The excess release of nitric oxide (NO) resulting from the activation of the inducible isoform of the NO synthase (iNOS, NOS2) is referred to as a "final mediator" of arterial hypotension during septic shock. However, NO is a janus-headed mediator during sepsis with both beneficial and deleterious effects. In the same model of endotoxin-challenged swine, iNOS inhibition (1) or infusing the NO donor SIN-1 (2) resulted in a comparable improvement of the intestinal microcirculatory perfusion and the arterial-intestinal PCO2 gap. Therefore, NOS inhibition in patients with septic shock is a matter of debate (3) despite the experimental evidence available even from clinically relevant large-animal models (4, 5). Although a phase II study using the competitive nonselective NOS inhibitor Nω-methyl-l-arginine (546C88; L-NMMA) showed promising results (6), the subsequent multicenter phase III trial was terminated prematurely because of increased mortality in the treatment arm (7). An accompanying editorial questioned the design of the latter study because protocol changes resulted in higher drug infusion rates and the inclusion of patients without hyperdynamic circulation (8). Moreover, this editorial emphasized that "the failure of nonselective NO inhibition does not necessarily mean failure of other more selective forms of NO inhibition" (8). In fact, treatment with L-NMMA was superior to fluid resuscitation alone in septic primates (9) but did not compare favorably with the standard clinical vasopressor treatment, i.e., norepinephrine, in other clinically relevant large-animal models (10-12). Therefore, several selective iNOS inhibitors were studied in a variety of experimental models of septic shock. Whereas pharmacological iNOS blockade and genetic iNOS deletion yielded controversial results in unresuscitated murine sepsis or endotoxemia (13-15), the same approach improved organ function (16, 17) and ultimately reduced mortality (18) in well-resuscitated murine models. In large-animal models, various selective iNOS inhibitors attenuated the organ dysfunction resulting from endotoxemia (19) or bacterial sepsis (20-22). Unfortunately and in contrast to the investigation of L-NMMA, none of these studies compared selective iNOS inhibition with the clinical standard norepinephrine titrated to the same macrohemodynamic targets. In this issue of Shock, Su et al. (23) now report on a comparison of the newly developed imidazopyridine derivative 2-[2-(4-methoxy-pyridin-2-yl)-ethyl]-3H-imidazol[4,5-b]pyridine (BYK191023), a highly iNOS-selective NOS blocker active both in vitro and in vivo, with norepinephrine and the combination of these two compounds during ovine fecal peritonitis-induced septic shock. With or without additional norepinephrine, BYK191023-treated animals showed less pulmonary artery hypertension and gas exchange impairment and higher visceral organ blood flow. These beneficial macrocirculatory effects coincided with lower plasma cytokine concentrations, and both attenuated hyperlactatemia and better maintained markers of microcirculatory perfusion. Thus, the data confirm previous findings in endotoxic and bacteremic swine receiving the selective iNOS inhibitors 1400W or L-NIL (19, 20). The findings by Su et al. (23) are fascinating because they originate from a long-term large-animal model, which is characterized by a hyperdynamic circulation with a sustained increase in cardiac output and progressive hypotension despite aggressive fluid resuscitation. Furthermore, peritonitis is one of the leading causes of sepsis in surgical intensive care, and the authors have the merit of using a "post-treatment design," i.e., drug administration started after the development of sepsis and titrated according to clinical end points. Nevertheless, several questions remain open. First, the authors did not report acid-base status, and thus it is unclear whether the smaller rise in blood lactate levels in the BYK191023-treated animals was caused by a less pronounced impairment of tissue ischemia as suggested by the higher proportion of perfused capillaries and/or to the reduced norepinephrine requirements. It is well established that norepinephrine per se causes hyperlactatemia because of activation of the Cori cycle (24). Second, the higher visceral organ blood flow in the BYK191023-treated ewes did not result in improved organ function or survival. This observation is particularly important with respect to renal dysfunction. Albeit iNOS inhibition attenuated tubular damage during human endotoxemia (25, 26), possibly caused by improved peritubular capillary perfusion (27), and ovine burn/smoke injury (28), other authors could only confirm this key role of NO in acute kidney injury resulting from ischemia/reperfusion (29) but not endotoxemia (30). Finally, the question remains whether iNOS really is the (only?) "bad guy" (31). Selective inhibition of the neuronal NOS isoform (nNOS, NOS1) also effectively attenuated ovine septic acute lung injury (32, 33). Again, nothing is simple and easy. Although iNOS-derived NO was shown to protect diastolic relaxation during polymicrobial murine sepsis (16), genetic NOS3 deletion impaired both systolic contraction and diastolic relaxation (34) because of impaired cardiomyocyte Ca2+ handling and mitochondrial ATP production (35).
Read full abstract