Cardiac failure in septic shock is not always corrected by vasoactive-inotropic drugs and fluid therapy [1]. Levosimendan is a new inodilator for primitive cardiac failure acting by sensitising troponin-C to calcium and opening the adenosine-triphosphate-sensitive potassium channels. In addition, it has phosphodiesterase-inhibiting properties [2]. Animal data (in vitro [2] and pre-treatment in vivo [3]) showed a positive effect when used in septic shock, as well as in different types of shock in humans, some of which were septic [4]. The drug’s mode of action in septic cardiac failure could also be different from calcium sensitivity [2].

We used levosimendan in two cases of septic shock with cardiac failure unresponsive to standard treatment (Table 1). Patient A was a 26-year-old, previous healthy female, admitted in shock due to bacterial pneumonia. Day 1: mechanical ventilation, fluid loading and continuous infusion of vasopressors improved the hemodynamic status but myocardial competence was getting worse with anuria, gut ischemia, intestinal bleeding, and cutaneous necrosis (Day 2).

Table 1 Best hemodynamic parameters found in the selected days in patient A and B. Days 1–2: standard septic-shock therapy. Day 3: before levosimendan infusion (pre-) and during 24-h treatment. Day 8 (patient A), day 6 (patient B): withdrawal of amine infusion. (NE norepynephrine, DB dobutamine).
Full size table

Patient B was a 79-year-old male, with dilated cardiomyopathy, admitted in shock due to acute cholangitis. Day 1: despite mechanical ventilation, fluid loading and continuous infusion of vasopressors, a severe hypo-dynamic state was present. Day 2: we tried to adjust vasopressor dosages without significant clinical change. In both cases, on day 3, a loading dose of levosimendan (12 μg/kg, but not in patient B because of atrial fibrillation at high frequency) followed by a 24-h continuous infusion at 0.1 μg·kg·min was administered.

A clinical improvement of shock was recorded at 2 h of levosimendan administration. There was improvement of pulmonary venous oxygen saturation and a decline in serum lactate levels. The peak of the hemodynamic effect was reached between 6–18 h allowing a concomitant decrease of vasoactive dosages until suspension on days 8 and 6 respectively, strongly suggesting a causal relationship. Of note, there was a trend toward pulmonary artery decrease as suggested by experimental data [3], not associated with worsening of pulmonary shunt. Patient A died from brain haemorrhage on day 23 after resolution of septic shock and severe pneumonia. Patient B was discharged to the ward after 14 days.

Our report confirms the effectiveness of levosimendan in human septic shock [4], supporting a potential application in a definite scenario (severe septic cardiac failure) and timing (unresponsiveness to standard therapy).