Richard Lehman’s journal blog 2 August 2010

Richard Lehman JAMA 21 July 2010 Vol 235 JAMA  28 July 2010  Vol 469
469    It is a solemn sight to see the great medical journals gathering to pronounce that rosiglitazone is dead. Like the bird of loudest lay in Shakespeare’s The Phoenix and the Turtle, JAMA leads the troop of mourners with this big observational comparison with pioglitazone which it published on its website a month ago. It is accompanied by a sensible editorial pointing out that since we now have observational studies and meta-analyses enough, all of them showing that pioglitazone is the safer of these two thiazolidinediones, why should we keep the worse drug in circulation? Not that pioglitazone is a particularly safe drug itself: it probably increases rates of myocardial infarction and certainly worsens heart failure, even while it decreases glycaemia. The fact is that glycaemia, however you measure it, is a very bad surrogate for real outcomes in treating diabetes: something we should have got wise to long ago.

419    There is a time to be born, as the Preacher says in Ecclesiastes; and that time is at 40 weeks’ gestation. This big trawl of US databases shows that just three weeks earlier than that, the risk of respiratory distress at birth is three times higher; at 34 weeks, it is 40 times higher. A needful reminder that even in the modern age, obstetrics is a risky business and that babies are usually better off left in the womb for as long as possible.

435    My English town is, alas, full of fat young people; more even than I saw on a visit to the USA a year ago. I worry for them, since getting rid of large amounts of fat is an impossible task for most human beings, and many will end up needing bariatric surgery. This just doesn’t seem right, but is the only effective intervention, and fortunately it is generally a safe one, at least in Michigan. This careful audit compares ordinary hospitals with “centers of excellence” and high volume surgeons with low volume surgeons. There is no difference between excellent and ordinary centres but there is a roughly twofold difference in the complication rates between low- and high-volume operators. In the brave new NHS, GPs will have to ensure that there is a high-volume bariatric surgery centre in every town.

304   Seasoned pedantic Journal Watchers – and I hope that includes most of you – will recollect that JAMA rarely uses the eponymous genitive whereas the NEJM always does. Hence the title of this JAMA review: Treatment of Primary Sjögren Syndrome. It would have been ” Sjögren’s” in the New England Journal. Never mind. Henrik Sjögren (1899-1986) was a Swedish ophthalmologist whose name has become attached to the sicca syndrome of indequate tears and saliva, first described by Jan Mikulicz-Radecki (1850-1905). Polish pride urges me to call it by its original and proper name, Mikulicz’s syndrome, but Mikulicz-Radecki himself would not have cared a jot. When asked his nationality, he replied: “I am a surgeon.” The evidence-based treatment for this syndrome, however you label it, is quite old-fashioned: not rituximab or TNF blockers but pilocarpine or cevimeline for sicca features and topical cyclosporine for dry eyes.

461   I’m told that some people find JAMA boring and old fashioned, but being that way myself, I like it. I used to look forward to the front covers and the essays about them until Therese Southgate finally retired. I occasionally have a laugh at the “poetry” – last week’s was a particular hoot. The papers are frequently good. A recent innovation – greatly daring – is the inclusion of brief commentaries on hot topics. These vary from dire to brilliant. I really like this one on colonoscopy vs sigmoidoscopy screening. If you want a an extreme example of how counterintuitive medicine can be, consider the fact that screening colonoscopy has never been shown to be superior to screening sigmoidoscopy, and that colonoscopy has been shown to reduce mortality from left sided bowel cancers but not right-sided ones. This is really difficult to get to the bottom of.

NEJM  29 July 2010  Vol 363

411    Metastatic prostate cancer usually responds well to castration by surgery (rare nowadays) or by androgen-deprivation therapy, but recurs within a couple of years or so. At that point median survival is between one and two years and there is no one standard treatment to resort to. I have a feeling that this trial of sipuleucel-T is not going to change that very much. It is an individualised immunotherapy technique rather than a drug: the patient’s own peripheral-blood mononuclear cells are harvested, primed to attack prostate specific antigens and then reintroduced in three injections. Weirdly, this produced no tumour shrinkage but an overall postponement of death by 4 months, observable in the remaining survivors compared with the placebo group for about 4 years. The editorial (p.479) queries the design of the trial and the plausibility of these findings, but on a more cheerful note lists a number of more promising drug trials for advanced prostate cancer.

