OUP user menu

Troponin-I, myoglobin, and mass concentration of creatine kinase-MB in acute myocardial infarction

A. Chiu, W.-K. Chan, S.-H. Cheng, C.-K. Leung, C.-H. Choi
DOI: http://dx.doi.org/10.1093/qjmed/92.12.711 711-718 First published online: 1 December 1999


Myoglobin, creatine kinase-MB (CKMB) mass concentration and troponin-I are newer biochemical markers for the diagnosis of acute myocardial infarction (AMI). We conducted a prospective study to formulate a model for the collective interpretation of these three markers in the diagnosis of AMI. Eighty-seven patients with AMI had serial serum samples taken to establish the time-frame sensitivity of individual markers. None of the markers had a good sensitivity within the first 4 h of infarction. Myoglobin and CKMB (mass) had sensitivities of 92.3% and 96.2%, respectively, at 4–8 h post infarct. CKMB (mass) and troponin-I had sensitivities >92% at 8–24 h. Troponin-I maintained sensitivity >93% until 72 h. A guideline was formulated based on the results. Our data suggest that troponin-I, myoglobin and CKMB (mass) yield satisfactory diagnostic sensitivity when used with reference to specific time frames. The combined use of these markers can provide valuable information for clinicians in managing AMI patients.


The basis of the conventional diagnosis of acute myocardial infarction (AMI) originated from the consensus of the World Health Organization Agreement.1 This includes a clinical account of chest pain, electrocardiographic representation and serial changes in cardiac enzymes. This statement was further elaborated in the American Heart Association (AHA) Scientific Statement (1996), which commented on the use of creatine kinase MB (CKMB) as the preferred cardiac marker for AMI diagnosis.2 CKMB is not an ideal marker as such because it lacks tissue specificity, has poor stoichiometric correlation with the extent of myocardial damage, and is only detectable in the circulation in a relatively short time-frame.3,,4 With the advancement of thrombolytic therapy5,,6 and primary percutaneous transluminal angioplasty,7,,8 there is a considerable difference in outcome if these treatments can be administered early in the course of the illness. There is therefore a urgent need for a rapid, sensitive and specific cardiac marker that can assist clinicians to make an early diagnosis of myocardial infarction. In recent years, much research has been done to look for such a marker. The latest markers available in market include troponin-I, CKMB (mass) and myoglobin.

Troponin-I is a contractile protein found almost solely in the myocardium. It is highly tissue-specific, and persists in the circulation for days. However, it is only detectable at 4 hours or more post injury.9 Myoglobin is a cytosolic haem protein that can be found in both cardiac and skeletal muscle. It is released early in the course of myocardial injury and can be detected within the circulation in <30 min. It is, however, excreted rapidly by renal clearance. The lack of tissue specificity and relatively short time-frame of detection therefore limit its significance as a useful marker.10 CKMB (mass) differs from conventional CKMB measurement in that the latter is an indirect measurement of the enzyme activity, while CKMB (mass) is a direct measurement of the enzyme concentration. It is detectable in the circulation from 4 to 8 h post injury.11 Each of these markers appears to have their own strengths and weaknesses. It would therefore be reasonable to combine these different tests, to maximize their diagnostic potential. To date, there are no well-established guidelines for the collective interpretation of these new cardiac markers when used together. The aim of our study was to evaluate the time-frame sensitivity of troponin-I, myoglobin, and CKMB mass in the diagnosis of AMI. Using the results, we attempted to formulate a guideline for interpretating the combined cardiac markers in everyday clinical application.



