The utility of high sensitivity troponins in clinical practice

Abstract

Cardiac troponins (cTn) are the gold standard biomarker of myocardial injury. The evidence for the use of cTn in clinical practice is well validated and used globally in the management of patients with suspected acute coronary syndromes (ACS). As a rule out test, the evidence for the use of cTn is unrivalled. When first utilised in clinical practice a common issue with cTn assays was the need for clinicians to wait 10-12 hours post suspected infarction before clinical decisions based on the cTn result. This period of waiting has been the driver for the improvement of the assays. To help improve efficiency within different health care systems for suspected ACS patients, high sensitivity cardiac troponin (hs-cTn) assays have been developed. These assays are now able to be detected 1 hour post infarction, thus allowing clinicians to make decisions on management at a much quicker pace than under the early cTn assays. The improvements in sensitivity thus allow clinicians to safely rule out ACS and discharge patients. However, as a ‘rule in’ test there are flaws in how clinicians interpret the hs-cTn levels. We are now aware that there are many different types of MI, however, in only type 1 myocardial infarction (T1MI) are there proven treatments, such as antiplatelet therapy and percutaneous coronary intervention (PCI), which improve prognosis. Despite this, a common flaw is to treat all patients with a raised cTn as a T1MI.
Despite the descriptions and prevalence of the different types of MI throughout the literature, in clinical practice most patients diagnosed as MI are treated and labelled as T1MI. Current data suggests 20-50% of MI patients are in fact type 2 MI (T2MI). The Fourth Universal Definition of MI recommends the use of the 99th percentile as the correct cut off to diagnose MI. The 99th percentile is defined by manufacturer’s data derived as part of the internal validation for the assay: subsequently this level is quoted and usually used as a clinical “upper limit of normal”. There are many variables that can affect an individual’s troponin level. This has a significant effect on the definition of the ULN for any assay. This is particularly important as each individual manufacturer will have a different inclusion and exclusion criteria when defining the reference population used to quantify the 99th centile. For example, it has been shown that the younger the reference population is, and the stricter the criteria that are used to define cardiac health are, the lower the 99th centile will be. This raises important questions. Firstly, is it appropriate to use hs-cTn as a binary marker to ‘rule in/rule out’ MI? Secondly, how should abnormal hs-cTn levels be defined; is it appropriate to use the 99th percentile from a young healthy population and apply the marker of abnormality for this population to the older heterogeneous population that presents to hospitals throughout the world?
The objectives of the studies presented in this thesis are as follows. Firstly, the prevalence of patients presenting with a tachyarrhythmia and associated hs-cTn rise will be assessed. The management of these patients will be assessed. The short and mid-term outcome in these patients will also be described. Secondly, the
99th percentile of high sensitivity cardiac troponin I concentration for an entire hospital population will be defined.
In chapter 3, I undertook retrospective analysis of 704 consecutive emergency admissions to UHS FT with either a diagnosis of MI or tachyarrhythmia. The clinical management of these patients was analysed. Furthermore, the tracked mortality for these patients was analysed. The study found the mortality rate of patients with a tachyarrhythmia and raised hs-cTn level is similar to that of T1MI. Furthermore, only one patient in the study population (0.14%) was diagnosed with T2MI, highlighting that T2MI is rarely diagnosed in clinical practice.
In chapter 4, I set out to define the 99th percentile for the hospital population. I undertook an observational study of hs-cTnI levels in 20,000 consecutive patients who utilised the hospital services. This was an all comers study that had never been undertaken before. I found that the 99th percentile for the population studied was 296 ng/L, more than 7 times the ULN quoted by the manufacturer. The study also showed 1 in 20 of all patients included had a raised hs-cTnI level.
The results from the studies described in this thesis highlight that there are many factors which can raise an individual’s cTn level. The evidence for the hs-cTn assays as a rule out test is robust, however, as a rule in tests questions remain. The work presented here demonstrates the use of hs-cTn ULN as a binary ‘rule in/rule out’ as a flawed concept. This has the potential for patients to be managed inappropriately and could lead to issues with patient safety. The work presented here is the stimulus for
further work in this field to establish more optimal hs-cTn cut off levels. Finally, the work presented here shows that hs-cTn can be a marker of risk.

More information

Identifiers

ORCID for Nicholas Curzen: orcid.org/0000-0001-9651-7829

Catalogue record

Export record