Ischemia-Modified Albumin Improves the Usefulness of Standard Cardiac Biomarkers for the Diagnosis of Myocardial Ischemia in the Emergency Department Setting

Saif Anwaruddin, MD; James L. Januzzi Jr, MD; Aaron L. Baggish, MD; Lee Lewandrowski, PhD, MPH; Kent B. Lewandrowski, MD


Am J Clin Pathol. 2005;123(1):140-145. 

In This Article


Of the 200 patients whose data were analyzed in a consecutive manner, 7 patients were excluded from the study because complete data about their clinical course could not be obtained. This left 193 patients for analysis. The majority of this population (172 [89.1%]) was admitted to the hospital following admission to the ED. Of the total population analyzed, 25 (13.0%) were judged as having myocardial ischemia at the time of admission, with 9 patients (4.7%) having acute MI and 16 patients (8.3%) with unstable angina pectoris.

Clinical characteristics of patients based on the presence or absence of myocardial ischemia differed significantly: patients with a diagnosis of ischemia were more likely to be older (mean ± SD age, 70.5 ± 16.9 years vs 65.4 ± 16.9 years; P < .001), to have had a previous diagnosis of coronary artery disease (17/25 [68%] vs 44/168 [26.1%]; P < .001), or to have undergone coronary revascularization (7/25 [28%] vs 17/168 [10.1%]; P = .002). Furthermore, patients with ischemia were more likely to have had previous congestive heart failure (5/25 [20%] vs 10/168 [6.0%]; P = .002). The presence of changes on the ECG suggestive of ischemia was more than twice as common among patients with a final diagnosis of an acute coronary syndrome (4/25 [16%] vs 12/168 [7.1%]; P = .05).

Assay for IMA, Cut Points, and Myocardial Ischemia. At the time of admission to the ED, the results from the Albumin Cobalt Binding Test for the detection of IMA were correlated with the adjudicated diagnosis of myocardial ischemia based on clinical data. Receiver operating characteristic curves demonstrated the assay for IMA to be highly sensitive but somewhat poorly specific for the detection of coronary ischemia (area under the ROC curve, 0.63; P = .01) Figure 1.

Receiver operating characteristic (ROC) analysis for the assay for ischemia-modified albumin. Although highly sensitive, the albumin cobalt assay was variably specific. The area under the ROC curve was 0.63 (P = .01).

With the receiver operating characteristic curve as a guide, exploratory analysis of different cut points was performed Table 1 . With a diagnostic threshold (cut point) of 80 U/mL for the diagnosis of ischemia, the IMA assay was 100% sensitive but demonstrated a low specificity, 20%. The negative predictive value for the assay at this cut point was 100% for excluding the diagnosis of coronary ischemia. In an attempt to optimize sensitivity and specificity, we examined different diagnostic thresholds. At a cut point of 85 U/mL, the assay was highly sensitive; however, the specificity remained low. Again, with this cut point for the diagnosis of ischemia, the negative predictive value for ischemia remained high. At a cut point of 90 U/mL, this assay had a sensitivity of 80% and specificity of 31%. Using 90 U/mL as a diagnostic threshold for ischemia did not affect the negative predictive value, which remained high. Last, a cut point of 100 U/mL had a sensitivity of 64%, a specificity of 66%, and a negative predictive value of 82%.

ECG and IMA Results. Among the 25 patients with ischemia, only 5 (25%) had evidence on the ECG for ischemia (ST segment depression or elevation or a new left bundle branch block). Among these patients, the IMA result was negative in 1 patient (using a cut point of 90 U/mL). Among the 20 patients with ischemia but negative findings on the ECG, an IMA result of more than 90 U/mL correctly identified 16 of 20.

Markers of Myonecrosis, With or Without IMA. At the time of admission to the ED, the combination of myoglobin, CK-MB, and TnI had a sensitivity of 57% for the diagnosis of myocardial ischemia in the same population Figure 2, with the majority of patients having elevated levels of 2 or 3 markers.

Venn diagram detailing patterns of myoglobin (Myo), creatine kinase MB (CK-MB), and troponin I (TnI) results among patients with an acute coronary syndrome. Among the 25 patients with an acute coronary syndrome, 8 had normal results for the triple screen of Myo/CK-MB/TnI; all but one of these patients had a positive ischemia-modified albumin test. Numbers of patients represented in the diagram are as follows: Myo/TnI+ and CK-MB/TnI, 1 each; Myo alone+, 5; Myo/CK-MB+, 3; all 3 assays+, 7.

By comparing myoglobin results with those for CK-MB, we found that myoglobin was elevated in 16 (64%) of patients with an acute coronary syndrome, while 11 (44%) had positive CK-MB results on admission to the ED. Furthermore, considering patients with 2 markers positive, myoglobin was elevated in 4 (16%) of patients, while CK-MB was elevated in 4 (16%). Last, among patients in whom only 1 marker was elevated on admission to the ED, myoglobin alone was elevated in 5 (20%) of patients subsequently given a diagnosis of acute coronary syndrome, in comparison with isolated CK-MB elevation, which was found in none of the patients with an ultimate diagnosis of acute coronary syndrome (P = .001).

When results of the IMA assay (using a diagnostic threshold for ischemia of 90 U/mL) were added to the panel of myoglobin, CK-MB, and TnI, the sensitivity for the diagnosis of myocardial ischemia of this combination was increased to 97%, with a negative predictive value of 92%. Patterns of biomarker results, incorporating IMA data, are detailed in Table 2 . It is interesting that among patients with ischemia in whom IMA was accompanied by only 1 other marker of myonecrosis, myoglobin was elevated universally, suggesting these patients were early in the course following the onset of acute myocardial ischemia.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.