Experimental studies suggest that the new short-acting intravenous anesthetic agent eltanolone does not markedly alter hemodynamics or cardiac function. However, because its intrinsic effects on myocardial performance and coronary blood flow are not yet known, they were examined in isolated blood-perfused rabbit hearts.
Coronary blood flow, myocardial contractility, relaxation, and oxygen consumption were measured during perfusion of hearts with 0.1 to 10 micrograms/ml eltanolone (n = 7) or its vehicle (n = 7). To determine whether the cardiac effects of eltanolone are mediated by indirect sympathetic activation, the same dose-response curve was studied in another group of five hearts depleted of catecholamine with reserpine treatment.
Coronary blood flow significantly increased with 10 micrograms/ml eltanolone and significantly decreased with 10 micrograms/ ml eltanolone vehicle. At eltanolone concentrations less than 10 micrograms/ml, myocardial contractility and relaxation remained unchanged but decreased at 10 micrograms/ml. Myocardial contractility and relaxation were not affected by perfusion of eltanolone vehicle alone. In eltanolone-perfused hearts, unchanged myocardial oxygen consumption was associated with significant increases in coronary venous oxygen content and tension, but in vehicle-perfused hearts, it was associated with reduced coronary venous oxygen content and tension. In catecholamine-depleted hearts, the variations in myocardial performance and coronary blood flow induced by eltanolone were similar to those observed in intact hearts.
Eltanolone (0.1 to 3 micrograms/ml) did not alter myocardial performance or coronary blood flow in isolated blood-perfused rabbit hearts. These effects were not due to an eltanolone-induced indirect sympathetic activation. Cardiac depression and coronary vasodilatation were only observed at concentrations of eltanolone far greater than those in clinical range.
Key words: Anesthetics, intravenous: eltanolone. Heart: coronary vessels; myocardial performance.
Eltanolone (3alpha hydroxy-5beta pregnan-20-one) is a progesterone metabolite with anesthetic properties and greater potency than thiopental or propofol. This short-acting steroid intravenous anesthetic induces smaller hemodynamic effects than those observed with equipotent doses of propofol. The induction of anesthesia with eltanolone resulted in a dose-dependent tachycardia without any changes in cardiac output, left atrial pressure, and systemic vascular resistances. However, marked decreases in cardiac output and systemic perfusion pressure are observed with high doses of eltanolone. In vivo, eltanolone also depresses cardiac function in humans. However, assessment of contractility and relaxation remains difficult in vivo, when loading conditions, heart rate, and myocardial oxygen consumption are changing simultaneously. In rat left ventricular papillary muscle, intrinsic myocardial contractility and relaxation were not affected by eltanolone or its vehicle. However, the contractile properties and calcium metabolism of rat hearts are different from those of other mammals. [7-9]Such species-related differences play an important role in the cardiac effects of anesthetic drugs. Consequently, this study used an isolated blood-perfused rabbit heart preparation. We used rabbit heart because the characteristics of rabbit myocardium are closer to those of human than are rat myocardium. To determine the intrinsic effects of eltanolone on myocardial performances and coronary vascular tone, increasing concentrations of eltanolone or its vehicle were infused into two groups of rabbit hearts. Because several products derived from progesterone have been shown to induce catecholamine release, the possibility that the cardiac effects of eltanolone are mediated by indirect sympathetic stimulation could not be precluded. Thus we also studied the direct cardiac effects of eltanolone in a group of hearts depleted of releasable norepinephrine. 
Materials and Methods
We cared for the animals according to the recommendations of the Guiding Principles in the Care and Use of Animals, and we did the study in accordance with the regulations of the official edict of the French Ministry of Agriculture.
Erythrocyte Perfusate Preparation
Outdated human erythrocytes were collected and centrifuged. The buffy coat and plasma were discarded and erythrocytes were washed with a solution of 150 mM NaCl. These operations were repeated twice and erythrocytes were stored at 4 degrees Celsius.
