Volume 146, Issue 1 , Pages 115-116, 7 January 2011
High diagnostic performance of a high-sensitivity cardiac troponin T assay in patients with suspected acute coronary syndrome
Article Outline
Keywords: Cardiac troponin, Troponin assay, High-sensitivity assay, Acute coronary syndrome, Unstable angina
Management of patients with chest pain remains challenging. Redefinition of acute myocardial infarction (MI) include a rise in cardiac troponin (cTn) above the 99th percentile measured in normal population samples, with a coefficient of variability (CV)
≤10% at the 99th percentile; this last condition cannot be achieved using standard assays of cTn [1].
High-sensitivity assays of cTn (HScTn) have been developed recently, enabling measurements of concentrations that are approximately 10 folds lower than previously measurable. While their accuracy in the detection of MI has been demonstrated [2], [3], their exact performance in the detection of unstable angina (UA) and overall acute coronary syndrome (ACS), as well as the possible additional value of serial measurements, remains to be established [4], [5].
Between June and November 2009, we enrolled consecutive patients admitted to the intensive care unit of Cochin Hospital, Paris, France, for suspected ACS. Patients presenting after a cardiac arrest or with ST-elevation myocardial infarction (STEMI) were not included.
Informed consent was granted from all patients; the study was approved by the local Ethics Committee.
Upon admission, all patients underwent a detailed clinical evaluation, EKG, screening blood tests, and chest roentgenogram. Blood samples were collected on admission, and after 3 and 6h; cTnI was measured immediately, and remaining plasma was stored at −40°C before HScTnT assay.
Plasmatic cTnI concentrations were immunoassayed on an X-pand® HM analyzer (Siemens Healthcare Diagnostics Inc., Newark, NJ). The 99th percentile and CV≤10% are achieved for 0.07ng/mL and 0.14ng/mL, respectively.
The plasma HScTnT concentrations were measured on an Elecsys® 2010 analyzer, using an electrochemiluminescence immunoassay (Roche Diagnostic, Meylan, France). The 99th percentile, with a CV≤10% is achieved for 14ng/L [6].
The final diagnosis of a) acute non-STEMI, b) UA, c) cardiac, non-coronary events, or d) non-cardiac events, was adjudicated by 2 cardiologists separately, based on all medical available charts except for HScTnT measurements. Acute MI was defined according to current guidelines [1]. UA was diagnosed based on 1) clinical manifestations, 2) a plasma cTnI concentration <0.14ng/mL, and 3) a) an EKG indicative of ongoing ischemia, or b) a >70% stenosis of an epicardial coronary artery. Confirmed ACS was defined as acute non-STEMI or UA.
Data were compared using Student's t-test for comparisons of continuous variables, and chi-square or Fisher's exact test for differences in frequency, as appropriate. As cTnI and HScTnT may not have a normal distribution, non parametric tests were used for comparisons;
Eighty-seven patients were enrolled; ACS was confirmed in 43 patients (17 patients with non-STEMI and 26 with UA), cardiac but non-coronary events in 6, and non-cardiac events in 38 patients. Baseline characteristics are reported in Table 1.
Table 1. Baseline characteristics and coronary risk factors in the overall population and in 43 patients with and 44 without confirmed acute coronary syndrome (ACS).
