Endotoxin, which is a part of the Gram-negative cell wall that is released during growth and lysis of Gram-negative bacteria, has become the best studied trigger of the host inflammatory response that underlies the pathogenesis of sepsis (1). Furthermore, bacterial mutants deficient of endotoxin have been shown to elicit only a partial response in comparison with that caused by wild-type bacteria (2). Therefore, several therapeutic strategies have been developed to reduce blood endotoxin concentration in severe sepsis. In the 1980s and 1990s, antibodies against endotoxin were developed and tested in clinical trials (3-5). Initially, positive results were reported, but in subsequent trials, these effects could not convincingly be confirmed (6, 7). Lacking ability to neutralize endotoxin was proposed to be the cause of this failure (8). Another approach has been the use of extracorporeal apheresis with the binding of endotoxin to different molecules (9, 10). Using these methods, it has been possible to demonstrate reductions in blood endotoxin levels in patients (10). Although positive clinical results with improved mortality were initially reported (11), several multicenter trials could not confirm these results (10, 12), and, surprisingly, not even the inflammatory response, as determined by IL-6 levels, was significantly changed.
The lack of clinical trial evidence of interventions to treat endotoxemia brings the response to endotoxin to a head, necessitating further knowledge of the biological response to endotoxin. Although there is an extensive literature featuring infusion of endotoxin to animals and humans, there are only limited data on the relationship between the magnitude of the endotoxin exposure and the biological response. To gain further insight into this relationship, detailed studies on dose responses are necessary, first in healthy animals and then in animals with different preexposure endotoxin levels and activated anti-inflammatory responses, to be followed by experiments in which the endotoxin load is effectively reduced from varying exposure levels at distinct time points. In a recent study evaluating the dose response in healthy animals, logarithmical increases in the endotoxin dose led to linear changes in blood pressure, hypoperfusion, and organ dysfunction, whereas cytokines such as TNF-α and IL-6 increased logarithmically (13). Mathematically, this implies that a greater change in the response to endotoxin will be achieved if a fixed dose is given to an individual with a low preexposure value than if the same dose would be given to patients in whom increased endotoxin levels already are present. The first case might be represented by a fulminating meningococcal infection and the latter by the increase in endotoxin that is due to absorption from the gut as a consequence of sepsis-induced hypoperfusion and organ dysfunction (14). However, it may be hypothesized that not only the dose but also the rate by which an organism is exposed to a given amount of endotoxin is of importance. Intuitively, this may be the case, but it has not yet been experimentally or clinically demonstrated. Therefore, the present investigation was undertaken to determine whether the administration rate is an important factor for the biological response and whether a fast increase in the endotoxin load elicits a stronger response than a more gradual one of the same total dose. The results of this study would also be of interest for the optimizing of endotoxin challenges in different animal models.
MATERIALS AND METHODS
The study included 18 Swedish "triple breed“ pigs of both sexes weighing 27 ± 0.6 kg (mean ± SEM). All animals were 12 to 14 weeks old and apparently healthy. The pigs were handled according to the guidelines of the Swedish National Board for Laboratory Animals and the European Convention on Animal Care. The experiment was approved by the Animal Ethics Committee of Uppsala University, Sweden. The following inclusion criteria were applied: no apparent preexisting diseases, Pao2 greater than 10 kPa, and a mean pulmonary arterial pressure (MPAP) of 20 mmHg or less at baseline, which was 30 min after the completed preparatory procedure (see below).
