CE-MARC 2: Lowering Unnecessary Angiography Rates With CMR

John M. Mandrola, MD; John P. Greenwood, MBChB, PhD, FRCP


September 28, 2016

This feature requires the newest version of Flash. You can download it here.

John M. Mandrola, MD: Hi, everyone. This is John Mandrola from on Medscape. I'm here in Rome, Italy, at the European Society of Cardiology (ESC) Meeting, and I am delighted to have Professor John Greenwood from the University of Leeds, who is the chief investigator of the CE-MARC 2 trial. Professor, welcome.

John P. Greenwood MBChB, PhD, FRCP: Thank you.

Dr Mandrola: Can you tell us a little bit about this trial,[1] starting with the background?

Dr Greenwood: One of the things we recognize in cardiology is that patients with suspected coronary artery disease often have invasive angiography quite early in their management pathway. Studies from the United States and around the world have shown that when we do invasive coronary angiograms, up to 60% of them can find no evidence of obstructive disease.[2] This clearly suggests that we are using angiography liberally and, perhaps, the rates of invasive catheterization are too high.

Paradoxically, if we look at a lot of the guidelines on how we investigate patients for suspected coronary artery heart disease, they use pretest risk models. Some of those models can overestimate risk, so again, they tend to push patients toward an invasive catheter approach rather than noninvasive diagnostic imaging.

We wanted to design a trial to see whether we could challenge the most contemporary set of guidelines that were available at the time and use advanced noninvasive imaging techniques to see whether we can reduce the rates of unnecessary angiography.

CE-MARC 2 was a prospective, multicenter, randomized, controlled trial that was funded by the British Heart Foundation comparing cardiovascular magnetic resonance (CMR) with the UK National Institute for Health and Care Excellence (NICE) guidelines[3] from 2010 and myocardial perfusion scintigraphy (MPS). MPS is the most widely used ischemia detection test worldwide.

Dr Mandrola: So it was a test of CMR vs the guidelines?

Dr Greenwood: That's correct. The main hypothesis was that we could reduce the rates of unnecessary angiography using CMR compared with the UK NICE guidelines. It was also compared with a strategy that involved MPS.

Dr Mandrola: Before we go to the methods, tell me about CMR.

Dr Greenwood: CMR is mature now. It's an imaging technique that has been used for over 15 years in routine clinical practice. We use it widely in the United Kingdom and Europe. I think it's a little bit less utilized in the United States.

CMR is used for a whole range of cardiovascular diseases. Obviously, we are very interested in coronary artery disease, and it is often used as a first-line test at our institution and many other institutions around the United Kingdom to diagnose and manage patients with suspected coronary heart disease.

Functional Assessment

Dr Mandrola: Is it an anatomical test?

Dr Greenwood: No; it's an anatomical and functional test, and that is very important. In addition to getting static anatomical images, we get dynamic images. We get first-pass perfusion images, we get functional images of the ventricle, we can image the coronary arteries, and we can look at the myocardium to see whether it is alive or dead. We get a multiparametric scan. It gives us many different viewpoints of the same disease process in a single test.

Dr Mandrola: What was the trial design and some of the methods?

Dr Greenwood: It was a prospective, multicenter, randomized, controlled trial across six UK centers comparing CMR-guided strategy with the UK NICE guidelines and an MPS-guided strategy.

Dr Mandrola: Who were the patients? What types of patients were they, and where did they come from?

Dr Greenwood: That's a very important question. We wanted to design a trial that was very generalizable, so we took all patients with a pretest likelihood ratio of 10%-90% of having coronary heart disease as per the UK NICE guidelines, and we managed them in that arm according to the guideline recommendations.

Dr Mandrola: Did you exclude patients with noncardiac chest pain?

Dr Greenwood: Yes. We also excluded the very low-risk patients—those with < 10% probability of having coronary artery disease. We know that the very low-risk patients can often produce more false-positives than disease detection because they are so ultra-low risk. Also, patients with noncardiac chest pain have alternative pathologies that you might want to investigate. So we were really focused on the broad group of de novo first-time presentation patients with suspected anginal chest pain.

Dr Mandrola: From emergency departments, chest pain centers, or the outpatient setting?

Dr Greenwood: These were stable outpatient angina patients or suspected angina patients.

Main Findings

Dr Mandrola: Tell us about the results.

