Effects of Carbon Monoxide Inhalation during Experimental Endotoxemia in Humans
http://www.100md.com
美国呼吸和危急护理医学 2005年第2期
Departments of Clinical Pharmacology, Medical and Chemical Laboratory Diagnostics
Anesthesiology and General Critical Care Medicine, Medical University of Vienna, Vienna, Austria
ABSTRACT
Data show that carbon monoxide (CO) exerts direct antiinflammatory effects in vitro and in vivo after LPS challenge in a mouse model. We hypothesized that CO may act as an antiinflammatory agent in human endotoxemia. The aim of this trial was to study the effects of CO inhalation on cytokine production during experimental human endotoxemia. The main study was a randomized, double-blinded, placebo-controlled, two-way cross-over trial in healthy volunteers. Each volunteer inhaled synthetic air (as placebo) and 500 ppm CO for 1 hour in random order with a washout period of 6 weeks and received a 2-ng/kg intravenous bolus of LPS after inhalation. Carboxyhemoglobin levels were assessed as a safety parameter. CO inhalation increased carboxyhemoglobin levels from 1.2% (95% confidence interval, 1.0 to 1.4%) to peak values of 7.0% (95% confidence interval, 6.5 to 7.7%). LPS infusion transiently increased plasma concentrations of tumor necrosis factor-, interleukin (IL)-6 (approximately 150-fold increases), and IL-8, as well as IL-1 and IL-1 mRNA levels (an approximately 200-fold increase). These LPS-induced changes were not influenced by CO inhalation. Inhalation of 500 ppm CO for 1 hour had no antiinflammatory effects in a systemic inflammation model in humans, as 250 ppm for 1 hour did in rodents.
Key Words: carboxyhemoglobin endotoxin tumor necrosis factor-
Carbon monoxide (CO) originates from the oxidation or combustion of organic matter (1). At high concentrations, exposure to CO is lethal, usually through accidental smoke or exhaust inhalation in enclosed spaces (2).
Cigarette smoke accounts for a major source of CO exposure in humans. In addition to uptake of exogenous gas, cells and tissues produce significant amounts of CO (3, 4). These two "kinds" of CO are different; one is produced locally, at the cellular level, as a component of heme degradation, and the other is incorporated systemically during exposure to various concentrations. Data show that CO exerts direct antiinflammatory effects after LPS challenge in vitro and in an in vivo mouse model (5). The key finding with the mouse model was that mice exposed to 250 ppm CO for 1 hour before LPS administration responded with significantly lower levels of proinflammatory cytokines (tumor necrosis factor- [TNF-] and interleukin [IL]-1) and higher levels of IL-10 than control mice. As a consequence, the role of CO in various rodent models was intensively studied and yielded encouraging results (6eC8). The latest results show that CO has a pharmacodynamic action in pigs undergoing cardiopulmonary bypass (9), extending our knowledge to larger animals. However, CO inhalation was combined with constant perfusion of the pigs' hearts with a CO-saturated cardioplegic solution, which obscures the relative contribution of CO inhalation to the beneficial effects.
On the basis of the rationale provided by these animal studies, we hypothesized that CO may act as an antiinflammatory agent in human endotoxemia as well, decreasing the release of proinflammatory cytokines such as TNF- in vivo. We therefore studied the effects of CO inhalation on systemic inflammation during experimental human endotoxemia.
The infusion of low doses of endotoxin into humans provides a well-standardized model to study the pharmacodynamics of drugs with supposed antiinflammatory properties in vivo (10). We used this model to characterize putative pharmacodynamic effects of CO in humans.
METHODS
This trial and subsequent amendments were approved by the Medical University of Vienna Ethics Committee, and written informed consent was obtained from all participants before enrollment.
Pilot Trial
Effects of CO inhalation on systemic inflammation have been investigated only in animals so far. Thus, we decided to conduct a pilot trial with nine healthy male nonsmokers, aged 19 to 40 years, to document the safety and tolerability of CO inhalation in humans. The pilot trial was designed as a randomized, placebo-controlled, four-way cross-over dose-escalation study. All volunteers inhaled placebo (synthetic air) and CO concentrations of 10, 50, and 250 parts per million (ppm; 250 ppm had a marked antiinflammatory effect in mice) for a time period of 1 hour randomly assigned to periods 1, 2, 3, and 4 with a washout phase of 1 week between periods. Having referred to the literature on maximal carboxyhemoglobin (HbCO) levels in smokers, we assumed that inhalation of CO, toxic at higher levels, would be innocuous as long as HbCO levels remained at 10% or less (11). All volunteers were asked to complete a "CO assessment sheet" before and immediately after inhalation, at which time we asked about the existence and intensity (none < mild < moderate < severe) of subjective symptoms such as headache, nausea, and vertigo. Because HbCO levels did not reach our predetermined objective of 10%, we invited the same nine volunteers (or adequate substitutes) to inhale further CO concentrations of 250 ppm for 2 hours and 500 ppm for 1 hour in a randomized, two-way cross-over design with a 1-week washout period in between.
Blood samples were collected before and directly after inhalation. Immediately thereafter, LPS (Escherichia coli lipopolysaccharide, 50 pg/ml; Sigma, Vienna, Austria) was added to the blood tubes, which were then incubated in a water bath at 37°C for 2 hours. After incubation, plasma was obtained by immediate centrifugation at 2,000 x g (15 minutes at 4°C) and stored in 0.5-ml aliquots at eC80°C until batch analysis of TNF-, IL-1 and IL-1, IL-6, and IL-8. IL-1 and IL-1 mRNAs were also isolated after 2 hours of incubation.
Main Trial
The main trial was conducted as a randomized, double-blind, placebo-controlled, two-way cross-over study at the Department of Clinical Pharmacology (Medical University of Vienna, Vienna, Austria). Each carbon monoxideeCnaive volunteer inhaled synthetic air (as placebo) and 500 ppm carbon monoxide in random order with a washout period of 6 weeks in between both treatments. To achieve similar TNF- peak levels compared with the study in mice (5), we chose an LPS dose of 2 ng/kg.
Study Subjects
Thirteen healthy male nonsmokers aged 18 to 40 years were invited to participate in this trial. Medical screening included medical history, physical examination, laboratory parameters, and virologic and standard drug screening. In addition, study subjects were tested for hereditary thrombophilia, that is, factor V Leiden and protein C and S deficiency, to minimize potential risks of endotoxin-induced coagulation activation as previously described (12). Exclusion criteria included regular or recent intake of medication including over-the-counter drugs, and clinically relevant abnormal findings in medical history or laboratory parameters.
Study Protocol
The experimental procedures of our endotoxin model have been described in detail in other trials (12, 13). In short, volunteers were admitted to our department at 8:00 A.M. after an overnight fast that was extended until 4 hours after LPS infusion. After a resting period of 15 minutes, basal blood samples were drawn from an indwelling catheter and vital signs (blood pressure, pulse rate, and oxygen saturation) were measured. Shortly thereafter, volunteers started to inhale 500 ppm carbon monoxide (Messer Austria GmbH, Gumpoldskirchen, Austria) or synthetic air (80% N, 20% O2; Messer Austria) as placebo for 1 hour while sitting in an upright position. For this purpose we used cushioned full face masks (Mirage NV full face mask Series II; Pulmomed, Vienna, Austria), which were fixed with Velcro straps. During inhalation venous blood samples were drawn every 15 minutes, by a physician not otherwise involved in the trial, to monitor HbCO levels. For safety reasons, additional HbCO measurements were performed 1.5 and 4 hours after inhalation. All samples were analyzed with a fully automatic spectral photometric measuring system (AVL 912 CO-Oxylite; AVL List, Graz, Austria) and cross-checked by staff of the emergency laboratory (ABL 700; Radiometer, Copenhagen, Denmark). An HbCO level greater than 10% was defined as a stop criterion. Immediately after inhalation volunteers received a bolus of national reference endotoxin (E. coli LPS, 2 ng/kg; U.S. Pharmacopeia, Rockville, MD). In addition, isotonic saline solution (administered at 200 ml/hour; Leopold Pharma, Vienna, Austria) was administered to all subjects within 5 minutes of LPS infusion and was continued over 6 hours to maintain adequate hydration.
Blood Sampling
Blood samples were collected by venipuncture into citrated or ethylenediaminetetraacetic acideCanticoagulated tubes (Vacutainer; BD Biosciences, Vienna, Austria) before inhalation, directly after inhalation and 1.5, 2, 4, 6, 8, and 24 hours after LPS administration. Plasma was obtained by immediate centrifugation at 2,000 x g (15 minutes at 4°C) and stored in 0.5-ml aliquots at eC80°C until batch analysis.
Cytokine Analysis
Plasma levels of TNF-, IL-6, and IL-10 were measured by (high) sensitivity enzyme immunoassays (12). Plasma IL-8 levels were determined with a highly sensitive chemoluminescence enzyme assay (detection limit, 0.8 pg/ml); all assays were purchased from R&D Systems (Abingdon, UK). Prothrombin fragment (F1+2) was used as a marker of in vivo thrombin generation, which was also measured by an enzyme immunoassay (Behring, Marburg, Germany).
Real-time Quantitative Reverse Transcriptase-Polymerase Chain Reaction
Because plasma levels of IL-1 are hardly detectable in humans even under endotoxemia (14) we decided to quantify IL-1 mRNA.
All blood samples were immediately processed to avoid storage-induced changes in mRNA levels (15). After isolating total RNA with an RNeasy minikit (Qiagen, Valencia, CA) according to the manufacturer's instructions, mRNA was directly transcribed into cDNA, using a TaqMan reverse transcription reagents kit (Applied Biosystems, Foster City, CA), and stored at eC80°C until analysis. IL-1 mRNA quantification was performed with the ABI PRISM 7700 (Applied Biosystems), using commercially available predesigned primers and probes (Applied Biosystems). We chose 18S as housekeeping gene for multiplexing (Applied Biosystems) because of its stable expression under endotoxemia (B. Jilma, unpublished data). IL-1 mRNA was normalized against the reference gene (18S) according to the 2eCCT method (16), and data are expressed as fold increase over baseline values.
Data Analysis
Data are expressed as means and 95% confidence interval (95% CI). All statistical comparisons of continuous variables were made with nonparametric tests. After repeated measures analysis of variance with treatment and period as independent variables, the Friedman analysis of variance and the Wilcoxon test for post hoc comparisons were used to assess time-dependent changes in outcome variables within groups. A two-tailed p value less than 0.05 was considered significant. All statistical calculations were performed with commercially available statistical software (Statistica version 6.1; StatSoft, Tulsa, OK). Power calculation was done as previously described (17).
RESULTS
Pilot Trial
CO inhalation was well tolerated by our nonsmoking volunteers, whose mean age was 25 years (range, 20 to 40 years), and whose body mass index (BMI) averaged 23.8 kg/m2 (range, 21.5 to 26.3 kg/m2). One volunteer reported a mild headache after inhalation of 250 ppm CO for 2 hours. Seven volunteers were not able to participate in the second part of the pilot trial because of individual time-scheduling problems. They were replaced by adequate substitutes. CO inhalation did not have any significant effect on vital parameters, leukocyte and neutrophil counts, as well as cytokine levels. Because inhalation of 250 ppm CO for 2 hours yielded roughly the same peak HbCO levels as 500 ppm CO for 1 hour, we chose a CO concentration of 500 ppm over 1 hour for the main trial. The results of the pilot trial are summarized in Table 1.
Main Trial
Demographic data.
All participating subjects were healthy male nonsmokers. Their mean age was 29 years (range, 18 to 38 years), and their BMI averaged 23.0 kg/m2 (range, 19.0 to 26.8 kg/m2).
Effects of CO on HbCO levels.
CO inhalation was well tolerated by our nonsmoking volunteers and no adverse events, as assessed by asking about subjective symptoms and monitoring vital signs, occurred. Baseline HbCO values of 1.2% (95% CI, 1.0 to 1.4%) increased five- to sixfold time dependently to peak values of 7.0% (95% CI, 6.5 to 7.7%) after 1 hour of inhalation of 500 ppm CO. As expected, inhalation of synthetic air did not alter basal HbCO levels significantly (Figure 1).
Vital parameters.
LPS infusion induced a transient increase in heart rate and a transient decrease in mean arterial blood pressure (all p < 0.05 versus time), as previously reported (18). CO inhalation did not have any influence on these LPS-induced transient changes (Figure 2). LPS infusion also induced a transient increase in body temperature that reached its maximum after 4 hours and was not altered by CO inhalation (Figure 2D).
Effects of CO on inflammatory cells and C-reactive protein.
As expected, neutrophil counts increased about threefold and monocyte counts sharply dropped compared with baseline levels (all p < 0.05 versus time) after LPS infusion. CO inhalation did not have any influence on these transient changes (Figure 3). Furthermore, C-reactive protein plasma levels increased (2.4 mg/dl; 95% CI, 2.1 to 2.7 mg/dl) 24 hours after LPS infusion in comparison with baseline values of less than 0.5 mg/dl. Again, the magnitude of C-reactive protein increase was not influenced by CO inhalation (data not shown).
Effects of CO on proinflammatory cytokines.
LPS infusion induced transient increases in plasma concentrations of TNF-, IL-6, IL-8, as well as IL-1 and IL-1 mRNA levels (all p < 0.05 versus time), which returned to baseline values after 24 hours as shown in previous trials (13, 19).
In subjects treated with placebo before LPS injection, plasma TNF- levels peaked 2 hours after LPS infusion (126 pg/ml; 95% CI, 98 to 155 pg/ml), whereas CO-treated subjects showed peak TNF- levels slightly earlier, 1.5 hours after LPS infusion (163 pg/ml; 95% CI, 68 to 258 pg/ml) (Figure 4A).
Peak IL-6 values were seen 2 hours after LPS infusion, with no significant differences between periods (267 pg/ml [95% CI, 173 to 361 pg/ml] in the placebo period versus 282 pg/ml [95% CI, 160 to 404 pg/ml] in the CO period) (Figure 4B).
IL-1 and IL-1 mRNA levels showed a distinct increase over baseline values 2 hours after LPS infusion: IL-1, a 170-fold increase (95% CI, 45- to 294-fold increase) in the placebo period versus a 202-fold increase (95% CI, 49- to 355-fold increase) in the CO period (Figure 5A); IL-1, a 199-fold increase (95% CI, 38- to 360-fold increase) in the placebo period versus a 134-fold increase (95% CI, 37- to 232-fold increase) in the CO period (Figure 5B).
Similar to IL-6, peak IL-8 values were detectable 2 hours after LPS infusion, with no significant differences between periods (202 pg/ml [95% CI, 103 to 301 pg/ml] in the placebo period versus 212 pg/ml [95% CI, 88 to 335 pg/ml] in the CO period), although IL-8 levels were slightly lower in the CO period at 4 hours (Figure 4C).
Effects of CO on IL-10.
IL-10 is an inhibitor of proinflammatory cytokine synthesis and as such can limit the inflammatory process, including the production of TNF- (20). CO inhalation, however, did not significantly influence IL-10 plasma levels after LPS infusion. Peak values were reached 2 hours after LPS injection and were slightly higher in CO-treated subjects (28 pg/ml; 95% CI, 18 to 37 pg/ml) compared with those treated with placebo (25 pg/ml; 95% CI, 15 to 34 pg/ml) (Figure 4D).
Effects of CO on coagulation.
