Levosimendan in Early Sepsis: When Good Ideas Give Poor Results

Septic shock is characterized by hypovolemia, decreased vascular tone, and myocardial depression. In addition, impairment in splanchnic perfusion may be observed.

Because tissue perfusion may remain altered after fluid optimization and correction of hypotension, inotropic drugs are frequently used to improve cardiac output. Dobutamine and recently levosimendan have been proposed for this purpose.

Several studies have shown that dobutamine effectively increases cardiac output in septic shock patients. In addition, β-adrenergic stimulation is associated with an increase in splanchnic perfusion. Multiple experimental and human studies have shown that dobutamine increases splanchnic perfusion in sepsis.^1–3 Unfortunately, dobutamine administration is also associated with undesirable effects, such as arrhythmias, increased metabolism, and slight immunosuppressive effects, so that there may be reluctance to use this drug.

Levosimendan is an interesting inotropic drug that is active through nonadrenergic pathways. It is an inotropic drug with associated vasodilatory properties, acting as a calcium sensitizer as well as a phosphodiesterase inhibitor and activator of adenosine triphosphate–dependent potassium channels.⁴ Levosimendan increases cardiac output in septic shock by increasing systolic and diastolic function,⁵ but its impact on splanchnic perfusion has not been well studied. In a pig endotoxic shock model, Oldner et al.⁶ reported that levosimendan improved gut and portal blood flows. Of note, the drug was given as a pretreatment, so these results may not apply to clinical scenarios. Dubin et al.⁷ reported that in endotoxic sheep, levosimendan preserved mesenteric perfusion better than dobutamine. Morelli et al.⁸ reported that, in patients with septic shock not responding to dobutamine, levosimendan decreased gastric Pco₂, suggesting that mesenteric perfusion improved in these patients. However, all these data were obtained in fluid-resuscitated animals and under normotensive conditions. Inotropic drugs are sometimes initiated very early in the course of sepsis.⁹ Accordingly, it is important to better evaluate the effect of inotropic drugs under these conditions.

In this issue of Anesthesia & Analgesia, Cunha-Goncalves et al.^10,11 present 2 companion articles evaluating the impact of levosimendan on global hemodynamics and splanchnic circulation in a pig model of early septic shock induced by endotoxin injection. In the first study,¹⁰ fluid resuscitation was minimal (crystalloids at a fixed rate of 10 mL · kg⁻¹ · h⁻¹). In control animals, mean arterial blood pressure and stroke volume decreased, and cardiac index was preserved because of a compensatory increase in heart rate. Whole-body oxygen consumption (VO₂) increased. Gut and hepatic oxygen delivery decreased but gut VO₂ was preserved, whereas liver VO₂ decreased. In the levosimendan-treated animals, levosimendan was begun at 90 min and continuously infused at a dose of 50 μg · kg⁻¹ · h⁻¹. Compared with endotoxic controls, levosimendan-treated animals exhibited a further decrease in mean arterial blood pressure and stroke volume. Cardiac output was preserved by a further increase in heart rate. Evolution of gut and liver perfusion and metabolism was similar to that of endotoxic control animals. Lactic acidosis developed in levosimendan animals but not in endotoxic control animals. The conclusion of this first article was that levosimendan failed to improve systemic and splanchnic perfusion.

The strong vasodilatory properties of the drug are probably involved in this negative response. Vasodilation leads to a decrease in cardiac preload, probably through blood pooling in large capacitance veins, so that stroke volume decreases, even in the presence of a further reduction in afterload. The fact that the fraction of flow directed to the splanchnic region was not affected by levosimendan suggests that levosimendan did not have specific splanchnic vasodilator effects or that the impact on splanchnic perfusion was counterbalanced by a decrease in perfusion pressure.

To evaluate the impact of perfusion pressure and fluid resuscitation, the authors conducted a second experimental study in which animals were more aggressively fluid resuscitated (addition of large amounts of dextrans and crystalloids) and in which norepinephrine was used to maintain mean arterial blood pressure more than 65 mm Hg.¹¹ The splanchnic effects of levosimendan and dobutamine were compared. Both drugs were initiated at 120 min, and the dose was modulated to achieve a 15% increase in cardiac output (levosimendan doses 25–50 μg · kg⁻¹ · h⁻¹ and dobutamine doses 10–20 μg · kg⁻¹ · min⁻¹). In control endotoxic animals, stroke volume decreased with time, despite maintenance of mean arterial blood pressure more than 65 mm Hg and higher amounts of fluid administration. Gut and liver oxygen delivery decreased, and again gut but not liver VO₂ was preserved. Lactate levels increased progressively. Shortly after their initiation, both dobutamine and levosimendan increased cardiac output by 15%, the predefined goal, but only dobutamine maintained it over time, whereas achievement of this goal was more transient with levosimendan. Stroke volume was preserved with dobutamine, whereas it decreased with levosimendan after a transient increase. Heart rate markedly increased in levosimendan-treated animals. Only dobutamine maintained splanchnic perfusion, which decreased with levosimendan. Interestingly, the fraction of flow directed to the splanchnic area decreased with both inotropic drugs. Arterial lactate levels further increased in levosimendan animals but were unchanged in dobutamine-treated animals.

This second study¹¹ also provided important information. The decrease in stroke volume and cardiac index at later stages in levosimendan-treated animals suggests that the shortening in diastolic time was too severe with levosimendan, inducing a reduction in preload and hence stroke volume. Because arterial blood pressure was similar in endotoxic controls and levosimendan-treated animals, it is likely that venous pooling in capacitance veins and afterload were similarly affected.

Why did the authors observe such unexpected negative results with levosimendan? Possibly the doses of levosimendan were too low or at least lower than in other studies showing more favorable effects of levosimendan.⁷ We cannot rule out that higher doses may produce more pronounced effects, but this is unlikely, because maximal doses of inotropic stimulation were achieved in a minority of animals, both in levosimendan and dobutamine groups. A more likely explanation is that persistent hypovolemia contributed to this paradoxical response, and that relative hypovolemia should be more aggressively corrected before initiating this drug. Of note, the tachycardia induced by levosimendan was somewhat unexpected, because other studies have shown that this drug induces less tachycardia than dobutamine.⁷ This excess tachycardia is probably a sign of concurrent hypovolemia, probably exacerbated by the vasodilatory properties of the drug. Dobutamine did not generate such a tachycardia, probably because its slight α-adrenergic properties helped to maintain preload. Tachycardia may have shortened diastolic time to such an extent that stroke volume could not be maintained, even though levosimendan has similar inotropic and even stronger lusitropic properties compared with dobutamine.⁵

Finally, both studies suggest that levosimendan is not a splanchnic vasodilatory drug, and that the beneficial impact of levosimendan reported in some studies is probably related to changes in cardiac output more than to specific splanchnic effects. This has already been reported with dobutamine.^3,12 Hence, using inotropic drugs to specifically improve the splanchnic circulation may be futile.

In conclusion, inotropic drugs, and especially levosimendan, should not be given before correction of hypovolemia and restoration of arterial blood pressure. Inotropic drugs may improve splanchnic perfusion, but this effect is mostly related to the increase in cardiac output.