Richard Lehman’s journal reviews—12 March 2018

Richard Lehman reviews the latest research in the top medical journals

richard_lehmanNEJM  8 Mar 2018

Increasing inhaled steroids in worsening asthma

Corticosteroids have long been the main rescue treatment for exacerbations of asthma: puffers first, then pills if needed. The logic of this tradition seems to be that steroid puffers are safer than steroid pills, especially for children. But that simply wasn’t the case in an excellent double-blinded trial

“In children (aged 5 to 11) with mild-to-moderate persistent asthma treated with daily inhaled glucocorticoids (fluticasone), quintupling the dose at the early signs of loss of asthma control did not reduce the rate of severe asthma exacerbations or improve other asthma outcomes and may be associated with diminished linear growth.” In fact the high-dose puffer group ended up with higher total steroid exposure plus a higher rate of exacerbations. “If you feel it’s getting worse, take more of your brown inhaler” is poor advice for children.

But what about adults? It may well be the same: we simply don’t know, because in this case the trial was much less rigorous and compared “a self-management plan that included an increase in the dose of inhaled glucocorticoids by a factor of 4 (quadrupling group) with the same plan without such an increase (non-quadrupling group), over a period of 12 months. The primary outcome was the time to a first severe asthma exacerbation, defined as treatment with systemic glucocorticoids or an unscheduled healthcare consultation for asthma.” There were 420 (45%) exacerbations in the quadrupling group compared with 484 (52%) in the non-quadrupling group. This squeezes into the bounds of statistical significance, but if 7% more people have to take a few days of prednisolone or go to see a doctor, this does not strike me as clinically very significant. Remember, the trial was not blinded. And in the end it’s mainly a choice of how you like to take your steroid.

Infarct in the outback

All journals—even the aristocratic NEJM—enjoy a headline-grabbing story. This week features a celebrated letter about a nurse who was 1000km from the nearest large hospital. He sent in two ECGs from a 44-year old man who was clearly having a large inferior MI with major arrhythmias. Fortunately he had on hand a full kit of MI treatments, including tenecteplase. So he set up drips on both arms, which happened to belong to him, and proceeded to thrombolize himself, defibs at the ready. Someone on Twitter added that he meanwhile killed a venomous brown snake with the IV pole but was too modest to mention it. The letter is written by three doctors who treated him later, so the identity of Infarct Dundee remains shrouded in mystery.

JAMA 6 Mar 2018

Opioids for back and lower limb joint pain?

Triallists who compare treatments for chronic pain should learn from real life and carry out large series of double-blinded n-of-1 trials. Participants would end up on the drugs that worked best for them personally, and researchers would discover the range of responses, and accumulate data on what might predict them. But although this seems obvious, such trials are actually rare. Instead, we get trials which use traditional randomised groups and obtain average results. But because pain is a symptom which would be cruel (and impractical) to ignore, there is usually an escape route permitting diversions from protocol. So it is with the SPACE trial of opioid vs non-opioid medications on pain-related function in patients with chronic back pain or hip or knee osteoarthritis pain. The group assigned to opioids generally stayed with different doses of various opioids, whereas the non-opioid group were allowed tramadol as a last resort. In fact only a few patients received it. Retention was lower in the opioid group, but overall follow-up was excellent at 98% after a year. “Treatment with opioids was not superior to treatment with non-opioid medications for improving pain-related function over 12 months. Results do not support initiation of opioid therapy for moderate to severe chronic back pain or hip or knee osteoarthritis pain.” This seems broadly true, though perhaps “initiation” should read “initial treatment.”

Single PSA screening and futile overdiagnosis

This is one of the most important studies ever done in British primary care. The Cluster Randomized Trial of PSA Testing for Prostate Cancer (CAP) was designed 18 years ago to determine the effects of a low-intensity, single invitation PSA test and standardized diagnostic pathway on prostate cancer–specific and all-cause mortality, while aiming to minimize overdetection and overtreatment. Now, after a median of 10 years follow-up, it shows no significant difference in prostate cancer mortality between screened and unscreened symptomless men (over 400,000 in the two arms). Yet there was a nearly 5-fold difference in the diagnosis of prostate cancer between groups in the first 18 months: 10.42 per 1000 person-years in the intervention group compared with 2.18 per 1000 person-years in the control group. Over subsequent years, the difference waned, but the headline message is clear: the harms of single PSA testing greatly outweigh any possible benefits, because it grossly over-detects harmless early neoplasia. PSA has Perfectly Stupid Attributes as a population screening tool. And as I said on Twitter, researchers who work to timescales like this to resolve real clinical questions are true heroes.

