- Email: [email protected]
Effects of contemporary troponin assay sensitivity on the utility of the early markers myoglobin and CKMB isoforms in evaluating patients with possible acute myocardial infarction
Clinica Chimica Acta 380 (2007) 213 – 216 www.elsevier.com/locate/clinchim
Short communication
Effects of contemporary troponin assay sensitivity on the utility of the early markers myoglobin and CKMB isoforms in evaluating patients with possible acute myocardial infarction B,B☆ Peter A. Kavsak a,⁎, Andrew R. MacRae b , Alice M. Newman c , Viliam Lustig b , Glenn E. Palomaki d , Dennis T. Ko c , Jack V. Tu c , Allan S. Jaffe e a
Department of Pathology and Molecular Medicine, McMaster University Medical Centre, 1200 Main St. W., HSC 2N52, Hamilton, ON, Canada L8N 3Z5 b Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada c Institute for Clinical Evaluative Sciences, University of Toronto, Canada d Department of Pathology, Women and Infants Hospital, Providence, RI, United States e Cardiovascular Division and Division of Laboratory Medicine, Mayo Clinic, Rochester, MN, United States Received 17 October 2006; received in revised form 28 November 2006; accepted 10 January 2007 Available online 23 January 2007
Abstract Background: The 2003 American Heart Association (AHA) definition for myocardial infarction (MI) requires an “adequate set” (i.e. at least 6 h between measurements) of biomarkers and specifically troponin for the diagnosis of MI. The aim of the present study was to assess the performance of myoglobin, the CKMB isoforms, and cardiac troponin I (cTnI) in specimens earlier than the requisite 6 h after presentation, in a population originally characterized using World Health Organization (WHO) criteria. Methods: In 1996, 228 acute coronary syndrome patients with an “adequate sample set” had their specimens assayed for CKMB isoforms and myoglobin. In 2003, the same specimens were analyzed with the AccuTnI™ troponin I assay and myoglobin (Beckman Coulter Access® immunoassay). Results: The clinical sensitivities for both myoglobin and the CKMB isoforms were N 90% when the population was classified by WHO criteria. However the sensitivities were b 70% when the ESC/ACC MI definition was used. Analyzing cTnI at earlier time points as long as there was at least 3 h between specimens or at least 1 specimen 6 h from pain onset did not misclassify subjects based on adverse outcomes in the year following their presentation. Conclusion: Contemporary assays for cTnI with increased analytical sensitivity reduce the utility of myoglobin and CKMB isoforms to rule-out an AMI. © 2007 Elsevier B.V. All rights reserved. Keywords: Acute myocardial infarction; Outcomes; Shorter interval; Emergency department
1. Introduction
☆ Sources of Support: Ontario Association of Medical Laboratories, Toronto, Ontario, Canada. The AccuTnI and Myoglobin reagent was contributed for the study by an unrestricted grant from Beckman-Coulter Inc. ☆☆ See editorial ‘Cardiac troponin monitoring for detection of myocardial infarction: Newer generation assays are here to stay’, F.S. Apple, doi:10.1016/j. cca.2007.01.002 and the ‘Letter to the editor: Authors' Response to Apple Editorial’, P.A. Kavsak, doi:10.1016/j.cca.2007.01.003. ⁎ Corresponding author. Tel.: +1 905 521 2100; fax: +1 905 521 2344. E-mail address: [email protected] (P.A. Kavsak).
0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2007.01.001
In 2003, the American Heart Association (AHA) published criteria based on the ESC/ACC recommendations that operationalized the use of biomarkers to diagnose acute myocardial infarction (AMI) for epidemiology and clinical research studies [1]. In brief, the diagnosis of AMI was considered to be present if cardiac troponin (cTn) measurements were diagnostic in a setting where ischemia was present. A diagnostic set of cTn markers was defined as containing at least 1 elevated measurement from a set of blood samples taken at least 6 h apart (i.e., from an “adequate set” of biomarkers) and manifesting a
214
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
Table 1 Specimen characteristics Hours from onset to presentation Number of specimens in adequate set Time interval for adequate set Time interval for at least 3 h interval Time interval for at least 1 h interval
Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR)
4.5 ± 5.2 3.0 (2.0–5.0) 4.8 ± 1.5 5.0 (4.0–6.0) 8.4 ± 4.3 7.0 (6.0–8.0) 4.8 ± 4.8 3.0 (3.0–4.0) 2.8 ± 5.3 1.0 (1.0–2.0)