Anesthesia and fluid administration
The animals had free access to water ad libitum until they were taken to the laboratory. To decrease the stress level before the induction of anesthesia, a single i.m. injection of 50 mg xylazin (Rompun Vet; Bayer, Leverkusen, Germany) was given to all animals as premedication before transport to the laboratory. General anesthesia was induced by injecting a mixture of 6 mg kg−1 tilétamin-zolazepam (Zoletil forte vet; Virbac Laboratories, Carros, France), 2.2 mg kg−1 xylazin, and 0.04 mg × kg−1 atropine (Atropin;Umeå NM Pharma, Stockholm, Sweden) intramuscularly. Anesthesia was then maintained with sodium pentobarbital (8 mg kg−1 h−1; Pentobarbitalnatrium, Apoteket, Umeå, Sweden), pancuronium bromide (0.26 mg kg−1 h−1; Pavulon, Organon, Oss, The Netherlands), and morphine (0.48 mg kg−1 h−1; Morphine, Pharmacia, Uppsala, Sweden) dissolved in 2.5% glucose solution given as a continuous infusion. Sodium chloride infusion was administered, resulting in a total fluid administration rate of 30 mL kg−1 h−1.
A bolus dose of 20 mg morphine was given intravenously before the tracheotomy procedure to secure a free airway during the experiment. The animals were artificially ventilated throughout the experimental procedure (Servo 900C; Siemens-Elema, Stockholm, Sweden). During surgical stimulation (i.e., catheter insertion), 30% oxygen in N2O was given, after which the gas mixture was set to fraction of inspired oxygen to 0.3 in medical air for the rest of the experiment. The ventilation after preparation was adjusted to yield a Paco2 between 5.0 and 5.5 kPa. Respirator settings were then kept constant throughout the experiment. The respiratory rate was 25 min−1 and the inspiratory-expiratory ratio was 1:3. The pigs were placed into prone position as soon as the preparation was completed, and a positive end-expiratory pressure of 5 cm H2O was maintained to counteract atelectasis development.
A cervical artery was catheterized for pressure monitoring and blood sampling. A central venous line and a 7-F Swan-Ganz catheter, equipped with a thermistor, were inserted through the internal jugular vein into the superior caval vein and into the pulmonary artery, respectively. A minor vesicotomy was performed, and a urinary catheter was introduced into the bladder. A heating pad (Operatherm 200W; KanMed, Bromma, Sweden) was set to 38°C throughout the experiment to decrease heat losses. After the preparation was completed, a 30-min stabilization time passed before baseline values were registered and baseline blood samples taken.
The pigs were randomly allocated to groups I (decelerating dose) or II (accelerating dose). Randomization was performed by the sealed-envelope method using the "randomization-in-block“ principle with a block size of five. Group I received endotoxin infusion (Escherichia coli: 0111:B4; Sigma Chemical, St. Louis, Mo) in doses of 4, 0.5, and 0.063 μg kg−1 h−1 during the first, second, and third hours of the experiment, respectively. Group II received endotoxin infusion in the reverse order, that is, in doses of 0.063, 0.5, and 4 μg × kg−1 × h−1 during the first, second, and third hours of the experiment, respectively. The endotoxin doses were chosen from our previous study in which the effects of continuous infusion of various amounts of endotoxin were investigated (13). In that study, 0.063 μg kg−1 h−1 resulted in a significant response, and the dose-response relationship was strong in the interval of 0.063 to 4 μg kg−1 h−1. These doses are associated with a low mortality rate, and the mortality that does occur is often during the first 2 h because of irreversible pulmonary hypertension. The pig model is considered especially suitable and relevant for endotoxin studies (15, 16). The major drawback, however, is a variable and sometimes dramatic early and transient increase in MPAP, resulting in decreasing cardiac output and arterial pressure, which starts within 30 min. These quickly reversible changes are not correlated to endotoxin doses in the interval used in the present study or to the ensuing inflammatory and circulatory responses (13, 17).
After the first 3 h, endotoxin infusion was terminated in both groups, and the pigs were observed for another 3 h. If the MPAP increased to levels in the magnitude of those of the MAP during the first hour of the experiment, a single dose of 0.2 mg adrenalin (i.v.) was given. Physiologic variables were registered, and blood samples were taken at every hour after baseline for 6 h, after which all surviving pigs were killed by potassium chloride injection (i.v.). This model of porcine endotoxemia has previously been described by our research group (13, 18).
