|
|
||||||||
a Department of Anesthesiology and Intensive Care, Helsinki University Central Hospital, Helsinki, Finland
c Department of Cardiothoracic Surgery, Helsinki University Central Hospital, Helsinki, Finland
e Department of Cardiology, Helsinki University Central Hospital, Helsinki, Finland
b Department of Anesthesiology, Turku University Hospital, Turku, Finland
f Department of Cardiovascular Surgery, Turku University Hospital, Turku, Finland
d Orion Pharma, Cardiology Unit, Clinical Research and Development, Espoo, Finland
Accepted for publication October 8, 2008.
* Address correspondence to Dr Eriksson, Department of Anesthesiology and Intensive Care, Haartmaninkatu 4, Helsinki, FIN-00290 HUS, Finland (Email: heidi.eriksson{at}hus.fi).
| Cardiothoracic anesthesiology:
The Annals of Thoracic Surgery CME Program is located online at http://cme.ctsnetjournals.org. To take the CME activity related to this article, you must have either an STS member or an individual non-member subscription to the journal.
|
| Abstract |
|---|
|
|
|---|
Methods: This prospective, randomized, placebo-controlled, double-blind study included 60 patients with 3-vessel coronary disease and left ventricular ejection fraction (LVEF) of less than 0.50. Levosimendan administration (12 µg/kg bolus, followed by an infusion of 0.2 µg/kg/min) was started immediately after induction anesthesia. Predefined strict hemodynamic criteria were used to assess the success of weaning. If weaning was not successful, CPB was reinstituted and an epinephrine infusion was started. If the second weaning attempt failed, intraaortic balloon pumping (IABP) was instituted.
Results: The groups had comparable demographics. The mean (standard deviation) preoperative LVEF was 0.36 (0.8) in both groups. The baseline cardiac index was 1.8 (0.3) L/min/m2 in the levosimendan group and 1.9 (0.4) L/min/m2 in the placebo group. The mean duration of CPB to primary weaning attempt was 104 (25) minutes in the levosimendan and 109 (22) minutes in the placebo group. Primary weaning was successful in 22 patients (73%) in the levosimendan group and in 10 (33%) in the placebo group (p = 0.002). The odds ratio for failure in primary weaning was 0.182 (95% confidence interval, 0.060 to 0.552). Four patients in the placebo group failed the second weaning and underwent IABP compared with none in the levosimendan group (p = 0.112).
Conclusions: Levosimendan significantly enhanced primary weaning from CPB compared with placebo in patients undergoing 3-vessel on-pump coronary artery bypass grafting. The need for additional inotropic or mechanical therapy was decreased.
| Drs Kivikko, Laine, and Mr Sarapohja disclose that they have a financial relationship with Orion Pharma.
|
Most patients with uncompromised preoperative heart function can be weaned from cardiopulmonary bypass (CPB) without inotropic agents [1]. However, in patients with preoperatively impaired ventricular function, weaning failure without medical or mechanical support may be seen in up to 70% to 80% [2–4].
Complicated weaning may lead to myocardial distension and damage, end-organ failure due to impaired perfusion, neurologic complications, increased operative room times, and need for multiple medications and mechanical support.
Traditional positive inotropic agents improve contractility by increasing intracellular concentrations of free calcium, either by increasing the intracellular concentrations of cyclic adenosine monophosphate (cAMP) by β-adrenergic stimulation (epinephrine, dobutamine, dopamine), or by blocking the degradation of cAMP (milrinone). However, contractility with these agents is increased at the expense of increased myocardial energy consumption, risk of ischemia, and arrhythmias [5].
Levosimendan does not increase the concentration of intracellular calcium. Instead, it induces a calcium-dependent conformational change of troponin C and enhances the rate and extent of contraction of cardiac myofilaments during systole [6]. Levosimendan increases cardiac output without increasing myocardial oxygen consumption in patients early after CPB for coronary artery bypass grafting (CABG) [7, 8]. It also exerts vasodilatory and possible antiischemic and cardioprotective effects by opening of the adenosine triphosphate-dependent potassium (KATP)-channel [9–11].
This study was designed to explore the effect of prophylactic levosimendan administration on weaning from and early recovery after CPB in patients with preoperatively impaired left ventricular function undergoing CABG.
| Patients and Methods |
|---|
|
|
|---|
The main criteria for inclusion were 3-vessel coronary artery disease, impaired left ventricular ejection fraction (LVEF
0.50 evaluated with left ventricular cineangiography or echocardiography
3 months before inclusion), or signs of acute ischemic congestive heart failure (CHF), or both. The main exclusion criteria were previous CABG, indication for any cardiac valve operation, weight exceeding 160 kg, severe chronic obstructive pulmonary disease, and administration of levosimendan within the preceding 30 days.
Anesthesia was performed using a standardized, protocol-specified regimen of intravenous agents (propofol, sufentanil, rocuronium). Ephedrine was allowed as a vasopressor during anesthesia induction and phenylephrine or vasopressin thereafter during the operation and during the weaning procedure to maintain mean arterial pressure of 60 mm Hg or higher. After intubation, a fiberoptic pulmonary artery catheter was inserted through the right internal jugular vein.
Ringer's solution was infused at 40 to 60 mL/h and additional fluid was given, if needed, to maintain pulmonary artery capillary wedge pressure (PCWP) at 10 to 12 mm Hg or at the baseline value. After completion of baseline hemodynamic measurements, levosimendan or a corresponding placebo infusion was started. A bolus dose of 12 µg/kg of levosimendan during a 10-minute interval was followed by a continuous infusion of 0.2 µg/kg/min for a total infusion period of 24 hours. Placebo was supplied in identical packaging to levosimendan, including a coloring agent to achieve a similar color to the levosimendan solution, and was diluted and infused according to an identical schedule to levosimendan.
A CPB circuit with a membrane oxygenator (Dideco Avant, Mirandola, Italy, or Affinity, Medtronic, Brooklyn Park, MN) was primed with 1500 to 2000 mL of Ringer's solution and 100 mL of 15% mannitol. A nonpulsatile flow of 2.4 L/min/m2 or more, PaO 2 exceeding 30 kPa, and hematocrit of 22 or higher were provided to maintain mixed venous oxygen saturation (SVO2) of 70%. Perfusion pressure was kept between 60 to 80 mm Hg. Mechanical ventilation of the lungs was stopped during CPB, and PaCO 2 was kept at 5 to 6 kPa by adjusting the fresh gas flow through the oxygenator. Blood glucose level was maintained between 4 to 8 mmol/L with short-acting insulin.
In all subjects, aortic cross-clamping with cold blood cardioplegia was used. The nasopharyngeal temperature was allowed to decrease during perfusion to 32° to 33°C. Active warming (bladder temperature
35.5°C and nasopharyngeal temperature < 37°C) was started 15 to 20 minutes before the anticipated removal of the aortic cross-clamp.
After the aorta was declamped, the heart was defibrillated, if necessary, and rescue medication (magnesium, lidocaine, amiodarone) was used, at the discretion of the investigator. Epicardial, preferably atrial, pacing to achieve a moderately elevated heart rate (80 to 88 beats/min) was used.
Criteria for attempting weaning consisted of reperfusion time of one-third or more of the aortic cross-clamp time, perfusion pressure exceeding 60 mm Hg at a pump flow of 2.4 L/m2/min; absence of acidosis (pH < 7.3 and base excess < –3); serum potassium, 5 to 6 mmol/L; and bladder temperature, 35.5°C or higher.
The subjects were weaned from the CPB gradually, keeping central venous pressure (CVP) at 6 to 8 mm Hg and SVO2 exceeding 60% while reducing the pump flow. At the end of the weaning procedure, CVP was allowed to be increased to a level of 12 mm Hg and PCWP to a level of 16 mm Hg, if indicated.
There were four criteria for successful weaning, assessed at 10 minutes after cessation of bypass:
All criteria were to be met to claim successful weaning. If weaning was claimed successful, administration of protamine and decannulation were started.
If weaning failed, CPB was restarted and an epinephrine infusion of 0.10 µg/kg/min was started. Weaning was reattempted after 10 minutes. If the second weaning attempt failed, CPB was again resumed and an intraaortic balloon pump (IABP) was inserted.
If CI decreased 2.2 L/m2 or more and SVO2 decreased 65% or more after weaning, rescue inotropic medication (ephedrine, epinephrine, or milrinone) was administered. After the operation was completed, propofol was continued at a sedative dose and phenylephrine infusion, if ongoing, was converted to norepinephrine.
Management of patients in the intensive care unit (ICU) was in accordance with institutional guidelines. Each patient was discharged from the ICU at the discretion of the intensive care specialist.
End Points
The primary end point of the study was the proportion of patients successfully weaned from CPB by the first attempt (criteria given in the previous section). Other variables studied included invasive hemodynamic measurements, need for vasoactive medication, 24-hour fluid balance, myocardial markers, pharmacokinetics, postoperative recovery, and safety data up to 31 days.
Statistical Methods
Sample size estimation was based on the assumptions that (1) 65% of placebo-treated patients fail at primary weaning and that (2) perioperative levosimendan would decrease weaning failures to 30%. With a total of 60 subjects, the two-sided
2 square test (nQuery Advisor 4.0, Statistical Solutions, Saugus, MA) had 80% power to detect treatment effect at nominal
= 0.05 level.
The primary efficacy variable was analyzed using stratified Cochran-Mantel-Haenszel test controlling for study center. The Mantel-Haenszel odds ratio for failure and corresponding 95% confidence interval were reported.
Similarities of baseline characteristics were statistically evaluated using two-group t test and the Fisher exact test for continuous and categoric variables, respectively. The continuous variables were compared between the two study groups using repeated-measures analysis of variance (ANOVA) model with effects for treatment, time point, and treatment–time point interaction. Significance of treatment effect was also evaluated at each time point separately using ANOVA contrasts for same model.
Markers for myocardial injury and natriuretic peptides violated the normality assumption for the ANOVA test. Therefore, the nonparametric Friedman test was applied.
Time to start of rescue medication, and other time-to-event type of end points were analyzed using Cox proportional hazards model with effects for treatment. Cumulative doses of selected medications were compared between treatment groups using the Mann-Whitney test (M-W).
| Results |
|---|
|
|
|---|
|
No patient in the levosimendan group, but 4 in the placebo group failed the second weaning attempt and underwent IABP (Fisher exact test, p = 0.112; Fig 1).
|
|
|
|
In the ICU, hypotension was treated with norepinephrine in 27 patients in the levosimendan group and 22 patients in the placebo group. The median (IQR) cumulative dose of norepinephrine was 12 mg (8 to 25 mg) in the levosimendan group and 13 mg (8 to 21 mg) in the placebo group (M-W, p = 0.755). Vasopressin was administered to 2 patients in the levosimendan group (cumulative doses of 19 and 66 mg) and to 1 patient in the placebo group (cumulative dose of 106 mg).
The 24-Hour Fluid Balance
No significant difference was detected between the treatment groups in fluid input, output, and fluid balance.
Markers of Myocardial Injury
After operation, N-terminal proatrial natriuretic peptide (NT-proANP) and N-terminal pro-B-type natriuretic peptide (NT-proBNP) increased in both groups without any significant difference between the groups.
There were no significant differences between the groups in the changes in P-creatine kinase-MB subunit (P-CK-MB). However, P-troponin T increased statistically significantly more in the placebo group than in the levosimendan group. The median (IRQ) P-troponin T concentration in µg/L was 0.03 (0.03 to 0.03) in both groups at baseline, 0.22 (0.16 to 0.41) in the levosimendan group and 0.37 (0.23 to 0.59) in the placebo group at the end of the procedure; 0.33 (0.26 to 0.91) and 0.69 (0.39 to 1.32) at 24 hours, and 0.23 (0.18 to 0.52) and 0.51 (0.29 to 0.81), respectively, at 2 days (p = 0.037). One patient in the levosimendan group and 3 patients in the placebo group had postoperative P-CK-MB levels that exceeded 100 µg/L.
Serum Lactate
The mean lactate value increased significantly more from baseline in the placebo group compared with the levosimendan group at 4 hours (p = 0.004) and at 6 hours (p = 0.015) after the aorta was declamped. The increase in lactate was also significantly higher in the placebo group than in the levosimendan group during the 4- to 24-hour period after aortic declamping (Fig 2).
Postoperative serum lactate levels exceeding 3 mmol/L were measured in 9 patients in the levosimendan group and in 15 patients in the placebo group. Postoperative serum lactate exceeding 6 mmol/L was measured in 2 patients in the levosimendan and 5 patients in the placebo group.
Pharmacokinetics
Levosimendan was rapidly absorbed and eliminated from plasma after the end of the study drug infusion. The mean ± SD peak concentration 98 ± 33 ng/mL was reached approximately 2 hours after the start of the study drug infusion. The mean half-life of elimination was 1.3 ± 0.4 hours. The active metabolites of levosimendan were formed slowly after the study drug infusion. The mean peak concentration of OR-1855 (6.6 ± 4.9 ng/mL) and OR-1896 (7.7 ± 6.3 ng/mL) were reached approximately 140 hours (6 days) after the start of the infusion.
Postoperative Recovery
The patients were extubated in a median (range) of 21 (10 to 792) hours in the levosimendan group compared with 23 (10 to 580) hours in the placebo group (Cox, p = 0.567). The median (range) duration of stay in the ICU was 2 (1 to 33) days in the levosimendan group and 2 (1 to 31) days in the placebo group (Cox, p = 0.658). The median for days alive and out of hospital for 31 days were 20 days in both groups. Two patients, both from the placebo group, died during the study period of ventricular fibrillation and multiorgan failure.
| Comment |
|---|
|
|
|---|
Levosimendan may facilitate weaning from CPB by both its inotropic and lusitropic properties. The direct inotropic effects have been demonstrated by intracoronary injections of levosimendan in failing human hearts [12]. Lusitropic effects of levosimendan (ie, improvement of left ventricular relaxation and filling) have been shown in patients with left ventricular hypertrophy due to aortic stenosis [13]. However, the methods used in this study do not allow identification of the mechanisms involved in improving myocardial performance observed during and after weaning from CPB in patients receiving levosimendan.
After CPB and ischemic cardioplegic arrest, the systolic performance of the heart is invariably depressed by postischemic stunning [14]. Stunning may easily be reversed by standard inotropic agents. Their basic mechanism of inotropic action is an increase in intracellular calcium. The associated increase in myocardial oxygen consumption may be harmful and may provoke ischemia, apoptosis, and cell death [15]. Levosimendan also reverses stunning effectively but it does not increase intracellular calcium [16, 17].
Levosimendan is suggested to exert cardioprotective effects by activation of KATP channels [11,18, 19]. We used total intravenous anesthesia to exclude the interference of inhalational anesthetics on the KATP channels [20]. Interestingly, the increase in postoperative P-troponin T was significantly smaller in the levosimendan group than in the placebo group in our study. Earlier an attenuation of P-troponin I release has been demonstrated in a small study in patients undergoing on-pump CABG receiving levosimendan pretreatment [19].
A few previous studies and case series suggest that levosimendan improves cardiac performance and loading conditions after CPB [21–26]. These studies may implicate an adjunctive role of levosimendan with other cardiac medications, as has been suggested for nesiritide [27].
In a recent study in cardiac surgical patients with a low preoperative ejection fraction, stroke volume was better maintained with the combination of dobutamine and levosimendan than with the combination of dobutamine and milrinone [28]. We showed the effects by performing a placebo-controlled study with levosimendan.
In the present study, the changes of hemodynamic variables after weaning from CPB were largely comparable in the two study groups. However, the prompt use of concomitant treatments to maintain hemodynamic parameters within predetermined limits probably influenced these changes: rescue inotropic medication was given to 63% of the patients in the levosimendan and to 87% of the patients in the placebo group, and PCWP and CVP were targeted to predetermined levels with fluid therapy according to the study protocol.
Still, cumulative phenylephrine requirements to counteract hypotension were higher in the levosimendan group during anesthesia and CPB, and both systemic and pulmonary vascular resistances and also systolic pulmonary artery pressure decreased significantly more over time in the levosimendan group. Levosimendan is known to act on the KATP channels of the peripheral resistance vessels, leading to vasodilation and eventually also hypotension, which might be a problem in vasodilatory conditions such as CPB. In the present study, however, hypotension was easily controlled with phenylephrine and norepinephrine. In the study by Lilleberg and colleagues [29], levosimendan decreased blood pressure and increased heart rate in heart failure patients, as was the case in the present study.
Interestingly, lactate levels were lower in the levosimendan-treated patients in our study, which may reflect improved tissue oxygenation. Lactate may be a relevant prognostic marker for outcome, because using lactate levels of less than 2 mmol/L as a goal to direct hemodynamic optimization in postoperative cardiac patients resulted in a shorter ICU stay and less organ damage [30]. We acknowledge, though, that the use of epinephrine may have influenced the lactate levels.
The plasma levels of the active levosimendan metabolites peaked at 6 days instead of the previously observed 2 to 4 days in the heart failure population [29]. The reason for this altered formation of the metabolites may be related to the use of prophylactic broad-spectrum antibiotics and a consequently diminished amount of intestinal bacteria, which are known to be essential for the formation of the metabolites [31].
Considering the potential clinical effect of our study, the main limitation is that it did not have an active control. Today, most cardiac surgical centers would use milrinone when problems in weaning from CPB are anticipated. Milrinone and other phosphodiesterase inhibitors have been shown to be efficacious for this purpose [1–3]. In experimental studies, however, milrinone increased myocardial oxygen consumption significantly more than levosimendan in equipotent doses [32, 33].
The criteria used for CPB weaning in the present study were more stringent compared with everyday practice and may have resulted in high epinephrine requirements in our patients. However, the criteria were objective and allowed rigorous comparison of the primary efficacy variable between the groups.
Despite the beneficial effect of levosimendan on the primary efficacy variable, no statistically significant differences were detected between the treatment groups in duration of stay in the ICU or the duration of hospitalization or death. In an unblinded study, the combination of dobutamine and levosimendan was associated with shorter intubation time than the combination of dobutamine and milrinone in patients with poor ventricular function after CPB [28]. However, the present study was planned and powered to detect a difference in the primary variable, which was weaning from CPB.
In conclusion, the present randomized, controlled study shows that prophylactic levosimendan facilitates weaning from CPB in patients with impaired preoperative left ventricular function undergoing CABG. However, studies designed and powered to detect relevant clinical outcomes are needed before justifiable practice advisories regarding its use in cardiac surgery can be made.
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |