The Journal of International Medical Research 2003; 31: 76 83 The Diagnostic Value of Troponin T and Myoglobin Levels in Acute Myocardial Infarction: a Study in Turkish Patients S VATANSEVER 1, V AKKAYA 1, O ERK 1, Ş ÖZTÜRK 1, MA KARAN 1, N SALMAYENLI 2, C TAŞÇIOĞLU 1 AND K GÜLER 1 1 Department of Internal Medicine, Division of Emergency Medicine, and 2 Department of Biochemistry, Istanbul Medical Faculty, Istanbul University, Istanbul, Turkey This study compares the diagnostic value of troponin T (TnT) and myoglobin with creatinine kinase (CK) for myocardial infarction (MI) in a tertiary care centre in a developing nation. The study group comprised 33 acute myocardial infarction patients and 27 healthy controls. Receiver operating characteristic curves for TnT, myoglobin and CK were drawn and areas under the curve calculated. At admission, myoglobin levels had greater diagnostic sensitivity than TnT or CK levels. After 2 h, myoglobin and TnT had equal sensitivity and specificity, whereas CK still had lower sensitivity than myoglobin and TnT. After 4 h there was no difference between the tests. It was concluded that myoglobin levels on admission and TnT at 2 h had the greatest diagnostic rate, whereas all the tests were similar after 4 h for MI. KEY WORDS: MYOGLOBIN; TROPONIN T; ACUTE MYOCARDIAL INFARCTION; RECEIVER OPERATING CHARACTERISTIC CURVES; TURKEY Introduction The diagnosis of myocardial infarction (MI) depends on evidence of myocyte necrosis provided by biochemical tests, clinical examination and electrocardiography (ECG). 1,2 The enzymes classically used as markers of myocyte necrosis are creatinine kinase (CK), aspartate aminotransferase and lactic dehydrogenase, which reach diagnostic levels long after the acute event and are not specific to myocardial damage. Hence, they are not suitable for early diagnosis. 3 Diagnosis or exclusion of MI as early as possible after admission to an emergency unit is essential for good prognosis. New and expensive biochemical tests have been developed and are used for this purpose. Troponin T (TnT), a regulatory protein of myocyte contractility specific to heart muscle was found to be a specific and sensitive marker of cardiac myocyte necrosis. 4 Myoglobin, another protein present in myocytes, has been found to peak early and hence be useful in the early diagnosis of MI, though with a lower specificity. 5 Nations with poor resources may have patient profiles different to those of developed countries; for example, the mean time of arrival at a medical facility may be longer and the clinical spectrum of the patients may show less variability. This may 76
reduce the need for expensive tests in the diagnosis of MI. The purpose of this study was to compare the diagnostic value of TnT and myoglobin with CK for MI in our local population in Turkey. Patients and methods PATIENTS AND CLINICAL EXAMINATION Patients with acute MI diagnosed by clinical presentation and ECG in the Emergency Unit of the University of Istanbul, Istanbul Faculty of Medicine, Turkey, were recruited to the study group. All had ST segment elevation greater than 1 mm in two or more contiguous leads on the ECG and diffuse retrosternal five dermatome chest pain. Time from onset of chest pain to admission and admission to diagnosis were noted. Venous blood samples were taken on admission and after 2 h, 4 h, 12 h and 24 h. The control group comprised members of the health profession, matched for age and gender. Blood samples were taken once after a 12 14-h fast. All patients were treated in accordance with the requirements of Good Clinical Practice. The Declaration of Helsinki s recommendations for guiding physicians in biomedical research involving human subjects were followed. LABORATORY ANALYSES Measurement of CK enzyme activity was done on a Cobas-Integra 700 automation system (F Hoffman-La Roche Diagnostics, Basel, Switzerland), using commercial kits (Roche Diagnostic Kit No. 2055317, Mannheim, Germany). The measurement interval, intraassay and inter-assay variability for CK were 0 1500 IU/l, 0.77% and 1.8%, respectively. The blood samples used for TnT and myoglobin measurement were centrifuged and the supernatant stored at 40 C. TnT and myoglobin were measured by electrochemiluminescence immune assay using a Roche Diagnostic Elecsys 2010 immune-assay device (F Hoffman-La Roche Diagnostics, Basel, Switzerland) and a commercial kit (Roche Diagnostic Kit No. 2178214 for myoglobin, No. 2017644 for TnT, Mannheim, Germany). The measurement interval, intra-assay and inter-assay variability were 0.01 25.00 ng/ml, 1.1% and 5.4% for TnT and 21 3000 ng/ml, 2.9% and 3.6% for