As expected, LPS injection was associated with a potent activation of the coagulation system (21) as reflected by increases in the plasma concentrations of prothrombin fragment F1+2, peaking after 4 hours [4.0 nmol/L; 95% CI, 3.2 to 4.8 nmol/L) (p < 0.05 versus time). This LPS-induced coagulant response was not significantly influenced by CO inhalation and showed the same time course in both groups. Peak levels in the CO period were 3.6 nmol/L (95% CI, 2.8 to 4.3 nmol/L) (Figure 3C).
DISCUSSION
Rodent models provide evidence that inhalation of relatively low concentrations of carbon monoxide have antiinflammatory effects (5eC7). In particular, inhalation of CO at 250 ppm for 1 hour significantly reduced TNF- and IL-1 in an experimental mouse endotoxemia model (5). This provided a sound rationale to investigate these putative antiinflammatory effects in humans.
Under the conditions that we used, we did not find a similar effect (5). We failed to show direct antiinflammatory effects of 500 ppm CO inhaled for 1 hour in experimental human endotoxemia, as measured by cytokine production. This discrepancy may have several explanations, some of which are likely interrelated.
First, blood of different species differs markedly in its affinity for CO. The affinity constants of mouse hemoglobin for CO and O2 differ by only 50-fold (22) as compared with a 217-fold higher affinity for CO than O2 shown by human hemoglobin (23). Second, the half-lives of HbCO in mice (30 minutes) and humans (5 hours) differ by 10-fold (24, 25). This demonstrates that CO dissociation from mouse hemoglobin is much faster. Third, saturation of hemoglobin differs markedly between species. For example, small animals such as hamsters, which are comparable in size to mice, have HbCO values exceeding 20% when breathing 250 ppm CO for 1 hour (26). In contrast, larger animals such as miniature pigs show HbCO values less than 10% after 1 hour (26), and HbCO values of about 12% were reached in pigs (70 kg) after inhalation of 250 ppm CO for 2 hours (27). This has been partly attributed to the lower respiratory minute volume per unit of body weight in large animals (26). Fourth, apart from species differences in hemoglobin affinity and HbCO half-lives, there are major differences in basic physiology between mice and humans that may explain our results in contrast to the findings in the rodent studies. In particular, mice have heart rates up to 325eC780 beats per minute and respiratory rates of 94eC163 breaths per minute (Louisiana Veterinary Medical Association, http://www.lvma.org/mouse.html), and are less sensitive to LPS (28). Therefore, saturation is much faster in mice than in humans and dissociation of CO is also approximately 10 times higher. This likely results in higher partial CO gas pressures in blood and tissues of mice compared with humans. One may argue that we should have used higher CO concentrations in humans. Perhaps lengthening the CO pretreatment time to account for these saturation/dissociation differences might allow for a more significant effect on LPS-induced cytokine production. At present it is difficult to assess the levels of CO present in tissues and cells after inhalation. If this were possible, a more direct comparison between species could be performed.
Although the evidence is scanty, many believe that it is the CO released from hemoglobin that raises CO levels in cells. One or more of the factors listed above may have been responsible for the decreased delivery of CO to cells other than red blood cells in our trial. We also note that we reached maximal HbCO levels of about 7%. Unfortunately, we were not able to compare these with maximal murine HbCO levels, because they were not assessed or reported (5).
In this initial study, we focused strongly on safety, using criteria based on the few human studies examining the effects of continuous CO inhalation on HbCO levels. Continuous CO inhalation has been performed in humans with similar HbCO cutoff limits. For example, volunteers breathed CO concentrations of 400 to 1,000 ppm until their HbCO levels reached 10 to 12% (29) and were then assigned to hyperbaric oxygen therapy, whereas our volunteers did not receive hyperbaric therapy. We reached HbCO levels of 7%, which are similar to the 8.3% (95% CI, 7.3 to 9.3%) HbCO levels reached in another continuous CO inhalation study using 1,000 ppm for 30 min and 100 ppm for 30 min (30). It could be that higher CO concentrations might be tolerated in healthy volunteers. However, HbCO levels as low as 2eC6% already decrease exercise capacity in patients with cardiovascular disease (31) and induce arrhythmias and ischemic ST-segment changes in patients with coronary artery disease (32eC34). In addition, HbCO levels as low as 5% lead to a detectable impairment of cognitive and psychomotor abilities as well as the visual threshold in the absence of subjective symptoms (35eC37). It has been estimated that the consequent reduction in venous or tissue oxygen tension at these HbCO levels is similar to that caused by the reduction of arterial oxygen tension at an altitude of 8,000 to 10,000 ft (38). These cognitive alterations pose more of a safety hazard than a health hazard, as complex psychological functions involving judgments, situational decisions, and responses may be affected (36). However, the data used to draw these conclusions were derived from nonlinear regression and were widely scattered. Because our volunteers were under observation for 8 hours and maximal HbCO levels were below 10%, we deliberately refrained from neurocognitive testing and instead asked all volunteers to report subjective symptoms such as headache, nausea, and vertigo by completing a CO assessment sheet.
Our endotoxin model is a well standardized and established inflammation model in humans. Although endotoxin doses of 2eC4 ng/kg body weight are typically used in this model, the lower dose was used in the current trial. The lower dose yields peak TNF- levels only 4- to 10-fold less than those seen in the mouse studies: peak TNF- levels averaged 162 pg/ml in our trial as compared with 600eC1,600 pg/ml in the mouse endotoxin study. We cannot entirely exclude that our inflammatory stimulus was too strong to allow detection of weak antiinflammatory properties of CO. C-reactive protein values 24 hours after LPS infusion averaged 2.4 mg/dl, a value that is typically seen with common colds.
Although our LPS model is a well established model, variability in individual cytokine response is marked; therefore we had the power to detect only a 50eC80% lower cytokine release in the CO period (17). This was, however, deemed sufficient in view of the fivefold different TNF- release in the mouse model (5). Although our clinical model cannot support an antiinflammatory role for CO in a human systemic inflammation model it should not discourage further investigation of putative antiinflammatory roles for CO in other clinical settings. Thus, the challenge in the clinical development of CO will be to identify the right indication, the right dose, and the right mode of administration for a potentially toxic agent. Our safety data may be useful in this context as a starting point for further dose escalation or alternative routes/methods of administration of CO.
Our results are also limited by the fact that all subjects included in the study were men. Studies have shown that inflammatory cytokine levels vary according to sex and that females have a favorable outcome in sepsis compared with males (39, 40).
In conclusion, inhalation of 500 ppm CO for 1 hour did not significantly modulate systemic cytokine production as 250 ppm did in a murine LPS model. Hence, marked species differences must be considered when developing CO as a potential medicinal product.
REFERENCES
Von Burg R. Carbon monoxide. J Appl Toxicol 1999;19:379eC386.
Weaver LK. Carbon monoxide poisoning [review]. Crit Care Clin 1999;15:297eC317, viii.
Coburn RF, Blakemore WS, Forster RE. Endogenous carbon monoxide production in man. J Clin Invest 1963;42:1172eC1178.
Marilena G. New physiological importance of two classic residual products: carbon monoxide and bilirubin. Biochem Mol Med 1997;61:136eC142.
Otterbein LE, Bach FH, Alam J, Soares M, Tao Lu H, Wysk M, Davis RJ, Flavell RA, Choi AM. Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway. Nat Med 2000;6:422eC428.
Moore BA, Otterbein LE, Turler A, Choi AM, Bauer AJ. Inhaled carbon monoxide suppresses the development of postoperative ileus in the murine small intestine. Gastroenterology 2003;124:377eC391.
Otterbein LE, Otterbein SL, Ifedigbo E, Liu F, Morse DE, Fearns C, Ulevitch RJ, Knickelbein R, Flavell RA, Choi AM. MKK3 mitogen-activated protein kinase pathway mediates carbon monoxide-induced protection against oxidant-induced lung injury. Am J Pathol 2003;163:2555eC2563.
