Printer Friendly

Cardiac troponin T for prediction of short- and long-term morbidity and mortality after elective open heart surgery.

Perioperative myocardial infarction (PMI) [5] in patients undergoing open heart surgery is a major determinant of postoperative morbidity and mortality (1-4). The detection of myocardial cell necrosis that may result from electrocardiographically inapparent perioperative acute myocardial infarction (AMI), from patchy areas of irreversible cell injury attributable to incomplete cardioprotection in diffuse coronary artery disease, or from contraction band necrosis attributable to catecholamine release is a challenging task.

Cardiac troponins have improved the sensitivity and specificity for detection of irreversible myocardial injury and have become the gold standard for AMI diagnosis. Myocardial infarction is now defined by a troponin concentration above the 99th percentile of a healthy reference population in the setting of myocardial ischemia (5).

Increased cardiac troponins have been reported to occur after virtually every open heart surgery. Release of troponins in the setting of open heart surgery not only reflects myocardial infarction but may also result from myocardial cell injury attributable to incomplete cardio-protection, reperfusion injury, unavoidable surgical trauma, and direct current defibrillation. The above definition of myocardial infarction derived from chest pain patients can therefore not be applied to patients undergoing open heart surgery.

The association between increased postoperative troponin concentrations and increased postoperative mortality and morbidity has been well established in recent trials (6-9). The optimum sampling time and cutoff value for detection of a critical amount of myocardial damage are, however, still unresolved. Furthermore, the added value of troponin measurement in addition to already existing elaborate risk scores is unclear.

The present study was designed to provide cardiac troponin T (cTnT) cutoff values by use of ROC curves and to assess its prognostic role in relation to established pre-and perioperative risk factors.

Materials and Methods


Between October 1998 and November 1999, a total of 204 patients scheduled to undergo elective open heart surgery at the Department for Cardiac Surgery at the University of Luebeck were recruited for the study. Patients undergoing nonelective cardiac surgery, those who had suffered an AMI with ST-segment elevation within the previous 2 weeks, and hemodynamically unstable patients were excluded.

Preoperative mortality risk was assessed by use of the prospectively validated severity score of the Cleveland Clinic Foundation (CCF) (10). In brief, this score considers preoperative risk factors that predict peri- and postoperative mortality. Factors include age, chronic obstructive pulmonary disease on medication, emergency procedure, preoperative serum creatinine >168 [micro]mol/L (1.9 mg/dL), severe left ventricular dysfunction, preoperative hematocrit <0.34, previous vascular surgery, reoperation, and presence of mitral valve regurgitation. Additional factors, including diabetes mellitus, corrected body weight, aortic stenosis, and cerebrovascular disease, were used to predict peri- and postoperative mortality.

The study was approved by the local ethics committee of the University of Luebeck. All patients gave written informed consent.


All patients underwent surgery involving median sternotomy and standard cardiopulmonary bypass techniques in moderate systemic hypothermia (32-35[degrees]C). Myocardial protection was achieved by use of a fixed combination of cold (4[degrees]C) crystalloid-based blood cardioplegia (Buckberg). This cardioplegia solution is composed of 50 g/L dextrose, 0.2 mol/L normal saline, 200 mL of 0.3 mol/L Tham, and 50 mL of standard citrate-phosphate-dextrose solution mixed with 60 mEq of KCl for the initial arresting solution and 30 mEq of KCl for the maintenance solution. The crystalloid-based cardioplegia is mixed in a 1:4 ratio with blood from the extracorporal circuit to create blood cardioplegia. This solution is then infused through the aortic root or directly into the coronary artery.

The following intraoperative variables were registered: type of surgery, cardiopulmonary bypass time, and aortic cross-clamp time.


Blood samples for measurement of cTnT were taken before cardiac surgery, 4 and 8 h after aortic cross-clamping, and every 24 h during the first postoperative week or until discharge.

cTnT was measured quantitatively by a one-step enzyme immunoassay based on electrochemiluminescence technology (Elecsys 2010; Roche). The lower detection limit of this assay is 0.01 [micro]g/L with a recommended diagnostic threshold of 0.03 [micro]/L for spontaneous AMI. The interassay CVs (between-day imprecision data set of at least 11 runs) were 20% at 0.015 [micro]/L, 10% at 0.03 [micro]g/L, and 5% at 0.08 [micro]/L.


