I'd like to thank the meeting organizers for inviting me to speak. Like everyone else, I'm sorry we can't be meeting in person. Today, I'll be speaking about the pros and pitfalls of implementing high sensitivity troponin assays. These assays are being used commonly throughout the world, and are being increasingly used in the United States, and will soon be in widespread use in the next year. I have no relevant disclosures. Throughout this talk, I'll be discussing and reviewing the various different troponin assay platforms, and highlighting recommendations for hospitals anticipating transitioning to high sensitivity assays. High sensitivity assays have a number of advantages. They have superior analytical characteristics that allow them to detect very low level of troponin values. At these very low levels, sensitivity is high as well as accuracy. Using these in the emergency department can help expedite early rule-in and rule-out of ED patients with potential myocardial infarction. However, there are a number of caveats to consider when using these assays. Reporting units are different, as are the assay cutoff values. As a result, one cannot translate one rule-out protocol to another. They are assay specific. Because of the increased sensitivity of these assays, more patients will be detected who have troponin elevations. However, it's important to consider that these elevations are not synonymous with myocardial infarction. Sensitivity, despite these high sensitivity assays, sensitivity is still lower in early presenters and may be negative in patients who come in very early after the onset of infarction. Even though we can detect very minor amounts of myocardial damage, unstable angina still exists. One of the issues that we have with all troponin assays is detection of troponin values in patients who have underlying myocardial disease, specifically patients who have chronic kidney disease. Therefore, rule-out protocols may not directly apply to these patients. Finally, education and preparation are essential for successfully transitioning to high sensitivity assays. First, reporting units are different. Current assays are reported in nanograms per milliliter. However, the high sensitivity assays are much more sensitive, resulting in values that may be difficult to interpret. For example, using the Abbott high sensitivity assay, the lower limit of detectability is 0.001 nanograms per milliliter, and the 99th percentile is 0.026 nanograms per milliliter. As a result, recommendations to reduce confusion or to report results in nanograms per liter. This would translate into the lower limit of detectability for the Abbott assay of one and the 99th percentile of 26. This will require clinicians to reset what they typically consider a large myocardial infarction. This is something we've already seen when we're receiving transfers from outside hospitals that are using these high sensitivity assays, in which our fellows are often confused at these what may seem to be excessively high values. Values will vary for all the cutoff values, including the lower limit detective value, 99th percentiles, which may have gender cutoffs, as well as the significant delta change value for rule-out protocols. To make things even more confusing, values may differ by country. For example, the Roche assay has a different 99th percentile in the United States and in Europe. Therefore, it's critical that you work with your lab to make sure that everyone is using the correct values and you're using one of the rule-out algorithms that has been validated. As one might expect that given the difference in assays, results may vary even when using the same sample. For example, the authors here applied different MI exclusion protocols specific to three different high sensitivity assays and found a significant level of discordance. Of note, most of the discordance was based on the patients with undetectable levels who then fell into the intermediate group of values ranging from undetectable to 99th percentile, those with fairly low values. This may relate to differences in accuracies at the very low end of measurement, as well as some differences in the ability to discern very tiny differences in results. Because these assays are more sensitive, more patients will be found who have increased troponin values. This includes all type of elevation, whether we're talking about type 1 MI as seen on the left side of these bar graphs, type 2 MIs, or patients with myocardial injury. Of the subgroup of patients who have troponin elevations as seen on the right, the proportion of patients who have type 1 MIs will actually be decreased. Concomitantly, the proportion of type 2 MI in myocardial injury will make up a greater proportion of those patients who are found to have troponin elevations. Given the increased frequency that we'll see troponin elevations, it is critical that everyone recognize that an elevated troponin is not synonymous with myocardial infarction. Outlined below here is the HIGH-US study, which did what many rule-out protocols do. It separates patients into three different groups. A low-risk rule-out group of patients who are eligible for immediate discharge. In the middle, an observation or continued evaluation group. Finally, on the right, a group called rule-in. However, it needs to be recognized that many of these patients who fall in this ruling group do not actually have MI. In this study, it was 69 percent, consistent with other studies which have ranged between 65-75 percent. Because the term rule-in may lead to clinical bias, an alternative definition used by the UT Southwestern group is to call this group patients who have abnormal troponin, given that after subsequent evaluation, many of these patients will have a diagnosis other than MI. I think it's important to reiterate that the detection of troponin even in the absence of MI is not a benign process. It identifies a high-risk group with a mortality that exceeds and often is double that for type 1 MI and is equal to the type 2 MI as illustrated in these three different studies here showing mortality at one and two years after troponin elevation is detected. It's likely that much of this mortality is driven by both cardiac and non-cardiac diagnoses, as these patients have frequent underlying comorbidities. Such as heart