Usefulness of high-sensitivity troponin T for the evaluation of patients with acute chest pain and no or minimal myocardial damage

Transcription

1 Usefulness of high-sensitivity troponin T for the evaluation of patients with acute chest pain and no or minimal myocardial damage Juan Sanchis, MD, a Alfredo Bardají, MD, b Xavier Bosch, MD, c Pablo Loma-Osorio, MD, d Francisco Marín, MD, e Pedro L. Sánchez, MD, f Julio Núñez, MD, a Arturo Carratalá, PhD, g and José A. Barrabés, MD h Valencia, Tarragona, Barcelona, Girona, Murcia, and Madrid, Spain Background Although high-sensitivity troponins allow early diagnosis of acute myocardial infarction, their role for identification of acute coronary syndrome in patients with normal conventional troponin remains unclear. Methods and results A total of 446 patients presenting to the emergency department with chest pain and normal troponin (common practice assays) in 2 serial samples were included. Both samples were also centrally analyzed for highsensitivity troponin T (hs-tnt) (Roche Diagnostics, Basel, Switzerland). Detection (N3 ng/l) and 99th percentile ( 14 ng/l) cutoffs of the maximum hs-tnt levels (hs-tntmax) were considered. The end points were acute coronary syndrome diagnosis and the composite of in-hospital revascularization or 30-day cardiac events. Results Acute coronary syndrome was adjudicated to 84 patients (19%), and 62 (14%) had the composite end point. In univariate setting, hs-tntmax N3 ng/l exhibited high sensitivity (87% and 92%, respectively) and negative predictive value (93% and 97%) for both end points, whereas hs-tntmax 14 ng/l provided high specificity (90% and 89%), although low positive predictive values (40% and 33%). After adjusting for clinical (pain characteristics and risk factors) and electrocardiographic data, there was a stepped increase of risk across hs-tntmax categories ( 3, N3 but b14, and 14 ng/l) for both end points; however, the discriminative capacity added was marginal (integrated discrimination improvement of 2.6% and 3.5%, respectively). Conclusions Clinical and electrocardiographic data remain the most important tools for the evaluation of patients with chest pain and with no or minimal myocardial damage. The main contribution of hs-tnt is the high negative predictive value of undetectable levels ( 3 ng/l). (Am Heart J 2012;164: e1.) High-sensitivity troponin has recently been introduced in clinical practice for the diagnosis and management of patients with acute chest pain. The main contribution of high-sensitivity assays compared with conventional methods relies on their greater precision to detect minor troponin elevations in the proximity of the 99th From the a Servicio de Cardiología Hospital Clínico Universitario, INCLIVA, Universitat de València, Valencia, Spain, b Servicio de Cardiología, Hospital Universitari Joan XXIII, IISPV, Universitat Rovira i Virgili, Tarragona, Spain, c Servicio de Cardiología, Hospital Cllinic, Universitat de Barcelona, Barcelona, Spain, d Servicio de Cardiología, Hospital Josep Trueta, Girona, Spain, e Servicio de Cardiología, Hospital Virgen Arrixaca, Murcia, Spain, f Servicio de Cardiología, Hospital Gregorio Marañón, Madrid, Spain, g Servicio de Bioquímica Clínica, Hospital Clínico Universitario, INCLIVA, Valencia, Spain, h Servicio de Cardiologia, Hospital Universitari Vall d'hebron, Institut de Recerca (VHIR), Universitat Autònoma de Barcelona, Barcelona, Spain. Submitted March 2, 2012; accepted May 21, Reprint requests: Juan Sanchis, MD, Cardiology Department, University Clinic Hospital, Medicine Department, University of Valencia, Blasco Ibáñez 17, València, Spain. sanchis_juafor@gva.es /$ - see front matter 2012, Mosby, Inc. All rights reserved. doi: /j.ahj percentile cutoff of a normal reference population and also on the earlier elevation that facilitates a more rapid decision making in the emergency department. 1,2 Several studies have evidenced the superiority of highsensitivity troponin over conventional troponin for diagnosis of acute myocardial infarction and for prognostic assessment in patients with acute coronary syndromes. 1-8 However, its usefulness for diagnosis of unstable angina in patients with chest pain and normal conventional troponin has not been fully demonstrated. 1,9 These patients show minor troponin raise or no elevation at all, and their correct classification, which has important implications, remains a challenge. 10 Because the diagnostic tools in the emergency department are scarce, often only consisting of the clinical history and electrocardiogram (ECG), the potential capacity of highsensitivity troponins for reclassifying patients from unstable angina to acute myocardial infarction and for ruling out the complete spectrum of acute coronary syndromes constitutes an attractive hypothesis.