myoglobin, respectively. STATISTICAL ANALYSIS The receiver operating characteristic (ROC) curves for CK, TnT and myoglobin were drawn and areas under the curve (AUC) and their SEs derived using statistical package SPSS for Windows 10.0 (SPSS Inc., Illinois, USA). ROC curves display the relationship between sensitivity and specificity for tests which have continuous outcomes. True-positive (sensitivity) rates are plotted against false-positive (1 specificity) rates, and the closer the curve is to the upper left corner, the more accurate the test. Diagnostic tests can be compared by visually analysing the ROC curves or, more accurately, by computing the AUC and using a modified Wilcoxon rank sum procedure as follows. The performance of different markers obtained from the same patient population was determined by comparing the area under the ROC using a method described by Hanley and McNeil. 6 The method consists of computing a ratio z using the equation: z = (A 1 A 2 )/ (SE 12 + SE 22 2* r * SE 1 *SE 2 ) where A 1 and A 2 are areas under the ROCs, SE 1 and SE 2 are the corresponding SEs and r is the estimated correlation coefficient between A 1 and A 2. The z values were estimated for various combinations of the indices, and the statistical significance (two-tailed P-value) determined from the standard z distribution tables. Subgroup analyses for onset of chest pain < 6 h or > 6 h before admission were undertaken using the equation: 77
z = (A 1 A 2 )/ (SE 12 + SE 22 ). The cut-off point of the test was taken as the point of minimum value of the function: (1 sensitivity) 2 + (1 specificity) 2. The results were considered statistically significant if P < 0.05 (z 1.96). Results The study group comprised 33 patients (29 men, four women, mean age 51 ± 11 years) and the control group 27 people (24 men, three women, mean age 51 ± 12 years). The location of the MI was acute anterior in 16 cases (48%), inferior and right ventricular in three (9%), inferior in nine (27%), inferoposterior in three (9%) and high lateral in two (6%) patients. Time from chest pain to diagnosis was 6.5 ± 4.6 h, and 19 patients (58%) arrived in the department within 6 h of the onset of chest pain. The changes in CK, TnT and myoglobin levels 0 h, 2 h, 4 h, 12 h and 24 h after admission are given in Figs 1 and 2. Myoglobin levels reached maximum 4 h after admission and had decreased abruptly by 12 h and 24 h after admission without reaching control levels. Maximum levels for CK and TnT were reached 12 h after admission and had started to decline by 24 h. Sensitivity and specificity of the tests at admission and 2 h later are given in Table 1. The most sensitive test at admission was myoglobin, but after 2 h TnT and myoglobin were equally sensitive. AUCs and SE for CK, TnT and myoglobin are reported in Table 2. On admission, myoglobin had greater diagnostic sensitivity than TnT (P = 0.034) or CK (P = 0.043; Fig. 3). The AUC curve for CK was greater than for TnT, but it did not reach Means of creatinine kinase and myoglobin with their 95% confidence interval 3000 2000 1000 Creatinine kinase Myoglobin 0 Controls Admission 2 h 4 h 12 h 24 h FIGURE 1: Changes in the mean levels of creatinine kinase enzyme activity and myoglobin, measured in patients admitted with acute myocardial infarction and controls. Blood was drawn from patients for these measurements at admission, and 2 h, 4 h, 12 h and 24 h after admission. Control blood was drawn after a 12 14-h fast. Creatinine levels peaked after 12 h, while myoglobin peaked after 4 h 78
6 Mean and 95% confidence interval of troponin T 5 4 3 2 1 0 Controls Admission 2 h 4 h 12 h 24 h FIGURE 2: Changes in troponin T levels measured in patients admitted with acute myocardial infarction and controls. Levels were measured in blood drawn from patients at admission and 2 h, 4 h, 12 h and 14 h after admission. Control blood was drawn after a 12 14-h fast. Troponin T levels peaked after 12 h statistical significance. Tests done 2 h after admission showed that TnT had the greatest diagnostic sensitivity, compared with CK (P = 0.003) and myoglobin. The difference between CK and myoglobin was still significant 2 h after admission (P = 0.04; Fig. 4). After 4 h, there was no difference in the diagnostic sensitivities between the tests, although the sensitivity of myoglobin began to decline. When patients were grouped according to duration of chest pain prior to admission (> 6 h or < 6 h), AUC for admission levels of TnT (Table 3; P = 0.33) and myoglobin were lower in patients admitted earlier, but did not reach statistical significance. Two hours after admission, there was no difference between the groups for TnT or myoglobin. Results obtained 4 h after admission showed no difference between the groups according to the duration of chest pain (Table 3). However, when the diagnostic accuracy of the tests (comparison of the AUC of the respective ROCs) was compared at TABLE 1: Sensitivity and specificity of creatinine kinase, troponin T and myoglobin at admission and 2 h after admission Admission 2 h after admission Sensitivity (%) Specificity (%) Sensitivity(%) Specificity (%) Creatinine kinase 63.6 89.9 78.8 89.9 Troponin T 75.8 89.9 97.0 89.9 Myoglobin 84.8 89.9 97.0 89.9 79
TABLE 2: Areas under the curve and their standard errors of creatinine kinase, troponin T and myoglobin according to time from admission Time from admission Creatinine kinase Troponin T Myoglobin 0 h 0.818 ± 0.054 0.767 ± 0.069 0.912 ± 0.038 2 h 0.911 ± 0.036 0.983 ± 0.013 0.978 ± 0.017 4 h 0.998 ± 0.003 1.000 ± 0.000 0.990 ± 0.009 12 h 0.998 ± 0.003 0.970 ± 0.030 0.974 ± 0.016 24 h 0.960 ± 0.029 1.000 ± 0.000 0.877 ± 0.047 admission according to duration of chest pain before admission, myoglobin seemed to be a more valuable test than either TnT (P = 0.085 and P = 0.09, respectively) or CK (P = 0.12 and P = 0.012). Discussion The purpose of this study was to examine the diagnostic accuracy of the laboratory tests according to time from patient admission. As 1.00 0.75 Sensitivity 0.50 0.25 Source of the curve Reference Myoglobin Troponin T Creatinine kinase 0.00 0.00 0.25 0.50 0.75 1.00 1 Specificity FIGURE 3: Receiver operating characteristic curves showing the relationship between specificity and sensitivity, drawn for creatinine kinase, troponin T and myoglobin at the time of admission. The closer the curve is to the upper left corner, the more accurate the test 80
1.00 0.75 Sensitivity 0.50 0.25 Source of the curve Reference Myoglobin Troponin T Creatinine kinase 0.00 0.00 0.25 0.50 0.75 1.00 1 Specificity FIGURE 4: Receiver operating characteristic curves showing the relationship between specificity and sensitivity, drawn for creatinine kinase, troponin T and myoglobin 2 h after admission. The closer the curve is to the upper left corner, the more accurate the test seen in Table 4, the cost of the tests in Turkey differs by an order of one, which compels their use in a rational way. The working hypothesis was that the time from the start of pain to patient admission would be longer in Turkey than elsewhere, obviating the advantage of the newer diagnostic tests. It was found, however, that on admission, the diagnostic rate for myoglobin was higher than either TnT or CK, and at 2 h after TABLE 3: Areas under the curve and their SEs for creatinine kinase, troponin T and myoglobin according to time from admission and the duration of chest pain on admission Duration of chest Time from admission pain Creatinine kinase Troponin T Myoglobin < 6 h > 6 h < 6 h > 6 h < 6 h > 6 h 0 h 0.800 ± 0.070 0.841 ± 0.063 0.712 ± 0.097 0.841 ± 0.091 0.877 ± 0.059 0.960 ± 0.026 2 h 0.893 ± 0.049 0.935 ± 0.042 0.983 ± 0.017 0.981 ± 0.016 0.967 ± 0.028 0.992 ± 0.010 4 h 1.000 ± 0.000 0.995 ± 0.007 1.000 ± 0.000 1.000 ± 0.000 0.994 ± 0.007 0.984 ± 0.017 12 h 0.996 ± 0.005 1.000 ± 0.000 1.000 ± 0.000 0.929 ± 0.069 0.994 ± 0.007 0.947 ± 0.031 24 h 0.984 ± 0.014 0.926 ± 0.064 1.000 ± 0.000 1.000 ± 0.000 0.918 ± 0.044 0.821 ± 0.081 81
TABLE 4: Kit costs for myocardial enzymes in Turkey Kit costs per Enzyme examination (US$) Creatinine kinase 0.32 Myoglobin 2.48 Troponin T 4.18 admission myoglobin had equivalent diagnostic sensitivity to TnT. The higher diagnostic rate of myoglobin at admission may be explained by its release kinetics after myocardial damage, 7 and the timing of enzyme rise was consistent with findings in other countries. 8 10 The ROC sensitivity and specificity curves seemed to be shifted to the left, i.e. to an earlier time from admission. This may be due to systemic underestimation of the time of onset of cardiac pain by the patients, or may be caused by comparing patients and controls from the two extremes sicker patients with MI versus normal controls. In our patient population, if the time estimations are correct, the favourable outcome of the newer tests may be artifactual due to early admission of patients to our centre, and may not be representative of the care received by Turkish MI patients generally. The only data on the patient population admitted to coronary care units with a diagnosis of MI in Turkey come from the Turkish Myocardial Infarction Registry Study (TÜMAR) (unpublished), undertaken between August 1998 and August 1999. The registry contains details of 3358 cases of acute MI presenting to 92 medical