Professor Wolfgang Koenig from The German Heart Centre in Munich, Germany, speaks to Cardio Debate & Radcliffe Cardiology about markers of cardiovascular risk during the ESC2016 Congress held in Rome, Italy.


This is a timely and interesting question, and I believe you are referring to the enormous efforts that we have seen over the past 20 years in finding new biomarkers, that would guide us to improve diagnostics, to improve risk stratification, to show progression or regression, etc.

And to answer the question we are not at a dead end, but we have faced a number of problems that we have not yet solved adequately.

So imagine at the beginning, there was a whole panel of biomarkers – almost every week you would have a new biomarker in the literature – at some point in time we have realised that simply showing an association between a biomarker and outcome is not enough when it comes to clinical utility.

So for clinical utility you need a number of other requirements where you really have to discriminate between healthy and diseased, and that you can reclassify patients in terms of risk stratification. And that has taken a while to improve the methodology.

And also a number of these biomarkers which are of pathophysiological interest are not robust enough to be introduced in the clinical setting because of pre-analytics issues and a couple of others. So because of this, only a relative few biomarkers over the past 15 years have made it into the clinical routine, and I think the best example is troponin. Troponins are extremely useful for us cardiologists to diagnose acute coronary syndrome and also tell us about prognosis.

In addition to this markers of haemodynamic stress, for example BNP or N-Terminal Pro-BNP, are important to give us an idea of the functioning of the heart.

And finally, markers of renal function – we have discovered those that go beyond creatinine in terms of precision, etc. So a couple of things have happened, but out of these many, many biomarkers that have been in the pipeline, only a very few today have made it into clinical practice.

Are there any biomarkers on the horizon of interest?

Certainly. I just mentioned a couple of them that have been established. So troponin, although it being established, still a lot of people are working on new assays, on higher sensitivity assays, they try to improve risk algorithm, etc. So a lot of research is still ongoing.

In terms of novel biomarkers, there are a number around that are still under consideration, like for example GDF-15, which is a member of the transforming growth factor family, but it consistently could be shown to be associated with hard cardiovascular end points, namely cardiovascular death or total mortality.

And more recently a Swedish group from Uppsala has shown that it may add significantly to established risk scores in predicting bleeding in patients on oral anticoagulants, for example. So this is a very important question.

And beyond that, new technologies open new horizons – for example, lipidomics, metabolomics and proteomics.

And just to mention one example out of the lipidomics area – so called ceremides – specific lipid particles have recently show to predict cardiovascular disease reasonably well, and add to existing biomarkers. So these have been tested in several populations and are already being offered by a US company, so you can order this test.

Is imaging a better option for risk stratification?

This is also an important question, and along the same lines as what I mentioned for blood biomarkers – there has also been developments in terms of imaging technology, and we now have a number of imaging technologies yet similarly only few have made it as a routine.

So, for example for years we were looking at carotid arteries measuring the so-called intima-media thickness complex. But this has changed over time, more and more studies have come out and we have really seen this is not as reliable as we initially thought.

In terms of other imaging methods that are suitable for risk stratification in the primary prevention setting, more recently calcium scoring came up. And this seems to be fairly reliable with very little radiation, at least these days. So you can use it in patients at intermediate risk where you’re not sure what to do with those in terms of the classical risk factors.

In addition to these non-invasive imaging methods there are also invasive imaging methods, but these are not suitable for screening methods. We use them in specific situations to evaluate high-risk patients during interventions, for example.

What research should be conducted to help us understand its role better and guide medical therapy?

This is a fairly complex question, and I think it has at least two sides.

One side is new technologies to create new, clinically relevant biomarkers that are reasonably robust for the clinical situation. On the other hand, obviously we need to improve our bioinformatics skills, to really assess adequately and digest that information that we get from our ‘omic’ technologies.

For example, there are several platforms available where you can measure almost 200 metabolites, with different pathophysiological pathways that are all of importance in terms of atherosclerosis. But to adequately manage this data and digest them, and then come up with a useful, final panel that may be clinically applicable – this is the question.

So besides, new technologies to detect new markers, we also need to improve our statistical skills, our bioinformatic skills, to come up with a final clinically-relevant set.