The term “acute coronary syndrome” (ACS) is applied to a broad spectrum of clinical presentations of chest pain, thus rendering accurate risk stratification crucial to rational decision making. Several clinical characteristics known at presentation have been shown to be particularly efficacious in predicting in-hospital risk, including electrocardiographic changes and post-infarct angina.1–3 Additionally, numerous recent studies have shown that troponin levels effectively differentiate between low and high risk without correction for other clinical parameters.4–10 The new ACC/AHA guidelines for unstable angina and non-ST elevation myocardial infarction (MI) recommend using both clinical criteria and cardiac troponin values to stratify all patients at presentation.11,12 This is especially important when evaluating the broad scope of patients at an intermediate level of risk. However, the utility of this approach has not been evaluated. One important question raised by Braunwald11 is whether models of risk based on both clinical variables and serum markers are more predictive than either separately among all risk categories; in other words, would combining troponin (Trp) levels with clinical risk models have supplemental predictive value over the whole spectrum of clinical risk? Recent studies from TIMI26 and PURSUIT27 suggest that there is supplemental value across all categories. However, it is possible that serum markers might be especially useful in better distinguishing risk within the high-risk categories; conversely, even though Trp is very useful when applied to a broad population, it may not be as discerning for characterizing risk in a low-risk subgroup. While it is logical that any variable assessing risk may have different values in narrowly selected populations, how one defines risk and selects a study population without using these variables, which biases the outcome, is problematic and likely explains why this question has never been assessed. Therefore, the current study evaluated the supplemental value of serial Trp I values when combined with a previously validated clinical model in a specifically selected intermediate- and high-risk patient population. These patients were categorized prospectively as high risk by their physicians (based on referral for coronary angiography within 72 hours of presentation for an acute coronary syndrome) and confirmed by the AHCPR risk assessment criteria. Patients at low risk were specifically excluded from this analysis by only including patients who underwent a coronary angiogram early in their hospital course, prior to stress testing, to test the relative value of the clinical model and Trp in predicting outcome in this subgroup. Methods Population inclusion criteria. The population was a group of 118 consecutive patients admitted over a 23-month period to a single tertiary care institution with high-risk unstable angina or non-Q wave MI who fulfilled all entry criteria. No patients were excluded who were qualified for entry. The patients had to be either intermediate or high risk by AHCPR guidelines. Patients transferred from an outside institution after receiving an invasive procedure were excluded. Additionally, coronary angiograms had to be performed during the first 72 hours of their hospitalization. Less than one-quarter of ACS patients during this time frame fulfilled these criteria (drawn from about 500 patients). Finally, all selected patients had at least three serial Trp I, CPK and CK-MB levels drawn at 0, 6 and 12 hours. Patients who presented with ST-elevation MI were excluded from this study. No patients meeting these inclusion criteria were omitted from analysis. Data collection. Patients were identified daily by a nurse coordinator both by admission log review and chart review. Patients were diagnosed with ACS if they had either: 1) acute ischemic type chest pain, either responsive to nitrates or associated with ST depression or T-wave inversion > 2 mm, occurring at rest or lasting longer than 20 minutes; or 2) progressive or accelerating angina characterized by exertional chest pain increasing in frequency, duration or at decreasing levels of exertion. All clinical information was collected prospectively and included: previous history (including previous cardiac history and procedures, pre-admission medications, and other co-morbid diseases); cardiac risk factors (hypertension, diabetes, hypercholesterolemia, family history and smoking status); presenting clinical symptoms (chest pain categorized by the CCSF class, shortness of breath, nausea/vomiting, diaphoresis and radiating pain); and the presenting ECG prior to any treatment or interventions in the emergency department. Clinical model. The RUSH model,2,3 a clinical score incorporating six clinical criteria (based on a variation of Braunwald’s anginal criteria)13 previously shown to be predictive of outcome in acute coronary syndromes, was evaluated in all patients. Four of the criteria originally designated by Braunwald for unstable angina were demonstrated by multivariate