Levosimendan versus placebo in cardiac surgery: a systematic review and meta-analysis

Levosimendan versus placebo in cardiac surgery: a systematic review and meta-analysis Abstract The aim of this meta-analysis was to review all published randomized clinical trials comparing levosimendan versus placebo in patients undergoing cardiac surgery. PubMed, EMBASE and the Cochrane library database of clinical trials were searched for prospective randomized clinical trials investigating the perioperative use of levosimendan versus placebo in patients undergoing adult cardiac surgery from 1 May 2000 to 10 April 2017. Binary outcomes from individual studies were analysed to compute individual and pooled risk ratios (RRs) with pertinent 95% confidence intervals (CIs). Fourteen randomized clinical trials with a total of 2243 patients were included in this review. Overall meta-analysis results demonstrated that levosimendan was associated with a significant reduction in 30-day mortality (RR = 0.71, 95% CI = 0.53–0.95; P = 0.023). Subgroup analysis showed that this benefit was confined to the moderate and low ejection fraction studies (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001), whereas no benefit was observed in the preserved ejection fraction studies (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78). Levosimendan also reduced the risk of renal replacement therapy (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015) and low cardiac output (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003). No significant differences were detected, between the levosimendan group and the placebo group, in terms of risk of myocardial injury (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44), intensive care unit stay (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055) and the use of ventricular assist device (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35). In conclusion, levosimendan was associated with a reduced risk of mortality, renal replacement therapy and low cardiac output syndrome in patients undergoing cardiac surgery. Levosimendan , placebo , Cardiac surgery , Meta-analysis INTRODUCTION Low cardiac output (LCO) syndrome after cardiac surgery is associated with increased mortality, delayed recovery, organ failure and prolonged intensive care unit stay. This syndrome is characterized by a decreased heart pump function, leading to reduced oxygen delivery and subsequent tissue hypoxia [1], in conjunction with signs of tissue hypoperfusion (cold periphery, clammy skin, confusion, oliguria and elevated lactate level) and in the absence of hypovolaemia [2]. LCO syndrome appears in approximately 20% of cardiopulmonary bypass surgeries [3]. Levosimendan, a calcium-sensitizing inotrope and an ATP-sensitive potassium channel opener, has been reported to be effective in decreasing LCO syndrome and mortality after cardiac surgery [4, 5]. The recent publication of 2 large randomized clinical trials (RCTs) [6, 7], which were unable to document any benefit of levosimendan compared with placebo in terms of survival, has strongly reopened the debate about its efficacy. The aim of this study was to evaluate the effect of levosimendan in comparison with placebo in adult patients undergoing cardiac surgery using a meta-analysis of RCTs. MATERIALS AND METHODS Search strategy PubMed, EMBASE and the Cochrane library database of clinical trials were searched for randomized clinical trials investigating the perioperative use of levosimendan versus placebo in patients undergoing cardiac surgery. The search terms used were ‘Levosimendan AND Cardiac Surgery’. The search was combined with filters to identify clinical trials, meta-analysis and reviews in the PubMed and EMBASE database. The filters for EMBASE searches were Cochrane review, systematic review, meta-analysis, randomized controlled trial, adult, humans, abstracts and year of publication 2000–2017. The filters for PubMed were clinical trial, meta-analysis, randomized controlled trial, review, systematic reviews and publication dates from 1 May 2000 to 10 April 2017. Searches were restricted to articles published in English. Search strategy for PubMed is described in Supplementary Material, Data S1, Table 1. Table 1: Description of the studies included in the meta-analysis First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days AVR: aortic valve replacement; CABG: coronary artery bypass graft; CBP: cardiopulmonary bypass; ECMO: extracorporeal membrane oxygenation; IABP: intra-aortic balloon pump; ICU: intensive care unit; inf: infusion; Levo: levosimendan; LV: left ventricular; LVEF: left ventricular ejection fraction; MV: mitral valve; MVR: mitral valve replacement; NR: not registered. Table 1: Description of the studies included in the meta-analysis First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days AVR: aortic valve replacement; CABG: coronary artery bypass graft; CBP: cardiopulmonary bypass; ECMO: extracorporeal membrane oxygenation; IABP: intra-aortic balloon pump; ICU: intensive care unit; inf: infusion; Levo: levosimendan; LV: left ventricular; LVEF: left ventricular ejection fraction; MV: mitral valve; MVR: mitral valve replacement; NR: not registered. Study selection Two investigators reviewed all available abstracts for potential inclusion. Studies that were prospective, randomized controlled trials (RCT) of perioperative administration of levosimendan in adults undergoing cardiac surgery were subjected for a full text review. Abstracts by identical authors were cross-checked to ensure that data were not duplicated. Divergences were resolved by consensus. Previously published meta-analyses and review articles were scanned to evaluate articles for inclusion. Population, intervention, comparison, outcomes and study design (PICOS) were the following: Participants: patients undergoing cardiac surgery Intervention: levosimendan Comparison: placebo Outcomes: mortality, renal replacement therapy (as per author definition), in-hospital myocardial injury (as per author definition), intensive care unit stay, low cardiac output syndrome (as per author definition) and the use of ventricular assist device Study design: RCTs The following inclusion criteria were used: randomized clinical trials comparing levosimendan treatment with placebo in adult patients undergoing cardiac surgery with cardiopulmonary bypass. The exclusion criteria were paediatric or congenital population, cardiac surgery carried out exclusively without cardiopulmonary bypass, cardiac surgery other than valve surgery or coronary surgery, animal studies and studies that did not include our primary end point or secondary end points. Reviews and meta-analysis were also excluded. If the abstract suggested that the study could potentially meet the inclusion criteria, then the full text article was reviewed. Data abstraction and study characteristics The following study characteristics were recorded: last name of the first author, year of publication, design details, number of participants, timing and dosing of levosimendan administration and duration of follow-up. Baseline population characteristics including type of cardiac surgery and preoperative left ventricular ejection fraction also were recorded (Table 1). Internal validity and risk of bias assessment The internal validity and risk of bias of included trials were appraised according to the Cochrane collaboration methods [8] by 2 independent reviewers with divergences resolved by consensus. Data analysis Computations were carried out using Statistical Package R, specifically the Meta package and MedCalc Statistical Software version 17.9.6 (MedCalc Software bvba, Ostend, Belgium; http://www.medcalc.org; 2017). Hypothesis of statistical heterogeneity was tested using Cochran Q test ( Xn2, where n is degrees of freedom), and inconsistency was measured using I2. Binary outcomes from individual studies were analysed to compute individual and pooled risk ratios (RRs) with pertinent 95% confidence intervals (CIs) [with equivalence set at 1, RR <1 favouring the levosimendan treatment and RR >1 favouring the control group using the Mantel–Haenszel fixed-effect method in case of low statistical inconsistency (I2 ≤25%) and using a random-effect method in case of high or moderate statistical inconsistency (I2 >25%)]. Sensitivity analysis was performed by removing each study, one by one, and repeating the meta-analysis for the primary end point to check for the influence of individual results on the overall results. Meta-analysis of the primary end point for 30-day mortality was divided in 2 sub-analyses: analysis carried out on studies including patients with moderate or low left ventricle ejection fraction (LVEF) and analysis carried out on studies including patients with preserved ejection fraction. Weighted mean differences and 95% CIs were computed for continuous variables using the same methods as just described. The risk of publication bias was assessed by visual inspection of funnel plots and linear regression analysis. Forest plots were built to provide a graphical representation of the analysis. The RR of each study is represented as a square (with lines extending to 95% CI) and the overall pooled RR as a diamond. The weight is the percentage that each individual study contributes to eventual outcome. The size of the square of each study in forest plot is proportional to the weight of that study in the analysis. Statistical significance was set at the 2-tailed 0.05 level for hypothesis testing and at 0.10 for heterogeneity testing. Unadjusted P-values are reported. This study was carried out in compliance with the Cochrane Collaboration, the Quality of Reporting of Meta-Analyses guidelines (QUOROM) [9] and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [10]. RESULTS The search strategy yielded 484 abstracts. Of these, 447 were excluded, and 37 were analysed as full text articles (Fig. 1). Among these, 14 were included [6, 7, 11–22] in our meta-analysis (Table 1), and 23 were excluded. Studies were excluded for the following reasons: 6 studies did not report interest end points [23–28], 5 studies compared the levosimendan group with the control group that was not placebo [29–33], 2 articles were study protocols [34, 35], 1 article was not in cardiac surgery setting [36], 1 article was about patients with end-stage renal disease [37], 1 article was a letter to the editor [38], 4 articles were reviews [39–42], 1 article was about off-pump cardiac surgery [43], 1 article was not a randomized trial [44] and 1 article [45] was a sub-study of the Eriksson trial [22]. Figure 1: View largeDownload slide Flow diagram. Figure 1: View largeDownload slide Flow diagram. Study characteristics The 14 included RCTs randomized 2243 patients (1122 