- Email: [email protected]
Application of albumin-adjusted ischemia modified albumin index as an early screening marker for acute coronary syndrome
Clinica Chimica Acta 384 (2007) 24 – 27 www.elsevier.com/locate/clinchim
Application of albumin-adjusted ischemia modified albumin index as an early screening marker for acute coronary syndrome Yong-Wha Lee a , Ho-Jung Kim b , Yoon-Haeng Cho c , Hee Bong Shin a , Tae-Youn Choi d , You Kyoung Lee a,⁎ a
Department of Laboratory Medicine, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, 1174 Jung-dong, Wonmi-gu, Bucheon 420-767, Republic of Korea b Department of Emergency Medicine, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, Bucheon, Republic of Korea c Department of Cardiology, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, Bucheon, Republic of Korea d Department of Laboratory Medicine, Soonchunhyang University Hospital and Soonchunhyang University College of Medicine, Seoul, Republic of Korea Received 19 January 2007; received in revised form 3 May 2007; accepted 3 May 2007 Available online 18 May 2007
Abstract Background: Existing cardiac markers are not sensitive for reversible myocardial ischemia. Ischemia modified albumin (IMA) has recently been shown to be an early and sensitive marker of myocardial ischemia. We established a newly standardized, albumin-adjusted IMA index that was more sensitive and accurate than the conventional IMA value. Methods: We enrolled 413 consecutive patients with symptoms suggestive of acute coronary syndrome (ACS). All patients were classified to either the ACS group (n = 129) or 4 other groups (n = 284). The ideal cutoff value of IMAwas calculated by the receiver operating characteristic (ROC) curve analysis. The albumin-adjusted IMA index was calculated from the results of correlation assay between serum albumin concentration and IMAvalue and re-applied. Results: The sensitivity and specificity of IMA for ACS were 93.0% and 35.6%, respectively, at 85.0 U/ml. IMA had a negative linear relationship with serum albumin and albumin-adjusted IMA index was calculated by using the following equation [IMA index = serum albumin conc. (g/ dl) × 23 + IMA (U/ml) − 100]. The sensitivity and specificity were 98.4% and 34.5%, respectively, at IMA index of 83.4. Conclusions: The use of the calculated albumin-adjusted IMA index is recommended to increase the sensitivity of the ACS diagnosis although IMA is a sensitive marker for the identification of ACS. © 2007 Elsevier B.V. All rights reserved. Keywords: Ischemia modified albumin; Albumin-adjusted IMA index; Acute coronary syndrome
1. Introduction For the diagnosis of acute coronary syndrome (ACS) among patients having chest pain, clinical manifestations, electrocardiogram (ECG), and cardiac marker test are performed [1,2]. Nevertheless, only approximately 30% of ACS patients manifest typical chest pain, and ECG is non-specific and normal finding can be seen in cardiac marker test in many of these cases. Myoglobin, creatine kinase-MB (CK-MB) and troponin have been used frequently as cardiac markers that are detectable in blood for conditions progressing to irreversible myocardial ⁎ Corresponding author. Tel.: +82 32 621 5941; fax: +82 32 621 5944. E-mail address: [email protected] (Y.K. Lee). 0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2007.05.003
necrosis, but they are limited by their insufficient sensitivity for reflecting the reversible ischemic condition in the initial onset period [3,4]. The recently introduced, ischemia-modified albumin (IMA) has been reported to be a marker capable of reflecting myocardial ischemic condition [5–7]. The amino terminal of plasma albumin normally has an affinity to cobalt, copper, and other heavy metal ions, and the N-terminal of albumin is modified during exposure to ischemic condition, due to the generation of oxidative free radicals and reactive oxygen species, resulting in the generation of IMA with a low binding affinity to heavy metals. Reports that IMA is increased within a few minutes after the onset of myocardial ischemia and continues to increase for 6–12 h suggest that it could be applied
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
effectively to the detection of myocardial ischemic condition prior to the progression to myocardial necrosis [8,9]. Therefore, we measured IMA by applying the albumin cobalt binding (ACB) by Inverness Medical Innovations (Waltham, MA) to evaluate its usefulness as an early screening marker of ACS in patients having chest pain. In this indirect measurement method, a constant amount of cobalt is allowed to react with the patient serum and thus bind to albumin in the serum, and the remaining non-bound cobalt is measured. Therefore, it was anticipated that IMA would be influenced by the albumin concentration in the patients' blood, and that its clinical usefulness as a marker of ACS could be evaluated by designating the value that the measured IMA value of the patients was adjusted by the serum albumin concentration of the patients. 2. Materials and methods 2.1. Subjects The study subjects were 413 cases who underwent IMA test (208 males, 205 females) and were selected among the adults who visited the emergency room for symptoms suspicious of ACS such as chest pain, epigastric pain, dyspnea, syncope, and arrhythmia from June 2005 to May 2006. The diagnosis was made by emergency medicine specialists and cardiologists and was based on the combination of clinical manifestation, ECG, cardiac markers, coronary angiography, and echocardiography. The IMA results were not available at the time of diagnosis. Patients were divided into 5 groups according to their diagnosis: ACS group including unstable angina (UA) and acute myocardial infarction (AMI) (group 1; n = 129), coronary artery diseases (CAD) excluding ACS (group 2; n = 91), cardiac diseases excluding CAD (group 3; n = 78), ischemia in areas other than the heart (group 4; n = 20), and other diseases (group 5; n = 95). Groups 2–5 were combined as the non-ACS groups.
2.2. Analytical characteristics of IMA Based on the established cut-off value of IMA, sensitivity and specificity were obtained. The ability to screen ACS was compared using the conventional cardiac marker panel (myoglobin, CK-MB, troponin-T and ECG) and the test panel which also included IMA.
25
screening ACS was established. SPSS 11.5 (SPSS Inc., Chicago, IL) was used for all statistical analyses and P b 0.05 was considered to be statistically significant.
3. Results 3.1. Clinical usefulness of IMA The IMA mean value ± SD of the ACS group, 101.7 ± 17.3 U/ ml, was significantly higher than the 89.5 ± 13.4 U/ml of the nonACS groups (P b 0.05). The significant differences were detected between group 5 and other non-ACS groups (Fig. 1). In ROC analysis, the area under the curve (AUC) was 0.76. For an optimum IMA cutoff of 85.0 U/ml, sensitivity and specificity were 93.0% (120/129) and 35.6% (101/284), respectively, and the negative and positive predictive values were 91.8% (101/110) and 39.6% (120/303), respectively (Fig. 2A). The sensitivity and specificity of the conventional cardiac marker panel (myoglobin, CK-MB and troponin T) for the diagnosis of ACS were 80.2% (73/91) and 57.0% (111/195), respectively, and were 94.5% (86/91) and 45.1% (88/195), respectively, with the IMA test included in the conventional panel. 3.2. Establishment of the albumin-adjusted IMA index In the experiments where serum albumin was diluted by 5 steps, IMA was significantly negatively correlated with serum albumin concentration (R2 = 0.99) (Fig. 3), according to the following equation: YIMA = − 23.1Xalbumin + 200. Further recalculation of the albumin-adjusted IMA index through our modification derived the formula: [albumin-adjusted IMA index = serum albumin concentration (g/dl) × 23 + IMA (U/ml)− 100]. In the ROC analysis of the applied albumin-adjusted IMA marker, AUC was 0.80, and for the ACS diagnosis with an applied cutoff of 83.4 the sensitivity and specificity were 98.4% (127/129) and 34.5% (98/284), respectively, and the negative
2.3. Effect of serum albumin on IMA measurement To examine whether IMA result varies according to serum albumin concentration the pooled sera obtained from healthy individuals were diluted in 5 steps using saline. Serum albumin and IMA were measured 4 times repeatedly (verification range: 0–4.9 g/dl). The albumin-adjusted marker was prepared by regression modification between serum albumin concentration and IMA measurement, and its clinical usefulness was re-evaluated.
2.4. Analytical methods Blood collected in a serum separation tube was transported immediately to the department of laboratory medicine, and serum was prepared for the cardiac marker tests, including the IMA test. IMA was measured using ACB test reagents by Cobas Integra 800 (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions. The cut-off suggested by the manufacturer was 85 U/ml. Troponin T, myoglobin and CK-MB were measured by electrochemiluminescence immunoassay using Modular E170 (Roche Diagnostics, Mannheim, Germany).
2.5. Statistics To examine whether IMA result differed significantly between the ACS group and the four non-ACS groups, ANOVA was performed. By performing receiver operating characteristics (ROC) analysis, the optimum cutoff for
Fig. 1. Distribution of IMA results for each group (group 1: ACS, group 2: coronary artery disease without ACS, group 3: cardiac disease without coronary artery disease, group 4: non-cardiac ischemic lesions, group 5: other diseases). Each bar presented as mean ± SD.