MAP was monitored and registered continuously. The central line was used for measurement of the central venous pressure. Cardiac output was assessed by the thermodilution method using the thermistor in the Swan-Ganz catheter. The average value of at least three serial cardiac output measurements was registered. Heart rate was continuously monitored using electrocardiography. Proximal airway pressure values and respiratory volumes were recorded from ventilator readings. Urine output rate was registered hourly.
Blood samples were taken at baseline and by the hour. Arterial and mixed venous blood gases (Pao2, venous partial pressure of oxygen), oxygen saturation (arterial oxygen saturation, venous oxygen saturation [Svo2]), and base excess (BE) were analyzed (ABL 5 and Hemoximeter; Radiometer, Brønhøj, Denmark). Blood was analyzed on a CELL-DYN 4000 (Abbott Scandinavia AB, Kista, Sweden) for hemoglobin (Hb), leukocytes, and platelets.
Calculations of left ventricular stroke work index (LVSWI), cardiac index (CI), oxygen delivery (DO2) and static pulmonary compliance were derived from their conventional formulae (19, 20).
IL-6 and TNF-α
Commercial sandwich enzyme-linked immunosorbent assay was used for the detection of IL-6 (Quantikine porcine IL-6, P6000; R&D Systems, Minneapolis, Minn) and TNF-α (Porcine TNF-α KSC3011; Biosource International, Nivelles, Belgium). For both IL-6 and TNF-α, the lower detection limit was 10 ng L−1. The assays had intra-assay coefficients of variation of less than 5% and a total coefficients of variation of less than 10%.
Calculations and statistics
To reduce interanimal variation and the number of animals in each group necessary for statistical analysis leukocyte count, platelet count, core temperature, MAP, CI, DO2, SvO2, Hb concentration, and pH, static pulmonary compliance and hourly diuresis were expressed as relative values in relation to that obtained at baseline just before the administration of endotoxin. Changes in these parameters approximated to normal distribution, as did changes in actual BE values. The cytokine concentrations were log-normally distributed, and, therefore, these values were logarithmically transformed. For the two nonsurviving animals, the last values obtained before their death were carried forward.
It has previously been shown by our research group and others that some of the changes induced by continuous infusion of endotoxin in doses used in the present experiment have a tendency to reverse during the last hours of a 6-h experiment (13). It was reasonable to believe that this trend should be more pronounced if the endotoxin infusion was stopped at 3 h. Therefore, a repeated-measures ANOVA during 1 to 6 h was chosen as the primary statistical analysis of the group effect instead of an analysis at 6 h alone. If mean values for a physiologic variable in groups I and II maintained the same intergroup relationship at 1 to 6 h, that is, their graphs did not cross more than once and then were less than 1 SEM away from the mean value of the other and yet no significant group difference was demonstrated at 1 to 6 h, the variable was assessed in a secondary analysis for significance of the difference from 1 h to a time point earlier than 6 h. Repeated-measures ANOVAs were also performed to test changes over time. A P value of less than 0.05 was considered significant. All values were expressed as mean ± SEM unless otherwise stated. The software STATISTICA (Stat. Soft. Inc., Tulsa, Okla) was used in the statistical calculations.
The pigs were comparable on all baseline data (Table 1). Despite an early dose of adrenalin after 20 and 30 min, respectively, two pigs died during the experiment: one in group I after the second hour and one in group II after the first hour of the experiment. Three surviving pigs in group I and two pigs in group II were given a single dose of adrenalin because they fulfilled the criteria for this intervention stated in the protocol at a time point ranging from 20 to 33 min after the start of the endotoxin infusion,. The results from these animals were well within the variation of those from the other animals in the group.