to levosimendan and 1121 to control). Levosimendan was administered as bolus followed by continuous infusions (6 RCTs), as continuous infusion only (6 RCTs) or as bolus only (2 RCTs) (Table 1). The dose varied from 10 to 24 µg/kg for intravenous bolus and from 0.1 to 0.2 µg/kg/min for continuous infusion (Supplementary Material, Data S1, Table 2). Five RCTs were multicentric [6, 7, 11, 12, 22] and 9 were unicentric [13–21]. Four RCTs enrolled exclusively patients with a preoperative LVEF <35% [6, 12, 16, 20]. Two studies did not report a median or mean ejection fraction: 1 study reported that 73% of patients in the placebo group and 77% of patients in the levosimendan group had an LVEF >50% [18] and another study reported that all study patients had left ventricle dysfunction with an LVEF >25% [11]. Five studies reported an LVEF of 36–50% in the preoperative sample population [7, 14, 17, 21, 22]. Three RCTs documented an LVEF of >50% in a sample population [13, 14, 19]. Study quality appraisal indicated that studies were of optimal quality, and 10 of them had a low risk of bias (Supplementary Material, Data S1, Table 3). In both the levosimendan group and the placebo group, additional standard inotropic treatment was administered according to hospital protocols. Quantitative data synthesis Overall analysis showed that the relative risk for postoperative mortality at 30 days was reduced by 29% in the levosimendan group when compared with the control group [67 of 1122 (6.0%) in the levosimendan group vs 96 of 1121 (8.6%) in the placebo group] (RR = 0.71, 95% CI = 0.53–0.95; P = 0.023; I2 = 22%; p for heterogeneity = 0.23; with 2243 patients and 14 studies included) (forest plot, Fig. 2). The funnel plot (Supplementary Material, Data S2, Fig. 1) and linear regression analysis showed no significative asymmetry (P = 0.15), and thus, no publication bias was detected by these methods. Subgroup analysis showed that survival benefit was confined to the low and moderate LVEF studies [23 of 739 (3.1%) in the levosimendan group vs 53 of 726 (7.3%) in the placebo group] (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001; I2 = 0%, p for heterogeneity = 0.47, with 1465 patients and 9 studies included). Survival benefit was not observed in the preserved LVEF studies [44 of 383 (11.5%) in the levosimendan group vs 43 of 395 (10.9%) in the placebo group] (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78; I2 = 0%; p for heterogeneity = 0.82, with 778 patients and 5 studies included). The sensitivity analysis (Supplementary Material, Data S2, Fig. 2) did not change the final result. Figure 2: View largeDownload slide Forest plot of the mortality meta-analysis. CI: confidence interval; EF: ejection fraction; RR: risk ratio. Figure 2: View largeDownload slide Forest plot of the mortality meta-analysis. CI: confidence interval; EF: ejection fraction; RR: risk ratio. In the levosimendan group, a significant reduction in the rate of renal replacement therapy was observed [49 of 995 (4.9%) in the levosimendan group vs 76 of 997 (7.6%) in the placebo group] (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015; I2 = 0%; p for heterogeneity = 0.78, with 1992 patients and 9 studies included) (forest plot, Fig. 3). The funnel plot (Supplementary Material, Data S2, Fig. 3) and linear regression analysis did not show asymmetry (P = 0.13), thus no publication bias was detected by these methods. The sensitivity analysis showed the same benefit (Supplementary Material, Data S2, Fig. 4). Figure 3: View largeDownload slide Forest plot of the renal replacement therapy meta-analysis. CI: confidence interval; RR: risk ratio. Figure 3: View largeDownload slide Forest plot of the renal replacement therapy meta-analysis. CI: confidence interval; RR: risk ratio. Figure 4: View largeDownload slide Forest plot of the myocardial injury meta-analysis. CI: confidence interval; RR: risk ratio. Figure 4: View largeDownload slide Forest plot of the myocardial injury meta-analysis. CI: confidence interval; RR: risk ratio. No significant reduction in risk of postoperative myocardial injury was detected in the levosimendan group [90 of 796 (11.3%)] vs the placebo group [99 of 787 (12.6%)] (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44; I2 = 0%; p for heterogeneity = 0.55, with 1583 patients and 9 studies included) (forest plot, Fig. 4). The funnel plot (Supplementary Material, Data S2, Fig. 5) and linear regression analysis demonstrated asymmetry (P = 0.043) showing a possible publication bias. The sensitivity analysis did not change the final result (Supplementary Material, Data S2, Fig. 6). Figure 5: View largeDownload slide Forest plot of the low cardiac output syndrome meta-analysis. CI: confidence interval; RR: risk ratio. Figure 5: View largeDownload slide Forest plot of the low cardiac output syndrome meta-analysis. CI: confidence interval; RR: risk ratio. Figure 6: View largeDownload slide Forest plot of ventricular assist device implantation meta-analysis. CI: confidence interval; RR: risk ratio. Figure 6: View largeDownload slide Forest plot of ventricular assist device implantation meta-analysis. CI: confidence interval; RR: risk ratio. Low cardiac output syndrome was reported in 6 studies. Levosimendan was shown to be effective in reducing low cardiac output syndrome when compared with placebo [107 of 723 (14.8%) in the levosimendan group vs 207 of 715 (29.0%) in the placebo group] (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003; I2 = 75%; p for heterogeneity = 0.003, with 1438 patients included and 6 RCTs) (forest plot, Fig. 5). The funnel plot (Supplementary Material, Data S2, Fig. 7) and linear regression analysis did not show asymmetry (P = 0.27), thus no publication bias was detected by these methods. The sensitivity analysis (Supplementary Material, Data S2, Fig. 8) did not change the final result. As a significative heterogeneity was detected in low cardiac output syndrome meta-analysis (I2 = 75%), we repeated the analysis omitting the largest trial [6]. The meta-analysis of the remaining 5 RCTs showed no heterogeneity (I2 = 0) and confirmed a significant reduction in risk of low cardiac output syndrome in the levosimendan group (Fig. 5). We carried out the analysis applying fixed effects model and random effects model obtaining the same results (RR = 0.29, 95% CI = 0.20–0.43; P < 0.001). The omitted trial [6] analysis also showed a significant reduction in terms of low cardiac output syndrome in the levosimendan group (RR = 0.71, 95% CI = 0.55–0.92). Figure 7: View largeDownload slide Forest plot of the intensive care unit stay meta-analysis. CI: confidence interval; MD: mean difference; SD: standard deviation. Figure 7: View largeDownload slide Forest plot of the intensive care unit stay meta-analysis. CI: confidence interval; MD: mean difference; SD: standard deviation. The levosimendan group did not show a significant reduction in risk of using ventricular assist device [47 of 572 (8.2%) in the levosimendan group vs 45 of 562 (8.0%) in the placebo group] (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35; I2 = 59%; p for heterogeneity = 0.088, with 1134 patients and 3 studies included) (forest plot, Fig. 6). The funnel plot (Supplementary Material, Data S2, Fig. 9) and linear regression analysis showed asymmetry (P = 0.026), documenting a possible publication bias. As a significative heterogeneity was detected in ventricular assist device meta-analysis (I2 = 59%), we repeated the analysis omitting the largest trial [6]. The meta-analysis of the remaining 2 RCTs showed no heterogeneity (I2 = 0) and detected a beneficial result for the levosimendan group (RR = 0.12, 95% CI = 0.02–0.94; P = 0.044) (Fig. 6). However, the analysis of the trial by Mehta et al. [6] did not show significant results (RR = 1.22, 95% CI = 0.81–1.83). (The sensitivity analysis is represented in Supplementary Material, Data S2, Fig. 10.) No significant reduction in intensive care unit (ICU) stay in the levosimendan group was observed (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055; I2 = 91%; p for heterogeneity <0.001, with 6 studies included) (forest plot, Fig. 7). The funnel plot (Supplementary Material, Data S2, Fig. 11) and linear regression analysis did not demonstrate asymmetry (P = 0.61), thus no publication bias was detected using these methods. Sensitivity analysis showed the same result (Supplementary Material, Data S2, Fig. 12). We also carried out a subgroup analysis to check whether the time of administration, dose and the presence of loading bolus influence the survival results. We carried out this analysis in patients with moderate or severe dysfunction because this group was the one benefitting from levosimendan administration in terms of survival. With regard to time of levosimendan administration, a statistically significant benefit was documented by the preoperative administration over the intraoperative one (RR = 0.46, 95% CI = 0.28–0.74; P = 0.002; I2 = 22%; p for heterogeneity = 0.28, with 1250 patients included and 5 RCTs), (Supplementary Material, Data S2, Fig. 13). With regard to the doses, 0.1 µg/kg/min was detected to be effective over 0.2 µg/kg/min (RR = 0.46, 95% CI = 0.28–0.74; P = 0.002; I2 = 22%; p for heterogeneity = 0.27, with 1263 patients included and 6 RCTs) (Supplementary Material, Data S2, Fig. 14). Also, we documented a beneficial effect of the initial loading bolus of levosimendan over the absence of loading dose (RR = 0.24, 95% CI = 0.10–0.56; P < 0.001; I2 = 0%; p for heterogeneity = 0.71, with 511 patients included and 5 RCTs) (Supplementary Material, Data S2, Fig. 15). DISCUSSION This meta-analysis shows that perioperative use of levosimendan is associated with a reduction in 30-day postoperative mortality in adult patients undergoing cardiac surgery. In particular, we documented this benefit in RCTs enrolling patients with low or moderate LVEF (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001), but not in RCTs including patients with preserved LVEF (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78). Furthermore, our analysis found levosimendan to be associated with a reduction in renal replacement therapy (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015) and low cardiac output syndrome (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003) compared with placebo. This meta-analysis failed to show any benefit of levosimendan in terms of decrease in myocardial injury (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44), ICU stay (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055) and the use of ventricular assist device (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35). These results are consistent with those of the previous 12 meta-analysis carried out on levosimendan used in surgical setting [46–57]. The first 10 of them, which did not include the 2 largest RCTs [6, 7], have been systematically reviewed by Pollesello et al. [58] in 2016. Briefly, levosimendan was reported to decrease postoperative mortality after cardiac surgery in the meta-analysis by Landoni et al. [47] [440 patients from 10 studies; odds ratio (OR)  = 0.35, 95% CI = 0.18–0.71; P = 0.003], Maharaj and Metaxa [48] (729 patients from 17 studies; OR = 0.40, 95% CI = 0.21–0.76; P = 0.005), Hernández et al. [49] (654 patients from 13 studies; OR = 0.36, 95% CI = 0.20–0.64; P = 0.001) and Zhou et al. [55] (1345 patients from 13 studies; OR = 0.41, 95% CI = 0.27–0.62; P < 0.001). Two meta-analysis showed, as we did, a beneficial effect in terms of postoperative survival only in the low ejection fraction trials: Harrison et al. [50] (1155 patients from 14 studies; risk difference −7.0%, 95% CI = −11.0% to −3.1%; P < 0.001) and Lim et al. [54] (965 patients from 14 studies; OR = 0.41, 95% CI = 0.24–0.77; P = 0.004). Levosimendan was associated with a lower incidence of acute kidney injury in meta-analysis by Niu et al. [51] (529 patients from 5 studies; OR = 0.44, 95% CI = 0.22–0.85; P = 0.02); Bove et al. [52] (3879 patients from 33 studies; RR = 0.79, 95% CI = 0.63–0.99; P = 0.048), Lim et al. [54] (965 patients from 14 studies; OR = 0.62, 95% CI = 0.40–0.95; P = 0.05) and Zhou et al. [55] (1345 patients from 13 studies OR = 0.51, 95% CI = 0.34–0.76; P = 0.001). Chen et al. [57] and Sanfilippo et al. [56] proposed a meta-analysis including the 2 largest trials [6, 7]. Both of them concluded that levosimendan decreases mortality in patients with preoperative ventricular systolic dysfunction. Our meta-analysis is based on 14 studies: 2 of them [6, 7] are large trials (sample population ≥506 patients), 2 midsize studies [12, 28] (sample population ≥252 patients) and 10 small studies [11, 13–17, 19–22] (sample population from 24 to 102 patients). Basically, large and midsize RCTs investigated clinical primary outcome (mortality, LCO syndrome or renal failure incidence), whereas small trials have analysed mostly haemodynamic outcomes (cardiac index and ejection fraction) or inotropic drug requirement. All small RCTs reported improved haemodynamic outcomes in patients treated with levosimendan. Mortality data analysed by small RCTs reached no statistical significance, likely because of the very limited number of events registered. Anyway, even not significant, small RCTs reported less mortality events in the levosimendan group in 6 studies over 10 (Fig. 2). Of the 2 midsized trials, 1 [18] was carried out on patients with preserved left ventricular function and reported no significant difference in mortality. The second one [12] was carried out on patients with left ventricle dysfunction and reported a significant decrease in mortality in those patients treated with levosimendan (P < 0.05). The 2 largest trials [6, 7] failed to show any difference in terms of 30-day mortality between the levosimendan and the placebo groups. In fact, the publication of these 2 large and solid trials has raised doubts about the utility of levosimendan in patients undergoing cardiac surgery. Even if apparently similar, these 2 trials present methodological and outcome analysis differences, which deserve further discussion. Landoni et al. [7] enrolled 506 patients. Inclusion criteria were preoperative LVEF <25%, high dose of inotropic support because of acute postoperative cardiac dysfunction (this being an indication of very late use of levosimendan) or preoperative intra-aortic balloon pump. Thus, most of the patients were enrolled in this study because of postoperative cardiogenic shock (65%), and only 4.3% of the population was randomized preoperatively because of low LVEF. In particular, among patients of the levosimendan group (n = 248), only 25.8% (n = 64) had left ventricle dysfunction (LVEF 25–40%). Because of the presence of a significant proportion of patients with postoperative cardiogenic shock, levosimendan was administered as infusion, without loading dose, to avoid arterial hypotension. Under such heterogenous circumstances, no differences were detected in 30- or 180-day mortality nor in the other explored outcomes. Mehta et al. [6] enrolled 849 patients; inclusion criterion was just 1: LVEF <35% documented 60 days before surgery. Just before surgery, loading and infusion doses were administered. Patients treated with levosimendan showed a significantly reduced incidence of LCO syndrome when compared with those treated with placebo (P = 0.007). Levosimendan was not associated with a reduced 30-day mortality (P = 0.45), but it showed a non-significative association (P = 0.12) with a reduced 90-day mortality. Our meta-analysis shows that the inclusion of these 2 trials does not neutralize the overall beneficial effect of levosimendan documented in terms of decreased mortality, renal replacement therapy and LCO syndrome RR. For this reason, the results of our meta-analysis suggest that perioperative levosimendan still represents an effective treatment in patients undergoing cardiac surgery with low LVEF. Limitations This meta-analysis carries the potential bias of this specific analysis format and those of the primary studies. In particular, in the ventricular assist device analysis, only 3 studies were included; high heterogeneity and possible publication bias, showed by the funnel plot and linear regression, were documented. The myocardial damage analysis was jeopardized by the different definitions of this clinical entity provided in the included studies. An ICU stay analysis showed a significant heterogeneity. Moreover, half of the studies in this analysis were small clinical trials, with sample size less than 50 patients. In addition, 3 of the studies had moderate risk of bias and 1 had high risk of bias. Subgroup meta-analysis on the different strategies of levosimendan administration is biased by the fact that those were applied to populations with different risk profiles. SUPPLEMENTARY MATERIAL Supplementary material is available at ICVTS online. Conflict of interest: none declared. REFERENCES 1 Vincent J-L , De Backer D. Circulatory shock . N Engl J Med 2013 ; 369 : 1726 – 34 . Google Scholar CrossRef Search ADS 2 Algarni KD , Maganti M , Yau TM. Predictors of low cardiac output syndrome after isolated coronary artery bypass surgery: trends over 20 years . Ann Thorac Surg 2011 ; 92 : 1678 – 84 . Google Scholar CrossRef Search ADS 3 Benjamin EJ , Blaha MJ , Chiuve SE , Cushman M , Das SR , Deo R. American Heart Association Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics-2017 update: a report from the American Heart Association . Circulation 2017 ; 135 : e146 – 603 . Google Scholar CrossRef Search ADS 4 Sunny YM , Karim HM , Saikia MK , Bhattacharyya P , Dey S. Comparison of Levosimendan, milrinone and dobutamine in treating low cardiac output syndrome following valve replacement surgeries with cardiopulmonary bypass . J Clin Diagn Res 2016 ; 10 : UC05 – 8 . 5 Toller W , Algotsson L , Guarracino F , Hormann C , Knotzer J , Lehmann A et al. Perioperative use of levosimendan: best practice in operative settings . J Cardiothorac Vasc Anesth 2013 ; 27 : 361 – 6 . Google Scholar CrossRef Search ADS 6 Mehta RH , Leimberger JD , van Diepen S , Meza J , Wang A , Jankowich R et al. Levosimendan in patients with left ventricular dysfunction undergoing cardiac surgery . N Engl J Med 2017 ; 376 : 2032 – 42 . Google Scholar CrossRef Search ADS 7 Landoni G , Lomivorotov VV , Alvaro G , Lobreglio R , Pisano A , Guarracino F et al. Levosimendan for hemodynamic support after cardiac surgery . N Engl J Med 2017 ; 376 : 2021 – 31 . Google Scholar CrossRef Search ADS 8 Higgins JPT , Green S. Cochrane Handbook for Systematic Reviews of Interventions. The Cochrane Collaboration. Version 5.1.0. http://handbook-5-1.cochrane.org (updated March 2011). 9 Moher D , Cook DJ , Eastwood S , Olkin I , Rennie D , Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of reporting of meta-analyses . Lancet 1999 ; 354 : 1896 – 900 . Google Scholar CrossRef Search ADS 10 Knobloch K , Yoon U , Vogt PM. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement and publication bias . J Craniomaxillofac Surg 2011 ; 39 : 91 – 2 . Google Scholar CrossRef Search ADS 11 Levin R , Degrange M , Porcile R , Salvagio F , Del Mazo C , Tanus E et al. Preoperative use of calcium sensitizer levosimendan reduces mortality and low cardiac output syndrome in patients with aortic stenosis and left ventricular dysfunction. 2008 World Congress of Cardiology Abstracts Circulation 2008 ; 118 : P322 . 12 Levin R , Degrange M , Del Mazo C , Tanus E , Porcile R. Preoperative levosimendan decreases mortality and the development of low cardiac output in high-risk patients with severe left ventricular dysfunction undergoing coronary artery bypass grafting with cardiopulmonary bypass . Exp Clin Cardiol 2012 ; 17 : 125 – 30 . 13 Leppikangas H , Järvelä K , Sisto T , Maaranen P , Virtanen M , Lehto P et al. Preoperative levosimendan infusion in combined aortic valve and coronary bypass surgery . Br J Anaesth 2011 ; 106 : 298 – 304 . Google Scholar CrossRef Search ADS 14 Tritapepe L , De Santis V , Vitale D , Santulli M , Morelli A , Nofroni I et al. Preconditioning effects of Levosimendan in coronary artery bypass grafting—a pilot study . Br J Anaesth 2006 ; 96 : 694 – 700 . Google Scholar CrossRef Search ADS 15 Tritapepe L , De Santis V , Vitale D , Guarracino F , Pellegrini F , Pietropaoli P et al. Levosimendan pre-treatment improves outcomes in patients undergoing coronary artery bypass graft surgery . Br J Anaesth 2009 ; 102 : 198 – 204 . Google Scholar CrossRef Search ADS 16 Erb J , Beutlhauser T , Feldheiser A , Schuster B , Treskatsch S , Grubitzsch H et al. Influence of levosimendan on organ dysfunction in patients with severely reduced left ventricular function undergoing cardiac surgery . J Int Med Res 2014 ; 42 : 750 – 64 . Google Scholar CrossRef Search ADS 17 Anastasiadis K , Antonitsis P , Vranis K , Kleontas A , Asteriou C , Grosomanidis V et al. Effectiveness of prophylactic Levosimendan in patients with impaired left ventricular function undergoing coronary artery bypass grafting: a randomized pilot study . Interact CardioVasc Thorac Surg 2016 ; 23 : 740 – 7 . Google Scholar CrossRef Search ADS 18 Lahtinen P , Pitkänen O , Pölönen P , Turpeinen A , Kiviniemi V , Uusaro A. Levosimendan reduces heart failure after cardiac surgery: a prospective, randomized, placebo-controlled trial . Crit Care Med 2011 ; 39 : 2263 – 70 . Google Scholar CrossRef Search ADS 19 Järvelä K , Maaranen P , Sisto T , Ruokonen E. Levosimendan in aortic valve surgery: cardiac performance and recovery . J Cardiothorac Vasc Anesth 2008 ; 22 : 693 – 8 . Google Scholar CrossRef Search ADS 20 Sharma P , Malhotra A , Gandhi S , Garg P , Bishnoi A , Gandhi H. Preoperative levosimendan in ischemic mitral valve repair . Asian Cardiovasc Thorac Ann 2014 ; 22 : 539 – 45 . Google Scholar CrossRef Search ADS 21 Ersoy O , Boysan E , Unal EU , Yay K , Yener U , Cicekcioglu F et al. Effectiveness of prophylactic levosimendan in high-risk valve surgery patients . Cardiovasc J Afr 2013 ; 24 : 260 – 4 . Google Scholar CrossRef Search ADS 22 Eriksson HI , Jalonen JR , Heikkinen LO , Kivikko M , Laine M , Leino KA et al. Levosimendan facilitates weaning from cardiopulmonary bypass in patients undergoing coronary artery bypass grafting with impaired left ventricular function . Ann Thorac Surg 2009 ; 87 : 448 – 54 . Google Scholar CrossRef Search ADS 23 Asaad OM , Hanafy MS. Levosimendan’s effect on coronary artery grafts blood flow in patients with left ventricular dysfunction, assessment by transit time flow meter . Egypt J Anaesth 2011 ; 27 : 45 – 53 . Google Scholar CrossRef Search ADS 24 Juhl-Olsen P , Jakobsen C-J , Rasmussen LA , Bhavsar R , Klaaborg K-E , Frederiksen CA et al. Effects of levosimendan in patients with left ventricular hypertrophy undergoing aortic valve replacement . Acta Anaesthesiol Scand 2015 ; 59 : 65 – 77 . Google Scholar CrossRef Search ADS 25 Jörgensen K , Bech-Hanssen O , Houltz E , Ricksten SE. Effects of levosimendan on left ventricular relaxation and early filling at maintained preload and afterload conditions after aortic valve replacement for aortic stenosis . Circulation 2008 ; 117 : 1075 – 81 . Google Scholar CrossRef Search ADS 26 Bragadottir G , Redfors B , Ricksten SE. Effects of levosimendan on glomerular filtration rate, renal blood flow, and renal oxygenation after cardiac surgery with cardiopulmonary bypass: a randomized placebo-controlled study . Crit Care Med 2013 ; 41 : 2328 – 35 . Google Scholar CrossRef Search ADS 27 Aksel’rod BA , Tolstova IA , Trekova NA , Kolpakov PE , Babaev MA , Belianko IE. Impact of preoperative levosimendan therapy on the volemic status and vascular tone of patients with chronic heart failure during anesthesia . Anesteziol Reanimatol 2009 ; 6 : 46 – 51 . 28 Abacilar AF , Dogan OF. Levosimendan use decreases atrial fibrillation in patients after coronary artery bypass grafting: a pilot study . Heart Surg Forum 2013 ; 16 : 287 – 94 . Google Scholar CrossRef Search ADS 29 Carev M , Karanovic N , Kocen D , Bulat C. Useful supplement to the best practice of using levosimendan in cardiac surgery patients: 2.5-mg intravenous bolus for cardiopulmonary resuscitation during perioperative cardiac arrest . J Cardiothorac Vasc Anesth 2013 ; 27 : e75 – 7 . Google Scholar CrossRef Search ADS 30 De Hert SG , Lorsomradee S , Vanden Eede H , Cromheecke S , Van der Linden PJ. A randomized trial evaluating different modalities of levosimendan administration in cardiac surgery patients with myocardial dysfunction . J Cardiothorac Vasc Anesth 2008 ; 22 : 699 – 705 . Google Scholar CrossRef Search ADS 31 Baysal A , Yanartas M , Dogukan M , Gundogus N , Kocak T , Koksal C. Levosimendan improves renal outcome in cardiac surgery: a randomized trial . J Cardiothorac Vasc Anesth 2014 ; 28 : 586 – 94 . Google Scholar CrossRef Search ADS 32 Tasouli A , Papadopoulos K , Antoniou T , Kriaras I , Stavridis G , Degiannis D et al. Efficacy and safety of perioperative infusion of levosimendan in patients with compromised cardiac function undergoing open-heart surgery: importance of early use . Eur J Cardiothorac Surg 2007 ; 32 : 629 – 33 . Google Scholar CrossRef Search ADS 33 Alvarez J , Bouzada M , Fernández AL , Caruezo V , Taboada M , Rodríguez J et al. Hemodynamic effects of levosimendan compared with dobutamine in patients with low cardiac output after cardiac surgery . Rev Esp Cardiol 2006 ; 59 : 338 – 45 . Google Scholar CrossRef Search ADS 34 Mehta RH , Van Diepen S , Meza J , Bokesch P , Leimberger JD , Tourt-Uhlig S et al. Levosimendan in patients with left ventricular systolic dysfunction undergoing cardiac surgery on cardiopulmonary bypass: rationale and study design of the levosimendan in patients with left ventricular systolic dysfunction undergoing cardiac surgery requiring cardiopulmonary bypass (LEVO-CTS) trial . Am Heart J 2016 ; 182 : 62 – 71 . Google Scholar CrossRef Search ADS 35 Caruba T , Hourton D , Sabatier B , Rousseau D , Tibi A , Hoffart-Jourdain C et al. Rationale and design of the multicenter randomized trial investigating the effects of levosimendan pretreatment in patients with low ejection fraction (≤40%) undergoing CABG with cardiopulmonary bypass (LICORN study) . J Cardiothorac Surg 2016 ; 11 : 127 . Google Scholar CrossRef Search ADS 36 Adamopoulos S , Parissis JT , Iliodromitis EK , Paraskevaidis I , Tsiapras D , Farmakis D et al. Effects of levosimendan versus dobutamine on inflammatory and apoptotic pathways in acutely decompensated chronic heart failure . Am J Cardiol 2006 ; 98 : 102 – 6 . Google Scholar CrossRef Search ADS 37 Atalay H , Temizturk Z , Azboy D , Colak S , Atalay A , Dogan OF et al. Levosimendan use increases cardiac performance after coronary artery bypass grafting in end-stage renal disease patients . Heart Surg Forum 2016 ; 19 : 230 – 6 . Google Scholar CrossRef Search ADS 38 De Santis V , Vitale D , Tritapepe L. Levosimendan and cardiac surgery . J Cardiothorac Vasc Anesth 2010 ; 24 : 210. Google Scholar CrossRef Search ADS 39 Elahi MM , Lam J , Asopa S , Matata BM. Levosimendan versus an intra-aortic balloon pump in adult cardiac surgery patients with low cardiac output . J Cardiothorac Vasc Anesth 2011 ; 25 : 1154 – 62 . Google Scholar CrossRef Search ADS 40 van den Brule J , Hoedemaekers C , Pickkers P. Clinical outcome benefits of pretreatment with levosimendan . Br J Anaesth 2009 ; 102 : 883 – 4 . Google Scholar CrossRef Search ADS 41 Toller W , Knez I. Medical support and surgery of the failing heart: levosimendan . Scand J Surg 2007 ; 96 : 121 – 4 . Google Scholar CrossRef Search ADS 42 García-González MJ , Domínguez-Rodríguez A. Effect of levosimendan treatment of myocardial stunning and low-output syndrome after cardiac surgery . Rev Esp Cardiol 2006 ; 59 : 851 – 2 . Google Scholar CrossRef Search ADS 43 Husedzinović I , Barisin S , Bradić N , Barisin A , Sonicki Z , Milanović R. Levosimendan as a new strategy during off-pump coronary artery bypass grafting: double-blind randomized placebo-controlled trial . Croat Med J 2005 ; 46 : 950 – 6 . 44 Temizturk Z , Azboy D , Atalay A , Atalay H , Dogan OF. The effects of levosimendan and sodium nitroprusside combination on left ventricular functions after surgical ventricular reconstruction in coronary artery bypass grafting patients . Open Cardiovasc Med J 2016 ; 10 : 138 – 47 . Google Scholar CrossRef Search ADS 45 Ristikankare A , Pöyhiä R , Eriksson H , Valtonen M , Leino K , Salmenperä M. Effects of levosimendan on renal function in patients undergoing coronary artery surgery . J Cardiothorac Vasc Anesth 2012 ; 26 : 591 – 5 . Google Scholar CrossRef Search ADS 46 Zangrillo A , Biondi-Zoccai G , Mizzi A , Bruno G , Bignami E , Gerli C et al. Levosimendan reduces cardiac troponin release after cardiac surgery: a meta-analysis of randomized controlled studies . J Cardiothorac Vasc Anesth 2009 ; 23 : 474 – 8 . Google Scholar CrossRef Search ADS 47 Landoni G , Mizzi A , Biondi-Zoccai G , Bruno G , Bignami E , Corno L et al. Reducing mortality in cardiac surgery with levosimendan: a meta-analysis of randomized controlled trials . J Cardiothorac Vasc Anesth 2010 ; 24 : 51 – 7 . Google Scholar CrossRef Search ADS 48 Maharaj R , Metaxa V. Levosimendan and mortality after coronary revascularisation: a meta-analysis of randomised controlled trials . Crit Care 2011 ; 15 : R140. Google Scholar CrossRef Search ADS 49 Hernández A , Miranda A , Parada A. Levosimendan reduces mortality in cardiac surgery: a systematic review and meta-analysis . Rev Esp Anestesiol Reanim 2012 ; 59 : 6 – 11 . Google Scholar CrossRef Search ADS 50 Harrison RW , Hasselblad V , Mehta RH , Levin R , Harrington RA , Alexander JH. Effect of levosimendan on survival and adverse events after cardiac surgery: a meta-analysis . J Cardiothorac Vasc Anesth 2013 ; 27 : 1224 . Google Scholar CrossRef Search ADS 51 Niu ZZ , Wu SM , Sun WY , Hou WM , Chi YF. Perioperative levosimendan therapy is associated with a lower incidence of acute kidney injury after cardiac surgery: a meta-analysis . J Cardiovasc Pharmacol 2014 ; 63 : 107 – 12 . Google Scholar CrossRef Search ADS 52 Bove T , Matteazzi A , Belletti A , Paternoster G , Saleh O , Taddeo D et al. Beneficial impact of levosimendan in critically ill patients with or at risk for acute renal failure: a meta-analysis of randomized clinical trials . Heart Lung Vessel 2015 ; 7 : 35 – 46 . 53 Greco T , Calabrò MG , Covello RD , Greco M , Pasin L , Morelli A et al. A Bayesian network meta-analysis on the effect of inodilatory agents on mortality . Br J Anaesth 2015 ; 114 : 746 – 56 . Google Scholar CrossRef Search ADS 54 Lim JY , Deo SV , Rababa’h A , Altarabsheh SE , Cho YH , Hang D et al. Levosimendan reduces mortality in adults with left ventricular dysfunction undergoing cardiac surgery: a systematic review an meta-analysis . J Card Surg 2015 ; 30 : 547 – 54 . Google Scholar CrossRef Search ADS 55 Zhou C , Gong J , Chen D , Wang W , Liu M , Liu B. Levosimendan for prevention of acute kidney injury after cardiac surgery: a meta-analysis of randomize controlled trials . Am J Kidney Dis 2016 ; 67 : 408 – 16 . Google Scholar CrossRef Search ADS 56 Sanfilippo F , Knight JB , Scolletta S , Santonocito C , Pastore F , Lorini FL et al. Levosimendan for patients with severely reduced left ventricular systolic function and/or low cardiac output syndrome undergoing cardiac surgery: a systematic review and meta-analysis . Crit Care 2017 ; 21 : 252. Google Scholar CrossRef Search ADS 57 Chen QH , Zheng RQ , Lin H , Shao J , Yu JQ , Wang HL. Effect of levosimendan on prognosis in adult patients undergoing cardiac surgery: a meta-analysis of randomized controlled trials . Crit Care 2017 ; 21 : 253. Google Scholar CrossRef Search ADS 58 Pollesello P , Parissis J , Kivikko M , Harjola VP. Levosimendan meta-analyses: is there a pattern in the effect on mortality? Int J Cardiol 2016 ; 209 : 77 – 83 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Interactive CardioVascular and Thoracic Surgery Oxford University Press

Levosimendan versus placebo in cardiac surgery: a systematic review and meta-analysis

Loading next page...
 
/lp/ou_press/levosimendan-versus-placebo-in-cardiac-surgery-a-systematic-review-and-0uTTXzgEas
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved.
ISSN
1569-9293
eISSN
1569-9285
D.O.I.
10.1093/icvts/ivy133
Publisher site
See Article on Publisher Site

Abstract

Abstract The aim of this meta-analysis was to review all published randomized clinical trials comparing levosimendan versus placebo in patients undergoing cardiac surgery. PubMed, EMBASE and the Cochrane library database of clinical trials were searched for prospective randomized clinical trials investigating the perioperative use of levosimendan versus placebo in patients undergoing adult cardiac surgery from 1 May 2000 to 10 April 2017. Binary outcomes from individual studies were analysed to compute individual and pooled risk ratios (RRs) with pertinent 95% confidence intervals (CIs). Fourteen randomized clinical trials with a total of 2243 patients were included in this review. Overall meta-analysis results demonstrated that levosimendan was associated with a significant reduction in 30-day mortality (RR = 0.71, 95% CI = 0.53–0.95; P = 0.023). Subgroup analysis showed that this benefit was confined to the moderate and low ejection fraction studies (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001), whereas no benefit was observed in the preserved ejection fraction studies (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78). Levosimendan also reduced the risk of renal replacement therapy (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015) and low cardiac output (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003). No significant differences were detected, between the levosimendan group and the placebo group, in terms of risk of myocardial injury (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44), intensive care unit stay (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055) and the use of ventricular assist device (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35). In conclusion, levosimendan was associated with a reduced risk of mortality, renal replacement therapy and low cardiac output syndrome in patients undergoing cardiac surgery. Levosimendan , placebo , Cardiac surgery , Meta-analysis INTRODUCTION Low cardiac output (LCO) syndrome after cardiac surgery is associated with increased mortality, delayed recovery, organ failure and prolonged intensive care unit stay. This syndrome is characterized by a decreased heart pump function, leading to reduced oxygen delivery and subsequent tissue hypoxia [1], in conjunction with signs of tissue hypoperfusion (cold periphery, clammy skin, confusion, oliguria and elevated lactate level) and in the absence of hypovolaemia [2]. LCO syndrome appears in approximately 20% of cardiopulmonary bypass surgeries [3]. Levosimendan, a calcium-sensitizing inotrope and an ATP-sensitive potassium channel opener, has been reported to be effective in decreasing LCO syndrome and mortality after cardiac surgery [4, 5]. The recent publication of 2 large randomized clinical trials (RCTs) [6, 7], which were unable to document any benefit of levosimendan compared with placebo in terms of survival, has strongly reopened the debate about its efficacy. The aim of this study was to evaluate the effect of levosimendan in comparison with placebo in adult patients undergoing cardiac surgery using a meta-analysis of RCTs. MATERIALS AND METHODS Search strategy PubMed, EMBASE and the Cochrane library database of clinical trials were searched for randomized clinical trials investigating the perioperative use of levosimendan versus placebo in patients undergoing cardiac surgery. The search terms used were ‘Levosimendan AND Cardiac Surgery’. The search was combined with filters to identify clinical trials, meta-analysis and reviews in the PubMed and EMBASE database. The filters for EMBASE searches were Cochrane review, systematic review, meta-analysis, randomized controlled trial, adult, humans, abstracts and year of publication 2000–2017. The filters for PubMed were clinical trial, meta-analysis, randomized controlled trial, review, systematic reviews and publication dates from 1 May 2000 to 10 April 2017. Searches were restricted to articles published in English. Search strategy for PubMed is described in Supplementary Material, Data S1, Table 1. Table 1: Description of the studies included in the meta-analysis First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days AVR: aortic valve replacement; CABG: coronary artery bypass graft; CBP: cardiopulmonary bypass; ECMO: extracorporeal membrane oxygenation; IABP: intra-aortic balloon pump; ICU: intensive care unit; inf: infusion; Levo: levosimendan; LV: left ventricular; LVEF: left ventricular ejection fraction; MV: mitral valve; MVR: mitral valve replacement; NR: not registered. Table 1: Description of the studies included in the meta-analysis First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days First author Year n Setting Exclusion criteria Mean LVEF % in the Levo group Mean LVEF % in the Placebo group Timing of Levo Dosing of Levo Follow-up Lahtinen 2011 200 Valve surgery ± CABG with CPB Endocarditis and aortic surgery NR NR Intraoperative  after anaesthetic induction Bolus: 24 µg/kg Inf: 0.2 µg/kg/min Duration: 24 h 6 months Anastasiadis 2016 32 Isolated CABG with minimally invasive CPB and LVEF <40% or with ischaemic mitral regurgitation Preoperative inotropic support or intra-aortic balloon pump, redo patients, concomitant valve surgery other than mitral valve repair, acute myocardial infarction with ST elevation in <14 days old and renal failure 35.7 ± 4.9 37.5 ± 3.4 Preoperative  24 h before surgery Inf: 0.1 µg/kg/min Duration: 24 h 30 days Landoni 2017 506 Cardiac surgery patients with perioperative cardiovascular dysfunction Adverse response to Levo, ECMO requirement and liver cirrhosis 50 (37–59) 50 (40–60) Intraoperative (to weaning from CPB) or postoperative (in the ICU within the first 24 h) Inf: 0.066 ± 0.031 µg/kg/min Duration: 33 ± 14.6 h 180 days Mehta 2017 849 CABG or CABG + AVR or MVR or combined procedures. All patients with LVEF ≤35% Intra-aortic balloon pump 26 (24–32) 27 (22–31) Preoperative  0.33 h before surgery Inf: 0.2 µg/kg/min during 1 h then 0.1 µg/kg/min 23 h Duration: >23.5–24.5 h 90 days Levin 2012 252 CABG with LVEF <25% IABP, inotropes, valve surgery or combined 17.56 ± 3.24 18.62 ± 2.12 Preoperative  24 h before surgery Bolus: 10 µg/kg 60 min Inf: 0.1 µg/kg/min 23 h Duration: 24 h 30 days Erikkson 2009 60 CABG with 3-vessel coronary disease and LVEF ≤50% Previous CABG and valve operation 36 ± 8 36 ± 8 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min 23 h and 50 min Duration: 24 h 31 days Leppikangas 2011 24 AVR with CABG and LVEF <50% or LV hypertrophy Allergy to Levo 69 ± 9 63 ± 9 Preoperative  24 h before surgery Bolus: 12 µg/kg 10 min Inf: 0.2 µg/kg/min Duration: 24 h 4 days Jarvela 2008 24 Isolated AVR and hypertrophy LV or with CABG Active endocarditis 50 ± 4 65 ± 5 Intraoperative  after induction anaesthesia Inf: 0.2 µg/kg/min Duration: 24 h 30 days Triptapepe 2006 24 Isolated multivessel CABG Unstable angina, valvular disease, renal failure, previous CABG and recent myocardial infarction 50 ± 7 52 ± 5 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min Intrahospital Triptapepe 2009 102 Isolated multivessel CABG Previous CABG and recent myocardial infarction 41.6 ± 10.7 44.1 ± 9.8 Intraoperative  before CPB Bolus: 24 µg/kg during 10 min 30 days Erb 2014 33 Isolated or combined CABG procedure with CPB and LVEF ≤30% 22 ± 4.5 22.4 ± 5.5 Intraoperative  after induction anaesthesia Inf: 12.5 mg 0.1 µg/kg/min Duration: NR 180 days Ersoy 2013 20 Valve surgery with severe pulmonary hypertension and LVEF <50% Liver insufficiency in child B or C 46.8 ± 10.9 49 ± 12.0 Intraoperative  after induction anaesthesia Bolus: 12 µg/kg 10 min Inf: 0.1 µg/kg/min Duration: 24 h Intrahospital Sharma 2013 40 MV repair and CABG and LVEF <30% Previous cardiac surgery, any other valve pathology, recent myocardial infarction and haemodynamic instability 23.55 ± 4.87 22.55 ± 0.92 Preoperative  24 h before surgery Inf: 200 μg/kg Duration: 24 h 30 days Levin 2008 77 AVR with ventricular dysfunction LVEF >25% NR NR NR Bolus: 10 μg/kg in 60 min inf: 0.1 μg/kg/min in 23 h Duration: 24 h 30 days AVR: aortic valve replacement; CABG: coronary artery bypass graft; CBP: cardiopulmonary bypass; ECMO: extracorporeal membrane oxygenation; IABP: intra-aortic balloon pump; ICU: intensive care unit; inf: infusion; Levo: levosimendan; LV: left ventricular; LVEF: left ventricular ejection fraction; MV: mitral valve; MVR: mitral valve replacement; NR: not registered. Study selection Two investigators reviewed all available abstracts for potential inclusion. Studies that were prospective, randomized controlled trials (RCT) of perioperative administration of levosimendan in adults undergoing cardiac surgery were subjected for a full text review. Abstracts by identical authors were cross-checked to ensure that data were not duplicated. Divergences were resolved by consensus. Previously published meta-analyses and review articles were scanned to evaluate articles for inclusion. Population, intervention, comparison, outcomes and study design (PICOS) were the following: Participants: patients undergoing cardiac surgery Intervention: levosimendan Comparison: placebo Outcomes: mortality, renal replacement therapy (as per author definition), in-hospital myocardial injury (as per author definition), intensive care unit stay, low cardiac output syndrome (as per author definition) and the use of ventricular assist device Study design: RCTs The following inclusion criteria were used: randomized clinical trials comparing levosimendan treatment with placebo in adult patients undergoing cardiac surgery with cardiopulmonary bypass. The exclusion criteria were paediatric or congenital population, cardiac surgery carried out exclusively without cardiopulmonary bypass, cardiac surgery other than valve surgery or coronary surgery, animal studies and studies that did not include our primary end point or secondary end points. Reviews and meta-analysis were also excluded. If the abstract suggested that the study could potentially meet the inclusion criteria, then the full text article was reviewed. Data abstraction and study characteristics The following study characteristics were recorded: last name of the first author, year of publication, design details, number of participants, timing and dosing of levosimendan administration and duration of follow-up. Baseline population characteristics including type of cardiac surgery and preoperative left ventricular ejection fraction also were recorded (Table 1). Internal validity and risk of bias assessment The internal validity and risk of bias of included trials were appraised according to the Cochrane collaboration methods [8] by 2 independent reviewers with divergences resolved by consensus. Data analysis Computations were carried out using Statistical Package R, specifically the Meta package and MedCalc Statistical Software version 17.9.6 (MedCalc Software bvba, Ostend, Belgium; http://www.medcalc.org; 2017). Hypothesis of statistical heterogeneity was tested using Cochran Q test ( Xn2, where n is degrees of freedom), and inconsistency was measured using I2. Binary outcomes from individual studies were analysed to compute individual and pooled risk ratios (RRs) with pertinent 95% confidence intervals (CIs) [with equivalence set at 1, RR <1 favouring the levosimendan treatment and RR >1 favouring the control group using the Mantel–Haenszel fixed-effect method in case of low statistical inconsistency (I2 ≤25%) and using a random-effect method in case of high or moderate statistical inconsistency (I2 >25%)]. Sensitivity analysis was performed by removing each study, one by one, and repeating the meta-analysis for the primary end point to check for the influence of individual results on the overall results. Meta-analysis of the primary end point for 30-day mortality was divided in 2 sub-analyses: analysis carried out on studies including patients with moderate or low left ventricle ejection fraction (LVEF) and analysis carried out on studies including patients with preserved ejection fraction. Weighted mean differences and 95% CIs were computed for continuous variables using the same methods as just described. The risk of publication bias was assessed by visual inspection of funnel plots and linear regression analysis. Forest plots were built to provide a graphical representation of the analysis. The RR of each study is represented as a square (with lines extending to 95% CI) and the overall pooled RR as a diamond. The weight is the percentage that each individual study contributes to eventual outcome. The size of the square of each study in forest plot is proportional to the weight of that study in the analysis. Statistical significance was set at the 2-tailed 0.05 level for hypothesis testing and at 0.10 for heterogeneity testing. Unadjusted P-values are reported. This study was carried out in compliance with the Cochrane Collaboration, the Quality of Reporting of Meta-Analyses guidelines (QUOROM) [9] and Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) [10]. RESULTS The search strategy yielded 484 abstracts. Of these, 447 were excluded, and 37 were analysed as full text articles (Fig. 1). Among these, 14 were included [6, 7, 11–22] in our meta-analysis (Table 1), and 23 were excluded. Studies were excluded for the following reasons: 6 studies did not report interest end points [23–28], 5 studies compared the levosimendan group with the control group that was not placebo [29–33], 2 articles were study protocols [34, 35], 1 article was not in cardiac surgery setting [36], 1 article was about patients with end-stage renal disease [37], 1 article was a letter to the editor [38], 4 articles were reviews [39–42], 1 article was about off-pump cardiac surgery [43], 1 article was not a randomized trial [44] and 1 article [45] was a sub-study of the Eriksson trial [22]. Figure 1: View largeDownload slide Flow diagram. Figure 1: View largeDownload slide Flow diagram. Study characteristics The 14 included RCTs randomized 2243 patients (1122 to levosimendan and 1121 to control). Levosimendan was administered as bolus followed by continuous infusions (6 RCTs), as continuous infusion only (6 RCTs) or as bolus only (2 RCTs) (Table 1). The dose varied from 10 to 24 µg/kg for intravenous bolus and from 0.1 to 0.2 µg/kg/min for continuous infusion (Supplementary Material, Data S1, Table 2). Five RCTs were multicentric [6, 7, 11, 12, 22] and 9 were unicentric [13–21]. Four RCTs enrolled exclusively patients with a preoperative LVEF <35% [6, 12, 16, 20]. Two studies did not report a median or mean ejection fraction: 1 study reported that 73% of patients in the placebo group and 77% of patients in the levosimendan group had an LVEF >50% [18] and another study reported that all study patients had left ventricle dysfunction with an LVEF >25% [11]. Five studies reported an LVEF of 36–50% in the preoperative sample population [7, 14, 17, 21, 22]. Three RCTs documented an LVEF of >50% in a sample population [13, 14, 19]. Study quality appraisal indicated that studies were of optimal quality, and 10 of them had a low risk of bias (Supplementary Material, Data S1, Table 3). In both the levosimendan group and the placebo group, additional standard inotropic treatment was administered according to hospital protocols. Quantitative data synthesis Overall analysis showed that the relative risk for postoperative mortality at 30 days was reduced by 29% in the levosimendan group when compared with the control group [67 of 1122 (6.0%) in the levosimendan group vs 96 of 1121 (8.6%) in the placebo group] (RR = 0.71, 95% CI = 0.53–0.95; P = 0.023; I2 = 22%; p for heterogeneity = 0.23; with 2243 patients and 14 studies included) (forest plot, Fig. 2). The funnel plot (Supplementary Material, Data S2, Fig. 1) and linear regression analysis showed no significative asymmetry (P = 0.15), and thus, no publication bias was detected by these methods. Subgroup analysis showed that survival benefit was confined to the low and moderate LVEF studies [23 of 739 (3.1%) in the levosimendan group vs 53 of 726 (7.3%) in the placebo group] (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001; I2 = 0%, p for heterogeneity = 0.47, with 1465 patients and 9 studies included). Survival benefit was not observed in the preserved LVEF studies [44 of 383 (11.5%) in the levosimendan group vs 43 of 395 (10.9%) in the placebo group] (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78; I2 = 0%; p for heterogeneity = 0.82, with 778 patients and 5 studies included). The sensitivity analysis (Supplementary Material, Data S2, Fig. 2) did not change the final result. Figure 2: View largeDownload slide Forest plot of the mortality meta-analysis. CI: confidence interval; EF: ejection fraction; RR: risk ratio. Figure 2: View largeDownload slide Forest plot of the mortality meta-analysis. CI: confidence interval; EF: ejection fraction; RR: risk ratio. In the levosimendan group, a significant reduction in the rate of renal replacement therapy was observed [49 of 995 (4.9%) in the levosimendan group vs 76 of 997 (7.6%) in the placebo group] (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015; I2 = 0%; p for heterogeneity = 0.78, with 1992 patients and 9 studies included) (forest plot, Fig. 3). The funnel plot (Supplementary Material, Data S2, Fig. 3) and linear regression analysis did not show asymmetry (P = 0.13), thus no publication bias was detected by these methods. The sensitivity analysis showed the same benefit (Supplementary Material, Data S2, Fig. 4). Figure 3: View largeDownload slide Forest plot of the renal replacement therapy meta-analysis. CI: confidence interval; RR: risk ratio. Figure 3: View largeDownload slide Forest plot of the renal replacement therapy meta-analysis. CI: confidence interval; RR: risk ratio. Figure 4: View largeDownload slide Forest plot of the myocardial injury meta-analysis. CI: confidence interval; RR: risk ratio. Figure 4: View largeDownload slide Forest plot of the myocardial injury meta-analysis. CI: confidence interval; RR: risk ratio. No significant reduction in risk of postoperative myocardial injury was detected in the levosimendan group [90 of 796 (11.3%)] vs the placebo group [99 of 787 (12.6%)] (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44; I2 = 0%; p for heterogeneity = 0.55, with 1583 patients and 9 studies included) (forest plot, Fig. 4). The funnel plot (Supplementary Material, Data S2, Fig. 5) and linear regression analysis demonstrated asymmetry (P = 0.043) showing a possible publication bias. The sensitivity analysis did not change the final result (Supplementary Material, Data S2, Fig. 6). Figure 5: View largeDownload slide Forest plot of the low cardiac output syndrome meta-analysis. CI: confidence interval; RR: risk ratio. Figure 5: View largeDownload slide Forest plot of the low cardiac output syndrome meta-analysis. CI: confidence interval; RR: risk ratio. Figure 6: View largeDownload slide Forest plot of ventricular assist device implantation meta-analysis. CI: confidence interval; RR: risk ratio. Figure 6: View largeDownload slide Forest plot of ventricular assist device implantation meta-analysis. CI: confidence interval; RR: risk ratio. Low cardiac output syndrome was reported in 6 studies. Levosimendan was shown to be effective in reducing low cardiac output syndrome when compared with placebo [107 of 723 (14.8%) in the levosimendan group vs 207 of 715 (29.0%) in the placebo group] (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003; I2 = 75%; p for heterogeneity = 0.003, with 1438 patients included and 6 RCTs) (forest plot, Fig. 5). The funnel plot (Supplementary Material, Data S2, Fig. 7) and linear regression analysis did not show asymmetry (P = 0.27), thus no publication bias was detected by these methods. The sensitivity analysis (Supplementary Material, Data S2, Fig. 8) did not change the final result. As a significative heterogeneity was detected in low cardiac output syndrome meta-analysis (I2 = 75%), we repeated the analysis omitting the largest trial [6]. The meta-analysis of the remaining 5 RCTs showed no heterogeneity (I2 = 0) and confirmed a significant reduction in risk of low cardiac output syndrome in the levosimendan group (Fig. 5). We carried out the analysis applying fixed effects model and random effects model obtaining the same results (RR = 0.29, 95% CI = 0.20–0.43; P < 0.001). The omitted trial [6] analysis also showed a significant reduction in terms of low cardiac output syndrome in the levosimendan group (RR = 0.71, 95% CI = 0.55–0.92). Figure 7: View largeDownload slide Forest plot of the intensive care unit stay meta-analysis. CI: confidence interval; MD: mean difference; SD: standard deviation. Figure 7: View largeDownload slide Forest plot of the intensive care unit stay meta-analysis. CI: confidence interval; MD: mean difference; SD: standard deviation. The levosimendan group did not show a significant reduction in risk of using ventricular assist device [47 of 572 (8.2%) in the levosimendan group vs 45 of 562 (8.0%) in the placebo group] (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35; I2 = 59%; p for heterogeneity = 0.088, with 1134 patients and 3 studies included) (forest plot, Fig. 6). The funnel plot (Supplementary Material, Data S2, Fig. 9) and linear regression analysis showed asymmetry (P = 0.026), documenting a possible publication bias. As a significative heterogeneity was detected in ventricular assist device meta-analysis (I2 = 59%), we repeated the analysis omitting the largest trial [6]. The meta-analysis of the remaining 2 RCTs showed no heterogeneity (I2 = 0) and detected a beneficial result for the levosimendan group (RR = 0.12, 95% CI = 0.02–0.94; P = 0.044) (Fig. 6). However, the analysis of the trial by Mehta et al. [6] did not show significant results (RR = 1.22, 95% CI = 0.81–1.83). (The sensitivity analysis is represented in Supplementary Material, Data S2, Fig. 10.) No significant reduction in intensive care unit (ICU) stay in the levosimendan group was observed (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055; I2 = 91%; p for heterogeneity <0.001, with 6 studies included) (forest plot, Fig. 7). The funnel plot (Supplementary Material, Data S2, Fig. 11) and linear regression analysis did not demonstrate asymmetry (P = 0.61), thus no publication bias was detected using these methods. Sensitivity analysis showed the same result (Supplementary Material, Data S2, Fig. 12). We also carried out a subgroup analysis to check whether the time of administration, dose and the presence of loading bolus influence the survival results. We carried out this analysis in patients with moderate or severe dysfunction because this group was the one benefitting from levosimendan administration in terms of survival. With regard to time of levosimendan administration, a statistically significant benefit was documented by the preoperative administration over the intraoperative one (RR = 0.46, 95% CI = 0.28–0.74; P = 0.002; I2 = 22%; p for heterogeneity = 0.28, with 1250 patients included and 5 RCTs), (Supplementary Material, Data S2, Fig. 13). With regard to the doses, 0.1 µg/kg/min was detected to be effective over 0.2 µg/kg/min (RR = 0.46, 95% CI = 0.28–0.74; P = 0.002; I2 = 22%; p for heterogeneity = 0.27, with 1263 patients included and 6 RCTs) (Supplementary Material, Data S2, Fig. 14). Also, we documented a beneficial effect of the initial loading bolus of levosimendan over the absence of loading dose (RR = 0.24, 95% CI = 0.10–0.56; P < 0.001; I2 = 0%; p for heterogeneity = 0.71, with 511 patients included and 5 RCTs) (Supplementary Material, Data S2, Fig. 15). DISCUSSION This meta-analysis shows that perioperative use of levosimendan is associated with a reduction in 30-day postoperative mortality in adult patients undergoing cardiac surgery. In particular, we documented this benefit in RCTs enrolling patients with low or moderate LVEF (RR = 0.44, 95% CI = 0.27–0.70; P < 0.001), but not in RCTs including patients with preserved LVEF (RR = 1.06, 95% CI = 0.72–1.56; P = 0.78). Furthermore, our analysis found levosimendan to be associated with a reduction in renal replacement therapy (RR = 0.66, 95% CI = 0.47–0.92; P = 0.015) and low cardiac output syndrome (RR = 0.40, 95% CI = 0.22–0.73; P = 0.003) compared with placebo. This meta-analysis failed to show any benefit of levosimendan in terms of decrease in myocardial injury (RR = 0.90, 95% CI = 0.69–1.17; P = 0.44), ICU stay (weighted mean differences = −0.57, 95% CI = −1.15 to 0.01; P = 0.055) and the use of ventricular assist device (RR = 0.42, 95% CI = 0.07–2.63; P = 0.35). These results are consistent with those of the previous 12 meta-analysis carried out on levosimendan used in surgical setting [46–57]. The first 10 of them, which did not include the 2 largest RCTs [6, 7], have been systematically reviewed by Pollesello et al. [58] in 2016. Briefly, levosimendan was reported to decrease postoperative mortality after cardiac surgery in the meta-analysis by Landoni et al. [47] [440 patients from 10 studies; odds ratio (OR)  = 0.35, 95% CI = 0.18–0.71; P = 0.003], Maharaj and Metaxa [48] (729 patients from 17 studies; OR = 0.40, 95% CI = 0.21–0.76; P = 0.005), Hernández et al. [49] (654 patients from 13 studies; OR = 0.36, 95% CI = 0.20–0.64; P = 0.001) and Zhou et al. [55] (1345 patients from 13 studies; OR = 0.41, 95% CI = 0.27–0.62; P < 0.001). Two meta-analysis showed, as we did, a beneficial effect in terms of postoperative survival only in the low ejection fraction trials: Harrison et al. [50] (1155 patients from 14 studies; risk difference −7.0%, 95% CI = −11.0% to −3.1%; P < 0.001) and Lim et al. [54] (965 patients from 14 studies; OR = 0.41, 95% CI = 0.24–0.77; P = 0.004). Levosimendan was associated with a lower incidence of acute kidney injury in meta-analysis by Niu et al. [51] (529 patients from 5 studies; OR = 0.44, 95% CI = 0.22–0.85; P = 0.02); Bove et al. [52] (3879 patients from 33 studies; RR = 0.79, 95% CI = 0.63–0.99; P = 0.048), Lim et al. [54] (965 patients from 14 studies; OR = 0.62, 95% CI = 0.40–0.95; P = 0.05) and Zhou et al. [55] (1345 patients from 13 studies OR = 0.51, 95% CI = 0.34–0.76; P = 0.001). Chen et al. [57] and Sanfilippo et al. [56] proposed a meta-analysis including the 2 largest trials [6, 7]. Both of them concluded that levosimendan decreases mortality in patients with preoperative ventricular systolic dysfunction. Our meta-analysis is based on 14 studies: 2 of them [6, 7] are large trials (sample population ≥506 patients), 2 midsize studies [12, 28] (sample population ≥252 patients) and 10 small studies [11, 13–17, 19–22] (sample population from 24 to 102 patients). Basically, large and midsize RCTs investigated clinical primary outcome (mortality, LCO syndrome or renal failure incidence), whereas small trials have analysed mostly haemodynamic outcomes (cardiac index and ejection fraction) or inotropic drug requirement. All small RCTs reported improved haemodynamic outcomes in patients treated with levosimendan. Mortality data analysed by small RCTs reached no statistical significance, likely because of the very limited number of events registered. Anyway, even not significant, small RCTs reported less mortality events in the levosimendan group in 6 studies over 10 (Fig. 2). Of the 2 midsized trials, 1 [18] was carried out on patients with preserved left ventricular function and reported no significant difference in mortality. The second one [12] was carried out on patients with left ventricle dysfunction and reported a significant decrease in mortality in those patients treated with levosimendan (P < 0.05). The 2 largest trials [6, 7] failed to show any difference in terms of 30-day mortality between the levosimendan and the placebo groups. In fact, the publication of these 2 large and solid trials has raised doubts about the utility of levosimendan in patients undergoing cardiac surgery. Even if apparently similar, these 2 trials present methodological and outcome analysis differences, which deserve further discussion. Landoni et al. [7] enrolled 506 patients. Inclusion criteria were preoperative LVEF <25%, high dose of inotropic support because of acute postoperative cardiac dysfunction (this being an indication of very late use of levosimendan) or preoperative intra-aortic balloon pump. Thus, most of the patients were enrolled in this study because of postoperative cardiogenic shock (65%), and only 4.3% of the population was randomized preoperatively because of low LVEF. In particular, among patients of the levosimendan group (n = 248), only 25.8% (n = 64) had left ventricle dysfunction (LVEF 25–40%). Because of the presence of a significant proportion of patients with postoperative cardiogenic shock, levosimendan was administered as infusion, without loading dose, to avoid arterial hypotension. Under such heterogenous circumstances, no differences were detected in 30- or 180-day mortality nor in the other explored outcomes. Mehta et al. [6] enrolled 849 patients; inclusion criterion was just 1: LVEF <35% documented 60 days before surgery. Just before surgery, loading and infusion doses were administered. Patients treated with levosimendan showed a significantly reduced incidence of LCO syndrome when compared with those treated with placebo (P = 0.007). Levosimendan was not associated with a reduced 30-day mortality (P = 0.45), but it showed a non-significative association (P = 0.12) with a reduced 90-day mortality. Our meta-analysis shows that the inclusion of these 2 trials does not neutralize the overall beneficial effect of levosimendan documented in terms of decreased mortality, renal replacement therapy and LCO syndrome RR. For this reason, the results of our meta-analysis suggest that perioperative levosimendan still represents an effective treatment in patients undergoing cardiac surgery with low LVEF. Limitations This meta-analysis carries the potential bias of this specific analysis format and those of the primary studies. In particular, in the ventricular assist device analysis, only 3 studies were included; high heterogeneity and possible publication bias, showed by the funnel plot and linear regression, were documented. The myocardial damage analysis was jeopardized by the different definitions of this clinical entity provided in the included studies. An ICU stay analysis showed a significant heterogeneity. Moreover, half of the studies in this analysis were small clinical trials, with sample size less than 50 patients. In addition, 3 of the studies had moderate risk of bias and 1 had high risk of bias. Subgroup meta-analysis on the different strategies of levosimendan administration is biased by the fact that those were applied to populations with different risk profiles. SUPPLEMENTARY MATERIAL Supplementary material is available at ICVTS online. Conflict of interest: none declared. REFERENCES 1 Vincent J-L , De Backer D. Circulatory shock . N Engl J Med 2013 ; 369 : 1726 – 34 . Google Scholar CrossRef Search ADS 2 Algarni KD , Maganti M , Yau TM. Predictors of low cardiac output syndrome after isolated coronary artery bypass surgery: trends over 20 years . Ann Thorac Surg 2011 ; 92 : 1678 – 84 . Google Scholar CrossRef Search ADS 3 Benjamin EJ , Blaha MJ , Chiuve SE , Cushman M , Das SR , Deo R. American Heart Association Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics-2017 update: a report from the American Heart Association . Circulation 2017 ; 135 : e146 – 603 . Google Scholar CrossRef Search ADS 4 Sunny YM , Karim HM , Saikia MK , Bhattacharyya P , Dey S. Comparison of Levosimendan, milrinone and dobutamine in treating low cardiac output syndrome following valve replacement surgeries with cardiopulmonary bypass . J Clin Diagn Res 2016 ; 10 : UC05 – 8 . 5 Toller W , Algotsson L , Guarracino F , Hormann C , Knotzer J , Lehmann A et al. Perioperative use of levosimendan: best practice in operative settings . J Cardiothorac Vasc Anesth 2013 ; 27 : 361 – 6 . Google Scholar CrossRef Search ADS 6 Mehta RH , Leimberger JD , van Diepen S , Meza J , Wang A , Jankowich R et al. Levosimendan in patients with left ventricular dysfunction undergoing cardiac surgery . N Engl J Med 2017 ; 376 : 2032 – 42 . Google Scholar CrossRef Search ADS 7 Landoni G , Lomivorotov VV , Alvaro G , Lobreglio R , Pisano A , Guarracino F et al. Levosimendan for hemodynamic support after cardiac surgery . N Engl J Med 2017 ; 376 : 2021 – 31 . Google Scholar CrossRef Search ADS 8 Higgins JPT , Green S. Cochrane Handbook for Systematic Reviews of Interventions. The Cochrane Collaboration. Version 5.1.0. http://handbook-5-1.cochrane.org (updated March 2011). 9 Moher D , Cook DJ , Eastwood S , Olkin I , Rennie D , Stroup DF. Improving the quality of reports of meta-analyses of randomised controlled trials: the QUOROM statement. Quality of reporting of meta-analyses . Lancet 1999 ; 354 : 1896 – 900 . Google Scholar CrossRef Search ADS 10 Knobloch K , Yoon U , Vogt PM. Preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement and publication bias . J Craniomaxillofac Surg 2011 ; 39 : 91 – 2 . Google Scholar CrossRef Search ADS 11 Levin R , Degrange M , Porcile R , Salvagio F , Del Mazo C , Tanus E et al. Preoperative use of calcium sensitizer levosimendan reduces mortality and low cardiac output syndrome in patients with aortic stenosis and left ventricular dysfunction. 2008 World Congress of Cardiology Abstracts Circulation 2008 ; 118 : P322 . 12 Levin R , Degrange M , Del Mazo C , Tanus E , Porcile R. Preoperative levosimendan decreases mortality and the development of low cardiac output in high-risk patients with severe left ventricular dysfunction undergoing coronary artery bypass grafting with cardiopulmonary bypass . Exp Clin Cardiol 2012 ; 17 : 125 – 30 . 13 Leppikangas H , Järvelä K , Sisto T , Maaranen P , Virtanen M , Lehto P et al. Preoperative levosimendan infusion in combined aortic valve and coronary bypass surgery . Br J Anaesth 2011 ; 106 : 298 – 304 . Google Scholar CrossRef Search ADS 14 Tritapepe L , De Santis V , Vitale D , Santulli M , Morelli A , Nofroni I et al. Preconditioning effects of Levosimendan in coronary artery bypass grafting—a pilot study . Br J Anaesth 2006 ; 96 : 694 – 700 . Google Scholar CrossRef Search ADS 15 Tritapepe L , De Santis V , Vitale D , Guarracino F , Pellegrini F , Pietropaoli P et al. Levosimendan pre-treatment improves outcomes in patients undergoing coronary artery bypass graft surgery . Br J Anaesth 2009 ; 102 : 198 – 204 . Google Scholar CrossRef Search ADS 16 Erb J , Beutlhauser T , Feldheiser A , Schuster B , Treskatsch S , Grubitzsch H et al. Influence of levosimendan on organ dysfunction in patients with severely reduced left ventricular function undergoing cardiac surgery . J Int Med Res 2014 ; 42 : 750 – 64 . Google Scholar CrossRef Search ADS 17 Anastasiadis K , Antonitsis P , Vranis K , Kleontas A , Asteriou C , Grosomanidis V et al. Effectiveness of prophylactic Levosimendan in patients with impaired left ventricular function undergoing coronary artery bypass grafting: a randomized pilot study . Interact CardioVasc Thorac Surg 2016 ; 23 : 740 – 7 . Google Scholar CrossRef Search ADS 18 Lahtinen P , Pitkänen O , Pölönen P , Turpeinen A , Kiviniemi V , Uusaro A. Levosimendan reduces heart failure after cardiac surgery: a prospective, randomized, placebo-controlled trial . Crit Care Med 2011 ; 39 : 2263 – 70 . Google Scholar CrossRef Search ADS 19 Järvelä K , Maaranen P , Sisto T , Ruokonen E. Levosimendan in aortic valve surgery: cardiac performance and recovery . J Cardiothorac Vasc Anesth 2008 ; 22 : 693 – 8 . Google Scholar CrossRef Search ADS 20 Sharma P , Malhotra A , Gandhi S , Garg P , Bishnoi A , Gandhi H. Preoperative levosimendan in ischemic mitral valve repair . Asian Cardiovasc Thorac Ann 2014 ; 22 : 539 – 45 . Google Scholar CrossRef Search ADS 21 Ersoy O , Boysan E , Unal EU , Yay K , Yener U , Cicekcioglu F et al. Effectiveness of prophylactic levosimendan in high-risk valve surgery patients . Cardiovasc J Afr 2013 ; 24 : 260 – 4 . Google Scholar CrossRef Search ADS 22 Eriksson HI , Jalonen JR , Heikkinen LO , Kivikko M , Laine M , Leino KA et al. Levosimendan facilitates weaning from cardiopulmonary bypass in patients undergoing coronary artery bypass grafting with impaired left ventricular function . Ann Thorac Surg 2009 ; 87 : 448 – 54 . Google Scholar CrossRef Search ADS 23 Asaad OM , Hanafy MS. Levosimendan’s effect on coronary artery grafts blood flow in patients with left ventricular dysfunction, assessment by transit time flow meter . Egypt J Anaesth 2011 ; 27 : 45 – 53 . Google Scholar CrossRef Search ADS 24 Juhl-Olsen P , Jakobsen C-J , Rasmussen LA , Bhavsar R , Klaaborg K-E , Frederiksen CA et al. Effects of levosimendan in patients with left ventricular hypertrophy undergoing aortic valve replacement . Acta Anaesthesiol Scand 2015 ; 59 : 65 – 77 . Google Scholar CrossRef Search ADS 25 Jörgensen K , Bech-Hanssen O , Houltz E , Ricksten SE. Effects of levosimendan on left ventricular relaxation and early filling at maintained preload and afterload conditions after aortic valve replacement for aortic stenosis . Circulation 2008 ; 117 : 1075 – 81 . Google Scholar CrossRef Search ADS 26 Bragadottir G , Redfors B , Ricksten SE. Effects of levosimendan on glomerular filtration rate, renal blood flow, and renal oxygenation after cardiac surgery with cardiopulmonary bypass: a randomized placebo-controlled study . Crit Care Med 2013 ; 41 : 2328 – 35 . Google Scholar CrossRef Search ADS 27 Aksel’rod BA , Tolstova IA , Trekova NA , Kolpakov PE , Babaev MA , Belianko IE. Impact of preoperative levosimendan therapy on the volemic status and vascular tone of patients with chronic heart failure during anesthesia . Anesteziol Reanimatol 2009 ; 6 : 46 – 51 . 28 Abacilar AF , Dogan OF. Levosimendan use decreases atrial fibrillation in patients after coronary artery bypass grafting: a pilot study . Heart Surg Forum 2013 ; 16 : 287 – 94 . Google Scholar CrossRef Search ADS 29 Carev M , Karanovic N , Kocen D , Bulat C. Useful supplement to the best practice of using levosimendan in cardiac surgery patients: 2.5-mg intravenous bolus for cardiopulmonary resuscitation during perioperative cardiac arrest . J Cardiothorac Vasc Anesth 2013 ; 27 : e75 – 7 . Google Scholar CrossRef Search ADS 30 De Hert SG , Lorsomradee S , Vanden Eede H , Cromheecke S , Van der Linden PJ. A randomized trial evaluating different modalities of levosimendan administration in cardiac surgery patients with myocardial dysfunction . J Cardiothorac Vasc Anesth 2008 ; 22 : 699 – 705 . Google Scholar CrossRef Search ADS 31 Baysal A , Yanartas M , Dogukan M , Gundogus N , Kocak T , Koksal C. Levosimendan improves renal outcome in cardiac surgery: a randomized trial . J Cardiothorac Vasc Anesth 2014 ; 28 : 586 – 94 . Google Scholar CrossRef Search ADS 32 Tasouli A , Papadopoulos K , Antoniou T , Kriaras I , Stavridis G , Degiannis D et al. Efficacy and safety of perioperative infusion of levosimendan in patients with compromised cardiac function undergoing open-heart surgery: importance of early use . Eur J Cardiothorac Surg 2007 ; 32 : 629 – 33 . Google Scholar CrossRef Search ADS 33 Alvarez J , Bouzada M , Fernández AL , Caruezo V , Taboada M , Rodríguez J et al. Hemodynamic effects of levosimendan compared with dobutamine in patients with low cardiac output after cardiac surgery . Rev Esp Cardiol 2006 ; 59 : 338 – 45 . Google Scholar CrossRef Search ADS 34 Mehta RH , Van Diepen S , Meza J , Bokesch P , Leimberger JD , Tourt-Uhlig S et al. Levosimendan in patients with left ventricular systolic dysfunction undergoing cardiac surgery on cardiopulmonary bypass: rationale and study design of the levosimendan in patients with left ventricular systolic dysfunction undergoing cardiac surgery requiring cardiopulmonary bypass (LEVO-CTS) trial . Am Heart J 2016 ; 182 : 62 – 71 . Google Scholar CrossRef Search ADS 35 Caruba T , Hourton D , Sabatier B , Rousseau D , Tibi A , Hoffart-Jourdain C et al. Rationale and design of the multicenter randomized trial investigating the effects of levosimendan pretreatment in patients with low ejection fraction (≤40%) undergoing CABG with cardiopulmonary bypass (LICORN study) . J Cardiothorac Surg 2016 ; 11 : 127 . Google Scholar CrossRef Search ADS 36 Adamopoulos S , Parissis JT , Iliodromitis EK , Paraskevaidis I , Tsiapras D , Farmakis D et al. Effects of levosimendan versus dobutamine on inflammatory and apoptotic pathways in acutely decompensated chronic heart failure . Am J Cardiol 2006 ; 98 : 102 – 6 . Google Scholar CrossRef Search ADS 37 Atalay H , Temizturk Z , Azboy D , Colak S , Atalay A , Dogan OF et al. Levosimendan use increases cardiac performance after coronary artery bypass grafting in end-stage renal disease patients . Heart Surg Forum 2016 ; 19 : 230 – 6 . Google Scholar CrossRef Search ADS 38 De Santis V , Vitale D , Tritapepe L. Levosimendan and cardiac surgery . J Cardiothorac Vasc Anesth 2010 ; 24 : 210. Google Scholar CrossRef Search ADS 39 Elahi MM , Lam J , Asopa S , Matata BM. Levosimendan versus an intra-aortic balloon pump in adult cardiac surgery patients with low cardiac output . J Cardiothorac Vasc Anesth 2011 ; 25 : 1154 – 62 . Google Scholar CrossRef Search ADS 40 van den Brule J , Hoedemaekers C , Pickkers P. Clinical outcome benefits of pretreatment with levosimendan . Br J Anaesth 2009 ; 102 : 883 – 4 . Google Scholar CrossRef Search ADS 41 Toller W , Knez I. Medical support and surgery of the failing heart: levosimendan . Scand J Surg 2007 ; 96 : 121 – 4 . Google Scholar CrossRef Search ADS 42 García-González MJ , Domínguez-Rodríguez A. Effect of levosimendan treatment of myocardial stunning and low-output syndrome after cardiac surgery . Rev Esp Cardiol 2006 ; 59 : 851 – 2 . Google Scholar CrossRef Search ADS 43 Husedzinović I , Barisin S , Bradić N , Barisin A , Sonicki Z , Milanović R. Levosimendan as a new strategy during off-pump coronary artery bypass grafting: double-blind randomized placebo-controlled trial . Croat Med J 2005 ; 46 : 950 – 6 . 44 Temizturk Z , Azboy D , Atalay A , Atalay H , Dogan OF. The effects of levosimendan and sodium nitroprusside combination on left ventricular functions after surgical ventricular reconstruction in coronary artery bypass grafting patients . Open Cardiovasc Med J 2016 ; 10 : 138 – 47 . Google Scholar CrossRef Search ADS 45 Ristikankare A , Pöyhiä R , Eriksson H , Valtonen M , Leino K , Salmenperä M. Effects of levosimendan on renal function in patients undergoing coronary artery surgery . J Cardiothorac Vasc Anesth 2012 ; 26 : 591 – 5 . Google Scholar CrossRef Search ADS 46 Zangrillo A , Biondi-Zoccai G , Mizzi A , Bruno G , Bignami E , Gerli C et al. Levosimendan reduces cardiac troponin release after cardiac surgery: a meta-analysis of randomized controlled studies . J Cardiothorac Vasc Anesth 2009 ; 23 : 474 – 8 . Google Scholar CrossRef Search ADS 47 Landoni G , Mizzi A , Biondi-Zoccai G , Bruno G , Bignami E , Corno L et al. Reducing mortality in cardiac surgery with levosimendan: a meta-analysis of randomized controlled trials . J Cardiothorac Vasc Anesth 2010 ; 24 : 51 – 7 . Google Scholar CrossRef Search ADS 48 Maharaj R , Metaxa V. Levosimendan and mortality after coronary revascularisation: a meta-analysis of randomised controlled trials . Crit Care 2011 ; 15 : R140. Google Scholar CrossRef Search ADS 49 Hernández A , Miranda A , Parada A. Levosimendan reduces mortality in cardiac surgery: a systematic review and meta-analysis . Rev Esp Anestesiol Reanim 2012 ; 59 : 6 – 11 . Google Scholar CrossRef Search ADS 50 Harrison RW , Hasselblad V , Mehta RH , Levin R , Harrington RA , Alexander JH. Effect of levosimendan on survival and adverse events after cardiac surgery: a meta-analysis . J Cardiothorac Vasc Anesth 2013 ; 27 : 1224 . Google Scholar CrossRef Search ADS 51 Niu ZZ , Wu SM , Sun WY , Hou WM , Chi YF. Perioperative levosimendan therapy is associated with a lower incidence of acute kidney injury after cardiac surgery: a meta-analysis . J Cardiovasc Pharmacol 2014 ; 63 : 107 – 12 . Google Scholar CrossRef Search ADS 52 Bove T , Matteazzi A , Belletti A , Paternoster G , Saleh O , Taddeo D et al. Beneficial impact of levosimendan in critically ill patients with or at risk for acute renal failure: a meta-analysis of randomized clinical trials . Heart Lung Vessel 2015 ; 7 : 35 – 46 . 53 Greco T , Calabrò MG , Covello RD , Greco M , Pasin L , Morelli A et al. A Bayesian network meta-analysis on the effect of inodilatory agents on mortality . Br J Anaesth 2015 ; 114 : 746 – 56 . Google Scholar CrossRef Search ADS 54 Lim JY , Deo SV , Rababa’h A , Altarabsheh SE , Cho YH , Hang D et al. Levosimendan reduces mortality in adults with left ventricular dysfunction undergoing cardiac surgery: a systematic review an meta-analysis . J Card Surg 2015 ; 30 : 547 – 54 . Google Scholar CrossRef Search ADS 55 Zhou C , Gong J , Chen D , Wang W , Liu M , Liu B. Levosimendan for prevention of acute kidney injury after cardiac surgery: a meta-analysis of randomize controlled trials . Am J Kidney Dis 2016 ; 67 : 408 – 16 . Google Scholar CrossRef Search ADS 56 Sanfilippo F , Knight JB , Scolletta S , Santonocito C , Pastore F , Lorini FL et al. Levosimendan for patients with severely reduced left ventricular systolic function and/or low cardiac output syndrome undergoing cardiac surgery: a systematic review and meta-analysis . Crit Care 2017 ; 21 : 252. Google Scholar CrossRef Search ADS 57 Chen QH , Zheng RQ , Lin H , Shao J , Yu JQ , Wang HL. Effect of levosimendan on prognosis in adult patients undergoing cardiac surgery: a meta-analysis of randomized controlled trials . Crit Care 2017 ; 21 : 253. Google Scholar CrossRef Search ADS 58 Pollesello P , Parissis J , Kivikko M , Harjola VP. Levosimendan meta-analyses: is there a pattern in the effect on mortality? Int J Cardiol 2016 ; 209 : 77 – 83 . Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press on behalf of the European Association for Cardio-Thoracic Surgery. All rights reserved. This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Interactive CardioVascular and Thoracic SurgeryOxford University Press

Published: Apr 30, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off