All patients with an initial diagnosis of suspected AMI from 1/5/98 to 30/11/98 were recruited. They were assessed by a member of the cardiac team or senior doctor on call, either at the accident and emergency department, or in a ward. The conventional WHO and AHA diagnostic criteria were followed. Chest pain was regarded as positive if the condition was sustained, and the clinical account convinced the clinician that the discomfort was genuine. Transmural infarct was defined as an ECG showing progression from no Q waves to definite Q waves (≤0.04 s, amplitude >1/4 R wave) or development of QS complexes. Non-Q infarct was defined as ECG showing ST segment depression or increase of at least 0.1 mV, and a T-wave inversion in at least two leads of the same vascular territory of at least 24 h duration.12 The cardiac marker criterion was regarded as positive if there were serial changes of CPK which demonstrated a rise-to-fall pattern with peak >200% of baseline and conventional CKMB >5% of total CPK. On admission, patients satisfying two out of the three criteria were considered as having myocardial infarction. The final diagnosis of myocardial infarction required all three criteria to be satisfied. Patients with chest pain and non-Q ECG pathology but not serial cardiac enzyme changes were diagnosed as having unstable angina.

Patients diagnosed as having, or being at risk of, myocardial infarction were admitted to our coronary care unit whenever possible. Each patient then had a blood test for conventional cardiac enzymes, CKMB and our cardiac panel of myoglobin, CKMB mass and troponin-I. The time from the onset of chest pain until the blood sample was recorded. Serial blood samples were taken at 6 h, 12 h, 24 h, 2 days and 3 days post admission whenever feasible. All other patient baseline characteristics, including age and sex, time of presentation, nature of infarction, thrombolytic administration and mortality information were retrieved from our hospital AMI registry.

Laboratory diagnostics

All blood samples were centrifuged at 3000 rpm for 5 min to separate the serum. Overnight specimens were stored at 4 °C until the next working day. All three new cardiac markers were measured on a multifunction immunoassay analyzer system (Opus, Dade Behring Diagnostics). The troponin-I assay was a sandwich enzyme-linked immunosorbent assay (ELISA) using two separate polyclonal antibodies directed against different epitopes unique to the cardiac form of troponin-I.13,,14 The upper reference limit was set at <2.0ng/ml, and the lower limit of detection was 0.5 ng/ml. CKMB mass was measured by sandwich-type ELISA immunoassay using anti-CKMB and anti-CKMM monoclonal antibodies.15 The upper reference limit was set at <5.0 ng/ml, and the lower limit of detection was 0.6 ng/ml. Myoglobin was measured by sandwich ELISA immunoassay using anti-myoglobin monoclonal antibody and anti-myoglobin polyclonal antibody. The upper reference limit was set at <90 ng/ml, and the lower limit of detection was 1 ng/ml. CKMB activity was measured by the enzyme inhibition method (Boehringer Mannheim), in which a specific antibody is used to inhibit the CK-M moiety without affecting the CK-B moiety. The activity contributed by the CK-BB moiety is assumed to be negligible. The calculated difference in activity from the total CK activity is then attributed to that of CK-MB.16 The upper reference limit of our laboratory was set at 5% of the total CK activity, and the lower limit of detection was 1 IU/l. Total creatine kinase concentration was measured by conventional enzymatic analysis (Cobas Mira Plus, Roche Diagnostic Systems). The upper reference limit was set at 154 IU/l, and the lower limit of detection was 5 IU/l.


There were 104 patients admitted during this period with diagnosis of AMI. Seventeen patients were excluded from the study because of uncertain sample time collection (8); incomplete data entry (3); haemolysed sample (1) and other violations of protocol (5). The excluded population did not differ from the study population in terms of gender, age distribution and type of myocardial infarction. In the remaining 87 patients, there were 59 men (67.8%) and 28 women (32.2%). One patient sustained two myocardial infarctions in two separate admissions, and the entry was counted twice. One patient sustained a reinfarction within the same admission, and the reinfarction was excluded for logistical reasons. The mean age was 66.6 years with a standard deviation of 10.5 years, and ranged from 39 to 88 years old. There were 75 cases of transmural infarction (86.2%) and 12 cases of non-Q myocardial infarction (13.8%). There were 66 patients (75.9%) who were given thrombolytics. The in-patient mortality rate was 10.3%. The time from onset of chest pain to the first sample being taken varied from 0.33 h to 42 h, and the mean time was 4.89 h, standard deviation 6.12 h. There were 11 patients (12.6%) with one sample taken; six patients (6.9%) with two samples taken; 38 patients (9.2%) with three samples; nine patients (10.3%) with four; 22 patients (25.3%) with five; and 31 patients (35.6%) with six. The number of samples taken depended mainly on the duration of stay within the coronary care unit. Patients were usually kept for 24 h or more, depending on their clinical status and hence the majority had at least three or more samples taken. Where less than three samples were taken, this was due to early transfer out from coronary care unit, discharge against medical advice, or death before the protocol was completed. The enzyme levels vs. time lapse distribution were plotted for the three markers (Figure 1). The time lapse was calculated as the time from the onset of chest pain (not admission) until the sample was taken. The time lapses were then grouped into different time frames which were arbitrarily set as 0 to 4 h, 4 to 8 h, 8 to 12 h, 12 to 24 h, 24 to 48 h, 48 to 72 h and beyond 72 h. The sensitivity of each individual cardiac marker and the combined approach were then determined within these different time frames. Using the combined approach, a sample was considered positive if any one of the three enzymes exceeded the upper reference limit. The 95% CIs were determined using the non-parametric method of Fleiss. (Table 1).

View this table:
Table 1 

Sensitivity of myoglobin, CKMB (mass), troponin-I and the combined approach in specific time frames

Hours since infarct...0–44–88–1212–2424–4848–72>72
Patients (n)34264176766967
Myoglobin (%)55.892.385.475.043.420.314.0
95% CI38.1–72.473.4–98.770.1–93.963.5–83.932.3–55.211.0–32.06.7–25.0
CKMB mass (%)
95% CI27.6–61.978.4–99.885.6–99.990.0–99.584.7–97.658.7–81.013.2–34.8
Troponin-I (%)35.380.792.797.496.197.193.0
95% CI20.3–53.460.0–92.779.0–98.190.0–99.588.1–99.089.0–99.582.2–97.4
Combined (%)61.896.297.697.498.798.694.7
95% CI43.6–77.378.4–99.885.6–99.990.0–99.591.9–99.991.1–99.984.4–98.4
Figure 1.

Plots of a myoglobin, b CKMB (mass) and c troponin-I vs. time for 87 patients with acute myocardial infarction.

The sensitivities obtained from our study varied from 35.3% to 100.0% for different enzymes at different time intervals. With regard to the first 4 h post onset of chest pain, none of the individual markers had a reliable sensitivity. The early sensitivity of myoglobin during 0–4 h was 55.9%, that of CKMB (mass) was 44.1%, and that of troponin-I was 35.3%. In the next interval 4–8 h post onset of chest pain), myoglobin attained its highest sensitivity of 92.3%. The sensitivity of CKMB (mass) during this interval was 96.2% and that of troponin-I was 80.8%. The combined sensitivity obtained was 96.2%. From 8–20 h post onset of chest pain, both CKMB (mass) and troponin-I had sensitivities of >90%, and the sensitivities of both attained 100% at 16–20 h. On the other hand, the sensitivity of myoglobin gradually decreased with the course of time: beyond 24 h, its sensitivity fell to 43.4%; beyond 48 h its sensitivity was 20.3%; and beyond 72 h only 14.0%. The sensitivity of CKMB (mass) beyond 24 h was 93.4%; beyond 48 h it fell to 71.0%; and beyond 72 h it was 22.8%. The sensitivity of troponin-I beyond 24 h was 98.7%; beyond 48 h, 98.6%; and beyond 72 h, 94.7%. The combined sensitivity obtained during the first 4 h post onset of chest pain was 61.8%. In all other time intervals, the combined sensitivity was >90%, with the exception of the 20–24 h interval.


Our present data demonstrate that the combined use of myoglobin, CKMB (mass) and troponin-I had high sensitivity from 4 to 72 h after onset of chest pain. The peak sensitivity of myoglobin (92.3%) occurred at 4–8 h, that of CKMB (mass) (97.6%) occurred at 8–12 h, and that of troponin-I (97.4%) was at 12–24 h (Figure 1). The clustering of points above the reference threshold symbolize those test results that were positive, which could be translated into a good diagnostic coverage for detection of AMI within this period. According to our study, however, none of the cardiac markers (myoglobin 55.8%, CKMB (mass) 44.1%, troponin-I 35.3%) was able to provide a satisfactory sensitivity from 0–4 h post onset of chest pain. This is in line with other investigators' findings when trying to establish the early sensitivities of cardiac markers.17 A negative combined result within this time frame does not exclude the diagnosis, because of the considerable false negative frequency. The diagnosis of early AMI remains clinical and electrocardiographical. At 4–8 h, the diagnostic emphasis should be based on myoglobin and CKMB (mass) as both tests have sensitivity of over 90%. In the interval 8–48 h, the emphasis should shift to CKMB (mass) and troponin-I. Beyond 48 h, only troponin-I retains reasonable sensitivity for the diagnosis of AMI.

Earlier studies of troponin-I, CKMB (mass) and myoglobin emphasized the validity of individual enzymes in the diagnosis of myocardial infarction. The sensitivity and specificity of these enzymes has been well demonstrated.9,15,18–,20 Presented with an increasing number of available cardiac markers, more recent studies have examined the combined sensitivity of these markers in improving the diagnostic accuracy for AMI.21–,23 Although temporal differences in sensitivity of different markers has been recognized, few papers have aimed to integrate the element of time into the interpretation of different markers.24–,26 The recent study by Chung et al.26 evaluated the feasibility of diagnosing AMI based on the combined result of CKMB mass, CKMB/total CK index, myoglobin and troponin-I. The proposed panel, however, required serial decisions based on enzymes being tested individually instead of collectively. In this study, we have attempted to use the cardiac panel as an integrated test rather than using an algorithmic approach. Based on the data obtained, we have constructed a guideline for interpretation of the cardiac panel result at different time frames. The design of the guideline was not meant to include complicated instructions and computations; instead it was meant to provide a diagnostic adjunct for clinicians while retaining their own autonomy in decision-making. Various international studies have also been evaluated21–,27 and taken into account, with modifications made according to logistic needs. The guideline (Table 2) will be used as the recommendation in our unit for the investigation of suspected AMI in the future.

View this table:
Table 2 

Guideline for cardiac panel interpretation

Time from onset of chest painInterpretation
0–4 hA negative result does not exclude a diagnosis of AMI. Positive result (myoglobin less specific, CKMB (mass) and troponin-I more specific) suggestive of AMI but needs to be assessed with ECG findings and clinical assessment. If in doubt, patient should have test repeated.
4–8 hBoth myoglobin and CKMB (mass) are sensitive markers for AMI within this time frame. Negative tropinin I does not exclude diagnosis of AMI as the level may not be detectable in the circulation yet.
8–24 hBoth CKMB (mass) and troponin-I are sensitive markers for diagnosis of AMI within this time frame. Negative myoglobin does not exclude AMI as the marker may have already been cleared from the circulation.

The latest cardiac profile, including CPK, myoglobin, CKMB (mass) and troponin-I, cost around HK$120.00 (US$15.40). There is a considerable monetary difference when compared with the conventional profile of CPK and CKMB activity which cost only HK$34.00 (US$4.40). The difference, however, should not be seen as that of the apparent cost alone. The superiority of the new cardiac profile lies not only in its improved sensitivity, but also in its ability to give more information to the clinician at times where decision is needed. The implementation of a better diagnostic tool would therefore imply better stratification of patients with suspected AMI, hence a better use of resources and avoidance of unnecessary coronary care unit admissions. The expense that could be saved probably outweighs the additional financial burden of the new cardiac profile. A `cardiac marker' committee composed of members from the cardiology division and the biochemistry department has been established in our hospital, and would regularly review the cost-effectiveness of the new cardiac profile use.