Actual data of laboratory and physiological parameters during the experiment are given in Table 1. According to the statistical plan, tests of significance should be performed on the relative values or changes that are depicted in Figures 1 and 2. The responses in cytokines, core temperature, leukocytes, and platelets are depicted in Figure 1. TNF-α and IL-6 increased in both groups in response to endotoxin infusion. After baseline values, mean TNF-α levels were significantly higher in group I at 1 to 3 h with no change in the opposite direction during the last 3 h of the experiment. Mean IL-6 concentrations in group I exceeded those in group II at 1 to 3 h and vice versa at 4 to 6 h. No difference was detected in the IL-6 concentration at 1 to 6 h, but the time concentration profiles were significantly different.
Mean core temperature was significantly higher in group I than in group II at 1 to 2 h. Not even a trend in the opposite direction was seen at 4 to 6 h.
Leukocytes decreased during the first hours of the experiment and increased thereafter. The levels of white blood cells were significantly lower in group I at 0 to 5 h. Platelets decreased slightly more in group I during the experiment, but the only significant difference between the groups was at 1 to 2 h.
Responses in physiological variables are displayed in Figure 2. MAP values were lower in group I after the start of the endotoxin infusion, a difference that was statistically significant at 1 to 4 h. Cardiac index, DO2, Svo2, and LVSWI all exhibited a biphasic course that decreased initially. No differences in CI and DO2 were noted between the groups. Both Svo2 and LVSWI were significantly lower in group I than in group II at 1 to 4 h. Hemoglobin peaked at 3 h for both groups without any significant difference between the groups.
Base excess and pH, which demonstrate changes in the cellular metabolism secondary to hypoperfusion, decreased from baseline in both groups. In group I, changes in BE were significantly more pronounced at 1 to 4 h, whereas the response in pH was significantly more marked throughout the experiment, probably because of a combined effect of hypoperfusion and pulmonary deterioration.
Organ dysfunction as displayed in a decrease in static pulmonary compliance developed faster and was more marked in group I, with significantly lower values throughout the experiment. Hourly diuresis did not demonstrate significant differences between the groups during the observation period.
The interaction between a mammal and Gram-negative bacteria during sepsis is a complex process. One of the early events in this condition is activation of the inflammatory pathways of the host by endotoxin liberated from the Gram-negative bacteria that may eventually lead to hypotension, hypoperfusion, and organ dysfunction, all of which are clinical hallmarks in severe sepsis and septic shock. The endotoxemic pig is a commonly used model to replicate human Gram-negative septic shock with all of the classic clinical signs. In the model used in this study, responses in temperature, cytokines, leukocytes, and platelets were chosen as indicators of the inflammatory and coagulation cascade systems, whereas BE and static pulmonary compliance represented hypoperfusion and organ dysfunction, respectively.
Linear responses in hypotension, BE, and static pulmonary compliance to logarithmic increases in the endotoxin dose were demonstrated in an earlier study (13). A similar dose-response relation was found for leukocytes and platelets, whereas the inflammatory cytokines TNF-α and IL-6 increased logarithmically (13).
The aim of the present study was to investigate whether, in addition to the dose, the rate of increase in endotoxin concentration affects the biological response. In group I, with the fast increase in endotoxin concentration, the highest dose was given initially during the first hour. In group II, there was a gradual increase by escalating the dose once an hour during a 3-h period. The dose in group I was decreased in an opposite manner, resulting in identical cumulative doses after 3 h. In another study using this animal model and continuous endotoxin infusion, all three chosen levels of endotoxin dosing were demonstrated to result in significant responses in an interval of the dose-response curve in which the slope is optimally steep (13). The duration of the endotoxin infusion in the present study was set to 3 h, which corresponds to previous results, demonstrating that all responses, except that of Hb, the indicator of plasma leakage, reached their peaks at different time points within 3 h (13). Accordingly, the postinfusion follow-up period was set to 3 h.