Serial 12-lead electrocardiograms (ECGs) were recorded preoperatively, at 12 h, and then daily postoperatively for 1 week or until discharge. New, persistent Q-waves [greater than or equal to] 0.04 ms or a R-wave reduction [greater than or equal to] 25% in at least two contiguous anterior leads were used as diagnostic criteria for PMI. Evaluation of the ECGs was performed by a cardiologist blinded to the results of cardiac marker analysis.


In-hospital complications included cardiac death, PMI, major hemorrhagia, ischemic stroke, and congestive heart failure with or without the need for intraaortic balloon counterpulsation (IABP).

Long-term follow-up was performed a mean of 28 months after hospital discharge by use of hospital records, primary physician questionnaires, and telephone contacts. The predefined primary endpoint was cardiac death. Secondary endpoints consisted of noncardiac death, myocardial infarction, and repeat coronary intervention [coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI)].


All data analyses were performed with the Statistical Package for Social Sciences (SPSS for Windows 8.0; SPSS Inc.) software. Continuous variables were compared by t-test when normally distributed or by one-way ANOVA in case of more than two groups. In case of nonparametric variables, comparisons were made by Mann-Whitney U-test. Categorical variables were tested by use of [chi square] or the Fisher exact test. Among related variables, those with the greatest univariate risk were selected for multivariate modeling. The final predictive model consisted of variables that retained a significant univariate association with the prespecified endpoint. Cutoff values for continuous variables were examined by use of ROC curves. These curves were constructed by plotting sensitivity vs 1--specificity of the variable. Sensitivity, specificity, positive and negative predictive values, and likelihood ratios were calculated.

A point score system was constructed with the independently predictive variables obtained from a second multivariate analysis. Points were assigned proportionally to the odds ratios of each category. Odds ratios were calculated for quintiles of cTnT concentrations, for quartiles of cross-clamp time, and per number of CCF risk factors (Table 1). After the points were summed, patients were categorized into four risk strata (3.5-10 points, 10.5-15 points, 15.5-20 points, and >20 points).

All statistical tests were two-tailed with a P value <0.05 indicating statistical significance.


The study population consisted of 204 patients. Baseline characteristics of the entire study cohort categorized by long-term survivors vs nonsurvivors are shown in Table 2. The spectrum of surgical interventions included CABG (n = 132), valve repair/ replacement (n = 27), combined CABG/valve surgery (n = 17), and miscellaneous (n = 28).

Complete follow-up was available for 180 patients (88.7%). Patients lost to follow-up did not differ with respect to baseline characteristics from those with complete follow-ups. At 30 days, all-cause and cardiac mortality were 3.4% and 2.9%, respectively. Early cardiac death was related to sustained ventricular tachycardia (n = 2) and progressive left ventricular failure unresponsive to IABP (n = 4). One additional noncardiac death resulted from hepatic failure. Perioperative myocardial Q-wave infarction occurred in 13 patients (6.8%) and was nonfatal in all cases.

After a mean of 28 months of follow-up, all-cause and cardiac mortality rates were 8.8% and 6.9%, respectively. Nonsurvivors were older, had higher preoperative CCF risk scores and longer cardiopulmonary bypass times, more often had undergone combined CABG/valve surgery, had developed more often severe postoperative heart failure, and had longer stays in the intensive care unit (ICU). Other major complications, including rates of PMI, major hemorrhagia, ischemic stroke, and rethoracotomy, were comparable among groups (Table 3).

A characteristic time release curve of serial cTnT was seen after cardiac surgery. Cumulative cTnT release over time was higher in nonsurvivors than in survivors. Concentration curves for survivors vs nonsurvivors began to diverge by 24 h and converged by 96 h after surgery (Fig. 1). Mean (SD) peak cTnT values [1.43 (1.56) vs 1.58 (0.86) [micro]g/L] and time to peak [26.9 (39.4) vs 35.4 (51.7) h] were not different between survivors and nonsurvivors.

There was a weak but statistically significant correlation between cardiac bypass time and cTnT concentrations (r = 0.15; P = 0.03 at 24 h) but not between cross-clamp time and cTnT concentrations.


Optimum discriminators for prediction of cardiac death were determined for all markers at each time point by use of ROC curves. The optimum time point was determined as 48 h after surgery. At this time, the area under the curve (AUC) was 0.73 (95% confidence interval, 0.64-0.77) for cTnT for prediction of outcome. At 48 h the cutoff concentration for cTnT was 0.46 mg/L (Fig. 2). cTnT concentrations above this cutoff were associated with a 6.7-fold higher long-term risk (P = 0.03) for subsequent cardiac death and a 11-fold higher risk (P = 0.009) for severe postoperative heart failure requiring mechanical support (Table 3). Regarding the occurrence of PMI, a ROC-optimized cTnT >1.26 [micro]g/L obtained 24 h after surgery was associated with a 6.3-fold higher risk (P <0.0005) for the evolution of new Q-waves.


