Richard Lehman’s journal review—16 December 2013

Richard LehmanNEJM  12 Dec 2013  Vol 369
2283    This week most of the NEJM is taken up with trials of genotyping to guide starting doses of vitamin K antagonists. Fair enough: this is a common clinical problem, and warfarin initiation is an important test case for genotyping as the gateway to personalised therapeutics. Dosing people with warfarin is currently a matter of trial and error: you grope your way forward, using lots of blood tests, and some patients end up on a dose of 1mg daily, others on a dose of 11mg, all for the same target INR. We know that a lot of this variation is governed by variants in the gene loci CYP2C9 and VKORC1. Now if genotyping can’t help in a situation as extreme as this, what use is it ever going to be as a guide to drug treatment? Here is the first trial of three: in 1015 patients during the first four weeks of newly started warfarin therapy, genotyping made absolutely no difference in achieving and maintaining the target INR.

2294   The next trial, from Europe, was smaller in size but longer in duration, and used a different genotyping platform. In 455 patients with atrial fibrillation being started on warfarin, slightly more of those in the genotyping group achieved their target INR at 12 weeks. The important outcomes—episodes of bleeding or thrombosis—were not measured in these trials.

2304   Finally, a similar trial was conducted in European countries where weirder coumarol anticoagulants are fashionable—acenocoumarol and phenprocoumon. These 548 patients fared exactly as well whether their dosing was guided by clinical factors or by genotyping. The journal runs two commentaries on these studies, one of which observes that, “The public’s expectations for pharmacogenetics may arguably be declining.” This is so true. Ever since I got my bus pass, I have used a lot of public transport, and never once have I overheard an upbeat conversation about pharmacogenetics. Neither of the articles can quite bring itself to make the final diagnosis. Yes, genotyping for vitamin K antagonist dosing may have appeared to stop breathing: yes, there is a dilated look about its pupils; no, we can’t feel a pulse; but it can’t be time to give up yet. Actually guys, if you don’t mind, I’m off for some coffee; and while I’m there I’ll fill out the death certificate.

JAMA  11 Dec 2013   Vol 310
2395   OK, you expect me to play the old curmudgeon, but I honestly try to think young and optimistic, and to see the positive side in technical progress. “Can mobile health technologies transform healthcare?” asks a viewpoint in this week’s JAMA. I think the authors want you to answer yes, but I’m frankly baffled. “mHealth could benefit ambulatory individuals in two general ways: (1) allow them to more easily and reliably self diagnose their acute symptoms, and (2) enhance monitoring, tracking, and communication of various biometric information (eg, blood pressure, glucose levels, spirometry values, oxygen saturation) for individuals with chronic medical conditions, enabling greater engagement and partnership in their care.” But hasn’t every study so far shown that (2) just increases anxiety and service utilization and has no effect on outcomes? But on they go. ” A host of other acute conditions could be addressed through novel technologies. For example, otitis media might be diagnosed using a smartphone based otoscope and urinary tract infections using at home urinalysis.” I’m really not getting this. If you kept dipping your mobile phone in your urine, wouldn’t that make it, well, a bit smelly after a while? And then you might not want to put it in your ear.

2407    I have never visited Spain, but I am told that everybody there sleeps all the afternoon and doesn’t have dinner till about 11 at night or go to bed until about 4 am; so I am surprised that any Spaniard has the opportunity to develop obstructive sleep apnoea. Yet the Spanish Sleep Network managed to find 194 patients with resistant hypertension and OSA, and proceeded to randomize them to continuous positive airway pressure or no CPAP. They found a 3mm drop in 24 hour SBP in the CPAP group after 12 weeks. Not exactly dramatic, but we know CPAP for OSA improves cardiovascular outcomes, so why bother with this surrogate measurement? Just give them all a pump and a mask, for their short nights and long siestas.

2416   It cures type 2 diabetes in over 60% of obese patients; they lose 40kg in weight; and their blood pressure and LDL-cholesterol fall to normal levels in a substantial proportion. So Roux-en-Y bariatric surgery is a fantastically successful intervention in the short to medium term. Yet if it were a drug, would we accept a 2% mortality rate? And the absence of long term outcome data? I hope not. And this study of outcomes from bariatric surgery at three years shows that not all individuals benefit equally, and that laparoscopic gastric banding is at best only half as effective as Roux-en-Y. So while I’m all for making bariatric surgery more widely available for obese people in metabolic disarray, I think we need to make every single patient the subject of longterm follow-up, measuring everything under the sun, including serum copper. Did you know that copper deficiency is a cause of irreversible neurological damage in some people following weight reduction surgery? There’s no limit to what you can learn by reading these reviews.

2435   And you don’t need to have your stomach bypassed to develop micronutrient deficiencies. Just taking an acid suppressant continuously will sometimes do nicely. Here’s a large case control study based on data from Kaiser Permanente Northern California looking at the association between vitamin B12 deficiency and long-term prescriptions for proton pump inhibitors and histamine 2 receptor antagonists. They detected a significantly increased risk of B12 deficiency with long term PPIs and a lesser risk with long-term H2As, but what we really need is a bigger better study based on a larger database like the CPRD. Anyone out there needing a quick paper in a major journal, here’s your chance.

2451   Statins! For everybody who wants them, say I. You probably disagree, so read this clinical evidence synopsis about statin therapy for primary prevention of cardiovascular disease. We can agree that more evidence is needed: so once again, young research fellows, get digging in those large databases. I think the biggest question is whether the increase in “diabetes” with statins has any long term significance or is merely an artefact of the way we arbitrarily define “diabetes.”

Lancet  14 Dec 2013  Vol 382
1981   When the Medicines Company ran its initial CHAMPION trials of cangrelor, a fast-acting intravenous platelet inhibitor, they failed to show benefit. Then earlier this year, the NEJM published the CHAMPION-PHOENIX trial which purported to show that “Cangrelor significantly reduced the rate of ischemic events, including stent thrombosis, during PCI, with no significant increase in severe bleeding.” But in fact it showed no such thing, because 63% of patients had received oral clopidogrel as well, as the NEJM editorial points out. Now the Lancet sees fit to publish a Medicines Company funded pooled patient data analysis from all the company’s CHAMPION trials with the conclusion that “Compared with control (clopidogrel or placebo), cangrelor reduced PCI periprocedural thrombotic complications, at the expense of increased bleeding.” So the bleeding bit is contradictory. But let’s leave the last word to the Lancet editorial entitled “Cangrelor: a new CHAMPION for percutaneous coronary intervention”, which ends “its favourable pharmacodynamic profile and effectiveness in reducing periprocedural events makes cangrelor a useful and welcome agent for interventional cardiologists and their patients.” The author then goes on to declare “I have worked as a consultant for AstraZeneca, Eli Lilly, and Sanofi, and my institution has received grant support from AstraZeneca and Boston Scientific.” So that’s all right then; his ringing endorsement of this product can’t be biased. Except that last year the Medicines Company and AstraZeneca joined forces to promote each other’s antiplatelet drugs, as reported in Forbes Weekly. How can the Lancet ignore conflicts of interest on such a scale?

1993   “Here we report MERS-CoV genomes obtained directly from 21 patients with MERS from across Saudi Arabia and assess the spatiotemporal distribution of MERS-CoV in Saudi Arabia.” MERS-CoV may be yesterday’s killer plague that never was, but its epidemiology is still interesting, and shows genomics at its batty best: “The closest relative to MERS-CoV was identified through phylogenetic analysis of a short fragment sequenced from the bat species Neoromicia zuluensis. However, in view of the estimated evolutionary rate of MERS-CoV, the most recent common ancestor between this isolate and MERS-CoV existed in bats more than 44 years ago. A recent serological study of dromedary camels in Oman and the Canary Islands found cross-reactive antibodies to MERS-CoV, but the investigators were unable to amplify any MERS-CoV-like viral sequences from the samples. A small fragment of sequence identical to the EMC/2012 MERS-CoV has been reported from a Taphozous perforatus bat captured in Saudi Arabia, suggesting a regionally relevant bat reservoir.”

BMJ  14 Dec 2013  Vol 347
Upper respiratory tract infections and back pain. Some Monday mornings it seems that general practice consists of nothing else. One way or another you have to blag your way into persuading people that doing nothing is the best course, and after trying it for 35 years, I’m not sure I am any more skilled than on day one. For the whole of that time, GPs have been lambasted for giving out too many antibiotics for URTIs, and we have duly reduced our prescriptions; at the same time, there has been a large drop in the number of people coming to see doctors for these conditions. So we are doing something right. And here is a massive piece of genuinely complex, truth-seeking primary care research into the complications of URTIs, of a kind that you would hardly ever find done in secondary or tertiary care. To me it’s worth the entire contents of the New England Journal and the Lancet for the past month. So what is the take home message? That serious complications of URTIs are rare, and you cannot tell who is going to get them. Antibiotics seem to make little difference, but as this was an observational study, you can’t be sure. Scores like CENTOR and the authors’ own FeverPAIN have little predictive value. Ditto 35 years of clinical experience.

The literature on upper respiratory infections in children is vast and heterogeneous, but Matthew Thompson and colleagues have trawled through it to find out how long URTIs usually last. You’ll no doubt already have read some of their findings. “In 90% of children, earache was resolved by seven to eight days, sore throat between two and seven days, croup by two days, bronchiolitis by 21 days, acute cough by 25 days, common cold by 15 days, and non specific respiratory tract infections symptoms by 16 days.” Maybe we should put these figures on a placard somewhere in the waiting room.

Metal-on-metal hip resurfacing has had a rotten press, and as a procedure for hip arthritis, it should probably never have been introduced. But according to this study, it prolongs life when compared with standard total hip replacement. The Kaplan-Meier survival curves are really quite extraordinary. Maybe we should all have this procedure in order to prolong our existence, never mind what it does to our hips.

There are lots of other goodies in this print issue of the BMJ, before the magazine goes into festive hibernation for two weeks. Nigel Hawkes talks to the heads of the “10 leading hospitals” of the NHS. I have only one quality criterion for a hospital: that it gives a high standard of secondary care in close cooperation with its primary care catchment. Any fool can do the distant, glamorous stuff. Chris Dowrick, whose book Beyond Depression was my best medical book read in 2013, writes beautifully about medicalising and medicating unhappiness. Steve Nissen writes with subdued rage about the way that GlaxoSmithKline has been allowed by the FDA to get away with “clearing” rosiglitazone by means of the RECORD trial, an open label study in which GSK had full control of the data. And you will not find a better guide anywhere than the clinical review which deals in extenso with the somewhat unseasonal subject of tick bite prevention and tick removal.

Plant of the Week: Rosa “Albéric Barbier

This may seem an odd time to be praising a rambling rose which has its main flowering in early summer, but ours is actually producing a few flowers this week. And it is a handsome beast with healthy dark green foliage that persists for almost the whole winter.

Alberic the Barbarian is a true Viking of a plant, never content unless it is invading new territory. Left unchecked, it can sprawl over about 80 square metres, but we try to keep our leading shoots to about 4 m. It can handle any amount of pruning. The flowers appear pale yellow and turn white, with a pleasant if not overwhelming scent in June. If its branches touch the ground they will readily form suckers which you can give away to friends with large spaces to cover.