NEJM 20 November 2014 Vol 371
1963 The melanoma trials last week got me thinking about how the current model of cancer drug research lets down trial participants and dying patients. Between 2002 and 2014, there have been 71 drug approvals by the Food and Drug Administration (FDA) for the treatment of metastatic and/or advanced and/or refractory solid tumours. The median gain in survival provided by these is 2.1 months. I discovered these figures in an excellent paper in JAMA Otolaryngology–Head & Neck Surgery, a journal that some of you may not read. I don’t know how it got hidden there.
In the UK, we are somewhat protected by the National Institute for Health and Care Excellence (NICE), but The Times filled its front page on Thursday with news that the delaying powers of NICE would be reduced, and that this would lead to a wave of innovative treatments based on a person’s genetic profile. If only. Cancer drugs are for the most part based on the genomics of the cancer line, not the person. And just one person in a hundred with non-small cell lung cancer (NSCLC) has a cancer line with a rearrangement of the chromosomes in the gene encoding ROS1 proto-oncogene receptor tyrosine kinase (ROS1). If you can identify these cancers, they will respond to crizotinib, a Pfizer drug currently blocked by NICE for the general treatment of NSCLC. Remember that 99% of these tumours will not respond to the drug, but for the lucky 1%, this trial shows that it is likely to produce a marked and sustained response. It’s possible that cancer research will have to carry on in this painstaking, incremental way, although I believe that it could be greatly accelerated by a non-competitive model with complete data sharing. In the meantime, any relaxation of regulation could simply unleash harm on the great majority of dying patients. To restate the obvious: the existing pattern of drug development threatens to make cancer care unaffordable, even in the richest countries.
1972 An observational study used a Swedish national registry to identify 33 915 people with type 1 diabetes, with a mean age of 36 and a mean disease duration of about 20 years. They followed them up for a little over eight years to discover how glycaemic control related to cardiovascular deaths, compared with a non-diabetic control group five times as large. Even for the best controlled group, with glycated haemoglobin below 7%, all cause deaths were 2.36 higher than in the non-diabetic group. Go up to 7.8% (62 mmol/mol) and the risk remains the same, but above that the increase becomes ever greater, and above 9.7 the hazard ratio for death is 8.5.
1994 Gene therapy is another big idea that has been around for decades, but is still in its early stages of application to real people. The severest form of hereditary haemophilia, type B, has been targeted for about a decade, but it has taken until now to work out the best way of getting a normal copy of the factor IX gene into enough cells without collateral damage. The treatment consists of a single intravenous injection of a novel serotype 8 pseudotyped, self-complementary adeno-associated virus 8 vector expressing a codon-optimized factor IX transgene (scAAV2/8-LP1-hFIXco). Just pause for a moment to reflect on the amount of cumulative scientific knowledge and technical effort that were compressed into this seemingly simple, one off injection. The 10 men in this phase 1 study have all shown lasting benefit at three years, without late toxic effects. This was a non-commercial trial for a rare condition, which otherwise costs $250 000 a year to treat. More of these, please.
2005 If, like me, you need cheering up about cancer therapeutics, there’s no better place to turn to than this update, which begins: “There have been substantial advances in the treatment of testicular cancer. Fifty years ago, a diagnosis of metastatic testicular cancer meant a 90% chance of death within 1 year. Today, a cure is expected in 95% of all patients who have received a diagnosis of testicular cancer and in 80% of patients with metastatic disease.” The downside is that the incidence of these cancers is increasing, especially in northern Europe, where 1% of men will receive a diagnosis of testicular cancer during their lifetime.
JAMA 19 November 2014 Vol 312
1988 And now we must turn to the left atrial appendage. As its name suggests, this is a bulge in the left atrium, somewhat fancifully described as “wind-sock shaped.” In non-valvular atrial fibrillation, this is the main site where clot forms. So if you close off the appendage, then you might be able to prevent stroke in AF without the need for anticoagulation. And according to the PROTECT-AF trial, this is indeed the case: “After 3.8 years of follow-up among patients with nonvalvular AF at elevated risk for stroke, percutaneous LAA closure met criteria for both noninferiority and superiority, compared with warfarin, for preventing the combined outcome of stroke, systemic embolism, and cardiovascular death, as well as superiority for cardiovascular and all-cause mortality.” Once we have better long term data, this could change the routine management of AF.
1999 If you have a heart attack in hospital, you are less likely to receive invasive treatment and much more likely to die. This finding emerges from a study of outcomes in 303 hospitals in California, and you can guess at some of the reasons why this should be so. The 5% of myocardial infarction patients who have their event in hospital are older and iller than those who are brought in by ambulance; and those whose MIs are immediately fatal in hospital are going to be counted in the figures, unlike patients who drop dead from heart attacks outside hospital.
2008 As you should know by now (since I mention it so frequently), half of old people with the syndrome of heart failure do not have a reduced systolic ejection fraction. The treatments that have some effect on people with impaired systolic function have no effect on them, at least in terms of hospital admission and survival. But a Swedish registry study suggests a possible benefit from β-blockade: “In patients with HFPEF, use of β-blockers was associated with lower all cause mortality but not with combined all cause mortality or heart failure hospitalization. B-blockers in HFPEF should be examined in a large randomized clinical trial.” But why? In these patients, whose mean starting age was 76, the difference in five year survival between those taking β-blockers or not was 45% vs 42%. This just scrapes into apparent statistical significance, provided you don’t round up 0.996 to 1 (hazard ratio [HR], 0.93; 95% CI, 0.86-0.996; P = .04). But consider that this is an observational study, not a randomised trial: people who are able to tolerate β-blockers are not the same as those who cannot; and no amount of propensity scoring can make them so, or balance the groups to a CI of 0.004. And then consider what you would want if you were 76 and had a stiff old heart that was beginning to cause you breathlessness and oedema. Would you want to take an extra pill a day, which might conceivably give you a 3% better chance of living out another five years, at the likely cost of more breathlessness and greater tiredness? My guess (confirmed by several studies) is that you would want something to make you feel better, even at the risk of shortening your remaining life. The drugs that need to be trialled in these patients might be β-adrenergic stimulants, not blockers.
2030 And now think about another therapeutic issue: dosing the entire population so that the herd, on average, can live a few weeks longer. How does individual choice fit in? What number needed to treat are you striving for, and what harms should individuals be asked to trade with it? Not that most real people make their choices this way, but never mind. I am talking, of course, about the polypill. I will just leave you with this summary of the evidence: “Polypills are associated with greater reductions in systolic blood pressure and total cholesterol compared with usual care, placebo, or active comparators, but also with a 19% higher risk of any adverse event. Due to limited power from available evidence, the association of polypills with all cause mortality or fatal and nonfatal CVD events is uncertain.”
Ann Intern Med 18 November 2014 Vol 161
690 I do wish there was a journal series that would remind working doctors—especially in primary care—of the risks they run of poisoning patients with common drugs. I was no saint in this respect myself, but after this week I cannot do any more harm as I shan’t be treating patients. What’s the harm of a bit of ibuprofen in someone with atrial fibrillation? A study of 150 900 Danish patients with AF gives us the answer. “Use of NSAIDs was associated with increased absolute risks for serious bleeding and thromboembolism across all antithrombotic regimens and NSAID types.” By using NSAIDs in these patients (and even more in those with heart failure) you are negating the effects of all the protective treatment they are taking.
