Long-Term Prognostic Value of Cardiac Troponin I and T Versus Creatine Kinase-MB Mass After Cardiac Surgery in Low-Risk Patients with Stable Symptoms

Long-Term Prognostic Value of Cardiac Troponin I and T Versus Creatine Kinase-MB Mass After Cardiac Surgery in Low-Risk Patients with Stable Symptoms Kjell Vikenes, MD, PhDa,*, Knut S. Andersen, MD, PhDa, Tor Melberg, MDc, Mikael Farstad, MD, PhDb, and Jan Erik Nordrehaug, MD, PhDa,d The long-term prognostic value of elevated cardiac biomarkers after elective cardiac surgery is not clear. The recent guidelines for diagnosing perioperative infarcts have advocated the use of similar thresholds for creatine kinase-MB (CK-MB) mass and the cardiac troponins. However, few previous data are available comparing these biomarkers after cardiac surgery, and it is not clear whether postoperative elevations of the troponins can be treated the same as elevations of CK-MB. We sought to compare the prognostic value of the cardiac troponins versus the CK-MB mass after elective cardiac surgery in low-risk patients with stable symptoms. A total of 204 consecutive patients undergoing cardiac surgery were included in the final analysis. Blood samples were drawn just before and 1 to 3 and 4 to 8 hours after the procedure, and every morning for 3 days thereafter. Patients with elevated baseline values were excluded. Using a cutoff value of 5 times the reference, patients with high and low values (controls) of CK-MB mass, cardiac troponin T (cTnT) and cardiac troponin I (cTnI) were compared. The median follow-up time was 92 months. None developed new Q-waves on the electrocardiogram. The incidence of the composite end point of all-cause mortality, readmission for acute coronary syndrome, and target vessel revascularization in the high CK-MB group was 41.2% compared to 21.8% in the controls (p ⴝ 0.004). The corresponding values for cTnT were 33.3% and 20.4% (p ⴝ 0.075) and for cTnI were 27.0% and 34.6% (p ⴝ 0.237). The p value in the isolated coronary artery bypass grafting subgroup (n ⴝ 156) was p ⴝ 0.043 for CK-MB, p ⴝ 0.137 for cTnT, and p ⴝ 0.795 for cTnI. High CK-MB (p ⴝ 0.001), ejection fraction (p ⴝ 0.002), and body mass index (p ⴝ 0.010) were the only variables independently related to reduced event-free survival. No such relation was found for high cTnT and cTnI. In conclusion, CK-MB was superior to the cardiac troponins (values >5 times the reference) in predicting long-term event-free survival after elective cardiac surgery in low-risk patients with stable symptoms undergoing coronary artery bypass grafting and/or valve surgery. © 2010 Elsevier Inc. All rights reserved. (Am J Cardiol 2010;106:780 –786) Recently, we have shown that creatine kinase-MB (CK-MB) values ⱖ5 times the reference predict worse long-term survival after elective cardiac surgery in stable, low-risk patients during a median follow-up of 95 months.1 The recent guidelines for diagnosing perioperative infarcts have advocated the use of similar thresholds for CK-MB and the cardiac troponins.2 However, few previous data are available comparing these biomarkers after cardiac surgery, and it is not clear whether postoperative elevations of the troponins can be treated the same as elevations of CK-MB. In the present study, we compared the long-term (⬎5 years) prognostic value of both cardiac troponin T (cTnT) and cardiac troponin I (cTnI) versus the CK-MB mass (using a

Departments of aHeart Disease and bClinical Biochemistry, Haukeland University Hospital, Bergen, Norway; cStavanger University Hospital, Stavanger, Norway; and dInstitute of Medicine, University of Bergen, Bergen, Norway. Manuscript received January 26, 2010; manuscript received and accepted April 23, 2010. *Corresponding author: Tel: (⫹047) 55-97-2220; fax: (⫹047) 55-975150. E-mail address: [email protected] (K. Vikenes). 0002-9149/10/$ – see front matter © 2010 Elsevier Inc. All rights reserved. doi:10.1016/j.amjcard.2010.04.038

cutoff value of 5 times the reference for all biomarkers) in patients with stable symptoms for ⱖ1 month before elective coronary artery bypass grafting (CABG) and/or valve replacement. Methods In the present prospective study, 210 consecutive patients with coronary artery and/or valve disease and stable symptoms were screened for inclusion from 1997 to 1998. Six patients were excluded. The exclusion criteria were elevated or missing cardiac troponin or CK-MB values just before surgery, not available for follow-up evaluations, recent acute coronary syndrome (ACS) of ⬍1 month, malignant disease, or creatinine ⬎200 ␮mol/L (2.3 mg/dl). Patients were compared according to biomarker values ⬍5 times the reference (controls) versus ⱖ5 times the reference (high biomarker group) at any point after surgery, in line with the recent guidelines.2 A few patients with elevated baseline values were still included in the present study if the other cardiac biomarkers were within the reference limit. All patients underwent angiography before surgery, according to the established www.ajconline.org

