Alex Nowbar’s journal reviews—20 June 2019

Alex Nowbar reviews the latest research from the top medical journals

Annals of Internal Medicine

White coat hypertension

This isn’t any old high blood pressure. This is blood pressure that is only high in the doctor’s office. It can be one of two varieties, treated or untreated. The former is called white coat effect and the latter white coat hypertension. The difference is important because many believe the latter not to be clinically significant, thus steering away from antihypertensive medication. Cohen et al show us otherwise by looking at the risk of death and cardiovascular events in these two groups compared with people with normal blood pressure. They carefully meta-analysed 27 observational studies, in which people had blood pressure monitoring in and out of the doctor’s office. They found no difference in risk between people with white coat effect and those with normal blood pressure, but found increased risk in those with white coat hypertension. The caveat is that confounding is still likely to be have been present no matter the quality of this analysis. Hence the results are unlikely to directly affect patient care.


Blood donor sex

This study finds that people receiving transfusions from female donors who were previously pregnant or from sex-discordant donors do not have higher mortality. It’s good news because blood for transfusion is a limited resource, and it’s hard enough to match for blood group, let alone sex considerations. Edgren et al conducted an analysis of two US cohorts and a Scandanavian cohort of blood transfusions.  Less than a quarter of the transfusions were from parous or previously pregnant donors. In-hospital mortality was no different whether the donor was female, previously pregnant, or sex-discordant. The authors also looked at longer term mortality, and again there was no difference based on these features. Unfortunately, pregnancy status was not known for many of the donors, so imputation had to be used, but this may be the best data we can get on this subject.


Percutaneous coronary intervention in patients with high bleeding risk

Stents warrant at least a month of dual antiplatelet therapy. In people with a high bleeding risk, the aim is to use the stent type least likely to block without adequate antiplatelet therapy—bare metal stent. Drug-coated balloons may be a new alternative. This Finnish randomised controlled trial compared bare-metal stents with percutaneous coronary intervention with a drug-coated balloon rather than a stent. The balloon versus stent question has been addressed before. But these balloons were coated with paclitaxel and iopromide. And there were two key criteria for the patients to be included. Firstly the patient’s lesion had to be suitable for balloon treatment technically; patients were only randomised once the lesion had already been pre-dilated and appeared suitable for ballooning. Secondly, they had to have at least one risk factor for bleeding, ranging from simply being over 80 years old to taking oral anticoagulants. Rissanen et al found the drug-coated balloon resulted in significantly fewer target vessel revascularisations, myocardial infarctions and deaths at nine months compared to the bare metal stent (1% versus 14%). However, the trial was terminated early for slow recruitment so there were only just over 100 patients in each group. For a cardiovascular event trial, this is rather small and only the patients were blinded, but the results still seem robust.

New predictor of acute kidney injury

This cohort study found that elevated concentrations of dickkopf-3 in the urine (a marker for renal tubular stress) before cardiac surgery were associated with increased risk of acute kidney injury. This study may hail a new biomarker for predicting acute kidney injury. A cynic might ask what good is predicting risk if we can’t modify it. I’d reply that, at least, it can be used to help plan postoperative care and, at most, it is a step towards finding an intervention that modifies risk because it better identifies patients in whom to target the intervention. An investor in the urine assay might get rather excited and say this is “personalised medicine.” Or worse still, they might say this is “precision medicine,” as if we’ve just been fumbling in the dark until this test showed us the light.


New drug for sickle cell disease

Voxelotor is a drug name you might be hearing in future. It works by blocking HbS polymerisation which is the process that causes issues in sickle cell disease. Vichinsky et al conducted a double-blind randomised controlled trial of this new therapy. There were numerous signs of it reducing the effects of sickle cell disease like haemolysis and anaemia at 24 weeks. My only misgiving (albeit a fairly big one) is whether the main reported endpoints (namely blood test results) have any bearing on the important things in life, like how the patient feels or how they do in terms of sickle crises, need for transfusion, hospital admissions, and so on. For crises, voxelotor made no significant difference. 35% of patients in the two voxelotor groups had a treatment-related adverse event compared to 25% of patients in the placebo group, which is perhaps something to write in the main paper, if not the abstract, rather than page 17 of the supplementary appendices. While most of these adverse events were not determined to be serious, diarrhoea is not what the doctor ordered.

