Because sepsis-induced myocardial dysfunction related to sepsis is at least partially related to a decrease in cardiac myofilament response to calcium, the use of the new myofilament-calcium sensitizer, levosimendan, has been proposed. In addition, arginine vasopressin is increasingly proposed as a vasopressor in septic patients, although data on its effects on cardiac function are still scarce. The aim of the current study was to assess, invasively and noninvasively, whether levosimendan, arginine vasopressin, and norepinephrine, either alone or combined, may modify sepsis-induced myocardial dysfunction and renal hemodynamics.
Thirty-six hours after lipopolysaccharide or saline administration, rabbits were studied either after slight sedation for echocardiography or after general anesthesia with sodium pentobarbital for the following measurements: aortic flow velocity and maximum acceleration of blood flow in the ascending aorta and renal macrocirculation and microcirculation.
Levosimendan improved, within 30 min of administration, both maximum acceleration of blood flow by 20 +/- 12% (n = 8; P < 0.05) and left ventricular shortening fraction by a similar extent. Furthermore, low doses of arginine vasopressin markedly deteriorated cardiac function via an afterload-independent mechanism, even when animals were pretreated with levosimendan, whereas norepinephrine showed no detrimental effects on cardiac function. The study also showed that norepinephrine often improved renal medullary blood flow, whereas arginine vasopressin consistently decreased it.
Levosimendan and norepinephrine both exert beneficial effects in endotoxemic animals and should be further explored in human sepsis trials.
ALTERATIONS in cardiac function and peripheral vascular tone are consistently present in severe sepsis and in septic shock1–5with or without fluid resuscitation.6,7A recent Consensus conference emphasized that early treatment of septic shock includes adequate fluid loading associated or not with pressors and/or cardiac inotropes.4Among vasopressors, dopamine or more frequently norepinephrine is the treatment of choice.8,9In North America, arginine vasopressin is emerging as an alternative therapy of vasodilatory shock due to systemic inflammatory response syndrome.10–12Although norepinephrine and arginine vasopressin seem to be the two most efficient vasopressors, doubts remain about their potential detrimental effects on septic cardiac dysfunction, especially via the increase in left ventricular afterload.
In studies focused on potential mechanisms that may contribute to sepsis-induced cardiomyopathy, it was recently hypothesized that, during sepsis, cardiac myofilaments phosphorylation decreases myofibrillar-calcium sensitivity.1,13The latter may contribute to the depression of cardiac contractility probably via a modulation of the regulatory action of troponin I on troponin C.1It was therefore proposed that an agent that may improve cardiac myofilament response to calcium (i.e. , a calcium-sensitizing agent) may specifically improve contractility in sepsis. Levosimendan is a calcium-sensitizing agent that is indicated for use in patients with acutely decompensated heart failure.14,15The calcium-sensitizing effects of levosimendan are mediated through its calcium-dependent binding to cardiac troponin C,16,17which produces an increase in the contractile force of the cardiac myocytes18without increases in intracellular calcium concentrations19and with little or no increase in myocardial oxygen consumption.20In a recent study, pretreatment with levosimendan in pigs subjected to endotoxin shock improved cardiac output and systemic and gut oxygen delivery.21Other authors showed that levosimendan may attenuate, ex vivo , myocardial dysfunction in hearts isolated from guinea pigs, 4 h after lipopolysaccharide administration.22In both previous studies, however, levosimendan was tested on animal models before or few hours after endotoxin administration. Therefore, our aim was to establish the role of levosimendan in treating sepsis-related myocardial failure by studying its effects on a chronic animal model of endotoxemia that was previously shown to mimic human septic cardiomyopathy.3,13
Accordingly, our aim was to assess the cardiovascular effects of levosimendan alone or of arginine vasopressin or norepinephrine either alone or, most importantly, combined with levosimendan in lipopolysaccharide-treated rabbits with a cardiovascular depression. Our study showed that (1) levosimendan improved cardiac systolic function in lipopolysaccharide-treated rabbits, and (2) arginine vasopressin, although at doses that only slightly increased mean arterial pressure, markedly deteriorated cardiac function even when combined with levosimendan, whereas norepinephrine showed no detrimental effects on cardiac function.
