Objective: To determine gender differences in the innate immune response and vascular reactivity during human endotoxemia. Design: Clinical experimental study. Setting: University medical center intensive care research unit. Subjects: Fifteen female and 15 male volunteers. Interventions: Intravenous injection of 2 ng/kg Escherichia coli lipopolysaccharide. Measurements and Main Results: C-reactive protein, leukocytes, and cytokines were measured at regular time intervals as indicators of inflammation. Heart rate and blood pressure were continuously monitored. Forearm blood flow and the responsiveness of forearm vessels to the intrabrachial infusion of norepinephrine (1–3–10–30 ng/min/dL) were measured before and 4 hrs after the administration of endotoxin using venous occlusion plethysmography. Differences were tested with repeated-measures analysis of variance. Females showed a more proinflammatory response to lipopolysaccharide than males, illustrated by a higher rise in C-reactive protein (42 ± 3 vs. 29 ± 3 mg/L, p = .002) and more leukocyte sequestration (leukopenia 1.8 ± 0.1 × 109 vs. 2.4 ± 0.1 × 109, p = .003). The increase in cytokine levels showed a more proinflammatory pattern in females as reflected by a higher increase in tumor necrosis factor-α (965 ± 193 vs. 411 ± 35 pg/mL, p < .0001), whereas the increase of the anti-inflammatory interleukin-10 was not significantly different (95 ± 15 pg/mL in females vs. 129 ± 15 pg/mL in males, p = .288). Females exhibited higher baseline levels (9.9 ± 1.1 vs. 7.0 ± 0.8 pg/mL in males, p = .042) and an augmented increase in lipopolysaccharide-binding protein, which may explain the more pronounced inflammatory response in females. The lipopolysaccharide-induced change in heart rate was not significantly different between the genders, whereas blood pressure decreased more in females (p < .0001). Lipopolysaccharide administration significantly attenuated the norepinephrine sensitivity in males (p = .002), whereas no lipopolysaccharide-induced effect was observed in females (p = .552; difference between groups, p = .045). Conclusions: During experimental human endotoxemia, females showed a more pronounced proinflammatory innate immune response associated with less attenuation of norepinephrine sensitivity. These findings may be relevant in view of the profound and incompletely explained differences in incidence and outcome of sepsis among male and female patients.
Gender influences in vivo human responses to endotoxin. Shock 26:538-543, 2006. To the Editor: Recently, Coyle et al. (1) reported their findings regarding gender differences during experimental human endotoxemia. In a subgroup of 16 males and 8 females, no significant differences in circulating white blood cells, cortisol, or cytokine levels were detected during the first 4 h after the administration of endotoxin (lipopolysaccharide [LPS]). The study appears well conducted, and one of the strong aspects includes measurements of blood levels of sex hormones. We conducted a similar study in which LPS was administered to 15 males and 15 females (2). In contrast to Coyle et al., we found several marked gender differences conform with a more proinflammatory response to LPS in females. This was illustrated by a significantly higher rise in C-reactive protein and more leukocyte sequestration during endotoxemia, and a more pronounced increase in proinflammatory cytokines in females as reflected by a higher rise in tumor necrosis factor α (TNF-α), interleukin (IL) 1β, and interferon γ, whereas IL-10 tended to be less induced in females (Table 1). Although Coyle et al. found a similar trend for the proinflammatory cytokine IL-8 (which may not have reached significance because of a type 2 error), this was not the case for TNF-α: no difference between females and males was found by Coyle et al., but an impressive difference in our study. The reason of this discrepancy is not clear. Coyle et al. examined slightly older subjects (28 ± 1 years) compared with our study (22 ± 1 years), and possibly, ethnic differences play a role, as 64% of subjects in the study of Coyle et al. were nonwhite, whereas we studied white subjects only. Also, differences in the extent of prehydration before the administration of LPS may be of importance, as it can influence immune response dramatically (3).Table 1: Markers of inflammation (peak values)The observed higher innate immune response in females in our study is consistent with our finding that females exhibited higher baseline levels and an augmented LPS-induced increase in LPS-binding protein (LBP). Lipopolysaccharide-binding protein mediates the transfer of endotoxin to CD14 and signaling of toll-like receptor 4, leading to the initiation of the immune response and resulting in the production of inflammatory cytokines (4). This may explain the strong initial proinflammatory response in females. Interestingly, estrogen treatment increases LBP mRNA (5), which provides another link between sex hormones and innate immunity. The studies of Coyle et al. and our group illustrate that even with a highly standardized setup, results that are partly conflicting can be obtained. In our view, the study of Coyle et al. does not rule out the possibility of marked gender difference in the innate immune response and underlines the need for further research on a larger scale. Lucas T. van Eijk, MD Mirrin J. Dorresteijn, MD Peter Pickkers, MD, PhD Radboud University Nijmegen Medical Centre Nijmegen, The Netherlands
Endotoxin administration to animals and humans is an accepted experimental model of Gram-negative sepsis, and endotoxin is believed to play a major role in triggering the activation of cytokines. In septic patients, the IL-12/IL-18/IFN-gamma axis is activated and correlates with mortality. Our aim was to investigate the effects of endotoxin administration in humans on the activation of the IL-12/IL-18/IFN-gamma axis. Seven healthy volunteers received E. coli endotoxin (O:113). Hemodynamics, temperature and the course of plasma concentrations of TNF-alpha, IL-1beta, IL-12, IL-18 and IFN-gamma were determined. Endotoxin administration resulted in the expected flu-like symptoms, a temperature of 38.8 +/- 0.3(o)C (p < 0.003), a decrease in mean arterial blood pressure of 14.8 +/- 1.8 mmHg (p < 0.0002) and an increase in heart rate of 27.5 +/- 4.8 bpm (p < 0.002) compared to baseline values. TNF-alpha increased from 16.6 +/- 8.2 to 927 +/- 187 pg/mL (p < 0.003). IL-1beta increased from 8.6 +/- 0.5 to 25.3 +/- 2.0 pg/mL (p < 0.0001). IL-12 showed no significant increase (8.2 +/- 0.2 to 9.3 +/- 0.8 pg/mL, p = 0.13), and all IL-18 measurements remained below the level of detection. In contrast, IFN-gamma showed an increase from 106.6 +/- 57.1 to 152.7 +/- 57.8 (p < 0.005). These results indicate that pathways other than the IL-12/IL-18 axis may induce IFN-gamma production in human endotoxemia.