analysis to have independent predictive value for major complications; these included recent MI, calcium- or beta-blocker use pre-admission, admission ST depression, and the need for the use of intravenous nitroglycerin on admission. In addition, age and history of diabetes were also found to be independently predictive of outcome and included in this model. A coefficient of risk was calculated for each of the above six criteria and these were combined in an additive fashion to arrive at a log odds score which could be used to calculate a percentage of risk for complication during the current hospitalization. This clinical score is a continuous measure, and indicates risk; low risk was defined as 15%. This scoring system was recently found to be a more accurate predictor of ischemic events for patients with ACS than the AHCPR guidelines.3,14 More importantly, this system identifies many more of the low-risk patients than previous scoring systems, and correlates strongly with resource utilization.15 Patients were initially evaluated and categorized by AHCPR guidelines14 into intermediate and high risk, as an objective, established measure of their overall risk, as noted in the inclusion criteria; only patients in these two risk categories were included. In this study, clinical risk was retrospectively determined using the RUSH model in the total patient group described above, and in each study subgroup. These values were not directly available to the clinician managing the patient. Troponin I evaluation. Trp I samples were drawn at 6-hour intervals starting at presentation (0, 6 and 12 hours). These were analyzed using the Bayer Assay on the Scintar Analyzer (Bayer Diagnostics), which measures Trp I levels via a chemi-luminescence assay; all tests were performed in the Rush-Presbyterian-St. Luke’s Medical Center clinical laboratory. A value of > 2.0 mg/dl is defined as the abnormal cut-off value in our laboratory. We then evaluated whether an abnormal Trp level added value to the risk score in predicting worse outcomes. In addition, we assessed whether a normal Trp level would improve the accuracy of identifying those at lower risk. Definition of outcome events. All patients were managed for ACS using standard medical strategies, including aspirin, heparin or enoxaparin in all patients, glycoprotein IIb/IIIa inhibitors in 53%, and nitrates, beta-blockers and calcium antagonists in some patients. The decision to proceed with coronary angiography versus stress testing, as well as the timing of angiography, were completely at the discretion of the attending physician. Major clinical cardiac events evaluated in this study were defined by the occurrence of any of these adverse outcomes in the following hierarchical order: 1) death; 2) myocardial infarction after the first 24 hours, defined as a rise in CPK > 200 mg/dl with an MB fraction of > 5% peaking after the first 24 hours following admission, or the development of new Q-waves > 24 hours after admission; 3) further increase in elevated troponin and/or CPK (> 3 times normal) if abnormal at time of admission; 4) recurrent chest pain; 5) new ST- and T-wave changes; and 6) sudden, acute development of congestive heart failure. If a patient had one or more of the above clinical outcomes, they were considered to have experienced an adverse outcome event. All patients were evaluated as the single most significant event if they fulfilled more than one definition. Data collection. Two nurse coordinators prospectively collected all clinical predictors and demographic information from the patient charts within 24 hours of hospital admission. Treatments and major in-hospital outcomes were obtained both by reviewing the patient’s chart and interviewing the patient’s physician on a daily basis. The data were subsequently entered in a relational database (Paradox; Ansa Software, Scott’s Valley, California). Neither of these coordinators was involved with the initial model development or trained in its parameters. Assignments of outcomes were verified in two ways: 1) chart audits in a 10% random sample of all admissions and complete chart review in all study patients; and 2) checking discharge ICD-9 codes obtained from the hospital information system. Lesion morphology: Ambrose criteria. Culprit stenosis morphology was evaluated to assess the high-risk nature of the study group and to correlate outcomes pathophysiologically. A segment was graded as having a culprit lesion if a lesion of >= 50% was visualized in two orthogonal views and was correlated with ECG changes or a regional wall motion abnormality on left ventriculography or echocardiogram. If these conditions were not fully met, a patient was categorized as having no identifiable culprit lesion. Culprit lesion morphology was evaluated in all patients (when identified) by an observer blinded to all clinical variables who reviewed all of the angiograms. Lesion morphology was graded using the Ambrose criteria, which systematically categorizes lesions based on stenosis eccentricity and luminal irregularity. A vessel