It’s too early to say whether this treatment has a net benefit for patients. A patient-reported outcome had been planned, but was removed from the study at the 12 week interim analysis for being too variable and the baseline scores being too low. Hardly reasons that could be described as watertight; they should still report what was found. It is only too common for things that improve markers of disease, not to actually help patients, and probably even more common for the medical profession to start using things without ever knowing either way.

Prevention of Type 1 diabetes

Everyone would be astounded if a drug could prevent type 1 diabetes. Is it going to be teplizumab, an anti-CD3 antibody? Herold et al tested a two week course of this intravenous drug in a double blind, randomised controlled trial of 76 people who had a relative with type 1 diabetes and were at high risk of developing the disease. To be included in the trial, they had to be over the age of 8 years and have diabetes-related autoantibodies and impaired glucose tolerance. The result was a clear reduction in progression to diabetes in the teplizumab group compared with placebo. Immune therapy might well modify the early course of type 1 diabetes, but more work is needed to assess safety and the optimal duration and frequency of treatments.

The infrastructure for this study and its precursor studies was provided by type 1 diabetes TrialNet which recruits thousands of people for type 1 diabetes research. This incredible initiative has moved from natural history studies to trials of potentially disease-modifying therapy. This is a great model of collaborative international research to emulate.

JAMA Internal Medicine

Light makes heavy?

Presence of artificial light at night while sleeping has been linked with obesity before. Park et al analysed data from over 40 000 women in the US and Puerto Rico, performed adjustments for confounders, and found the same link. The risk increase was statistically significant, but relatively small. This is fascinating and finding the mechanism even more so. Obesity is more than just overconsumption of calories, there is a complex interaction between neurological and metabolic processes. In terms of interventions to combat obesity, it’s easy to, say, avoid television and switch the lights off, but, on the basis of these data, I wouldn’t expect that to have much impact as it doesn’t get to the core of what is making the light lead to weight gain. Are there features of personality, mental state, or economic status associated with use of artificial light at night that could be contributing? These factors can’t be fully accounted for in an observational study.

Asthma self-management support

Federman et al conducted a randomised controlled trial of the SAMBA intervention for people over 60 with asthma compared to usual care in the US. The intervention was a year of individually tailored support from an asthma care coach with focus on medication adherence and inhaler technique to help tackle the psychosocial, financial, and cognitive barriers to good asthma control. There was a greater improvement in asthma control scores and quality of life in the intervention group compared to usual care, plus a reduction in emergency department visits. Who wouldn’t support such a marvellous intervention? Oddly, quite a few patients apparently, 30% of those approached to take part in this trial refused or did not turn up to the baseline interview. There is also the small matter of cost given that the protocol permitted flexibility of in-person and telephone encounters. However no information was provided on the cost of this intervention.


Coronary CT for diagnosis of stable chest pain

This non-invasive test is growing in popularity especially since it is now recommended in anyone with possible angina (although in reality thought to be best when the chance of having obstructive coronary artery disease is intermediate). Thus, the floodgates were opened, but justification isn’t too far behind.

Haase et al performed a meta-analysis of individual patient data from diagnostic accuracy studies of coronary CT angiography for stable chest pain. As usual, the usefulness of the test rides on the pre-test probability. This study identifies the best range of pre-test probability for which to use CT angiography was 7-67%. This translates to an extension from CTs for people with an intermediate pre-test probability to those with low or intermediate. The study also confirms that CT performs better in men and less well in older patients. Test accuracy was based on the presence of obstructive coronary artery disease at the invasive coronary angiogram, defined as at least a 50% reduction in vessel diameter. So yes, CT is an excellent test for deciphering coronary anatomy and is likely to avoid invasive tests for many, but no, it should not be for everyone.

Alex Nowbar is a clinical research fellow at Imperial College London.