Sato K, Balla J, Otterbein L, Smith RN, Brouard S, Lin Y, Csizmadia E, Sevigny J, Robson SC, Vercellotti G, et al. Carbon monoxide generated by heme oxygenase-1 suppresses the rejection of mouse-to-rat cardiac transplants. J Immunol 2001;166:4185eC4194.
Lavitrano M, Smolenski RT, Musumeci A, Maccherini M, Slominska E, Di Florio E, Bracco A, Mancini A, Stassi G, Patti M, et al. Carbon monoxide improves cardiac energetics and safeguards the heart during reperfusion after cardiopulmonary bypass in pigs. FASEB J 2004;18:1093eC1095.
Suffredini AF, Hochstein HD, McMahon FG. Dose-related inflammatory effects of intravenous endotoxin in humans: evaluation of a new clinical lot of Escherichia coli O:113 endotoxin. J Infect Dis 1999;179:1278eC1282.
Marshall MD, Kales SN, Christiani DC, Goldman RH. Are reference intervals for carboxyhemoglobin appropriate A survey of Boston area laboratories. Clin Chem 1995;41:1434eC1438.
Jilma B, Marsik C, Mayr F, Graninger MT, Taylor FB Jr, Ribel MC, Erhardtsen E, Handler S, Eichler HG. Pharmacodynamics of active site-inhibited factor VIIa in endotoxin-induced coagulation in humans. Clin Pharmacol Ther 2002;72:403eC410.
Derhaschnig U, Reiter R, Knobl P, Baumgartner M, Keen P, Jilma B. Recombinant human activated protein C (rhAPC; drotrecogin alfa [activated]) has minimal effect on markers of coagulation, fibrinolysis, and inflammation in acute human endotoxemia. Blood 2003;102:2093eC2098.
Marsik C, Halama T, Cardona F, Wlassits W, Mayr F, Pleiner J, Jilma B. Regulation of Fas (APO-1, CD95) and Fas ligand expression in leukocytes during systemic inflammation in humans. Shock 2003;20:493eC496.
Rainen L, Oelmueller U, Jurgensen S, Wyrich R, Ballas C, Schram J, Herdman C, Bankaitis-Davis D, Nicholls N, Trollinger D, et al. Stabilization of mRNA expression in whole blood samples. Clin Chem 2002;48:1883eC1890.
Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2eCCT method. Methods 2001;25:402eC408.
Stolley PD, Strom BL. Sample size calculations for clinical pharmacology studies. Clin Pharmacol Ther 1986;39:489eC490.
van Deventer SJ, Buller HR, ten Cate JW, Aarden LA, Hack CE, Sturk A. Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 1990;76:2520eC2526.
van den Blink B, Branger J, Weijer S, Gupta A, van Deventer SJ, Peppelenbosch MP, van der Poll T. P38 mitogen activated protein kinase is involved in the downregulation of granulocyte CXC chemokine receptors 1 and 2 during human endotoxemia. J Clin Immunol 2004;24:37eC41.
Huhn RD, Radwanski E, Gallo J, Affrime MB, Sabo R, Gonyo G, Monge A, Cutler DL. Pharmacodynamics of subcutaneous recombinant human interleukin-10 in healthy volunteers. Clin Pharmacol Ther 1997;62:171eC180.
Pernerstorfer T, Hollenstein U, Hansen JB, Stohlawetz P, Eichler HG, Handler S, Speiser W, Jilma B. Lepirudin blunts endotoxin-induced coagulation activation. Blood 2000;95:1729eC1734.
Brittain T, Sutherland J, Greenwood C. A study of the kinetics of the reaction of ligands with the liganded states of mouse embryonic haemoglobins. Biochem J 1986;234:151eC155.
Engel RR, Rodkey FL, O'Neal JD, Collison HA. Relative affinity of human fetal hemoglobin for carbon monoxide and oxygen. Blood 1969;33:37eC45.
Vreman HJ, Mahoney JJ, Stevenson DK. Carbon monoxide and carboxyhemoglobin. Adv Pediatr 1995;42:303eC334.
Watson ES, Jones AB, Ashfaq MK, Barrett JT. Spectrophotometric evaluation of carboxyhemoglobin in blood of mice after exposure to marijuana or tobacco smoke in a modified Walton horizontal smoke exposure machine. J Anal Toxicol 1987;11:19eC23.
Klimisch HJ, Chevalier HJ, Harke HP, Dontenwill W. Uptake of carbon monoxide in blood of miniture pigs and other mammals. Toxicology 1975;3:301eC310.
Lavitrano M, Smolenski RT, Musumeci A, Maccherini M, Slominska E, Florio E, Bracco A, Mancini A, Stassi G, Patti M, et al. Carbon monoxide improves cardiac energetics and safeguards the heart during reperfusion after cardiopulmonary bypass in pigs. FASEB J 2004;18:1093eC1095.
Redl H, Bahrami S, Schlag G, Traber DL. Clinical detection of LPS and animal models of endotoxemia. Immunobiology 1993;187:330eC345.
Takeuchi A, Vesely A, Rucker J, Sommer LZ, Tesler J, Lavine E, Slutsky AS, Maleck WH, Volgyesi G, Fedorko L, et al. A simple "new" method to accelerate clearance of carbon monoxide. Am J Respir Crit Care Med 2000;161:1816eC1819.
Hausberg M, Somers VK. Neural circulatory responses to carbon monoxide in healthy humans. Hypertension 1997;29:1114eC1118.
Allred EN, Bleecker ER, Chaitman BR, Dahms TE, Gottlieb SO, Hackney JD, Pagano M, Selvester RH, Walden SM, Warren J. Short-term effects of carbon monoxide exposure on the exercise performance of subjects with coronary artery disease. N Engl J Med 1989;321:1426eC1432.
Allred EN, Bleecker ER, Chaitman BR, Dahms TE, Gottlieb SO, Hackney JD, Pagano M, Selvester RH, Walden SM, Warren J. Effects of carbon monoxide on myocardial ischemia. Environ Health Perspect 1991;91:89eC132.
Mall T, Grossenbacher M, Perruchoud AP, Ritz R. Influence of moderately elevated levels of carboxyhemoglobin on the course of acute ischemic heart disease. Respiration (Herrlisheim) 1985;48:237eC244.
Sheps DS, Herbst MC, Hinderliter AL, Adams KF, Ekelund LG, O'Neil JJ, Goldstein GM, Bromberg PA, Ballenger M, Davis SM, et al. Effects of 4 percent and 6 percent carboxyhemoglobin on arrhythmia production in patients with coronary artery disease. Res Rep Health Eff Inst 1991;41:1eC46; discussion 47eC58.
Somers VK, Mark AL, Zavala DC, Abboud FM. Contrasting effects of hypoxia and hypercapnia on ventilation and sympathetic activity in humans. J Appl Physiol 1989;67:2101eC2106.
Schulte JH. Effects of mild carbon monoxide intoxication. Arch Environ Health 1963;38:524eC530.
McFarland RA, Roughton FJW, Halperin MH, Niven JJ. The effects of carbon monoxide and altitude on visual thresholds. J Aviat Med 1944;15:381eC394.
Turino GM. Effect of carbon monoxide on the cardiorespiratory system. Carbon monoxide toxicity: physiology and biochemistry. Circulation 1981;63:253AeC259A.
Schroder J, Kahlke V, Staubach KH, Zabel P, Stuber F. Gender differences in human sepsis. Arch Surg 1998;133:1200eC1205.
