Secondary Logo

Journal Logo

Original Article

First clinical experience with the rapid-, short-acting amiodarone derivative E 047/1 after cardiac surgery

Gombotz, H.*; Vicenzi, M.*; Mahla, E.*; Rehak, P.; Metzler, H.*

Author Information
European Journal of Anaesthesiology: January 2002 - Volume 19 - Issue 1 - p 23-31

Abstract

Introduction

Depending on the inclusion criteria used, perioperative ventricular dysrhythmias occur in 44% of patients with known coronary artery disease or with high risk of coronary artery disease undergoing major non-cardiac surgery and in up to 66% of patients undergoing coronary artery bypass (CABG) surgery [1-5]. These ventricular dysrhythmias may appear at any time throughout the perioperative period, with an accumulation within 36 h after surgery [4]. Because haemodynamic stability after open-heart surgery is fragile, significant ventricular dysrhythmias may become dangerous. They need to be treated promptly and consistently. Therefore antidysrhythmic drugs are administered in up to 23% of patients after CABG-surgery and valve replacement in proportion to the predicted mortality (STS National Database 1999, http://www.ctsnet.org/doc/2986) although the use of these drugs as well as the choice of agent are controversially discussed.

Amiodarone has been successfully used in cardiac surgery in the prophylaxis and treatment of supra- as well as ventricular dysrhythmias [3,6-9]. However, intensivists and anaesthesiologists involved in perioperative intensive care still frequently refrain from using amiodarone. This is partly due to the fairly slow onset of action, the protracted half time of elimination, the occurrence of partially active metabolites, and its interaction with a wide range of concomitant medication [10,11]. Furthermore, amiodarone can make the management of patients during anaesthesia and surgery more difficult [10,12]. Especially for cardiopulmonary bypass results concerning safety are somewhat controversial [3,6,13-16].

In contrast to amiodarone, E 047/1 (EBEWE Pharmaceuticals, Unterach, Austria) as an iodine-free derivative lacks metabolites and has a short half time of elimination of approximately 6h [17]. In conscious dogs with ventricular tachycardia after myocardial infarction, E 047/1 recently demonstrated superior antidysrhythmic properties when compared to lidocaine, bretylium, flecainide and amiodarone [18]. The present clinical phase II investigation reports the first use of E 047/1 in patients. It focuses on the antidysrhythmic properties and haemodynamic influences of E 047/1 immediately after cardiac surgery when serious, haemodynamically destabilising ventricular dysrhythmias are present.

Methods

Patient population

Testing E 047/1 was performed prospectively in two consecutive phase II open, clinical studies at a single-centre intensive care unit specializing in the perioperative care of patients undergoing cardiac surgery. The initial prospective study included 15 treatment patients who underwent CABG surgery; the subsequent prospective, randomized study included 20 additional patients after CABG and/or valve repair (or replacement) procedures. Both studies acquired the approval of the Ethics Committee of the Medical Faculty and were in accordance with the guidelines of good clinical practice. All study procedures and protocols were monitored by an independent external institution (Schlieper, Schöndorf, Germany).

After written informed consent, patients (aged 18 years and over, body weight within the range >45 kg or <110 kg) with a preoperative left ventricular ejection fraction of at least 40%, with serum K+ >3.5 and serum Mg2+ >0.8 mmol L−1 - scheduled for one of the above mentioned elective surgical procedures - were included in the study. Patients excluded from the study were: those with preoperative heart failure (New York Heart Association class III or IV), or those with any type of known ventricular dysrhythmia or with long QT syndrome; those with clinically relevant disorders in cardiac impulse formation or conduction; those with severe obstructive disorders of the respiratory system; or patients who had received amiodarone treatment within the 6 months prior to surgery.

After surgery patients with serum K+ >3.5 mmol L−1 and serum Mg2+ > 0.8 mmol L−1 and a core temperature > 36.5°C having any of the following types of new-onset ventricular dysrhythmia, or a combination thereof were treated according to the study protocol:

  • continuously > 5 premature ventricular beats min−1 (Lown 2) for 1 h
  • polytopic premature ventricular beats (Lown 3a)
  • ventricular bigeminy (Lown 3b)
  • couplets, triplets, runs of premature ventricular contractions (Lown 4a)
  • non-sustained ventricular tachycardia (Lown 4b).