2 American Heart Journal Volume 164, Number 2 Sanchis et al 195 Figure 1 Chest pain Normal troponin ECG suggestive of ischemia (n = 58) ECG not suggestive of ischemia (n = 388) Non-invasive Initial coronary Exercise test Initial coronary Non-invasive imaging test (n=16) angiogram (n = 42) (n = 352) angiogram (n = 29) imaging test (n = 7) n = 20 n = 6 n = 48 Total coronary angiogram (n = 48) Total coronary angiogram (n = 86) Total non-invasive imaging test (n = 27) n = 9 Patient diagnostic algorithm. The Elecsys high-sensitivity troponin T (hs-tnt) (Roche Diagnostics, Basel, Switzerland) is an assay that meets the 2 criteria for high-sensitivity level 4 assay: (a) total imprecision at the 99th percentile 10% and (b) measurable concentration below the 99th percentile at a concentration value above the assay's limit of detection for at least 50% of healthy individuals. 11 In a prior study, its diagnosis accuracy for unstable angina in patients with normal conventional troponin, although modest, was superior to other highsensitivity assays. 1 The present multicentric study includes a series of patients presenting to the emergency department with acute chest pain and normal troponin with the assays used in normal clinical practice of each hospital. Patients were managed according to a standardized protocol, and hs-tnt was centrally and blindingly determined. The aim was to evaluate hs-tnt for the diagnosis of acute coronary syndrome, which is the most important objective of the decision process in the emergency departments, in this population with no or minimal myocardial damage. Methods The PITAGORAS study was a prospective Spanish multicenter study designed by the Working Group on Ischemic Heart Disease and Coronary Units of the Spanish Society of Cardiology. It aimed to evaluate the ability of hs-tnt for diagnosis of acute coronary syndrome in patients presenting to the emergency department with acute chest pain and showing normal serial troponin with the assays used in normal clinical practice of each hospital. The online Appendix A lists the participant sites and investigators and the online Appendix B Supplemental Table the troponin assays used in each hospital. The inclusion criterion was the presentation to the emergency department with acute chest pain of possible coronary origin at criterion of the cardiologist on duty, being the pain onset within the previous 24 hours. Exclusion criteria were the following: (a) persistent ST-segment elevation in the ECG, (b) troponin elevation in any of 2 serial determinations (at arrival and 6-8 hours later), (c) prior diagnosis of ischemic heart disease by either the finding of significant stenosis in a prior coronary angiogram or previously documented myocardial infarction, (d) left bundle-branch block or other noninterpretable ECG or inability to perform an exercise test, (e) structural heart disease different to ischemic heart disease, and (f) concomitant heart failure or significant bradyarrhythmia (b55 beat/min) or tachyarrythmia (N110 beat/min) at admission. Because the primary end point of the study (diagnosis of acute coronary syndrome) was mainly based on angiographic findings (see below), patients with previously documented ischemic heart disease were excluded because the result of the coronary angiogram would lead to the diagnosis of acute coronary syndrome in all cases. A total of 446 patients were included. The ethical committee of each participant hospital approved the study, and the included patients signed the corresponding writing consent. Patient management Figure 1 shows the diagnostic algorithm. All patients underwent clinical assessment and ECG at admission. Those patients with an ECG suggestive of ischemia (ST-segment depression