If the dose of endotoxin was the sole determinant of the biologic response, a greater response would be expected in group I during the first 3 h, which then would be followed by an intersection of the response curves during the subsequent 3 h of the experiment when identical total endotoxin doses have been given. However, such a response with no group effect but a highly significant group-by-time interaction was only observed for IL-6. For the rest of the endotoxin dose-dependent variables, significantly greater biologic responses were recorded in group I either initially or during the whole experiment, and in none of them was an opposite response, similar to that of IL-6, observed during the last part of the experiment. No significant dose responses in our model have been demonstrated in CI and DO2, probably because of variation in confounding factors such as, for example, preexperimental hydration status or stress level of the laboratory animals (13). Accordingly, in the present study, no differences between the groups were noted in these variables. It may theoretically be argued that differences at 6 h were relatively limited and, therefore, that the infusion rate does not matter. Still, it must be emphasized that a nonlethal endotoxin model was used in this study, and, as pointed out in the statistical plan, some of the changes induced by the moderate endotoxin doses tend to reverse within 6 h. If higher doses in a lethal model would have been given, it is most likely that the more pronounced circulatory failure, hypoperfusion, and organ dysfunction in group I would have translated into a worse outcome with increased mortality. The mortality that occurred, one case from each group, was due to noncytokine-driven early increases in MPAP that did not reverse. The use of the dichotomous variable mortality as the primary end point in an LD50 model would have required many more animals and higher endotoxin doses that are less clinically relevant, and, therefore, such a model was not chosen.
Thus, our results indicate that the biological response is increased if the organism is exposed to a fixed amount of endotoxin more quickly. This was not only found in the inflammatory variables, that is, TNF-α, leukocytes, and core temperature, but also, and even more evidently, in the more delayed parameters, that is, hypotension, BE, pulmonary dysfunction, and pH. The half-life of endotoxin has been reported to vary from 15 min to a few hours (21-23). Taking this fact into account, the concentration of endotoxin should be higher after 3 h in group II than after 1 h in group I. In contrast, however, the response is more pronounced in group I, emphasizing that the response is not only dependent on the concentration but also on the rate by which this concentration is achieved.
The explanation of these findings is not clear. Concerning the half-life, it may be argued that the area under the endotoxin concentration curve was greater in group I, and that the differences observed could at least partially be explained by this event. Even if this explanation cannot completely be excluded, the magnitude of such an effect would be minimal because IL-6, which is strongly correlated to the biological activity of endotoxin, did not differ between the groups. A more probable explanation is that a slow increase in endotoxin administration may lead to tolerance. The development of endotoxin tolerance, which has been demonstrated in several studies (24, 25), can easily be observed in experimental models with continuous infusions of endotoxin (13). Another possibility is that the dose response represents an example of the modified Weber-Fechner dose action law, with an intracellular signaling that is greater when the ligand binding to the receptor increases with a higher rate (26, 27).
Similar to the physiological response, the TNF-α reaction in this study was dependent on the endotoxin administration rate, whereas the IL-6 response was not. Data from our porcine model therefore support the notion that IL-6, in contrast to TNF-α, is not a key mediator in the important reactions leading to inflammation, hypotension, hypoperfusion, and organ dysfunction. Consistent with our results, several studies using different animal models have demonstrated that infusion of TNF-α results in the same signs and symptoms as endotoxin (28). Regarding IL-6, data are limited. In the mouse, it has been found that, although being a marker for the outcome in septic shock, IL-6 contributes only marginally to the pathogenesis leading to circulatory collapse and death (29). However, our data indicate that the IL-6 response is a good indirect parameter of the endotoxin load irrespective of the administration rate. This finding is particularly noteworthy given that the quantitative measurement of plasma endotoxin with the chromogenic modification of the limulus amebocyte assay has been shown to be notoriously difficult (30).
The finding that biological responses are dependent on both the dose and the administration rate of endotoxin is, of course, important when setting up experimental endotoxemic models in which gradual increases to a certain dose will be associated with a lower response than if the final dose is being set directly. This result is in agreement with a recent publication by Schmidhammer et al. (17) in which a model using a continuously increasing endotoxin infusion was described. In that study, a systemic inflammatory response and progressive acute lung injury were achieved without causing an early increase in MPAP. However, no comparison with any other endotoxin infusion regimen was made, and the lower MPAP was only related to previous experience. Changes in MPAP occur early and not as a consequence of the inflammatory response. Therefore, MPAP was not considered a parameter of interest in the present study. However, the study by Schmidhammer etal. emphasizes the advantage of gradually increasing the initial endotoxin infusion rate in those animals that may respond with increases in MPAP.