Coronary Artery Disease/Prognostic Value of Biomarkers After Cardiac Surgery

781

Table 1 Baseline characteristics of all patients (n ⫽ 204) Characteristic

All patients†

Controls vs High Biomarker Group* CK-MB

Age (y) (range) Men Ejection fraction储 Surgery Bypass only Aortic valve Mitral valve Body mass index (kg/m2) Smoking status Never Ever Current smoker Exsmoker Hypercholesterolemia†† Diabetes mellitus Hypertension Current use of aspirin Previous myocardial infarction Anterior Inferior Bypass ⫹ valve New York Heart Association class 0–1 2 3 Not known Previous bypass surgery Previous coronary angioplasty Creatinine ␮mol/L mg/dl Hospital stay (d)

‡

cTnT§

cTnI¶

70.9 (31-92) 157 (77.0%) 62.3 ⫾ 13.2

70.7 vs 71.3 96 vs 61 61.5 vs 63.5

69.1 vs 71.6 42 vs 115 62.6 vs 62.2

70.9 vs 70.9 64 vs 93 60.9 vs 63.2

156 (77.6%) 38 (18.0%) 10 (4.9%) 25.2 (3.5)

97 vs 59Ⲇ 15 vs 23** 3 vs 7 25.2 vs 25.3

49 vs 107** 4 vs 34** 1 vs 9 25.3 vs 25.2

68 vs 88** 8 vs 30** 3 vs 7 25.3 vs 25.2

88 (43.1%) 116 (56.9%) 49 (24.1%) 67 (32.8%) 154 (76.2%) 13 (6.4%) 57 (28.1%) 152 (74.5%) 83 (40.7%) 34 (16.7%) 40 (19.6%) 9 (4.4%)

45 vs 43 72 vs 42 32 vs 17 42 vs 25 96 vs 58** 8 vs 5 31 vs 26 94 vs 58 55 vs 28 25 vs 9 23 vs 17 7 vs 2

21 vs 67 34 vs 82 16 vs 33 18 vs 49 45 vs 109 2 vs 11 11 vs 46 48 vs 104** 26 vs 57 12 vs 22 11 vs 29 3 vs 6

25 vs 63** 53 vs 63** 18 vs 31 35 vs 32** 58 vs 96 7 vs 6 18 vs 39 64 vs 88 34 vs 49 14 vs 20 13 vs 27 7 vs 2

19 (9.3%) 123 (60.3%) 60 (29.4%) 2 (1.0%) 9 (4.4%) 18 (8.8%)

9 vs 10 70 vs 53 39 vs 21

3 vs 16 35 vs 88 15 vs 45

10 vs 11 4 vs 79 24 vs 36

2 vs 7 12 vs 6

4 vs 5 10 vs 8

2 vs 7 9 vs 9

95.4 ⫾ 36.6 1.1 ⫾ 0.4 10.4 ⫾ 4.1

96.0 vs 94.5 1.1 vs 1.1 10.4 vs 10.5

91.1 vs 97.0 1.0 vs 1.1 10.3 vs 10.5

98.6 vs 93.4 1.1 vs 1.1 10.6 vs 10.3

* Data are presented as absolute number of patients. Data are presented as n (%) or mean ⫾ SD. ‡ CK-MB ⬍25 ␮g/L (n ⫽119) versus ⱖ25 ␮g/L (n ⫽ 85). § cTnT ⬍0.75 ␮g/L (n ⫽ 54) versus ⱖ0.75 ␮g/L (n ⫽ 150). ¶ cTnI ⬍0.50 ␮g/L (n ⫽ 78) versus ⱖ0.50 (n ⫽ 126). 储 Angiography. Ⲇ p ⬍0.001; ** p ⬍0.05. †† Treated hypercholesterolemia or cholesterol level ⱖ6.5mmol/L in untreated patients. †