with a concentric narrowing or a smooth eccentric lesion is considered a stable lesion. A vessel containing an eccentric lesion with an overhanging edge, irregular luminal borders or dual overhanging edges is considered an unstable morphology, implying the presence of ruptured plaque and/or thrombosis. In addition, a total occlusion (TIMI 0 or 1 flow) without bridging collaterals was considered the culprit lesion when ECG and wall motion abnormalities were suggestive. Patients were then grouped into two categories based on the morphology of lesions identified: group 1 comprised stable lesion morphologies and group 2 comprised unstable lesion morphologies. Statistical analysis. Variables are presented as means ± one standard deviation. Univariate comparisons were made using Chi-square tests for categorical variables and two-tailed t-tests for continuous variables. P-values of Adverse events. The patient population was comprised of 118 intermediate- or high-risk ACS patients. Fifty patients (42%) experienced 79 adverse clinical events, including 19.5% of patients who experienced either death or MI. These adverse events were (listed non-inclusively): one death, twenty-three clinically apparent new MIs, fifteen further increases in troponin and CPK above admission levels, nineteen recurrent chest pain episodes, twenty acute ST-T wave changes, and 1 episode of acute pulmonary edema related to ischemia requiring intubation. All of these patients were among those with urgent revascularization, which included 43 percutaneous transluminal coronary angioplasties and 11 coronary artery bypass graft surgeries. The high incidence of adverse events confirms the high-risk nature of this group, demonstrating the validity of the concept behind the study design. Lesion morphology. Angiographically, ninety-eight out of 118 patients had identifiable culprit lesions, while 20 patients had no clearly identifiable culprit lesion. Of the identifiable lesions, fifty-four patients had stable lesion morphologies (group 1) and 44 patients had unstable morphology (group 2), including 40 with eccentric and irregular morphologies and 4 with acute total occlusions. The median time to angiography was within 48 hours of hospital admission. AHCPR guideline analysis. When analyzed by the AHCPR risk and prediction model of death and non-fatal MI in patients with symptoms suggestive of unstable angina, a total of 66 patients (56%) fell in the AHCPR intermediate-risk group and 52 patients (44%) fell in the high-risk group, as required by the inclusion criteria. When the Trp values were compared between the AHCPR intermediate- and high-risk groups, there was no significant difference (p = 0.25). Furthermore, when event rates were compared between these two categories, no significant difference was seen (p = 0.21). Clinical score and Trp. The clinical scores of the entire study population, all of whom had angiography within 72 hours, are shown in Figure 1. The event group had a clinical score of 12.7 ± 12.4% and the no event group had a clinical score of 13.2 ±10.2% (p = 0.64). The magnitude of these RUSH model scores demonstrates that the total group, and each subgroup, are at intermediate to high risk for clinical events. When the troponin positive patients are compared to the troponin negative patients, clinical scores of 14.2 ± 13.2% and 11.7 ± 9.3%, respectively, were observed (p = 0.21). These values slightly underestimated the observed 19.5% rate of death and MI. Nevertheless, the patients fall into a high-intermediate to high-risk group and the groups are indistinguishable on the basis of these clinical parameters. Thus, the clinical score is highly predictive for identifying those at higher risk within a broadly inclusive set of ACS patients. None of the individual components that comprise the clinical score was significantly different in the event versus the no event group (Table 1) and the same holds true when comparing the Trp positive and negative groups as well (Table 2). In the Trp I positive group, a total of 64% had cardiac events (death, MI, recurrent angina, CHF, enzyme elevation and/or ECG changes), while 36% had no events (p Coronary morphology correlations. Patients with positive Trp I had unstable (group 2) lesions in 30/46 (65.2%), whereas patients with negative Trp had unstable lesion morphologies in 14/52 (26.9%) (p Current study. In this high-risk unstable angina/non-Q wave MI population, based on presenting clinical criteria and the need for angiography, the presence of elevated Trp I was found to be a sensitive and specific predictor of adverse in-hospital events, as well as angiographic morphology of the culprit lesion. In addition, a normal Trp I was predictive of a lower risk of clinical events. That the Trp level was effective in further identifying those patients at highest risk confirms Braunwald’s suggestion of using both clinical models and serum markers together in assessing risk. The supplemental value of serum markers to the clinical model seen in this study applies only to the intermediate- and high-risk patients as selected in the protocol. It cannot be extended on the basis of these data to low-risk patients (where it is likely that other relationships exist), or to those at moderate risk stabilized medically and not undergoing angiography