26
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
Fig. 3. Distribution of IMA levels according to serum albumin concentration. Negative linear relationship between IMA level and serum albumin concentration is observed.
Fig. 2. ROC curves for IMA (A) and albumin-adjusted IMA index (B) in a diagnosis of acute coronary syndrome.
and positive predictive values were 98.0% (98/100) and 40.6% (127/313), respectively (Fig. 2B). With the albumin-adjusted IMA marker included in the conventional panel, the sensitivity and specificity for the diagnosis of ACS were 98.9% (90/91) and 36.9% (72/195), respectively (Table 1). 4. Discussion IMA was found to be sufficiently sensitive as the first screening marker capable of determining the patients admitted for chest pain and ACS suspicious symptoms requiring whether intensive care or emergency treatment. Furthermore, the sensitivity was additionally improved by adjusting the IMA measurement with the serum albumin of the patient. For the screening of ACS, the sensitivity and negative predictive value of IMA were higher than those of the conventional cardiac markers, myoglobin, CK-MB and troponin T, and its specificity was lower, which were similar to the results of previous studies (sensitivity, 80–90%; negative prediction rate, 85–92%; specificity, 31–49%) [5–7]. The increase of IMA sensitivity
could be explained by its theoretical increase under ischemic condition prior to the development of tissue necrosis. In our study, among 129 cases confirmed as ACS using follow-up test, 13 cases were negative for the conventional cardiac marker test but were nevertheless positive for IMA increase. The specificity and positive predictive value of IMA were very low, because IMA reflects not only cardiac ischemia but also various ischemic conditions of the brain, G–I tract or other organs, as well as various diseases. In our study, IMA was increased in a few patients only with liver diseases or muscle diseases without other causalities and IMA levels had no significance among non-ACS groups, suggesting that IMA elevation could be influenced by various factors and reflects the in vivo condition dynamically [10]. The established IMA cutoff differed depending on institutions, measurement methods, and assay purpose. Higher cutoff levels of 98.5 U/ml and 110 U/ml were used in some reports [7,11]. To screen effectively for ACS, which has to be ruled out in chest pain patients, we established 85.0 U/ml as the most ideal IMA cutoff based on ROC analysis. At the manufacturer's recommended cutoff of 90 U/ml, sensitivity was decreased to Table 1 Comparisons of diagnostic characteristics for the diagnosis of acute coronary syndrome (n = 413) Markers
Sensitivity, %
Specificity, %
NPV, %
PPV, %
Triple cardiac markers a IMA b Triple cardiac markers and IMA Albumin-adjusted IMA index c Triple cardiac markers and albuminadjusted IMA index
80.2 93.0 94.5 98.4 98.9
57.0 35.6 45.1 34.5 36.9
86.0 91.8 94.6 98.0 98.6
46.5 39.6 44.6 40.6 42.3
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; IMA, ischemia modified albumin. a Triple cardiac markers: myoglobin + creatine kinase-MB+ cardiac troponin T. b IMA cutoff: 85.0 U/ml. c Albumin-adjusted IMA index cutoff: 83.4.
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
78.3% and specificity to 59.4%. However, compared with the screening ability of conventional cardiac markers, its differentiating power when used alone remained insufficient. The initial rule-out of ACS from diagnosis was beyond the basic purpose of the IMA test. As the sensitivity and negative predictive value of IMA were superior to the conventional cardiac markers while its specificity was lower, the usefulness its combination with the conventional cardiac panel was evaluated. The inclusion of IMA in the conventional panel increased the sensitivity to 94.5%. This confirmed that the IMA test should always be performed in parallel with the conventional cardiac panel to compensate for the false negative IMA result for ACS and to distinguish between cardiac ischemia and necrosis. We considered IMA to be an emergency test. Therefore, to minimize the preparation time needed to conduct the test separately from the analysis equipment used for the conventional cardiac marker tests, the IMA test was performed on the Integra 800 simultaneously with the conventional cardiac marker tests. The results were available approximately 40 min after receipt of the samples, with a further 15 min required for analysis, for a total of 1 h. The IMA test follows the indirect measurement principle, after the reaction of a constant amount of cobalt with albumin in the serum of patient, to measure the cobalt unbound to the Nterminal of albumin. Therefore, it is anticipated that the amount of cobalt able to bind to albumin would be low at low serum albumin concentration, and the albumin-adjusted marker was thus designed. This design differed from the reagent manufacturer's report that IMA measurement may vary depending on the value of serum albumin; nonetheless, it was explained to be a statistically insignificant change [12]. However, the experimental, 5-step dilution of serum albumin validated the presence of a linear inverse correlation, and IMA was decreased by 23 U/ ml with a serum albumin increase of 1 g/dl. In the index calculation formula [albumin-adjusted IMA marker = serum albumin concentration (g/dl) × 23 + IMA (U/ml) − 100], ‘− 100’ was designated as the value able to show the closest value to the absolute IMA measurement. Serum albumin concentration may show great variance among individuals and even within an individual due to various reasons, and thus it was thought that serum albumin concentration should be included in the interpretation of IMA measurement. The result of the re-application of the albuminadjusted IMA index to our data showed that in a comparison of the screening ability of IMA measurement, the sensitivity and the negative and positive predictive values were increased at the cutoff of 83.4, while the specificity was slightly decreased. However, when combined with the conventional test panel, the
27
sensitivity was increased to 98.9%. The IMA result could vary according to the concentration of serum albumin, which suggests the serum albumin measurement should be combined with the IMA measurement, and that the albumin-adjustment index should be established in each laboratory. In conclusion, the IMA test was confirmed to have superior sensitivity and negative predictive value for the early screening of ACS in patients admitted for chest pain. However, its diagnostic specificity was low, indicating that the IMA test should be performed in parallel with the conventional tests for the mandatory exclusion of ACS. It is suggested that the serum albumin-adjusted IMA index should be applied to ACS screening. References [1] Lee TH, Cook EF, Weisberg M, Sargent RK, Wilson C, Goldman L. Acute chest pain in the emergency room. Identification and examination of lowrisk patients. Arch Intern Med 1985;145:65–9. [2] Rouan GW, Lee TH, Cook EF, Brand DA, Weisberg MC, Goldman L. Clinical characteristics and outcome of acute myocardial infarction in patients with initially normal or nonspecific electrocardiograms (a report from the Multicenter Chest Pain Study). Am J Cardiol 1989;64:1087–92. [3] Brogan Jr GX, Hollander JE, McCuskey CF, et al. Evaluation of a new assay for cardiac troponin I vs. creatine kinase-MB for the diagnosis of acute myocardial infarction. Biochemical Markers for Acute Myocardial Ischemia (BAMI) Study Group. Acad Emerg Med 1997;4:6–12. [4] Gomez MA, Anderson JL, Karagounis LA, Muhlestein JB, Mooers FB. An emergency department-based protocol for rapidly ruling out myocardial ischemia reduces hospital time and expense: results of a randomized study (ROMIO). J Am Coll Cardiol 1996;28:25–33. [5] Anwaruddin S, Januzzi Jr JL, Baggish AL, Lewandrowski EL, Lewandrowski KB. Ischemia-modified albumin improves the usefulness of standard cardiac biomarkers for the diagnosis of myocardial ischemia in the emergency department setting. Am J Clin Pathol 2005;123:140–5. [6] Roy D, Quiles J, Aldama G, et al. Ischemia modified albumin for the assessment of patients presenting to the emergency department with acute chest pain but normal or non-diagnostic 12-lead electrocardiograms and negative cardiac troponin T. Int J Cardiol 2004;97:297–301. [7] Kang SY, Suh JT, Lee WI. Clinical usefulness of ischemia modified albumin in acute coronary syndrome. Korean J Lab Med 2005;25:306–11. [8] Gidenne S, Ceppa F, Fontan E, Perrier F, Burnat P. Analytical performance of the Albumin Cobalt Binding (ACB) test on the Cobas MIRA Plus analyzer. Clin Chem Lab Med 2004;42:455–61. [9] Christenson RH, Duh SH, Sanhai WR, et al. Characteristics of an Albumin Cobalt Binding Test for assessment of acute coronary syndrome patients: a multicenter study. Clin Chem 2001;47:464–70. [10] Zapico-Muniz E, Santalo-Bel M, Merce-Muntanola J, Montiel JA, Martinez-Rubio A, Ordonez-Llanos J. Ischemia-modified albumin during skeletal muscle ischemia. Clin Chem 2004;50:1063–5. [11] Kim JY, Yoon J, Jung IH, et al. A diagnostic value of ischemia modified albumin in patients with suspected acute coronary syndrome with normal EKG and cardiac markers. Korean Circ J 2005;35:928–33. [12] Internal data from 510k premarket notification submission to USFDA. FDA Ref No. K023824;2003 [Feb 14].