The clinical relevance of our findings, however, is not clear. However, because endotoxin infusion can cause sepsis-related symptoms in healthy persons (31), it is reasonable to assume that similar mechanisms may operate in humans. Furthermore, high plasma endotoxin concentrations have been found to correlate to multiple organ failure and mortality in meningococcal disease (32). The results of the present and our earlier study (13) indicate that not only the endotoxin load and the concentrations connected with this but also the rate by which these concentrations are achieved and the preexposure endotoxin levels may be of importance for the biological response and the severity of the disease. This observation may offer an explanation for the distinct and powerful response that is caused by rapidly growing endotoxin-releasing bacteria in an otherwise healthy individual, whereas the response is less evident in a patient with sepsis-induced mucosal changes and increased endotoxin absorption from the gut, although the total load from the gut over time may be much greater. Consequently, preexposure level and rate of concentration increase may add to the list of factors known to confound the correlation between endotoxin levels and the clinical picture (33) and influence the maximally achievable effects of potential endotoxin elimination and other antiendotoxin strategies. These speculations, however, need further investigation.
The authors thank Monika Hall and Anders Nordgren for excellent technical assistance.
1. Alexander C, Rietschel ET: Bacterial lipopolysaccharides and innate immunity. J Endotoxin Res
2. Pridmore AC, Wyllie DH, Abdillahi F, et al.: A lipopolysaccharide-deficient mutant of Neisseria meningitidis
elicits attenuated cytokine release by human macrophages and signals via toll-like receptor (TLR) 2 but not via TLR4/MD2. J Infect Dis
3. Ziegler EJ, McCutchan JA, Fierer J, et al.: Treatment of gram-negative bacteremia and shock with human antiserum to a mutant Escherichia coli
. N Engl J Med
4. Ziegler EJ, Fisher CJ, Sprung CL, et al.: Treatment of gram-negative bacteremia and septic shock with HA-1A human monoclonal antibody against endotoxin. A randomized, double-blind, placebo-controlled trial. The HA-1A Sepsis
Study Group. N Engl J Med
5. Greenman RL, Schein RM, Martin MA, et al.: A controlled clinical trial of E5 murine monoclonal IgM antibody to endotoxin in the treatment of gram-negative sepsis
. The XOMA Sepsis
Study Group. JAMA
6. McCloskey RV, Straube RC, Sanders C, Smith SM, Smith CR: Treatment of septic shock with human monoclonal antibody HA-1A. A randomized, double-blind, placebo-controlled trial. CHESS Trial Study Group. Ann Intern Med
7. Angus DC, Birmingham MC, Balk RA, et al.: E5 murine monoclonal antiendotoxin antibody in gram-negative sepsis
: a randomized controlled trial. E5 Study Investigators. JAMA
8. Bengsch S, Boos KS, Nagel D, Seidel D, Inthorn D: Extracorporeal plasma treatment for the removal of endotoxin in patients with sepsis
: clinical results of a pilot study. Shock
9. Teramoto K, Nakamoto Y, Kunitomo T, et al.: Removal of endotoxin in blood by polymyxin B immobilized polystyrene-derivative fiber. Ther Apher
10. Staubach KH, Boehme M, Zimmermann M, Otto V: A new endotoxin adsorption device in Gram-negative sepsis
: use of immobilized albumin with the MATISSE adsorber. Transfus Apher Sci
11. Suzuki H, Nemoto H, Nakamoto H, et al.: Continuous hemodiafiltration with polymyxin-B immobilized fiber is effective in patients with sepsis
syndrome and acute renal failure. Ther Apher
12. Vincent JL, Laterre PF, Cohen J, et al.: A pilot-controlled study of a polymyxin B-immobilized hemoperfusion cartridge in patients with severe sepsis
secondary to intra-abdominal infection. Shock
13. Lipcsey M, Larsson A, Eriksson MB, Sjölin J: Inflammatory, coagulatory and circulatory responses to logarithmic increases in the endotoxin dose in the anaesthetised pig
. J Endotoxin Res
14. Marshall JC: Lipopolysaccharide: an endotoxin or an exogenous hormone? Clin Infect Dis
41(Suppl 7):S470-S480, 2005.
