Interview of Professor Paul Collinson - thoughts on past and future of cardiac disease diagnosis

Troponin and BNP have generated a paradigm shift in heart attack diagnosis

Q: First things first – have you already decided how you are going to spend the award money, $20,000?

A: It will allow me to fund some of our biochemists to go to the American Association of Clinical Chemistry meeting this year. I have given them the challenge, have an abstract accepted and you will go to Atlanta! It will also help fund some of the other more speculative research ideas that we have. So I will be spending it on my hobbies, but that will include some scuba diving!

Q: What made you choose cardiac troponin as the topic of your research?

A: My conversion to cardiac troponin came when I was working with Peter Stubbs. He became interested in the whole idea of rapid diagnosis and using novel markers. We did a study which was validating the measurement of cardiac troponin T as a diagnostic test. It was very straightforward and the test seemed to work, however, we had 4 sets of patient case notes we couldn’t work out whether these patients had had a heart attack or not. According to the conventional criteria they had not. According to the troponin results, they had. We were trying to work out how we were going to resolve this contradiction. Then, I noticed that each set of case notes had been stamped “deceased”. This made me realize that troponin could be a prognostic marker and it was finding people who had been missed as they did not fulfil the conventional criteria for diagnosis of a heart attack.

Q: You have many roles and responsibilities including managing, research and actual clinical work. How do you divide your time between all these roles?

A: I would say that I use about half of my time seeing patients, half of my time doing research and half of my time doing routine work… I used to do a lot of management work but within the last years I have fortunately been able to cut down my management roles and concentrate more on research and clinical work which I prefer to management.

Q: What do you enjoy most in your work?

A: I like treating patients and getting them better. That is really why I chose medicine in the first place. I also like the research I do, because my research is very applied. I’m afraid I’m not a theoretical researcher, I like to know that what I am going to do is really going to make a difference. I really enjoy the fact that technology is changing quickly and enabling me to do things now that I could not do even one year ago, let alone five years ago. In some respects I feel that it is a shame that I will not be practicing in 20 years’ time because I think that the potential will be there for really some phenomenal uses of technology.

So I like the interaction with patients and I love my research. I have always enjoyed what I have done. I feel exceptionally privileged to have had a job that pays me for my hobby. That’s why I have been quite prepared to work for long hours and also in the evenings and weekends, because it’s not work, it’s pleasure.

Q: You have published over 250 articles. Can you name one or two articles that you are especially proud of? 

A: I think that the two articles I’m most proud of are the work I did together with Peter Stubbs. The first was on the measurement of troponin on the patients presenting with defined acute myocardial infarction. It showed that troponin elevation when you were admitted puts you in an adverse prognostic group. This was the first publication on that subject. It was published in Circulation in 1996 and we were very pleased with ourselves.

The second one was when we were looking at the long-term follow-up of patients who had a diagnosis of unstable angina and were positive for troponin. It was published in BMJ in 1996. It was the foundation for the push we made to convert the UK from being backwards in measuring cardiac enzymes to measuring troponin as the routine standard test.


Outside work, Paul’s interests include scuba diving and photographing sharks. Photo: Tiger shark, photographed in November 2014, Cocos Island, Costa Rica. Photograph by Paul Collinson.

Q: How did troponin change heart attack (MI) diagnosis?

The advent of cardiac troponin was a complete shift in the way hearts attacks, myocardial infarction (MI) was diagnosed. For the first time we had a test that proved, when all other tests were negative, that the patient had had significant ischemic myocardial injury with a long-term adverse outcome. That was the first step. The second step came when series of studies were done looking at the patients with a diagnosis of unstable angina and looked at the impact of treatment. All the studies showed that if you treated people who had unstable angina, but who were troponin positive, with a variety of therapies including anti-thrombotic drugs and re-vascularization, their outcome was improved. Conversely, if you treated people who were troponin negative there was no difference in outcome whether they were treated or not. This meant was that we had a test that not only predicted a high risk of death but also that you could do something about it.