rising or falling pattern [1]. We have previously reported the increase in the prevalence of AMI using these criteria [2,3]. For the present study, we sought to determine the potential diagnostic utility of myoglobin and the CKMB isoforms when a contemporary highly sensitive cTn assay is used. In previous studies, using myoglobin and isoforms, less sensitive cTn assays were used or high cutoff values were employed which may have enhanced the apparent value of these markers [4–7]. This data set also allowed us to assess if earlier measurements of cTn could be used to identify adverse outcomes following the initial event. 2. Patients and methods The study population and its characteristics have been previously reported [2,3,8–10]. Briefly, consecutive patients presenting in 1996 to the emergency department of a community hospital with symptoms suggestive of cardiac ischemia were recruited for a retrospective cardiac marker study. The present study includes only those patients who provided at least two specimens with at least one specimen obtained 6 or more hours after presentation, thus conforming to the AHA adequate set of biomarkers criterion [1]. A retrospective classification of AMI was made if either of the two specimens in the adequate set contained a cTnI concentration N99th percentile (i.e., N0.04 μg/L with the AccuTnI assay, Beckman Coulter Inc., Chaska, MN), [11] consistent with the 2000 ESC/ACC AMI definition [12]. This classification served as the reference for the biomarker and outcome analyses. At the time of study enrolment in 1996, CKMB mass measurement (Abbott IMx) was the only biomarker used for AMI classification in our population and the criteria used were the prevailing WHO criteria [13]. In addition to regular emergency department ordering (e.g. specimens at presentation and at subsequent 12 h intervals), specimens were also collected based on time from onset of symptoms. These specimens were collected hourly until 6 h, and thereafter at 9, 12, 24, and 48 h after onset [2]. These additional specimens were analyzed in 1996 for the “early markers” of CKMB isoforms (MB2 and the ratio of MB2/MB1 with Helena Cardio-Rep® test), and myoglobin (Dade Behring OPUS Plus). The clinical diagnosis for AMI in 1996 was based on electrocardiograms obtained at presentation and 30 min thereafter, together with CKMB mass assay results (i.e., CKMB ≥ 8 μg/L) in the 12-hourly clinical specimens. The study population was subsequently adjudicated by a cardiologist and a specialist in emergency medicine, independently, viewing the expanded set of all the clinical and study CKMB results together, to yield an “expert diagnosis” based on WHO criteria [2,13]. All specimens were then stored predominantly at −70 °C with some periods when, during hospital restructuring, the samples were stored at − 20 °C [2]. In 2003, the specimens were thawed for the first time and analyzed for cTnI and myoglobin with the Beckman Coulter Access® immunoassay system. To verify the stability of the analytes, the myoglobin concentrations obtained in 1996 were compared with the values generated in 2003 in 1141 specimens from our 228 subjects. Linear regression analysis yielded a correlation coefficient, r = 0.99 with a slope = 1.062 and intercept of 3.9 μg/L. The stability of cTnI as measured with the AccuTnI assay has been well documented with respect to multiple freeze-thaw cycles and duration of storage (e.g., 11 ± 2 years) [14,15].
The criteria for a positive result for myoglobin, CKMB isoforms, the AHA definition, and the ESC/ACC definition were as follows: a myoglobin N65 μg/L (females) or N105 μg/L (males) in all specimens within and including the adequate set pair; a simultaneous MB2 N2.5 IU/L at 37 °C and a ratio of MB2/ MB1 N1.60 in all specimens within and including the adequate set pair; for the AHA a cTnI concentration N99th percentile with at least a 20% proportional change in concentration from the 1st specimen (i.e. the initial sample collected at presentation) in the adequate set pair [2]; and for the ESC/ACC a cTnI concentration N99th percentile in either specimen in the adequate set. The strategy for the decreased time interval cTnI analysis has been previously described [3]. The reference time interval used to select the two adequate set specimens was 6 h between specimens following the AHA criteria (i.e., the presentation specimen and a specimen collected at least 6 h thereafter). We applied the 20% change criterion (i.e., our operationalized AHA definition of change) for the specimen pair. Alternative, shorter-interval pairings of specimens were also examined. Three-hour and one-hour interval pairings consisted of the presentation specimen and the first subsequent specimen collected at least 3 later, or at least 1 h later, respectively. Intervals from the time of onset of symptoms were also examined. Six hours from onset pairings used the presentation specimen and the next specimen (at least 1 h later) that was collected at least 6 h after the onset of symptoms of cardiac ischemia. Three hours from onset pairings were similarly constructed. The time of onset of symptoms was identified by the patient or an accompanying person. The 20% change criterion was applied for each specimen pair. For the adverse outcome assessments the primary outcome was the combined endpoint of death/re-AMI at 1 year after presentation. Research ethics board approval was obtained and linkages with the registered persons database (RPDB) for mortality outcomes and the Canadian Institute for Health Information database (CIHI) for recurrent AMI outcomes were obtained [16]. The Kappa statistic was used to determine agreement between different criteria and the binomial probability of an event was used to test the hypothesis that the probability of death/AMI for the different timing intervals was the same as those classified by the ESC/ACC definition with a 6 h separation between samples.
3. Results Details on the demographics, in-hospital course treatment, and outcomes for the 228 subjects conforming to the AHA case definition have been previously reported [8]. Briefly this population (median age 67 years) presented early with a median time of 3 h from onset of symptoms to presentation specimen (Table 1). There was a median 5 specimens (in the interval between, and including, the adequate set) used for myoglobin and CKMB isoform analysis. The adjudicated WHO AMI Table 2 Performance (agreement) of the WHO diagnosis, CKMB isoforms, Myoglobin, and AHA definition compared to the ESC/ACC definition of AMI ESC/ACC Definition for AMI
WHO diagnosis CKMB isoforms Myoglobin AHA definition
yes no yes no yes no yes no
yes
no
Kappa
Sensitivity
Specificity
NPV
46 45 62 29 57 34 83 8
0 137 22 115 19 118 0 137
0.551
51%
100%
75%
0.528
68%
84%
80%
0.502
63%
86%
78%
0.926
91%
100%
94%
AHA = American Heart Association. AMI = Acute myocardial infarction. ESC/ACC = European Society of Cardiology/American College of Cardiology. NPV = Negative predictive power. WHO = World Health Organization.
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
215
Table 3 Outcomes for the different AMI specimen timing intervals as compared to ESC/ACC definition AMI protocol
Number (%) Positive by definition
Death/AMI at 1 year for Positive group
p-value
Number (%) Negative by definition
Death/AMI at 1 year for Negative group
p-value
ESC/ACC At least 6 h interval (AHA) At least 6 h from onset At least 3 h interval At least 3 h from onset At least 1 h interval
91 (39.9%) 83 (36.4%) 79 (34.6%) 79 (34.6%) 66 (28.9%) 63 (27.6%)
38.5% 39.8% 38.0% 39.2% 34.8% 34.9%
reference 0.808 0.929 0.887 0.546 0.564
137 (60.1%) 145 (63.6%) 149 (65.4%) 149 (65.4%) 162 (71.1%) 165 (72.4%)
12.4% 13.1% 14.8% 14.1% 17.9% 18.2%
reference 0.800 0.383 0.533 0.034 0.025
AHA = American Heart Association. AMI = acute myocardial infarction. ESC/ACC = European Society of Cardiology/American College of Cardiology.
diagnosis in 1996 resulted in 46 subjects being classified as AMI (20% AMI prevalence), of whom 24 had ST-segment elevation MI (STEMI). When myoglobin and the CKMB isoforms were assessed with the adjudicated WHO definition of AMI, the sensitivity of both myoglobin and CKMB isoforms was 91% (42/46 with 40 identified by both myoglobin and CKMB isoforms). The negative predictive power (NPV) of these analytes was 97%. However, when the subjects were reclassified with the ESC/ACC definition, the AMI prevalence doubled to 91 of 228 (40%), and the sensitivity and NPV for both myoglobin and CKMB isoforms were substantially diminished (Table 2). The agreement between our operationalized AHA case definition and the ESC/ACC definition was high, suggesting that the 20% proportional change criteria did not misclassify patients. Outcomes for patients classified as positive for AMI and negative for AMI were obtained for each of the different timing intervals and compared to the ESC/ACC definition (Table 3). Patients who were classified as negative for AMI with either the one-hour interval pairing or the three-hour from onset pairing had significantly more adverse events compared to those who were negative by the ESC/ACC definition, requiring the 6 h interval. All other pairings were not significantly different from the ESC/ACC definition in terms of adverse outcome frequency.