The perfusion medium was reconstituted by mixing the erythrocytes with modified Krebs-Henseleit bicarbonate buffer containing 118 mM NaCl, 5.9 mM K sup +, 2.5 mM free Ca sup ++, 0.5 mM MgSO4, 1.17 mM NaPO4H2, 28 mM NaHCO3, 11 mM glucose, 0.9 mM lactate, and 0.5% bovine serum albumin. The resulting suspension was passed through a 40-micro filter (Pall Ultipor SQ 40S; Pall Biomedical, Saint Germain en Laye, France). The filtered erythrocyte solution was oxygenated with a membrane oxygenator (VPCML; Cobe Cardiovascular, Arvada, CO); that is, a gas mixture comprising 20% oxygen, 5% carbon dioxide, and 75% nitrogen. After rewarming to 37 degrees Celsius, electrolyte concentrations were adjusted to achieve physiologic concentrations and sodium bicarbonate was added to obtain a standard acid-base balance (pH between 7.35 and 7.45).
New Zealand male albino rabbits weighing 2 to 2.5 kg were anesthetized with ether. Thoracotomy was performed and the heart and aortic arch were excised rapidly and placed in cold (4 degrees Celsius) isotonic saline solution. The pericardium was removed under immersion and the aorta was prepared for cannulation. The heart was mounted on an aortic cannula and retrograde perfusion was performed according to the Langendorff technique, with a hydrostatic perfusion pressure of 75 mmHg. As previously reported, the apparatus was modified to record coronary blood flow (CBF) continuously, to limit erythrocyte sedimentation, and to reduce the filling volume of the circuit. The column used to set perfusion pressure was replaced by a syringe with a plunger containing mercury. The plunger was attached to a displacement transducer that controlled the speed of a coronary pump that reflected CBF. Coronary driving pressure was computed from the signal pressure obtained from a small catheter that was positioned above the aortic valves and connected to a pressure transducer (Statham P10 EZ; Spectramed, Bilthoven, The Netherlands). The whole apparatus was enclosed in a thermostatic chamber at 37.5 degrees Celsius. A constant heart rate was maintained by atrial pacing (Medtronic 5375; Medtronic, Minneapolis, MO). The coronary sinus was drained by a small catheter inserted into the pulmonary artery. A cannulated fluid-filled balloon connected to a Statham P10 EZ pressure transducer by a rigid catheter was inserted into the left ventricle through a left atrial incision. The intraventricular balloon was inflated with normal saline to maintain left ventricular volume constant and to obtain a left ventricular end-diastolic pressure of 5 to 10 mmHg. Left ventricular systolic pressure, left ventricular end-diastolic pressure, and heart rate were recorded, and the maximal positive (dP/dt max) and negative (dP/dt min) left ventricular pressure derivatives were electronically derived from the left ventricular signal. Because intraventricular volume was held constant, dP/dt max and dP/dt min were inotropic and lusitropic indices, respectively. Atrial pacing maintained heart rate constant at 100 to 130 bpm.
Blood Gas Measurements
Arterial (PaO2) and coronary venous (PvO2) oxygen tension, arterial (PaCO2) and coronary venous (PvCOsub 2) carbon dioxide tension, and pH were measured with standard electrodes at 37 degrees Celsius (BG3 System; Instrument Laboratory, Milano, Italy). The arterial hemoglobin concentration and arterial and coronary venous saturation (SaO2and SvO2) were measured with a hemoximeter (IL482; Instrument Laboratory). Arterial and coronary venous oxygen content (CaO2and CvO2) and myocardial oxygen consumption were derived from a standard formula.