1. Braunwald E, Mark D, Jones R. Unstable angina: Diagnosis and management in clinical practice guideline, number 10. AHCPR Publication No. 94-0602. 1994, U.S. Department of Health and Human Services: Rockville. 2. Calvin JE, Klein LW, VandenBerg BJ, et al. Risk stratification in unstable angina: Prospective validation of the Braunwald classification in 393 consecutive patients. JAMA 1995;273:136‚Äì141. 3. Calvin J, Klein L, Vandenberg G, et al. Validated risk stratification model accurately predicts low risk in patients with unstable angina. J Am Coll Cardiol 2000;36:1803‚Äì1808. 4. Wu A, Abbas S, Green A, et al. Prognostic value of cardiac troponin T in unstable angina pectoris. Am J Cardiol 1995;76:970‚Äì972. 5. Hamm W, Rabkilde J, Gerhardt W, et al. The prognostic value of serum troponin T in unstable angina. N Engl J Med 1992;327:146‚Äì150. 6. Hamm W, Goldmann B, Heeschen C. Emergency room triage of patients with acute chest pain by means of rapid testing for cardiac troponin T or troponin I. N Engl J Med 1992;337:1648‚Äì1653. 7. Zimmerman J, Fromm R, Meyer D, et al. Diagnostic marker cooperative study for the diagnosis of myocardial. Circulation 1999;99:1671‚Äì1677. 8. Lindahl B, Andrent B, Ohisson J, et al., for the FRISC Study Group. Risk stratification in unstable coronary artery disease: Additive value of troponin T determinations and pre-discharge exercise tests. Eur Heart J 1997;18:762‚Äì770. 9. Olatidoye A, Wu A, Feng Y-J, Waters D. Prognostic role of troponin T versus troponin I in unstable angina pectoris for cardiac events with meta-analysis comparing published studies. Am J Cardiol 1998;81:1405‚Äì1410. 10. Heeschen C, Hamm C, Goldmann B, et al. Troponin concentrations for stratification patients with acute coronary syndromes in relation to therapeutic efficacy of tirofiban. Lancet 1999;354:1757‚Äì1762. 11. Hamm C, Braunwald E. A classification of unstable angina revisited. Circulation 2000;102:118‚Äì122. 12. Braunwald E, Antman E, Beasley J, et al. ACC/AHA guidelines for the management of patients with unstable angina and non-ST-segment elevation myocardial infarction. J Am Coll Cardiol 2000;36:970‚Äì1062. 13. Braunwald E. Unstable angina: A classification. Circulation 1989;80:410‚Äì414. 14. U.S. Department of Health and Human Services. National Heart, Lung and Blood Institute. Unstable angina: Diagnosis and management. AHCPR publication No. 94-0602: 1994. 15. Calvin J, Klein L, Vandenberg B, et al. Clinical predictors easily obtained at presentation predict resource utilization in unstable angina. Am Heart J 1998;136:373‚Äì381. 16. Goldman L, Weinberg M, Weisberg M, et al. A computer-derived protocol to aid in the diagnoses of emergency room patient with acute chest pain. N Engl J Med 1982;307:588‚Äì596. 17. Goldman L, Cook E, Brand D, et al. A computer protocol to predict myocardial infarction in emergency department patients with chest pain. N Engl J Med 1988;318:797‚Äì803. 18. Murphy J, Connell P, Hampton J. Predictors of risk in patients with unstable angina admitted to a district general hospital. Br Heart J 1992;67:395‚Äì401. 19. Tierney W, Roth B, Psaty B, et al. Predictors of myocardial infarction in emergency room patients. Crit Care Med 1985;13:526‚Äì531. 20. Iliadis EA, Klein LW, Vandenberg BJ, et al. Clinical practice guidelines in unstable angina improve clinical outcomes by assuring early intensive medical treatment. J Am Coll Cardiol 1999;34:1689‚Äì1695. 21. Class S, Christopherson C, Hursey T, et al. The Agency for Health Care Policy and Research guidelines predict subsequent ischemic events but not extent of disease or culprit lesion morphology in patients with unstable angina. J Am Coll Cardiol 1999;33:337A. 22. Ahmed W, Bittl J, Braunwald E. Relation between clinical presentation and angiographic findings in unstable angina pectoris, and comparison with that in stable angina. Am J Cardiol 1993;72:544‚Äì550. 23. Dangas G, Mehran R, Wallenstein S, et al. Correlation of angiographic morphology and clinical presentation in unstable angina. J Am Coll Cardiol 1997;29:519‚Äì525. 24. Servi SD, Arbustini E, Marsica F, et al. Correlation of angiographic morphology between clinical and morphologic findings in unstable angina. Am J Cardiol 1997;77:128‚Äì132. 25. Calvin JE, Klein LW, Van den Berg E, Parrillo JE. The intermediate coronary care unit admission: A preliminary study. Heart Disease 2001;3:18‚Äì23. 26. Antman EM, Cohen M, Bernink PJLM, et al. The TIMI risk score for unstable angina/non-ST elevation MI: A method for prognostication and therapeutic decision making. JAMA 284;7:835‚Äì842. 27. Boersma E, Pieper KS, Steyerberg EW, et al. Predictors of outcome in patients with acute coronary syndromes without persistent ST-segment elevation: Results from an international trial of 9,461 patients. Circulation 2001;101:2557‚Äì2567.