Q: You’ve introduced and developed troponin testing in the UK and greatly influenced to the fact that routine troponin tests were taken into use in UK. Can you tell a little about that?

This was how I first started working in the area of health economics. With troponin there was a lot of evidence, both from us and from others, showing that you could use troponin testing to expedite management. What we needed to do was to convince the hospital managers that a troponin test, which at that time was more than ten times more expensive than the diagnostic tests in use, significantly shortened the time patients with chest pain needed to stay in the hospital. With the old tests, everyone stayed for 2-3 days. By analyzing the cardiac troponin we were able to send people home after one day if their test result was negative. This shorter time spent in the hospital brought significant savings and easily paid off the cost of a single troponin test. 

Whenever anything is done in medicine it is essential to look at the whole process of care because it is the process of care that drives the cost, not the individual small subsections of it..

Q: You’ve done the same with BNP tests as well, and you’ve and been involved in developing and implementing the UK National Institute for Health Excellence (NICE) guidelines for cardiac diseases. Was it the same story as with troponin?

A: With BNP the story was not to replace an existing test but to introduce a test that did not exist. The routine was that if you went to your GP and you said you’re a bit short of breath, your GP sent you for an echocardiogram. Unfortunately, making a diagnosis of heart failures is quite difficult and about 50-60% of such patients have a completely normal echocardiogram. That is a waste. Our task was to educate that the level of BNP in blood could be used to distinguish whether an echocardiogram was needed. One echocardiogram costs about £100 and one BNP test is about £30. By doing one BNP you can save doing two echocardiograms which brings savings in heart failure management without compromising patient care.

First we did some audits with our local GPs and the feedback was very positive. To validate this new strategy we carefully looked at the data and did a big data pooling project. By analyzing the data we were able to pull out the decision levels to enable accurate rule-out. This work contributed to the NICE guidelines that came out a while ago. It has been nice to see that the European Society of Cardiologists embrace the use of BNP as a test for both acute and chronic heart failure diagnosis.


Photo: Grey Reef Shark, photographed in November 2011, Turks and Caicos

Q: Have you done the follow-up, do you see the cost reduction when using BNP testing?

A: Yes, people have done studies and showed that it works.

Q: The NICE guidelines are used in UK. Are similar guidelines used elsewhere?

A: NICE guidelines are acknowledged as being cost effective evidence based guidelines and are looked at around the world. But it also depends to some extent on the health system. If you’re dealing with a health system which accepts the fact that funding is central (and limited) and is episode based then they are applicable. But if you are a cardiologist somewhere where you’re paid on a fee for service basis, you will not want to do a BNP because it means that you will not be doing an echocardiogram. NICE guidelines are most effective in countries which are more sensible about their health care expenditure.

Q: When looking back, how has the diagnostics of cardiac diseases improved over the years? Has it been through major breakthrough discoveries or mainly small steps?

A: It is both. We had a complete shift that came with the routine ability to measure troponin and BNP. They were big changes in behavior. But we also have small changes that go with it. The history of troponin shows this very nicely. First we had imperfect tests for myocardial infarction and then along comes troponin, “the answer to your prayers” which said heart attack yes or no. Then, it became apparent that troponin was elevated in a range of other clinical conditions and not just a heart attack. In order to distinguish these other conditions from a heart attack we were back to serial measurements and rate of change. Now we are in a constant cycle to needing to educate people how to use troponin and BNP tests and to understand that these tests are just one part of the clinical diagnosis, not the whole truth.

Troponin has generated a paradigm shift in heart attack diagnosis. One part of the technology shift has been for rapid provision of immunoassays. A test that used to take a week because it was carried out by hand nowadays takes just 20 minutes. The bad side of this is that doctors order a lot of tests that they would not have ordered before and whose results are not relevant for the immediate treatment of the patient. The second part of the technology shift has been the development of the point of care (POC) testing. As we move to sensitive POC troponin assays my prophecy would be that the majority of troponin testing will move from out from the labs to next of the patient and generate immediate patient management changes.

Q: What would you say are the main pain points in the current cardiac disease diagnostics?

