Anesth Analg 1999;89:829
© 1999 International Anesthesia Research Society
CARDIOVASCULAR ANESTHESIA
Pericardial Cardiac Troponin I Release After Coronary Artery Bypass Grafting
Jean-Luc Fellahi, MD*,
Philippe Léger, MD*,
Eddy Philippe, MD*,
Martine Arthaud, PhD
,
Bruno Riou, MD, PhD*,
Iradj Gandjbakhch, MD
, and
Pierre Coriat, MD*
*Département d’Anesthésie-Réanimation,
Laboratoire de Biologie des Urgences, and
Service de Chirurgie Cardiothoracique et Vasculaire, Groupe Hospitalier Pitié-Salpêtrière, Paris VI University, Paris France
Address correspondence and reprint requests to Jean-Luc Fellahi, MD, Département d’Anesthésie-Réanimation, C.H.P. Saint-Martin, 18 rue des Roquemonts, F-14050 Caen, Cedex, France.
 |
Abstract
|
|---|
Pericardial fluid can reflect the composition of cardiac interstitium in myocardial ischemia. This study investigated the hypothesis that pericardial cardiac troponin I (CTnI) measurements could be a more accurate marker of perioperative myocardial infarction (MI) than serum CTnI after coronary artery bypass grafting (CABG). Postoperative arterial and pericardial blood samples were taken in 102 subjects undergoing elective CABG allocated to one of three groups according to the 12-lead electrocardiogram (ECG) abnormalities observed during the first postoperative 24 h: Group 1 = normal ECG; Group 2 = nonspecific ECG abnormalities; and Group 3 = perioperative Q-wave MI. Peak pericardial CTnI concentrations were much higher than peak serum concentrations in all subjects and significantly greater in Group 3 than in Groups 1 and 2 (1,318 ± 1,810 ng/mL vs 367 ± 339 ng/mL and 558 ± 608 ng/mL, respectively; P < 0.01). However, no significant difference between groups occurred at any time for pericardial/serum CTnI ratios, indicating that time courses of CTnI were not different in pericardial fluid and serum. A significant correlation was found between serum and pericardial CTnI concentrations (R = 0.70, P < 0.001). Pericardial CTnI was not more accurate than serum CTnI in predicting Q-wave MI as shown by the low value of the area under the receiver-operator characteristic curve (=0.71). Peak and early pericardial CTnI were also not accurate in predicting an increase of serum CTnI greater than a cutoff value of 19 ng/mL. Thus, pericardial CTnI measurements were less useful than serum CTnI measurements in the diagnosis of perioperative MI after CABG.
Implications: Although cardiac troponin I concentrations were much higher in pericardial fluid than in serum and significantly increased in subjects who experienced perioperative Q-wave myocardial infarction, pericardial cardiac troponin I measurements were of less value than serum cardiac troponin I measurements for the diagnosis of perioperative myocardial infarction after coronary artery bypass grafting and cannot be recommended in routine clinical practice.
|
Introduction
|
|---|
Myocardial infarction (MI) is a common cause of morbidity and mortality in patients undergoing cardiac surgery. However, the detection of perioperative MI is difficult in clinical practice, since it often occurs without changes in blood pressure or heart rate and increased values of MB creatine kinase have low specificity for the diagnosis of myocardial cell injury (1). Cardiac troponin I (CTnI) is a regulatory protein with a high specificity for myocardial damage (1,2), particularly useful for the diagnosis of perioperative MI during noncardiac surgery (3). However, the upper limit for "normal" CTnI release after cardiac surgery (associated with an uneventful outcome) is not yet well established, since multiple intraoperative causes of myocardial ischemia may occur, including aortic cross-clamping and cardiopulmonary bypass (CPB). Moreover, it could vary according to the cardiac surgical procedure (4). Pericardial fluid is an ultrafiltrate of plasma (5) that also reflects the composition of cardiac interstitium in cardiac ischemia, and the production and release of large molecules in a diseased myocardium (6). Indeed, molecules with a mass of less than 40,000 Da can diffuse through the epicardium into the pericardial space (7). Mebazaa et al. (8) have advocated that pericardial fluid analysis could be of interest for the detection of perioperative myocardial ischemia in subjects undergoing cardiac surgery.
