Richard Lehman’s journal blog, 19 July 2010

Richard LehmanJAMA  14 July 2010  Vol 304
163    The classic hero of palliative care used to be the personal doctor who turned up in the middle of the night to administer symptom relief and consolation. Most doctors have done that from time to time, and found it immensely satisfying and immensely tiring. So what, by contrast, is the effect of centralised telephone-based care management coupled with automated symptom monitoring on pain and depression in patients with cancer? Also good, is the verdict of this randomised trial from Indiana, where there are lots of scattered rural communities, making heroism less practical. It resulted in better pain control and relief of depression than “usual care”. Yet palliative care ideally comprises not only such routine systems for averting symptom crises, but also good systems of immediate care, and always the availability of human comfort.

172    Most childhood cancers are now followed by long term survival, but at a high cost. Here is the latest report from the cohort of nearly 18,000 British children followed up after cancer treatment from 1940 onwards. Overall, there were 11 times the expected number of deaths in these survivors. Of these nearly two-thirds were due to recurrent or progressive initial malignancy, but that still leaves an awful excess of second cancers and cardiovascular disease. Even after thirty years, the death rate is tripled.

194    It’s often stated that before antiretroviral treatment, HIV infection was uniformly fatal, but that is not quite true. About one infected person in 200 failed to develop progressive disease while remaining untreated. These people (LTNPs – long-term non-progessors) are intensively studied, of course, for what they might be able to tell us about mechanisms to protect everybody from this virus. They fall into two broad categories according to their detectable viral load and CD-4 counts but most remain healthy without any treatment for 20 years. Read on.

NEJM  15 July 2010  Vol 363

211    ANCA-associated vasculitis is nasty: in the days before it was called that, it killed most people who had it, by causing renal failure and respiratory tract inflammation and necrosis complicated by infection. Cyclophosphamide and high-dose corticosteroids are effective treatments, but mortality in the first year still exceeds 15%. Then along came rituximab, a monoclonal antibody aimed at knocking out B-cells expressing CD20. It was hoped that this would be a magic bullet treatment that would avoid all the dangerous effects of cytotoxic drugs and steroids. Here are two trials in patients with newly presenting Wegener’s granulomatosis, Churg-Strauss syndrome and other ANCA-associated diseases. The first demonstrates non-superiority; the second (p.221) demonstrates non-inferiority. In other words, rituximab treatment is still associated with a 15%+ first year mortality and similar adverse effects to existing standard regimens using cyclophosphamide and prednisolone. The niche use of rituximab will be for patients who fail to respond to these – see editorial on p.285.

233    By now, most of you should be aware that ACCORD was a big trial of various interventions in type 2 diabetes that was stopped early because of excess mortality in the group randomised to tight control of glycaemia. Death is generally agreed to be the most important and definite end-point in any study. Diabetic retinopathy is not quite in the same league, and a good deal harder to define. ACCORD used a definition based on three points in a visual acuity scale and/or the need for intervention (laser photocoagulation or vitrectomy). Just as in UKPDS and other studies, tighter glycaemic control achieved a small but significant reduction in these end-points. But the biggest reduction (about 40%) was seen in patients randomised to receive fenofibrate in addition to statins. This is seized upon with enthusiasm by an ophthalmologist writing on p.287 . Reading this editorial and last week’s Lancet review, I’m beginning to worry that diabetic retinopathy can cause tunnel vision.

245    If anything can cause a company’s profits to BLOOM, it’s a new obesity drug. The BLOOM (Behavioural Modification and Lorcaserin for Overweight and Obesity Management) trial was funded by Arena Pharmaceuticals, who will be hoping for vast returns on the latest drug to target the serotonin receptor. Those with supernaturally good memories and profound knowledge of clinical pharmacology (OK, you can put your hand down, Jeff Aronson) will remember that there are actually three such receptors and that previous anti-obesity drugs such as fenfluramine and dexfenfluramine targeted them non-specifically. They worked fairly well for appetite suppression but were withdrawn because they could cause valvular heart defects and pulmonary hypertension. This is because cells around the heart valves and in the pulmonary vasculature contain 5HT2B receptors whereas the receptor you need to hit for appetite suppression is 5HT2C. Lorcaserin is powerfully specific for this receptor and Arena went out of their way to check their trial subjects regularly with echocardiograms which prove that it doesn’t cause heart valve problems in the first two years. Whereas it certainly does help people lose weight and will be advertised as blooming wonderful if and when it gets it licence.

266    In reviews of acute pulmonary embolism I look for two things: mention of it as a common cause of exacerbations in heart failure and COPD, and guidance about which patients need long-term anticoagulation. This article by two Italian authors doesn’t fully satisfy either criterion. There’s little mention of HF or COPD and although they say that “extended treatment requires a reassessment of the patient’s risk-benefit ratio at periodic intervals” they fail to tell us how to calculate these risks and benefits.