Patients in an unacceptably unstable haemodynamic condition, with signs of respiratory distress syndrome, being treated with epinephrine (> 0.12 μg kg−1 min−1) or dopamine (> 10 μg kg−1 min−1) or with additional antidysrhythmic treatment were excluded from the study. In case of failure to reduce existing dysrhythmias within 20 min, onset of new or life-threatening dysrhythmias (pro-dysrhythmia) or severe hypotension (mean arterial pressure < 60 mmHg) the study was terminated.

Study protocol

All patients received their routine medication until the evening before surgery. Only β-adrenoceptor blocking agents were continued until the day of surgery. After premedication with flunitrazepam (0.03 mg kg−1) and fentanyl (100 μg) arterial and central venous catheter and a pulmonary artery catheter (Edwards 93A-131H-7F, USA), were placed percutaneously prior to induction of anaesthesia. Induction of anaesthesia was performed with etomidate (0.4 mg kg−1) and fentanyl (10 μg kg−1) followed by intubation of the trachea after administration of vecuronium (0.1 mg kg−1) and pancuronium for further relaxation. Anaesthesia was supplemented with additional dosages of fentanyl (50-100 μg) and isoflurane up to 0.8 vol% inspired concentrations as clinically appropriate. Cardiopulmonary bypass was performed using nonpulsatile perfusion with membrane oxygenators (Bard HF 5701, 2.4 L m−2 ± 20%, MAP between 40 and 80 mmHg) and 1.5 L crystalloid prime in moderate hypothermia (28-30°C oesophageal). St. Thomas' Hospital solution was used for cardioplegic arrest. Depending on the haemodynamic status, the fluid remaining in the oxygenator was retransfused into the patient before decannulation.

After surgery had been completed the patients were transferred to the intensive care unit where the study took place. Crystalloids and colloids were administered to maintain PAWP in the range 8-15 mmHg, and packed red cells to maintain a haematocrit of more than 28%. Intubated patients were well-sedated (incremental doses of midazolam and fentanyl) and controlled ventilation of the lungs administered with Dräger EVITA® ventilators (Dräger, Lübeck, Germany) to maintain arterial carbon dioxide in the range 4.7-6.0 kPa and arterial oxygen tension greater than 10.7 kPa. Awake patients were breathing oxygen by mask and additional analgesia was provided with intermittent boluses of piritramide.

E 047/1 (2-butyl-3-benzofuranyl) (5-(2-(diethylamino)ethoxy)-2-thienyl)-methanone HCl, provided by EBEWE Pharmaceuticals, Unterach, Austria, was dissolved in water with polysorbate 80 (Tween 80) and was diluted to 50 mL 0.9% NaCl immediately before being applied with a syringe pump via a central venous catheter. In the pilot study, 1 mg kg−1 was administered over the course of 10 min, followed by a continuous infusion of 1 mg kg−1 h−1 for 2 h by syringe drivers. In the ensuing study, the 20 patients were randomized to receive either 2 or 3 mg kg−1 as bolus, followed by the continuous infusion of 1 mg kg−1 h−1 over 2 h. This randomization was blinded to the investigators until the trial was completed and all data were scored. Because of restrictions determined by the clinical phase I study in healthy volunteers the total dose per patient was limited to 500 mg [17]. A preoperative electrocardiogram (ECG) was obtained in all patients. After surgery ECG and pressure tracings were monitored and recorded continuously using a CER patient monitoring system (Hewlett Packard, Andover, Massachusetts, USA). Electrocardiogram tracings were recorded for up to 2h after drug initiation for the pilot study and up to 12 h for the randomized trial. Heart rate (HR), arterial, pulmonary artery, central venous and pulmonary artery wedge pressure, and, using the thermodilution technique, cardiac output was determined. Cardiac index, stroke index, systemic and pulmonary resistance index, left cardiac work index, left ventricular stroke work index, right cardiac work index and right ventricular stroke work index were calculated using standard formulae. Measurement points were at control (during the hour before drug initiation), 10, 30, 60, 120 min and 24 h after the bolus. Blood samples were drawn at the above time points to determine catecholamine and plasma drug concentrations (HPLC determination of E 047/1 was performed after solid phase extraction on C18 reversed columns, evaporation of the eluate, reconstitution and injection into the chromatographic system by means of an autosampler. HPLC was performed on a C18 stationary phase with a mixture of acetonitrile, water and perchlorate as eluent and UV-detection at 340 nm). Cardiac conduction parameters were determined in both studies preoperatively and within 1 h before the bolus and 120 min, 24 and 48 h after the bolus in the pilot and 60, 120, 240 and 480 min after the bolus in the randomized trial by using preoperative ECG and ECG tracings obtained with the CER patient monitoring system. The ECG tracings were scored every minute for single premature ventricular contractions, couplets, triplets, non-sustained and sustained ventricular tachycardia and the underlying rhythm (SA rhythm, atrial fibrillation).

Data evaluation and statistics

The primary outcome variable was the incidence of ventricular dysrhythmias. Secondary outcome variables were changes in haemodynamics, conduction intervals and levels of plasma E 047/1 and catecholamines. Data for the pilot study and the randomized trial were evaluated separately since they were performed separately and differed slightly in the observation points and periods. All data were analyzed for normal distribution. Accordingly, they are reported as means and standard deviations or as medians and quartiles. Area under the curve (AUC) was analyzed with the Wilcoxon (before and after treatment), Mann and Whitney tests for comparison of group differences in the randomized trial. Haemodynamic, plasma catecholamine and drug concentrations, as well as conduction parameters, were analyzed by analyses of variance for repeated measures (effect: time in the pilot study; main effects: group: time in the randomized study) (Super ANOVA and StatView, Abacus Concepts, Berkeley, California, USA). The significance level was set at P < 0.01.

Results

For the recruitment of 15 patients for the pilot study, 224 patients had to comply with all the preoperative inclusion criteria. This yielded a 6.7% incidence of dysrhythmias when the postoperative inclusion criteria of this study were applied. To recruit 20 patients for the randomized trial, 380 patients were needed to fulfil the preoperative criteria (5.6% incidence of dysrhythmias). Demographic characteristics and baseline data of both trials are listed in Table 1. Although moderately hypothermic cardiopulmonary bypass was applied in each case, all patients had normal core temperatures (37.1 ± 0.2°C) at the point of final inclusion into the study.

Table 1
Table 1:
Demographic data.

Outcome variables

At the time of final inclusion the patients in this study had between 6 and 12 (or more) ventricular ectopics per minute, irrespective of whether they occurred as singular premature beats, as couplets, triplets or as runs of ventricular tachycardia (Figs 1, 2). The majority of patients had a wide variety of polymorphic ventricular ectopics, up to runs of ventricular tachycardia. For the single-dose pilot trial, the time course and number of ventricular dysrhythmias before and during the administration of E 047/1 is presented in Figure 1. Within the first 2-3 min of application E 047/1 induced a decrease of ventricular dysrhythmias to between none and four per minute, a decrease that held for the duration of treatment. The area under the curve presented as median and quartiles decreased from 434 (322, 855) to 114 (9, 477, P < 0.01) events per hour. Haemodynamic variables, plasma concentrations and conduction parameters of the pilot trial are presented in Tables 2a and 2b. None of the parameters were influenced by the administration of E 047/1 except for the plasma concentration of the study drug, which rose steeply after the start of infusion.

Figure 1
Figure 1:
Total number of VES in the pilot trial. Depicts the median incidence of postoperative, haemodynamically compromising ventricular dysrhythmias (VES tot) irrespective of whether these occur as single premature beats, mono or polymorphic in couplets, triplets or runs of tachycardia. The x-axis represents the observation time in minutes. The y-axis represents the ectopics per minute. E 047/1 was initiated as a 1 mg kg−1 body weight bolus at minute 0 and was continued as an infusion of 1 mg kg−1 h−1 until minute 120.
Figure 2
Figure 2:
Depicts the median incidence of postoperative, haemodynamically compromising ventricular dysrhythmias (VES tot) irrespective of whether these occur as single premature beats, mono or polymorphic in couplets, triplets or tachycardic runs. The x-axis represents the observation time in minutes. The y-axis represents the incidence of ectopics per minute. E 047/1 was initiated as a 2 mg kg−1 and 3 mg kg−1 body weight bolus respectively at minute 0, and was continued as an infusion of 1 mg kg−1 h−1 until minute 120.
Table 2a
Table 2a:
Pilot trial - haemodynamic variables (mean ± SD).
Table 2b
Table 2b:
Pilot trial - PQ and QTc intervals (mean ± SD).