The aim of our study was to compare serum and pericardial CTnI release after elective coronary artery bypass grafting (CABG) to evaluate the clinical relevance of pericardial fluid analysis as a complementary made of evaluating perioperative myocardial damage.
|
Methods
|
|---|
After obtaining approval from our University Ethics Committee, 102 consecutive patients undergoing elective CABG with extracorporeal circulation for symptomatic coronary disease were prospectively enrolled in the study over a 4-mo period in a single center. Exclusion criteria included emergency surgery, reoperative procedures, concomitant valvular repair or replacement, and subjects with a preoperative-paced rhythm or intraventricular conduction block. Anesthetic management was similar in all subjects. Premedication consisted of intramuscular morphine 0.1 mg/kg, atropine 0.5 mg, and an oral benzodiazepine. Anesthesia was induced with sufentanil (1–3 µg/kg) and midazolam (0.05–0.3 mg/kg) and maintained with a continuous infusion of sufentanil and midazolam and administration of isoflurane in oxygen. Muscular relaxation was induced with pancuronium (0.1–0.15 mg/kg). Standard CPB technique was used in all patients with moderate systemic hypothermia (28°–33°C), moderate hemodilution (hematocrit 20%–28%), and a nonpulsatile pump flow of 2.4 L · min-1 · m-2. Myocardial protection was achieved by cold (4°C) potassium crystalloid (Plégisol®; Abbott, Rungis, France) or cold (10°C) blood cardioplegia and by additional topical cooling. Cardioplegic solution was infused into the aortic root until cardiac arrest occurred and reinjected every 20 min during aortic cross-clamping. After the primary surgical procedure, patients were warmed to an esophageal temperature of 36.5°–37°C, and the heart was defibrillated after cardiac reperfusion if sinus rhythm did not resume spontaneously. After termination of CPB, catecholamines and/or antischemic drugs were used when necessary, at the discretion of the attending anesthesiologist. All patients were admitted postoperatively to the cardiac intensive care unit (ICU) for at least 24 h, and postoperative care was provided by clinicians not involved in the study and blinded to all research information.
Patients were allocated to one of three groups, according to the 12-lead electrocardiogram (ECG) abnormalities observed during the first postoperative 24 h: Group 1 = normal ECG; Group 2 = nonspecific ECG abnormalities; and Group 3 = perioperative acute MI. In all patients, ECG recordings were obtained preoperatively, immediately postoperatively, and 12 and 24 h later. Patients who developed symptoms suggestive of myocardial ischemia had additional ECG recordings according to clinical requirements. ECGs were assessed by two independent blinded experts who were unaware of the clinical and biochemical information. In the case of a disagreement, a third blinded expert participated in a discussion with the first two and a consensus was reached. Diagnostic criteria for perioperative MI were the appearance of new Q-waves of more than 0.04 s and 1 mm deep or a reduction in R waves of more than 25% in at least two continuous leads of the same vascular territory. Acquired conduction defects, ST segment depression of at least 1 mm or elevation of at least 2 mm, T-wave inversion, and transient left bundle-branch block were considered as nonspecific criteria for MI diagnosis and referenced in Group 2. If several abnormalities were observed for the same patient, the most serious was taken into account for the group assignment. When patients were paced with epicardial atrial or ventricular pacing for inadequate spontaneous cardiac conduction, the pacing was discontinued during ECG recording.
Serial arterial and pericardial blood samples were simultaneously drawn into dry tubes after admission to the ICU at 0 (H0), 3 (H3), 6 (H6), 12 (H12), and 24 (H24) h postoperatively. Previously, an arterial blood sample had been drawn in the operating room before CPB as a control value. Pericardial samples were easily obtained by means of a connector with a Luer-lock site inserted on the pericardial chest tube at the end of surgery. After 24 h, most pericardial drains were removed, and the study was discontinued. The samples were centrifuged at 2,000 rpm for 15 min, and serum was stored at -30°C, thawed once, and assayed in batches. Assays were performed by a technician who was unaware of the clinical and ECG data. CTnI concentrations were measured by a sensitive and highly specific immunoenzymometric assay developed by Dade BehringTM (Rueil-Malmaison, France) that uses two monoclonal antibodies specific for CTnI that recognize different epitopes. The results are expressed in units of concentration (ng/mL), and the assay allows the detection of CTnI within the range of 0.35–50 ng/mL, needing appropriate dilutions when necessary. Values greater than or equal to 0.6 ng/mL were considered positive. The cutoff value for postoperative serum CTnI release associated with no change in ECG was defined as the mean ± 2 SD of peak CTnI concentrations in Group 1.
Statistical analysis was performed using NCSS 6.0 software (Statistical Solution Ltd., Cork, Ireland). Data are expressed as mean ± SD. Comparisons of means within and between groups were performed using a two-way analysis of variance for repeated measurements completed in case of significance by the Student-Newman-Keuls multiple-comparison test. Qualitative data were compared using 
2 tests, and the number of grafts between groups were compared using the Kruskal-Wallis test. A linear regression of serum versus pericardial CTnI was performed. To analyze the accuracy of both peak and early pericardial CTnI concentrations in predicting subjects with either perioperative Q-wave MI or increased serum CTnI concentrations (irrespective of ECG), receiver-operator characteristic (ROC) curves were plotted and the area under the curve (Ac) was calculated. Ac is thought to be a precise and valid measure of diagnostic accuracy in that it is not influenced by decision biases and prior probabilities (9). A P value of less than 0.05 was considered significant.
|
Results
|
|---|
Two subjects were excluded because of control CTnI values showing a recent MI (24.6 ng/mL and 5.9 ng/mL, respectively) confirmed intraoperatively for one of them by the discovery of a fresh thrombus in the right coronary artery. Another subject was excluded because of failure to cross-clamp the aorta, and eight subjects were excluded because of incomplete samples for pericardial CTnI analysis. Table 1 shows the preoperative, operative, and postoperative data of the 91 remaining subjects. Thirty-nine (43%) subjects had normal postoperative ECG and were allocated to Group 1; 45 (49%) subjects had nonspecific ECG abnormalities and were allocated to Group 2; 7 (8%) subjects had a perioperative Q-wave MI (4 inferior, 2 anterior, and 1 septal) and were allocated to Group 3. Control values of serum CTnI were less than 0.6 ng/mL in all 91 subjects. There were no significant differences for all perioperative data between the three groups (Table 1).
Postoperative serum and pericardial CTnI concentrations in the three groups of subjects are shown in Table 2. Serum and pericardial CTnI peaked at H12 in Groups 1 and 2 and at H24 in Group 3. Peak serum concentrations were significantly higher in subjects who experienced perioperative MI (Group 3) when compared with Groups 1 and 2 (64.2 ± 85.1 ng/mL vs 7.1 ± 5.9 ng/mL and 8.8 ± 9.1 ng/mL, respectively, P < 0.01). The cutoff value (as previously defined) was 19 ng/mL. Pericardial concentrations of CTnI were much higher than serum concentrations in the three groups at all postoperative times, and peak concentrations in pericardial fluid were also significantly increased when a perioperative MI (Group 3) occurred (1,318 ± 1,810 ng/mL vs 367 ± 339 ng/mL in Group 1 and 558 ± 608 ng/mL in Group 2; P < 0.01). Postoperative pericardial/serum CTnI ratios in the three groups of subjects are shown in Table 2. No significant difference between groups occurred at any time. A significant correlation was found between serum and pericardial CTnI concentrations (Fig. 1).