15. Redl H, Bahrami S, Schlag G, Traber DL: Clinical detection of LPS and animal models of endotoxemia. Immunobiology
16. Dodds WJ, Abelseth MK: Criteria for selecting the animal to meet the research need. Lab Anim Sci
17. Schmidhammer R, Wassermann E, Germann P, Redl H, Ullrich R: Infusion of increasing doses of endotoxin induces progressive acute lung injury but prevents early pulmonary hypertension in pigs. Shock
18. Eriksson M, Larsson A, Lundkvist K, Loof L: Endotoxaemic liver injury in a porcine model--relation to tumour necrosis factor-alpha release and survival. Scand J Infect Dis
19. Clark CA, Harman EM: Hemodynamic monitoring: pulmonary artery catheters. In Civetta JM, Taylor RW, Kirby RR, (eds.): Critical Care
. 3rd ed. Philadelphia: Lippincott Williams & Wilkins, pp 293-302, 1988.
20. Morrison PR, Meyer MP: Constants for use in body surface area formula: mammals. In Altman PL, Dittmer DS, (eds.): Biology Data Book
. Washington, DC: Federation of American Society of Experimental Biology, pp 120-122, 1964.
21. Mathison JC, Ulevitch RJ: The clearance, tissue distribution, and cellular localization of intravenously injected lipopolysaccharide in rabbits. J Immunol
22. DiLuzio NR, Crafton CG: Influence of altered reticuloendothelial function on vascular clearance and tissue distribution of S. enteritidis
endotoxin. Proc Soc Exp Biol Med
23. Carey FJ, Braude AI, Zalesky M: Studies with radioactive endotoxin. III. The effect of tolerance on the distribution of radioactivity after intravenous injection of Escherichia coli
endotoxin labeled with Cr51. J Clin Invest
24. Fan H, Cook JA: Molecular mechanisms of endotoxin tolerance. J Endotoxin Res
25. Cross AS: Endotoxin tolerance-current concepts in historical perspective. J Endotoxin Res
26. Miyazaki M, Sawada S, Nishide T, Iwanaga K, Morimoto K, Kakemi M: Bioavailability assessment of arginine-vasopressin (AVP) using pharmacokinetic-pharmacodynamic (PK-PD) modeling in the rat. Biol Pharm Bull
27. Bachem A: The modified Fechner-Weber law as a complex law of dose-action. Am J Psychol
28. Tracey KJ, Beutler B, Lowry SF, et al.: Shock and tissue injury induced by recombinant human cachectin. Science
29. Libert C, Vink A, Coulie P, et al.: Limited involvement of interleukin-6 in the pathogenesis of lethal septic shock as revealed by the effect of monoclonal antibodies against interleukin-6 or its receptor in various murine models. Eur J Immunol
30. Marshall JC, Walker PM, Foster DM, et al.: Measurement of endotoxin activity in critically ill patients using whole blood neutrophil dependent chemiluminescence. Crit Care
31. Martich GD, Boujoukos AJ, Suffredini AF: Response of man to endotoxin. Immunobiology
32. Brandtzaeg P, Kierulf P, Gaustad P, et al.: Plasma endotoxin as a predictor of multiple organ failure and death in systemic meningococcal disease. J Infect Dis
33. Opal SM: The clinical relevance of endotoxin in human sepsis
: a critical analysis. J Endotoxin Res