Materials and Methods
Reactives and Drugs
The endotoxin mixture was prepared with Escherichia coli , Salmonella enteritidis , and Salmonella minnesota lipopolysaccharides (Sigma Chemical Co., St. Louis, MO) as previously described.3,13The day of experiment, levosimendan powder was solubilized with 0.05 m NaHCO3and then diluted in 0.9% saline to obtain a final solution of 200 μg/ml. A 1-mg/ml stock solution of arginine vasopressin (Sigma Chemical Co., St. Louis, MO) was prepared and stored at −80°C until use. Norepinephrine (Noradrénaline Aguettant®; Laboratoire Aguettant; Fougères, France) was prepared from a stock solution of 2 mg/ml.
Care of the animals conformed to the recommendations of the Helsinki Declaration and the study was performed in accordance with the regulations of the official edict of the French Ministry of Agriculture. Fifty-seven male White New Zealand rabbits (Charles River, L’Arbresle, France) weighing 2.2 ± 0.1 kg were housed in individual cages in a controlled environment with free access to food and water for at least 4 days before experimentation. Thirty-six hours before starting treatment and hemodynamic measurements, 600 μg/kg lipopolysaccharide mixture or saline (control) was injected via a 22-gauge catheter placed in a marginal ear vein, as previously described.3,13Animals were fasted the night before experimentation.
Overall mortality was 19% (n = 9) in lipopolysaccharide group, quasi-exclusively during the 36 h after lipopolysaccharide administration, and 0% in control. Therefore, data presented here are the summary of the 38 animals that survived in lipopolysaccharide group and 10 control animals.
In the main series of rabbits, general anesthesia was induced by intravenous injection of sodium pentobarbital (CEVA Santé Animale, Libourne, France). The dose of sodium pentobarbital required to achieve an optimal anesthesia without deleterious hemodynamic effects was 21 ± 6 mg/kg in the lipopolysaccharide group and 35 ± 6 mg/kg in the control group (P < 0.0001) for the induction of anesthesia and 6 ± 2 mg · kg−1· h−1in the lipopolysaccharide group and 13 ± 1 mg · kg−1· h−1in control group (P < 0.0001) to maintain anesthesia. After tracheostomy, animals were mechanically ventilated with 100% oxygen (Rodent ventilator; Harvard Apparatus, Boston, MA). Rectal temperature was maintained near 38.5°C by using a warming pad. Catheters were inserted in right carotid and right jugular vein, and a 20-MHz pulsed Doppler flow velocity probe (4 mm internal diameter) was positioned around the ascending aorta.
A surface laser Doppler probe (Transonic System, Ithaca, NY) was placed on the renal convexity between the capsule and the cortex. A 26-gauge laser Doppler needle probe (Transonic Needle System) was inserted in the medulla after making a small hole in the capsule of at least 1 cm far away from the cortical probe. The two laser Doppler probes were connected to a Doppler flow meter (BLF21D; Transonic System). The mean signals from renal artery, cortex, and medulla were measured continuously.
In another series of spontaneously breathing rabbits, transthoracic echocardiography (Vivid7 General Electric; Horten, Norway) with a 12-MHz transducer was performed during light sedation with sodium pentobarbital.
Hemodynamic Variables and Calculations
In the series of rabbits under general anesthesia, the following variables were measured. Heart rate (beats/min), arterial blood pressure (mmHg), systolic aortic blood flow velocity (sAoV; cm/s), systolic and diastolic renal artery blood flows (ml/min), and cortical and medullary blood flows (in arbitrary units, Tissue Perfusion Units) were continuously measured at a sampling rate of 1,000 s−1, and recorded on a computer (Macintosh Power PC6745) with an analogic/digital transducer (Biopac Systems MP100, Goleta, CA). Baseline aortic blood flow zero velocity was verified during the diastolic time. Aortic acceleration was calculated as the sAoV first derivative. Peak values of sAoV and aortic acceleration (maximal aortic acceleration [Gmax]) were used as indexes of systolic function.23–25
In the series of spontaneously breathing rabbits, transthoracic echocardiography was applied in parasternal long-axis view at a frequency of 120 Hz. It allowed the following measurements in M mode: (1) left ventricular end-diastolic diameter, defined as the largest left ventricle diameter; (2) left ventricular end-systolic diameter, defined as the smallest left ventricle diameter; and (3) left ventricle fraction shortening (%), as
left ventricular end diastolic diameter − left ventricular end systolic diameter/left ventricular end diastolic diameter × 100%.
The echocardiographer was blinded to treatment.
