NonSTSegment Elevation Acute Coronary Syndromes

The history, physical examination, and baseline ECG all can be used to risk-stratify patients with non-ST-segment elevation acute coronary syndromes. Woodlief and colleagues developed a regression model in 1384 patients in the GUSTO-IIa trial that identified age, Killip class, systolic blood pressure, and previous hypertension as significant predictors of 30-d mortality (89). In 393 patients with unstable angina, Calvin and colleagues found previous infarction, lack of b-blocker or calcium-channel blocker therapy, ST-segment depression on the presenting ECG, and diabetes to be predictors of death or acute MI (90).

As described previously, in the absence of ST-segment elevation on the initial ECG, the diagnosis of acute MI vs unstable angina is largely made, in retrospect, on the basis of serial CK-MB testing. However, because even small infarcts as measured by CK-MB sampling confer worse outcomes, and the best outcomes in these patients are likely to be obtained when specific treatments are started early, it is clearly important to identify these groups as soon as possible. The use of the sensitive specific cardiac markers discussed previously may aid in diagnosis, risk stratification, and management of this diverse group of patients as well. Specifically, TnT and Tnl have now been studied extensively as indicators of prognosis in patients with non-ST-segment elevation acute coronary syndromes.

In an enzyme substudy of the Fragmin during Instability in Coronary Artery Disease (FRISC) trial, cTnT was measured at baseline in 976 patients who presented within 12 h of symptom onset (91). At 5 mo there was a correlation between the combined rate of death or MI and the level of TnT measured in the serum at baseline; cTnT < 0.06 ng/mL, 4.3%; 0.06-0.18 ng/mL, 10.5%; >18 ng/mL, 16.1%. The cTnT level, age, hypertension, number of antianginal drugs, and ECG changes were identified in multivariable analysis as the most important independent predictors of risk in this population of unstable angina patients.

The GUSTO-IIa TnT substudy evaluated the use of a single baseline measure of cTnT compared with the baseline ECG and CK-MB as a risk marker in 855 patients across the spectrum of acute coronary syndromes (20). Of the 755 patients who had all three studies at baseline, 36% were cTnT-positive and 32% had elevated CK-MB. As in the FRISC analysis, the probability of short-term mortality correlated with the serum concentration of cTnT at baseline; when the result of the cTnT test was considered as a dichotomous variable, 30-d mortality was 11.8% in the cTnT-positive patients compared with 3.9% in cTnT-negative patients. This relationship of TnT status to outcome held across all ECG categories (ST-segment elevation, ST-segment depression, T-wave inversion/normal, and confounding factors), and the incidence of in-hospital complications was also higher in the cTnT-positive patients. When the result of the baseline ECG, cTnT, and CK-MB were evaluated in an unadjusted mortality model, baseline cTnT had the largest v2 value, followed by the ECG and the CK-MB. However, when the mortality model was adjusted for the presence of the other two variables (which were forced in first), the ECG was the strongest predictor of 30-d mortality (v2 = 11.5, p = 0.009), followed by the baseline cTnT result (v2 = 9.2, p = 0.027). In the adjusted model, the baseline CK-MB added no significant information after the ECG and cTnT results were considered.

TnI also has been evaluated as a risk marker in acute coronary syndrome patients. In a retrospective analysis of serum from 1404 acute coronary syndrome patients enrolled in the TIMI-3 trial, cTnI was positive in 41% of patients (92). The risk of mortality increased with increasing levels of cTnI; when troponin I status (positive > 0.04 ng/mL) was analyzed as a dichotomous variable, mortality was significantly higher in positive than in negative patients (3.7 vs 1.0%). In a multivariable mortality model, ST-segment depression (p < 0.001), age >65 yr (p = 0.026), and the baseline cTnI status (p = 0.03) were independent predictors of mortality.

Newby and colleagues have shown that obtaining serial measures of TnT adds significantly to the result of the baseline measure in determining the risk of both in-hospi-tal and 30-d events in the same GUSTO-IIa TnT substudy population (93). Further, the results of both baseline and serial cTnT testing remained predictive of events in the GUSTO-IIa TnT substudy cohort at 1 yr (mortality 14.2% in baseline cTnT-positive patients vs 5.8% in negative patients; 9.6 vs 5.6% for any positive). Similarly, Stubbs and colleagues showed a significant relationship between baseline cTnT measures and death, combined death/MI, and revascularization at a median follow-up of 3 yr (94). However, when the event rates between 30 d and 1 yr were evaluated, there was no significant difference in mortality over this period for either the baseline or any positive result on serial testing. Based on these results, the increased risk identified by cTnT testing in acute coronary syndrome patients appears to be for events that occur early, suggesting that the next step is to identify treatment strategies that can favorably alter this risk when applied early.

Analyses of Tn testing in populations of unstable angina patients enrolled in several large clinical trials suggest that Tn measurement shortly after presentation may be useful to define subgroups of patients who would benefit most from early medical or percutaneous intervention strategies. Such use of Tn testing may improve clinical outcome as well as facilitate cost-effective use of expensive medical and interventional therapies.

The FRISC study of a low molecular weight heparin treatment strategy suggested that Tn-positive subgroups achieved greater benefit from treatment than Tn-negative patients. In the FRISC study, cTnT-positive (>0.1 ng/mL) patients had greater reduction in the 40-d incidence of death or MI (14.2 vs 7.4%) with long-term administration of dalteparin than did those who were cTnT-negative (<0.1 ng/mL) where there was no difference with or without treatment (95). A similar analysis in 1265 unstable angina patients receiving percutaneous intervention who were randomized to treatment with abciximab or placebo in the Chimeric c7E3 Antiplatelet Therapy in Unstable Angina Refractory to Standard Treatment (CAPTURE) trial suggested that cTnT analysis might be used to identify a subgroup of patients who would realize the most benefit from abciximab treatment (96). This evaluation showed a 60% reduction in death or nonfatal MI at 6 mo among cTnT-positive patients treated with abciximab. However, among patients who were cTnT-negative, there was no benefit of abciximab treatment relative to placebo.

Similar results have been reported by both the Platelet Receptor Inhibition in Ischemic Syndrome Management (PRISM) and the PARAGON-B TnT Substudy investigators with 2 small-molecule glycoprotein IIb/IIIa inhibitors, tirofiban and lamifiban, respectively (97,98). In PRISM, 2222 patients were randomized to treatment with tirofiban plus aspirin vs heparin plus aspirin had Tn levels determined at a mean of 8.4 h after symptom onset. Among patients who were cTnI-positive, the rate of death or MI was 4.3% among tirofiban-treated patients compared with 13.0% among those receiving standard therapy with heparin and aspirin. Among cTnI-negative patients, there was no difference in the occurrence of death or MI by treatment group. Results were similar for TnT, and importantly, this differential treatment effect was observed for both medically treated patients and those treated with percutaneous coronary revascularization. In the PARAGON B TnT substudy, the 30-d rate of death or non-fatal MI was 19.0% in cTnT-positive patients who received placebo compared with 11.0% in those who received lamifiban. The corresponding results among cTnT-negative patients were 10.3% for placebo and 9.6% for lamifiban, suggesting that the majority of the overall reduction of events in lamifiban-treated patients (9.9%) vs placebo (9.9%) occurred among cTnT-positive patients. A systematic overview of these results is displayed in Fig. 3.

Fig. 3. Systematic overview of treatment effect by Tn status. Shown are odds ratios with 95% confidence intervals for death or MI among Tn-negative and Tn-positive patients and for interaction of Tn status with treatment effect for PRISM, CAPTURE, PARAGON-B, and combined trials. Values to left of 1.0 indicate a benefit of GP IIb/IIIa inhibition. Reprinted with permission from ref. 98.

Fig. 3. Systematic overview of treatment effect by Tn status. Shown are odds ratios with 95% confidence intervals for death or MI among Tn-negative and Tn-positive patients and for interaction of Tn status with treatment effect for PRISM, CAPTURE, PARAGON-B, and combined trials. Values to left of 1.0 indicate a benefit of GP IIb/IIIa inhibition. Reprinted with permission from ref. 98.

Comparisons of TnT with TnI as risk markers have been attempted. However, the lack of standardized assays and use of different cut-off values for the same assay in different studies make the results difficult to interpret. Luscher and colleagues compared cTnT with cTnI by the Sanofi assay in 491 patients with unstable angina and found them to identify similar groups of patients at high risk for 30-d death or MI (99).

Using the Dade Stratus II cTnI assay at a cut point of 1.5 ng/mL in comparison to cTnT in the GUSTO-IIa TnT substudy cohort, both markers when measured at baseline predicted 30-d mortality, but cTnT provided the most prognostic information. In a mortality model with the ECG and the other marker forced in first, only cTnT still provided additional prognostic information (62). Further, the area under the receiver operator characteristics (ROC) curve, which is independent of the cut point used for an assay, was larger for cTnT compared to cTnI (0.68 vs 0.64).

Serial bedside tests for cTnT and cTnI at baseline and 4 h were evaluated by Hamm and colleagues with similar performance (sensitivity and negative predictive value) by both markers. In a logistic regression model, both markers were strong predictors of 30-d death or MI and remained so even after ST-segment depression on the initial ECG was forced into the model first (45). The individual chi-square for cTnI in this model was larger, however.

The results of these studies suggest that the prognostic information demonstrated in individual studies of cTnT or cTnI should not be generalized to other assays or testing conditions. Differences in analytical characteristics and measurement precision for the different cTnI assays may translate into important differences in their ability to risk-stratify the acute coronary syndrome patient. Differences in the characteristics of the marker proteins themselves, including the timing and magnitude of release after both myocardial necrosis and injury, may also translate into the differences in prognostic ability seen in individual studies performed under different testing circumstances and in different patient populations. In an effort to update and standardize the diagnosis of MI and the use of cardiac marker assays, the European Society of Cardiology/American College of Cardiology Joint Committee for the Tn testing, but also on assay precision and standardization (100). They have recommended that each assay should be defined as positive at values above the 99th percentile of normal controls and also that acceptable imprecision should be reflected by a coefficient of variation < 10%. Further study with standardized assays under well-defined clinical circumstances will be needed to characterize whether there are any important differences between these markers in diagnosis and risk stratification.

Your Heart and Nutrition

Your Heart and Nutrition

Prevention is better than a cure. Learn how to cherish your heart by taking the necessary means to keep it pumping healthily and steadily through your life.

Get My Free Ebook

Post a comment