| Patients | ||||
|---|---|---|---|---|
| All (n=87) | With ACS (n=43) | Without ACS (n=44) | p | |
| Age, years | 60.3±13.7 | 64.2±2.1 | 56.5±1.9 | 0.008 |
| Men/women | 56/31 | 32/11 | 24/20 | 0.053 |
| Coronary risk factors | ||||
| Hypertension | 44 (50.6) | 22 (51.3) | 22 (50.0) | 0.914 |
| Hypercholesterolemia | 35 (42.5) | 22 (51.2) | 15 (34.1) | 0.107 |
| Diabetes | 18 (20.7) | 11 (25.6) | 7 (15.9) | 0.265 |
| Obesity | 28 (32.2) | 15 (34.9) | 13 (29.6) | 0.594 |
| Family history | 29 (33.3) | 12 (27.9) | 17 (38.6) | 0.289 |
| Smoking | 23 (26.4) | 10 (23.3) | 13 (29.6) | 0.722 |
| Previous ACS | 18 (20.7) | 9 (20.9) | 9 (20.5) | 0.956 |
| Drug regimen | ||||
| Aspirin | 33 (37.9) | 17 (39.5) | 16 (36.4) | 0.761 |
| Clopidogrel | 19 (21.8) | 9 (20.9) | 10 (22.7) | 0.839 |
| Beta-adrenergic blocker | 32 (36.8) | 15 (34.9) | 17 (38.6) | 0.717 |
| ACE inhibitor or ARB | 32 (36.8) | 18 (41.9) | 14 (31.8) | 0.331 |
| Statin | 35 (40.2) | 18 (41.9) | 17 (38.6) | 0.759 |
| Delay between chest pain and admission, h | 7.6±10.2 | 7.5±8.6 | 7.6±11.7 | 0.953 |
| Duration prior to admission <3h | 27 (31.0) | 14 (32.6) | 13 (29.6) | 0.819 |
| Electrocardiogram | ||||
| Recorded during chest pain | 23 (26.4) | 11 (25.6) | 12 (27.3) | 0.858 |
| Abnormal findings | ||||
| Q wave | 10 (11.5) | 6 (14.0) | 4 (9.1) | 0.477 |
| ST segment depression | 13 (14.9) | 10 (23.3) | 3 (6.8) | 0.039 |
| T wave inversion | 32 (37.2) | 20 (47.6) | 12 (27.3) | 0.051 |
| Procedures | ||||
| Coronary angiography | 71 (81.6) | 38 (88.4) | 33 (75.0) | 0.107 |
| Angioplasty | 27 (38.0) | 27 (62.8) | 0 (0) | |
| Stent implantation | 25 (35.2) | 25 (58.1) | 0 (0) | |
Admission measurements of cTnI and HScTnT were both significantly higher in patients with ACS (p<0.001 versus other patients). However, the area under the ROC curve was 0.83 (95%CI: 0.73–0.91) for cTnI measured on admission and repeated at 6h, 0.88 (95%CI: 0.81–0.96) for HScTnT measured on admission and 0.92 (95%CI: 0.86–0.98) for HScTnT repeated at 3h (p=0.156 and p=0.011 versus cTnI, respectively).
Using a cut-off value of 0.07ng/mL (99th percentile with a CV>10%), cTnI accurately detected 21 patients (49%) with ACS when measured on admission, and 27 patients (62.8%) when repeated at 3 or 6h. HScTnT measured on admission identified adequately 33 patients (77%) with ACS. A second measurement >14ng/L at 3h detected 3 additional patients, and a >30% increase in HScTnT between admission and 3h detected 4 additional patients; repeated measurement at 6h had no additional value (Table 2).
Table 2. Predictive value of serial measurements of cardiac troponins (cTn) in the diagnosis of confirmed acute coronary syndrome (ACS).