centres from 23 cities. In that study, 83% of patients had ST elevation on admission and 71% were admitted within 6 h of onset of chest pain (median 3.5 h). This compares with all our patients having ST elevation but fewer being admitted within 6 h of chest pain (median = 4.7 h), suggesting that our patient population is representative of MI patients admitted to coronary care units in Turkey. Comparing the data from our study with time from chest pain to admission figures from the ENACT 11 (65% within 12 h, 116 patients from Turkey) and PRAIS-UK 12 (median 3 h [IQR 1.5 8.2]) studies suggests that the time from chest pain to admission is similar in Turkey to elsewhere. Can the results from our study be extrapolated to cases admitted with equivocal ECG findings? Our study population represents 83% of the patients admitted to coronary care units in Turkey according to the TÜMAR data, but patients admitted with equivocal ECG findings represent the minority of coronary care patients in Turkey. The ENACT study also states that the final diagnosis is MI in 20% of patients with a provisional diagnosis of acute coronary syndrome and 10% of patients with a normal ECG. 11 We are confident that if MI is not apparent on admission, follow-up and myoglobin analysis for 4 h after admission would give the best rate of diagnosis. For patients with equivocal ECG findings, myocardial necrosis may develop during the follow-up period with concomitant increase in biochemical markers for up to 12 h after admission, which decreases the validity of this extrapolation. 13 To conclude, in our Turkish population at admission, analysis of myoglobin had the greatest diagnostic rate for MI. Myoglobin was equivalent to TnT 2 h after admission and all tests were equivalent 4 h after admission. Received for publication 20 November 2002 Accepted 20 December 2002 Copyright 2003 Cambridge Medical Publications 82
References 1 The Joint European Society of Cardiology/ American College of Cardiology Committee: Myocardial infarction redefined a consensus document of The Joint European Society of Cardiology/American College of Cardiology Committee for the Redefinition of Myocardial Infarction. Eur Heart J 2000; 21: 1502 1513. 2 Panteghini M: Biochemical assessment of myocardial damage with new diagnostic tools. Cardiologia 1999; 44: 419 425. 3 Penttila I, Penttila K, Rantanen T: Laboratory diagnosis of patients with acute chest pain. Clin Chem Lab Med 2000; 38: 187 197. 4 Lang K, Borner A, Figulla HR: Comparison of biochemical markers for the detection of minimal myocardial injury: superior sensitivity of cardiac troponin-t ELISA. J Intern Med 2000; 247: 119 123. 5 Panteghini M, Apple FS, Christenson RH, Dati F, Mair J, Wu AH: Use of biochemical markers in acute coronary syndromes. IFCC Scientific Division, Committee on Standardization of Markers of Cardiac Damage. International Federation of Clinical Chemistry. Clin Chem Lab Med 1999; 37: 687 693. 6 Hanley J, McNeil BJ: A method of comparing the areas under receiver operating characteristics curves derived from the same cases. Radiology 1983; 148: 839 843. 7 Mair J: Tissue release of cardiac markers: from physiology to clinical applications. Clin Chem Lab Med 1999; 37: 1077 1084. 8 Jurlander B, Clemmensen P, Wagner GS, Grande P: Very early diagnosis and risk stratification of patients admitted with suspected acute myocardial infarction by the combined evaluation of a single serum value of cardiac troponin-t, myoglobin, and creatinine kinase MB mass. Eur Heart J 2000; 21: 382 389. 9 Panteghini M, Pagani F, Bonetti G: The sensitivity of cardiac markers: an evidencebased approach. Clin Chem Lab Med 1999; 37: 1097 1106. 10 Hudson MP, Christenson RH, Newby LK, Kaplan AL, Ohman EM: Cardiac markers: point of care testing. Clin Chim Acta 1999; 284: 223 237. 11 Fox KAA, Cokkinos DV, Decker J, Keil U, Maggioni A, Steg G on behalf of the ENACT investigators: The ENACT Study: a pan- European survey of acute coronary syndromes. Eur Heart J 2000; 21: 1440 1449. 12 Collinson J, Flather MD, Fox KAA, Findley I, Rodrigues E, Dooley P, et al. for the PRAIS-UK Investigators: Clinical outcomes, risk stratification and practice patterns of unstable angina and myocardial infarction without ST elevation: Prospective Registry of Acute Ischaemic Syndromes in the UK (PRAIS-UK). Eur Heart J 2000; 21: 1450 1457. 13 Jernberg T, Lindahl B, Wallentin L: The combination of a continuous 12-lead ECG and troponin T. Eur Heart J 2000; 21: 1464 1472. Address for correspondence Dr S Vatansever Kaptanpaşa mah, Halit Ziya Türkkan Sok, Famas Blokları B blok 6/20, Şişli-İstanbul, Turkey. E-mail: sezai@istanbul.edu.tr 83