In the rabbits under general anesthesia, after a recovery period of 30–45 min after surgery, three consecutive hemodynamic measurements were performed over 30 min to assess the stability of the preparation. For each variable, the average value between the three recordings was considered as baseline. Twenty milliliters saline was infused in 5 min to assess cardiac response to fluid challenge. Hemodynamic measurements were repeated at the end of fluid challenge.
In a first set of experiments, rabbits in the lipopolysaccharide group (n = 8) and control rabbits (n = 6) received a continuous infusion of levosimendan at a dose previously described in animals (200 μg · kg−1· h−1)21during 4 h. In these animals, levosimendan plasma concentration, measured as previously described,26was 286 ± 92 ng/ml at 60 min and slightly further increased to 336 ± 84 ng/ml at 240 min.
In a second set of experiments, the effects of pressor agents were tested on cardiac performance in lipopolysaccharide group. Thus, only moderate doses of norepinephrine and arginine vasopressin, with a moderate impact on blood pressure to limit the effect of afterload modification, were tested. Lipopolysaccharide-treated animals (n = 26) were randomized and separated in four groups, receiving, respectively, levosimendan (200 μg · kg−1· h−1) alone during 30 min followed by the combination of levosimendan and norepinephrine (1 μg · kg−1· min−1, usual dose in human studies) for an additional 90 min (n = 7); saline solution during 30 min followed by norepinephrine (1 μg · kg−1· min−1) alone during 90 min (n = 6); levosimendan (200 μg · kg−1· h−1) alone during 30 min followed by the combination of levosimendan with arginine vasopressin (0.005 U · kg−1· min−1, determined in pilot studies to induce similar changes in blood pressure than norepinephrine [1 μg · kg−1· min−1] in lipopolysaccharide-treated rabbits) for an additional 90 min (n = 6); or saline solution during 30 min followed by arginine vasopressin (0.005 U · kg−1· min−1) alone during 90 min (n = 7).
Hemodynamic measurements were repeated every 15 min during treatment. Blood samples were drawn from the carotid artery during the baseline period (before fluid challenge) and at the end of the study period for biologic assessment. The following measurements were performed in plasma: glucose, lactate (spectrophotometric technique, Ektachem 950 IRC system; Johnson & Johnson, Strasbourg, France); blood gas (ABL-30; Radiometer, Copenhagen, Denmark); arginine vasopressin27and troponin I (by immunoenzymatic technique as previously described in rabbits28).
Spontaneously breathing rabbits included four lipopolysaccharide-treated animals and four control animals. Echocardiographic measurements were performed before and 30 min after levosimendan administration (200 μg · kg−1· h−1).
Results were expressed as mean ± SD. Intergroup and intragroup differences were tested, in data presented in all figures, by two-way and one-way analyses of variance for repeated measures, respectively. A Mann–Whitney test was performed when intergroup differences were tested at one time point, as in table 1. A P value less than 0.05 was considered significant.
Myocardial Effects of Lipopolysaccharide Treatment
In vivo injection of lipopolysaccharide altered body weight and metabolic parameters, 36 h after treatment, compared with saline rabbits, as summarized in table 1.
Figure 1shows that cardiac systolic function was impaired 36 h after lipopolysaccharide treatment as previously reported.3,13Fluid challenge–induced increase in sAoV was smaller in the lipopolysaccharide group than in the control group (P < 0.05). In addition, the index of cardiac systolic function Gmax was lower (table 1) and remained so after fluid challenge in the lipopolysaccharide group compared with the control group (fig. 1; P < 0.05).
Lipopolysaccharide-induced impairment in myocardial function was confirmed by echocardiography in spontaneously breathing rabbits: Left ventricular shortening fraction was impaired in lipopolysaccharide-treated (n = 4) compared with control (n = 4) animals (28 ± 2 vs. 36 ± 1%, respectively; P < 0.05, Mann–Whitney test).
Figure 1also shows that mean arterial pressure was lower, although not significant, in the lipopolysaccharide group than in the control group and remained lower after fluid challenge.