The time course and number of summarized ventricular dysrhythmias of the randomized trial before and during the administration of E 047/1 are presented in Figure 2. E 047/1 administered in either dose rapidly reduced ventricular dysrhythmias at least as effectively as in the pilot trial (AUC presented as median and quartiles: 565 (478, 701) to 33 (8, 238, P < 0.05) after a 2 mg bolus; 482 (339, 482) to 95 (13, 540, P < 0.01) events per hour after a 3 mg bolus)). Approximately 4-6 h after drug termination, dysrhythmias reappeared to varying degrees in the majority of patients. In three of the 20 patients, the incidence of dysrhythmias returned to inclusion criteria levels. These patients required further antidysrhythmic drug therapy, but E 047/1 could not be re-administered. The two bolus doses of the randomized trial were equally potent with respect to the antidysrhythmic effect. Haemodynamic variables, plasma concentrations and conduction parameters of the randomized trial are presented in Tables 3a and 3b. In contrast to the results of the pilot trial, there were mild effects on some haemodynamic variables from both the 2 and the 3 mg bolus in the randomized trial. Specifically, there was an increase of MPAP, CVP and PAWP and a decrease of LCWI (Table 3a). Control QTc after surgery was prolonged compared to preoperative QTc. E 047/1 did not influence this QTc prolongation (Table 3b). Plasma concentrations of E 047/1 were generally higher than in the pilot trial. In no case did treatment have to be terminated prematurely because exclusion or termination criteria became applicable.

Table 3a
Table 3a:
Prospective randomised study - haemodynamic parameters (mean ± SD).
Table 3b
Table 3b:
Prospective randomized study - PQ and QTc intervals (mean ± SD).

Discussion

Perioperative cardiac rhythm disturbances are often the result of transient physiological imbalances and do not necessarily require treatment. In general they impair haemodynamics modestly. However, long runs of frequent ventricular extrasystoles or frequent interpolated ventricular extrasystoles in patients with structural heart disease may produce myocardial ischaemia or haemodynamic deterioration [19,20]. Especially in the frequently fragile haemodynamic situation immediately after cardiac surgery, frequent or sustained ventricular dysrhythmias (as well as atrial fibrillation) can be deleterious. The reduced tolerance of dysrhythmias in this setting may be partially due to diminished coronary blood flow, decreased ventricular compliance and ventricular hypertrophy, reduced stroke volume, and a greater dependence of cardiac output on the co-ordinated, sequential contraction of atria and ventricles [19,20]. The underlying mechanisms of postoperative dysrhythmias frequently remain undiscovered in the individual patient, can resolve spontaneously after a certain period of time, or do not permit timely causal remedy. Therefore, to restore adequate heart rhythm and haemodynamics, suspected co-morbid conditions should be controlled as a first step followed by at least temporary additional treatment with specific antidysrhythmic agents [21,22]. This specific clinical situation may be somewhat different than in non-surgical patients with chronic ventricular dysrhythmias. The majority of surgical patients do not in general require specific antidysrhythmic drug treatment beyond the postoperative period. However, once acute postoperative treatment is indicated, antidysrhythmic substances should ideally have a rapid onset of action, should be easily titratable, should be eliminated quickly when discontinuation is indicated and should impose little haemodynamic compromise. All these pharmacodynamic requirements are not fulfilled by amiodarone but appear to be closely met by E 047/1 in this investigation [17,23].

In the present phase II study, the incidence of postoperative, haemodynamically relevant ventricular dysrhythmias of 5-7% is lower than previously published [3-5]. This difference appears to be due to our fairly stringent definition of dysrhythmias, the omission of pre- and intraoperative dysrhythmias, and the exclusion of patients with very poor ventricular function or poor general medical condition. Also, normothermia, adequate electrolyte, acid-base and fluid equilibrium, and, last but not least, adequate analgesia and sedation were strictly maintained in our patients. The small haemodynamic effects of the two higher dose-regimens on MPAP, CVP, PAWP and LCWI are obviously clinically not relevant and cannot be explained by the drug itself nor by the solvent (polysorbate) used.