Stork S, Baumann K, von Schacky C, Angerer P. The effect of 17-estradiol on MCP-1 serum levels in postmenopausal women. Cardiovasc Res 2002;53:642eC649.(Florian B. Mayr, Alexande)
Anesthesiology and General Critical Care Medicine, Medical University of Vienna, Vienna, Austria
ABSTRACT
Data show that carbon monoxide (CO) exerts direct antiinflammatory effects in vitro and in vivo after LPS challenge in a mouse model. We hypothesized that CO may act as an antiinflammatory agent in human endotoxemia. The aim of this trial was to study the effects of CO inhalation on cytokine production during experimental human endotoxemia. The main study was a randomized, double-blinded, placebo-controlled, two-way cross-over trial in healthy volunteers. Each volunteer inhaled synthetic air (as placebo) and 500 ppm CO for 1 hour in random order with a washout period of 6 weeks and received a 2-ng/kg intravenous bolus of LPS after inhalation. Carboxyhemoglobin levels were assessed as a safety parameter. CO inhalation increased carboxyhemoglobin levels from 1.2% (95% confidence interval, 1.0 to 1.4%) to peak values of 7.0% (95% confidence interval, 6.5 to 7.7%). LPS infusion transiently increased plasma concentrations of tumor necrosis factor-, interleukin (IL)-6 (approximately 150-fold increases), and IL-8, as well as IL-1 and IL-1 mRNA levels (an approximately 200-fold increase). These LPS-induced changes were not influenced by CO inhalation. Inhalation of 500 ppm CO for 1 hour had no antiinflammatory effects in a systemic inflammation model in humans, as 250 ppm for 1 hour did in rodents.
Key Words: carboxyhemoglobin endotoxin tumor necrosis factor-
Carbon monoxide (CO) originates from the oxidation or combustion of organic matter (1). At high concentrations, exposure to CO is lethal, usually through accidental smoke or exhaust inhalation in enclosed spaces (2).
Cigarette smoke accounts for a major source of CO exposure in humans. In addition to uptake of exogenous gas, cells and tissues produce significant amounts of CO (3, 4). These two "kinds" of CO are different; one is produced locally, at the cellular level, as a component of heme degradation, and the other is incorporated systemically during exposure to various concentrations. Data show that CO exerts direct antiinflammatory effects after LPS challenge in vitro and in an in vivo mouse model (5). The key finding with the mouse model was that mice exposed to 250 ppm CO for 1 hour before LPS administration responded with significantly lower levels of proinflammatory cytokines (tumor necrosis factor- [TNF-] and interleukin [IL]-1) and higher levels of IL-10 than control mice. As a consequence, the role of CO in various rodent models was intensively studied and yielded encouraging results (6eC8). The latest results show that CO has a pharmacodynamic action in pigs undergoing cardiopulmonary bypass (9), extending our knowledge to larger animals. However, CO inhalation was combined with constant perfusion of the pigs' hearts with a CO-saturated cardioplegic solution, which obscures the relative contribution of CO inhalation to the beneficial effects.
On the basis of the rationale provided by these animal studies, we hypothesized that CO may act as an antiinflammatory agent in human endotoxemia as well, decreasing the release of proinflammatory cytokines such as TNF- in vivo. We therefore studied the effects of CO inhalation on systemic inflammation during experimental human endotoxemia.
The infusion of low doses of endotoxin into humans provides a well-standardized model to study the pharmacodynamics of drugs with supposed antiinflammatory properties in vivo (10). We used this model to characterize putative pharmacodynamic effects of CO in humans.
METHODS
This trial and subsequent amendments were approved by the Medical University of Vienna Ethics Committee, and written informed consent was obtained from all participants before enrollment.
Pilot Trial
Effects of CO inhalation on systemic inflammation have been investigated only in animals so far. Thus, we decided to conduct a pilot trial with nine healthy male nonsmokers, aged 19 to 40 years, to document the safety and tolerability of CO inhalation in humans. The pilot trial was designed as a randomized, placebo-controlled, four-way cross-over dose-escalation study. All volunteers inhaled placebo (synthetic air) and CO concentrations of 10, 50, and 250 parts per million (ppm; 250 ppm had a marked antiinflammatory effect in mice) for a time period of 1 hour randomly assigned to periods 1, 2, 3, and 4 with a washout phase of 1 week between periods. Having referred to the literature on maximal carboxyhemoglobin (HbCO) levels in smokers, we assumed that inhalation of CO, toxic at higher levels, would be innocuous as long as HbCO levels remained at 10% or less (11). All volunteers were asked to complete a "CO assessment sheet" before and immediately after inhalation, at which time we asked about the existence and intensity (none < mild < moderate < severe) of subjective symptoms such as headache, nausea, and vertigo. Because HbCO levels did not reach our predetermined objective of 10%, we invited the same nine volunteers (or adequate substitutes) to inhale further CO concentrations of 250 ppm for 2 hours and 500 ppm for 1 hour in a randomized, two-way cross-over design with a 1-week washout period in between.
Blood samples were collected before and directly after inhalation. Immediately thereafter, LPS (Escherichia coli lipopolysaccharide, 50 pg/ml; Sigma, Vienna, Austria) was added to the blood tubes, which were then incubated in a water bath at 37°C for 2 hours. After incubation, plasma was obtained by immediate centrifugation at 2,000 x g (15 minutes at 4°C) and stored in 0.5-ml aliquots at eC80°C until batch analysis of TNF-, IL-1 and IL-1, IL-6, and IL-8. IL-1 and IL-1 mRNAs were also isolated after 2 hours of incubation.
Main Trial
The main trial was conducted as a randomized, double-blind, placebo-controlled, two-way cross-over study at the Department of Clinical Pharmacology (Medical University of Vienna, Vienna, Austria). Each carbon monoxideeCnaive volunteer inhaled synthetic air (as placebo) and 500 ppm carbon monoxide in random order with a washout period of 6 weeks in between both treatments. To achieve similar TNF- peak levels compared with the study in mice (5), we chose an LPS dose of 2 ng/kg.
Study Subjects
Thirteen healthy male nonsmokers aged 18 to 40 years were invited to participate in this trial. Medical screening included medical history, physical examination, laboratory parameters, and virologic and standard drug screening. In addition, study subjects were tested for hereditary thrombophilia, that is, factor V Leiden and protein C and S deficiency, to minimize potential risks of endotoxin-induced coagulation activation as previously described (12). Exclusion criteria included regular or recent intake of medication including over-the-counter drugs, and clinically relevant abnormal findings in medical history or laboratory parameters.
Study Protocol
The experimental procedures of our endotoxin model have been described in detail in other trials (12, 13). In short, volunteers were admitted to our department at 8:00 A.M. after an overnight fast that was extended until 4 hours after LPS infusion. After a resting period of 15 minutes, basal blood samples were drawn from an indwelling catheter and vital signs (blood pressure, pulse rate, and oxygen saturation) were measured. Shortly thereafter, volunteers started to inhale 500 ppm carbon monoxide (Messer Austria GmbH, Gumpoldskirchen, Austria) or synthetic air (80% N, 20% O2; Messer Austria) as placebo for 1 hour while sitting in an upright position. For this purpose we used cushioned full face masks (Mirage NV full face mask Series II; Pulmomed, Vienna, Austria), which were fixed with Velcro straps. During inhalation venous blood samples were drawn every 15 minutes, by a physician not otherwise involved in the trial, to monitor HbCO levels. For safety reasons, additional HbCO measurements were performed 1.5 and 4 hours after inhalation. All samples were analyzed with a fully automatic spectral photometric measuring system (AVL 912 CO-Oxylite; AVL List, Graz, Austria) and cross-checked by staff of the emergency laboratory (ABL 700; Radiometer, Copenhagen, Denmark). An HbCO level greater than 10% was defined as a stop criterion. Immediately after inhalation volunteers received a bolus of national reference endotoxin (E. coli LPS, 2 ng/kg; U.S. Pharmacopeia, Rockville, MD). In addition, isotonic saline solution (administered at 200 ml/hour; Leopold Pharma, Vienna, Austria) was administered to all subjects within 5 minutes of LPS infusion and was continued over 6 hours to maintain adequate hydration.
