• adverse events
• advanced heart failure
• Levosimendan
• outcomes

Background and aims: Repeated, planned infusions of levosimendan (rp-Levo) are utiiized mostly in an in-hospital setting in some patients with advanced/refractory heart failure, to maintain clinical stability and end organ function, Our study aims to evaluate the effectiveness and safety of monthly levosimendan infusion in outpatients with AHF, and to identify potential predictors of therapeutic efficacy.
Methods: 75 patients, of whom 43 (53%) were listed for Heart Transplant ( HTx), were treated with rp-Levo (0.05-0.15 mcg/kg/min for 24-48 hours every 4 weeks) at the day-service of a single center, from Jan 2012 to Feb 2019. Patients went home after the first 2 hours of infusion. Actuarial survival and freedom from death or treatment failure, hospital readmissions, trend of laboratory parameters, predictors of treatment failure, and adverse events were evaluated.
Results: During a median follow-up of 16 months (IQR 11-31), 24 patients (32 %) died, of whom 20 on medical therapy, 3 after Left Ventricular Assist Device (LVAD) implant and 1 after HTX. Kaplan Meier overall actuarial survival was 85% at 1 year, with a significant advantage in HTX-listed pts (Logrank χ 2= 14.609, p<0.001) versus those treated for symptoms control. After a median of 12 (IQR 6-19) months, 41 patients (55%) met the composite endpoint of death or any treatment failure (LVAD or urgent HTx, switch to another inotrope for refractory HF), without differences between HTX listed patients and those treated for symptoms. 45 (60%) patients were hospitalized at least one time for a total of 60 hospitalization (0.8 hospitalizations per patients) for a median of 13 (0-33) days spent in hospital (DIH) during treatment. Time to first hospitalization was 187 (135-350) days. The DIH of the first 6 months on rp-Levo was significantly lower with respect to the 6 months preceding rp-Levo start, either including ( p= 0.018 ) or excluding (p< 0.001 ) the last hospitalization on rp-Levo. No significant increase in the levels of NT-proBNP, creatinine or bilirubin was found up to 1 year. The change (Δ) of NT-proBNP at 3 months was the most accurate parameter in predicting treatment failure (sensitivity 75%, specificity 60%, Area Under the Curve= 0.700, cut off= - 14%), being the sole significant predictor of treatment failure at multivariate Cox Analysis (HR 1.009, 95%C.I. 1.000-1.017, p= 0.044). Over a total of 1121 administrations, there were 33 potentially drug related adverse events ( 13 VT/VF, tachyarrhythmias, 7 supraventricular tachyarrhythmias, 2 bradyarrhythmias, 4 symptomatic hypotensions and 6 acute renal failures). Only 9 of them in the first 10 days from levosimendan infusion.
Conclusions: Rp-Levo may be safely delivered in an outpatient setting for maintaining clinical stability and possibly reducing or delaying readmissions in pts with AHF. The trend of NT-proBNP levels on treatment may help in predicting treatment failure. Heart replacement therapies such as LVAD and HTx are still fundamental to modify the prognosis of patients with advanced to refractory disease.