Alex Nowbar’s research reviews—29 October 2018

Alex Nowbar reviews the latest research from the top medical journals


A delirium don’t—antipsychotics for delirium in intensive care

Some clinicians describe delirium as organ failure in which the brain is the failing organ—a horrifying thought. Even more horrifying is the lack of treatment for it. So I commend Girard et al for conducting a large double-blind randomised controlled trial of antipsychotics in the intensive care setting. The strategy they used was titration of the drugs every 12 hours based on the presence or absence of delirium using the confusion assessment method for intensive care and based on side effects e.g. extra-pyramidal symptoms. Patients received haloperidol, ziprasidone, or placebo for a median of four days. Antipsychotics were not found to improve outcomes of delirium, survival, or length of stay. This is a stark reminder that we have no good interventions for delirium. Adverse events weren’t too alarming in the treatment groups, but in the absence of supporting evidence for the drugs, best practice for now might be non-pharmacological measures.

The cosmic consequences—MRI brain changes after space flight

Any knowledge I have of astronaut’s health comes from a film called The Right Stuff (originally a book). It’s about US test pilots in the 60s. It shows the medical testing they underwent to check if they were fit to go into space. The scientists were checking to see if they had “the right stuff”. But the scientists clearly had no idea what this was (for example one of the medical tests involved a barium enema). In other words, knowledge of the effects of going into space on the body was almost none.

This NEJM letter to the editor shows us we are not much further on in understanding these effects today. Ombergen et al looked at T1 weighted MRI brains of 10 astronaut before and after a space flight. You might rightly be wary of this pre and post comparison because there is no control group. However, the authors report that they account for the possibility of aging effects that may occur over the interval between preflight and postflight by longitudinal data from matched controls.

They found a loss of grey-matter volume after the flight, but longer term follow-up showed this returned to preflight levels. They also found that CSF volume increased in the subarachnoid space and continued to do so months after a return to Earth. But what does this mean for the individuals concerned?

The letter ends aptly, “Further investigation is required in order to determine the overall clinical significance of the findings”. This is worth remembering for any imaging study that doesn’t report clinical outcomes.

A concept crush—Pantoprazole for gastrointestinal protection in intensive care

This large Danish double blind randomised controlled trial of pantoprazole versus placebo is another pragmatic study set in intensive care. The trial was negative for the primary endpoint of death at 90 days. “Stress” gastrointestinal bleeding is a common problem in intensive care. It is easy to make the conceptual link that a proton pump inhibitor would prevent this, but conceptual linking isn’t good enough. We need evidence like this to measure the benefits and risks of what we do so that we make informed decisions.

There was no difference in adverse events between the groups in this study (although they presented the secondary outcomes without p values as they didn’t adjust for multiple comparisons). The overall interpretation of this study is fairly flexible. You could conclude that giving pantoprazole is a waste of resources. Or you could conclude that even if pantoprazole is unlikely to have meaningful benefit, giving it to selected patients won’t be especially harmful either. Perhaps the real take-home message is that interventions that seem to make sense conceptually aren’t necessarily of benefit.

An iron investigation—Proactive or reactive treatment of anaemia in haemodialysis

Treatment of iron deficiency anaemia in haemodialysis is a controversial and high stakes area of research. Adverse events with intravenous iron compared to oral iron have featured prominently in previous studies.

The PIVOTAL trial was an open-label non-inferiority randomised controlled trial in 50 UK sites comparing high dose proactive intravenous iron with low dose reactive intravenous iron to treat anaemia in patients on maintenance haemodialysis. The primary endpoint was a composite of nonfatal myocardial infarction, nonfatal stroke, hospitalization for heart failure, or death. They found the high dose regimen was noninferior to the low dose regimen. The trial was funded by the pharma company that sells the iron products used.

With pharma funded professional medical writing assistance, the authors introduce the study aptly by saying “Rigorous scientific evaluation of the use of high doses of iron in patients undergoing hemodialysis has been limited, which has resulted in marked variation in its use among individual practitioners and across countries”. Sadly unblinded trials such as this aren’t “rigorous scientific evaluation”. Merely knowing which arm patients were in could affect patient and staff behaviour and therefore impact on the outcomes. I’m afraid blinded endpoint assessment isn’t enough.