There was no difference in peak cTnT concentrations among the four groups: patients undergoing CABG, patients undergoing valve surgery, patients undergoing combined CABG/valve surgery, and patients undergoing miscellaneous other procedures.

ANOVA revealed differences in cTnT concentrations at time points 24 h (P = 0.009), 48 h (P = 0.007), and 72 h (P = 0.006) after surgery, with the lowest values being noted among patients undergoing CABG only and the highest values among patients undergoing miscellaneous procedures. ROC discriminators were not determined for the individual groups of different surgical procedures because all but one cardiac death occurred in patients undergoing either CABG or combined CAGB/valve surgery, and there were no significant differences regarding peak cTnT and cTnT values at all time points between these two groups.

Several multivariate models were constructed for prediction of long-term prognosis (Table 4). In the final model, long-term cardiac death was best predicted by the preoperative sum risk score. Intra- and postoperative variables added independent prognostic information. Thus, a cross-clamping time exceeding 65 min was associated with a 6.6-fold higher risk of cardiac death, and a cTnT concentration >0.46 [micro]g/L at 48 h after cardiac surgery predicted a 4.9-fold higher long-term cardiac mortality.

Individual sum point scores ranged from 3.5 to 27 points (Table 5). Distribution of patients in the four risk strata was as follows: 77 patients (38%) had 3.5-10 points, 73 patients (36%) had 10.5-15 points, 43 patients (21%) had 15.5-20 points, and 10 patients (5%) had >20 points. Corresponding long-term mortality rates were 1.3%,2.7%, 14%, and 40%, respectively.


When a patient with typical anginal chest pain has a cardiospecific marker in his or her blood, diagnosis of myocardial infarction must be made. For diagnostic classification the ECG recording is of little help. In fact, even patients with normal or only nonspecifically altered ECGs but with an increased blood concentration of a cardiospecific troponin carry high short- and long-term cardiovascular risks that require appropriate medical and interventional treatment. Thus, in internal medicine departments, measurement of cardiac troponins is regarded as the biochemical gold standard for diagnostic classification of chest pain patients (5).

In patients undergoing cardiovascular surgery, such a categorization into AMI and non-AMI patients based on release of cardiac constituents is no longer appropriate. There are many reasons for increased concentrations of cardiac molecules other than regional myocardial necrosis, which is the common feature of AMI after coronary thrombosis. After surgery involving cardiopulmonary bypass, reversible and irreversible cell injury may occur because of incomplete cardioprotection in diffuse coronary disease or massive hypertrophy as well as unavoidable surgical trauma. Classification of these patients as having suffered PMI would be inappropriate. However, irrespective of the causes of irreversible injury, myocardial cell necrosis still might markedly affect the short- and long-term outcomes of patients undergoing open heart surgery (11,12). Thus, it is probably more relevant to assess the total amount of perioperative myocardial cell injury as a continuum variable instead of classification as PMI or not.

The diagnosis of PMI has remained an issue of ongoing debate. Imaging by cardiac ultrasound is often technically not possible and does not allow the differentiation of scar and myocardial infarction (13,14). More sophisticated methods, such as magnetic resonance imaging or scintigraphy, are not easily applicable in the perioperative and intensive care setting (15). The ECG is commonly altered, and more specific ECG criteria such as QRS changes are of low sensitivity (16). Considering these limitations, the biochemical analysis of cardiospecific molecules becomes an attractive alternative for estimation of myocardial injury as long as they are used as indicators of cell injury but not as indicators of myocardial infarction. Troponins are the gold standard for biochemical testing for myocardial cell injury because they are truly cardiospecific. Because these molecules have a short half-life in the circulation of 90 min, in the case of cTnT an increased concentration at 48-72 h after onset of injury results from disintegration of the contractile machinery that is observed only in irreversible cell death (17).

We were the first to describe troponin release in patients after open heart surgery, for which we used a research assay for cTnT (17). In the present study, cTnT was observed in all patients undergoing surgery. The release corresponded to QRS changes in the ECG and correlated to the duration of cross-clamping. These findings were confirmed by others using different troponins and assay formats.

When ECG criteria, i.e., QRS changes, were used to classify patients with PMI, a strong association was observed between blood concentrations of troponin and ECG findings (17-21). This was also observed in the present study, in which 6.8% of all patients showed QRS criteria for AMI. This rate is consistent with the results from the CASS registry, which indicated a PMI rate of 6.4%(4).