423    Every year I am supposed to undergo three hours of Completely Pointless Retraining, or CPR for short. Fortunately I have devised a cunning plan to avoid this, which I am not at liberty to share with you. I was incensed to learn that our Instructor was still insisting that we spend part of these 3 hours learning to perform rescue breathing when it has been abundantly clear for several years that chest compression alone will achieve the same result as chest compression interrupted by rescue breathing, i.e. death in nine out of ten cases. When I become local health supremo with a budget of £1bn I will insist that everybody is allowed no more than 15 minutes every 5 years for out-of-hospital CPR training, consisting of a reminder of how to compress the chest. I will cite this US/UK study and the very similar one from Sweden which follows it (p.434).

454   Another tempting career to supplement my NHS pension is that of sham acupuncturist. There will be a huge market for this in the new GP-led NHS. It is a well-trialled treatment for low back pain, as this review reminds us: cheap, harmless and moderately effective, unlike practically all other interventions for back pain. There is absolutely no need to know your yaoyangguan from your huantiao; just buy some sterile needles and get going as soon as you like. It helps to have a plausible manner and to look Chinese, though I must say I rather fall down on the latter.

Lancet  31 July 2010

333    JUPITER took many forms in ancient Rome, and the trial named after him tries to do the same. This was the study that compared the effect of rosuvastatin with placebo in people with no history of cardiovascular disease or diabetes or raised LDL-cholesterol but with a CRP of 2mg/L or above. The rosuvastatin-treated group had 44% fewer events and their LDL-cholesterol showed the expected fall. But so did their HDL-cholesterol, which should have cancelled out some of the benefit. The paper here points out that it did not: when LDL-cholesterol is low, they argue, HDL-cholesterol no longer matters. But actually all you can say is that it failed to matter in people who were given a certain dose of a drug called rosuvastatin: you can’t generalise from this into the overall effects of subfractions of cholesterol. I do find all this special pleading by lipidologists depressingly circular: the editorial on p.305 is even worse. For a mind-clearing antidote read the brilliant overview just posted on the BMJ website, “Shifting views on lipid-lowering therapy” by Harlan Krumholz and Rodney Hayward.

BMJ  31 July 2010  Vol 341

233    It is impossible to read a completely clear account of the breast screening controversy, because the issue is intrinsically murky. But Klim McPherson’s analysis of the data is the clearest you will ever get. It won’t change your mind on this subject if you already have a strong opinion, but then it doesn’t set out to do that. Instead it points out the huge range of uncertainty in the existing studies, making it almost impossible to give women meaningful figures on which to base a choice. It’s a real tragedy that cancer screening programmes are still being rolled out with the same inadequate amount of evidence, and with accompanying information which plays down the uncertainties. I have just thrown my bowel cancer screening kit in the bin.

237    Lifestyle Over and Above Drugs in Diabetes (LOADD) is a principle we can all agree on, at least in the earlier stages of type 2. Every drug we use to lower HbA1c  in this condition has its disadvantages and we simply don’t know long-term  effects of some of the polypharmacy that has become popular recently, such as metformin plus a dipeptidyl-peptidase inhibitor. We know from the observational study of Currie et al earlier this year that overall, drug-based strategies used in the UK to reduce HbA1c below 7.5 tend to increase mortality. The LOADD study from Dunedin in New Zealand sought to get HbA1c  even lower – below 7 – using diet and exercise alone. They succeeded, and may even have improved the long-term outlook of some of their patients by so doing. We shall never know, since the trial was both underpowered  (n=93) and heterogeneous (the exact drug treatments are not specified).

239    Vaccines that work: there’s a nice cheery topic when most of the rest of medicine is a confusing and subject to the whims of lunatic politicians. Give them quadrivalent human papillomavirus vaccine and it does just what it says on the syringe: it provides 96% protection against low grade cervical, vulval, and vaginal intraepithelial neoplasia and anogenital warts.

240   There is a time to be born, as the preacher says in Ecclesiastes (the second time I’ve said that in this blog); and that time is between 0900 and 1700 on a weekday. Babies delivered at term outside these hours in Scotland are 60-70% more likely to die from anoxia. This is quite a shocking figure, and I guess we may be in for more shocks when we at last get place-of-birth morbidity figures for England and Wales later this year.