Application of albumin-adjusted ischemia modified albumin index as an early screening marker for acute coronary syndrome Yong-Wha Lee a , Ho-Jung Kim b , Yoon-Haeng Cho c , Hee Bong Shin a , Tae-Youn Choi d , You Kyoung Lee a,⁎ a
Department of Laboratory Medicine, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, 1174 Jung-dong, Wonmi-gu, Bucheon 420-767, Republic of Korea b Department of Emergency Medicine, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, Bucheon, Republic of Korea c Department of Cardiology, Soonchunhyang University Bucheon Hospital and Soonchunhyang University College of Medicine, Bucheon, Republic of Korea d Department of Laboratory Medicine, Soonchunhyang University Hospital and Soonchunhyang University College of Medicine, Seoul, Republic of Korea Received 19 January 2007; received in revised form 3 May 2007; accepted 3 May 2007 Available online 18 May 2007
Abstract Background: Existing cardiac markers are not sensitive for reversible myocardial ischemia. Ischemia modified albumin (IMA) has recently been shown to be an early and sensitive marker of myocardial ischemia. We established a newly standardized, albumin-adjusted IMA index that was more sensitive and accurate than the conventional IMA value. Methods: We enrolled 413 consecutive patients with symptoms suggestive of acute coronary syndrome (ACS). All patients were classified to either the ACS group (n = 129) or 4 other groups (n = 284). The ideal cutoff value of IMAwas calculated by the receiver operating characteristic (ROC) curve analysis. The albumin-adjusted IMA index was calculated from the results of correlation assay between serum albumin concentration and IMAvalue and re-applied. Results: The sensitivity and specificity of IMA for ACS were 93.0% and 35.6%, respectively, at 85.0 U/ml. IMA had a negative linear relationship with serum albumin and albumin-adjusted IMA index was calculated by using the following equation [IMA index = serum albumin conc. (g/ dl) × 23 + IMA (U/ml) − 100]. The sensitivity and specificity were 98.4% and 34.5%, respectively, at IMA index of 83.4. Conclusions: The use of the calculated albumin-adjusted IMA index is recommended to increase the sensitivity of the ACS diagnosis although IMA is a sensitive marker for the identification of ACS. © 2007 Elsevier B.V. All rights reserved. Keywords: Ischemia modified albumin; Albumin-adjusted IMA index; Acute coronary syndrome
1. Introduction For the diagnosis of acute coronary syndrome (ACS) among patients having chest pain, clinical manifestations, electrocardiogram (ECG), and cardiac marker test are performed [1,2]. Nevertheless, only approximately 30% of ACS patients manifest typical chest pain, and ECG is non-specific and normal finding can be seen in cardiac marker test in many of these cases. Myoglobin, creatine kinase-MB (CK-MB) and troponin have been used frequently as cardiac markers that are detectable in blood for conditions progressing to irreversible myocardial ⁎ Corresponding author. Tel.: +82 32 621 5941; fax: +82 32 621 5944. E-mail address: [email protected] (Y.K. Lee). 0009-8981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.cca.2007.05.003
necrosis, but they are limited by their insufficient sensitivity for reflecting the reversible ischemic condition in the initial onset period [3,4]. The recently introduced, ischemia-modified albumin (IMA) has been reported to be a marker capable of reflecting myocardial ischemic condition [5–7]. The amino terminal of plasma albumin normally has an affinity to cobalt, copper, and other heavy metal ions, and the N-terminal of albumin is modified during exposure to ischemic condition, due to the generation of oxidative free radicals and reactive oxygen species, resulting in the generation of IMA with a low binding affinity to heavy metals. Reports that IMA is increased within a few minutes after the onset of myocardial ischemia and continues to increase for 6–12 h suggest that it could be applied
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
effectively to the detection of myocardial ischemic condition prior to the progression to myocardial necrosis [8,9]. Therefore, we measured IMA by applying the albumin cobalt binding (ACB) by Inverness Medical Innovations (Waltham, MA) to evaluate its usefulness as an early screening marker of ACS in patients having chest pain. In this indirect measurement method, a constant amount of cobalt is allowed to react with the patient serum and thus bind to albumin in the serum, and the remaining non-bound cobalt is measured. Therefore, it was anticipated that IMA would be influenced by the albumin concentration in the patients' blood, and that its clinical usefulness as a marker of ACS could be evaluated by designating the value that the measured IMA value of the patients was adjusted by the serum albumin concentration of the patients. 2. Materials and methods 2.1. Subjects The study subjects were 413 cases who underwent IMA test (208 males, 205 females) and were selected among the adults who visited the emergency room for symptoms suspicious of ACS such as chest pain, epigastric pain, dyspnea, syncope, and arrhythmia from June 2005 to May 2006. The diagnosis was made by emergency medicine specialists and cardiologists and was based on the combination of clinical manifestation, ECG, cardiac markers, coronary angiography, and echocardiography. The IMA results were not available at the time of diagnosis. Patients were divided into 5 groups according to their diagnosis: ACS group including unstable angina (UA) and acute myocardial infarction (AMI) (group 1; n = 129), coronary artery diseases (CAD) excluding ACS (group 2; n = 91), cardiac diseases excluding CAD (group 3; n = 78), ischemia in areas other than the heart (group 4; n = 20), and other diseases (group 5; n = 95). Groups 2–5 were combined as the non-ACS groups.