Ann Intern Med 6 Mar 2018

Glycated haemoglobin in type 2 diabetes

Ten years ago, three major long-term trials showed that in people with type 2 diabetes that was stable over 8-10 years, there was no benefit from reducing the HbA1c below 7%. Silence followed. A year later, GPs were incentivized by the Quality and Outcomes Framework to reduce HbA1c below 7% in half their patients with type 2 diabetes. I wrote a protesting editorial with Harlan Krumholz in The BMJ, but this payment incentive stood for 3 years until it quietly disappeared. This taught me a lot about evidence-based medicine. It can be pathetically weak when confronted with an established narrative that appeals to patients and doctors alike and sells lots of drugs. Now in 2018 the latest American College of Physicians guidance on Hemoglobin A1c Targets for Glycemic Control With Pharmacologic Therapy for Nonpregnant Adults With Type 2 Diabetes Mellitus embodies an evidence-based, shared knowledge approach, aiming for a level of 7-8% HbA1c in most patients. Hurrah! I hope we will soon see similar guidance from interventional cardiologists on the role of stents in stable angina, 11 years after COURAGE and half a year after ORBITA.

The Lancet 10 Mar 2018

Fatty liver: the untapped market

Nearly a third of men in the richest countries have fatty livers: women and middle-income countries are racing to catch up. How about that for a market? An unlucky few of these will develop fibrosis and of these an unlucky few will die of liver failure or hepatocellular carcinoma. So you have your advertising copy sorted too. But by the same token any drug given to such a vast number of people will have to be very safe and will need some very lengthy hard-end point trials. And there is an excellent cheap candidate drug already—pioglitazone, which can reverse cirrhosis as well as steatosis. But never mind all that. Let’s read about a 116-patient phase 2 trial of an injectable experimental drug, lasting 12 weeks. In that time, a third of the participants experienced diarrhoea and three-quarters lost more than 5% of their liver fat. “Larger clinical trials of longer duration are now needed to fully assess the safety and efficacy of NGM282.” Might The Lancet not have found better use for its space in the meantime?


When single-author reviews are good, they tend to be very good, and this one by David Dodick is excellent. I bet he gets migraines himself, because it is written with real understanding and a passion for completeness. I read every migraine review that I come across, on behalf of someone who gets them much worse than I do: but if you read only one, make it this (if you can get a copy).

The BMJ 10 Mar 2018

Vitamin D and cancer in Japan

I also tend to read lots about vitamin D, though I’m beginning to think that a moratorium on further observational studies might be a good idea. This one shows that in a large cohort of Japanese adults, cancer incidence since 1990 was lowest in those with the highest levels of vitamin D. I like sunshine and fish. Wake me up when there is a positive randomised trial of vitamin D pills for anything.

It’s a cancer drug: who needs evidence?

Two medical students under Vinay Prasad’s guidance examine how the American National Comprehensive Cancer Network (NCNN) arrives at its recommendations. “Just 23% (10/44) of these recommendations are supported by evidence from randomized controlled trials, while most do not provide references or are based on small, uncontrolled studies or case reports. In contrast, among 83 consecutive cancer drug approvals by the FDA over a similar period, 58% (n=48) were based on evidence from randomized controlled trials.” Note that the FDA is hardly a shining example: 42% of their recommendations were not based on RCTs, and many of those that were relied on “progression-free” survival or other surrogates rather than actual survival. There is no such thing as compassionate capitalism, and oncology drug pricing is the proof. Even more worrying is that commercial capture of regulation is now an accepted fact. We have learned helplessness, and our health systems and cancer patients pay the price.

Plant of the Week: Iris reticulata

English gardens just now are looking rather flattened and muddy following the Beast from the East. It is hard to believe that within a month, all heaven will break loose.

In our own garden, the oriental hellebores provide the greatest glory so far, though some of the ornate doubles show a bit of frost damage. From time to time, we’ve planted the odd packet of Iris reticulata bulbs, but they tend to disappear, perhaps because we are very successful at raising slugs in our limy clay.

Nearby friends have greater success by planting theirs at the edge of a lawn. They really are the prettiest little things: they come in some striking shades of clear blue if you can find the right ones. If you like a bit of purple in your blue, or care for bright yellow, you won’t have to go any further than your nearest garden centre. But you should have done that back in October.