Lancet  17 July 2010  Vol 376

163   Droves of healthy people come to see doctors all year round to have blood pressure checks. If it’s off target, their GP sees them every few weeks to make adjustments. Neither the timing, the place nor the health professional involved reflects any real logic. This ground-breaking study (TASMINH2) addresses these realities by passing management to the patient whose blood pressure is monitored at home with a reliable automatic device linked by an automated modem to the GP practice. If it remains high, the patient is given advice and if necessary additional drug treatment to reduce it. The group randomised to this intervention showed usefully better control of systolic BP at the end of a year. If this technology became widespread, we would save many GP appointments and improve control in most of our hypertensive patients.

173    Recently we were urged by our prescribing adviser to change everyone possible from celecoxib to a cheaper combination of NSAID with omeprazole if gastric protection is needed. This was largely cost-driven; but there is also an appreciable difference in cardiovascular risk between celecoxib and naproxen and ibuprofen, for example. Both celecoxib and diclofenac carry approximately the same twofold risk of cardiovascular events, which may be why Pfizer chose this NSAID as the comparator in the CONDOR study looking at the risk of gastrointestinal events. They have demonstrated to their own satisfaction that celecoxib is a safer choice than diclofenac plus omeprazole in patients with a history of GI bleeding. But for most patients needing NSAIDs, neither celecoxib nor diclofenac would be the best choice for overall long-term safety.

180    Most of us are vitamin D deficient for at least part of the year, some more than others. The overriding determinant is skin exposure to sunlight, followed way behind with diet, and some way behind that with genetic determinants. This genome-wide study was even more labour-intensive than most, covering 34 000 subjects in five discovery cohorts, plus five “in-silico replication cohorts” (I think this quaint term refers to stored samples)  and five de-novo replication cohorts. Unlike many of these prodigious exercises in gene gnomery, this one produced results: variants at three loci reached genome-wide significance, and together they can account for a 2.7-fold difference in the risk of vitamin D deficiency. However, this brings no nearer the day when we will get round to detecting and treating this important and ubiquitous condition. These gene gnomes need to get out more, and so probably do the rest of us.

BMJ  17 July 2010  Vol 341
135    This interesting observational study discovers that opioid addicts who regularly take prescribed methadone are less likely to die than those who do not, even though they continue to inject street opioids. This is purely an artefact of the way we criminalise drug addiction. Addicts who remain in regular contact with health professionals by supplementing their street use with methadone are displaying a less risky behaviour pattern than those who don’t.

139    The latest in Michael Goldacre et al’s regular analysis of early career choices by doctors compared with their eventual specialties. The pico format kills this sort of thing stone dead. You can’t loll in an armchair squinting at the tables one by one, then flicking back and forth to make sense of them. All you’re given in the printed BMJ is a bare little summary table showing that the longer you’ve been a doctor, the more likely you are to have found your eventual specialty.

140   Thank goodness the BMJ still allows readers the full text of its clinical reviews. Obstetric anal sphincter injury is an important topic which I haven’t ever read a good review about, but I fear that this one may suffer from some over-enthusiasm on the part of its expert authors. On the basis of another review published in 2003, they declare that: “A thorough clinical examination of the perineum and vagina, including a digital rectal examination, should be performed after every vaginal delivery to improve the detection rate of anal sphincter injury.”

Following every vaginal delivery? Not just that, but “visual inspection should be combined with palpation, including a pill rolling movement using the index finger in the rectum and the thumb over the anal sphincter to assess muscle bulk.” I do hope they have some evidence that all this anal fixation can result in better outcomes.

146   It’s always a source of pleasure to me when one of my original suggestions for a BMJ series to be called “commoner than you think?” appears under the eventual series title of Easily Missed, and this week it’s bronchiectasis. As with many of these topics, the reason we have become more aware of its prevalence is better imaging – no more horrible endoscopic contrast bronchograms, just a nice CT scan. But  the patient may still be in constant danger of misdiagnosis by unfamiliar doctors, especially when the basal crackles are bilateral. Some should be protected by a tattoo over a marked area of their backs, reading “These crackles are always here: I do not need furosemide”.

Arch Intern Med  12 July 2010  Vol 170

1135    Vitamin D again. The InCHIANTI cohort study finds a small but significant relationship between vitamin D levels and the risk of dementia. The average MMSE score in these 70+ year olds in Tuscany began surprisingly low at under 25 and declined slightly faster according to quartile of serum 25-hydroyvitamin D. It’s interesting to note that 80% of these D-deficient Tuscans were the black-clad women who stay at home to cook meals for the men. The vitamin-rich men sit with their friends in the sun on benches dressed in dark suits with open neck white shirts, uttering the odd word every half hour.

Plant of the Week: Eryngium alpinum

This Sea Holly is an old favourite, introduced to England from Europe in 1597. It can be one of the best things in the garden at this time of year, if you give it a good start in well-drained soil in full sun. Once established it will grow in limy rubble quite happily.

The flowers are large and look prickly, but unlike all other sea holly flowers, they are not. Their excellent blue makes a nice contrast to the yellows and oranges of the day-lilies which provide most of the garden colour at this time of the year.