Beneficial Effects of Levosimendan on Cardiac Performance in Lipopolysaccharide-treated Rabbits
As shown in figure 2, the calcium-sensitizer levosimendan improved myocardial contractile function in both lipopolysaccharide-treated and control animals. In lipopolysaccharide-treated animals, levosimendan improved both sAoV and Gmax by +13 ± 8% and 20 ± 12%, respectively (P < 0.05 for both), as early as 30 min after the start of treatment. Improvement in contractile function persisted during the whole study period. During the same period, levosimendan induced a decrease in mean arterial pressure in all studied animals (fig. 2) and an increase in heart rate: +33 ± 13 beats/min, corresponding to 6 ± 10% in lipopolysaccharide-treated animals, and +19 ± 7 beats/min, corresponding to 6 ± 2% in control animals (P < 0.05 for both).
A beneficial effect of levosimendan on cardiac performance was confirmed by echocardiography, showing an improvement in left ventricle shortening fraction in lipopolysaccharide-treated and control rabbits (combined data, n = 8, from 32 ± 4 to 36 ± 2%; P < 0.05).
Hemodynamic Effect of Norepinephrine, Arginine Vasopressin, and Levosimendan in Lipopolysaccharide-treated Rabbits
Mean arterial blood pressure increased significantly with norepinephrine by +7 ± 9 mmHg (n = 13; P = 0.01) but not with arginine vasopressin by +7 ± 12 mmHg (n = 13; P = not significant). However, diastolic arterial blood pressure improved with both norepinephrine (+8 ± 10 mmHg) and arginine vasopressin (+12 ± 12 mmHg; P < 0.05 for each drug).
Figure 3shows cardiac effects of norepinephrine or arginine vasopressin associated or not with levosimendan in lipopolysaccharide-treated animals. Although norepinephrine never altered myocardial function indexes over 90 min of treatment, arginine vasopressin induced a striking deterioration in both sAoV and Gmax in all arginine vasopressin–treated rabbits even when pretreated with the inotropic agent levosimendan (all P < 0.05). In addition, norepinephrine did not alter heart rate when associated or not with levosimendan, whereas arginine vasopressin did decrease heart rate by −21 ± 22 beats/min over 90 min in all arginine vasopressin–treated animals (P < 0.05). Furthermore, ejection time remained unchanged over the treatment period in all norepinephrine-treated animals (137 ± 38 ms before and 129 ± 37 ms after 90 min of treatment) and in all arginine vasopressin–treated animals (119 ± 32 ms before and 124 ± 36 ms after 90 min of treatment).
Renal Effects of the Vasoactive Drugs
As shown in table 1, no significant difference was observed in renal blood flow in basal conditions between lipopolysaccharide-treated and control rabbits. Levosimendan alone did not alter systolic or diastolic renal blood flow or cortical or medullary perfusion within 30 min of administration in lipopolysaccharide-treated rabbits (data not shown). Arginine vasopressin and norepinephrine also did not alter systolic (fig. 4) or diastolic (data not shown) renal blood flow, regardless of the use of levosimendan in lipopolysaccharide-treated animals. However, as shown in figure 4, norepinephrine did increase medullary tissue perfusion in the absence of levosimendan (P < 0.05) but not when combined with levosimendan, whereas arginine vasopressin did consistently decrease medullary perfusion whether it was used with or without levosimendan (P < 0.05). Neither arginine vasopressin nor norepinephrine altered cortical perfusion (data not shown).
As shown in table 1, plasma arginine vasopressin concentration was lower in lipopolysaccharide-treated rabbits compared with control rabbits.
Interestingly, differences were observed in the time course of glycemia and lactatemia among the four groups of lipopolysaccharide-treated animals (n = 26). Although glycemia was similar among the four groups at baseline (9.4 ± 2.4 mm), norepinephrine-treated animals had a higher glycemia (13.3 ± 3.8 mm) compared with arginine vasopressin–treated rabbits (9.1 ± 2.2 mm; P < 0.01) after 90 min of treatment. By contrast, although lactate was similar among the four groups at baseline (2.9 ± 2.2 mm), norepinephrine-treated animals had a lower lactate concentration (3.3 ± 3.0 mm) compared with arginine vasopressin–treated rabbits (4.8 ± 2.9 mm; P < 0.05) after 90 min of treatment. Troponin I did not differ significantly among the four groups at any time (data not shown).
Our study showed that (1) levosimendan improved cardiac function in endotoxemic rabbits; (2) low doses of arginine vasopressin markedly deteriorated cardiac function, even when animals were pretreated with levosimendan, whereas norepinephrine showed no detrimental effects on cardiac function; and (3) norepinephrine and arginine vasopressin had opposite effects on renal medullary blood flow.
Levosimendan is a myofilament-calcium sensitizer that improves cardiac contractility with coronary and peripheral vasodilatation.29The primary mechanism of action for levosimendan is through Ca2+sensitization of contractile proteins, although it does not affect the intracellular Ca2+concentrations.18,19Our group showed that the profound myocardial depression observed in sepsis was mostly related to intrinsic alterations in cardiac myocytes function rather than direct cardiodepressing effects of circulating agents.2,3,13Our group and others showed that intrinsic alteration of cardiac myocyte function in sepsis was not related to changes in intracellular calcium concentration but rather to a decrease in myofilaments’ response to calcium in several animal species.1,2,13We further showed that the latter was related to an increased phosphorylation of troponin I2,13that modulated the regulatory action of troponin I on troponin C. In the current study, we showed both by invasive measurements in anesthetized animals and by echocardiography in spontaneously breathing animals that levosimendan markedly improved sepsis-induced myocardial dysfunction. However, as shown in figure 2, levosimendan decreased mean arterial pressure in our endotoxemic rabbits over the 4-h study period. Although one may suggest that levosimendan-induced improvement in indexes of cardiac performance could be related to decrease in left ventricle afterload, figure 2shows that the increase in maximum acceleration of aortic blood flow in the ascending aorta was maximum at 15 min at the time when mean arterial pressure was only slightly decreased in endotoxemic animals. In addition, although mean arterial pressure further decreased after 30 min, neither aortic velocity nor maximum acceleration changed. Accordingly, our data showed that levosimendan directly improved cardiac performance in our endotoxemic animal model, regardless of the level of arterial pressure. Further experiments should be performed in septic patients to confirm the beneficial effects of levosimendan on sepsis-induced myocardial depression. Care should be taken to evaluate and maintain, if needed, organ perfusion pressure with vasoactive agent, optimally norepinephrine, as shown below.
Arginine vasopressin is increasingly used to restore arterial blood pressure in human septic shock, and promising data recently were recently obtained in animal studies of hemorrhagic shock and cardiopulmonary arrest.12,30Relative vasopressin deficiency was repeatedly found in patients with severe sepsis.31,32A similar result was found in our animal model, with baseline plasma arginine vasopressin concentration lower in lipopolysaccharide-treated than in control animals (table 1). Few human studies dealt with the use of arginine vasopressin in septic shock, and data on myocardial function are still missing. Two randomized studies have shown that 0.01–0.04 U/min arginine vasopressin was effective in the correction of sepsis-induced hypotension.10,33Although cardiac output remained unchanged in the two studies, this does not exclude an intrinsic deterioration of myocardial contraction. To assess arginine vasopressin effects on myocardial contraction in sepsis, we performed beat-by-beat measurements of systolic aortic velocity and maximal aortic acceleration, previously described as good indexes of myocardial contraction.23–25Our study shows that despite a slight and nonsignificant increase in mean arterial pressure (+7 mmHg), arginine vasopressin alone induced a marked decrease in heart rate as previously described34but more importantly deteriorated the indexes of myocardial contraction: sAoV and Gmax. Because ejection time remained constant during arginine vasopressin infusion, the decrease in sAoV also represented a deterioration in stroke volume. Therefore, altogether, the decrease in both heart rate and stroke volume associated with the administration of arginine vasopressin imply a marked deterioration in cardiac output in our lipopolysaccharide-treated rabbits. Interestingly, pretreatment with the positive inotropic agent levosimendan could not prevent arginine vasopressin–induced deterioration in cardiac function. This should be further investigated and might be at least partially related to a similar effect of both arginine vasopressin and levosimendan on myocardial adenosine triphosphate–sensitive potassium channels.35,36Accordingly, our study demonstrated that arginine vasopressin induced an afterload-independent deterioration of cardiac function. This is further supported by the lack of deterioration in cardiac function with norepinephrine alone despite a similar increase in mean arterial pressure in our endotoxemic rabbits. This difference between arginine vasopressin and norepinephrine might be related to a potential β-receptor agonist effect of norepinephrine37that was not described with arginine vasopressin. Of note, the norepinephrine-induced increase in glycemia was also likely related to a β-receptor agonist effect. Although troponin I was not altered by arginine vasopressin or norepinephrine during the study period, one cannot exclude a direct vasoconstrictive effects on coronary circulation in septic rabbits that was recently described with arginine vasopressin in isolated perfused normal rabbit hearts.38Whether the effect of arginine vasopressin on cardiac contractile function is related to reduced coronary perfusion or to a direct effect on cardiac myocytes should be investigated.