2.2. Analytical characteristics of IMA Based on the established cut-off value of IMA, sensitivity and specificity were obtained. The ability to screen ACS was compared using the conventional cardiac marker panel (myoglobin, CK-MB, troponin-T and ECG) and the test panel which also included IMA.
25
screening ACS was established. SPSS 11.5 (SPSS Inc., Chicago, IL) was used for all statistical analyses and P b 0.05 was considered to be statistically significant.
3. Results 3.1. Clinical usefulness of IMA The IMA mean value ± SD of the ACS group, 101.7 ± 17.3 U/ ml, was significantly higher than the 89.5 ± 13.4 U/ml of the nonACS groups (P b 0.05). The significant differences were detected between group 5 and other non-ACS groups (Fig. 1). In ROC analysis, the area under the curve (AUC) was 0.76. For an optimum IMA cutoff of 85.0 U/ml, sensitivity and specificity were 93.0% (120/129) and 35.6% (101/284), respectively, and the negative and positive predictive values were 91.8% (101/110) and 39.6% (120/303), respectively (Fig. 2A). The sensitivity and specificity of the conventional cardiac marker panel (myoglobin, CK-MB and troponin T) for the diagnosis of ACS were 80.2% (73/91) and 57.0% (111/195), respectively, and were 94.5% (86/91) and 45.1% (88/195), respectively, with the IMA test included in the conventional panel. 3.2. Establishment of the albumin-adjusted IMA index In the experiments where serum albumin was diluted by 5 steps, IMA was significantly negatively correlated with serum albumin concentration (R2 = 0.99) (Fig. 3), according to the following equation: YIMA = − 23.1Xalbumin + 200. Further recalculation of the albumin-adjusted IMA index through our modification derived the formula: [albumin-adjusted IMA index = serum albumin concentration (g/dl) × 23 + IMA (U/ml)− 100]. In the ROC analysis of the applied albumin-adjusted IMA marker, AUC was 0.80, and for the ACS diagnosis with an applied cutoff of 83.4 the sensitivity and specificity were 98.4% (127/129) and 34.5% (98/284), respectively, and the negative
2.3. Effect of serum albumin on IMA measurement To examine whether IMA result varies according to serum albumin concentration the pooled sera obtained from healthy individuals were diluted in 5 steps using saline. Serum albumin and IMA were measured 4 times repeatedly (verification range: 0–4.9 g/dl). The albumin-adjusted marker was prepared by regression modification between serum albumin concentration and IMA measurement, and its clinical usefulness was re-evaluated.
2.4. Analytical methods Blood collected in a serum separation tube was transported immediately to the department of laboratory medicine, and serum was prepared for the cardiac marker tests, including the IMA test. IMA was measured using ACB test reagents by Cobas Integra 800 (Roche Diagnostics, Mannheim, Germany) according to the manufacturer's instructions. The cut-off suggested by the manufacturer was 85 U/ml. Troponin T, myoglobin and CK-MB were measured by electrochemiluminescence immunoassay using Modular E170 (Roche Diagnostics, Mannheim, Germany).
2.5. Statistics To examine whether IMA result differed significantly between the ACS group and the four non-ACS groups, ANOVA was performed. By performing receiver operating characteristics (ROC) analysis, the optimum cutoff for
Fig. 1. Distribution of IMA results for each group (group 1: ACS, group 2: coronary artery disease without ACS, group 3: cardiac disease without coronary artery disease, group 4: non-cardiac ischemic lesions, group 5: other diseases). Each bar presented as mean ± SD.
26
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
Fig. 3. Distribution of IMA levels according to serum albumin concentration. Negative linear relationship between IMA level and serum albumin concentration is observed.