A recent study found no increased risk of haemodynamic compromise following short-term amiodarone therapy in patients undergoing CABG or valve surgery with fentanyl-isoflurane anaesthesia [14]. Nevertheless, in other studies amiodarone had to be discontinued because of the induction of bradycardia and excessive prolongation of QTc [3,7,24]. However, QTc prolongation tends to be more pronounced with long-term oral than with short-term intravenous treatment with amiodarone [10]. Evaluating the exact cellular mechanisms for the antidysrhythmic properties of E 047/1 is far beyond the scope of this investigation. However, it should be emphasized that HR, PQ and QTc conduction intervals were not altered by E 047/1. Therefore the authors conclude that E 047/1 has no clinically relevant β-blocking properties or calcium channel antagonism. Furthermore, the typical class III effect does not occur when E 047/1 is administered in the short-term.

The above mentioned electrophysiological findings diverge in part from results with isolated, perfused guinea pig hearts where E 047/1 increased conduction intervals. With patch clamp techniques the inward rectifier potassium current is more strongly decreased by E 047/1 than by amiodarone. The fast and the slow component of the delayed rectifier potassium current are completely blocked by E 047/1 in contrast to partial inhibition by amiodarone. The tail current of the slow component is totally blocked and the sodium current is less influenced than by amiodarone. Calcium currents are also significantly inhibited [17]. While translation of in vitro results to the clinical situation is complicated and prone to erroneous interpretations, our hypothesis is that some of these contrasts are likely due to species differences, the lack of autonomic influences in vitro, different times of drug exposure and difficulties comparing molar concentrations in perfusion substances to drug concentrations (free and protein bound) in human plasma.

Study limitations

Several limitations of this study need to be addressed. For one, a placebo-controlled trial was considered unethical. Studies comparing treatments e.g., with lidocaine or amiodarone would require much larger sample sizes to detect inter-group differences. Furthermore, this trial was conducted in a single centre unit specializing in the perioperative management of patients undergoing cardiac surgery. Patients after cardiac surgery are particularly susceptible to exhibiting transient ventricular dysrhythmias and suffering from ensuing haemodynamic deterioration. It is also difficult under perioperative conditions to clarify the cause of ventricular dysrhythmias in the individual patient. Finally, E 047/1, like amiodarone, requires a solvent such as polysorbate 60 or 80. This is necessary to keep the substance from adhering to glass or plastic surfaces. Polysorbate has been shown to be a vasodilator and may therefore influence the haemodynamic effects of amiodarone and E 047/1 [18,25].

Conclusions

This first clinical investigation of the short-acting amiodarone derivative E 047/1 clearly demonstrates its high efficacy in the treatment of serious ventricular dysrhythmias after cardiac surgery and is thus of interest to anaesthetists. Modest haemodynamic effects were only observed at higher doses, whereas even the lowest dose used appeared to be very effective in treating serious haemodynamically destabilising ventricular arrhythmia. The antidysrhythmic effect was seen within a few minutes of drug initiation and is therefore comparable to the pharmacokinetic properties of lidocaine. Prodysrhythmia never occurred, nor did other undesired effects.

Acknowledgement

This investigation was partially funded by EBEWE Pharmaceuticals, Unterach, Austria.