routines at our institution.3,4 Additional treatment strategies were determined by a consensus decision among the cardiologists and cardiac surgeons. Only patients with low risk were included in the present study. The decisions were determined by the clinical presentation, operability, and co-morbidity. The regional ethics committee approved the study, and all patients provided written informed consent. Predefined major adverse clinical events included the composite end point of all-cause mortality (⬎24 hours after the procedure), readmission for ACS, and target vessel revascularization. ACS (unstable angina pectoris, ST-segment elevation myocardial infarction, and non–ST-segment elevation myocardial infarction) were defined according to the guidelines.5 All patients underwent standard CABG and/or valve surgery according to the established routines at our hospital.1 Left internal mammary artery bypass was used to the

left anterior descending artery, and saphenous vein grafts were used to the other coronary arteries and/or valve surgery during cardiopulmonary bypass with moderate general hypothermia (32°C). Cardiac arrest was induced by infusion of cold (8°C) oxygenated blood cardioplegia administered every 20 minutes during the ischemic period. The vast majority of the patients were given cardioplegia using the anterograde route. The postoperative shed mediastinal blood was not retransfused. In ⬎95% of the patients, the proximal anastomosis was made after declamping the aorta, with a side clamp on the aorta. Venous blood samples were prospectively collected just before surgery from all patients. Blood samples were also taken at 1 to 3 and 4 to 8 hours after the procedure, and every morning thereafter for 3 days. The CK-MB mass (reference limit ⬍5 ␮g/L) was assayed using an established immunoassay, the Technicon Immuno 1 System (Bayer

782

The American Journal of Cardiology (www.ajconline.org)

Table 2 Procedural data for all patients and controls versus high biomarker group Characteristic

All Patients†

Controls vs High Biomarker Group* ‡

Procedural time (min) All patients (n ⫽ 204) Bypass surgery (n⫽156) Valve surgery (n ⫽ 35) Bypass ⫹ valve (n ⫽ 13) Bypass time (min) All patients Bypass surgery Valve surgery Bypass ⫹ valve Aortic cross clamp time (min) All patients Bypass surgery Valve surgery Bypass ⫹ valve Site of coronary artery stenosis** Left main stem Left anterior descendant Circumflex/marginal Right coronary artery No. of grafts** 1 2 3 4 5 Left internal mammary artery graft** Coronary artery disease** One vessel Two vessel Three vessel

CK-MB

cTnT§

cTnI¶

163 (85–318) 157 (85–318) 176 (110–247) 193 (120–253)

154 vs 175储 151 vs 169Ⲇ 166 vs 183 183 vs 197

148 vs 168Ⲇ 147 vs 162Ⲇ 162 vs 178 156 vs 196

149 vs 171Ⲇ 144 vs 167Ⲇ 174 vs 177 195 vs 192

82 (28–163) 74 (28–158) 104 (50–163) 119 (72–144)

72 vs 96储 68 vs 86储 93 vs 111 110 vs 122

68 vs 88储 65 vs 79储 91 vs 106 106 vs 120

70 vs 90储 65 vs 81储 96 vs 106 120 vs 118

50 (15–119) 43 (15–106) 72 (40–119) 82 (51–108)

43 vs 61储 39 vs 51储 62 vs 79Ⲇ 76 vs 84

39 vs 54Ⲇ 37 vs 46Ⲇ 58 vs 74Ⲇ 63 vs 83

41 vs 56储 37 vs 47储 65 vs 74 73 vs 84

25 (16.0%) 147 (94.2%) 137 (87.8%) 132 (84.6%)

18 vs 7 97 vs 50 88 vs 52 86 vs 46

9 vs 16 48 vs 99 39 vs 98 37 vs 95

12 vs 13 63 vs 84 53 vs 84Ⲇ 55 vs 77

1 (0.6%) 16 (10.3%) 114 (73.1%) 77 (49.4%) 15 (9.6%) 147 (94.2%)

1 vs 0 14 vs 2 71 vs 43 44 vs 33Ⲇ 11 vs 4 96 vs 51

1 vs 0 9 vs 7Ⲇ 34 vs 80 19 vs 58 3 vs 12 47 vs 100

1 vs 0 6 vs 10 51 vs 63 34 vs 43 4 vs 11 67 vs 80

2 (1.3%) 30 (19.2%) 124 (79.5%)