Fig. 2. ROC curves for IMA (A) and albumin-adjusted IMA index (B) in a diagnosis of acute coronary syndrome.
and positive predictive values were 98.0% (98/100) and 40.6% (127/313), respectively (Fig. 2B). With the albumin-adjusted IMA marker included in the conventional panel, the sensitivity and specificity for the diagnosis of ACS were 98.9% (90/91) and 36.9% (72/195), respectively (Table 1). 4. Discussion IMA was found to be sufficiently sensitive as the first screening marker capable of determining the patients admitted for chest pain and ACS suspicious symptoms requiring whether intensive care or emergency treatment. Furthermore, the sensitivity was additionally improved by adjusting the IMA measurement with the serum albumin of the patient. For the screening of ACS, the sensitivity and negative predictive value of IMA were higher than those of the conventional cardiac markers, myoglobin, CK-MB and troponin T, and its specificity was lower, which were similar to the results of previous studies (sensitivity, 80–90%; negative prediction rate, 85–92%; specificity, 31–49%) [5–7]. The increase of IMA sensitivity
could be explained by its theoretical increase under ischemic condition prior to the development of tissue necrosis. In our study, among 129 cases confirmed as ACS using follow-up test, 13 cases were negative for the conventional cardiac marker test but were nevertheless positive for IMA increase. The specificity and positive predictive value of IMA were very low, because IMA reflects not only cardiac ischemia but also various ischemic conditions of the brain, G–I tract or other organs, as well as various diseases. In our study, IMA was increased in a few patients only with liver diseases or muscle diseases without other causalities and IMA levels had no significance among non-ACS groups, suggesting that IMA elevation could be influenced by various factors and reflects the in vivo condition dynamically [10]. The established IMA cutoff differed depending on institutions, measurement methods, and assay purpose. Higher cutoff levels of 98.5 U/ml and 110 U/ml were used in some reports [7,11]. To screen effectively for ACS, which has to be ruled out in chest pain patients, we established 85.0 U/ml as the most ideal IMA cutoff based on ROC analysis. At the manufacturer's recommended cutoff of 90 U/ml, sensitivity was decreased to Table 1 Comparisons of diagnostic characteristics for the diagnosis of acute coronary syndrome (n = 413) Markers
Sensitivity, %
Specificity, %
NPV, %
PPV, %
Triple cardiac markers a IMA b Triple cardiac markers and IMA Albumin-adjusted IMA index c Triple cardiac markers and albuminadjusted IMA index
80.2 93.0 94.5 98.4 98.9
57.0 35.6 45.1 34.5 36.9
86.0 91.8 94.6 98.0 98.6
46.5 39.6 44.6 40.6 42.3
Abbreviations: NPV, negative predictive value; PPV, positive predictive value; IMA, ischemia modified albumin. a Triple cardiac markers: myoglobin + creatine kinase-MB+ cardiac troponin T. b IMA cutoff: 85.0 U/ml. c Albumin-adjusted IMA index cutoff: 83.4.
Y.-W. Lee et al. / Clinica Chimica Acta 384 (2007) 24–27
78.3% and specificity to 59.4%. However, compared with the screening ability of conventional cardiac markers, its differentiating power when used alone remained insufficient. The initial rule-out of ACS from diagnosis was beyond the basic purpose of the IMA test. As the sensitivity and negative predictive value of IMA were superior to the conventional cardiac markers while its specificity was lower, the usefulness its combination with the conventional cardiac panel was evaluated. The inclusion of IMA in the conventional panel increased the sensitivity to 94.5%. This confirmed that the IMA test should always be performed in parallel with the conventional cardiac panel to compensate for the false negative IMA result for ACS and to distinguish between cardiac ischemia and necrosis. We considered IMA to be an emergency test. Therefore, to minimize the preparation time needed to conduct the test separately from the analysis equipment used for the conventional cardiac marker tests, the IMA test was performed on the Integra 800 simultaneously with the conventional cardiac marker tests. The results were available approximately 40 min after receipt of the samples, with a further 15 min required for analysis, for a total of 1 h. The IMA test follows the indirect measurement principle, after the reaction of a constant amount of cobalt with albumin in the serum of patient, to measure the cobalt unbound to the Nterminal of albumin. Therefore, it is anticipated that the amount of cobalt able to bind to albumin would