Richard Lehman’s journal review—25 April 2016

richard_lehmanNEJM 21 April 2016 Vol 374

Aliskiren in Cardioland

1521 What does the R in the RAA pathway stand for? I used to pose this question in lectures several times a year, believing all that I had been told about the importance of the renin-angiotensin-aldosterone pathway in heart failure. I’d explain that we had drugs which blocked the AA but not the R. Then, a few years ago, along came aliskiren, a direct renin inhibitor. But early trials raised fears that poor Alice had come too late to the party. And this trial in people with chronic heart failure and reduced ejection fraction really pushes her down the rabbit hole. “In patients with chronic heart failure, the addition of aliskiren to enalapril led to more adverse events without an increase in benefit. Noninferiority was not shown for aliskiren as compared with enalapril. (Funded by Novartis; ATMOSPHERE)” shouted the Queen of Hearts. “Not non-inferior, not non-inferior” squawked the Mad Hatter. Alice went to look for her friend the Walrus, because she couldn’t stand the atmosphere any longer.
AF ablation: frying or freezing?
OL “Non-inferior!” shouted the Walrus when Alice arrived. “Upon my soul, balloon cryoablation is non-inferior to radiofrequency ablation for drug-refractory paroxysmal atrial fibrillation! Have you ever heard such a thing, my dear?” And all the oysters began to sing:
“Our hearts were in a flutter
Till they brought us the cold balloon!
What in the name of decency
Is all this radiofrequency
When you can bring us the cold balloon!
O bring us the cold balloon!
Our hearts will be over the moon
You can never do it too soon
O bring us your freezing balloon!”
But poor Alice was not comforted. She remembered that she herself was not non-inferior. She was inferior. The whole family of Kirens would now live in disgrace, while cold balloons became the rage with interventional cardiologists. How tiring it is to live in Cardioland, she thought, taking out her handkerchief.

Post-op AF: rate or rhythm control?
OL Alice decided to have a cup of tea, but unfortunately that meant seeing the March Hare and the Mad Hatter again. “Riddles if you want any tea!” said the Hare, so Alice replied, “I’ll ask you one this time. If two things are equally good at something, which one is non-inferior?” “Not non-inferior?” squawked the Mad Hatter. “No, just non-inferior” replied Alice. “You mean superior” sighed the Dormouse, and fell asleep. “What’s the difference between rate control and rhythm control?” said the Hare. “Don’t change the question,” said Alice. “But what if it’s a trial to see if one strategy or the other is non-inferior for post-operative atrial fibrillation? Or superior? Or not non-inferior but inferior?” shouted the Mad Hatter. “What it they’re both the same?” offered the Hare. “Oh shut up and give me some tea,” cried Alice, “A Canadian study has just showed that they’re both equally good at resolving AF at 60 days though the patients get switched over a lot depending on whether they can’t stand amiodarone or don’t respond to beta-blockers. Yes, just a spot of milk thanks.”

How classy is this hole in my foot?
OL Two months ago, the NEJM published the now famous “research parasites” editorial which I described as shooting itself in the foot. A hasty note of clarification followed. But the issue won’t go away, and now the journal has put on its website two open access letters which offer lofty-sounding justifications for caution about data sharing. One is mostly about the recognition of “data generators” (as opposed to parasites) so that the original investigators are given due recognition for the primary work of conducting the trial.

I don’t think this was ever in question, but if this terminology helps to allay the alleged fear that “data-sharing requirements may discourage researchers from initiating and participating in clinical trials” then well and good. The second letter raises the spectre of de-identifying not just individuals but also institutions in embedded trials that use routinely collected health care data. The arguments are a bit far-fetched and it seems to me that solutions already exist. Once again the letter comes from Harvard, Boston, Mass. and includes a warning that “this consideration may dissuade clinicians and health systems from participating.”

I do think the NEJM is having trouble deciding whether to wear the hole in its foot with pride, as opposed to hoping it will heal up and that people will cease to notice.