Patients who developed new Q-waves on postoperative ECGs showed higher peak values and characteristic release curves of cTnT with a first peak on day 1 and a second, smaller peak occurring on day 4 after surgery (Fig. 3) (17). The likelihood to develop Q-wave changes was 6.3-fold higher with a ROC-optimized cTnT >1.26 [micro]g/L at 24 h after surgery (AUC = 0.69; 95% confidence interval, 0.61-0.78). This cutoff is considerably higher and the AUC lower than in a previous report on patients undergoing elective CABG (18). Bonnefoy et al. (18) reported optimum cutoff values of 0.3 [micro]g/L at 12 h for cTnT (AUC = 0.81) and of 20 [micro]g/L for creatine kinase-MB (AUC = 0.84).


The discrepancy between our findings and those of Bonnefoy et al. (18) may result from the more liberal criteria for PMI. When more patients are classified as having AMI, the troponin concentrations in the remaining group of non-AMI patients must be lower.

Montgomery et al. (22) observed increased mortality rates after open heart surgery in infants with higher preoperative cTnI concentrations. cTnI concentrations within 24 h after surgery also tended to be higher in nonsurvivors than in survivors. Consistent with our data, cTnI was significantly higher in those patients who died from progressive cardiac failure. Also consistent with our results, Eigel et al. (23) reported that after elective CABG, cTnI >0.495 [micro]g/L at the end of cardiopulmonary bypass was associated with an adjusted 17-fold risk of sustaining an adverse outcome, defined as myocardial infarction and/or perioperative death. Greenson et al. (6) demonstrated in a mixed-case cohort undergoing elective cardiac surgery that peak cTnI concentrations added to the predictive value of echocardiography, Q-waves, or significant ST-segment changes. In an attempt to determine cutoff values for patients with or without untoward events, cTnI concentrations >60 [micro]g/L were predictive of new wall motion abnormalities or Q-waves, use of IABP, significant arrhythmias, and longer ICU stays (6).

In our series, cTnT measurement at 48 h after surgery was the best predictor for severe cardiac failure and high postoperative mortality. A cTnT value exceeding 0.46 [micro]g/L at 48 h was associated with an 11-fold higher risk for severe hemodynamic deterioration and a 6.7-fold risk for cardiac death. In support of this, Koh et al. (8) found that cTnT measured in the coronary effluent of patients undergoing elective CABG surgery was related to ischemic time as well as to delayed recovery of left ventricular function and oxidative metabolism (8). Despite an absence of PMI, net release of cTnT was inversely related to left ventricular stroke work index at 3 h after surgery (8). cTnT concentrations at 48 h likely reflect irreversible cell injury.

In agreement with previous studies showing that troponin release correlated with the complexity of congenital heart disease and type of surgery (22, 24), in our study cTnT concentrations were significantly different among the four groups of patients undergoing CABG, combined CABG/valve surgery, valve surgery, or miscellaneous procedures. The lowest values were in the group undergoing CABG, and the highest values were observed with more complex procedures such as the Yacoub and Ross procedures.

The choice of cardioplegic solution is of importance for the provision of adequate cardioprotection and hence may well influence the release of cTnT (25). However, in the present study, to exclude bias, all patients uniformly received a standard protocol using a fixed combination (1:4) of cold crystalloid (St. Thomas) and cold blood cardioplegia (Buckberg).

Several models for preoperative risk assessment have been developed over the years (10, 26, 27). One of the best validated and most widely used preoperative risk scores is the one described by Higgins et al. in 1992 (10). In the present study, multivariate analysis identified three independent risk factors that were subsequently used to develop a risk score system. These variables comprised the preoperative CCF score, the intraoperative cross-clamp time, and the postoperative cTnT value at 48 h after surgery. Points were assigned to each item depending on the odds ratios of the category. There was a progressive increase in the rate of long-term mortality from 1.4% to 40% with increasing point score category. Thus, a simple point score considering pre-, intra-, and postoperative risk predictors appears very useful for risk stratification in patients undergoing open heart surgery. However, before routine use can be recommended, the present risk score system needs to be tested in a large validation cohort.

The choice of stringent ECG criteria for diagnosis of PMI constitutes the main limitation of this study. Surprisingly, we found that none of the patients who suffered a PMI died until the end of follow-up. The reasons for this may include intensified medical therapy resulting from increased attention of the attending physician. We also cannot exclude that the incidence of new Q-waves was possibly overestimated in the present study, because Q-waves on ECG are difficult to interpret postoperatively (28).