Our study further assessed effects of levosimendan, arginine vasopressin, and norepinephrine on renal macrocirculation and microcirculation. Data on the effects of levosimendan on renal function are few and remain controversial.21,39Pagel et al. 39showed that levosimendan increases renal medullary blood flow and decreases cortical blood flow in normal dogs, whereas Ordner et al. 21showed no change in renal blood flow in a very early phase of sepsis. Our study confirmed the absence of effects of levosimendan on both macrocirculation and microcirculation in septic animals despite a marked improvement in cardiac output. The consequences of levosimendan on renal function, including creatinine clearance and glomerular filtration, should be further investigated in septic animals. Our study further compared the effects of arginine vasopressin and norepinephrine on renal hemodynamics. As shown in figure 4, despite a similar and modest increase in mean arterial pressure, arginine vasopressin and norepinephrine had divergent effects on renal medullary blood flow: Norepinephrine alone markedly increased renal medullary circulation, whereas arginine vasopressin alone rapidly deteriorated it. Norepinephrine is increasingly used as first-line therapy to restore mean arterial pressure in septic patients. Increasing recent evidences demonstrated no harmful and even beneficial effects of norepinephrine or phenylephrine on renal function in animals40and in humans during sepsis.41,42Our study further shows that such improvement in renal function with no change in renal blood flow could be related to a change in renal blood flow partition toward medullary regions of the kidneys.
By contrast, arginine vasopressin worsened medullary renal blood flow, with no change in cortical renal blood flow, in our endotoxemic rabbits. Previous works showed that the vasoconstrictive effects of arginine vasopressin induce an increase in arterial blood pressure together with a decrease in both cortical and medullary blood flow in normal animals.43,44Albert et al. 34recently confirmed the latter in normal rabbits but more importantly showed that arginine vasopressin increased mean arterial blood pressure, renal blood flow velocity, mostly in its diastolic component, and cortical renal blood flow in acutely endotoxemic rabbits, whereas medullary flow was unchanged. Furthermore, Sun et al. 45showed an arginine vasopressin increase in urine output in the early phase of sepsis in sheep. Interestingly, our current study extends these data by showing that a low dose of arginine vasopressin that barely increased mean arterial blood pressure may decrease medullary blood flow in a more severe model of chronic endotoxemic animals. Further studies are needed to investigate whether the decrease in renal medullary flow observed in our study could be related to a direct vasoconstrictive effect of arginine vasopressin on renal medullary vessels or indirectly to the arginine vasopressin-induced decrease in cardiac output and how this impacts renal function and urine output.
A couple of limitations of the current study are noteworthy. Intravenous lipopolysaccharide does not fully represent a model of sepsis with a bacterial source. However, we previously showed3,13that our model mimics the extend of depression and the time course of myocardial dysfunction previously described in septic patients.46Anesthesia did little to interfere with hemodynamic results presented above. The major results, namely lipopolysaccharide-induced impairment and the beneficial effect of levosimendan, both on myocardial function, were confirmed by echocardiography in slightly sedated, spontaneously breathing animals. Assessment of myocardial function in our study focused on systolic aortic velocity, ejection time, and maximal aortic acceleration using Doppler techniques as well as shortening fraction by echocardiography. Those parameters exclusively explore left ventricular systolic function with some dependency on load changes. Results should be confirmed with the pressure–volume loop technique that may even further explore myocardial diastolic parameters as relaxation and ventricular compliance.
In summary, our study explored cardiac and renal effects of new cardiovascular agents, including arginine vasopressin and levosimendan, with potential use in severe sepsis and septic shock in humans. Levosimendan improved, arginine vasopressin deteriorated, and norepinephrine had no effect on cardiac contractility in endotoxemic animals. In addition, norepinephrine and arginine vasopressin had opposite effects on renal medullary blood flow. Accordingly, levosimendan and norepinephrine both exert beneficial effects in endotoxemic animals and should be explored in human sepsis trials.
The authors thank Daniel Bichet, M.D. (Head of Research Center, Department of Medicine, Hôpital du Sacré-Coeur de Montréal, Université de Montréal, Quebec, Canada), for the measurement of plasma arginine vasopressin in rabbits.