Arch Intern Med  26 July 2010 

1191    Rosiglitazone increases the risk of myocardial infarction. We’ve known that for about 3 years, but oddly nobody has yet proved that it increases cardiovascular mortality. This latest meta-analysis gives the overall figures with and without the infamous RECORD study, which I won’t go on about. Although there are many similarities, the data on this drug are not as damning as they were for another discredited drug, rofecoxib. But just as rofecoxib disappeared because there was celecoxib, so rosiglitazone will disappear because there is pioglitazone: and the debate will only continue in the law courts, between patient group lawyers and the manufacturers.

1256   I love the “Less is More” series, and not just for its whacky title. As far as I’m concerned, the less PSA testing we do, the better, and that’s the moral of this paper – sort of. The fact is that we don’t really know how to use this test, except to monitor the progression of disseminated prostate cancer. We don’t really know how best to treat prostate cancer, full stop. We can’t reliably separate out cancers that will do harm and those that will stay dormant. So the American way is to treat them all aggressively as soon as possible, as illustrated by this stratification of prostate treatment by PSA level. The “cancers” detected with PSA levels below 4 get the most aggressive treatment of all. This and much else is chewed over in the Invited Commentary which follows the paper – good, but not as good as the BMJ commentary on breast screening.

Plant of the Week: Belamcanda chinensis

This slightly outrageous border plant sits out the earlier part of the season looking like an iris of moderate size. Then, just as most other border plants are looking seedy, it throws up a stalk from which emerge a series of star-shaped flowers of bright orange mottled with crimson.

I am not sure whether this plant has any vices. It has survived a brutal winter and seems to like a dry summer. I expect it splits as easily as most of its cousins in the iris family. Plant it wherever your late summer garden threatens to look sedate and tasteful.

Poem of the Week: The Phoenix and the Turtle by W Shakespeare

Good luck with making sense of this poem. In a way, you don’t need to. Two tips: the “turtle” is a turtle dove, and “the bird of loudest lay / On the sole Arabian tree” should be a phoenix, but later in the poem it is clear that it can’t be, since the phoenix turns out to be dead. Now carry on.

Let the bird of loudest lay,
On the sole Arabian tree,
Herald sad and trumpet be,
To whose sound chaste wings obey.
But thou, shriking harbinger,
Foul pre-currer of the fiend,
Augur of the fever’s end,
To this troop come thou not near.
From this session interdict
Every fowl of tyrant wing,
Save the eagle, feather’d king:
Keep the obsequy so strict.
Let the priest in surplice white,
That defunctive music can,
Be the death-divining swan,
Lest the requiem lack his right.
And thou, treble-dated crow,
That thy sable gender mak’st
With the breath thou giv’st and tak’st,
‘Mongst our mourners shalt thou go.
Here the anthem doth commence:
Love and constancy is dead;
Phoenix and the turtle fled
In a mutual flame from hence.
So they lov’d, as love in twain
Had the essence but in one;
Two distincts, division none:
Number there in love was slain.
Hearts remote, yet not asunder;
Distance, and no space was seen
‘Twixt the turtle and his queen;
But in them it were a wonder.
So between them love did shine,
That the turtle saw his right
Flaming in the phoenix’ sight:
Either was the other’s mine.
Property was thus appall’d,
That the self was not the same;
Single nature’s double name
Neither two nor one was call’d.
Reason, in itself confounded,
Saw division grow together;
To themselves yet either-neither,
Simple were so well compounded
That it cried how true a twain
Seemeth this concordant one!
Love hath reason, reason none
If what parts can so remain.
Whereupon it made this threne
To the phoenix and the dove,
Co-supreme and stars of love;
As chorus to their tragic scene.
Beauty, truth, and rarity.
Grace in all simplicity,
Here enclos’d in cinders lie.
Death is now the phoenix’ nest;
And the turtle’s loyal breast
To eternity doth rest,
Leaving no posterity:–
‘Twas not their infirmity,
It was married chastity.
Truth may seem, but cannot be:
Beauty brag, but ’tis not she;
Truth and beauty buried be.
To this urn let those repair
That are either true or fair;
For these dead birds sigh a prayer.

Published 1601