Lancet 22 November 2014 Vol 384
1849 Last week I shrugged my shoulders as I told you about the latest meta-analysis of anticoagulant regimens for use during percutaneous coronary intervention (The BMJ): “In patients undergoing primary PCI, unfractionated heparin plus GpIIb/IIIa inhibitor and LMWH plus GpIIb/IIIa inhibitor were most efficacious, with the lowest rate of major adverse cardiovascular events, whereas bivalirudin was safest, with the lowest bleeding.” I remarked that these meta-analyses change with each new piece of evidence, and this week’s print Lancet illustrates the point. The HEAT-PPCI trial concludes: “Compared with bivalirudin, heparin reduces the incidence of major adverse ischaemic events in the setting of PPCI, with no increase in bleeding complications. Systematic use of heparin rather than bivalirudin would reduce drug costs substantially.” OK, so bivalirudin may not be the safest, with the lowest rate of bleeding. Can a mere Toyota owner dispute the findings of a study, which had among its sponsors The Bentley Drivers Club (UK)?
1859 Now might be a good time for me to look back on 35 years in general practice, as I put away my doctor’s bag and increasingly take my turn at the other side of the consultation desk. The basic human encounter has changed little, except for the obstacles now placed in its way. First there is telephone triage, typically done by a nurse after a brief period of training. Then there is the appointments system, typically run to make the day workable for the inadequate staff available. Then there is the computer, loaded with whatever priorities emerge from a political battle between NICE, politicians, the BMA, single condition groups, and anybody except acutely ill people or people with complex problems, ie the people who see GPs. The basic job of the GP remains to be able, amiable, and available. The problem is that it is humanly impossible to be all those things during every minute of every 12 hour day in the current system. Telephone triage is supposed to be a way to relieve GPs of the immediate pressures of the day, but I don’t see any evidence from the study reported here, and it is a constant source of complaint from patients. “Our findings show that, compared with usual care, introduction of GP triage or nurse triage was associated with an increase in the number of primary care contacts in the 28 days after a patient’s request for a same-day GP consultation. Although introduction of GP triage was associated with an increase in overall GP workload compared with usual care, we recorded a substantial reduction in GP face-to-face contacts.” Oh great.
The BMJ 22 November 2014 Vol 349
Whenever I think about vitamin D—which is quite often—I find that my head starts spinning. This is because pretty well all the literature goes round in circles. Low vitamin D is associated with lots of bad outcomes, but is that because bad health lowers levels of vitamin D, or because low vitamin D induces bad health? What are the thresholds? At what point in life do these matter most? Before birth perhaps, in the case of multiple sclerosis? Et cetera. One way to try and cut the Gordian knot is to use genetics. Some people are programmed by their genes to have low levels, and a Danish mendelian randomisation study concludes that: “Genetically low 25-hydroxyvitamin D concentrations were associated with increased all cause mortality, cancer mortality, and other mortality but not with increased cardiovascular mortality. These findings are compatible with the notion that genetically low 25-hydroxyvitamin D concentrations may be causally associated with cancer and other mortality but also suggest that the observational association with cardiovascular mortality could be the result of confounding.” Confound it, my head is beginning to spin again. As an antidote, I would recommend a study of estimated vitamin D levels in goodies and baddies from Lord of the Rings, published in the Medical Journal of Australia.
Margaret McCartney’s excellent column this week has the title “We lack the tools to help patients decide about statins.” It’s true that we lack the ideal tool, but there are several which, with suitable tweaking, could be used in real life primary care. The first and most comprehensive is Victor Montori’s from the Mayo clinic.
A single page tool that is easy to use and which incorporates a wide range of risk reducing options has been developed by James McCormack.
The most recent, which compares statins directly with dietary modification and activity, is the Option Grid, which I helped to develop.
Soon there will be a package directly developed by NICE. But what we really need is a two click tool that allows GPs to instantly calculate the figures for individual people, based on their health record, and print off a personalised decision aid that makes sense to users at all levels of education. Long term risk reduction is a matter for shared decision making, not for pronouncements by experts who never see patients.
Plant of the Week: Mahonia japonica
This is the big bush with spiky leaves that has sprays of wonderfully scented yellow flowers when everything else looks dead. I have already praised it several Novembers in the past. Pause to inhale the lily of the valley fragrance whenever you come across this shrub. But beware of impaling your eyes on its leaves, and be warned that there are many winter flowering mahonia hybrids of similar appearance, which have scentless flowers. People who plant these lack not just scent but sense.