The hemodynamics of septic shock are characterized by peripheral arteriolar vasodilation, and this condition leads to a hyperdynamic state with low systemic vascular resistance, high cardiac output, hypotension and inadequate tissue perfusion1, 2). In clinical practice, there is no optimal therapeutic option for septic shock patients who are unresponsive to adequate fluid resuscitation and inotropics.

Nitric oxide (NO) has been shown to play a key role in the pathogenesis of septic shock3-5). Endotoxin and various cytokines such as interleukin (IL)-1, and tumor necrosis factor (TNF)-α in the circulation of septic shock patients stimulate the synthesis of NO by activating inducible nitric oxide synthase (iNOS)3, 6-8). The increased NO stimulates the soluble guanyl cyclase enzyme in the vascular smooth muscle, and this leads to the production of cyclic GMP that in turn results in endothelial smooth muscle relaxation. The final consequence of all of this is vasodilatation and arterial hypotension9).

Methylene blue (MB), is an inhibitor of the guanylate cyclase enzyme, and it has been studied as a potential vasopressor in septic shock. In fact, the administration of MB has been reported to transiently increase blood pressure in patients with septic shock10-15). Moreover, when MB was continuously infused into patients with septic shock, it counteracted myocardial depression, maintained oxygen transport and reduced adrenergic support compared with conventional treatment alone16,17). As for the underlying mechanisms of its hemodynamic effect on septic shock, MB has been suggested not only to hamper the action of NO by the inhibition of soluble guanyl cyclase of the vascular smooth muscle, but also to inhibit the production of NO18,19).

Considering that endogenous NO stimulates the expression of pro-inflammatory cytokines through the NF-kB pathway20-22), MB may also affect the endogenous production of the cytokines. However, it is not known whether the transient increase in blood pressure observed with the administration of MB is associated with the production of pro-inflammatory cytokines. Moreover, the acute hemodynamic effect of MB in septic shock patients who are unresponsive to the usual treatment has not been well reported on. We hypothesized that MB would increase the blood pressure by attenuating the synthesis of NO, IL-1 and TNF-α along with blocking the NO-induced vasodilation via guanylate cyclase inhibition in septic shock. The aims of this study were to assess the acute effects of MB on the hemodynamics and on the production of NO and pro-inflammatory cytokines in patients with refractory septic shock.

This study was approved by the institutional Ethics Committee of Asan Medical Center. An informed consent was obtained from each of the patients or from their next of kin.

MB induced an increase in the SVR resulting in the transient increase of the MAP in our patients with refractory septic shock, and this was without a decrease in the cardiac index. MB also induced an increase in the PVR, which resulted in an increase of the PAP. However, there was no deterioration of gas exchange. These findings were consistent with previous reports10-15). MB increased the left ventricular stroke work index without changing the cardiac filling pressure and with increasing the systemic vascular resistance. The myocardial function could be improved by MB because MB increased the sensitivity of the cardiovascular system to catecholamines by inhibiting the excessive production of cGMP23). If these effects could be maintained by MB infusion, MB could be a treatment option to the usual clinical practices that are employed for salvaging refractory septic shock patients. However, the effect of a single infusion of MB was not sustained in our patients for more than 4 h. Moreover, we could not maintain the elevated MAP with continuous MB infusion in two of our patients. Preiser et al. showed that the hemodynamic effects of MB were achieved with repeated does (4 mg/kg)10). Mikhail et al. demonstrated that continuously infused MB improved the cardiovascular function and reduced the requirement for adrenergic support16). A possible explanation for these different results might be the difference in the severity of the study patients. Those previous studies included septic shock patients who were unresponsive to more than 0.05 or 5 µg/kg/min of norepinephrine or dopamine, respectively. In our study, we included septic shock patients who were unresponsive to more than 0.1 or 20 µg/kg/min of norepinephrine or dopamine, respectively.

It is well known that NO is a causative molecule for the peripheral vasodilation observed in septic shock. The endotoxin and various pro-inflammatory cytokines, such as IL-1 and TNF-α, in the circulation of septic shock patients stimulate the synthesis of NO by activating inducible nitric oxide synthase (iNOS)3,6-8). On contrary, NO, which is an important endogenous regulatory molecule, has been implicated in both the pro-inflammatory and anti-inflammatory processes20). The enhanced NO generation may play an autoregulatory role in amplifying the synthesis of pro-inflammatory cytokines through the activation of NF-kB24). Considering these results, MB might alter the endogenous production of the pro/anti-inflammatory cytokines. In the research literature, MB reduced the endotoxin-induced TNF-α synthesis in Kupffer cells25), and MB decreased the plasma levels of the stable NO metabolites, nitrites and nitrates, as well as cGMP19). To determine whether the MB-induced increase in the SVR was associated with the endogenous production of pro/anti-inflammatory cytokines in septic shock patients, we assessed the acute changes of cytokines in the patients' blood after a single infusion of MB. We found that a single infusion of MB did not affect the endogenous production of IL-1, IL-10 and TNF-α during the study period, although the TNF-α and IL-10 concentrations in septic shock patients were 7 and 20 times higher, respectively, in the septic shock patients than in the normal volunteers. These results are consistent with Memis and colleagues' report in which MB infusion was shown not to change the cytokine levels or the clinical outcome in severe sepsis26). Memis measured the plasma levels of cytokines at baseline (15 min before start of the study), at immediately after MB infusion and at 24 and 48 h after MB infusion. We measured the blood level of the cytokines only at 2 h after MB infusion because such confounding factors as the underlying disease progression can affect the blood levels of the cytokines. In addition, TNF-α levels are known to rise rapidly after the injection of bacteria or endotoxin; in the animal models of shock and gram-negative sepsis, they reach peak levels at 60~90 min and then decrease 27). The IL-1 levels rise slowly and peak at 180 min28). Similar kinetics have been observed in human subjects who were injected with endotoxin29). We could not exclude the possibility of the late effect of MB on the endogenous production of the cytokines. However, Mikhail et al. noted that TNF-α and IL-6 levels remained unchanged16) after 24 h of MB treatment in their endotoxemic sheep model. Instead, the MB improvement of the cardiopulmonary function was associated with a reduction of the arachidonic acid metabolites thromboxane B2 and prostaglandin F1α30). MB seems to block prostacyclin production by a mechanism that is independent of inhibition of soluble guanyl cyclase31). Since cytokine levels rise and fall with different time courses in response to an inflammatory insult, further studies are needed about the time frame after MB infusion.

MB is known to generate superoxide anions, and this could antagonize the action of nitric oxide-induced vasodilatation by decreasing the available nitric oxide32,33). We found that a single infusion of MB did not decrease the endogenous production of NO. This result differs with a previous report that showed MB's inhibitory effect on the production of NO19). However, in animal models of sepsis induced by bowel perforation, the plasma NO level decreased after NOS inhibitor, but the plasma and urine NO levels were not changed after 4 h of MB administration34), and this is a finding consistent with our study. Martin et al. showed that MB is an inhibitor of the soluble guanylate cyclase activity that counteracts the effects of nitric oxide and other nitrovasodilators in the endothelium and vascular smooth muscle35). MB prevents the endotoxin- or IL-1-induced vasorelaxation by abolishing guanosine 3' and 5'-cyclic monophosphate production and MB reverses the endotoxin induced or cytokines induced hypotension and myocardial depression23,36). Our result supports the concept that MB exerts its antagonistic effect on NO by inhibiting the soluble guanyl cyclase without altering the NO production. There is another concern that although methylene blue may be able to reverse hypotension, it may not be able to improve organ blood flow and oxygen delivery37). Yet we found that the arterial lactate concentration was significantly decreased 30 min after MB infusion compared with its baseline, which suggests there was better tissue oxygenation.

Six patients among 11 MB responders and only 1 patient among 9 MB non-responders survived, suggesting a possible prognostic indicator of a single MB infusion. However, large randomized controlled studies are required to determine whether the initial hemodynamic response to MB infusion could be a prognostic index for patients with septic shock. The limitations of our study are the small number of patients and the absence of a control group to assess the effect of a similar volume of isotonic saline. In addition, we did not check the level of cGMP and the guanylate cyclase activity.

In conclusion, although a single administration of MB transiently elevated the MAP by increasing the SVR, endogenous productions of NO, IL-1, IL-10 and TNF-α were not altered during the study period in the patients with refractory septic shock. The acute vasopressor effect of MB seems to be mainly due to inhibition of guanyate cyclase activity. Further studies are required to determine whether MB could be an alternative to increasing vasopressor support for patients with refractory septic shock.