Introduction

Since the first publications on cardiac troponin (cTn) in the 1970s, around 20.000 scientific reports on this powerful biomarker of myocardial injury have been published. Over time, cTn assays improved their analytical performance and became the designated biomarker for ruling out or ruling in acute myocardial infarction (MI). However, with increasing analytical quality the complexity of the clinical diagnostic use of especially the high-sensitivity cardiac troponin (hs-cTn) assays increased as well.

Laboratory (AACC, IFCC) and clinical expert groups (ACC, AHA, ESC) regularly update guidelines for the use of cTn assays [1-5]. Based on these guidelines this paper will discuss:

1. Classification of cTn assays
2. Interpretation of cTn results
3. Challenges associated with hs-cTn testing
4. Awareness and education of clinical staff

1. Classification of cTn assays

The current definition of cTn assays is based on analytical properties and not on clinical data nor on the type of cardiac troponin (e.g. troponin I or T). Currently, two classes of cTn assays are considered adequate [2-4].

1. Clinically usable: Contemporary assays which show an acceptable imprecision at their 99th percentile cut-off (CV >10% and ≤ 20%). Because of their low risk for false classification of patients with suspected MI, these assays are classified as “clinically usable” [3].
2. Guideline acceptable: Assays characterized by superior imprecision at their 99th percentile cut-off (CV ≤ 10%). Assays in this class which can measure troponin concentrations in over 50% of a healthy control population (e.g. measurable results > LoD in both men and women) qualify as hs-cTn assays.

2. Interpretation of cTn results

Finding a significant rise or fall of cTn levels remains key for the diagnosis of AMI. However, even a rise or fall of cTn values in serial samples drawn at admission and, depending on the assay used, one, three or six hours later, will not suffice for the diagnosis of MI, but merely indicate an acute myocardial injury which can be caused by acute coronary syndrome (ACS) as well as by non-ACS related pathologies [1,4].

From the Fourth Universal Definition of MI [1]:

• Biomarker criteria for Myocardial Injury
  Detection of an elevated cTn value above the 99th percentile URL is defined as myocardial injury. The injury is considered acute if there is a rise and/or fall of cTn values.
• 
  Clinical Criteria for Myocardial Infarction
  The clinical definition of MI denotes the presence of acute myocardial injury detected by abnormal cardiac biomarkers in the setting of evidence of acute myocardial ischemia.

Consequently, algorithms have been developed to demonstrate clinical or functional symptoms of acute myocardial ischemia, in combination with a significant change in cTn levels. While ischemia is mostly indicated by irregularities in ECG-tracings, the latter is substantiated with at least one of two serial results above the 99th percentile URL [1]. It is recommended to develop standardized serial sampling protocols for patients admitted with suspected ACS, appropriate for the cTn assay in use [5].

Finally, larger insults to the myocardium will result in higher circulating concentrations of cTn and a higher risk for adverse clinical outcome. Accordingly, cTn provides excellent prognostic power, as shown in ED patients with suspected ACS where cTn levels can identify patients with ongoing MI and/or increased risk for major adverse cardiovascular events within 30 days post-MI [6].

More detailed information is contained in table 1, which provides a comprehensive overview of the technical and diagnostic features of hs-cTn testing versus testing with a contemporary cTn assay.

3. Challenges associated with hs-cTn testing

Advantages

Because of higher analytical sensitivity and precision of hs-cTn assays, any increased cTn levels or change in cTn levels can be detected earlier and with higher confidence. These analytical advantages provide the core for improved diagnostic rule-out, rule-in algorithms for myocardial injury and support the clinical management of patients with suspected ACS. Especially accelerated diagnostic protocols (ADPs), based on faster diagnosis of acute cardiac injury using hs-cTn assays, are considered for supporting fast-tracking of patients admitted to ED with symptoms suggestive of acute coronary syndrome. Typical examples of ADPs are the 0h/1h rule-in and rule-out algorithm for diagnosing the presence or absence of acute coronary injury based on very high or very low results respectively, immediately after admission of a patient to ED. Also, the time interval between two serial samples providing optimal diagnostic accuracy for acute cardiac injury and MI is typically shorter than found with contemporary assays.

The increased ability to measure lower levels of cTn also resolved a gender difference and the need to establish and apply gender-specific 99th percentiles URL. Although the diagnostic advantages are clear, this finding also necessitated additional analytical requirements for optimal use of hs-cTn assays (see table 1).

Disadvantages

The downside of hs-cTn testing is the decreased specificity for MI compared to contemporary assays. Of patients admitted to ED without suspected ACS, about one of eight will present with an increased hs-cTn result, but only one of 200 will have MI [7].

Many factors and co-morbidities can cause cTn concentrations going up, including age, chronic kidney disease, hypoxia, chronic heart failure, anemia, and more [5]. Many patients admitted with suspected ACS are often burdened with mild levels of such co-morbidities, especially the elderly, and express borderline cTn levels even without the presence of AMI. Not surprisingly, hs-cTn testing shows moderate incremental diagnostic value in in elderly patients, in patients with co-morbidities, and in ED patients admitted several hours after onset of ACS symptoms, as demonstrated in several reports in which the diagnostic accuracy of a hs-cTn assay was not superior to a contemporary POC assay [8,9].

In addition, the typically high negative predictive values (NPV) of contemporary and POC cTn assays leave little room for improvement by hs-cTn in ruling out MI. Moreover, the higher specificity and higher positive predictive values (PPV) of contemporary assays appears to be associated with higher dismissal rates and reduction of cost when compared to a clinical algorithm based on hs-cTn [10]. Finally, patients with suspected ACS are mostly admitted several hours after onset of chest pain, are typically older, often have a history of CAD, and often carry additional morbidities and risk factors. In those cases, when a patient background does not warrant early discharge anyway as would be suggested by an early negative cTn result based on hs-cTn testing, then cTn testing based on a three or six hour sampling protocol with a contemporary cTn assay will provide similar diagnostic value.

4. Awareness and education of clinical staff

Many pre-analytical variables and analytical confounders in the testing of cTn with especially hs-cTn assays have been described, including inter-laboratory variability and inter-assay variability. Moreover, several interfering substances have been identified to affect especially hs-cTn assay results [7, 11].

Also, the side-by-side use of hs-cTn and contemporary assays is not recommended [2-4]. POC testing of cTn with its obvious advantage of short turnaround times and high NPV for MI, in combination with laboratory hs-cTn testing is feasible when the analytical differences between the systems is appreciated. Any kinetic interpretation of serial results obtained from different cTn-assays is NOT recommended [1,3,5].

Clinicians responsible for the clinical management of patients with suspected ACS require profound knowledge of the assay and interpretation of cTn results as described earlier. Especially product information pertaining to the assay(s) in use, and some reports from expert groups [1-5] provide a solid basis to ensure qualified use of cTn in patients with suspected ACS. Further refinements for optimal use of especially hs-cTn assays should be anticipated.

In case a clinical institution is not able (yet) to ensure implementation of the analytical quality and organizational requirements associated with hs-cTn testing, then a transition to hs-cTn testing might be premature, potentially create confusion, and ultimately cause patient harm [5].

Area of use	Valid for all cTn assays		Source of data
Marker for myocardial injury	cTn is a sensitive and specific marker for myocardial injury. However, the etiology of myocardial injury is often unclear and can be both chronic and acute in character, making cTn a sensitive but rather unspecific biomarker for AMI. Typical non-cardiac causes for increased cTn include aortic dissection, PE, shock, HF, and CRI.		[1,3,6]
Prognostic value of increased cTn levels	The level of cTn correlates with the risk of AMI, as well as the risk for future MACE and cardiovascular death.		[6]
Side-by-side use of various cTn assays	Current cTn assays are not harmonized. Consequently, dynamic interpretation of serial results from different cTn assay is not recommended.		[1-5]
Validation of 99th percentiles and total imprecision	Enroll individuals eligible for 99th percentile validation to extensive pre-selection to reduce risk of co-morbidities affecting the outcome.		[3,4,12]
	Contemporary cTn assays	hs-cTn assays
Determination of 99th percentile(s)	One assay-specific 99th percentile provided by manufacturer	Two gender-specific 99th percentiles provided by manufacturer per assay Eventually, other population specific 99th percentiles might be appropriate, e.g. age and ethnicity	[2]
Daily QC-testing	2 QC levels required: Close to 99th percentile of the assay At the upper level of the analytical range	3 QC levels required: Between LoD and the lowest 99th percentile Higher but close (within 20%) to the highest 99th percentile At the upper level of the analytical range of the assay	[5]
Annual validation of key analytical limits	99th percentile and total imprecision at the 99th percentile	LoD, gender specific 99th percentiles and total imprecision at these 99th percentiles	[5,11]
Reporting	Results should be reported in µg/L with 2 significant figures (e.g. 0.023 µg/L), while QC values should be reported with 3 significant figures (e.g. 0.0234 µg/L)	Results should be reported in ng/L in whole numbers, e.g. 23 ng/L, while QC values should be reported with 1 decimal point (e.g. 8.6 ng/L)	[5]
Rule-out of myocardial injury (NPV)	Consistently low levels, or levels expressing minimal change over three to six hours, allow rule-out of AMI with NPV of about 95%	A cTn level	[1, 13]
Rule-in of myocardial injury (PPV)	Elevated cTn levels, demonstrating a significant rise and/or fall in cTn levels over a period of three to six hours, with PPV of about 80%	A high cTn level at admission, or elevated cTn levels, demonstrating a 20% relative change in cTn levels over a period of one to three hours, with PPV >80%	[1, 14]

NPV: Negative predictive value; PPV: Positive predictive value; AMI: Acute myocardial infarction; PE: Pulmonary embolism; HF: Heart failure; CRI: Chronic renal insufficiency; MACE: Major adverse cardiovascular event.
Table 1. Contemporary cTn testing versus hs-cTn testing

Summary

• Hs-cTn testing supports safer (higher NPV) and earlier (higher sensitivity) rule-out of MI. However, lower numbers of MI-negative patients will benefit from this faster algorithm (lower PPV).
• Hs-cTn testing supports earlier (higher sensitivity and precision) rule-in of MI with a lower risk of missing-out on patients having MI (higher NPV). However, higher numbers of MI-negative patients will be referred to extensive follow-up diagnostics (lower specificity and PPV), especially the elderly patients, and other patients with increased baseline cTn levels.
• Because an early discharge of patients with a clinical history of, or increased risk for ACS is hardly feasible, such patients will hardly benefit from faster cTn testing.
• From a socio-economic stand, hs-cTn assays can be expected to improve healthcare for patients admitted with suspected ACS, but presumably at higher cost.
• Finally, using hs-cTn assays is associated with higher complexity and a need for more quality features and training. Moreover, future data might indicate the need for additional analytical requirements to ensure optimal diagnostic use of hs-cTn assays.