Activity Transcript

Allan Jaffe, MD: Hello. I'm Dr Allan Jaffe, Professor of Cardiology and a Professor of Laboratory Medicine and Pathology, at the Mayo Clinic in Rochester, Minnesota. Welcome to this program, "Recent Advances in Diagnostic Cardiology: Harnessing the Power of High-sensitivity Troponin Assays. Joining me today is Dr Nick Mills, who's the Chair of Cardiology at the University of Edinburgh in Scotland. Welcome Nick. 

Nicholas Mills, MD, PhD: Thank you, Allan. 

Dr Jaffe: Our program today is to improve the knowledge about high-sensitivity cardiac troponin (hs-cTn) assays that are used in detecting myocardial infarction (MI) and to increase our confidence in the integration of the use of these assays into clinical practice. By way of background, the diagnosis of MI requires myocardial injury that is acute, meaning a rise and/or fall of cardiac troponin values, and clinical criteria, meaning that the setting is one where there is evidence of acute myocardial ischemia. 

This is depicted in this conceptual slide in the universal definition where you can see that there is an up-slope and then a peak and a down-slope, and one needs to be able to anticipate and find the rising and falling values with at least one value above the 99th percentile. 

The hs-cTn assays allow us to anticipate that these criteria will either be met or not met, such that we can use earlier metrics that predict whether or not this is going to occur and rule-in, or going to occur and rule out MI.  

I would be remiss if I didn't mention the fact that although we're going to talk about MI today, there's tremendous utility in the use of hs-cTn assays for a variety of different circumstances. This is simply one example of the anticipation of mortality and the prediction of mortality in patients who have COVID disease showing that elevations of cTn I in this instance improves the ability to prognosticate. But this is just one example of many in primary and secondary prevention, stable coronary disease, heart failure, renal disease, and there are a large number of these that eventually we need to discuss such that they are understood. 

Nick, let's talk about some of the key characteristics to keep in mind about hs-cTn assays as they're presently used. 

This first slide looks at a study done by Fred Apple, where they took samples from people who were putative normals, corrected for a variety of biomarker comorbidities - hemoglobin A1C, EGFRs, and NT-proBNP - and then determined the percentage of individuals who had measurable values with each of these assays as shown in the bars. Then the percentage of individuals who had values above the limit of detection of the assays shown by the dots. The bottom line for this is that all the assays are different. They are different in terms of their ability to detect normals, but they're also very different in terms of what the key metric values are that one has to use.

Nick, what about sex-specific cutoffs?

Dr Mills: Thanks Allan. I think the other important observation from these studies is that consistently, for all assays, the 99th percentile differs in men and women. In particular the 99th percentile is substantially lower in women than men. Therefore, in order to apply these tests consistently and identify the same proportion of men with myocardial injury, the manufacturers and the universal definition recommend that we use sex-specific 99th percentiles. 

Dr Jaffe: Some key metrics to keep in mind that we've already mentioned: myocardial injury occurs when one has a value above the 99th percentile; if it is rising and/or falling, it is acute; if it has not changed, it's mostly chronic. But the 99th percentile is assay- and sample-dependent, and we have already seen the data from the universal sample bank, which allows one to make comparisons between men and women. 

It is axiomatic, and we should say that all assays have analytic problems. They are specific to the given assay. Troponin T, for example, can detect some individuals who have chronic skeletal muscle disease, but there are macro kinases, heterophilic antibodies, and a variety of other analytic products that are usually specific for a given assay. The good news is that they're not very common, and if you find a value that does not comport what you think is appropriate for a given patient, if you check with your lab, the lab ought to be able to troubleshoot the problem and give you better information. Nick, let's review how you use these high-sensitivity assays in clinical practice. 

Dr Mills: I think the first thing to say is that the additional characteristics of being more precise at that key decision threshold, the 99th percentile, means if you are using this test as a simple rule-in or rule-out test, then their performance is better. You're less likely to get analytical false positives at that threshold. 

But the real value is that we can now quantify cTn concentrations in almost everyone right down to the limited detection of the assay at concentrations way below the threshold that we now use to diagnose MI. As a consequence of that, a number of prospective studies have created a large evidence base, which has informed the use of other thresholds as are mentioned to predict the final diagnosis. That consists of either using the limit of detection itself or an optimized rule-out threshold just above it to make earlier decisions in the emergency department. Rule out patients with a single test where previously we might've admitted them for serial testing. There's also some work trying to inform how to make rapid rule-in decisions at higher thresholds. 

Ultimately, this evidence base has sort of coalesced around the use of validated early diagnostic pathways like the one recommended by the European Society of Cardiology. I call these multi-threshold pathways because they incorporate both rule-out, rule-in, and the diagnostic threshold along with a series of delta values with retesting between an hour and 2 hours following presentation. Ultimately these pathways do a good job of identifying low-risk patients and high-risk patients who need to come into hospital, but they also identify a group in the middle with intermediate values in the so-called observed zone. We need more research, I think, to understand how best to further assess this group. Now in Scotland, we also use a variation of this pathway, slightly more paired back version, where we use separate rule-in and rule-out thresholds so that ultimately, we make a decision in 3 quarters of patients to triage them to an outpatient setting, and we admit a quarter for further assessment within 3 hours of presentation to the emergency department. 

We've built up a lot of experience, really, with the high-sensitivity assays on how to use them safely and effectively in just over 80,000 consecutive patients studied in Scotland, and the use of these early rule-out pathways we can be confident in. They reduce length of stay, for sure. They also increase the proportion of patients discharged directly from the emergency department, moving care from the medical unit to the front door. Overall we now discharge three-quarters of all patients with suspected MI, and crucially, follow-up of these patients demonstrates that the compression of care, this accelerated pathway, it does not harm patients, and the serious adverse events at one year are similar before and after the adoption of these accelerated pathways. I think we can be confident about the adoption of these tests. That's our experience with I, I know that there is a good experience also with the hs-cTn assay. 

Dr Jaffe: Well, there is a little bit of attention there because high-sensitivity of T, if it substitutes for the prior fourth-generation assay, causes a large number of increases, and that has led people to be concerned. But in point of fact, the majority of the studies would suggest that despite the increased frequency of elevated values, that one doesn't end up with increasing rates of angiography, of stress testing, or of cost by and large, no matter which assay you use that is high-sensitivity, these accelerated protocols markedly reduce the time that it takes to rule-out and does not tend to impact tremendously on cost or on utilization of testing. 

Dr Mills: Allan, do you have some caveats or advice that you'd like to share just about the use of these assays in clinical practice? 

Dr Jaffe: Yes. I have a couple of concerns and one of them comes from a study done by your group. That has to do with the single sample rule-out when you have a very low value, these patients are clearly low risk. You've shown, I think correctly, that if you take patients who come in very early, that the predictive accuracy can go down from 99% to 95%, and that ends up with a metric that at least many emergency departments are uncomfortable with. I would also point out for this metric that if you take patients who are high risk or elderly, you get different results as well. So I think we have to be careful about that.  

The second is that depending upon which study you look at, there are a substantial number of individuals who come in late after the onset of their clinical syndrome. 

They may be on the down slope of the time concentration curve, which is much slower. The consequences that, in some studies like one from SWEDEHEART, as many as 26% of patients with MI may not have a delta that can be anticipated or measured over a very short period of time. We need to be careful about those patients. Finally, although I would argue troponin ought to go up every single time one has an acute ischemic episode, we have multiple good cases of unstable angina still existing. So, although I can't explain the biology, the phenomenon is clear. 

How about you Nick? Any practical advice on the implementation of these pathways? 

Dr Mills: Thanks, Allan. Yes, I do. I couldn't agree more with your caveats and, just to digress briefly, I had a patient in my cath lab table on Wednesday morning who had a diagnosis of unstable angina, and he's currently undergoing coronary bypass grafting. One of the observations that I've made, this is anecdotal since the introduction of these tests, is that patients with unstable angina are perhaps a little more likely to have intermediate, but not elevated troponin concentrations as this patient did. But I think we need to always keep in mind the importance of unstable angina. Moving on to some suggestions for adoption of these pathways, that the thing that I always emphasize is that clinical assessment for cases like the unstable angina and the electrocardiograms has to take priority over these pathways. If you have a patient with ST-segment-elevation MI, you measure high sensitivity cardiac troponin, as many as 1 in 4 will have a troponin concentration within normal reference range before that artery is opened and the myocardium is reperfused. This is particularly important if you've got someone who you're suspicious about a posterior infarct. Secondly, having gone through these pathways and ruled-out or identified that high-risk patient, it's really important to remember that this is not when you stop your clinical assessment. So, patients who are ruled-out still require clinical assessment and follow-up. Those in intermediate concentrations in the observed zone, perhaps even more so, because they have an event rate of around 1 in 20, at 1 year. They may not have had an acute infarct on this occasion, but they are certainly at higher risk in the future. I think that it goes that that message is doubly important for those who you admit with myocardial injury whom the final diagnosis is not an acute coronary syndrome. 

Here myocardial injury almost always reflects important underlying structural coronary disease, and the assessment does not stop once you've ruled out MI. These patients may benefit from further investigations and care. Then just a couple of practical points with the adoption either of a high-sensitivity test or a new accelerated pathway, really important to involve the multidisciplinary team, emergency medicine, cardiology, and laboratory medicine. Education -- make sure that the staff are comfortable with the purpose of these pathways. Lastly, before you consider adopting an early rule-out pathway, know your assay. Often these are threshold; these thresholds are assay dependent. And monitor its performance once you've implemented. 

Dr Jaffe: Thank you, Nick. We've covered a lot of information, but to summarize there's some key takeaways. Hs-cTn assays will facilitate more rapid and thus better and more cost-effective care if one understands how to use them properly. We started to do that today, and I hope that was helpful. 

Dr Mills: Agreed. I think once we've learned how to harness these tests to their full capacity in the emergency department to make better decisions, I am confident that we will start to use them much more widely to form care in our outpatient clinics. 

Dr Jaffe: Thank you, Nick. And thank the audience for joining us today. 

This is a verbatim transcript and has not been copyedited.