Because cTnT concentrations in blood depend on many variables, including mode of cardioprotection, severity of disease, and complexity of the procedure, the discriminator value derived in our trial may not be applicable to other patient cohorts. Therefore, the derived indices should be tested prospectively in different settings. Moreover, the prognostic ability of our point score system has not yet been tested in a validation cohort.

In conclusion, to provide optimum medical care to patients after open heart surgery it is mandatory to identify patients at high risk for future adverse events. The postoperative release of cTnT reflects the actual amount of myocardial necrosis caused not only by Q-wave myocardial infarction and surgical trauma but also by smaller infarcts, which often are inapparent to the clinician. We were able to demonstrate that relatively low concentrations of cTnT that did not produce an electrocardiographically apparent Q-wave myocardial infarction were associated with severe functional impairment of left ventricular function and higher short- and long-term mortality. A single measurement of cTnT 48 h after surgery added independent prognostic information to established clinical risk factors. Further studies are needed to determine whether patients with high postoperative cTnT release may benefit from specific medical management.


(1.) Force T, Hibberd P, Weeks G, Kemper AJ, Bloomfield P, Tow D, et al. Perioperative myocardial infarction after coronary artery bypass surgery. Clinical significance and approach to risk stratification. Circulation 1990;82:903-12.

(2.) Gray RJ, Matloff JM, Conklin CM, Ganz W, Charuzi Y, Wolfstein R, et al. Perioperative myocardial infarction: late clinical course after coronary artery bypass surgery. Circulation 1982;66:1185-9.

(3.) Namay DL, Hammermeister KE, Zia MS, DeRouen TA, Dodge HT, Namay K. Effect of perioperative myocardial infarction on late survival in patients undergoing coronary artery bypass surgery. Circulation 1982;65:1066-71.

(4.) Chaitman BR, Alderman EL, Sheffield LT, Tong T, Fisher L, Mock MB, et al. Use of survival analysis to determine the clinical significance of new Q-waves after coronary bypass surgery. Circulation 1983;67:302-9.

(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. Eur Heart J 2000;21:1502-13.

(6.) Greenson N, Macoviak J, Krishnaswamy P, Morrisey R, James C, Clopton P, et al. Usefulness of cardiac troponin I in patients undergoing open heart surgery. Am Heart J 2001;141:447-55.

(7.) Immer FF, Stocker F, Seiler AM, Pfammatter JP, Bachmann D, Printzen G, et al. Troponin-I for prediction of early postoperative course after pediatric surgery. J Am Coll Cardiol 1999;33:1719-23.

(8.) Koh TW, Hooper J, Kemp M, Ferdinand FD, Gibson DG, Pepper JR. Intraoperative release of troponin T in coronary venous and arterial blood and its relation to recovery of left ventricular function and oxidative metabolism following coronary artery surgery. Heart 1998;80:341-8.

(9.) Fellahi JL, Gue X, Richomme X, Monier E, Guillou L, Riou B. Short-and long-term prognostic value of postoperative cardiac troponin I concentration in patients undergoing coronary artery bypass grafting. Anesthesiology 2003;99:270-4.

(10.) Higgins TL, Estafanous FG, Loop FD, Beck GJ, Blum JM, Paranandi L. Stratification of morbidity and mortality outcome by preoperative risk factors in coronary artery bypass patients. A clinical severity score. JAMA 1992;267:2344-8.

(11.) Hamm CW, Ravkilde J, Gerhardt W, Jorgensen P, Peheim E, Ljungdahl L, et al. The predictive value of serum troponin T in unstable angina. N Engl J Med 1992;327:146-50.

(12.) Cannon CP, Weintraub WS, Demopoulos LA, Vicari R, Frey MJ, Lakkis N, et al., for the TACTICS (Treat Angina with Aggrastat and Determine Cost of Therapy with an Invasive or Conservative Strategy) Trial-Thrombolysis in Myocardial Infarction 18 Investigators. Comparison of early invasive and conservative strategies in patients with unstable coronary syndromes treated with the glycoprotein IIb/IIIa inhibitor tirofiban. N Engl J Med 2001;344: 1879-87.

(13.) Uday J. Myocardial infarction during coronary artery bypass surgery. J Cardiothorac Vasc Anesth 1992;6:612-23.

(14.) Swaanenburg JC, Klaase JM, DeJongste MJ, Zimmerman KW, ten Duis HJ. Troponin I, troponin T, CK-MB activity and CK-MB mass as markers for the detection of myocardial contusion in patients who experienced blunt trauma. Clin Chim Acta 1998;272:171-81.