2 vs 0 22 vs 8 78 vs 46

2 vs 0 14 vs 16 33 vs 91Ⲇ

1 vs 1 18 vs 12Ⲇ 49 vs 75Ⲇ

* Data are presented as absolute number of patients. † Data are presented as minutes (range) or n (%). ‡ CK-MB ⬍25 ␮g/L (n ⫽119) versus ⱖ25 ␮g/L (n ⫽ 85). § cTnT ⬍0.75 ␮g/L (n ⫽ 54) versus ⱖ0.75 ␮g/L (n ⫽ 150). ¶ cTnI ⬍0.50 ␮g/L (n ⫽ 78) versus ⱖ0.50 (n ⫽ 126). 储 p ⬍0.001; Ⲇ p ⬍0.05. ** Bypass patients only. Bypass ⫽ coronary artery bypass grafting.

Business Group Diagnostics, Tarrytown, New York) according to the manufacturer’s instructions. CTnI was analyzed using a one-step, immunoenzymometric assay (ERIA, Diagnostics Pasteur, Marnes-la-Coquette, France).6 The detection limit was ⬍0.10 ␮g/L for healthy donors, and our reference was similar. CTnT was analyzed according to a revised procedure using microtiter plates, with which our laboratory was familiar, with a reference value of ⬍0.15 ␮g/L.7 The reference values of the cardiac troponins were defined as the mean baseline values plus 3 times the SEM. An electrocardiogram was taken before and at least once daily after the procedure until discharge, and any changes were assessed according to the Minnesota Code for defining myocardial infarction according to the presence of new Q-waves.8 During follow-up, all available medical records, as well as the Central Population Register and local patient administrative registries, were in-

vestigated in detail for the predefined end points and other adverse clinical events. The data are presented as the mean ⫾ SD, if not otherwise stated. Student’s t test was used when comparing continuous variables and the chi-square test (or Fisher’s exact test) for analysis of discrete variables (Statistical Package for Social Sciences, version 15, SPSS, Chicago, Illinois). The Kaplan-Meier method and log-rank test were used to evaluate differences in event-free survival. The independent ability of the cardiac markers to predict worse outcome was assessed using univariate and multivariate logistic regression analyses. All clinical relevant baseline and procedural variables were tested in the univariate analysis. Variables with a univariate p ⬍0.10 were adjusted for in the logistic regression multivariate analysis. The differences were considered significant with a 2-sided p ⬍0.05, if not otherwise stated.

Coronary Artery Disease/Prognostic Value of Biomarkers After Cardiac Surgery

783

Table 3 Temporal pattern of cardiac biomarkers before and after cardiac surgery Marker CK-MB (⬍5 ␮g/L) All (n ⫽ 204) Bypass (n ⫽ 156) Valve (n ⫽ 35) Bypass ⫹ valve (n ⫽ 13) cTnI (⬍0.10 ␮g/L) All (n ⫽ 204) Bypass (n ⫽ 156) Valve (n ⫽ 35) Bypass ⫹ valve (n ⫽ 13) cTnT (⬍0.15 ␮g/L) All (n ⫽ 204) Bypass (n ⫽ 156) Valve (n ⫽ 35) Bypass ⫹ valve (n ⫽ 13)

Before

1–3 h

4–8 h

16–20 h

40–44 h

64–68 h

High Value* (%)

2.0 (1.2) 1.9 (1.2) 2.2 (1.3) 1.7 (0.9)

23.6 (15.6) 21.4 (14.6) 31.0 (17.8) 29.2 (15.0)

22.8 (17.0) 20.4 (16.0) 32.2 (19.9) 26.2 (9.5)

17.4 (19.5) 15.1 (20.8) 22.4 (15.7) 18.2 (6.2)

5.1 (6.5) 4.6 (6.5) 6.9 (6.4) 6.5 (4.2)

3.0 (6.8) 1.7 (20.8) 3.2 (2.9) 11.2 (26.7)

42.0 35.0 66.9 69.2

0.03 (0.04) 0.04 (0.04) 0.07 (0.04) 0.04 (0.03)

0.62 (0.60) 0.54 (0.54) 0.84 (0.68) 1.04 (0.73)

0.85 (0.90) 0.79 (0.96) 1.00 (0.67) 1.15 (0.59)

0.61 (0.85) 0.55 (0.90) 0.74 (0.64) 0.88 (0.63)

0.40 (1.05) 0.38 (1.18) 0.46 (0.44) 0.49 (0.34)

0.40 (0.45) 0.22 (0.45) 0.27 (0.21) 0.63 (0.85)