Blood Sampling
Blood samples were collected by venipuncture into citrated or ethylenediaminetetraacetic acideCanticoagulated tubes (Vacutainer; BD Biosciences, Vienna, Austria) before inhalation, directly after inhalation and 1.5, 2, 4, 6, 8, and 24 hours after LPS administration. Plasma was obtained by immediate centrifugation at 2,000 x g (15 minutes at 4°C) and stored in 0.5-ml aliquots at eC80°C until batch analysis.
Cytokine Analysis
Plasma levels of TNF-, IL-6, and IL-10 were measured by (high) sensitivity enzyme immunoassays (12). Plasma IL-8 levels were determined with a highly sensitive chemoluminescence enzyme assay (detection limit, 0.8 pg/ml); all assays were purchased from R&D Systems (Abingdon, UK). Prothrombin fragment (F1+2) was used as a marker of in vivo thrombin generation, which was also measured by an enzyme immunoassay (Behring, Marburg, Germany).
Real-time Quantitative Reverse Transcriptase-Polymerase Chain Reaction
Because plasma levels of IL-1 are hardly detectable in humans even under endotoxemia (14) we decided to quantify IL-1 mRNA.
All blood samples were immediately processed to avoid storage-induced changes in mRNA levels (15). After isolating total RNA with an RNeasy minikit (Qiagen, Valencia, CA) according to the manufacturer's instructions, mRNA was directly transcribed into cDNA, using a TaqMan reverse transcription reagents kit (Applied Biosystems, Foster City, CA), and stored at eC80°C until analysis. IL-1 mRNA quantification was performed with the ABI PRISM 7700 (Applied Biosystems), using commercially available predesigned primers and probes (Applied Biosystems). We chose 18S as housekeeping gene for multiplexing (Applied Biosystems) because of its stable expression under endotoxemia (B. Jilma, unpublished data). IL-1 mRNA was normalized against the reference gene (18S) according to the 2eCCT method (16), and data are expressed as fold increase over baseline values.
Data Analysis
Data are expressed as means and 95% confidence interval (95% CI). All statistical comparisons of continuous variables were made with nonparametric tests. After repeated measures analysis of variance with treatment and period as independent variables, the Friedman analysis of variance and the Wilcoxon test for post hoc comparisons were used to assess time-dependent changes in outcome variables within groups. A two-tailed p value less than 0.05 was considered significant. All statistical calculations were performed with commercially available statistical software (Statistica version 6.1; StatSoft, Tulsa, OK). Power calculation was done as previously described (17).
RESULTS
Pilot Trial
CO inhalation was well tolerated by our nonsmoking volunteers, whose mean age was 25 years (range, 20 to 40 years), and whose body mass index (BMI) averaged 23.8 kg/m2 (range, 21.5 to 26.3 kg/m2). One volunteer reported a mild headache after inhalation of 250 ppm CO for 2 hours. Seven volunteers were not able to participate in the second part of the pilot trial because of individual time-scheduling problems. They were replaced by adequate substitutes. CO inhalation did not have any significant effect on vital parameters, leukocyte and neutrophil counts, as well as cytokine levels. Because inhalation of 250 ppm CO for 2 hours yielded roughly the same peak HbCO levels as 500 ppm CO for 1 hour, we chose a CO concentration of 500 ppm over 1 hour for the main trial. The results of the pilot trial are summarized in Table 1.
Main Trial
Demographic data.
All participating subjects were healthy male nonsmokers. Their mean age was 29 years (range, 18 to 38 years), and their BMI averaged 23.0 kg/m2 (range, 19.0 to 26.8 kg/m2).
Effects of CO on HbCO levels.
CO inhalation was well tolerated by our nonsmoking volunteers and no adverse events, as assessed by asking about subjective symptoms and monitoring vital signs, occurred. Baseline HbCO values of 1.2% (95% CI, 1.0 to 1.4%) increased five- to sixfold time dependently to peak values of 7.0% (95% CI, 6.5 to 7.7%) after 1 hour of inhalation of 500 ppm CO. As expected, inhalation of synthetic air did not alter basal HbCO levels significantly (Figure 1).
Vital parameters.
LPS infusion induced a transient increase in heart rate and a transient decrease in mean arterial blood pressure (all p < 0.05 versus time), as previously reported (18). CO inhalation did not have any influence on these LPS-induced transient changes (Figure 2). LPS infusion also induced a transient increase in body temperature that reached its maximum after 4 hours and was not altered by CO inhalation (Figure 2D).
Effects of CO on inflammatory cells and C-reactive protein.
As expected, neutrophil counts increased about threefold and monocyte counts sharply dropped compared with baseline levels (all p < 0.05 versus time) after LPS infusion. CO inhalation did not have any influence on these transient changes (Figure 3). Furthermore, C-reactive protein plasma levels increased (2.4 mg/dl; 95% CI, 2.1 to 2.7 mg/dl) 24 hours after LPS infusion in comparison with baseline values of less than 0.5 mg/dl. Again, the magnitude of C-reactive protein increase was not influenced by CO inhalation (data not shown).
Effects of CO on proinflammatory cytokines.
LPS infusion induced transient increases in plasma concentrations of TNF-, IL-6, IL-8, as well as IL-1 and IL-1 mRNA levels (all p < 0.05 versus time), which returned to baseline values after 24 hours as shown in previous trials (13, 19).
In subjects treated with placebo before LPS injection, plasma TNF- levels peaked 2 hours after LPS infusion (126 pg/ml; 95% CI, 98 to 155 pg/ml), whereas CO-treated subjects showed peak TNF- levels slightly earlier, 1.5 hours after LPS infusion (163 pg/ml; 95% CI, 68 to 258 pg/ml) (Figure 4A).
Peak IL-6 values were seen 2 hours after LPS infusion, with no significant differences between periods (267 pg/ml [95% CI, 173 to 361 pg/ml] in the placebo period versus 282 pg/ml [95% CI, 160 to 404 pg/ml] in the CO period) (Figure 4B).
IL-1 and IL-1 mRNA levels showed a distinct increase over baseline values 2 hours after LPS infusion: IL-1, a 170-fold increase (95% CI, 45- to 294-fold increase) in the placebo period versus a 202-fold increase (95% CI, 49- to 355-fold increase) in the CO period (Figure 5A); IL-1, a 199-fold increase (95% CI, 38- to 360-fold increase) in the placebo period versus a 134-fold increase (95% CI, 37- to 232-fold increase) in the CO period (Figure 5B).
Similar to IL-6, peak IL-8 values were detectable 2 hours after LPS infusion, with no significant differences between periods (202 pg/ml [95% CI, 103 to 301 pg/ml] in the placebo period versus 212 pg/ml [95% CI, 88 to 335 pg/ml] in the CO period), although IL-8 levels were slightly lower in the CO period at 4 hours (Figure 4C).
Effects of CO on IL-10.
IL-10 is an inhibitor of proinflammatory cytokine synthesis and as such can limit the inflammatory process, including the production of TNF- (20). CO inhalation, however, did not significantly influence IL-10 plasma levels after LPS infusion. Peak values were reached 2 hours after LPS injection and were slightly higher in CO-treated subjects (28 pg/ml; 95% CI, 18 to 37 pg/ml) compared with those treated with placebo (25 pg/ml; 95% CI, 15 to 34 pg/ml) (Figure 4D).
Effects of CO on coagulation.
As expected, LPS injection was associated with a potent activation of the coagulation system (21) as reflected by increases in the plasma concentrations of prothrombin fragment F1+2, peaking after 4 hours [4.0 nmol/L; 95% CI, 3.2 to 4.8 nmol/L) (p < 0.05 versus time). This LPS-induced coagulant response was not significantly influenced by CO inhalation and showed the same time course in both groups. Peak levels in the CO period were 3.6 nmol/L (95% CI, 2.8 to 4.3 nmol/L) (Figure 3C).