JAMA 19 April 2016 Vol 315

Back to cardioland
1580 It’s back to Cardioland in this week’s JAMA, though I’ll spare you any more Alice. Cardioland is an ineffably weird place where drugs can cost anything the market can bear, even if they have never been shown to change any outcomes. Take evolocumab, which can lock you into paying $14,000 per year in the USA, or about half that in the UK, just to see your levels of LDL-cholesterol come down on blood tests. Since statins can do this at hundreds of times less cost, the market for this PCSK9 inhibitor must lie in people who cannot take statins but need to lower their LDL-C. Thus billions of dollars stand to be made by persuading people that they are intolerant to statins and that evolocumab will lower their LDL-C better than any competing agent. The standard competing agent is ezetimibe. This Steve Nissen-run trial (GAUSS-3) began by ascertaining if people were intolerant to atorvastatin vs placebo in two-week cross-over periods. If they were, they were randomized to 2:1 to subcutaneous evolocumab (420 mg monthly) or oral ezetimibe (10 mg daily). Quite predictably, their LDL-C dropped more with the first than the second. In real life, this would have been matched by their bank accounts. As I have already pointed out, evolocumab is an antibody with no yet-known benefits and a very strong affinity to dollars.

In days gone by, myocardial infarction made fortunes in Cardioland. But as MI gets rarer, and the marginal benefits of each intervention grow smaller, the chances of a windfall have declined. Still, GlaxoSmithKline were sufficiently convinced of the blockbuster potential of losmapimod to take it all the way to a phase 3 trial conducted by the TIMI group (LATITUDE-TIMI 60) conducted at 322 sites in 34 countries. Losmapimod is an inhibitor of p38 Mitogen-activated protein kinase (MAPK)–stimulated inflammation, which is implicated in atherogenesis, plaque destabilization, and maladaptive processes in MI. It may indeed combat all these malificent things but it makes absolutely no difference in human beings who have had heart attacks. Unless GSK can find a different use for it, they may have to rename it losmakimod.

Programming life & death in melanoma
1600 If you find Cardioland disquieting, stay away from Oncoland. You need about $100K just to get in. Once there, you really have no idea what might happen. Say you have advanced malignant melanoma. This was a certain death sentence until very recently. But the programmed death 1 (PD-1) pathway limits immune responses to melanoma and can be blocked with the humanized anti-PD-1 monoclonal antibody pembrolizumab. A lucky few even seem to go into permanent remission. But as this overview of randomized and non-randomized studies of pembrolizumab shows, two-thirds of patients get no benefit at all. And at present there is no way of predicting who these will be. I won’t go into detail, as it is dealt with in a clear editorial. Here again is a clear case of the number-needed-to-treat being a very poor guide for individuals or decision makers. The NNT is not at all spread evenly like honey on toast. It is more like a life-and-death lottery with few winners, while everyone invests those few $100Ks each.

Ann Intern Med 19 April 2016 Vol 164
The defeat of rheumatoid arthritis
523 When I began writing these reviews 18 years ago, rheumatology seemed a rather sleepy specialty, and I even got into hot water by contrasting our local service with the rapid access clinics that I read about in the literature. One by one, new treatments appeared and early “aggressive” treatment with tumour necrosis α blocking agents became popular. About this time a group of Dutch rheumatologists set up a highly pragmatic trial to see if there would be any long-term differences between four common strategies for treating RA “to target”, i.e. to suppress joint inflammation clinically and serologically. These were: Sequential monotherapy (strategy 1), step-up combination therapy (strategy 2), or initial combination therapy with prednisone (strategy 3) or with infliximab (strategy 4), all followed by targeted treatment aiming at low disease activity. You don’t need to remember any of that, because there was lots of cross-over and they all had the same results at 10 years. “Conclusion: In patients with early RA, initial (temporary) combination therapy results in faster clinical improvement and targeted treatment determines long-term outcomes. Drug-free remission, with prevention of functional deterioration and clinically relevant radiographic damage, and normalized survival are realistic outcomes.” So quietly, without understanding its aetiology, we have come to conquer rheumatoid arthritis over the last decade or so. If that isn’t good news, I don’t know what is.