60.5 54.8 80.0 76.9

0.12 (0.05) 0.12 (0.05) 0.12 (0.06) 0.11 (0.05)

1.01 (1.08) 0.96 (1.13) 1.15 (0.84) 1.28 (1.13)

1.63 (1.60) 1.66 (1.68) 1.56 (1.28) 1.62 (1.20)

0.79 (0.92) 0.74 (0.98) 0.96 (0.70) 0.92 (0.57)

0.59 (0.90) 0.52 (0.94) 0.85 (0.67) 0.77 (0.71)

0.46 (0.80) 0.40 (0.86) 0.66 (0.56) 0.70 (0.63)

70.7 66.2 82.9 92.3

Data are presented as mean ⫾ SD. * Peak values ⱖ5 times reference level.

1,0

1,0

0,8

Controls *

High CK-MB group**

0,6

0,4

0,2

0,0

Event-Free Survival

Event-Free Survival

0,8

Controls * High TnT group **

0,6

0,4

0,2 0,00

20,00

40,00

60,00

80,00

100,00

Time (months) * CK-MB mass < 25 µg/L

** CK-MB mass ≥25µg/L

Figure 1. Kaplan-Meier curves of event-free survival for CK-MB (n ⫽ 204, log-rank test, p ⫽ 0.004).

0,0 0,00

20,00

40,00

60,00

80,00

100,00

Time (months) * TnT < 0.75 µg/L ** TnT ≥ 0.75 µg/L

Results Tables 1 and 2 list the patient characteristics and procedural data. Of the 204 patients (mean age 71 years, range 31 to 92) enrolled in the final analysis, 38 patients underwent aortic valve surgery and 10 patients underwent mitral valve surgery; 13 patients underwent both CABG and valve surgery. One patient had a creatinine value ⬎150 ␮g/L (1.7 mg/dl), and 2 patients (1.1%) had an ejection fraction of ⬍30%. The number of patients with elevated biomarkers (ⱖ5 times the reference) versus controls was 85 versus 119 for CK-MB, 150 versus 54 cTnT, and 126 versus 78 cTnI. The aortic cross-clamp time was significantly longer in the high value group of all biomarkers in both the total cohort and the isolated CABG subgroup. In the subgroup of patients (n ⫽ 156) who had undergone isolated CABG (mean age 69.8 years, range 41 to 89), no significant differences were found between the controls and the high CK-MB

Figure 2. Kaplan-Meier curves of all patients for cTnT (n ⫽ 204, log-rank test, p ⫽ 0.075).

groups regarding patient characteristics comparing the same variables as listed in Table 1. The left internal mammary artery was used in 94.2% of the isolated CABG patients, with no significant difference between the controls and the high biomarker groups. In the total study cohort, both CK-MB and the cardiac troponins peaked rapidly within 4 to 8 hours after surgery (Table 3). The mean peak CK-MB was 16.0 ␮g/L in the controls versus 46.0 ␮g/L in the high CK-MB group (p ⬍0.001). The corresponding values for cTnT and cTnI were 0.38 and 2.31 ␮g/L (p ⬍0.001) and 0.28 and 1.37 ␮g/L (p ⬍0.001). One patient had a peak CK-MB value less than the reference limit, and 3 and 4 patients had a cTnT and cTnI value less than the reference limit, re-

784

The American Journal of Cardiology (www.ajconline.org) Table 5 Event-free survival according to quartiles

1,0

0,8

Event-Free Survival

High TnI group ** Controls * 0,6

Quartile

CK-MB

cTnT

cTnI

Lower Second Third Upper*

82.0% (n ⫽ 50) 76.4% (n ⫽ 55) 67.3% (n ⫽ 49) 54.0% (n ⫽ 50)

81.3% (n ⫽ 48) 66.7% (n ⫽ 54) 66.7% (n ⫽ 51) 66.7% (n ⫽ 51)

70.8% (n ⫽ 48) 65.4% (n ⫽ 52) 75.5% (n ⫽ 53) 68.6% (n ⫽ 51)

* p ⫽ 0.003 for CK-MB, p ⫽ 0.101 for cTnT, and p ⫽ 0.812 for cTnI for overall patient cohort (n ⫽ 204; upper versus lower quartiles). 0,4

0,2

0,0 0,00

20,00

40,00

60,00

80,00

100,00

Time (months) * TnI < 0.50 µg/L ** TnI ≥ 0.50 µg/L

Figure 3. Kaplan-Meier curves of all patients for cTnI (n ⫽ 204, log-rank test, p ⫽ 0.237).