3 196 Sanchis et al American Heart Journal August 2012 Table I. Baseline characteristics of the patient population All patients (N = 446) Group 1 (n = 154) Group 2 (n = 229) Group 3 (n = 63) Age (y) 60 ± ± ± ± 12 Men 262 (59%) 78 (51%) 143 (62%) 41 (65%) Effort-related pain 144 (32%) 42 (27%) 74 (32%) 28 (44%) 2 pain episodes in the 153 (34%) 48 (31%) 73 (32%) 32 (51% last 24 hours Current smokers 111 (25%) 36 (23%) 65 (28%) 10 (16%) Hypertension 242 (54%) 70 (46%) 126 (55%) 46 (73%) Hypercholesterolemia 205 (46%) 61 (40%) 110 (48%) 34 (54%) Diabetes 89 (20%) 21 (14%) 48 (21%) 20 (32%) Family history of ischemic 61 (14%) 25 (16%) 30 (13%) 6 (10%) heart disease Peripheral artery disease 6 (1.3%) 1 (0.6%) 5 (2.2%) 0 Cerebrovascular disease 13 (2.9%) 3 (1.9%) 8 (3.5%) 2 (3.2%) Prior antiplatelet treatment 89 (20%) 19 (12%) 54 (24%) 16 (25%) Prior statins 129 (29%) 29 (19%) 78 (34%) 22 (35%) Prior β-blockers 50 (11%) 15 (10%) 24 (11%) 11 (18%) Admission systolic blood 144 ± ± ± ± 23 pressure (mm Hg) Admission diastolic blood 80 ± ± ± ± 15 pressure (mm Hg) Admission heart rate 76 ± ± ± ± 16 (beat/min) ECG ischemia 58 (13%) 13 (8%) 29 (13%) 16 (25%) Group I: hs-tntmax 3 ng/l. Group II: 3 N hs-tntmax b 14 ng/l. Group III: hs-tntmax 14 ng/l. 0.5 mm or T-wave inversion 2 mm in 2 contiguous leads) were hospitalized and planned for coronary angiogram. The remaining patients were subjected to a noninvasive stress test for detection of ischemia (standard exercise ECG test or imaging stress test according to local practice). Those patients with a positive stress test result were planned for coronary angiography, whereas those with a negative result were discharged. In patients not reaching at least 85% of the maximal predicted heart rate in an exercise test, subsequent evaluation (either by noninvasive or invasive methods) was encouraged. The following clinical data were collected at admission: (a) characteristics of chest pain, such as effort-related pain in the last 2 weeks or 2 pain episodes in the last 24 hours; (b) coronary risk factors; (c) prior history of peripheral artery disease; (d) prior history of cerebrovascular disease; and (e) prior treatment with antiplatelet drugs, β-blockers, or statins. High-sensitivity troponin measurement The 2 blood samples used for troponin assessment were used for hs-tnt measurement as well. For this purpose, blood was collected in serum tubes and, after centrifugation, frozen at 80 C until they were assayed in a central laboratory blinded to the clinical results. High-sensitivity troponin T measurement was performed using the Elecsys hs-tnt assay (Roche Diagnostics). This assay is a modification of the fourth-generation troponin T assay to improve its limit of quantitation and precision. Limit of detection has been determined to 3 ng/l, and the 99th percentile of a healthy reference population, at 14 ng/l. 1 End points The primary end point was the diagnosis of acute coronary syndrome at the index episode. This diagnosis was adjudicated by 3 cardiologists blinded to the results of hs-tnt. The criteria for the diagnosis of acute coronary syndrome were significant stenosis in the coronary angiogram, positive stress test in the absence of coronary angiogram, or cardiac events at 30-day follow-up. To increase the robustness of the diagnosis, the protocol strongly encouraged performing coronary angiography in patients with baseline ECG changes suggestive of ischemia or with a positive stress test result. Conversely, acute coronary syndrome was not considered in the case of a normal coronary angiogram or a negative maximal stress test. The secondary end point was the composite of in-hospital revascularization or cardiac events at 30-day follow-up, defined by death, acute myocardial infarction (chest pain with raised troponin), or postdischarge revascularization. Acute myocardial infarction was also defined if creatine kinase MB mass increased to 3 times the upper limit of normal after coronary angioplasty or to 5 times the upper limit of normal after coronary bypass surgery. Creatine kinase MB was measured at 8 and 24 hours after a revascularization procedure. Statistical analysis Continuous data are presented as mean ± SD or median (25-75 interquartile intervals), and categorical data, as numbers and percentages. Univariate analysis was performed to investigate the relation between the clinical variables and ECG as well as creatinine values at admission with the end points. Those related variables were included in a logistic multivariable model. The discriminatory ability of hs-tnt values in the arrival blood sample (hs-tnt1), the second blood sample (hs-tnt2), and the maximum value of both (hs-tntmax) was tested by the area under the receiver operating characteristic curve (AUC) for the end points. The absolute (dtnt) and percentage (%dtnt)