(15.) Burns RJ, Gladstone PJ, Tremblay PC, Feindel CM, Salter DR, Lipton IH, et al. Myocardial infarction determined by technetium-99m pyrophosphate single-photon tomography complicating elective coronary artery bypass grafting for angina pectoris. Am J Cardiol 1989;63:1429-34.

(16.) Espinoza J, Lipski J, Litwak R, Donoso E, Dack S. New Q-waves after coronary artery bypass surgery for angina pectoris. Am J Cardiol 1974;33:221-4.

(17.) Katus HA, Schoeppenthau M, Tanzeem A, Bauer HG, Saggau W, Diederich KW, et al. Non-invasive assessment of perioperative myocardial cell damage by circulating cardiac troponin T. Br Heart J 1991;65:259-64.

(18.) Bonnefoy E, Filley S, Kirkorian G, Guidollet J, Roriz R, Robin J, et al. Troponin I, troponin T, or creatine kinase-MB to detect perioperative myocardial damage after coronary artery bypass surgery. Chest 1998;11:482-6.

(19.) Gensini GF, Fusi C, Conti AA, Calamai GC, Montesi GF, Galanti G, et al. Cardiac troponin I and Q-wave perioperative myocardial infarction after coronary artery bypass surgery. Crit Care Med 1998;26:1986-90.

(20.) Alyanakian MA, Dehoux M, Chatel D, Seguret C, Desmonts JM, Durand G, et al. Cardiac troponin I in diagnosis of perioperative myocardial infarction after cardiac surgery. J Cardiothorac Vasc Anesth 1998;12:288-94.

(21.) Mair J, Larue C, Mair P, Balogh D, Calzolari C, Puschendorf B. Use of cardiac troponin I to diagnose perioperative myocardial infarction in coronary artery bypass grafting. Clin Chem 1994;40:2066-70.

(22.) Montgomery VL, Sullivan JE, Buchino JJ. Prognostic value of pre-and postoperative cardiac troponin I measurement in children having cardiac surgery. Pediatr Dev Pathol 2000;3:53-60.

(23.) Eigel P, van Ingen G, Wagenpfeil S. Predictive value of perioperative cardiac troponin I for adverse outcome in coronary artery bypass surgery. Eur J Cardiothorac Surg 2001;20:544-9.

(24.) Swaanenburg JC, Loef BG, Volmer M, Boonstra PW, Grandjean JG, Mariani MA, et al. Creatine kinase MB, troponin I, and troponin T release patterns after coronary artery bypass grafting with or without cardiopulmonary bypass and after aortic and mitral valve surgery. Clin Chem 2001;47:584-7.

(25.) Pichon H, Chocron S, Alwan K, Toubin G, Kaili D, Falcoz P, et al. Crystalloid versus cold blood cardioplegia and cardiac troponin I release. Circulation 1997;96:316-20.

(26.) Parsonnet V, Dean D, Bernstein AD. A method of uniform stratification of risk for evaluating the results of surgery in acquired adult heart disease. Circulation 1989;79:13-12.

(27.) Nashef SA, Roques F, Michel P, Gauducheau E, Lemeshow S, Salamon R. European system for cardiac operative risk evaluation (EuroSCORE). Eur J Cardiothorac Surg 1999;16:9-13.

(28.) Svedjeholm R, Dahlin LG, Lundberg C, Szabo Z, Kagedal B, Nylander E, et al. Are electrocardiographic Q-wave criteria reliable for diagnosis of perioperative myocardial infarction after coronary surgery? Eur J Cardiothorac Surg 1998;13:655-61.


[1] Johns Hopkins University, Department of Cardiology, Baltimore, MD. Medizinische Universitit zu Lubeck, Departments of [2] Cardiothoracic Surgery and [3] Cardiology, Lubeck, Germany.

[4] Medizinische Universitatsklinik Heidelberg, Department of Cardiology, Heidelberg, Germany.

[5] Nonstandard abbreviations: PMI, perioperative myocardial infarction; AMI, acute myocardial infarction; cTnT, cardiac troponin T; CCF, Cleveland Clinic Foundation; ECG, electrocardiogram; IABP, intraaortic balloon counter-pulsation; CABG, coronary artery bypass grafting; PCI, percutaneous coronary intervention; ICU, intensive care unit; and AUC, area under the curve.

* Address correspondence to this author at: Medizinische Universitatsklinik Heidelberg, Abteilung fur Innere Medizin III, Bergheimer Strasse 58, 69115 Heidelberg, Germany. Fax 49-6221-56-5516; e-mail

Received January 16, 2004; accepted June 1, 2004.