Table 4 Logistic regression analysis, with adverse clinical events* during followup as dependant variable (all patients, n ⫽ 204) Marker

CK-MB mass Ejection fraction Body mass index

Univariate

Multivariate

OR

95% CI

p Value

OR

95% CI

p Value

2.10 1.03 0.91

1.12–3.94 0.95–0.99 0.80–0.98

0.003 0.004 0.015

3.27 1.04 0.90

1.67–6.41 0.94–0.99 0.77–0.96

0.001 0.002 0.010

Multivariate p value of CK-MB mass (with cutpoint of 25 ␮g/L) when all relevant baseline variables reaching univariate p value ⬍0.10 were included in analysis. Corresponding univariate and multivariate p values for cTnT were 0.078 and 0.056 and cTnI were 0.248 and 0.201. * Composite end point of all-cause mortality, readmission for ACS, and target vessel revascularization. CI ⫽ confidence interval.

spectively. A typical increase and decrease of all 3 markers was seen, but only CK-MB returned to baseline levels within 3 days after surgery. The mean peak values of the biomarkers were greatest in the combined group (Table 3) and lowest in the isolated CABG group, with the valve group in between. The all-cause mortality rate for all patients at 30 days was 0.49%, confirming the low risk of the population. During a median follow-up period of 92 months (mean 81.3), 78.2% of the controls versus 58.8% for the high CK-MB group were free of death and/or readmission for ACS and target vessel revascularization (p ⫽ 0.004; Figure 1). The freedom of all-cause mortality rate was 83.2% versus 74.1% for the controls and high biomarker group, respectively (p ⫽ 0.088). In the isolated CABG subgroup, 79.4% and 57.4% of the controls and high CK-MB group were free of death, readmission for ACS, and target vessel revascularization,

respectively (p ⫽ 0.004). In the high cTnT group and controls in the overall cohort, the rate of adverse events was 79.6% and 66.7%, respectively (p ⫽ 0.075; Figure 2). The corresponding values for the high cTnT and controls in the CABG subgroup were 79.6% and 68.2% (p ⫽ 0.137). No significant difference was seen between the high cTnI group and the controls, either in the overall cohort (p ⫽ 0.237; Figure 3) or in the isolated CABG subgroup (p ⫽ 0.795). In the small valve subgroup, none of the biomarkers predicted the outcome. On univariate logistic analysis, all relevant variables for event-free survival in Tables 1 and 2 were tested. Only body mass index (p ⫽ 0.015), ejection fraction (p ⫽ 0.004), CK-MB greater than the cutpoint of 25 ␮g/L (p ⫽ 0.003) and cTnT ⬎0.75 ␮g/L were significant (p ⬍0.10). On multivariate analysis, CK-MB was highly significant (p ⫽ 0.001; Table 4). The corresponding multivariate p values of cTnT and cTnI were 0.056 and 0.201. In the isolated CABG subgroup, the multivariate p value of CK-MB, cTnT, and cTnI was 0.001, 0.078, and 0.751, respectively. None of the biomarkers predicted increased mortality during follow-up in the isolated CABG group. When comparing the upper and lower quartiles (Table 5), the p value for the overall patient cohort was 0.003 for CK-MB, 0.101 for cTnT, and 0.812 for cTnI. The corresponding p values for the isolated CABG subgroup were 0.009, 0.171, and 0.880. If the lower and upper tertiles were compared, the corresponding p values of the overall cohort were 0.003 (isolated CABG subgroup, p ⫽ 0.015), 0.086 (isolated CABG subgroup, p ⫽ 0.169), and 0.793 (isolated CABG subgroup, p ⫽ 0.812). Discussion The main issue of the present study was to compare the prognostic value of the cardiac biomarkers around the similar cutoff value of 5 times the reference in a low-risk cohort of patients undergoing routine cardiac surgery. CK-MB, in contrast to the cardiac troponins (ⱖ5 times the reference), significantly predicted worse event-free survival after cardiac surgery in both the total cohort and the isolated CABG subgroup during a median follow-up of 92 months. The present results also showed that in these stable patients at clinical and procedural low risk, the biomarkers increased to greater than the reference level in nearly all patients. We recently showed that CK-MB values ⱖ5 times the reference predict worse outcomes in low-risk patients who undergo CABG and/or valve replacement.1 Five other previous studies have also assessed the prognostic value of