tional change) [2] or changing the time interval between specimens (e.g. from at least 6 h to at least 3 h between specimens) [3] on the prevalence of AMI. In the present study we determined that those subjects classified as negative for AMI by the two shortest protocols (i.e. at least 1 h between specimens or at least 3 h after the onset of symptoms) had significantly more adverse events than those negative by the ESC/ACC definition. This is likely due to an insufficient amount of time elapsing before measuring cTnI and as such, a failure to detect a subsequent rise above the cutoff values. However, protocols with an interval between specimens of at least 3 h, or with one specimen at least 6 h after onset of symptoms did not significantly misclassify subjects with respect to adverse outcomes in the first year following their presentation. In addition, decreasing the time required to make a diagnosis for AMI provided similar outcomes to those classified by the ESC/ACC criteria as long as the sample interval was at least 3 h and/or the second sample was at least 6 h after the onset of symptoms. In conclusion, these data, from patients presenting early after pain onset and fulfilling the AHA adequate set requirement, demonstrate that the use of contemporary cardiac troponin assays render the use of myoglobin and CKMB isoforms measurements unnecessary. Acknowledgments
4. Discusssion The present study confirms earlier work suggesting that the diagnostic utility for myoglobin and CKMB isoforms for the early assessment of AMI has been significantly diminished with the 2000 ESC/ACC criteria [17,18]. This study is different from others in that the subjects were selected based on their conformity to the AHA case definition. Moreover, this “adequate set” of samples was analyzed with a highly sensitive cTn assay using the 99th percentile cutoff for AMI designation. Previous reports have used less sensitive cTn assays and higher cutoff values [4–7] or the study population presented much later after onset of symptoms than our population [17,18]. Thus, our study not only supports the conclusions of these earlier reports [17,18] but also provides important additional validation of this concept for subjects who present to the emergency department early after onset of symptoms. We have previously reported the effects of either enforcing a change criterion in troponin measurements (e.g. 20% propor-
This work was supported by the Ontario Association of Medical Laboratories' Research Trust Small Grants program. Unrestricted grants for reagents were also contributed by Beckman Coulter Inc. (AccuTnI and myoglobin) and Helena (CKMB isoforms). References [1] Luepker RV, Apple FS, Christenson RH, et al. AHA Scientific Statement. Case definitions for acute coronary heart disease in epidemiology and clinical research studies. Circulation 2003;108:2543–9. [2] Kavsak PA, MacRae AR, Lustig V, et al. The Impact of the ESC/ACC redefinition of myocardial infarction and new sensitive troponin assays on the frequency of acute myocardial infarction. Am Heart J 2006;152: 118–25. [3] MacRae AR, Kavsak PA, Lustig V, et al. Assessing the requirement for the six-hour interval between specimens in the American Heart Association classification of myocardial infarction in epidemiology and clinical research studies. Clin Chem 2006;52:812–8.
216
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
[4] Fesmire FM, Christenson RH, Fody EP, Feintuch TA. Delta creatine kinase-MB outperforms myoglobin at two hours during the emergency department identification and exclusion of troponin positive non-ST-segment elevation acute coronary syndromes. Ann Emerg Med 2004;44:12–9. [5] Penttila K, Koukkunen H, Halinen M, et al. Myogloblin, creatine kinase MB isoforms and creatine kinase MB mass in early diagnosis of myocardial infarction in patients with acute chest pain. Clin Biochem 2002;35: 647–53. [6] Newby LK, Storrow AB, Gibler B, et al. Beside multimarker testing for risk stratification in chest pain units: the chest pain evaluation by creatine kinase-MB, myoglobin, and troponin I (CHECKMATE) study. Circulation 2001;103:1832–7. [7] Sallach SM, Nowak R, Hudson MP, et al. A change in serum myoglobin to detect acute myocardial infarction in patients with normal troponin I levels. Am J Cardiol 2004;94:864–7. [8] Kavsak PA, MacRae AR, Palomaki GE, et al. Health outcomes categorized by current and previous definitions of acute myocardial infarction in an unselected cohort of troponin naïve emergency department patients. Clin Chem 2006;52:2028–35. [9] Kavsak P, Bhargava R, Lustig V, et al. Performance of CKMB isoforms in a population defined by the 2003 American Heart Association case definition for acute coronary heart disease. Clin Chem 2005;51:A68 [abstr]. [10] Kavsak P, Lustig V, MacRae AR. The clinical performance of myoglobin under the new definition of AMI. Clin Biochem 2004;37:736 [abstr]. [11] Apple FS, Quist HE, Doyle PJ, Otto AP, Murakami MM. Plasma 99th percentile reference limits for cardiac troponin and creatine kinase MB
[12]
[13]
[14] [15]
[16]
[17]
[18]
mass for use with European Society of Cardiology/American College of Cardiology Consensus Recommendations. Clin Chem 2003;49:1331–6. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redefined — a consensus document of the Joint European Society of Cardiology/American College of Cardiology Committee for the Redefinition of Myocardial Infarction. J Am Coll Cardiol 2000;36:959–69. Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas A, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case fatality rates in 38 populations from 21 countries in four continents. Circulation 1994;90:583–612. Venge P, Lindahl B, Wallentin L. New generation cardiac troponin I assay for the access immunoassay system. Clin Chem 2001;47:959–61. Zethelius B, Johnston N, Venge P. Troponin I as a predictor of coronary heart disease and mortality in 70-year-old men: a community-based cohort study. Circulation 2006;113:1071–8. Austin PC, Daly PA, Tu JV. A multicentre study of the coding accuracy of hospital discharge administrative data for patients admitted to cardiac care units in Ontario. Am Heart J 2002;144:290–6. Eggers KM, Oldgren J, Nordenskjold A, Lindahl B. Diagnostic value of serial measurements of cardiac markers in patients with chest pain: limited value of adding myoglobin to troponin I for exclusion of myocardial infarction. Am Heart J 2004;148:574–81. Ilva T, Eriksson S, Lund J, et al. Improved early risk stratification and diagnosis of myocardial infarction, using a novel troponin I assay concept. Eur J Clin Invest 2005;35:112–6.