DISCUSSION
Rodent models provide evidence that inhalation of relatively low concentrations of carbon monoxide have antiinflammatory effects (5eC7). In particular, inhalation of CO at 250 ppm for 1 hour significantly reduced TNF- and IL-1 in an experimental mouse endotoxemia model (5). This provided a sound rationale to investigate these putative antiinflammatory effects in humans.
Under the conditions that we used, we did not find a similar effect (5). We failed to show direct antiinflammatory effects of 500 ppm CO inhaled for 1 hour in experimental human endotoxemia, as measured by cytokine production. This discrepancy may have several explanations, some of which are likely interrelated.
First, blood of different species differs markedly in its affinity for CO. The affinity constants of mouse hemoglobin for CO and O2 differ by only 50-fold (22) as compared with a 217-fold higher affinity for CO than O2 shown by human hemoglobin (23). Second, the half-lives of HbCO in mice (30 minutes) and humans (5 hours) differ by 10-fold (24, 25). This demonstrates that CO dissociation from mouse hemoglobin is much faster. Third, saturation of hemoglobin differs markedly between species. For example, small animals such as hamsters, which are comparable in size to mice, have HbCO values exceeding 20% when breathing 250 ppm CO for 1 hour (26). In contrast, larger animals such as miniature pigs show HbCO values less than 10% after 1 hour (26), and HbCO values of about 12% were reached in pigs (70 kg) after inhalation of 250 ppm CO for 2 hours (27). This has been partly attributed to the lower respiratory minute volume per unit of body weight in large animals (26). Fourth, apart from species differences in hemoglobin affinity and HbCO half-lives, there are major differences in basic physiology between mice and humans that may explain our results in contrast to the findings in the rodent studies. In particular, mice have heart rates up to 325eC780 beats per minute and respiratory rates of 94eC163 breaths per minute (Louisiana Veterinary Medical Association, http://www.lvma.org/mouse.html), and are less sensitive to LPS (28). Therefore, saturation is much faster in mice than in humans and dissociation of CO is also approximately 10 times higher. This likely results in higher partial CO gas pressures in blood and tissues of mice compared with humans. One may argue that we should have used higher CO concentrations in humans. Perhaps lengthening the CO pretreatment time to account for these saturation/dissociation differences might allow for a more significant effect on LPS-induced cytokine production. At present it is difficult to assess the levels of CO present in tissues and cells after inhalation. If this were possible, a more direct comparison between species could be performed.
Although the evidence is scanty, many believe that it is the CO released from hemoglobin that raises CO levels in cells. One or more of the factors listed above may have been responsible for the decreased delivery of CO to cells other than red blood cells in our trial. We also note that we reached maximal HbCO levels of about 7%. Unfortunately, we were not able to compare these with maximal murine HbCO levels, because they were not assessed or reported (5).
In this initial study, we focused strongly on safety, using criteria based on the few human studies examining the effects of continuous CO inhalation on HbCO levels. Continuous CO inhalation has been performed in humans with similar HbCO cutoff limits. For example, volunteers breathed CO concentrations of 400 to 1,000 ppm until their HbCO levels reached 10 to 12% (29) and were then assigned to hyperbaric oxygen therapy, whereas our volunteers did not receive hyperbaric therapy. We reached HbCO levels of 7%, which are similar to the 8.3% (95% CI, 7.3 to 9.3%) HbCO levels reached in another continuous CO inhalation study using 1,000 ppm for 30 min and 100 ppm for 30 min (30). It could be that higher CO concentrations might be tolerated in healthy volunteers. However, HbCO levels as low as 2eC6% already decrease exercise capacity in patients with cardiovascular disease (31) and induce arrhythmias and ischemic ST-segment changes in patients with coronary artery disease (32eC34). In addition, HbCO levels as low as 5% lead to a detectable impairment of cognitive and psychomotor abilities as well as the visual threshold in the absence of subjective symptoms (35eC37). It has been estimated that the consequent reduction in venous or tissue oxygen tension at these HbCO levels is similar to that caused by the reduction of arterial oxygen tension at an altitude of 8,000 to 10,000 ft (38). These cognitive alterations pose more of a safety hazard than a health hazard, as complex psychological functions involving judgments, situational decisions, and responses may be affected (36). However, the data used to draw these conclusions were derived from nonlinear regression and were widely scattered. Because our volunteers were under observation for 8 hours and maximal HbCO levels were below 10%, we deliberately refrained from neurocognitive testing and instead asked all volunteers to report subjective symptoms such as headache, nausea, and vertigo by completing a CO assessment sheet.
Our endotoxin model is a well standardized and established inflammation model in humans. Although endotoxin doses of 2eC4 ng/kg body weight are typically used in this model, the lower dose was used in the current trial. The lower dose yields peak TNF- levels only 4- to 10-fold less than those seen in the mouse studies: peak TNF- levels averaged 162 pg/ml in our trial as compared with 600eC1,600 pg/ml in the mouse endotoxin study. We cannot entirely exclude that our inflammatory stimulus was too strong to allow detection of weak antiinflammatory properties of CO. C-reactive protein values 24 hours after LPS infusion averaged 2.4 mg/dl, a value that is typically seen with common colds.
Although our LPS model is a well established model, variability in individual cytokine response is marked; therefore we had the power to detect only a 50eC80% lower cytokine release in the CO period (17). This was, however, deemed sufficient in view of the fivefold different TNF- release in the mouse model (5). Although our clinical model cannot support an antiinflammatory role for CO in a human systemic inflammation model it should not discourage further investigation of putative antiinflammatory roles for CO in other clinical settings. Thus, the challenge in the clinical development of CO will be to identify the right indication, the right dose, and the right mode of administration for a potentially toxic agent. Our safety data may be useful in this context as a starting point for further dose escalation or alternative routes/methods of administration of CO.
Our results are also limited by the fact that all subjects included in the study were men. Studies have shown that inflammatory cytokine levels vary according to sex and that females have a favorable outcome in sepsis compared with males (39, 40).
In conclusion, inhalation of 500 ppm CO for 1 hour did not significantly modulate systemic cytokine production as 250 ppm did in a murine LPS model. Hence, marked species differences must be considered when developing CO as a potential medicinal product.
REFERENCES
Von Burg R. Carbon monoxide. J Appl Toxicol 1999;19:379eC386.
Weaver LK. Carbon monoxide poisoning [review]. Crit Care Clin 1999;15:297eC317, viii.
Coburn RF, Blakemore WS, Forster RE. Endogenous carbon monoxide production in man. J Clin Invest 1963;42:1172eC1178.
Marilena G. New physiological importance of two classic residual products: carbon monoxide and bilirubin. Biochem Mol Med 1997;61:136eC142.
Otterbein LE, Bach FH, Alam J, Soares M, Tao Lu H, Wysk M, Davis RJ, Flavell RA, Choi AM. Carbon monoxide has anti-inflammatory effects involving the mitogen-activated protein kinase pathway. Nat Med 2000;6:422eC428.
Moore BA, Otterbein LE, Turler A, Choi AM, Bauer AJ. Inhaled carbon monoxide suppresses the development of postoperative ileus in the murine small intestine. Gastroenterology 2003;124:377eC391.
Otterbein LE, Otterbein SL, Ifedigbo E, Liu F, Morse DE, Fearns C, Ulevitch RJ, Knickelbein R, Flavell RA, Choi AM. MKK3 mitogen-activated protein kinase pathway mediates carbon monoxide-induced protection against oxidant-induced lung injury. Am J Pathol 2003;163:2555eC2563.