Coronary Artery Disease/Prognostic Value of Biomarkers After Cardiac Surgery

CK-MB after CABG and/or valve surgery.9 –13 However, few data are available on the long-term prognostic value of the cardiac troponins,14 and the recommendation of their use to diagnose perioperative myocardial infarction has primarily been based on expert consensus.2 To our knowledge, the present study is the first to report on the long-term prognostic value (⬎5 years) comparing cTnT, cTnI, and CK-MB mass in low-risk, stable patients in the same study population. Numerous studies have shown the high cardiospecificity of the troponins, and the prognostic value of the troponins in patients with ACS has been well documented.15–18 However, the elevation of the cardiac troponins in these patients with unstable symptoms has been considered to indicate coronary thrombosis as the causative mechanism of increased risk. The present cohort consisted of low-risk patients with stable symptoms for ⱖ1 month before surgery. The lack of predictive value of the troponins might thus be explained by the absence of thrombotic activity. Furthermore, the molecular size of the biomarkers might be of importance.19 CK-MB is a molecule that is nearly 3 times as large as cTnI, and cTnT is somewhat larger than cTnI. CK-MB might be a more robust biomarker in this setting, because its release might to a larger degree represent cell death compared to cTnT and, in particular, compared to cTnI, which might leak more easily through a damaged cell membrane. Possibly, it is the degree of micronecrosis that predicts subsequent adverse events during follow-up in these stable patients. A recent study by Sabatine et al,20 using an ultrasensitive troponin I assay to assess transient myocardial ischemia in stable patients in response to a stress test also gives some support to the concept of reversible ischemia. We could assess the temporal patterns of the biomarkers with certainty in the present study, and the serial prospective blood samples identified the peak values of the cardiac markers with certainty. In all included patients, normal baseline values were identified, and blood samples were also collected immediately after the procedure to ensure that no factors other than the surgical procedure itself had influenced the biomarker levels. The patients had stable symptoms without evidence of acute cardiac events for ⱖ30 days before surgery. The present study was also strengthened by the very long follow-up time, equal to 1,600 patient-years, although it could be argued that some events occurred so late they were unlikely to have been mediated by perioperative biomarker release. The study was limited because it was a singlecenter study with a limited sample size. Furthermore, the most recent troponin assays were not available at study inclusion. This will often be so in long-term follow-up studies and might have influenced the results. However, we have used these assays in several experimental and clinical studies, and the cardiac sensitivity and specificity have been high compared with the CK-MB mass. The recent guidelines for diagnosing perioperative myocardial infarction2 have advocated the use of similar thresholds (a reference of 5-fold or 10-fold) of CK-MB and the troponins. These recommendations were based on prognostic data and expert consensus. Our aim was to assess whether this expert consensus was valid in the present study

