what is the actual yield in terms of lives saved? The answers do not emerge clearly even though this is an exemplary systematic review and meta-analysis of individual participant data.

The Lancet  2 Sep 2017  Vol 390

The Importance of Being PURE

The PURE study is a five-continent observational study of diet in relation to cardiovascular disease in mortality in nearly 150,000 people. It is probably the best of its kind ever done, and concludes that “High carbohydrate intake was associated with higher risk of total mortality, whereas total fat and individual types of fat were related to lower total mortality. Total fat and types of fat were not associated with cardiovascular disease, myocardial infarction, or cardiovascular disease mortality, whereas saturated fat had an inverse association with stroke.” But in these columns I have often implied that most dietary studies are bunk, and no observational study should be taken at face value unless it can be backed by randomised trial evidence. Which this cannot be. So believing PURE means sacrificing some intellectual purity. And I think that in this case I am willing to, because this seems to be the closest we are likely to get to the “truth” about diet and vascular disease. Follow-up was meticulous and households were studied as units, and the median time was 7.4 years. Above all, the main message was negative: there was no specific connection between carbohydrate intake and cardiovascular disease: the association was with total all-cause mortality. By contrast, eating more fat—including saturated fat—was associated with lower cardiovascular disease, meaning that we can abandon the saturated fat-CVD hypothesis with some certainty.

So what does “healthy food” look like? Plausibly, it looks like fruit, vegetables, and legumes. “Higher fruit, vegetable, and legume consumption was associated with a lower risk of non-cardiovascular, and total mortality. Benefits appear to be maximum for both non-cardiovascular mortality and total mortality at three to four servings per day (equivalent to 375–500 g/day).” Splendid. I’ll settle for fagioli con tonno, with lots of olive oil, followed by rib-eye steak swimming in a buttery sauce, a salad, fruit, a few cheeses, followed by strawberries and cream (no sugar).

The BMJ 2 Sep 2017  Vol 358

Michelin STARD diagnostic study abstracts

Primary care doctors should be good diagnosticians. They should know when to start diagnosing, when to stop diagnosing, and when not to attempt diagnosing at all. This is half the art of medicine. We leave the other half to the more simple-minded. Unfortunately they often confuse diagnosing with diagnostic testing, frequently leading to circularity of thought and permanent mislabelling. Back in primary care we then have to pick up the pieces and try to join up Humpty Dumpty all over again. There is a special science of diagnostic accuracy testing, which betrays its connection with alchemy in the following sentence: “STARD for Abstracts is a list of 11 quintessential items, to be reported in every abstract of a diagnostic accuracy study.” In alchemy there could only be one quintessence, the element that unites air, fire, earth, and water. But I digress. The STARD list is actually very useful, because the reporting of diagnostic accuracy studies is generally all over the place. I long for the day when most of these studies can be applied directly to the populations of primary care. Maybe this is a first step.

Plant of the Week: Phaseolus coccineus

Since the PURE study makes special mention of legumes, let us consider the merits of the scarlet runner bean. These begin with the bright scarlet flowers, for which these plants were first introduced into British and North American gardens from their native habitat in the mountains of Central America. Although scarlet has fallen out of favour in modern gardening, it is a lovely colour to lighten up the palette of late summer.

Until a few decades ago, it was still common practice to cut off the nascent bean-pods and throw them away, in order to encourage the production of more red flowers. Now it is more common to ignore the flowers in favour of the pods. They are well-flavoured but often fatally stringy. I am usually willing to take the risk, but it is certainly a bore having to trim them on both sides of each pod. Even then you can find yourself chewing on plaques of lignin from the sides.

The native American tribes saved themselves the bother by just eating the roots. In various parts of southern Europe the pods are discarded in favour of the large white or purple beans which are used for particular dishes. There is an attractive-sounding Segovian dish called Judiones de la Granja, in which the beans are mixed with pig’s ears, pig’s trotters, and chorizo, amongst other ingredients. All part of a healthy PURE diet.