failure, renal failure, and other non-cardiac diagnoses that contribute to mortality. Another caveat is that although we typically require a significant change in troponin values to identify patients with MI, not all non-STEMIs will have this on short-term sampling. In this cohort of approximately 2,000 non-STEMI patients, up to 25 percent of patients failed to have a significant delta change in either relative or absolute troponin value when assessed over a six-hour period. Potential explanations are that the patient presented late, or as a result of the patient having intermittent ischemia without subsequent troponin release. Misclassification of these patients as non-STEMI who actually have chronic myocardial injury is another possibility. Although when the authors compared those who did and did not have a significant increase in delta troponin, there was no significant difference in clinical characteristics, including the subgroup of patients who had definitive cardiac testing. This indicates that longer testing with additional sampling outside of a six-hour window may be necessary to provide an accurate diagnosis in these patients. Despite the improved sensitivity in these assays, patients who present early after infarction onset still can have negative values. For example, on the study on the left of 911 non-STEMI patients, sensitivity was only 95 percent in the patients who presented within one hour, increasing to 98.5 percent in those who presented at one to two hours. Although these sensitivities seem high, it's important to recognize that this 94.6 percent sensitivity translates into about a five and a half percent MI missed rate. This also is true for patients who present with STEMI. For example, on the slide on the right of 925 patients who presented with STEMI, overall, the initial sample was completely negative, less than five nanograms per liter in 2.2 percent of all patients and less than 99th percentile in 14 percent. When they looked at those who presented within two hours, six percent had completely normal troponin values, with a quarter of patients having values less than the 24th, the 99th percentile. This lower sensitivity related to sample timing is important when one is interpreting many of the published studies. As most of these were clinical trials, the timing of the first sample is after study enrollment, usually one to two hours after the initial evaluation because of these delays. Earlier assessment is likely to occur in clinical practice. This does have implications as some centers have moved towards sampling at the time of initial triage to try to expedite evaluation, which can occur prior to clinical evaluation. And some have even proposed having EMS sample troponins to assist in the early triage process as well. Although this reduced sensitivity is often focused on the initial sample, the same limitations will apply for early presenters, particularly those who present with only one to two hours and those who are using a zero to one hour rule-out protocol. Although some have hypothesized that as troponin sensitivity increases, unstable angina would disappear. Unstable angina has decreased, but it's not gone. Depending on the study, the proportion of patients who still receive a diagnosis of unstable angina can be as high as one in six patients. It has long been recognized that patients who have renal dysfunction, as defined in most studies, as a GFR less than 60, are more likely to have elevated troponin. These elevations are likely related to chronic myocardial damage and not related necessarily to excretion through the kidneys, as troponins are typically cleared through the lymphatic system. This increased frequency of troponin elevation is likely to be associated with chronic myocardial damage. This increased frequency of troponin elevation remains true for high sensitivity assays, as noted on the graph on the left, although this was limited to smaller numbers of patients. It does appear that these patients who have CKD who do have very low troponin values at the time of presentation would still be typically considered for early discharge as they have a similar low event rate. However, as seen on the graph on the right, the proportion are eligible for early discharge is considerably lower than we look at all comers and has often been less than 20% in studies that have looked at this subgroup. Therefore, protocols that should take into account that these patients are more likely to have higher values, those that exceed the 90th percentile, and will likely make them less eligible for early discharge, potentially resulting in overall longer length of stays. Finally, collaboration, preparation, and education are essential for a successful transition to high sensitivity troponins and will require collaboration among multiple different groups, not just emergency medicine and cardiology, but will require input from the lab and hospital medicine, as well as other stakeholders who are involved in the management and care of these patients. Some important recommendations on how to make the successful transition to high sensitivity assays are included in this article by Jim Genuzzi and colleagues, which provides a wealth of information, including department level checklists that can be used. In addition, updated guidelines for the evaluation of chest pain are expected to be released soon. As a supplement to these guidelines, there will be an accompanying expert consensus decision pathways, which will provide some additional recommendations on how to implement these new guidelines and assays. In conclusion, high sensitivity troponins offer substantial advantages for the early evaluation of patients with potential ACS. Variations in assay characteristics, rule out protocols, and patient presentations need to be taken into account when assessing the individual emergency department patient. You need to know your assay and use a validated ACS troponin algorithm. Clinical assessment still remains important for interpreting these high sensitivity troponin values. I think even 13 years later, as was well stated by Dr. Alpert in this editorial, we need to keep in mind that lab tests don't make diagnosis. Doctors do. Thank you.

high sensitivity troponin assays

pros and pitfalls

myocardial infarction

education and preparation

clinical assessment