Sepsis-induced immunosuppression is a frequent cause of opportunistic infections and death in critically ill patients. A better understanding of the underlying mechanisms is needed to develop targeted therapies. Circulating bile acids with immunosuppressive effects were recently identified in critically ill patients. These bile acids activate the monocyte G-protein coupled receptor TGR5, thereby inducing profound innate immune dysfunction. Whether these mechanisms contribute to immunosuppression and disease severity in sepsis is unknown. The aim of this study was to determine if immunosuppressive bile acids are present in endotoxemia and septic shock and, if so, which patients are particularly at risk.
During septic shock, the vasoconstrictor response to norepinephrine is seriously blunted. Animal experiments suggest that hyperpolarization of smooth muscle cells by opening of potassium (K) channels underlies this phenomenon. In the present study, we examined whether K-channel blockers and/or nitric oxide (NO) synthase inhibition could restore norepinephrine sensitivity during experimental human endotoxemia.Volunteers received 2 ng/kg Escherichia coli endotoxin intravenously. Forearm blood flow (FBF) was measured with venous occlusion plethysmography. Infusion of 4 dose steps of norepinephrine into the brachial artery decreased the FBF ratio (ratio of FBF in the experimental arm to FBF in the control arm) to 84 +/- 4%, 70 +/- 4%, 55 +/- 4%, and 38 +/- 4% (mean +/- SEM) of its baseline value. After endotoxin administration, norepinephrine-induced vasoconstriction was attenuated (FBF ratio, 101 +/- 4%, 92 +/- 4%, 83 +/- 6%, and 56 +/- 7%; n = 30; P = 0.0018; pooled data). Intrabrachial infusion of the K-channel blocker tetraethylammonium (TEA) completely restored the vasoconstrictor response to norepinephrine from 104 +/- 5%, 93 +/- 7%, 93 +/- 12%, and 69 +/- 12% to 89 +/- 9%, 73 +/- 4%, 59 +/- 5%, and 46 +/- 8% (n = 6; P = 0.045). Other K-channel blockers did not affect the response to norepinephrine. The NO synthase inhibitor N(G)-monomethyl-l-arginine (L-NMMA; 0.2 mg x min(-1) x dL(-1) intra-arterially) also restored the norepinephrine sensitivity. In the presence of L-NMMA, TEA did not have an additional effect on the norepinephrine-induced vasoconstriction (n = 6; P = 0.9).The K-channel blocker TEA restores the attenuated vasoconstrictor response to norepinephrine during experimental human endotoxemia. Coadministration of L-NMMA abolishes this potentiating effect of TEA, suggesting that NO mediates the endotoxin-induced effect on vascular K channels. In the absence of an effect of the selective adenosine triphosphate-dependent K-channel blocker tolbutamide, we conclude that the blunting effect of endotoxin on norepinephrine-induced vasoconstriction is caused by NO-mediated activation of calcium-activated K channels in the vascular wall.
In agreement with the study by Fazekas and colleagues [1] in a recent issue of Critical Care, an increase in the concentration of carbon monoxide (CO) has been observed after surgery and cardiopulmonary bypass and during sepsis [2]. Although infl ammation induces heme oxygenase and the above-mentioned conditions do lead to infl ammation, a clear association in humans has not been established, underlining the relevance of the remarks made by Cove and Pinsky [3] in the same issue of Critical Care. We would like to present data that illustrate that infl ammation does increase CO in humans in vivo. During experimental endotoxemia in humans, a controlled immune response is evoked, leading to increased levels of several pro- and anti-infl ammatory markers [4]. Recent experiments demonstrated that, in 38 healthy male volunteers infused with US Standard Reference Endotoxin (National Institutes of Health, Bethesda, USA) obtained from Escherichia coli O:113 at a dose of 2 ng/kg, tumor necrosis factor-alpha concentrations increase at t = 2 hours to a median of 610 pg/mL (interquartile range of 400 to 853 pg/mL; Friedman test P <0.001) and arterial carboxyhemoglobin levels increase by 42% at 4 hours after lipopolysaccharide infusion (Friedman test P = 0.0057) (Figure 1). During these experiments, subjects breathed ambient air in a climate-controlled room; therefore, Cove and Pinsky’s claim that an increase of CO is due to an increase in inhalation is unlikely. Interestingly, the anti-infl ammatory eff ects of CO have also been studied in this model. Although CO has had clear benefi cial eff ects in several animal studies, inhaled CO during experimental endotoxemia failed to infl uence the infl ammatory response in humans [5], making it
