How many ways can a troponin assay come up with a false positive or a false negative? In of AP Cardiology, Andrew Perry, MD, tallies it up with Robert Christenson, PhD, of the University of Maryland School of Medicine and the University of Maryland Medical Center in Baltimore.
A transcript of the podcast follows.
Perry: Hey, everyone. Andrew here. I hope you're enjoying the show. The episode today is inspired by a case that I encountered on the wards. We had a patient who came to the emergency room multiple times to be evaluated for falls or fractures. When she would come, they would end up checking troponin and it would be elevated. In this setting, I learned about how there can be false-positive troponins and interference in the troponin assays.
I reached out to Dr. Robert Christenson. He's a professor at the University of Maryland School of Medicine. Suffice to say that he really is an expert when it comes to using cardiac troponins in the diagnosis of acute myocardial infarction. He has chaired numerous committees, wrote many guidelines upon the use of cardiac troponin, involvement with the ACC, AHA, national laboratories, and hosts a Core lab for many clinical trials. I hope you enjoy the episode. If you are enjoying the show in general, please leave a review on iTunes or your podcast app or better yet, tell your friends and share it on social media. With that, we'll get started.
...
Perry: This is AP Cardiology and this is your host, Andrew Perry.
Christenson: Right, so my name is Rob Christenson. I'm a Professor of Pathology at the University of Maryland School of Medicine and also Medical Director of Core Laboratories and Point of Care at the University of Maryland Medical Center.
Perry: Great! Let me start by presenting a case to introduce our topic for today. I saw a woman in her 50s. She had hepatitis C and she had presented with generalized weakness and falls a few months prior to my meeting her. Her EKG at that time had some non-specific T-wave changes in the lateral leads. Her troponin was elevated to 32, so she was immediately sent to the cath lab at that time and had essentially a normal coronary angiogram except for some mild calcification in the proximal left anterior descending artery. Her echocardiogram was also normal, well, maybe some mild wall hypokinesis of the lateral wall, so she was discharged with a presumptive diagnosis of myocarditis and sent home on all these cardiac meds: beta blockers, ACE inhibitors, aspirin, statin.
She returns a few weeks later where I meet her where she presents after a leg fracture and her troponin is elevated to 10 at that time. During both of these presentations, just note that she denied any chest pain, and we checked a rheumatoid factor which was elevated to 1,100 and the CK-MB was normal. We did some send out testing to Mayo and they used a separate troponin assay, and the troponin there was totally normal. The first question for you is about how common it is for there to be interference with these troponin assays.
Christenson: Interferences with the cardiac troponin assays, particularly the newer-generation assays, the high-sensitivity version of cardiac troponin, is relatively rare. I guess your patient, the first time that she was in the hospital, what was her second troponin? You mentioned it. There was one troponin collected that was elevated I think you said. I just wondered what the second value was.
Perry: Sure. She went from about 32 down to about 28.
Christenson: Okay. The second time you said 10, so that was also above the 99th percentile at your institution?
Perry: Yeah, our 99th percentile goes to 0.03.
Christenson: That makes that 10 be like 300-fold higher than that 99th percentile? That is out in la-la land as is the one on repeat testing when she came in a couple weeks later. That kind of helps say the magnitude of just how elevated she is, which is very substantial. It would be 10,000 by the new nanogram per liter units. This would be I think relatively rare, although not unheard of. I think every institution has seen this sort of elevation. The repeat testing for both presentations to the ED, the change between the first and the second value -- I'm assuming the second value was probably some three to six hours after the first value -- did not show that dynamic rise that you see with or a fall in the case that they're presenting very late. From the value of the rheumatoid factor, I would, at first, go to heterophile antibodies and certainly the information of you sending the assay out and having it be completely below the 99th percentile for the patient is also valuable information about looking for an interference like heterophile antibodies, like may have been the case here with this individual.
Perry: Could you explain about these heterophile antibodies and how they interact with the troponin assays, how those work on a fundamental level?
Christenson: Sure. Assays for cardiac troponin are all immunoassays. We rely on our tools of developing antibodies for these tests. Antibodies are great. They're very sensitive. They're very specific. They bind very strongly and very avidly to cardiac troponin, so they're really a wonderful tool. However, they do have frailties. One of which is that there are two antibodies, generally, that are used in these formats. One is called the capture antibody, which is against one amino acid sequence for cardiac troponin. These were cardiac troponin I assays, so it would be against likely a cardiac troponin I amino acid sequence. There's a second antibody, which has a detection strategy bound to it. Oftentimes, it's just chemiluminescence or another very bright signal producer.
What can happen is that there's a couple of ways that these can either cause false positives or false negatives. In this case, it sounds like it's a false positive with the elevated value 300-fold higher than the 99th percentile at your institution. What happens here is that the antibody that is native to the individual, that is in the patient's sample, binds to both the capture antibody, a portion of the capture antibody, and also to the detector antibody, the one that's linked to the signal-producing entity.
What this does is during the assay's wash step where you wash away all of the unbound detection, this crosslinking by the patient's own antibody doesn't allow the signal to be washed away because, like I say, it's linked between the capture and the signal antibody. That would mean that when you go to measure the signal, it acts as if it's bound to cardiac troponin. That is there's a signal which is proportional to concentration of the analyte, in this case, cardiac troponin I, that shows up in your instrument and gives us a falsely, very high reading.
Perry: I think you suspected the heterophile antibodies because those have been associated with rheumatoid factor before. In fact, I think I saw case reports of rheumatoid factor itself being implicated as the interfering antibody that's doing this extra binding.
Christenson: Right, and that's what I would suspect in this case, particularly with the evidence that you have from the other lab ... which shows 300-fold above at your lab ... and then something completely within the normal limits at the other laboratory.
Perry: What are the other areas for interference with these assays or other areas for error in those troponin assays?
Christenson: There are a few things and it really depends on which assay you're looking at or which one we're examining. One that's very clear and very common is hemolysis. Hemolysis, contents of red blood cells when released into the solution can interfere with these tests. Depending on the format of the assay, they can either be false positive or false negative interferences. We talked about heterophile antibodies. There are also some autoantibodies that are against the troika, the TIC complex from cardiac troponin, which can also cause interferences to these assays. Other assays can be susceptible to so-called "microclots." If either you draw a blood sample -- let's say lithium heparin, for example -- and the sample is not mixed properly, so a preanalytical error where the sample is not mixed in accordance with the manufacturers' recommendations, you can get these microclots as the sample sort of slowly clots. That can cause these species, these microclots, which interfere with some assays, and they uniformly cause an increase in signal.
We've talked about hemolysis. We've talked about some autoantibodies, the heterophile in your case. Then with cardiac troponin T assays, in particular, it's a bit controversial, but there's thought that with skeletal muscle injury, that there can be a cross-reactivity with regenerating skeletal muscle tissue that can cause a false positive with cardiac troponin T assays. Again, there's capture of the signal antibody, and the capture antibody can react and show a false positive with that skeletal muscle disease with cardiac troponin. Again, that's controversial, but there is a building body of evidence for that.
Probably the most hip sort of interference is . It turns out that many individuals take biotin as a supplement. It's believed to be a beauty aid and it improves nails and so on, so you can buy this in health food stores or right across the counter at a pharmacy. Biotin can easily be a falsely positive interference in competitive immunoassays or it can be a negative interferent, give falsely negative values in other immunoassays such as sandwich immunoassays, which are commonly troponin. With biotin and with these other interferences as well, the interference isn't particularly limited to cardiac troponin. There can be interferences to other immunoassays, as well, that are susceptible to any of these interferences.
Perry: Gotcha. I think I also read about elevated alkaline phosphate levels because the detector antibody has an alkaline phosphatase enzyme. Is that another way that these can be interfered with?
Christenson: It's certainly been reported. I believe that most of the manufacturers, and of course, the manufacturers are working to mitigate or eliminate all of these interferences with various strategies, but I have certainly seen that as an interference in the past, but I believe it's mostly mitigated, although that is something that one has to consider with a falsely elevated or falsely decreased value for a patient.
Perry: Could you also comment about you mentioned that there are ways that the manufacturers try to help mitigate these? What are some of those ways in which the manufacturers for these assays try to decrease the potential for error on those?
Christenson: There's a lot of innovative ways that our colleague scientists in industry have thought about this. One way would be to, perhaps, incubate. Let's take biotin, for example, again, which is one that is under a lot of scrutiny now, where if you could have a pretreatment that would eliminate biotin. Of course, one way to do that would be to put streptavidin in there. One of the strongest bonds in nature is the biotin-avidin binding, so if you were to sort of pretreat in some way with avidin, you could pull out the native biotin that was in the patient's serum and then use that treated sample in the assay after you've taken measures to remove all of the biotin that might be the interference.
Perry: Gotcha.
Christenson: Similar ways that in development of antibody strategies to eliminate the heterophile antibodies or autoantibodies that might be present as well ... For the microclots, just making sure that the assays are well mixed, that the preanalytical piece is intact so that the microclots don't form to begin with, but the other is to spin it down or wait for a longer time for it to clot and then remeasure after a very hard spin just to validate that there was or was not microclots present.
Perry: I see. What are ways that a lab or even a clinician can help investigate whether the troponin assay is being interfered with? What are tips or tricks that you can do quickly without any specialized testing?
Christenson: A very valuable tool for that is the telephone. So knowing assays that your colleagues within a reasonable distance or an area are available is a valuable tool. Doing just what you had articulated that you all had done when you saw this very high sample in this patient who was not suffering acute myocardial infarction. That's one way. There are other tools that you can buy. Scantibodies is one which can help clear and that works reasonably well. Also to make it very clear, to make sure that the word is out, that patients who have high doses of biotin to make sure that in draw stations, etc., that patients are notified if they are taking biotin supplements to let their physician know, and then the half-life of biotin is such that it can be cleared pretty quickly. There are some patients such as MS patients, for example, where biotin is used . It seems to help with their disease management. Those patients would have very high levels and may have to wait several days or some time before the biotin would be cleared for them. These are just some of the strategies that are used to help mitigate.
Perry: I appreciate those bits about biotin. That was something I did not know or had come across when I was doing my preparation here. I think that's all that I have. I appreciate your time in visiting with me.
is a cardiology fellow at the University of Washington Medical Center in Seattle.