| Cause of hospitalization | Predictive values | |||
|---|---|---|---|---|
| Troponin assay | ACS (n=43) | Non ACS (n=44) | Positive (%) | Negative (%) |
| Cardiac troponin Ia (cTnI) | ||||
| >0.14ng/mL on admission | 10 (22.7) | 0 (0) | 100 | 57.1 |
| >0.07ng/mL on admission | 21 (48.8) | 1 (2.3) | 95.4 | 66.2 |
| >0.07ng/mL at 3h | 27 (62.8) | 1 (2.3) | 96.4 | 72.9 |
| >0.07ng/mL at 6h | 26 (60.4) | 1 (2.3) | 96.3 | 71.7 |
| High-sensitivity cardiac troponin Tb (HScTnT) | ||||
| >14ng/L on admission | 33 (76.7) | 6 (13.6) | 84.6 | 79. |
| >14ng/L at 3h | 37 (86.1) | 5 (11.4) | 88.1 | 86.7 |
| >14ng/L at 6h | 37 (86.1) | 6 (13.6) | 86.0 | 86.4 |
| >14ng/L on admission or at 3h | 37 (86.1) | 6 (13.6) | 86.0 | 86.4 |
| >14ng/L on admission or >30% increase within 3h | 38 (88.4) | 11 (25) | 77.6 | 86.8 |
a0.07 |
b14 |
Concentrations of admission HScTnT were higher in patients with acute non-STEMI as the final diagnosis (488±1189ng/L) than in patients with a) UA (47±55ng/L), b) cardiac, non-coronary events (65±102ng/L), and c) non-cardiac events (6±5ng/L). In addition, HScTnT increased by 539% between admission and 3h in patients with acute non-STEMI, versus 58% in patients with UA, while it decreased by 0.3% in patients with final diagnoses of non-cardiac events, and by 8% in patients with non-coronary, cardiac events. Sensitivity in the detection of UA was 65% for HScTnT measured on admission and 77% when repeated at 3h.
In this study, we deliberately examined the capacity of HScTnT to detect confirmed ACS. We assume that identifying ACS is clinically more pertinent than acute MI solely. UA is largely represented (30%) in patients with chest pain related symptoms, and both acute non-STEMI and UA have poor prognosis in the absence of prompt diagnosis and treatment [7].
As an echocardiogram was performed in >2/3 of patients, and angiograms were performed in >80% of patients, we assume that our adjudicated diagnostic is very robust.
Two retrospective studies investigated patients with UA and reported increased accuracy with repeated measurements (3 to 8h) [4], [5]. Our study hereby confirms the additional accuracy offered by duplicate measurements at baseline and 3h later.
Serial assay of HScTnT in our study contributed to valuable information for diagnosing ACS in both following situations: [1] The observation of elevated HScTnT concentrations, either on admission or 3h later, and [2] an increase in the concentrations of HScTnT between the 2 measurements,
Our study is limited by the enrolment of relatively few patients. Its results, therefore, are preliminary and need to be confirmed. Since our study was observational, it did not measure the possible clinical effects of a more accurate detection of ACS.
Acknowledgment
The investigational kits for the assays of HScTnT were provided by Roche Diagnostics (Roche Diagnosis, Meylan, France), who were not involved in any part of the study including its design, data analysis, or preparation of the manuscript.
The authors of this manuscript have certified that they comply with the Principles of Ethical Publishing in the International Journal of Cardiology [8].
References
- . Universal definition of myocardial infarction. J Am Coll Cardiol. 2007;50:2173–2195
- Sensitive troponin I assay in early diagnosis of acute myocardial infarction. N Engl J Med. 2009;361:868–877
- Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Engl J Med. 2009;361:858–867
- . High-sensitivity cardiac troponin T for early prediction of evolving non-ST-segment elevation myocardial infarction in patients with suspected acute coronary syndrome and negative troponin results on admission. Clin Chem. 2010;56:642–650
- . Detection of myocardial injury in patients with unstable angina using a novel nanoparticle cardiac troponin I assay: observations from the PROTECT-TIMI 30 Trial. Am Heart J. 2009;158:386–391
- . Analytical validation of a high-sensitivity cardiac troponin T assay. Clin Chem. 2010;56:254–261
- Ticagrelor versus clopidogrel in patients with acute coronary syndromes. N Engl J Med. 2009;361:1045–1057
- . Ethical authorship and publishing. Int J Cardiol. 2009;131:149–150
PII: S0167-5273(10)00799-0
doi:10.1016/j.ijcard.2010.09.084
© 2010 Elsevier Ireland Ltd. All rights reserved.
Volume 146, Issue 1 , Pages 115-116, 7 January 2011