After aortic cannulation, a 60-min recovery period was allowed for stabilization. Eltanolone (Pharmacia, Stockholm, Sweden) was mixed with five samples of control perfusate to obtain the following concentrations: 0.1 (0.31 micro Meter), 0.3 (0.94 micro Meter), 1 (3.14 micro Meter), 3 (9.42 micro Meter), and 10 micro/ml (31.41 micro Meter). These concentrations included values measured in vivo during anesthesia with eltanolone. [14,15]After baseline measurements, perfusate containing 0.1 micro gram/ml eltanolone was infused until stabilization of myocardial performances and CBF. Next, control perfusate was infused again and time allowed for baseline values to recover. This protocol was then applied with the other eltanolone concentrations tested: 0.3, 1, 3, and 10 micro gram/ml. Before drug perfusion, and during the last minute of perfusion, arterial and coronary venous blood samples were collected for blood gas analysis. Seven hearts were included in this subset (eltanolone group).
In seven other rabbit hearts (vehicle group) interspersed among the seven hearts included in the eltanolone group, we tested the vehicle of eltanolone, an Intralipid-like fat emulsion (Pharmacia), using the same protocol as that previously described. The concentrations of vehicle corresponded to those tested in the eltanolone group.
Arterial and coronary blood samples were also collected in this group for blood gas analysis. The effects of eltanolone were also assessed in an other group of five hearts that were depleted of catecholamines by a single intravenous infusion of 5 mg/kg reserpine 36 h before hearts were excised. 
The hearts were weighed at the end of each experiment. Data normalized for weight refer to wet weight.
Comparisons among several means were performed using repeated-measures analysis of variance and the Newman-Keuls test. Data are expressed as means +/- SEM, and P < 0.05 was considered significant.
The pH and blood gas values and the electrolyte composition of the perfusion medium were pH, 7.36 +/- 0.1; PaO2, 157 +/- 2 mmHg; PaCO2, 41 +/- 1 mmHg; and 144 +/- 3 mM Na sup +, 103 +/- 2 mM Cl sup -; 3.5 +/- 0.7 mM K sup +; 28 +/- 7 mM NaHCO3; 2.5 +/- 0.1 mM Ca2+.
Left ventricular end-diastolic pressure, heart rate, and coronary driving pressure did not change during the study (Table 1). Return to control values of CBF, myocardial performances, and oxygen consumption was obtained between each perfusion of eltanolone or its vehicle (Table 1). During these four washout phases, the cardiac effects of eltanolone dissipated in 9 +/- 1 min, 10 +/- 2 min, 10 +/- 2 min, and 13 +/- 1 min, respectively.
Effect of Eltanolone in Its Vehicle and of the Eltanolone Vehicle Alone on Myocardial Performances
When hearts were perfused with eltanolone concentrations less than 10 micro gram/ml, left ventricular systolic pressure, dP/dtmax, and dP/dtmindid not change. In contrast, a significant decrease in myocardial performances occurred with 10 micro gram/ml sup -1, the largest concentration tested. In hearts perfused with vehicle, left ventricular systolic pressure and dP/dtmaxand dP/dtmindid not change significantly (Table 2).
Effects of Eltanolone in Its Vehicle and of the Eltanolone Vehicle Alone on Coronary Blood Flow and Myocardial Oxygen Consumption
Perfusion of eltanolone concentrations of 0.1 to 3 micro gram/ml sup -1 did not change CBF. However, with 10 micro gram/ml sup -1, a significant increase in CBF was observed, reflecting coronary vasodilatation. Myocardial oxygen consumption did not change during the study at any eltanolone concentration, but PvO2and CvO2increased significantly at 10 micro gram/ml. In the vehicle group, CBF did not change at vehicle concentrations of 0.1 to 3 micro gram/ml, but 10 micro gram/ml vehicle diminished CBF, reflecting coronary vasoconstriction. These variations were combined with unchanged myocardial oxygen consumption and significant decreases in PVO2and CvO2at 3 and 10 micro gram/ml vehicle concentrations (Table 3).
Effects of Eltanolone on Myocardial Performance and Coronary Blood Flow in Hearts Depleted of Catecholamines by Reserpine Treatment
In catecholamine-depleted hearts, left ventricular end-diastolic pressure, CP and heart rate did not change during the study. Eltanolone concentrations of 0.1 to 3 micro gram/ml did not alter dP/dt sub max or dP/dtmin, but both parameters decreased with 10 micro gram/ml eltanolone. Left ventricular systolic pressure decreased with 3 and 10 micro gram/ml eltanolone. A significant increase in CBF, reflecting coronary vasodilatation, occurred during infusion of 10 micro gram/ml eltanolone only (Table 4).
Our study showed that at concentrations as great as 3 micro gram/ml, eltanolone does not have direct myocardial or coronary effects. Coronary vasodilatation and myocardial depression only occurred for eltanolone concentrations tested corresponding to several times the clinical concentrations. [14,15]
During anesthesia with eltanolone, the cardiovascular pattern shown in dogs was a combination of tachycardia, unchanged cardiac output and systemic vascular resistances, and significant reductions in left ventricular dP/dtmaxand segmental wall thickening. Depression of cardiac function was also demonstrated in eltanolone-fentanyl anesthesia of patients with coronary diseases. However, in vivo assessment of myocardial contractility and relaxation is difficult when the heart rate and the heart loading conditions are changing. In an isolated heart preparation, left ventricular volume and heart rate can be maintained at constant levels, allowing accurate assessments of left ventricular contractility (dP/dtmax) and relaxation (dP/dtmin). The present study was conducted in a blood-perfused model because, unlike buffer-perfused hearts, this model provides physiologic values for PaO2and CaO2[13,17,18]and preserves myocardial metabolism. Such differences in the type of the perfusate were shown to affect markedly the action of propofol on the heart. In addition, because the pharmacologic effects of anesthetic drugs are species specific, we did this study using rabbit hearts, the characteristics of which are closer to those of human heart than of rat hearts. [7-9,20,21]The isomyosin form of normal rabbit and human ventricle is predominantly V3, whereas normal rat ventricle mainly consists of the V1 isomyosin form. This was associated with lower maximal shortening velocity (Vmax) in healthy rabbit and healthy human ventricle than in healthy rat myocardium. The action potential of rat ventricle is short-lived and lacks the plateau observed in rabbit or human hearts. In addition, the calcium released from the sarcoplasmic reticulum has a crucial role in the contraction of the rat ventricle, whereas calcium-induced calcium release from sarcoplasmic reticulum is implicated less in the contraction of the rabbit ventricle. Finally, Calcium2+ reuptake by sarcoplasmic reticulum during the relaxation phase is lower in rabbit and human myocytes than in the rat myocardium. [8,21]
In agreement with Riou and associates, we could not show that eltanolone concentrations of 0.1-3 micro gram/ml had any cardiac depressant effects. This similarity in results for different animal species suggests that eltanolone, over a large range of concentrations, has no direct inotropic or lusitropic effects.
Because eltanolone is a metabolite of progesterone, the transient indirect sympathetic activation demonstrated with several products derived from progesterone might have influenced our results. Indeed, the rabbit heart has very dense sympathetic innervation and a high myocardial catecholamine content, [24-26]and the sympathetic nerve terminals remain functional in isolated heart preparations. [24-26]Thus, an intrinsic negative inotropic effect of eltanolone might be masked by this indirect sympathetic activation, and the overall effect of these opposite influences might be to leave myocardial performances unchanged in most of the concentrations we tested. Because these possibilities have not been confirmed for eltanolone by specific studies, we also studied the effects of similar concentrations of eltanolone in five other catecholamine-depleted hearts obtained from reserpinized rabbits. The protocol we used has been shown to reduce catecholamine content in rabbit hearts [12,27,28]and to eliminate the cardiac and blood pressure response to tyramine. [27,28]However, the effects of eltanolone on myocardial contractility, relaxation, and coronary vessels were not modified by reserpine treatment. Overall, the present results showed (1) that concentrations of eltanolone less than 10 micro gram/ml or equivalent concentrations of its vehicle tested in the present study have no direct inotropic or lusitropic effects, and (2) that this maintenance of myocardial performance was not mediated by sympathetic activation.
However, because the concentration of protein in our medium is low, the free concentration of eltanolone may be greater than that observed in humans during anesthesia with eltanolone. Studies measuring the serum concentrations of eltanolone when anesthesia is induced showed that the median value in the serum is close to 1 micro gram/ml but also that peak concentration in some patients may reach 6 micro gram/ml [14,15]after a bolus of 0.75 mg/kg. Preliminary data suggest that erythrocyte binding of eltanolone is absent (Dr. A. Jonasson, Pharmacia, written communication, May 4, 1996). Free eltanolone concentrations have not been published yet. Assuming that a peak value of eltanolone of 6 micro gram/ml and protein binding of 99%, the free concentration of eltanolone in vivo may be close to 0.06 micro gram/ml. The concentrations we tested might be considered high. Because the albumin concentration is low in our medium, we assume that the free concentration of eltanolone may be close to the dose infused. Nevertheless, no significant changes in the parameters studied were observed until 10 micro gram/ml, suggesting that very high concentrations of free eltanolone is devoided of any cardiac effect. However, with 10 micro gram/ml eltanolone, which corresponds to several times the clinical concentrations, [14,15]myocardial performance decreased significantly both in hearts treated with reserpine and hearts not so treated. Because the eltanolone vehicle altered neither myocardial contractility nor relaxation, this cardiac depression was due to the eltanolone molecule per se. This depression might be mediated by impairment of the maximum capacity of the sarcoplasmic reticulum to load or release calcium. 
We observed a significant decrease in CBF when a concentration of vehicle equivalent to 10 micro gram/ml was perfused. This vasoconstriction by the eltanolone vehicle could be due to the presence of high free fatty-acid concentration, [29,30]because the vehicle of eltanolone contained 20% soy bean emulsion, 7% diacetylated monoglycerides, and 1.8% phospholipids. The albumin concentration remained constant in the perfusate medium, and consequently the highest free fatty-acid concentrations were obtained with the highest concentrations of vehicle. In this group, the myocardial oxygen consumption was maintained constant by a greater oxygen extraction, which was reflected by lower CvO sub 2 and PvO2. In contrast, CBF remained unchanged with eltanolone concentrations as great as 3 micro gram/ml but increased with 10 micro gram/ml eltanolone. Thus, at the highest concentration studied, the direct vasodilating effect of eltanolone per se offset the direct vasoconstricting action of its vehicle. Because myocardial oxygen consumption did not vary, this vasodilatation was associated with a significant increase in CvO2and PvO2. Similar vasodilatation was observed in catecholamine-depleted hearts, thus precluding the possibility of catecholamine release in coronary effects induced by eltanolone.
Several points must be considered when assessing the clinical relevance of our findings. This in vitro study only addressed the intrinsic effects of eltanolone on the myocardium. However, in a clinical setting, cardiac function is also affected by several cardiac and noncardiac factors, whereas CBF depends primarily on myocardial metabolic status. In addition, this study was conducted in the rabbit heart because its characteristics are more similar to those of humans hearts than are those of others mammals. However, some species-related effects cannot be dismissed for eltanolone. [9,10,23]
In conclusion, this study showed that in isolated blood-perfused hearts, 0.1 to 3 micro gram/ml eltanolone did not have direct myocardial or coronary vascular effects. In contrast, 10 micro gram/ml eltanolone induced coronary vasodilatation and myocardial depression in isolated blood-perfused rabbit hearts. Unlike what occurred with other compounds derived from progesterone, the cardiac effects of eltanolone were not mediated by indirect sympathetic activation. These results contrast with the impairment of myocardial performance observed in vitro with most anesthetics agents. [31,32]