NEJM  23 Jan 2014  Vol 370 301    It’s hard to think of two places less alike than the icy expanses of Alaska and the hot claustrophobic depths of a South African gold mine. But back in the 1960s, people who huddled for warmth in Alaska often spread tuberculosis among themselves, and a randomised trial showed that transmission could be reduced by 50% if you gave everybody daily isoniazid. This inspired researchers in South Africa to try the same approach with workers in their gold mines, where a similar epidemic smoulders on, in part because of the 25% prevalence of HIV in this community of men exposed to intense heat and dust and sharing sleeping quarters. Alas, in the face of all these factors, mass screening and the near-universal use of isoniazid had no significant effect on the continuing TB epidemic. 311   In the Wolf’s Glen scene in Weber’s opera Der Freischütz, the baddies who are in league with the devil forge a bullet which will always hit its target, to an accompaniment of oohs and ahs and spooky noises. They invoke the demon Samiel to put the right spell on the bullet: “Segn’es sieben, neun und drei, Daß die Kugel tüchtig sei! Samiel, Samiel, herbei!” It’s this scene that led to the expression “magic bullet,” first used in pharmacology by Paul Ehrlich 120 years ago. It was first applied to drugs designed to bind to specific receptors, and ultimately to the creation of monoclonal antibodies with supposedly absolute accuracy of aim. Now in what we call Alzheimer’s disease, amyloid-beta plaques accumulate in the brain and are thought to play a pathogenic role. So two big pharma companies, Eli Lilly and Janssen, raced each other to the magic Glen, invoked the right demons, and came up with competing magic bullets which bind to soluble amyloid. Bang! Off went the first one. Oops, a miss. It hit the amyloid but made no impression on the Alzheimer’s. And now for the second shot. Oh dear, same again. Time to march home together to the sound of horns and do some folk singing. The two failed bullets, by the way, were called solanezumab and bapineuzumab: fine names for operatic demons. JAMA  22/29 Jan 2014  Vol 311 This issue of JAMA almost qualifies for the epithet “historic” because it contains so many good pieces about clinical trials and the regulation of drugs and devices. When the editor Howard Bauchner was challenged to sign his journal up to AllTrials a year ago, he made a statement about all the very wise people whom he would be consulting about this over the next six months. I don’t know what the upshot was, but by putting out this themed issue, Howie has done the cause more good than any other editor except the BMJ‘s. 361   But I should qualify that: Rita Redberg, editor of JAMA Intern Med appears in two terrific pieces this week, one in the BMJ about the Stryker Wingspan device, and here in JAMA “Opening the FDA Black Box.” What a contrast with a piece that appeared in JAMA two years ago, drooling about how hard working and well intentioned everybody at the FDA is! The fact is that an organisation can be 90% full of hard-working and well intentioned people, but can still be undermined by industry influence and political trimming. Look around you: in academe, commissioning groups, Royal Colleges, even NICE and the US Institute of Medicine. It is a fact of life, but sometimes you have to take sides. And JAMA has taken the right side here. This editorial lists the FDA’s shortcomings, but is not wholly negative. The agency has the potential for change and openness and shows some signs of moving in that direction. Would that the same could be said for our own Medicines and Healthcare products Regulatory Agency (MRHA). 368   When I went to Yale for a year in the summer of 2011, one of the first people I met was a first year medical student in shorts called Nick Downing. We shared a large space at the top of a building which housed lab rats in its cellar, and in that hot summer we couldn’t escape their pervasive odour—wafted up to us by convection in the stairwell. Since then Nick has produced first-author papers in NEJM, the Archives of Internal Medicine and now this landmark analysis of FDA approvals in JAMA (you can get it free). I suspect they might give him a medical degree. The Goodman/Redberg editorial summarises his findings: “Despite the FDA requirement for evidence from a minimum of 2 randomized clinical trials supporting an effect on health outcomes, 37% of product approvals were based on only 1 trial, 53% of cancer trials were nonrandomized, and an active comparator was used in only 27% of non–infectious disease trials. Surrogate end points were used in almost all approvals via the accelerated approval process and in 44% of nonaccelerated approvals. Trials were comparatively short, with most lasting less than 6 months, even those assessing chronic treatments for chronic diseases. Cancer drugs, perhaps predictably, were more often approved via the accelerated process and with weaker designs.” 378   ” The most remarkable aspect of the report by Sacks et al is that it was written at all. Details about unapproved drugs are sparse; the FDA interprets the law as prohibiting the agency from sharing data for unapproved drugs or even from releasing the disapproval letter, the latter despite recommendations from its own transparency task force. Sacks et al have partially peeled back this curtain with memos and materials that in the past have been inaccessible to those outside of the FDA—materials reporting details that only the FDA can provide.” Goodman and Redberg again. I will spare you the details of this paper: basically only one drug approved by the FDA over 12 years was voluntarily withdrawn for safety reasons. For most of the other drugs and devices subsequently taken off the market by the FDA, there were clear signals which should have made the agency pause before granting approval in the first place. 385   I told you the other day that you could go into the market with any medical device you wanted, from a cotton wool bud to an implantable defibrillator, and you would probably get approval if you had the right glossy brochure. This is called the Premarket Approval (PMA) process. If you are already in the business, it is even easier than that. “Many cardiac implantable electronic device models currently used by clinicians were approved via the PMA supplement process, not as original PMAs. Most new device models are deemed safe and effective without requiring new clinical data, reinforcing the importance of rigorous postapproval surveillance of these devices.” Or rather a complete overhaul of the system and reshaping of the agencies, so that patients are not made into experimental subjects without any rights or protection. Has anyone even bothered to look at our own MRHA? There are other goodies in this issue of JAMA too: a must-read piece on “Improving the Drug Development Process: More Not Less Randomized Trials” (sic) by Djulbegovic and Ioannidis. Ioannidis again among the authors of a useful guide to subgroup analysis and Richard Smith of United Healthcare introducing an Oral History of Evidence Based Medicine. Lancet  25 Jan 2014  Vol 383 321   One magic bullet that seems to have found its mark is adalimumab, a monoclonal antibody which binds to tumour necrosis factor-α. It also binds strongly to dollars—over 9 billion of them in 2012. So its manufacturers AbbVie can feel pretty laid back about the OPTIMA trial of adalimumab with methotrexate in early rheumatoid arthritis. ” Outcomes were much the same whether adalimumab was continued or withdrawn in patients who initially responded to adalimumab plus methotrexate.” But since a larger number of patients responded initially to MTX with adalimumab than to MTX alone, specialists and patient organisations paid by AbbVie will join in telling the world that this is the best initial treatment for RA. Or am I wrong? I would be very happy to be told otherwise: I am simply assuming business as usual. 333   In 2005, the FDA approved the Stryker Wingspan stent system for use in stenosed intracranial arteries, on the basis of a single-arm study of 45 patients. Eleven thousand of these devices have been sold in the USA since then, but the fact is that they can harm patients and should never have been licensed at all. This is shown by the first adequately powered randomized controlled trial of the device, SAMMPRIS. Patients on “aggressive medical treatment”, i.e. statins and antiplatelet agents, did better than those who were stented, yet the FDA has not pulled Wingspan. Just how far does industry capture of regulation extend? Read the BMJ article about this device by Rita Redberg and colleagues. BMJ  25 Jan 2014  Vol 347 The BMJ is popular in the USA, and getting more so, but there may be a backlash over here if it continues to feature spellings like “color” on its website and run titles like “Off-hour presentation and outcomes of patients with acute myocardial infarction.” I don’t know if “off-hour” is an expression used in the USA: I have never heard it either there or here. In fact all this American systematic review tells us is what we already knew from excellent single studies like Harlan Krumholz’s comprehensive analyses of Medicare data a few years ago: have your heart attack near a hospital on a weekday morning. If you follow your clinical judgement, you will sometimes miss things, patients will suffer, and you will feel guilty. If you follow diagnostic decision aids, the same will happen, perhaps more frequently, perhaps less. I’ve been a great fan of The Rational Clinical Examination series in JAMA over the years, but I would find it hard to think of an instance where it has decisively changed my practice. I think the wheel may be turning full circle with this systematic review of “red flags” to alert doctors to the possibility of cancer and spinal fractures in patients with low back pain. These flags are either blindingly obvious—the patient has cancer or is an old lady with osteoporosis—or else they aren’t predictive at all. The studies examined are so remote from real life that they rarely begin to do what every doctor does, which is combine features to look at the whole picture: and not one of them includes the test of time. Forget these so-called red flags and use your brain instead. It has better design features. But there are some things the human brain is not well designed for, and one is balancing several numerical risk factors simultaneously. How frequently should you do cervical cytology for a 32 year old who has had four sexual partners from the age of 15 but screened negative for HPV three years ago, compared with a 45 year old who had CIN1 at 35 but has had subsequent clear cytology on three occasions and has not been tested for HPV? And so on. This Swedish study of follow-up after HPV testing study throws some more data into the equation, but it is quite impossible for a human being to put it into context. It needs a well programmed computer. And instead of a fixed recall system for cervical cytology, we need a more sophisticated system of personal risk estimation which factors in age and HPV status amongst other things for each individual. Maybe we already have: not doing daytime general practice, I’m losing touch with this sort of thing, and I can’t say I’m missing it. Plant of the Week: Clematis cirrhosa var balearica I’m sure I’ve praised this plant at least twice before in these reviews, but how can you not, when it’s the prettiest thing in the winter garden? Lovely dark cut leaves, and an abundance of dangling creamy flowers, freckled with purple. Some even claim fragrance for these, but I can’t pick up any on our young plant. Give it a few mild winters, and it will perhaps completely cover an ugly electricity pole we have in our front garden. There are several examples of C cirrhosa around here which have come through some very hard winters, so we are hopeful. Somewhere in your garden, you must have space for this beautiful climber, which can grow to 4 m. Get one now.