At present, the greatest value of rapid diagnosis of AMI lies in the benefit of prompt thrombolysis.28,,29 For patients with evident ST segment elevation, the additional information of raised cardiac marker should not interfere with the decision of thrombolytic administration. For patients with non-diagnostic ECGs, the indeterminate quality can be due to pre-existing ECG changes such as left bundle branch block or left ventricular aneurysm; or because of difficult recognition such as high lateral or posterior infarctions; or because of previous infarctions. In such circumstances, the cardiac panel result would be of great value in deciding whether the patient is suffering from AMI. In patients presenting with unstable angina or non-Q infarct, collectively known as acute coronary syndrome, raised cardiac markers, especially troponin-I, are associated with a worse prognosis and outcome.30,,31 Although current evidence does not support the use of thrombolytic agents in these situations,32 the cardiac panel would provide valuable information for risk stratification of these patients.

As only patients with confirmed myocardial infarction were recruited, the greatest limitation of our study was that the specificity of the cardiac markers cannot be discerned. We have chosen not to include patients with suspected myocardial infarction or clinical chest pain because it was not our primary intention to determine the specificity as such. Numerous studies have shown that troponin-I has high tissue specificity, and its level has been already evaluated in patients with acute skeletal muscle injury, chronic muscular dystrophy, chronic renal failure and arduous athletes. None of these entities were associated with false positive elevation of troponin-I.33–,35 Conventional CKMB assay, in contrast, was frequently raised in these situations because the enzyme is expressed in small amounts in skeletal tissues as well.36–,38 Moreover, in circumstances where the total CK is normal, the relevance of raised CKMB fraction is yet to be determined.39 Both these factors limit the utility of CKMB alone in the diagnosis of AMI. In our study, troponin-I persisted in the circulation beyond 72 h post myocardial damage. Other studies have shown that troponin-I persists in the circulation for up to 127 h.14,40,,41 CKMB, however, is only detectable within a time frame of about 30 h.42,,43 The contrast in release kinetics is due to differences in intracellular compartmentation between the two markers. CKMB is a cytosolic protein that appears early in the circulation following myocardial injury. Troponin-I, in contrast, is only found in small amounts within the cytosolic compartment, but is abundant within myofibrils. The release of troponin-I from myofibrils depends on the proteolytic disintegration of the contractile apparatus, which is a time-consuming process. Hence the marker appears late after onset of injury, persists for a prolonged period.44–,46 For patients with delayed presentation of AMI, troponin-I would therefore be superior to CKMB in making retrospective diagnosis. For patients that suffer reinfarction shortly after their initial insult, however, the persistence of troponin-I within the circulation can be a disadvantage. In such circumstances, it would be more difficult to observe the change in marker level if the patient sustained reinfarction before troponin-I returned to baseline. The nadir could be overlooked if the frequency of sample collection was not able to capture the transition. In our study, one patient sustained a reinfarction within the same admission, and his results were excluded for this reason. Other workers that investigated troponin-I47,,48 have expressed similar opinions. Addition of CKMB (mass) in conjunction with troponin-I in the standard cardiac panel would be beneficial, because CKMB (mass) is cleared from the circulation much earlier than troponin-I. Any reinfarction that occurred >24 h from initial injury should have observable changes in CKMB (mass). CKMB (mass) also amends the lack of early sensitivity with troponin-I. According to our results, CKMB (mass) attained sensitivity of >90% earlier than troponin-I. CKMB (mass) and troponin-I together have complementary benefits, and the combined cardiac panel appears to be a promising replacement for conventional assay in the future diagnosis of AMI.

Looking ahead, newer-generation analyzers with modified methods of sample preparation, should reduce machine runtime to less than 15 min. The advantage of such analyzers is that there is built-in centrifuge, in which whole blood instead of serum sample can be used, and they can be easily used by clinical rather than laboratory personnel. Results can therefore be generated without time spent on serum separation, computer interfacing or document transfer between departments. The operational cost of such analyzers would be considerably higher, but their strategic value lies in their ability to provide immediate results for decision-making. Such an analyzer, in addition to a well-equipped unit with serial ECG monitoring, could maximize the diagnostic efficacy for AMI. The practicability of such units has been recently shown.49

In conclusion, our present data suggested that troponin-I, myoglobin and CKMB (mass) yield satisfactory diagnostic sensitivity when used in reference to specific time frame. The combined use of these markers can provide valuable information for clinicians in the diagnosis of AMI within the context of a logistic guideline.


View Abstract