be low at low serum albumin concentration, and the albumin-adjusted marker was thus designed. This design differed from the reagent manufacturer's report that IMA measurement may vary depending on the value of serum albumin; nonetheless, it was explained to be a statistically insignificant change [12]. However, the experimental, 5-step dilution of serum albumin validated the presence of a linear inverse correlation, and IMA was decreased by 23 U/ ml with a serum albumin increase of 1 g/dl. In the index calculation formula [albumin-adjusted IMA marker = serum albumin concentration (g/dl) × 23 + IMA (U/ml) − 100], ‘− 100’ was designated as the value able to show the closest value to the absolute IMA measurement. Serum albumin concentration may show great variance among individuals and even within an individual due to various reasons, and thus it was thought that serum albumin concentration should be included in the interpretation of IMA measurement. The result of the re-application of the albuminadjusted IMA index to our data showed that in a comparison of the screening ability of IMA measurement, the sensitivity and the negative and positive predictive values were increased at the cutoff of 83.4, while the specificity was slightly decreased. However, when combined with the conventional test panel, the
27
sensitivity was increased to 98.9%. The IMA result could vary according to the concentration of serum albumin, which suggests the serum albumin measurement should be combined with the IMA measurement, and that the albumin-adjustment index should be established in each laboratory. In conclusion, the IMA test was confirmed to have superior sensitivity and negative predictive value for the early screening of ACS in patients admitted for chest pain. However, its diagnostic specificity was low, indicating that the IMA test should be performed in parallel with the conventional tests for the mandatory exclusion of ACS. It is suggested that the serum albumin-adjusted IMA index should be applied to ACS screening. References [1] Lee TH, Cook EF, Weisberg M, Sargent RK, Wilson C, Goldman L. Acute chest pain in the emergency room. Identification and examination of lowrisk patients. Arch Intern Med 1985;145:65–9. [2] Rouan GW, Lee TH, Cook EF, Brand DA, Weisberg MC, Goldman L. Clinical characteristics and outcome of acute myocardial infarction in patients with initially normal or nonspecific electrocardiograms (a report from the Multicenter Chest Pain Study). Am J Cardiol 1989;64:1087–92. [3] Brogan Jr GX, Hollander JE, McCuskey CF, et al. Evaluation of a new assay for cardiac troponin I vs. creatine kinase-MB for the diagnosis of acute myocardial infarction. Biochemical Markers for Acute Myocardial Ischemia (BAMI) Study Group. Acad Emerg Med 1997;4:6–12. [4] Gomez MA, Anderson JL, Karagounis LA, Muhlestein JB, Mooers FB. An emergency department-based protocol for rapidly ruling out myocardial ischemia reduces hospital time and expense: results of a randomized study (ROMIO). J Am Coll Cardiol 1996;28:25–33. [5] Anwaruddin S, Januzzi Jr JL, Baggish AL, Lewandrowski EL, Lewandrowski KB. Ischemia-modified albumin improves the usefulness of standard cardiac biomarkers for the diagnosis of myocardial ischemia in the emergency department setting. Am J Clin Pathol 2005;123:140–5. [6] Roy D, Quiles J, Aldama G, et al. Ischemia modified albumin for the assessment of patients presenting to the emergency department with acute chest pain but normal or non-diagnostic 12-lead electrocardiograms and negative cardiac troponin T. Int J Cardiol 2004;97:297–301. [7] Kang SY, Suh JT, Lee WI. Clinical usefulness of ischemia modified albumin in acute coronary syndrome. Korean J Lab Med 2005;25:306–11. [8] Gidenne S, Ceppa F, Fontan E, Perrier F, Burnat P. Analytical performance of the Albumin Cobalt Binding (ACB) test on the Cobas MIRA Plus analyzer. Clin Chem Lab Med 2004;42:455–61. [9] Christenson RH, Duh SH, Sanhai WR, et al. Characteristics of an Albumin Cobalt Binding Test for assessment of acute coronary syndrome patients: a multicenter study. Clin Chem 2001;47:464–70. [10] Zapico-Muniz E, Santalo-Bel M, Merce-Muntanola J, Montiel JA, Martinez-Rubio A, Ordonez-Llanos J. Ischemia-modified albumin during skeletal muscle ischemia. Clin Chem 2004;50:1063–5. [11] Kim JY, Yoon J, Jung IH, et al. A diagnostic value of ischemia modified albumin in patients with suspected acute coronary syndrome with normal EKG and cardiac markers. Korean Circ J 2005;35:928–33. [12] Internal data from 510k premarket notification submission to USFDA. FDA Ref No. K023824;2003 [Feb 14].