A: From a patient’s point of view it is the speed of getting the answer. We know what can be achieved in theory but this is quite often not achieved in practice. I think speed is not just the speed of the diagnostic technology but that a large part of it comes from the process of care. People often seem to forget that patient focused care is also about the process and how the patients flow through the system. It is in everyone’s interest to make it better: patients will like it, clinicians can be more effective and it will also save money so the managers will like it too.

One of the reasons I like working with the concept of cardiovascular diagnostics and chest pain is that they are relatively clean areas. It is relatively straightforward to put in to place pathways that will be efficient and patient focused.

For the laboratory or the physician it is about looking at the entire process of what is going on and seeing how the testing fits into, how it can be optimized. There is no point in providing a high sensitivity troponin test in 10 minutes if it is going to take 2 hours for someone to make a decision on the diagnosis. That is why I think information technology is potentially a great enabler. For example the troponin test results could come to your smart phone immediately they are available.

Q: Are new markers needed or is better understanding of older markers creating new opportunities for practicing clinicians?

A: The vision is that a patient comes into the hospital, some tests are made and they immediately give the answer to everything. I am not personally convinced that we will ever get to that point. The problem is that many of the tests that have been proposed are not disease specific but they are only revealing if you are ill or not. If you get a positive result from a test you should also know what to do with it. I think we can do much better with what we have already as has been demonstrated recently by Christian Müller’s group using very rapid changes in cardiac troponin for early diagnosis and the studies by Rick Body in Manchester.

There is a problem also with people not using existing knowledge. We have just performed a large pan European audit and found that only about half of the people we surveyed were using the recommended 99 percentile cut-off for troponin. So, there is still a lot of education needed. I would say that when 95% of all clinicians know and understand troponin, then we can worry about new markers. I would like to see people using the tests they already have properly. We have got to the point where we have a test that is relatively cheap, which we understand really well and which has a wide set of applications. Of course we should always be looking for new things. We also should be skeptical of what we have and look for suitable ways to improve. But, I would very much like us to sort out and use to best advantage what we have already.

Q: So what you’re saying is that the troponin and BNP tests are not always used correctly?

A: There seems to be a gap between what is perceived as the use of troponin and BNP as markers and what the actual use of them is in real life. The paradox is that it is the older and more experienced clinicians that tend to be better users of these markers because they also tend to think clinically. The problem is that the younger clinicians are just doing the tests and not looking at the patient. It feels as though in the cardiological community, people are drifting away from being clinical to just being number, investigation and intervention focused. So, it is a question of how do we persuade clinicians to re-establish their roots as clinicians and not as investigation technicians?

Q: Do you see any major changes in CHD diagnostics in the next 5-10 years?

A: I suspect imaging is going to become more central as we will be able to use lower radiation doses. Better direct imaging of the coronary artery trees will be useful. In my own field I am looking forward to having non-invasive imaging so that I can see somebody, look at their cardiovascular risks but also have a good feeling of whether or not they have early arterial damage. I have people who are referred to me in their 80’s who have high cholesterol but have no evidence of coronary disease and it is quite apparent that they will take their high cholesterol to the grave, instead of the high cholesterol taking them to the grave. I suspect imaging will never ultimately be as sensitive as biomarkers but will be a supplement to rather than a replacement for biomarkers. I also think that the technology to deliver at the bedside is going to progress. The blood volumes required and the analytical technology are going to get smaller and smaller.

There is no doubt that new detection technologies will come along, we may have biosensors that you can implant and will give you real-time measurement. The trend is for all technologies to get smaller, faster and cheaper but what is important is what is wrapped around it – the clinical understanding of the testing and the information technology that goes with it. The IT side of it is going to have two aspects: One is to utilize communication technology and the other will be to put it to a more holistic assessment of the patient. It is going to be about utilization and display, on conversion of data to information. That is what I have always been interested in: making data into information. Information enables you to make decisions. You have to process data and put the data into context.


Hammerhead shark, photographed in November 2014, Cocos Island, Costa Rica.



All photographs courtesy of Paul Collinson.

Do you need samples?

We are happy to make you an offer