Short communication
Effects of contemporary troponin assay sensitivity on the utility of the early markers myoglobin and CKMB isoforms in evaluating patients with possible acute myocardial infarction B,B☆ Peter A. Kavsak a,⁎, Andrew R. MacRae b , Alice M. Newman c , Viliam Lustig b , Glenn E. Palomaki d , Dennis T. Ko c , Jack V. Tu c , Allan S. Jaffe e a
Department of Pathology and Molecular Medicine, McMaster University Medical Centre, 1200 Main St. W., HSC 2N52, Hamilton, ON, Canada L8N 3Z5 b Department of Laboratory Medicine and Pathobiology, University of Toronto, Toronto, ON, Canada c Institute for Clinical Evaluative Sciences, University of Toronto, Canada d Department of Pathology, Women and Infants Hospital, Providence, RI, United States e Cardiovascular Division and Division of Laboratory Medicine, Mayo Clinic, Rochester, MN, United States Received 17 October 2006; received in revised form 28 November 2006; accepted 10 January 2007 Available online 23 January 2007
Abstract Background: The 2003 American Heart Association (AHA) definition for myocardial infarction (MI) requires an “adequate set” (i.e. at least 6 h between measurements) of biomarkers and specifically troponin for the diagnosis of MI. The aim of the present study was to assess the performance of myoglobin, the CKMB isoforms, and cardiac troponin I (cTnI) in specimens earlier than the requisite 6 h after presentation, in a population originally characterized using World Health Organization (WHO) criteria. Methods: In 1996, 228 acute coronary syndrome patients with an “adequate sample set” had their specimens assayed for CKMB isoforms and myoglobin. In 2003, the same specimens were analyzed with the AccuTnI™ troponin I assay and myoglobin (Beckman Coulter Access® immunoassay). Results: The clinical sensitivities for both myoglobin and the CKMB isoforms were N 90% when the population was classified by WHO criteria. However the sensitivities were b 70% when the ESC/ACC MI definition was used. Analyzing cTnI at earlier time points as long as there was at least 3 h between specimens or at least 1 specimen 6 h from pain onset did not misclassify subjects based on adverse outcomes in the year following their presentation. Conclusion: Contemporary assays for cTnI with increased analytical sensitivity reduce the utility of myoglobin and CKMB isoforms to rule-out an AMI. © 2007 Elsevier B.V. All rights reserved. Keywords: Acute myocardial infarction; Outcomes; Shorter interval; Emergency department
1. Introduction
☆ Sources of Support: Ontario Association of Medical Laboratories, Toronto, Ontario, Canada. The AccuTnI and Myoglobin reagent was contributed for the study by an unrestricted grant from Beckman-Coulter Inc. ☆☆ See editorial ‘Cardiac troponin monitoring for detection of myocardial infarction: Newer generation assays are here to stay’, F.S. Apple, doi:10.1016/j. cca.2007.01.002 and the ‘Letter to the editor: Authors' Response to Apple Editorial’, P.A. Kavsak, doi:10.1016/j.cca.2007.01.003. ⁎ Corresponding author. Tel.: +1 905 521 2100; fax: +1 905 521 2344. E-mail address: [email protected] (P.A. Kavsak).
0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2007.01.001
In 2003, the American Heart Association (AHA) published criteria based on the ESC/ACC recommendations that operationalized the use of biomarkers to diagnose acute myocardial infarction (AMI) for epidemiology and clinical research studies [1]. In brief, the diagnosis of AMI was considered to be present if cardiac troponin (cTn) measurements were diagnostic in a setting where ischemia was present. A diagnostic set of cTn markers was defined as containing at least 1 elevated measurement from a set of blood samples taken at least 6 h apart (i.e., from an “adequate set” of biomarkers) and manifesting a
214
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
Table 1 Specimen characteristics Hours from onset to presentation Number of specimens in adequate set Time interval for adequate set Time interval for at least 3 h interval Time interval for at least 1 h interval
Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR) Mean ± SD Median (IQR)
4.5 ± 5.2 3.0 (2.0–5.0) 4.8 ± 1.5 5.0 (4.0–6.0) 8.4 ± 4.3 7.0 (6.0–8.0) 4.8 ± 4.8 3.0 (3.0–4.0) 2.8 ± 5.3 1.0 (1.0–2.0)
rising or falling pattern [1]. We have previously reported the increase in the prevalence of AMI using these criteria [2,3]. For the present study, we sought to determine the potential diagnostic utility of myoglobin and the CKMB isoforms when a contemporary highly sensitive cTn assay is used. In previous studies, using myoglobin and isoforms, less sensitive cTn assays were used or high cutoff values were employed which may have enhanced the apparent value of these markers [4–7]. This data set also allowed us to assess if earlier measurements of cTn could be used to identify adverse outcomes following the initial event. 