Previously published online at DOI: 10.1373/clinchem.2004.031468
Table 1. Baseline characteristics according to survival status.

 Survivors Nonsurvivors
 (n = 190) (n = 14)

Mean (SD) age, years 63.3 (11.3) 71.8 (7.2)
Age > 70 years, n (%) 60 (31.6%) 9 (64.3%)
Male sex, n (%) 123 (64.7%) 10 (71.4%)
Risk factors, n (%)
 Hypertension 125 (65.8%) 12 (85.7%)
 Diabetes mellitus 43 (22.6%) 5 (35.7%)
 Smoking 71 (37.4%) 2 (14.3%)
 BMI >30 kg/[m.sup.2] 33 (17.4%) 3 (21.4%)
 Hyperlipidemia 99 (52.1%) 10 (71.4%)
History, n (%)
 Recent AMI 4 (2.1%)
 CABG 9 (4.7%) 2 (14.3%)
 PCI 25 (13.2%) 1 (7.1%)
CAD, n (%)
 None 33 (17.4%)
 1 vessel 14 (7.3%) 2 (14.3%)
 2 vessels 26 (13.7%) 2 (14.3%)
 3 vessels 117 (61.6%) 10 (71.4%)

 OR (a) (CI) P

Mean (SD) age, years 0.003
Age > 70 years, n (%) 2.04 (1.3-3.2) 0.03
Male sex, n (%) 1.1 (0.8-1.6) 0.8
Risk factors, n (%)
 Hypertension 1.3 (1.0-1.7) 0.2
 Diabetes mellitus 1.6 (0.7-3.3) 0.3
 Smoking 0.4 (0.1-1.4) 0.09
 BMI >30 kg/[m.sup.2] 1.2 (0.4-3.5) 0.7
 Hyperlipidemia 1.4 (0.9-1.9) 0.2
History, n (%)
 Recent AMI 1.0
 CABG 3.01 (0.7-12.6) 0.2
 PCI 0.5 (0.08-3.7) 1.0
CAD, n (%)
 None 0.1
 1 vessel 1.9 (0.5-7.7) 0.3
 2 vessels 1.04 (0.3-3.9)
 3 vessels 1.2 (0.8-1.6) 0.6

(a) OR, odds ratio; CI, confidence
interval; BMI, body mass index.

Table 2. Intra- and postoperative characteristics
according to survival status.

 Survivors Nonsurvivors
 (n = 190) (n = 14)

Mean (SD) ICU stay, days 1.9 (1.8) 3.9 (3.4)
Mean (SD) hospital stay, days 10.5 (5) 13.6 (7.4)
Mean (SD) cross-clamp time, min 61.2 (50.3) 83.4 (34)
Mean (SD) perfusion time, min 101.2 (33.2) 118.6 (37)
Preoperative risk, n (%)
 Low-moderate 113 (59.5%) 1 (7.1%)
 High 77 (40.5%) 13 (92.9%)
Postoperative complications, n (%)
 Major bleeding 12 (6.3%)
 Rethoracotomy 11 (5.8%)
 Severe LV failure 9 (4.7%) 3 (21.4%)
 IABP 7 (3.7%) 3 (21.4%)
 Stroke 2 (1.1%)
 Q-wave AMI 13 (6.8%)

 OR (a) (CI) P

Mean (SD) ICU stay, days 0.007
Mean (SD) hospital stay, days 0.24
Mean (SD) cross-clamp time, min 0.006
Mean (SD) perfusion time, min 0.045
Preoperative risk, n (%)
 Low-moderate 0.12 (0.02-0.8) <0.001
 High 2.3 (1.8-2.9) <0.001
Postoperative complications, n (%)
 Major bleeding 1.0
 Rethoracotomy 1.0
 Severe LV failure 4.5 (1.4-14.8) 0.039
 IABP 5.8 (1.7-20.1) 0.023
 Stroke 1.0
 Q-wave AMI 0.6

(a) OR, odds ratio; CI, confidence interval; LV, left ventricle.