Dr Greenwood: The primary endpoint was the rates of unnecessary angiography across the three groups, which we defined by invasive catheterization. We used the international invasive reference standard of fractional flow reserve (FFR) to prove that there was hemodynamic significant disease. Where that wasn't clinically possible—for safety reasons, for example, in critical stenosis—then we used quantitative coronary angiography (QCA). Importantly, all of the FFR data and all of the QCA data were sent to a blinded independent lab in Glasgow for analysis, because it was very important that we had a robust primary endpoint.

Dr Mandrola: What were the findings of the primary endpoint?

Dr Greenwood: The findings were that a CMR strategy highly statistically significantly reduced the rates of unnecessary coronary angiography compared with guidelines-directed care. There was no difference at all between the CMR strategy and the MPS strategy. Both of them significantly reduced rates of coronary angiography. In the guidelines group, we found that the rate of unnecessary angiography was about 29%, and in the noninvasive functional imaging groups, it was about 7%.

Dr Mandrola: That's a pretty low hazard ratio, comparatively. You also looked at a secondary endpoint.

Dr Greenwood: That's correct. We built into the trial some safety secondary endpoints. We wanted to look at short- and medium-term major adverse cardiac events. What we saw was that there was no difference in major adverse cardiovascular events across the three groups at a minimum of 12 months of follow-up. Obviously, we will be following these groups of patients for a longer time.

Another important thing that we built into the trial was the disease detection rates, and no difference was found in the rates of having positive coronary angiography across the three arms of the trial.

Dr Mandrola: Did all patients get catheterization? Can you get a true estimate of the false-positives?

Dr Greenwood: No. That's another important point. In this trial, about 20% of the patients had invasive catheterization. If you want to calculate the sensitivity and specificity, then you need to know the true denominator.

We did that study with the CE-MARC 1 trial, which was published in the Lancet in 2012.[4] We did a head-to-head comparison of CMR vs MPS with angiography as the reference standard, and every patient was scheduled for every test of that study. The results showed that CMR was statistically significantly better in sensitivity and negative predictive value over MPS.

Dr Mandrola: Let's move into the discussion. Tell me about the NICE guidelines and the pretest likelihood calculation.

Dr Greenwood: The UK NICE guidelines were the most contemporary set of guidelines available when the trial started in 2010. They used the Duke Clinical Score to estimate pretest probability. That score was generated about three decades ago, so perhaps it doesn't reflect the contemporary population of patients, contemporary risk factor burden, and contemporary treatments.

Interestingly, there have been updated models to the Duke score. Tessa Genders and colleagues has published a modified version[5] that confirms that the Duke score overestimates coronary heart disease risk. In fact, those are now incorporated into the latest 2013 ESC guidelines.

Dr Mandrola: It's not surprising that the results came out this way?

Dr Greenwood: In hindsight, you might see the results and say that a lot of them were driven by the desire to catheterize patients who were deemed high-risk. In some way, that does reflect our clinical practice. When we talk to our cardiology colleagues, many of them still say, "If I really think they have angina and they are high-risk, I just want to catheterize them." So, it does reflect practice.

It also tells us that we need to be very careful when we construct guidelines, because sometimes the data that we use and the expert opinion that creates guidelines can change the mechanism of investigation. We shouldn't be afraid to do trials against the guidelines to test them out in a randomized controlled fashion.

Dr Mandrola: Can you speak about the controversy about the difference between finding anatomical coronary disease and actual coronary disease that causes ischemia, and how that might relate to outcomes?

Dr Greenwood: Yes. There is a wide range of noninvasive imaging techniques. You can broadly divide them into anatomical techniques or functional techniques.

Anatomical techniques are very good at ruling out disease. For example, cardiac CT is a fantastic test for ruling out obstructive disease. What becomes more important as your disease progresses is actually knowing whether it's producing ischemia. If we look at all of our international guidelines for revascularization, we should only be revascularizing patients if they really have evidence of ischemia.

Dr Mandrola: Symptomatic ischemia, right?

Dr Greenwood: Yes, symptomatic ischemia, or asymptomatic ischemia in some of the guidelines, but we should be very focused on ischemia. If you really want to look at ischemia detection, that is when you see the benefits of a functional test over an anatomical test.

Dr Mandrola: Thanks for being with us.

Dr Greenwood: You're welcome. It was a pleasure.

Dr Mandrola: That's it from the ESC Meeting. This is John Mandrola, from on Medscape.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: