Levosimendan use in patients with acute heart failure and reduced ejection fraction with or without severe renal dysfunction in critical cardiac care units: a multi-institution database study

Background Acute heart failure is a life-threatening clinical condition. Levosimendan is an effective inotropic agent used to maintain cardiac output, but its usage is limited by the lack of evidence in patients with severely abnormal renal function. Therefore, we analyzed data of patients with acute heart failure with and without abnormal renal function to examine the effects of levosimendan. Methods We performed this retrospective cohort study using data from the Chang Gung Research Database (CGRD) of Chang Gung Memorial Hospital (CGMH). Patients admitted for heart failure with LVEF ≤ 40% between January 2013 and December 2018 who received levosimendan or dobutamine in the critical cardiac care units (CCU) were identified. Patients with extracorporeal membrane oxygenation (ECMO) were excluded. Outcomes of interest were mortality at 30, 90, and 180 days after the cohort entry date. Results There were no significant differences in mortality rate at 30, 90, and 180 days after the cohort entry date between the levosimendan and dobutamine groups, or between subgroups of patients with an estimated glomerular filtration rate (eGFR) ≥ 30 mL/min/1.73 m2 and eGFR < 30 mL/min/1.73 m2 or on dialysis. The results were consistent before and after propensity score matching. Conclusions Levosimendan did not increase short- or long-term mortality rates in critical patients with acute heart failure and reduced ejection fraction compared to dobutamine, regardless of their renal function. An eGFR less than 30 mL/min/1.73 m2 was not necessarily considered a contraindication for levosimendan in these patients.

treat AHF with reduced LVEF to improve cardiac output. Levosimendan is a Ca 2+ -sensitizer of the myocardium, which induces vasodilatation and increases cardiac output [2][3][4]. It is particularly effective in patients with acute decompensated heart failure by improving left ventricular ejection fraction (LVEF), decreasing B-type natriuretic peptide levels, improving renal function, and decreasing the need for renal replacement therapy in different settings [5]. Previous studies have demonstrated a comparable effect on improving cardiac output in heart failure patients to dobutamine, and it may result in a greater improvement in glomerular filtration rate than dobutamine in patients with cardiorenal syndrome [6][7][8][9][10][11][12]. However, evidence supporting the use of levosimendan in patients with severe renal dysfunction (i.e., an estimated glomerular filtration rate (eGFR) of < 30 mL/ min/1.73 m 2 ) is limited, particularly in patients with AHF and low cardiac output. In this study, we hypothesized that severe renal dysfunction may frequently be complicated with AHF, and that levosimendan may be beneficial in these patients. Therefore, the aim of this study was to analyze patients with AHF and reduced LVEF with or without abnormal renal function to examine the effects of levosimendan.

Data source
We performed this retrospective cohort study using data from the Chang Gung Research Database (CGRD) of Chang Gung Memorial Hospital (CGMH), which includes two medical centers, two regional hospitals, and three district hospitals around Taiwan. The total numbers of outpatient, emergency room and inpatient visits in 2018 in Taiwan were 420,283, 20,252 and 3172 per day, respectively [13]. Among the hospitals in Taiwan, CGMH has the largest volume of both inpatient (12.4%) and outpatient (21.2%) services in Taiwan. The CGRD comprises patient data derived from initial electronic medical records, and it was established for administrative and insurance purposes for CGMH. Disease diagnoses were coded using International Classification of Disease, Ninth Revision, Clinical Modification codes before 2016, and International Classification of Disease, Tenth Revision codes thereafter. Previous studies have provided additional detailed information on the CGRD [14,15]. The Institutional Review Board of the Chang Gung Medical Foundation approved this study.

Study design
Patients admitted for heart failure with decreased EF (defined as LVEF ≤ 40% by echocardiography) between January 2013 and December 2018 who received levosimendan or dobutamine in the critical cardiac care units (CCU) were identified from the CGRD. Patients who received levosimendan during the hospital stay were classified as the levosimendan group, regardless of whether they received dobutamine or other inotropic agents prior to the administration of levosimendan. All patients were prescribed inotropes because of heart failure with low cardiac output and systemic hypoperfusion (i.e., cardiogenic shock). Patients in the levosimendan group received continuous intravenous levosimendan at a dose of 0.05-0.2 µg per kilogram of body weight per minute for 72 h (without a loading dose). The date of cohort entry in the levosimendan group was defined as the day of the first levosimendan infusion during the hospital stay. To eliminate immortal time bias, the cohort entry date of the patients in the dobutamine group was assigned based on their counterparts in the levosimendan group [33]. Patients in the levosimendan group were matched with those in the dobutamine group by sex, age, eGFR, and LVEF. We excluded patients with extracorporeal membrane oxygenation (ECMO). The remaining patients were further classified according to their renal function using plasma creatinine and Modification of Diet in Renal Disease (MDRD) eGFR [16].

Covariates and outcomes
We analyzed the following covariates: age, sex, LVEF, eGFR, inotropic agent use during the admission, acute myocardial infarction (AMI) or percutaneous coronary intervention (PCI) during admission, length of intensive care unit stay, hospital days, comorbidities (atrial fibrillation, diabetes mellitus, hypertension and dyslipidemia), medications (aspirin, clopidogrel, ticagrelor, beta-blockers, angiotensin-converting enzyme inhibitors (ACEis), angiotensin II receptor blockers (ARBs), mineralocorticoid receptor antagonists (MRAs), digoxin, amiodarone and ivabradine), and baseline laboratory data (hemoglobin, alanine aminotransferase, B-type natriuretic peptide, HCO3, total bilirubin, blood urea nitrogen, sodium, potassium, platelets, hematocrit, white blood count, albumin, lactate and international normalized ratio). All patients underwent transthoracic echocardiography in the emergency department or after they had been admitted to a CCU. We followed the guidelines of the American Society of Echocardiography and used the biplane method to calculate the LVEF. The patients were defined as having a comorbidity if they had at least two outpatient diagnoses or one inpatient diagnosis for that comorbidity prior to the cohort entry date. Medication and laboratory data for 3 months prior to the cohort entry date were extracted. The outcomes of interest were mortality at 30, 90, and 180 days after the cohort entry date. We also evaluated in-hospital mortality after dosing. Patient follow-up was censored 6 months after the cohort entry date, last visit at CGMH, date of death, or December 31, 2018, whichever occurred first.

Statistical analysis
We evaluated differences in baseline characteristics between groups using standardized differences (STDs), where an absolute value of > 0.2 was considered to be a substantial difference. We performed propensity score matching (PSM) based on the propensity score derived from multivariable logistic regression to balance the two groups. Each patient in the levosimendan group was matched to two counterparts in the dobutamine group, if possible. Propensity scores were calculated using age, sex, LVEF, eGFR, use of inotropic agents (dopamine, norepinephrine, and epinephrine), AMI, myocarditis and mechanical ventilation during the index admission, intraaortic balloon pumping, Sequential Organ Failure Assessment (SOFA) score and Acute Physiology and Chronic Health Evaluation III (APACHE III) score. PSM was performed separately for those with an eGFR ≥ 30 mL/ min/1.73 m 2 and those with an eGFR < 30 mL/min/1.73 m 2 or dialysis. A univariate Cox proportional hazard model was used to compare the mortality rates between groups. Statistical significance was set at a two-sided P value of < 0.05, and no adjustment of multiple testing (multiplicity) was made in this study. We used SAS (version 9.4; SAS Institute, Cary, NC, USA) to perform all statistical analyses.

Results
After matching, 102 and 567 patients with eGFR > 30 mL/ min/1.73 m 2 were classified into the levosimendan and dobutamine groups, respectively, and 52 and 374 patients with eGFR ≤ 30 mL/min/1.73 m 2 or dialysis were classified into the levosimendan and dobutamine groups ( Fig. 1), respectively.

Baseline characteristics (supplemental digital content)
Before PSM, patients in the levosimendan group were more likely to receive dopamine and less likely to receive epinephrine during hospital stay. The usage rate of norepinephrine was comparable. Myocarditis, AMI, PCI, and IABP were more common in the levosimendan group. After PSM, 63 patients in the levosimendan group with an eGFR ≥ 30 mL/min/1.73 m 2 had 2 counterparts and 19 patients had only 1 counterpart, resulting in a total of 145 patients in the dobutamine group. After PSM, 38 patients in the levosimendan group with an eGFR < 30 mL/ min/1.73 m 2 or dialysis had 2 counterparts and 5 patients had only 1 counterpart, resulting in a total of 81 patients in the dobutamine group. No significant group differences were observed in age, sex, LVEF, eGFR, inotropic agent use, and AMI after PSM (Tables 1, 2).

Discussion
To the best of our knowledge, this is the first study to comprehensively investigate the short-and long-term survival of critical patients with both AHF and severe renal dysfunction who received levosimendan. We found that the patients who received levosimendan had similar survival rates to the patients who received dobutamine. Moreover, the differences in mortality rate were consistently nonsignificant between the two groups up to 180 days of follow-up, regardless of whether their eGFR was > 30 or < 30 mL/min/1.73 m 2 or they were on maintenance dialysis.
Previous studies have reported that the use of levosimendan in patients with heart failure can improve cardiac output, urine amount, and eGFR. These results have been demonstrated in patients with acute or chronic left ventricular systolic dysfunction [6,7,10,11,[17][18][19][20]. However, few studies have established the efficacy and safety of levosimendan in patients with severe renal dysfunction. Despite a lack of convincing evidence, depressed eGFR was thought to reduce the renal excretion of levosimendan leading to fatal arrhythmia, and several studies about levosimendan even excluded patients with an eGFR of < 30 mL/min/1.73 m 2 [11,21]. The Taiwan Food and Drug Administration (FDA) also   [22,23]. In contrast, the neutral effect on myocardial diastolic function and arrhythmogenic potential may suggest that levosimendan is a suitable alternative to dobutamine [24][25][26]. Levosimendan plays an important role in treating patients with severe renal dysfunction in our clinical practice. In our CCUs, patients with acutely deteriorating heart failure are treated using diuretic agents, nitroglycerin, and ACEis (or angiotensin II antagonists) first, and we only prescribe inotropic agents for those with signs of systemic tissue hypoperfusion, including oliguria or a rapidly worsening plasma creatinine level. Because low cardiac output and systemic hypoperfusion in patients with reduced left ventricular systolic function is a critical situation, our intensivists are not limited with regard to the choice of treatment. It depends on their clinical judgement, including whether or not they combine inotropic agents with norepinephrine. The initial eGFR was ≥ 30 mL/min/1.73 m 2 in most of our patients, but then became further depressed due to deteriorating cardiac function. We consider that renal dysfunction is reversible and not a contraindication for using levosimendan. For patients with an eGFR < 30 mL/min/1.73 m 2 , levosimendan is prescribed based on detailed discussions among the patients, their family members, and our critical medical teams in our clinical practice. The available medical evidence is discussed at these meetings [27][28][29][30], and the patients are carefully monitored to avoid adverse events. In this study, 53.9% of the patents in the levosimendan group received dopamine or dobutamine before the initiation of levosimendan in the index admission. This suggests that the patients were poor responders to dopamine or dobutamine. In our practice, some poor responders will receive mechanical cardiac support, such as veno-arterial ECMO or a temporary ventricular assist device (Centrimag), instead of levosimendan [34][35][36]. In the current study, none of the patients received a durable ventricular assist device or heart transplantation, and this situation is common in the medical environment in Taiwan.
The average LVEF of our study patients was < 30%. Before PSM, more patients in the levosimendan group had AMI and received PCI in the index admission, and dopamine and intra-aortic balloon pumps were more frequently used in this group. There was also a higher rate of myocarditis, however the difference was small. After PSM, the rates of AMI, PCI, mechanical ventilation, IABP, APACHE III, and SOFA scores between the two groups were similar. There were no significant differences in mortality rates up to 6 months between the two groups. In the subgroup with an eGFR < 30 mL/ min/1.73 m 2 , the survival rate of the patients who received levosimendan was also equivalent to the patients who received dobutamine. Our results suggest that levosimendan did not increase mortality even in these critical AHF patients with severe renal dysfunction.

Limitations
There are several limitations to this study. First, because of the retrospective and real-practice design, we could not limit our intensivists in their choices of inotropic agents. The prescription of levosimendan was based on the clinical judgement of the intensivists. For example, the rate of dopamine use was higher in the levosimendan group at baseline, so we used PSM to minimize bias. In total, 53.9% of the patients who received levosimendan were prescribed dobutamine or dopamine before their index date, and the outcomes may have been affected if the clinicians decided to continue dobutamine or dopamine at that time. These results can be only confirmed by well-designed prospective studies; however, it would be ethically wrong if the patients are considered to be poor responders to initial inotropic agents and clinicians do not adjust their therapeutic policies. Second, we excluded patients with ECMO implantation from our study. Venoarterial ECMO is considered to be an emergency lifesaving method for patients with AHF complicated with refractory cardiogenic shock. However, refractory cardiogenic shock is a contraindication for levosimendan. Moreover, the disease severity in the patients who needed ECMO was highly heterogeneous, ranging from postcardiotomy shock to ECMO-assisted cardiopulmonary resuscitation for prolonged cardiac arrest. Therefore, differences in mortality rate between those who received levosimendan and dobutamine may have been affected by the use ECMO. Third, we used ICD codes to identify patients with heart failure and other comorbidities. This is common in database studies, however it can lead to misestimations [31,32]. To minimize possible bias,  . 3 All-cause mortality rates of the patients with eGFR < 30 mL/min/1.73 m 2 or dialysis who received levosimendan versus dobutamine alone before (a) and after (b) propensity score matching we only enrolled patients who received inotropic agents and had a reduced ejection fraction, thereby reinforcing the diagnosis of heart failure. Fourth, we used pre-dosing plasma creatinine to estimate GFR. Because creatinine is not an instant parameter reflecting renal function, the baseline GFR may have been overestimated in this study.
Collecting urine for 24 h increases the accuracy of GFR values; however, this cannot always be done for such critical patients.

Conclusions
The critical AHF patients with or without severe renal dysfunction who received levosimendan had similar survival rates compared to the patients who received dobutamine while patients with veno-arterial ECMO were excluded. According to our results, an eGFR < 30 mL/ min/1.73 m 2 is not necessarily a contraindication for levosimendan.