RE: How useful are troponin and cardiac pro-BNP in diagnosing heart disease?

Mark Patteson — Thu, 16 Apr 2020 17:08:15 GMT

Sonia

This is a common question. Sadly the answer isn't a short one, it's a bit like saying how useful are radiographs. But here's my stab at it. Opinion differs. Of course, if they differ a lot I think that it's wrong, but see what discussion this prompts.

With these assays, as with any test, the first question is what question are you asking when you run the tests. You have said how useful they are – which covers a range. I will take to mean when should you do them

But a key part of this is knowing the answer is to understand why we do screening tests.

What is an ideal test?

One that identifies every animal with disease. This means a test with no false negatives – i.e. a highly sensitive test
One that doesn't identify any animals as having the disease when they do not. A test with no false positives – i.e. a very specific test
A test that is easily available and requires little user skill.
A test that is cheap to perform.

So the great joy of biomarkers are that they often fulfill 1, 3 and 4. But usually not 2. (OK it woudl depend what value you used)

In an ideal world, we would have lots of statistics on which to base our advice. Sadly, there is still a real dearth when it comes to biomarkers. If we had enough data we would be able to say how sensitive and how specific a test is in every population, age, breed etc. We need to know that, because the sensitivity and specificity vary depending on the population you are looking at.

Its also important to recognise that some tests are useful if you are looking at a population – so, for example, to look at the effects of an intervention in a group with a certain condition. It may help to distinguish two groups. But I am afraid that doesn’t necessarily mean it’s helpful in an individual patient.

How does this translate into practice? The answer is of course – it depends.

So, if you are asking how useful running a troponin would be to look for evidence of heart disease in a 1 year old chihuahua the answer is - it isn’t. If you are saying is it a clue to prognosis in a cat with known cardiomyopathy that’s a very different question.

And of course these two biomarkers are different. NT proBNP is released by stressed myocardium – usually this is stretch but it can be pressure, so it will go up in most acquired heart disease that we see and some congenital disease. But it doesn’t tell us what sort of disease is present. The degree of change can vary, but often it can increase by several times, so a small change say just may not matter. And there can be quite a lot of variation in value even in the same animal day to day (or lab to lab!). However, the degree of elevation is a guide to severity.

Troponin is released by damaged myocardium. In humans, this often results from an infarct due to coronary artery disease where the increase can be factors of 100s of times normal. We can see it go up with cardiomyopathy, but often the increase is small.

If people ask me when I use these assays, I think I can give some useful insight as to what I do now and what I hope we might be able to do one day:

Cats –a bedside proBNP snap test is available, which is set with quite a low value so it’s a sensitive test and is therefore useful for screening. So if you see a cat with a heart murmur in the clinic you will know that there is about a 50% chance that the cat actually has heart disease – it depends on the age, breed etc, the higher the risk group the more likely it is to mean that heart disease is present. So auscultation is a screening test, and if you add a proBNP that is another screening test that will add to the information. If it is negative it’s likely that the cat doesn’t have heart disease. If it’s positive it's now considerably more likely that heart disease is present and for me that means that the cat needs an echo. If the cat in your consult had a gallop sound, then its 90% likely that there is heart disease so adding a proBNP doesn’t really add much – better to save the cost and go straight to echo. And personally, I wouldn’t use troponin as a screening test in cats.

Canine DCM – for me I would recommend using proBNP as a way to screen at-risk breeds at an age that is likely to be appropriate, - say 3-5 years. The high-risk breeds that we are most concerned about are Dobermans, IWH, Great Danes, Boxers, Leonburgers, Newfoundlands, Dogue de Bordeauxs, and any other giant breed dog. It beats me why someone spends money to vaccinate an 8 year old Doberman that is at a very low risk of dying from an infectious disease, but ignores screening for heart disease because the tests is too expensive, even though the patient has a 40% chance of dying from DCM. Which most likely is occult – so test. Test, test, test.

Our second tier of concern is those breeds that can get DCM but the prevalence is much less. These would include Labradors, Weimeraners, Setters and Pointers, Spaniels, St Bernards and any other large breed dog. I think that screening is still valid for these, maybe for age 6 onwards or if there is a family history. Because the prevalence is much lower than in Dobermans, it’s much more likely that a sensitive test will give false positives. Your clients need to understand that. It doesn’t mean it’s a waste of time – if you had a 4% risk of having cardiomyopathy yourself I would think you would like to know – and echo is a much more expensive test in most instances. The advantage of echo is that it’s not just sensitive, it’s also the most specific test we have.

Troponin may be raised in DCM, but its not specific and it’s not as sensitive as proBNP so I don’t use it routinely.

Mitral valve disease in dogs – for me the best screening test is a stethoscope. To tell how bad the disease is we don’t as yet have a way to delineate stage B1, B2, C etc based on a biomarker, but we do have the echo parameters that are derived from the EPIC study that dictate when we start treatment (onset of B2). And we recognise congestive heart failure clinically and start diuretics etc at that point. As yet a biomarker isn’t a substitute. Yes, it’s interesting, but it hasn’t answered any questions, it’s added to the cost and I knew the dog had the disease because I heard the typical murmur. ProBNP values are very labile, so for example a Labrador with mild B1 MVD may have a proBNP of thousands – but the high value doesn’t mean it has DCM or even that the MVD needs treating. You can’t base treatment on proBNP (yet! This may come…..)

This will change and there is a lot more that we could discuss on this subject, but I suspect that is more than enough!

And a word of warning – these tests are only any good if they are performed well. ProBNP is very labile, a problem that has been addressed in part with the newer assays, but if it’s not done right you can get the wrong answer – which in my book is worse than no answer. Tropinin I is more stable – but again every test needs to be interpreted in the light of the other information you have. If a test results doesn’t make sense – it may be wrong – check it.

by all means ask more

Mark