4 American Heart Journal Volume 164, Number 2 Sanchis et al 197 Table II. Area under the receiver operating characteristic curves for the diagnosis of acute coronary syndrome and the composite end point (in-hospital revascularization or 30-day cardiac events) Acute coronary syndrome AUC Composite end point AUC Figure 2 1,00 hs-tnt , P = , P =.0001 hs-tnt , P = , P =.0001 hs-tntmax 0.713, P = , P =.0001 dtnt 0.530, P = , P =.6 %dtnt 0.522, P = , P =.8 change between hs-tnt1 and hs-tnt2 were also analyzed. According to the results of this analysis, hs-tntmax, which had the best discriminatory capacity, was considered for further analysis. Sensitivity, specificity, positive and negative predictive values of detection (N3 ng/l), and 99th percentile ( 14 ng/l) hs-tntmax cutoffs for the end points were calculated. Likewise, hs-tntmax was introduced in the multivariable models as continuous variable and as categorical variable with 3 categories: 3 ng/l, N3 but b14 ng/l, and 14 ng/l. The odds ratio (OR) and 95% CIs were estimated. Next, we proceeded to calculate the added discriminatory value of hs-tntmax against the base model created with clinical and ECG variables using the following measures: (1) increment in the AUC and (2) integrated discrimination improvement (IDI) index. Statistical analysis was performed using SPSS 18.0 software version (SPSS, Inc, Chicago, IL) and STATA 11.1 (STATACorp, College Station, TX). This work was supported by an unrestricted grant from Roche Diagnostics. Drs Sanchis, Núñez, and Bosch were also supported by the Instituto de Salud Carlos III (FEDER), Red HERACLES. Drs Sánchez and Barrabés were also supported by the Instituto de Salud Carlos III (FEDER), Red RECAVA. The authors are solely responsible for the design and conduct of this study, all study analyses, and drafting and editing of the manuscript. Sensitiv vity,75,50,25 0,00 0,00,25 1,00 related to revascularization procedures at the initial hospitalization, whereas 2 patients had postdischarge spontaneous myocardial infarction. Revascularization procedures were performed in 61 patients (59 during the initial hospitalization)., Specificity it Hs-TnTmax Hs-TnT2 Hs-TnT1 Performance of hs-tnt1 (high-sensitive troponin at arrival), hs-tnt2 (high-sensitive troponin 6-8 hours after arrival), and hs-tntmax (maximum high-sensitive troponin values) for acute coronary syndrome diagnosis.,75 Results Patient population Table I shows the characteristics of the patient population, and Figure 1, the diagnostic algorithm. Median time (25-75 interquartile intervals) from chest pain onset to first blood sample was 4 hours (3-7 hours). Median values (25-75 interquartile intervals) of hs-tnt1, hs-tnt2, and hs-tntmax were, respectively, 3.8 (3-7.6 ng/l), 4.3 (3-9.1 ng/l), and 4.53 (3-9.3 ng/l). With respect to dtnt and %dtnt, they were 0 ng/l ( 0.28 to 1.04) and 0% ( 3.6 to 14.6). A final diagnosis of acute coronary syndrome was adjudicated to 84 patients (19%), and 62 (14%) had the composite end point. Acute coronary syndrome diagnosis was established based on the angiographic findings in 75 patients (89%) by a positive stress test in 8 patients (who refused a coronary angiogram) and by the occurrence of cardiac events in 1 patient who was discharged after a negative exercise test. No patient died. A total of 6 patients had an acute myocardial infarction, in 4 of them Relation between hs-tnt and outcomes Table II and Figure 2 show the AUC of hs-tnt1, hs- TnT2, hs-tntmax, dtnt, and %dtnt, as continuous variables, for the outcomes. The hs-tnt1, hs-tnt2, and hs-tntmax were significantly related with the end points in contrast to dtnt and %dtnt. According to the AUC values, hs-tntmax was the best discriminator. The distribution of patients according to the prespecified detection (N3 ng/l) and 99th percentile ( 14 ng/l) limits of hs-tntmax was as follows: 154 patients (35%) had hs-tntmax 3 ng/l; 229 (51%), hs-tntmax N3 but b14 ng/l; and 63 (14%), hs-tntmax 14 ng/l. Table III presents the sensitivity, specificity, positive predictive value, and negative predictive value of hs-tntmax N3 ng/l and hs-tntmax 14 ng/l for the end points. To point out, hs-tntmax N3 ng/l exhibited high sensitivity (87% and 92%) and negative predictive value (93% and 97%) for acute coronary syndrome diagnosis and the composite secondary end point, respectively. On the other hand, the main contribution of hs-tntmax 14 ng/l was its