Sato K, Balla J, Otterbein L, Smith RN, Brouard S, Lin Y, Csizmadia E, Sevigny J, Robson SC, Vercellotti G, et al. Carbon monoxide generated by heme oxygenase-1 suppresses the rejection of mouse-to-rat cardiac transplants. J Immunol 2001;166:4185eC4194.
Lavitrano M, Smolenski RT, Musumeci A, Maccherini M, Slominska E, Di Florio E, Bracco A, Mancini A, Stassi G, Patti M, et al. Carbon monoxide improves cardiac energetics and safeguards the heart during reperfusion after cardiopulmonary bypass in pigs. FASEB J 2004;18:1093eC1095.
Suffredini AF, Hochstein HD, McMahon FG. Dose-related inflammatory effects of intravenous endotoxin in humans: evaluation of a new clinical lot of Escherichia coli O:113 endotoxin. J Infect Dis 1999;179:1278eC1282.
Marshall MD, Kales SN, Christiani DC, Goldman RH. Are reference intervals for carboxyhemoglobin appropriate A survey of Boston area laboratories. Clin Chem 1995;41:1434eC1438.
Jilma B, Marsik C, Mayr F, Graninger MT, Taylor FB Jr, Ribel MC, Erhardtsen E, Handler S, Eichler HG. Pharmacodynamics of active site-inhibited factor VIIa in endotoxin-induced coagulation in humans. Clin Pharmacol Ther 2002;72:403eC410.
Derhaschnig U, Reiter R, Knobl P, Baumgartner M, Keen P, Jilma B. Recombinant human activated protein C (rhAPC; drotrecogin alfa [activated]) has minimal effect on markers of coagulation, fibrinolysis, and inflammation in acute human endotoxemia. Blood 2003;102:2093eC2098.
Marsik C, Halama T, Cardona F, Wlassits W, Mayr F, Pleiner J, Jilma B. Regulation of Fas (APO-1, CD95) and Fas ligand expression in leukocytes during systemic inflammation in humans. Shock 2003;20:493eC496.
Rainen L, Oelmueller U, Jurgensen S, Wyrich R, Ballas C, Schram J, Herdman C, Bankaitis-Davis D, Nicholls N, Trollinger D, et al. Stabilization of mRNA expression in whole blood samples. Clin Chem 2002;48:1883eC1890.
Livak KJ, Schmittgen TD. Analysis of relative gene expression data using real-time quantitative PCR and the 2eCCT method. Methods 2001;25:402eC408.
Stolley PD, Strom BL. Sample size calculations for clinical pharmacology studies. Clin Pharmacol Ther 1986;39:489eC490.
van Deventer SJ, Buller HR, ten Cate JW, Aarden LA, Hack CE, Sturk A. Experimental endotoxemia in humans: analysis of cytokine release and coagulation, fibrinolytic, and complement pathways. Blood 1990;76:2520eC2526.
van den Blink B, Branger J, Weijer S, Gupta A, van Deventer SJ, Peppelenbosch MP, van der Poll T. P38 mitogen activated protein kinase is involved in the downregulation of granulocyte CXC chemokine receptors 1 and 2 during human endotoxemia. J Clin Immunol 2004;24:37eC41.
Huhn RD, Radwanski E, Gallo J, Affrime MB, Sabo R, Gonyo G, Monge A, Cutler DL. Pharmacodynamics of subcutaneous recombinant human interleukin-10 in healthy volunteers. Clin Pharmacol Ther 1997;62:171eC180.
Pernerstorfer T, Hollenstein U, Hansen JB, Stohlawetz P, Eichler HG, Handler S, Speiser W, Jilma B. Lepirudin blunts endotoxin-induced coagulation activation. Blood 2000;95:1729eC1734.
Brittain T, Sutherland J, Greenwood C. A study of the kinetics of the reaction of ligands with the liganded states of mouse embryonic haemoglobins. Biochem J 1986;234:151eC155.
Engel RR, Rodkey FL, O'Neal JD, Collison HA. Relative affinity of human fetal hemoglobin for carbon monoxide and oxygen. Blood 1969;33:37eC45.
Vreman HJ, Mahoney JJ, Stevenson DK. Carbon monoxide and carboxyhemoglobin. Adv Pediatr 1995;42:303eC334.
Watson ES, Jones AB, Ashfaq MK, Barrett JT. Spectrophotometric evaluation of carboxyhemoglobin in blood of mice after exposure to marijuana or tobacco smoke in a modified Walton horizontal smoke exposure machine. J Anal Toxicol 1987;11:19eC23.
Klimisch HJ, Chevalier HJ, Harke HP, Dontenwill W. Uptake of carbon monoxide in blood of miniture pigs and other mammals. Toxicology 1975;3:301eC310.
Lavitrano M, Smolenski RT, Musumeci A, Maccherini M, Slominska E, Florio E, Bracco A, Mancini A, Stassi G, Patti M, et al. Carbon monoxide improves cardiac energetics and safeguards the heart during reperfusion after cardiopulmonary bypass in pigs. FASEB J 2004;18:1093eC1095.
Redl H, Bahrami S, Schlag G, Traber DL. Clinical detection of LPS and animal models of endotoxemia. Immunobiology 1993;187:330eC345.
Takeuchi A, Vesely A, Rucker J, Sommer LZ, Tesler J, Lavine E, Slutsky AS, Maleck WH, Volgyesi G, Fedorko L, et al. A simple "new" method to accelerate clearance of carbon monoxide. Am J Respir Crit Care Med 2000;161:1816eC1819.
Hausberg M, Somers VK. Neural circulatory responses to carbon monoxide in healthy humans. Hypertension 1997;29:1114eC1118.
Allred EN, Bleecker ER, Chaitman BR, Dahms TE, Gottlieb SO, Hackney JD, Pagano M, Selvester RH, Walden SM, Warren J. Short-term effects of carbon monoxide exposure on the exercise performance of subjects with coronary artery disease. N Engl J Med 1989;321:1426eC1432.
Allred EN, Bleecker ER, Chaitman BR, Dahms TE, Gottlieb SO, Hackney JD, Pagano M, Selvester RH, Walden SM, Warren J. Effects of carbon monoxide on myocardial ischemia. Environ Health Perspect 1991;91:89eC132.
Mall T, Grossenbacher M, Perruchoud AP, Ritz R. Influence of moderately elevated levels of carboxyhemoglobin on the course of acute ischemic heart disease. Respiration (Herrlisheim) 1985;48:237eC244.
Sheps DS, Herbst MC, Hinderliter AL, Adams KF, Ekelund LG, O'Neil JJ, Goldstein GM, Bromberg PA, Ballenger M, Davis SM, et al. Effects of 4 percent and 6 percent carboxyhemoglobin on arrhythmia production in patients with coronary artery disease. Res Rep Health Eff Inst 1991;41:1eC46; discussion 47eC58.
Somers VK, Mark AL, Zavala DC, Abboud FM. Contrasting effects of hypoxia and hypercapnia on ventilation and sympathetic activity in humans. J Appl Physiol 1989;67:2101eC2106.
Schulte JH. Effects of mild carbon monoxide intoxication. Arch Environ Health 1963;38:524eC530.
McFarland RA, Roughton FJW, Halperin MH, Niven JJ. The effects of carbon monoxide and altitude on visual thresholds. J Aviat Med 1944;15:381eC394.
Turino GM. Effect of carbon monoxide on the cardiorespiratory system. Carbon monoxide toxicity: physiology and biochemistry. Circulation 1981;63:253AeC259A.
Schroder J, Kahlke V, Staubach KH, Zabel P, Stuber F. Gender differences in human sepsis. Arch Surg 1998;133:1200eC1205.
Stork S, Baumann K, von Schacky C, Angerer P. The effect of 17-estradiol on MCP-1 serum levels in postmenopausal women. Cardiovasc Res 2002;53:642eC649.(Florian B. Mayr, Alexande)