View larger version (22K):
[in this window]
[in a new window]
|
Figure 1. Relationship between serum and pericardial cardiac troponin I (CTnI) concentrations (n = 455 simultaneous measures in 91 subjects).
|
|
Figure 2 shows the comparison of serum and pericardial CTnI when used to predict perioperative Q-wave MI. Pericardial CTnI was not more accurate than serum CTnI in predicting MI, as shown by a low value of the Ac (Ac = 0.71). Even early pericardial CTnI (i.e., 3 h after admission into the ICU) was not very accurate as a predictive factor (Fig. 2). Figure 3 shows the accuracy of pericardial peak and early CTnI in predicting an increase of serum CTnI greater than the cutoff value (19 ng/mL). None of these variables were very accurate, as reflected by the low values of Ac.
View larger version (12K):
[in this window]
[in a new window]
|
Figure 2. Receiver-operator characteristic curves showing the relation between sensitivity and 1-specificity in determining the predictive value of peak serum cardiac troponin I (CTnI) (A), peak pericardial cardiac troponin I (B), and early pericardial cardiac troponin I (C) concentrations for the diagnosis of postoperative Q-wave myocardial infarction. Dotted diagonal line is the no-discrimination curve. Ac = area under the receiver-operator characteristic curve.
|
|
View larger version (13K):
[in this window]
[in a new window]
|
Figure 3. Receiver-operator characteristic curves showing the relation between sensitivity and 1-specificity in determining the predictive value of peak pericardial cardiac troponin I (CTnI) (A) and early pericardial cardiac troponin I (B) concentrations for an increase in serum cardiac troponin I concentration greater than the cutoff value of 19 ng/mL. Dotted diagonal line is the no-discrimination curve. Ac = area under the receiver-operator characteristic curve.
|
|
|
Discussion
|
|---|
The present study demonstrated that pericardial CTnI concentrations could be easily measured at the bedside in patients undergoing elective CABG, as long as the pericardial chest tube was maintained for postoperative mediastinal bleeding. These concentrations were much higher than serum concentrations (30–50 times) during the first postoperative 24 h, whatever the outcome of the subject. The occurrence of perioperative Q-wave MI was responsible for both a dramatic increase in pericardial and serum CTnI concentrations. However, pericardial/serum ratios were not significantly different between the three groups, indicating that time courses of CTnI release were not different in pericardial fluid and serum. Moreover, the ROC curves showed the lack of accuracy of pericardial CTnI in predicting either a Q-wave MI or an increase in serum CTnI. Thus, pericardial CTnI measurements were of less value than serum troponin in this study for the diagnosis of postoperative MI. These results are somewhat different from those recently obtained by Mebazaa et al. (8), who found a greater increase in pericardial CTnI levels when severe myocardial ischemia occurred after cardiac surgery. However, a very small heterogeneous group of subjects scheduled for CABG or valvular replacement was investigated by these authors.
Pericardial fluid is classically considered to be an ultrafiltrate of plasma (5), but it may also reflect the composition of cardiac interstitium in ischemic cardiac disease (6). Pericardial fluid may contain molecules up to 40,000 Da (7), including CTnI (23,500 Da) and cardiac troponin T (37,000 Da), which can migrate from the interstitial myocardium through the epicardium into the pericardial space when myocardial cell damage occurs. Atrial natriuretic peptide, another molecule less than 40,000 Da produced and released by the heart, has also been found in greater concentration in the pericardial fluid of subjects with heart disease (10). Conversely, MB creatine kinase is too large (86,000 Da) to go through the epicardium into pericardial fluid. Thus, pericardial troponin analysis has been proposed as a marker for myocardial viability and function after cardiac surgery. However, even if pericardial fluid is not solely an ultrafiltrate of plasma, the present study does not support the assumption that pericardial CTnI release can be a more accurate marker than serum CTnI release for the detection of perioperative MI after CABG.
Troponin analysis has already been shown to be a reliable tool in diagnosing early perioperative myocardial cell damage during noncardiac surgery (11). The assay for CTnI currently seems to be the most sensitive and specific marker of myocardial injury (12). However, the cutoff value for perioperative MI is not yet fully established after cardiac surgery and could differ according to the surgical procedure itself, being higher after CABG than after aortic valvular replacement (4). Moreover, the CTnI serum levels are lower in perioperative non–Q-wave MI than in perioperative Q-wave MI (13). Non–Q-wave MIs are the most frequent in the perioperative period (14). Unfortunately, they present the greatest diagnostic difficulty since we do not have a highly sensitive and specific method of diagnosis. In the present study, the stringent ECG criteria used to diagnose MI were expected to screen out nontransmural MI, so that the true incidence of perioperative MI was probably underestimated. Significant increases in serum and pericardial CTnI concentrations sometimes occurred in subjects allocated to Groups 1 or 2, in the absence of clinical events or specific ECG abnormalities. It is likely that CTnI is more sensitive in detecting myocardial damage than serial ECGs. Conversely, nonspecific ECG abnormalities were frequently observed without increased serum or pericardial CTnI levels, accounting for the very low specificity of serial ECGs in diagnosing perioperative myocardial ischemia. Continuous ECG Holter monitoring might be of greater specificity in detecting prolonged postoperative ST segment changes leading to definite cell injury. The detection of new abnormalities in segmental wall motion by two-dimensional echocardiography could have also been used for this purpose (15). However, up to 37% of myocardial segments can be uninterpretable because of poor quality or drop-out of the image (16), and acute changes in left ventricular loading conditions have been shown to induce new segmental wall motion abnormalities in the absence of ischemia (17,18), reflecting less than ideal specificity of this diagnostic tool.
Our study was discontinued after the first postoperative 24 h because most pericardial chest tubes were removed at this time. Consequently, a later time point was not evaluated. It is a limitation of the study since peak serum and pericardial CTnI concentrations could have been found later in some subjects. A recent study, however, reported a peak serum CTnI concentration occurring within the first postoperative 24 h after cardiac surgery whatever the outcome of the patient (19), meaning that the distribution of blood and pericardial fluid samples over 24 h would adequately evaluate most of the time course of CTnI release.
In conclusion, this study demonstrated that pericardial CTnI concentrations could be easily measured at the bedside in subjects undergoing CABG and were much higher than simultaneous serum concentrations. The occurrence of postoperative Q-wave MI was associated with a dramatic increase in both serum and pericardial CTnI concentrations, and similar time courses of CTnI release were observed, so that pericardial/serum CTnI ratios did not significantly change. Early and peak pericardial CTnI concentrations did not accurately predict the postoperative increase in serum CTnI concentrations or the occurrence of a Q-wave MI. Thus, pericardial CTnI measurements were of less value than serum measurements in the diagnosis of perioperative myocardial injury after CABG.
|
Acknowledgments
|
|---|
Thanks to Dade Behring for kindly providing all the facilities necessary for measurement of CTnI levels. Special thanks to all the nursing staff for their patience and technical assistance, and to Dr. Mary Regan for her help in the editing of the manuscript.
|
References
|
|---|
-
Adams JE III, Bodor GS, Davila-Roman VG, et al. Cardiac troponin I: a marker with high specificity for cardiac injury. Circulation 1993;88:101–6.[Abstract/Free Full Text]
-
Mair J, Wagner I, Puschendorf B, et al. Cardiac troponin I to diagnose myocardial injury. Lancet 1993;341:838–9.[ISI][Medline]
-
Adams JE III, Sicard GA, Allen BT, et al. Diagnosis of perioperative myocardial infarction with measurement of cardiac troponin I. N Engl J Med 1994;330:670–4.[Abstract/Free Full Text]
-
Etievent JPH, Chocron S, Toubin G, et al. Use of cardiac troponin I as a marker of perioperative myocardial ischemia. Surg 1995;59:1192–4.
-
Gibson T, Segal MB. A study of the composition of pericardial fluid, with special reference to the probable mechanism of fluid formation. J Physiol (Lond) 1978;277:367–77.[Abstract/Free Full Text]
-
Fujita M, Ikemoto M, Kishishita M, et al. Elevated basic fibroblast growth factor in pericardial fluid of patients with unstable angina. Circulation 1996;94:610–3.[Abstract/Free Full Text]
-
Page E, Upshaw-Earley J, Goings G. Permeability of rat atrial endocardium, epicardium, and myocardium to large molecules: stretch-dependent effects. Circ Res 1992;71:159–73.[Abstract/Free Full Text]
-
Mebazaa A, Bchir A, Tournois C, et al. Intérêt du dosage de la troponine I dans le liquide de drainage péricardique pour le diagnostic des ischémies sévères en post-opératoire de chirurgie cardiaque [abstract]. Fr Anesth Réanim 1997;16:R205.
-
Hanley JA, McNeil BJ. The meaning and use of the area under a receiver operating characteristic (ROC) curve. Radiology 1982;143:29–36.[Abstract/Free Full Text]
-
Amano J, Suzuki A, Sunamori M, et al. Atrial natriuretic peptide in the pericardial fluid of patients with heart disease. Clin Sci 1993;85:165–8.[Medline]
-
Metzler H, Gries M, Rehak P, et al. Perioperative myocardial cell injury: the role of troponins. Br J Anaesth 1997;78:386–90.[Abstract/Free Full Text]
-
Brown CS, Bertolet BD. Cardiac troponin: see ya later, CK! Chest 1997;111:2–4.[Free Full Text]
-
Mair J, Larue C, Mair P, et al. Use of cardiac troponin I to diagnose perioperative myocardial infarction in coronary artery bypass grafting. Clin Chem 1994;40:2066–70.[Abstract]
-
Badner NH, Knill RL, Brown JE, et al. Myocardial infarction after noncardiac surgery. Anesthesiology 1998;88:572–8.[ISI][Medline]
-
Force T, Kemper AJ, Bloomfield P, et al. Non-Q wave perioperative myocardial infarction: assessment of the incidence and severity of regional dysfunction with quantitative two-dimensional echocardiography. Circulation 1985;72:781–9.[Abstract/Free Full Text]
-
Bergquist BD, Leung JM, Bellows WH. Transesophageal echocardiography in myocardial revascularization. I. Accuracy of intraoperative real-time interpretation. Analg 1996;82:1132–8.[Abstract]
-
Buffington CW, Coyle RJ. Altered load dependence of postischemic myocardium. Anesthesiology 1991;75:464–74.[ISI][Medline]
-
Seeberger MD, Cahalan MK, Rouine-Rapp K, et al. Acute hypovolemia may cause segmental wall motion abnormalities in the absence of myocardial ischemia. Anesth Analg 1997;85:1252–7.[Abstract]
-
Alyanakian MA, Dehoux M, Chatel D, et al. Cardiac troponin I in diagnosis of perioperative myocardial infarction after cardiac surgery. J Cardiothorac Vasc Anesth 1998;11:288–94.
Accepted for publication May 28, 1999.
This article has been cited by other articles:
|
|
|
|
 
A. L. Fernandez, J. B. Garcia-Bengochea, J. Alvarez, and J. R. Gonzalez Juanatey
Biochemical markers of myocardial injury in the pericardial fluid of patients undergoing heart surgery
Interactive CardioVascular and Thoracic Surgery,
June 1, 2008;
7(3):
373 - 377.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
C. Simon, F. Capuano, A. Roscitano, U. Benedetto, C. Comito, and R. Sinatra
Cardiac Troponin I vs EuroSCORE: Myocardial Infarction and Hospital Mortality
Asian Cardiovasc Thorac Ann,
April 1, 2008;
16(2):
97 - 102.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
P. Botha, D. V. Nagarajan, P. S. Lewis, and J. Dunning
Can cardiac troponins be used to diagnose a perioperative myocardial infarction post cardiac surgery?
Interactive CardioVascular and Thoracic Surgery,
September 1, 2004;
3(3):
442 - 449.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
L. Holmvang, B. Jurlander, C. Rasmussen, J. J. Thiis, P. Grande, and P. Clemmensen
Use of Biochemical Markers of Infarction for Diagnosing Perioperative Myocardial Infarction and Early Graft Occlusion After Coronary Artery Bypass Surgery
Chest,
January 1, 2002;
121(1):
103 - 111.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