5 198 Sanchis et al American Heart Journal August 2012 Table III. Sensitivity, specificity, positive predictive value and negative predictive value of the detection (N3 ng/l), and 99th percentile ( 14 ng/l) limits of hs-tnt in the whole population (N = 446) Sensitivity Specificity PPV NPV Acute coronary syndrome hs-tntmax N hs-tnt1 N hs-tntmax Composite end point hs-tntmax N hs-tnt1 N hs-tntmax PPV, Positive predictive value; NPV, negative predictive value. specificity (90% and 89%), although with low positive predictive value (40% and 33%). We also tested the potential value of a single determination at admission (hs- TnT1), this analysis showing that hs-tnt1 N3 ng/l had a high sensitivity (81% and 85%) and negative predictive value (91% and 95%), albeit inferior to hs-tntmax N3 ng/l. Multivariable analysis In the multivariable analysis, adjusted by clinical data and ECG findings (base model), hs-tntmax evaluated as continuous variable was independently associated with the diagnosis of acute coronary syndrome (per nanogram per liter, OR 1.04, 95% CI , P =.009) and the composite end point (per nanogram per liter, OR 1.04, 95% CI , P =.007). When hs-tntmax was categorized as 3, N3 but b14, and 14 ng/l, there was a stepped increase of risk across hs-tntmax categories for both end points (Tables IV and V). However, the incremental discriminative ability added to the models by hs-tntmax was modest, being statistically significant in terms of IDI (2.6% for acute coronary syndrome diagnosis and 3.5% for the composite end point) but nonsignificant in terms of AUC. Different troponin assays were used in the different hospitals for patient screening. Three hospitals used a high-sensitivity assay (troponin I, ADVIA Dimension EXL LOCI Module; Siemens, Erlangen, Germany). After adjusting the multivariable models for the troponin assay used, the association of hs-tntmax with acute coronary syndrome diagnosis (hs-tntmax N3 but b14 ng/ L: OR 2.5, 95% CI , P =.02; hs-tntmax 14 ng/l: OR 3.9, 95% CI , P =.006) and the composite end point (hs-tntmax N3 but b14 ng/l: OR 4.5, 95% CI , P =.004; hs-tntmax 14 ng/l: OR 6.5, 95% CI , P =.002) persisted. Subgroup without ischemia in the ECG We separately analyzed the subgroup with a nonischemic ECG because these represent the most challenging patients for initial decision making. A total of 338 patients had a nonischemic initial ECG. Acute coronary Table IV. Multivariable model for prediction of acute coronary syndrome Variable OR 95% CI P Base model Age (per y) Male sex Effort-related pain Family history Diabetes Peripheral artery disease ECG ischemia AUC = hs-tntmax 3 b hs-tntmax b 14 versus hs-tntmax 3 hs-tntmax 14 versus hs-tntmax 3 AUC = 0.810, P =.18 IDI = 2.6 ( ) OR, Odds ratio; CI, confidence interval; P, Statistical significance; hs-tntmax, maximum high-sensitive troponin value (ng/l); AUC, area under the receiver operating characteristic curve; IDI, integrated discrimination improvement. Compared with the base model. syndrome diagnosis was established in 56 patients (14%), and 36 (9.3%) presented the composite end point. The hs- TnTmax N3 ng/l retained a high sensitivity (84% and 90%) and negative predictive value (94% and 97%) for both end points, respectively. The hs-tntmax 14 ng/l showed 89% specificity for both acute coronary syndrome diagnosis and the composite end point but low positive predictive values (23% and 19%). Discussion Main findings The results of the present study indicate that hs-tnt was associated to acute coronary syndrome diagnosis in patients presenting to the emergency department with chest pain and no or minimal myocardial damage. Nevertheless, the discrimination power added over clinical and ECG data was marginal. The hs-tnt cutoff at the 99th percentile ( 14 ng/l) showed a low positive predictive value that limited the rule in. Undetectable hs- TnT levels ( 3 ng/l), however, provided a high negative predictive value that could contribute to the rule out. Chest pain with normal conventional troponin Diagnosis of acute coronary syndrome and risk stratification in patients with acute chest pain and normal conventional troponin is challenging. Clinical history is of paramount importance. The present study demonstrates the robustness of the predictive models that included clinical and ECG variables. In fact, hs-tnt added marginal discriminatory capacity. Many of the clinical data found predictive in this study, such as effort related chest pain, 2 pain episodes in the last 24 hours, age, male sex, diabetes, and family history, were also predictive in other

6 American Heart Journal Volume 164, Number 2 Sanchis et al 199 Table V. Multivariable model for prediction of the composite end point (in-hospital revascularization or 30-day cardiac events) Variable OR 95% CI P Base model Age (per y) Male sex Effort-related pain episodes in the last 24 h Family history Prior statins ECG ischemia AUC = hs-tntmax 3 b hs-tntmax b 14 versus hs-tntmax 3 hs-tntmax 14 versus hs-tntmax 3 AUC = 0.809, P =.18 IDI = 3.5 ( ) Compared with the base model. studies In addition, prior statin treatment and peripheral artery disease, as indicators of high cardiovascular risk, were also predictors. However, the negative predictive value of clinical and ECG data is not usually sufficient for decision making. Therefore, a stress test or imaging technique is commonly required to drive their further management This implies an important logistic constraint. Whether high-sensitivity troponin could modify this scenario and facilitate patient management remains to be elucidated. High-sensitivity troponin and acute coronary syndrome Prior studies have demonstrated the role of highsensitivity troponin for early diagnosis of acute myocardial infarction. 1-8 However, its accuracy for diagnosis of unstable angina in patients with normal conventional troponin was limited. 1,9 On one hand, some patients can present unstable angina with normal troponins, that is, below the 99th percentile cutoff. On the other hand, lowering troponin cutoff to the 99th percentile can increase the number of false-positive results. Among the high-sensitive assays tested, hs-tnt was the more precise for unstable angina diagnosis in a prior study conducted by Reichlin et al. 1 Our study focused on much lower risk patients than the series of Reichlin et al, which had a 17% prevalence or myocardial infarction. 1 In a recent study, hs-tnt elevation predicted mortality in patients with unstable angina and normal conventional troponin. 18 In contrast, although we found a stepped increase of risk across hs-tntmax categories of progressive higher levels, the 99th percentile cutoff ( 14 ng/l) provided a low positive predictive value. Alternatively, the hs-tnt detection cutoff (N3 ng/l) optimizes the negative predictive value. A prior study showed that the absence of detectable hs-tnt levels ruled out acute myocardial infarction with very high negative predictive value, although the capacity for diagnosis of unstable angina was not tested. 19 In the present study, the absence of detectable hs-tnt levels after 2 determinations showed negative predictive values of 93% for acute coronary syndrome diagnosis and of 97% for the composite end point. The high negative predictive value of undetectable hs-tnt could be of clinical use. This should not mean the discharge of all these patients without stress test evaluation but could help physicians in deciding, after integrating this information with clinical and ECG variables, early discharge of some of them without waiting for stress testing. Dynamic changes of high-sensitivity troponin levels have also been studied, and the usefulness for diagnosis of acute myocardial infarction has been demonstrated. 10,20,21 Different δ changes have been proposed based on the type of assay as well as on the consideration of absolute or relative change. 5,22-24 However, we did not observe any association between either absolute or percentage changes of hs-tnt and the outcomes. Probably, in situations with minor hs-tnt raise, as in our patients, the small magnitude of the hs-tnt change could be within the variability limits of the assay and remain undetected. Limitations Different troponin assays and with different cutoffs, as in normal practice, were used to include the patients. On the other hand, our study was not focused on the whole spectrum of acute coronary syndrome, as non STsegment elevation acute myocardial infarction was excluded, but on the more challenging patients for diagnosis, those with normal conventional troponin. Patients with prior documented coronary artery disease or with contraindication to exercise test were also excluded. The adjudication of the causal relation between symptoms and angiographic findings in patients with well-known prior coronary lesions and no other signs of myocardial involvement (ie, ECG changes or troponin elevation) is particularly confusing. However, in the case of new symptoms with presumably new coronary lesions, the casual relation seems to be more plausible. Our criteria for unstable angina diagnosis were also used in other studies. 1 The reason to also exclude those patients with contraindications to exercise was with the aim of using the same noninvasive test (exercise test) in all the hospitals. As a whole, according to the previous data from one of the participant centers, the population included represents 50% of all patients presenting to the emergency department with chest pain and normal conventional troponin. 12,14 The relatively high incidence of acute coronary syndrome in our study may be explained, in part, by the fact that all patients were screened by a cardiologist, making it likely that very-low-risk patients were not included. Finally, the second blood sample for troponin measurement was

7 200 Sanchis et al American Heart Journal August 2012 obtained 6 to 8 hours after presentation because the study was designed before the publication of the current guidelines that recommend only 3 hours of delay. 10 Conclusions A comprehensive clinical evaluation is the most important tool for diagnosis and risk stratification of patients with acute chest pain and with no or minimal myocardial damage. The main contribution of hs-tnt seems to be the high negative predictive value of undetectable levels ( 3 ng/l). Including the criterion of undetectable hs-tnt for the rule out could help decision making and risk stratification of these patients. Disclosures Conflict of interest: Dr Marín has received funding for research from Abbot Vascular, Madrid, Spain. References 1. Reichlin T, Hochholzer W, Bassetti St, et al. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. N Eng J Med 2009;361: Keller T, Zeller T, Peetz D, et al. Sensitive troponin I assay in early diagnosis of acute myocardial infarction. N Eng J Med 2009;361: Bonaca M, Scirica B, Sabatine M, et al. Prospective evaluation of the prognostic implications of improved assay performance with a sensitive assay for cardiac troponin I. J Am Coll Cardiol 2010;55: Lindahl B, Venge P, James St. The new high-sensitivity cardiac troponin T assay improves risk assessment in acute coronary syndromes. Am Heart J 2010;160: Giannitsis E, Becker M, Kurz K, et al. 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: Mills NL, Churchhouse A, Lee KK, et al. Implementation of a sensitive troponin I assay and risk of recurrent myocardial infarction and death in patients with suspected acute coronary syndrome. JAMA 2011; 305: Reiter M, Twerenbold R, Reichlin T, et al. Early diagnosis of acute myocardial infarction in the elderly using more sensitive cardiac troponin assays. Eur Heart J 2011;32: Reiter M, Twerenbold R, Reichlin T, et al. Early diagnosis of acute myocardial infarction in patients with pre-existing coronary artery disease using more sensitive cardiac troponin assays. Eur Heart J 2012;33: Januzzi JJ, Bamberg F, Lee H, et al. High-sensitivity troponin T concentrations in acute chest pain patients evaluated with cardiac computed tomography. Circulation 2010;121: Hamm Ch W, Bassand P, Agewall St, et al. ESC guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J 2011;32: Apple FS, Collison PO. Analytical characteristics of hig-sensitivity cardiac troponin assays. Clin Chem 2012;58: Sanchis J, Bodi V, Nuñez J, et al. New risk score for patients with acute chest pain, non ST-segment deviation and normal troponin concentrationss. A comparison with the TIMI risk score. J Am Coll Cardiol 2005;46: Sanchis J, Bodí V, Núñez J, et al. Usefulness of pain presentation characteristics for predicting outcome in patients presenting to the hospital with chest pain of uncertain origin. Emerg Med J 2011;28: Sanchis J, Bodí V, Núñez J, et al. Identification of very low risk chest pain using clinical data in the emergency department. Int J Cardiol 2011;150: Sanchis J, Bodí V, Núñez J, et al. Usefulness of early exercise testing and clinical risk score for prognostic evaluation in chest pain units without preexisting evidence of myocardial ischemia. Am J Cardiol 2006;97: Jeetley P, Burden L, Stoykova B, et al. Clinical and economic impact of stress echocardiography compared with exercise electrocardiography in patients with suspected acute coronary syndrome but negative troponin: a prospective randomized controlled study. Eur Heart J 2007;28: Goldstein JA, Gallagher MJ, O'Neill WW, et al. A randomized controlled trial of multi-slice coronary computed tomography for evaluation of acute chest pain. J Am Coll Cardiol 2007;49: Ndrepepa G, Braun S, Mehilli J, et al. Prognostic value of sensitive troponin T in patients with stable and unstable angina and undetectable conventional troponin. Am Heart J 2011;161: Body R, Carley S, McDowell G, et al. Rapid exclusion of acute myocardial infarction in patients with undetectable troponin using a high-sensitivity assay. J Am Coll Cardiol 2001;58: Thygesen K, Mair J, Katus H, et al. Recommendations for the use of cardiac troponin measurement in acute cardiac care. Eur Heart J 2010;31: Jaffe AA, Ordoñez-Llanos J. High sensitivity troponin in chest pain and acute coronary syndromes. A step forward? Rev Esp Cardiol 2010;63: Apple FS, Pearce LA, Smith SW, et al. Role of monitoring changes in sensitive cardiac troponin I assay results for early diagnosis of myocardial infarction and prediction of risk of adverse events. Clin Chem 2009;55: Kavsak PA, Ko DT, Wang X, et al Universal myocardial infarction definition change criteria for risk stratification by use of a high-sensitivity cardiac troponin I assay. Clin Chem 2010;56: Reichlin T, Irfan A, Twerenbold R, et al. Utility of absolute and relative changes in cardiac troponin concentrations in the early diagnosis of acute myocardial infarction. Circulation 2011;124:

8 American Heart Journal Volume 164, Number 2 Sanchis et al 200.e1 Appendix A. List of Spanish participating sites and Investigators The following investigators participated in the PITA- GORAS study: Hospital Clínico Universitario, Valencia: Juan Sanchis, Arturo Carratalá, Lidia Abellán, Mauricio Pellicer and Cristina Gómez-Monsoliu. Hospital Valle de Hebrón, Barcelona: Jose A Barrabés, Albert Alonso and Luzma Cruz. Hospital Joan XXIII, Tarragona: Alfredo Bardají and Angels Vilanova. Hospital Gregorio Marañón, Madrid: Pedro L Sánchez, Mercedes García and Carolina Puerta. Hospital Meixoeiro, Vigo: Fran Calvo and Carmen Farreirós. Hospital Clínic, Barcelona: Xavier Bosch, Jose L Bedini and Bernar González. Hospital Josep Trueta, Girona: Pablo Loma, Manel Ramíre and Marina Fontant. Hospital Clínico, Valladolid: Alberto San Román, Encarna Lago and Inmaculada Vidriales. Hospital Virgen Arrixaca, Murcia: Francisco Marín, Ana Romero and Pedro Martínez.. Hospital General Asturias: Pablo Avanzas and Francisco Álvarez. Hospital Germans Trías i Pujol, Badalona: Silvia Serrano and Amparo Galán. Appendix B. Supplemental table Supplemental Table. Participant hospitals and troponin assays used Hospital Patients Troponin assay Josep Trueta. Girona Clínic. Barcelona Meioxeiro. Vigo Virgen Arrixaca. Murcia Clínico. Valladolid Gregorio Marañón. Madrid Clínico. Valencia Valle Hebrón. Barcelona Germans Trías i Pujol Badalona General Central. Oviedo 50 Troponin T (4th 107 Troponin I. ADVIA Dimension EXL LOCI Module. Siemens 21 Troponin I. ADVIA Dimension EXL LOCI Module. Siemens 83 Troponin T (4th 9 Troponin T (4th 36 Troponin T (4th 36 Troponin I. Dimension (Dade Behring). Siemens 86 Troponin T (4th 2 TnI. ADVIA Dimension EXL LOCI Module. Siemens 16 Troponin T (4th 99th cutoff AMI cutoff AMI, Acute myocardial infarction. Troponin values in ng/l.