2. Patients and methods The study population and its characteristics have been previously reported [2,3,8–10]. Briefly, consecutive patients presenting in 1996 to the emergency department of a community hospital with symptoms suggestive of cardiac ischemia were recruited for a retrospective cardiac marker study. The present study includes only those patients who provided at least two specimens with at least one specimen obtained 6 or more hours after presentation, thus conforming to the AHA adequate set of biomarkers criterion [1]. A retrospective classification of AMI was made if either of the two specimens in the adequate set contained a cTnI concentration N99th percentile (i.e., N0.04 μg/L with the AccuTnI assay, Beckman Coulter Inc., Chaska, MN), [11] consistent with the 2000 ESC/ACC AMI definition [12]. This classification served as the reference for the biomarker and outcome analyses. At the time of study enrolment in 1996, CKMB mass measurement (Abbott IMx) was the only biomarker used for AMI classification in our population and the criteria used were the prevailing WHO criteria [13]. In addition to regular emergency department ordering (e.g. specimens at presentation and at subsequent 12 h intervals), specimens were also collected based on time from onset of symptoms. These specimens were collected hourly until 6 h, and thereafter at 9, 12, 24, and 48 h after onset [2]. These additional specimens were analyzed in 1996 for the “early markers” of CKMB isoforms (MB2 and the ratio of MB2/MB1 with Helena Cardio-Rep® test), and myoglobin (Dade Behring OPUS Plus). The clinical diagnosis for AMI in 1996 was based on electrocardiograms obtained at presentation and 30 min thereafter, together with CKMB mass assay results (i.e., CKMB ≥ 8 μg/L) in the 12-hourly clinical specimens. The study population was subsequently adjudicated by a cardiologist and a specialist in emergency medicine, independently, viewing the expanded set of all the clinical and study CKMB results together, to yield an “expert diagnosis” based on WHO criteria [2,13]. All specimens were then stored predominantly at −70 °C with some periods when, during hospital restructuring, the samples were stored at − 20 °C [2]. In 2003, the specimens were thawed for the first time and analyzed for cTnI and myoglobin with the Beckman Coulter Access® immunoassay system. To verify the stability of the analytes, the myoglobin concentrations obtained in 1996 were compared with the values generated in 2003 in 1141 specimens from our 228 subjects. Linear regression analysis yielded a correlation coefficient, r = 0.99 with a slope = 1.062 and intercept of 3.9 μg/L. The stability of cTnI as measured with the AccuTnI assay has been well documented with respect to multiple freeze-thaw cycles and duration of storage (e.g., 11 ± 2 years) [14,15].
The criteria for a positive result for myoglobin, CKMB isoforms, the AHA definition, and the ESC/ACC definition were as follows: a myoglobin N65 μg/L (females) or N105 μg/L (males) in all specimens within and including the adequate set pair; a simultaneous MB2 N2.5 IU/L at 37 °C and a ratio of MB2/ MB1 N1.60 in all specimens within and including the adequate set pair; for the AHA a cTnI concentration N99th percentile with at least a 20% proportional change in concentration from the 1st specimen (i.e. the initial sample collected at presentation) in the adequate set pair [2]; and for the ESC/ACC a cTnI concentration N99th percentile in either specimen in the adequate set. The strategy for the decreased time interval cTnI analysis has been previously described [3]. The reference time interval used to select the two adequate set specimens was 6 h between specimens following the AHA criteria (i.e., the presentation specimen and a specimen collected at least 6 h thereafter). We applied the 20% change criterion (i.e., our operationalized AHA definition of change) for the specimen pair. Alternative, shorter-interval pairings of specimens were also examined. Three-hour and one-hour interval pairings consisted of the presentation specimen and the first subsequent specimen collected at least 3 later, or at least 1 h later, respectively. Intervals from the time of onset of symptoms were also examined. Six hours from onset pairings used the presentation specimen and the next specimen (at least 1 h later) that was collected at least 6 h after the onset of symptoms of cardiac ischemia. Three hours from onset pairings were similarly constructed. The time of onset of symptoms was identified by the patient or an accompanying person. The 20% change criterion was applied for each specimen pair. For the adverse outcome assessments the primary outcome was the combined endpoint of death/re-AMI at 1 year after presentation. Research ethics board approval was obtained and linkages with the registered persons database (RPDB) for mortality outcomes and the Canadian Institute for Health Information database (CIHI) for recurrent AMI outcomes were obtained [16]. The Kappa statistic was used to determine agreement between different criteria and the binomial probability of an event was used to test the hypothesis that the probability of death/AMI for the different timing intervals was the same as those classified by the ESC/ACC definition with a 6 h separation between samples.
3. Results Details on the demographics, in-hospital course treatment, and outcomes for the 228 subjects conforming to the AHA case definition have been previously reported [8]. Briefly this population (median age 67 years) presented early with a median time of 3 h from onset of symptoms to presentation specimen (Table 1). There was a median 5 specimens (in the interval between, and including, the adequate set) used for myoglobin and CKMB isoform analysis. The adjudicated WHO AMI Table 2 Performance (agreement) of the WHO diagnosis, CKMB isoforms, Myoglobin, and AHA definition compared to the ESC/ACC definition of AMI ESC/ACC Definition for AMI
WHO diagnosis CKMB isoforms Myoglobin AHA definition
yes no yes no yes no yes no
yes
no
Kappa
Sensitivity
Specificity
NPV
46 45 62 29 57 34 83 8
0 137 22 115 19 118 0 137
0.551
51%
100%
75%
0.528
68%
84%
80%
0.502
63%
86%
78%
0.926
91%
100%
94%
AHA = American Heart Association. AMI = Acute myocardial infarction. ESC/ACC = European Society of Cardiology/American College of Cardiology. NPV = Negative predictive power. WHO = World Health Organization.
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
215
Table 3 Outcomes for the different AMI specimen timing intervals as compared to ESC/ACC definition AMI protocol
Number (%) Positive by definition
Death/AMI at 1 year for Positive group
p-value
Number (%) Negative by definition
Death/AMI at 1 year for Negative group
p-value
ESC/ACC At least 6 h interval (AHA) At least 6 h from onset At least 3 h interval At least 3 h from onset At least 1 h interval
91 (39.9%) 83 (36.4%) 79 (34.6%) 79 (34.6%) 66 (28.9%) 63 (27.6%)
38.5% 39.8% 38.0% 39.2% 34.8% 34.9%
reference 0.808 0.929 0.887 0.546 0.564
137 (60.1%) 145 (63.6%) 149 (65.4%) 149 (65.4%) 162 (71.1%) 165 (72.4%)
12.4% 13.1% 14.8% 14.1% 17.9% 18.2%
reference 0.800 0.383 0.533 0.034 0.025
AHA = American Heart Association. AMI = acute myocardial infarction. ESC/ACC = European Society of Cardiology/American College of Cardiology.
diagnosis in 1996 resulted in 46 subjects being classified as AMI (20% AMI prevalence), of whom 24 had ST-segment elevation MI (STEMI). When myoglobin and the CKMB isoforms were assessed with the adjudicated WHO definition of AMI, the sensitivity of both myoglobin and CKMB isoforms was 91% (42/46 with 40 identified by both myoglobin and CKMB isoforms). The negative predictive power (NPV) of these analytes was 97%. However, when the subjects were reclassified with the ESC/ACC definition, the AMI prevalence doubled to 91 of 228 (40%), and the sensitivity and NPV for both myoglobin and CKMB isoforms were substantially diminished (Table 2). The agreement between our operationalized AHA case definition and the ESC/ACC definition was high, suggesting that the 20% proportional change criteria did not misclassify patients. Outcomes for patients classified as positive for AMI and negative for AMI were obtained for each of the different timing intervals and compared to the ESC/ACC definition (Table 3). Patients who were classified as negative for AMI with either the one-hour interval pairing or the three-hour from onset pairing had significantly more adverse events compared to those who were negative by the ESC/ACC definition, requiring the 6 h interval. All other pairings were not significantly different from the ESC/ACC definition in terms of adverse outcome frequency.
tional change) [2] or changing the time interval between specimens (e.g. from at least 6 h to at least 3 h between specimens) [3] on the prevalence of AMI. In the present study we determined that those subjects classified as negative for AMI by the two shortest protocols (i.e. at least 1 h between specimens or at least 3 h after the onset of symptoms) had significantly more adverse events than those negative by the ESC/ACC definition. This is likely due to an insufficient amount of time elapsing before measuring cTnI and as such, a failure to detect a subsequent rise above the cutoff values. However, protocols with an interval between specimens of at least 3 h, or with one specimen at least 6 h after onset of symptoms did not significantly misclassify subjects with respect to adverse outcomes in the first year following their presentation. In addition, decreasing the time required to make a diagnosis for AMI provided similar outcomes to those classified by the ESC/ACC criteria as long as the sample interval was at least 3 h and/or the second sample was at least 6 h after the onset of symptoms. In conclusion, these data, from patients presenting early after pain onset and fulfilling the AHA adequate set requirement, demonstrate that the use of contemporary cardiac troponin assays render the use of myoglobin and CKMB isoforms measurements unnecessary. Acknowledgments
4. Discusssion The present study confirms earlier work suggesting that the diagnostic utility for myoglobin and CKMB isoforms for the early assessment of AMI has been significantly diminished with the 2000 ESC/ACC criteria [17,18]. This study is different from others in that the subjects were selected based on their conformity to the AHA case definition. Moreover, this “adequate set” of samples was analyzed with a highly sensitive cTn assay using the 99th percentile cutoff for AMI designation. Previous reports have used less sensitive cTn assays and higher cutoff values [4–7] or the study population presented much later after onset of symptoms than our population [17,18]. Thus, our study not only supports the conclusions of these earlier reports [17,18] but also provides important additional validation of this concept for subjects who present to the emergency department early after onset of symptoms. We have previously reported the effects of either enforcing a change criterion in troponin measurements (e.g. 20% propor-
This work was supported by the Ontario Association of Medical Laboratories' Research Trust Small Grants program. Unrestricted grants for reagents were also contributed by Beckman Coulter Inc. (AccuTnI and myoglobin) and Helena (CKMB isoforms). References [1] Luepker RV, Apple FS, Christenson RH, et al. AHA Scientific Statement. Case definitions for acute coronary heart disease in epidemiology and clinical research studies. Circulation 2003;108:2543–9. [2] Kavsak PA, MacRae AR, Lustig V, et al. The Impact of the ESC/ACC redefinition of myocardial infarction and new sensitive troponin assays on the frequency of acute myocardial infarction. Am Heart J 2006;152: 118–25. [3] MacRae AR, Kavsak PA, Lustig V, et al. Assessing the requirement for the six-hour interval between specimens in the American Heart Association classification of myocardial infarction in epidemiology and clinical research studies. Clin Chem 2006;52:812–8.
216
P.A. Kavsak et al. / Clinica Chimica Acta 380 (2007) 213–216
[4] Fesmire FM, Christenson RH, Fody EP, Feintuch TA. Delta creatine kinase-MB outperforms myoglobin at two hours during the emergency department identification and exclusion of troponin positive non-ST-segment elevation acute coronary syndromes. Ann Emerg Med 2004;44:12–9. [5] Penttila K, Koukkunen H, Halinen M, et al. Myogloblin, creatine kinase MB isoforms and creatine kinase MB mass in early diagnosis of myocardial infarction in patients with acute chest pain. Clin Biochem 2002;35: 647–53. [6] Newby LK, Storrow AB, Gibler B, et al. Beside multimarker testing for risk stratification in chest pain units: the chest pain evaluation by creatine kinase-MB, myoglobin, and troponin I (CHECKMATE) study. Circulation 2001;103:1832–7. [7] Sallach SM, Nowak R, Hudson MP, et al. A change in serum myoglobin to detect acute myocardial infarction in patients with normal troponin I levels. Am J Cardiol 2004;94:864–7. [8] Kavsak PA, MacRae AR, Palomaki GE, et al. Health outcomes categorized by current and previous definitions of acute myocardial infarction in an unselected cohort of troponin naïve emergency department patients. Clin Chem 2006;52:2028–35. [9] Kavsak P, Bhargava R, Lustig V, et al. Performance of CKMB isoforms in a population defined by the 2003 American Heart Association case definition for acute coronary heart disease. Clin Chem 2005;51:A68 [abstr]. [10] Kavsak P, Lustig V, MacRae AR. The clinical performance of myoglobin under the new definition of AMI. Clin Biochem 2004;37:736 [abstr]. [11] Apple FS, Quist HE, Doyle PJ, Otto AP, Murakami MM. Plasma 99th percentile reference limits for cardiac troponin and creatine kinase MB
[12]
[13]
[14] [15]
[16]
[17]
[18]
mass for use with European Society of Cardiology/American College of Cardiology Consensus Recommendations. Clin Chem 2003;49:1331–6. Alpert JS, Thygesen K, Antman E, Bassand JP. Myocardial infarction redefined — a consensus document of the Joint European Society of Cardiology/American College of Cardiology Committee for the Redefinition of Myocardial Infarction. J Am Coll Cardiol 2000;36:959–69. Tunstall-Pedoe H, Kuulasmaa K, Amouyel P, Arveiler D, Rajakangas A, Pajak A. Myocardial infarction and coronary deaths in the World Health Organization MONICA Project. Registration procedures, event rates, and case fatality rates in 38 populations from 21 countries in four continents. Circulation 1994;90:583–612. Venge P, Lindahl B, Wallentin L. New generation cardiac troponin I assay for the access immunoassay system. Clin Chem 2001;47:959–61. Zethelius B, Johnston N, Venge P. Troponin I as a predictor of coronary heart disease and mortality in 70-year-old men: a community-based cohort study. Circulation 2006;113:1071–8. Austin PC, Daly PA, Tu JV. A multicentre study of the coding accuracy of hospital discharge administrative data for patients admitted to cardiac care units in Ontario. Am Heart J 2002;144:290–6. Eggers KM, Oldgren J, Nordenskjold A, Lindahl B. Diagnostic value of serial measurements of cardiac markers in patients with chest pain: limited value of adding myoglobin to troponin I for exclusion of myocardial infarction. Am Heart J 2004;148:574–81. Ilva T, Eriksson S, Lund J, et al. Improved early risk stratification and diagnosis of myocardial infarction, using a novel troponin I assay concept. Eur J Clin Invest 2005;35:112–6.