A tricky trial—Ventilation weaning strategies in intensive care

The BREATHE trial was another randomised controlled trial set in intensive care. It must have been challenging to conduct a trial in patients requiring mechanical ventilation who were difficult to wean. They found that noninvasive protocolised weaning was no better than an invasive strategy in reducing the time to liberation from ventilation. This endpoint is important because weaning off ventilation is resource-heavy. The patient group was mainly pneumonia and postsurgical respiratory failure. Only 3.9% of patients were admitted for COPD which was the primary reason for admission in previous trials of noninvasive ventiliation to facilitate weaning. This is important because invasive ventilation for COPD has become less common since these trials. This trial was a thoughtful approach to a tricky area.


A migraine marvel—Erenumab for difficult-to-treat episodic migraine

The LIBERTY trial was a pharma-sponsored double blind randomised controlled trial of an injectable monoclonal antibody that inhibits the calcinton gene related peptide receptor. It enrolled patients with episodic migraine in whom 2 to 4 preventative treatments had not worked. The primary endpoint was the proportion of patients with a 50% or greater reduction in the mean number of monthly migraine days. At 12 weeks 30% of patients in the erenumab group had reached the primary endpoint compared to 14% in the placebo group (p=0.002). This is certainly a good signal, but the patient group was quite specific and the results can’t be generalised beyond that.

A gouty glory—Nurse led care encouraging shared-decision making for gout

Doherty et al randomised 517 patients to nurse-led care involving patient education and engagement with treat-to-target urate-lowering or GP-led usual care. The primary endpoint was the percentage of participants whose urate levels were less than 360 micromols at 2 years. This was achieved in 95% in the nurse-led group and only 30% in the usual care group (p<0.001). They present a detailed cost-effectiveness analysis in the supplementary appendix which is interesting, but perhaps a little premature. People who take part in trials may be more likely to engage with care and adhere to medications (95% adherence at 2 years was seen in this trial) than other patients with gout. The intervention was delivered by specially trained research nurses so can’t really be scaled up in its current form. The cost-effectiveness of this intervention once adapted for widespread implementation may be different. Still, this study is a good start.


A confounder corker—ACE inhibitors and lung cancer

The headline in the media: “Blood pressure drugs linked to raised risk of lung cancer”. As a member of the public, I think that a reasonable response from others to reading this might include fear of blood pressure drugs if they aren’t on them, fear of lung cancer if they or their family are on them and, sadly but understandably, loss of confidence in the medical profession. The public expect drugs have been checked for danger. They won’t be aware that there wasn’t an increased risk of lung cancer in the randomised controlled trials (even though follow-up was only a few years). And they won’t be aware of the magnitude of benefit of ACE inhibitors for prevention of stroke, cardiovascular disease, and death. But perhaps the headline is justified. Let’s delve into this cohort study from Hicks et al to find out.

They found ACE inhibitors were associated with a 14% increased risk of lung cancer and risk was particularly elevated with longer term use i.e. after 5 or 10 years. The comparator group was patients on ARBs. They adjust for many confounders that could have caused the relationship such as smoking. The incidence of lung cancer in the ARB group was 1.2 per 1000 person years and 1.4 in the ACE inhibitor group. Yes, that is some extra cases of lung cancer given how common ACE inhibitor use is, but it is hardly alarming. They discuss that the cough associated with ACE inhibitors could have led to lung investigations thus increasing the rate of lung cancer detection, but they conclude this hypothesis has no merit (although I disagree). They did not comment on whether the detected lung cancers were of clinical significance which is probably the only thing the matters.

The study is well done, but the data will always be limited by being only observational. Perhaps what would have helped for the media coverage would have been further patient and public involvement in the dissemination plan for these results to ensure proportionality and context.

Annals of Internal Medicine

A biomarker bonanza—Chronic kidney disease after intensive blood pressure lowering

I imagine many a patient has been tormented after a blood test by being labelled with chronic kidney disease despite being asymptomatic. Blood pressure lowering drugs can often drive these results. This case-control study nested in the SPRINT trial provides some good data on the pathological significance of this development of CKD by looking at urinary biomarkers of kidney damage. (SPRINT was a large randomised controlled trial comparing intensive versus standard blood pressure lowering.) It found that new cases of CKD in patients having blood pressure lowering did not show signs of intrinsic kidney damage. In fact it seemed the kidneys were better off. The reduction in eGFR may be benign and related to the drugs reducing blood flow to the kidneys but this may not be damaging. The authors conclude “the perception of a tradeoff between cardiovascular benefits and kidney harms during intensive blood pressure lowering may be misguided”. This may not be randomised data, but it’s a good research question to be addressing.

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

Competing interests: None declared