Lancet 23 April 2016 Vol 387
Antibiotic CV lines in children
1732 Paediatric intensive care is a place where you don’t want your child to be. The mortality in this study conducted in 14 English paeds ICUs was 6-8% in all three groups: those who received standard central venous catheters, those who received heparin-coated CV lines, and those whose lines were coated with minocycline and rifampicin. But although there were no differences in mortality, the kids who had the antibiotic coated CV lines had significantly fewer bloodstream infections.

Malaria: the path to global eradication
1775 In the last 100 years, over 100 countries have eliminated malaria. Even in the last ten years, there have been huge reductions globally. 35 countries are now well on the way to national eradication. This review of progress towards total elimination is wonderfully detailed and upbeat. The target year is 2040, for those who may be around to see it.

Stem cells for HF: a result at last?
OL We all want stem cell treatment for heart failure to succeed. But for about 15 years, it hasn’t. It reached the point a couple of years ago where I said further studies should not appear in major journals unless they had something new to show. Darrel Francis and his team produced a rather witty study showing that the more positive the results of these stem cell trials, the higher the likelihood that they contained internal data inconsistencies. I hope he will do the same analysis for this one. On the face of it, it was well designed and blinded, and just about reached statistical significance in the first 12 months for a clinical composite end-point of all-cause death and cardiovascular admission to hospital: risk ratio 0•63 [95% CI 0•42–0•97]. The Vericel Corporation sponsored this phase 2b trial of their product called ixmyelocel-T, which takes bone marrow aspirates and enhances the right cells and puts them into the myocardium by endocardial injection. If this really does work in the longer term, then it may be a useful treatment option for a few people with dilated ischaemic cardiomyopathy. But I don’t see it becoming practical and affordable for the other millions.

The BMJ 23 April 2016 Vol 353
Dying for better care
For me, the best thing in this week’s The BMJ is Scott Murray’s editorial on end of life care. I’ve known Scott for over 20 years, during which we’ve both advocated for improved care for people dying from heart disease. I think it’s become clear that this can’t be done simply by more referrals of heart failure patients to palliative specialists: it requires a transformation of goals of care so that everyone involved accepts that there are many difficult processes of dying, and in all of them the patient’s aims must come first. So Scott’s suggestion is that palliative care specialists should spend half their time training other doctors to care for people dying from any cause. I absolutely agree.

Beta-blockers for all systolic heart failure
When guidelines were young and I was in my early 40s, I helped to draw up the first heart failure guideline for Oxfordshire. Among the drugs which should never be given were beta-adrenergic blockers. A couple of years later, we changed that to the opposite, and so it has remained. And since then, evidence has been accumulating that β-blockers postpone mortality in all groups with systolic heart failure, and are tolerable to most people, including those with COPD, when introduced slowly. But it seems to me that many questions remain. We know that most elderly people with HF would rather have symptom improvement than a longer life, and the evidence would suggest that β-agonists might do that better than β-blockers. Among the β-blockers, we only know the types and doses used in clinical trials, so we can’t say if one type or dose is “optimal” since the comparative evidence is just not there. This individual participant meta-analysis can only tell us what is in the randomised trials. It’s a classic case of evidence-based medicine not equating with medicine-based evidence. True EBM will only become possible when patient-centred goals determine the research agenda.

Plant of the Week: Viola labradorica

In the last week, Ben Goldacre has been tweeting for garden advice from plant nerds. He has nearly half a million twitter followers, so such nerds must be well represented among them. I am honoured that he has accepted my advice about wisteria.

But here is my tip for a novice gardener: plant the Labrador violet. It looks so nice with its purple-brown leaves and abundant little purple flowers in spring. Within a couple of years, your garden will be full of the little things. Within five, you will be sick of the sight of them. You will have pulled most of them out, but they will keep appearing in every neglected corner. This is just as it should be. It is what they are for.