POISE Study

This study showed that beta blockers (metoprolol) given 2 – 4 hours prior to  surgery and continued for 30 days led to a reduction in  non fatal myocardial infarction (3.6% vs 5.1%). However this benefit was offset by an increase in stroke rate (1.0% vs 0.5%). Also total mortality in the metoprolol group was higher (3.1% vs 2.3%). 

Commentators on this paper are suggesting that commencing beta-blockers peri-operatively for noncardiac surgery to prevent cardiovascular events (in at-risk patients) may lead to more harm than benefit.

 

 

Here follows an edited transcript of a presentation given by Prof Paul Myles, one of the POISE trial investigators.

This is also available as a podcast (10 minutes) Click here

I’m Dr Rod Tayler, a VMO Anaesthetist at the Alfred Hospital in Melbourne. I’m talking with Professor Paul Myles, Professor and Director of the Department of Anaesthesia and Peri-operative Medicine at the Alfred Hospital and Monash University. We are discussing the POISE trial. That is the Peri-Operative ISchaemic Evaluation trial, which was published in the Lancet in May 2008.

Paul, what was the trial looking into?

  The purpose of the trial was to test whether or not complete peri-operative beta blockade in patients at risk of coronary or cardiac events, could lead to a reduction in adverse outcomes. The purpose was to design the study in a routine peri-operative or anaesthetic setting. And it aimed to represent a broad range of patients in numerous hospitals at many sites around the world. That is, designed to be very much a practical or pragmatic trial.

Paul, what was the thinking in the anaesthetic world about routinely starting high risk patients on beta blockers prior to the POISE results being published? 

This has been a controversial area and is really illustrated by the fact that if you asked colleagues around the country and around the world you would find marked variation in practice. There had been a couple of small but influential trials published back in the mid to late 1990s suggesting that peri-operative beta blockers – atenolol or bisoprolol, could potentially improve outcomes and survival  in this setting. But there were also some other studies that were negative and these didn’t receive the same level of exposure or press that the two influential studies did. The two studies I’m referring to are the Mangano study published in the New England Journal in 1996 and the Polderman’s study from the Netherlands looking at bisoprolol in vascular surgery. This was published in the New England Journal in 1999. Both of these studies showed a marked reduction in mortality, short term and long term with beta blockers. It was these two studies that I think heavily influenced practice guidelines. These guidelines were consensus statements put out by the American Heart Association and the American College of Cardiologists that recommended peri- operative beta blockers, particularly in patients who had risk factors for coronary artery disease or were undergoing vascular surgery. These consensus guidelines together with those influential trials were considered by some groups to be so important that peri-operative beta blockers became a standard of care or a mark of quality of hospital care, at least in some parts of North America. Other anaesthetists, myself included, were never convinced by the weight of evidence, and considered that in fact the data was conflicting and there needed to be a more comprehensive large scale study before there should be recommendations to change practice. That is essentially why the POISE study was set up.

 

And what did the POISE trial show?

 

The POISE trial is a most interesting study. It does require careful consideration and debate before we fully understand what it shows and what its implications are. One thing I think is clear is that there is a reduction in myocardial infarction. That is one clear end point. This replicates what was demonstrated in those earlier studies. I think it is now beyond dispute that beta blockers in the cardiac and non cardiac surgical settings, and also in non surgical settings such as medical patients, are an important and effective intervention that should be used in many circumstances.

 But the trial showed some other findings that I think are at least if not more important, and I think are particularly relevant of those of us looking after surgical patients. This is where it differs from the medical (ie non surgical) situation. We know as anaesthetists that peri-operative beta blockers increase the risk of bradycardia during surgery and in the intensive care setting afterwards there is a higher incidence of hypotension. We know that. And therefore in this setting with patients with known coronary artery disease they are also at risk of other complications – organ failure, and in particular stroke.  This is where I think the POISE study is very important.  The POISE study demonstrates that overall there is an increased mortality and a higher incidence of stroke with peri-operative beta blockers. Therefore we have to weigh up the potential benefits to reduce myocardial infarction with the now demonstrated increase risks of stroke and death. And that’s were we are now.

 

The dose of metoprolol used in the trial has been criticized by some commentators.

Yes I think this is an unfair criticism If the POISE study had been a negative study , the first point that would of been made by any commentators would have been that  not enough beta blocker had been given. That is you didn’t properly beta block the patients. We know from both medical studies and the smaller trials in the surgical settings in the past that many patients who have been put on beta blockers are not adequately  beta blocked. And many experts in the world strongly recommended that the target heart rate should be around about 60 beats per minute. Certainly less than 75 beats per minute. And in fact if you read the previous guidelines from the American Heart Association and the American College of Cardiologists, if you read the detail of those documents both in 2002 and updated as recently as December last year, they once again make that point , that you have to target a heart rate of around about 60. Now to do that you need to give quite a lot of beta blocker.

It’s different in the medical world because the average physician will up-titrate beta blocker therapy in a new patient. Slowly and carefully introducing the new therapy. We haven’t got that luxury in a peri-operative setting, because patients arrive often on the day of surgery or perhaps in the days earlier, or perhaps in a semi-urgent or urgent situation and we have to beta block them acutely. If you accept the guidelines then the dose was the correct dose to use in the trial.

Now it needs to be appreciated that the trial protocol allowed practitioners to stop the beta blocker on any particular day or to halve the beta blocker dose if the patient had unwanted bradycardia or hypotension. In fact only a relative small proportion of the 8000 patients needed that to be done. So in fact the dose used was probably the correct dose, it did allow for a dose reduction if there was unwanted bradiacardia or hypotension, and it did I think adhere to the guidelines.

One of the other things which has been criticized about the trial is the timing of the initial dose.  What are your thoughts on that?

Well again, this is where I guess an ideal setting might differ from the real world. The medical experience of beta blockade is different from the surgical peri-operative situation that we are familiar with. In our setting we don’t have the luxury of these patients being seen in the weeks before surgery and having a slow titration of beta blockade.

The Mangano study was the pivotal study back in 1996. They beta blocked patients with atenolol an hour or so before surgery and that was really replicated in the POISE study . So I think it was a practical, realistic, and representative technique of peri-operative beta blockade.

Is there a need for more clinical trials with regard to this issue?

Well I think there is. I think the peri-operative beta blocker story is a very good example of many interventions in anaesthesia in peri-operative medicine.  That is how we change our practice and what evidence is required before we change our practice. There now have been many excellent beta blocker studies and now the definitive POISE  trial. Even so, we have some unanswered questions. As anaesthetists we need to understand there’s rarely a single landmark study that answers all questions once and for all. I think this study, the POISE trial, raises a few new questions. That is, could a lower dose of beta blockers be used more safely so you gain all the benefits of avoiding myocardial infarction but also avoid the risks of severe hypotension, stroke and  increased mortality.

It also raises a question about patients who are all ready on beta blockers. The current recommendations by experts are still stating that the POISE results don’t apply to people who are already on beta blockers. In my own practice I am now more cautious about maintaining  peri operative beta blockade, certainly at higher doses and certainly in patients who are at greater risk of major blood loss or fluid shifts or a stormy post operative course. So these are the studies that need to be done: lower doses, titrated doses, more targeted studies in patients who are in fact at very low risk of stroke or  hypotension but at higher  risk of unstable coronary syndromes.

 

Thanks Paul for your comments on the POISE trial and the peri-operative use of beta blockers.