References

1. O'Kelly B, Browner WS, Massie B, Tubau J, Ngo L, Mangano DT. Ventricular arrhythmias in patients undergoing noncardiac surgery. The Study of Perioperative Ischemia Research Group. JAMA 1992; 268: 217-221.
2. Mahla E, Rotman B, Rehak P, et al. Perioperative ventricular dysrhythmias in patients with structural heart disease undergoing noncardiac surgery. Anesth Analg 1998; 86: 16-21.
3. Butler J, Harriss DR, Sinclair M, Westaby S. Amiodarone prophylaxis for tachycardias after coronary artery surgery: a randomised, double blind, placebo controlled trial. Br Heart J 1993; 70: 56-60.
4. Ferraris VA, Ferraris SP, Gilliam HS, Berry WR. Predictors of postoperative ventricular dysrhythmias: a multivariate study. J Cardiovasc Surg 1991; 32: 12-20.
5. Smith RC, Leung JM, Keith FM, Merrick S, Mangano DT. Ventricular dysrhythmias in patients undergoing coronary artery bypass graft surgery: incidence, characteristics, and prognostic importance. Study of Perioperative Ischemia (SPI) Research Group. Am Heart J 1992; 123: 73-81.
6. Rady MY, Ryan T, Starr NJ. Preoperative therapy with amiodarone and the incidence of acute organ dysfunction after cardiac surgery. Anesth Analg 1997; 85: 489-497.
7. Raja P, Hawker RE, Chaikitpinyo A, et al. Amiodarone management of junctional ectopic tachycardia after cardiac surgery in children. Br Heart J 1994; 72: 261-265.
8. Installe E, Schoevaerdts JC, Gadisseux P, Charles S, Tremouroux J. Intravenous amiodarone in the treatment of various arrhythmias following cardiac operations. J Thorac Cardiovasc Surg 1981; 81: 302-308.
9. Di Biasi P, Scrofani R, Paje A, Cappiello E, Mangini A, Santoli C. Intravenous amiodarone vs propafenone for atrial fibrillation and flutter after cardiac operation. Eur J Cardiothorac Surg 1995; 9: 587-591.
10. Balser JR. The rational use of intravenous amiodarone in the perioperative period. Anesthesiology 1997; 86: 974-987.
11. Jafari-Fesharaki M, Scheinman MM. Adverse effects of amiodarone. Pacing Clin Electrophysiol 1998; 21: 108-120.
12. Van Mieghem W, Coolen L, Malysse I, Lacquet LM, Deneffe GJ, Demedts MG. Amiodarone and the development of ARDS after lung surgery. Chest 1994; 105: 1642-1645.
13. Chin C, Feindel C, Cheng D. Duration of preoperative amiodarone treatment may be associated with postoperative hospital mortality in patients undergoing heart transplantation. J Cardiothorac Vasc Anesth 1999; 13: 562-566.
14. White CM, Dunn A, Tsikouris J, et al. An assessment of the safety of short-term amiodarone therapy in cardiac surgical patients with fentanyl-isoflurane anesthesia. Anesth Analg 1999; 89: 585-589.
15. Mickleborough LL, Maruyama H, Mohamed S, et al. Are patients receiving amiodarone at increased risk for cardiac operations? Ann Thorac Surg 1994; 58: 622-629.
16. Chelimsky FC, Middlekauff HR, Stevenson WG, et al. Amiodarone therapy does not compromise subsequent heart transplantation. J Am Coll Cardiol 1992; 20: 1556-1561.
17. Mayrleitner M. E 047/1. Drugs of the Future 2000; [in press].
18. Kulier AH, Novalija E, Hogan Q, et al. The effects of the new antiarrhythmic E 047/1 on postoperative ischemia-induced arrhythmias in dogs. Anesth Analg 1999; 89: 1393-1399.
19. Steinbach KK, Merl O, Frohner K, et al. Hemodynamics during ventricular tachyarrhythmias. Am Heart J 1994; 127: 1102-1106.
20. Bethge KP. Classification of arrhythmias. J Cardiovasc Pharmacol 1991; 17 (Suppl 6): S13-S19.
21. Atlee JL. Recognition and management of arrhythmias. Preexcitation, ventricular arrhythmias, and heart block. In: Arrhythmias and Pacemakers. Philadelphia, USA: W.B. Saunders, 1996, 389-449.
22. Johnson RG, Shafique T, Sirois C, Weintraub RM, Comunale ME. Potassium concentrations and ventricular ectopy: a prospective, observational study in post-cardiac surgery patients. Crit Care Med 1999; 27: 2430-2434.
23. Connolly SJ. Evidence-based analysis of amiodarone efficacy and safety. Circulation 1999; 100: 2025-2034.
24. Hohnloser SH, Meinertz T, Dammbacher T, et al. Electrocardiographic and antiarrhythmic effects of intravenous amiodarone: results of a prospective, placebo-controlled study. Am Heart J 1991; 121: 89-95.
25. Gough WB, Zeiler RH, Barreca P, El Sherif N. Hypotensive action of commercial intravenous amiodarone and polysorbate 80 in dogs. J Cardiovasc Pharmacol 1982; 4: 375-380.
Keywords:

ANTI-ARRHYTHMIA AGENTS, amiodarone; CARDIOVASCULAR DISEASES, heart diseases, arrhythmia; PATIENT CARE, critical care; PERIOPERATIVE CARE, postoperative care; SURGICAL PROCEDURES, OPERATIVE, extracorporeal circulation, cardiopulmonary bypass

© 2002 European Academy of Anaesthesiology