The most common and serious postoperative complications identified were low cardiac output syndrome and kidney failure. Perioperative cardioprotective and renoprotective strategies consist of maintaining organ perfusion and oxygen supply to tissues. Appart from diuretics, renal replacement therapy is another therapeutic option for acute kidney injury. When response to inotropic support is not adequate, intra-aortic balloon pump (IABP) counterpulsation and other ventricular assist devices can be used. However, none of our patients required any of these extracorporeal therapies [10]. The effect of levosimendan could be two-fold: first, its inotropic effects may improve cardiac output. Secondly, in a setting of low cardiac output, levosimendan has a direct effect on renal perfusion, as it is a KATP channel opener with a vasodilator effect on arteries. In addition, in a setting of renal hypoperfusion due to decreased cardiac output, levosimendan provides kidney protection by blocking mitochondrial KATP channels.
According to the literature, the incidence of kidney failure in LCOS is 20 to 30%, which is consistent with our results.
The drug used for treating LCOS was an independent factor of kidney failure. In our study, the incidence of kidney failure at 48 h after treatment completion was lower in the patients who were administered the calcium sensitizer for LCOS, as compared to those who received beta-agonists. Similar results have been obtained in previous comparative studies of levosimendan and beta-agonists or placebo. Levin et al. performed a study in 252 patients with severe left ventricular dysfunction undergoing coronary artery bypass grafting with cardiopulmonary bypass who developed LCOS. Preoperative levosimendan was compared with placebo, and the incidence of kidney failure was found to be lower with levosimendan [11]. Baysal et al. randomized 128 patients with FEVI < 45% undergoing mitral valve surgery to two treatment arms. The first group received a loading dose of 6 mcg/kg/min of levosimendan after removal of the aortic cross-clamp followed by an infusion of 0.1mcg/kg/min in combination or not with standard inotropics (dobutamine, adrenaline and noradrenaline). The control group only received standard inotropics. Postoperative serum creatinine levels were lower in the first group, who also showed a better glomerular filtration rate [12]. In contrast with previous works, our study focused on cardiac surgery patients who developed organ dysfunction caused by hypoperfusion induced by postoperative LCOS.
Differences were also observed in the incidence of kidney failure –not significant but with a clear tendency– based on the timing of levosimendan administration. Early administration and the agent selected were found to be determinant. Differences may have not reached significance because our sample size was not calculated for this endpoint.
Balzer et al. conducted a retrospective study in 46 patients who underwent coronary artery bypass grafting with creatinine < 2 mg/dl and a left ventricular systolic function < 35% and/or LOCS. The incidence of acute kidney failure was significantly lower in patients who received early levosimendan therapy vs. those who received late-start levosimendan therapy [13]. Treskatsch et al. performed a study in 159 cardiac surgery patients with FEVI < 35% and/or LCOS, who were divided into two groups based on the timing of levosimendan administration. Early administration was found to be related to a lower incidence of kidney failure and a reduced need for renal replacement therapy [14]. In a multicenter, randomized, double-blind, placebo-controlled LICORN study, Caruba et al. included 340 patients from 13 hospitals with FEVI < 40% undergoing CABG with cardiopulmonary bypass. The authors investigated the effects of levosimendan administered prior to anesthesia vs placebo and the requirement of renal replacement therapy while in the ICU. A lower incidence was demonstrated in the levosimendan group [15]. In our study, the incidence of kidney failure at 24 h showed a decreasing tendency following early administration of levosimendan vs beta-agonists at diagnosis of LCOS, with differences becoming significant at 48 h of diagnosis. Sample homogeneity in terms of cardiac function and oxygen supply was ensured by measuring central venous saturation and cardiac index, guaranteeing that the hemodynamic therapy administered for LCOS was similar in the two groups. All LCOS patients received inotropic therapy, and their cardiac index and venous saturation were within normal limits. The incidence of renal dysfunction was lower in the levosimendan group. This finding supports our hypothesis that the protective effect of levosimendan on renal function is independent from its cardioprotective effect. Bragadottir et al. documented that levosimendan has beneficial effects on renal blood flow, glomerular filtration rate and renal vascular resistance, as a result of its action on KATP channels in renal afferent arterioles [7]. Levosimendan improves renal function (defined as a reduction in serum creatinine) at 24 h after therapy completion and enhances diuresis at ICU discharge by reducing serum creatinine levels.
Although our results are consistent with those obtained by other authors such as Fedele et al., who documented the benefits of levosimendan on renal function [16], conflicting results have been obtained in other studies. Thus, Mehta et al. examined data from 849 patients with severely reduced ejection fraction who underwent cardiac surgery requiring cardiopulmonary bypass support (66% had coronary artery bypass surgery). With a third of patients having mild preoperative chronic kidney failure, no significant differences were found in the need for renal replacement therapy at 30 days between patients who received preoperative levosimendan vs placebo [17]. Similar results were reported by Landoni in a multicenter, randomized, double-blind study in 506 patients with perioperative ventricular dysfunction (preoperative FEVI < 25%; need for IABP or elevated postoperative score). In this study, most patients were recruited in the postoperative period and presented LCOS requiring hemodynamic support with high doses of inotropics that could be secondary to LCOS-induced cardiogenic shock. No differences have been reported either between levosimendan and placebo in the incidence of kidney failure as measured by the AKI scale, or in the need for renal therapy [18]. Inconsistencies could be due to the population of patients. In our group, patients prevailingly had mitral and mitral-tricuspid dysfunction, where the incidence of postoperative right ventricular dysfunction is higher. Also, the sample did not include any patient with cardiogenic shock. In a previous study, our group evaluated a case series of patients who underwent valve replacement/ repair surgery, who were administered levosimendan preoperatively. The hemodynamics of patients with preoperative right ventricular dysfunction improved as a result of improved postoperative right ventricular and renal function [19]. As described by Damman et al., an elevated renal venous pressure causes a reduction in perfusion pressure and renal blood flow. When ventricular function improves, renal venous pressure decreases, thereby exerting a renoprotective effect [20]. In two studies previously published by our group, preoperative levosimendan was found to protect renal function in patients at high risk for cardio-renal dysfunction, as assessed by the AKI scale and postoperative N-GAL levels. Thus, the results of our study may be explained by both, the potential renoprotective effects of levosimendan and the underlying heart disease of our patients, whose biventricular –and, especially, right ventricular– function improved.
The possibilities of a relationship between the effect on potassium channels, as well as the possible properties related to pharmacological post-conditioning, in addition to the improvement of renal perfusion through the treatment of low cardiac output, should be the starting point again clinical trials that have the ability to confirm our findings.
Limitations
The main limitation of this study was that patients were not randomized to treatment groups. The reason is that we preferred that physicians treated their patients with the medications they were more familiar with. This may have caused bias in the interpretation of results, as physicians will probably favor the conventional therapy. A triple-blinded study was not possible due to the type of study. However, the team conducting statistical analysis was blinded to the group analyzed. The use of certain biochemical parameters with greater sensitivity and specificity (N-GAL or Cystatin C), would also be useful to evaluate their action.