Table 3. Characteristics and outcome according to cTnT cutoff at 48 h.

 cTnT >0.46 cTnT <0.46
 [micro]g/L [micro]g/L
 (n = 91) (n = 112)

Mean (SD) age, years 64.5 (12.3) 63.3 (10.3)
Age >70 years, n (%) 35 (38.5%) 33 (29.5%)
Type of surgery, n (%)
 CABG 48 (52.7%) 77 (68.8%)
 Valve surgery 43 (47.3%) 35 (31.2%)
Intraoperative variables
 Mean (SD) cross-clamp time, min 67.7 (32.6) 58.4 (59.9)
 Mean (SD) bypass time, min 111.7 (39.1) 94.7 (26.6)
CCF risk score, n (%)
 Low-moderate risk 45 (49.5%) 69 (61.6%)
 High risk 46 (50.5%) 43 (38.4%)
Major events, n (%)
 Death at 30 days 6 (6.6%)
 Cardiac death at 30 days 5 (5.5%)
 Death at 28 months 14 (15.4%) 3 (2.7%)
 Cardiac death at 28 months 11 (12.1%) 2 (1.8%)
 Q-wave AMI (b) 7/32 (21.9%) 6/172 (3.5%)
 Heart failure 8 (8.8%) 3 (2.7%)
 Severe heart failure (IABP or shock) 9 (9.9%) 1 (0.9%)
 Stroke 1 (1.1%) 1 (0.9%)
 Major bleeding 7 (7.7%) 5 (4.5%)
Mean (SD) ICU stay, days 2.7 (2.6) 1.5 (1.2)
Mean (SD) hospital stay, days 12.0 (5.7) 9.8 (4.6)

 OR (a) (CI) P

Mean (SD) age, years 0.4
Age >70 years, n (%) 1.3 (0.9-1.9) 0.24
Type of surgery, n (%)
 CABG 0.77 (0.6-0.96) 0.03
 Valve surgery 1.51 (1.06-2.1) 0.03
Intraoperative variables
 Mean (SD) cross-clamp time, min 0.2
 Mean (SD) bypass time, min 0.001
CCF risk score, n (%)
 Low-moderate risk 1.3 (0.9-1.8) 0.16
 High risk 0.8 (0.6-1.05) 0.16
Major events, n (%)
 Death at 30 days 0.007
 Cardiac death at 30 days 0.017
 Death at 28 months 5.7 (1.7-19.4) 0.001
 Cardiac death at 28 months 6.7 (1.5-29.8) 0.033
 Q-wave AMI (b) 6.3 (2.2-17.4) 0.001
 Heart failure 5.7 (1.7-19.4) 0.11
 Severe heart failure (IABP or shock) 11.1 (1.4-85.8) 0.009
 Stroke 1.2 (0.08-19.4) 1.0
 Major bleeding 1.7 (0.6-5.2) 0.51
Mean (SD) ICU stay, days 0.001
Mean (SD) hospital stay, days 0.003

(a) OR, odds ratio; CI, confidence interval.

(b) Optimal ROC cutoff for Q-wave AMI: cTnT
at 24 h >1.26 [micro]g/L (AUC = 0.69).

Table 4. Multivariate analysis of
predictors for long-term death.

 Variable B value SE OR (a) 95% CI P

Sex (male vs female) -0.41 0.68 0.66 0.17-2.53 0.54
Type of surgery (CABG vs -0.99 1.11 0.37 0.04-3.28 0.37
cTnT (>0.46 [micro]g/L vs 1.59 0.80 4.93 1.02-23.9 0.047
 <0.46 [micro]g/L)
Preoperative cTnT 0.08 0.69 1.09 0.28-4.2 0.9
 >0.03 [micro]g/L
Cross-clamp time 1.88 0.68 6.57 1.73-24.9 0.006
 (>65 min vs <65 min)
CCF risk score (per point) 0.39 0.18 1.47 1.04-2.08 0.03

(a) OR, odds ratio; CI, confidence interval.

Table 5. Point score system for estimation of long-term prognosis.

Variable Points

[cTnT.sub.48 h] ([micro]g/L)
 <0.25 1.5
 0.25-0.49 3.0
 0.5-0.74 4.5
 0.75-0.99 6.0
 >1.00 9.0
Clamp time (min)
 <42 2.0
 42-53 4.0
 54-74 6.0
 >75 8.0
CCF risk score (per point) 1.5
COPYRIGHT 2004 American Association for Clinical Chemistry, Inc.
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Evidence-Based Laboratory Medicine and Test Utilization
Author:Lehrke, Stephanie; Steen, Henning; Sievers, Hans H.; Peters, Hanno; Opitz, Armin; Muller-Bardorff, M
Publication:Clinical Chemistry
Date:Sep 1, 2004
Previous Article:Generic scheme for independent performance assessment in the molecular biology laboratory.
Next Article:Brain- and heart-type fatty acid-binding proteins in the brain: tissue distribution and clinical utility.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters