Richard Lehman’s journal review – 23 May 2011

Richard LehmanJAMA  18 May 2011  Vol 305
1969   “Gastroesophageal reflux disease (GERD) is a chronic, relapsing disease with symptoms that have negative effects on daily life.” Actually it only became a “disease” about ten years ago: before that it was classed as a nuisance, treated with antacids by many, proton pump inhibitors by some, and surgery by the desperate. As you get older and fatter, you get more of it. The European LOTUS trial simply took 554 patients with established GERD/GORD and randomised them to continued esomeprazole or laparoscopic fundoplication. At five years, outcomes are nearly identical in this group of PPI responders; but the trial we really need is in PPI non-responders.

1996   Most of the older anti-epilepsy drugs carry a significant risk of major birth defects, but this whole-population study from Denmark carries excellent news for mothers who need to take anticonvulsants: no such added risk could be observed from 1996 to 2008 in infants exposed in-utero to lamotrigine, oxcarbazepine, topiramate, gabapentin, or levetiracetam during the first trimester.

NEJM  19 May 2011  Vol 364
1897   Here’s a landmark paper which many readers will already know about, since it has been on the NEJM website for several weeks. It will save the sight of thousands of patients with neovascular age-related macular degeneration and will avert a drug bill of billions of pounds in advanced health systems, simply by demonstrating that one drug manufactured by Genentech is as good as another drug manufactured by Genentech. The two drugs both block vascular endothelial growth factor A: bevacizumab is marketed for the treatment of advanced cancers while ranibizumab is marketed for the treatment of “wet” AMD.  Ophthalmologists quickly realised that by using the first drug off-licence, they could cut costs by a factor of 40 or more: but they did not have a randomised trial to back the practice. This head-to-head trial confirms what Genetech most feared, and naturally enough, the company left the triallists to pay the costs of the treatments themselves, if outside the insurance held by the patient. One can only smile.

1920   From time to time, the moony element selenium rises in the medical firmament and then wanes into obscurity once more. It is an anti-oxidant and has been proposed as beneficial in all the many situations where oxidation is a bad thing – the latest being mild Graves’ orbitopathy. Rather weirdly, 100μg of Se daily seems to help to slow the bulging of the eyes.

1932   Genomics and the Eye. I know little about either, but for a long time I have suspected that genomics is bad for the eyes, as it involves so much close work. I raced to the last paragraph, to see if the authors could point to anything else I might have missed. Well, no: but they do their best to celebrate this noble organ of the senses: “a patient with a disease that injures only a few thousand neurons in the fovea can describe this injury to his physician in great detail, and the physician can in turn view these neurons in the living patient at microscopic resolution by taking advantage of the near perfect optics of the anterior portion of the eye. These natural optics also contribute to a surgical accessibility that is unmatched by any other part of the central nervous system. This latter attribute will be a tremendous advantage for clinician scientists seeking to translate all the recent progress in gene-transfer and stem-cell biology into effective therapies for their patients with genetic eye diseases.” How splendid.

Lancet  21 May 2011  Vol 377
1749   An increasing number of patients with colorectal cancer will be frail and elderly, but most chemotherapy trials either exclude these patients or recruit too few of them to guide individual treatment. Shared decision-making with people who are near the end of life, whatever doctors may do, can be problematic, because it is so difficult to advise individuals about the specific gains or horrors of each option. Here at least there is a trial which attempts to guide oncologists looking after frail elderly patients with metastatic bowel cancer: the results are not definitive but help to show that fluouracil and oxaliplatin are reasonable choices for most patients who want chemo.

1760   I rubbed my eyes when I read the abstract of this trial. I have been rather rude about marketing trials in The Lancet recently: things seemed to have got as bad as they could; I had vowed to try and stay polite from now on. Yet here is a  findings section in an industry-funded trial that gives some of the raw data but no risk ratios, let alone their 95% confidence intervals; followed by an interpretation section which reads: “The data show pirfenidone has a favourable benefit risk profile and represents an appropriate treatment option for patients with idiopathic pulmonary fibrosis. Funding InterMune.” If you now look at the Kaplan-Meier charts for progression-free survival (Fig.3) you will see curves that differ maximally at about 70 weeks (by about 5-15% of participants in the two trials) but coalesce completely by 96 weeks. For some of this period, active treatment improved walking distance to a clinically meaningful extent. The editorial on the study ends by noting that “pirfenidone is so far the only approved drug that will be commercially available in Europe, raising hopes for many patients with this life-threatening lung disease.” Help me: I am trying to stay polite. This stuff will raise hopes for many patients with a uniformly fatal condition, at a cost of $30-50K per year. The data show no significant mortality benefit, and slight functional improvement to some patients. Was this a good licensing decision by the European Commission? Should a responsible medical journal really be providing the manufacturers with the sound-bites that I have quoted?

BMJ  21 May 2011  Vol 342
1093   It’s a relief to turn to the BMJ, particularly as this issue is dominated by the regulation of medical devices, and the scandalous lack of regulation in the UK. Peter Wilmshurst writes a stinging editorial to accompany lengthy and excellent investigative pieces by Deb Cohen and others: it ends with “Regulators such as the General Medical Council should have the power to impose sanctions on doctors who fail to disclose conflicts of interest or who fail to report adverse events related to devices or drugs. Journals should demand that authors publish the actual amount paid to individual investigators involved in trials and should themselves publish annually the amounts they receive from individual companies from reprint sales and advertising.” Now I wonder if The Lancet agrees with this.

1134    Several decades ago, when The Lancet had a sense of humour (see, I’m really trying to be polite about The Lancet), it published an editorial asking why total hip replacement required orthopaedic surgeons. In those days, the idea of non-doctors performing surgery was regarded as a joke, and what better butt for a joke than our colleagues who saw bones, ha hah. If only most physicians were half as useful, say I. Anyway, things have moved on, and in the developing world clinical officers now often perform Caesarian sections: very well too, according to this meta-analysis of controlled studies. Maternal and perinatal death rates do not differ from operations carried out by medical doctors, though the latter have lower rates of wound infection and dehiscence.

1135   For ages, you may remember, the argument went back and forth about the co-prescription of proton pump inhibitors and clopidogrel. It made pharmacological sense to worry that by inhibiting CYP2C19, PPIs would block the transformation of clopidogrel into its active metabolite. But then a randomised trial produced COGENT proof that this does not affect clinical outcomes. Now there is no particular reason to suspect that PPIs will block the action of aspirin, and yet this Danish study of patients following first myocardial infarction raises suspicions that this may indeed be the case. The methodology of this study is described as a nationwide retrospective cohort study with propensity score matching. Not an everyday method, but one that generates a clear Kaplan-Meier chart to compare the risk of cardiovascular death, myocardial infarction or stroke in Danes taking aspirin after their first MI, with or without PPIs. It would be impossible to design a RCT to address this particular group of patients but it would certainly be of interest to do a prospective trial on say patients taking aspirin with or without PPIs for stroke prevention.

Ann Intern Med  17 May 2011  Vol 154
672    For two years I have been trying to work out what we should actually be doing for patients with type 2 diabetes once we’re forced to move beyond metformin and sulfonylureas. This is the latest meta-analysis which is supposed to help us. Don’t hold your breath: “Evidence suggests no clear differences in benefit between drugs when adding a third agent to metformin and sulfonylurea therapy. Patient preferences and characteristics should guide the choice of drug.” Yes, yes: but how do we share these decisions when we ourselves don’t know what these drugs really achieve in terms of meaningful outcomes? At least I can assure you that there is a wonderful team of people debating this on a practically daily basis.

Plant of the Week: Rhododendron luteum

Bodnant in North Wales is one of the rare places in the world where gardening becomes sublime. We visited it a few days ago in light rain, and the colours of the rhododendrons glowed in the moist light and the valley of towering trees shone softly in innumerable textures of green. At no scale is this place less than ravishing: the very undergrowth seems to come from another world. And throughout the magic ravine there was a constant smell which hung in the wet air: not a heavy perfume as from roses but something more vegetal, coming from the damp earth and from some other place.

We tracked it down to the small, rather sparse yellow-flowered rhododendron which forms part of the understory of the great valley. It has a scent like a pheromone, pulling you towards it: there is no other smell quite like it. Other yellow rhododendrons carry this smell, just as they carry its colour genes: but the original species is the best. Now that our camellias have died from frosting of their roots in pots, we will get some of these yellow rhododendrons: they are hardy to US zone 2, which is practically tundra. If you are lucky enough to garden in acid soil, you can plant these at will: they may not be the showiest of this garish tribe, but their small orange-yellow flowers will act on your senses like no others.