785

population. We did not aim to define a diagnostic threshold for myocardial infarction. The exact diagnostic cutoff value would be difficult to define even using receiver operating characteristic analysis because biomarker release is caused by the surgery itself and also because of the lack of a precise diagnostic reference standard. In our study, none of the patients developed new Q-waves after surgery. It could be argued that higher thresholds of the troponins would have performed better; for that reason, we did compare the upper and the lower quartiles, as well as the upper and lower tertiles. The upper tertile reflected somewhat greater thresholds of the troponins but was quite similar to the fivefold threshold of CK-MB. Only CK-MB demonstrated a significant difference between the upper and lower tertiles, further supporting the better prognostic value of CK-MB compared to the troponins in low-risk patients undergoing routine heart surgery. 1. Vikenes K, Andersen KS, Melberg T, Farstad M, Nordrehaug JE. Long time prognostic value of CK-MB mass in low-risk patients with stable angina scheduled for cardiac surgery. Cardiology 2009;113: 122–131. 2. Thygesen K, Alpert JS, White HD. Joint ESC/ACCF/AHA/WHF Task Force for the redefinition of myocardial infarction: universal definition of myocardial infarction. Circulation 2007;116:2634 –2653. 3. Nygård O, Nordrehaug JE, Refsum H, Ueland PM, Farstad M, Vollset SE. Plasma homocysteine levels and mortality in patients with coronary artery disease. N Engl J Med 1997;337:230 –236. 4. Vikenes K, Farstad M, Nordrehaug JE. Serotonin is associated with coronary artery disease and cardiac events. Circulation 1999;100:483– 489. 5. The Joint European Society of Cardiology/American College of Cardiology Committee. Myocardial infarction redefined—a consensus document of the Joint European Society of Cardiology/American College of Cardiology Committee for the redefinition of myocardial infarction. J Am Coll Cardiol 2000;36:959 –969. 6. Larue C, Calzolari C, Bertinchant JP, Leclercq F, Grolleau R, Pau B. Cardiac-specific immunoenzymometric assay of troponin I in the early phase of acute myocardial infarction. Clin Chem 1993;39: 972–979. 7. Hetland Ø, Gøransson L, Nilsen DWT. Cardiac troponin T immunoassay on biotin-streptavidin-coated microplates: preliminary performance in acute myocardial infarction. Scand J Clin Lab Invest 1995; 55:701–713. 8. Prineas RJ, Crowe RS, Blackburn H. The Minnesota Code Manual of Electrocardiographic Findings. Bristol: John Wright, 1982:203. 9. Klatte K, Chaitman BR, Theroux P, Gavard GA, Boyce S, Bartels C, Jessel A; GUARDIAN Study Group. Increased mortality after coronary artery bypass graft surgery is associated with increased levels of postoperative creatine kinase-myocardial band isoenzyme release. J Am Coll Cardiol 2001;38:1070 –1077. 10. Costa MA, Lichtenstein SV, Foley DP, de Valk V, Roose PCH, van Geldorp TR, Macaya C, Castanon JL, Fernandez-Avilez F, Gonzales JH, heyer G, Unger F, Serruys PW; ARTS Study Group. Incidence, predictors, and significance of Abnormal Cardiac Enzyme Rise in Patients Treated with Bypass Surgery in the Arterial Revascularization Therapies Study (ARTS). Circulation 2001;104: 2689 –2693. 11. Brener SJ, Lytle BW, Schneider JP, Ellis SG, Topol EJ. Association between CK-MB elevation after percutaneous or surgical revascularisation and three-year mortality. J Am Coll Cardiol 2002;40:1961– 1967. 12. Steuer J, Hörte L-G, Lindahl B, Ståhle E. Impact of perioperative myocardial injury on early and long-term outcome after coronary artery bypass grafting. Eur Heart J 2002;23:1219 –1227. 13. Newall N, Oo AY, Palmer ND, Grayson AD, Hine TJ, Stables RH, Fabri BM, Ramsdale DR. Intermediate and high perioperative cardiac enzyme release following isolated coronary artery bypass surgery are

786

14. 15. 16. 17.

The American Journal of Cardiology (www.ajconline.org) independently associated with higher one-year mortality. J Cardiothorac Surg 2006;1:20. Petaja L, Salmenpera M, Pulkki K, Pettila V. Biochemical injury markers and mortality after coronary artery bypass grafting: A systematic review. Ann Thorac Surg 2009;87:1981–1992. Adams JE III, Bodor GS, Davila-Roman VG, Delmez JA, Apple FS, Ladenson JH, Jaffe AS. Cardiac troponin I, a marker with high specificity for cardiac injury. Circulation 1993;88:101–106. Adams JE III, Schectman KB, Landt Y, Ladenson JH, Jaffe AS. Comparable detection of acute infarction by creatine kinase MB isoenzyme and cardiac troponin I. Clin Chem 1994;40:1291–1295. Mair J, Morandell D, Genser N, Lechleitner P, Dienstl F, Puschendorf B. Equivalent early sensitivities of myoglobin, creatine kinase MB

mass, creatine kinase isoform ratios, and cardiac troponins I and T for acute myocardial infarction. Clin Chem 1995;41:1266 –1272. 18. Vikenes K, Andersen KS, Farstad M, Nordrehaug JE. Temporal pattern of cardiac troponin I after thoracotomy and lung surgery. Int J Cardiol 2004;96:403– 407. 19. Feng YJ, Chen C, Fallon JT, Knibbs DR, Waters DD, Wu AH. Comparison of cardiac troponin I, creatine kinase-CK, and myoglobin for detection of acute ischemic myocardial injury in a swine model. Am J Clin Pathol 1998;110:70 –77. 20. Sabatine MS, Morrow DA, De Lemos JA, Jarolim P, Braunwald E. Detection of acute changes in circulating troponin in the setting of transient stress test-induced myocardial ischaemia using an ultrasensitive assay: results from TIMI. 35. Eur Heart J 2009;30:162–169.