During resuscitation from hemorrhagic shock, the degree of the accumulated “oxygen debt” indicates the severity of shock (1), and its rapid repayment can alleviate organ dysfunction and injury (2). Hyperoxia, i.e., ventilation with FIO2 = 1.0, may potentially accelerate repayment of the oxygen debt: in fact, we previously demonstrated that hyperoxia allowed for achieving markedly higher tissue oxygen tensions during progressive hemorrhage and subsequent resuscitation with retransfusion of shed blood and fluid administration (3). This better maintenance of tissue oxygenation may counterbalance hyperinflammation resulting from tissue hypoxia, e.g., due to hemorrhagic shock, and thus attenuate organ dysfunction and tissue injury (4). In agreement with this rational, we previously showed that hyperoxia attenuated heart, kidney, and liver injury and ultimately improved survival in healthy swine during resuscitation from near-lethal hemorrhagic shock (5). However, shorter exposure to hyperoxia (12 rather than 24 h) after less severe [mean blood pressure target 40 mm Hg over 3 h (6) vs. 35 mm Hg over 4 h (5)] hemorrhagic shock did not exert any organ-protective effects (6). Moreover, it is well established that hyperoxia can have deleterious side effects due to excess formation of reactive oxygen species (ROS) (7), slowly impairing nitric oxide (NO) availability (8), which ultimately causes macro- and microcirculatory vasoconstriction, in particular in the cerebral and coronary vascular beds (7). In fact, hyperoxia aggravated myocardial ischemia in otherwise healthy pigs with acute left anterior coronary artery stenosis induced by inflation of a hydraulic occluder (9). In sharp contrast, we showed in swine with pre-existing coronary artery disease (CAD) undergoing hemorrhage and resuscitation that hyperoxia during resuscitation mitigated heart injury as indicated by lower troponin blood levels. However, parameters of cardiac function as assessed using left ventricular (LV) pressure conductance catheterization remained unaffected (10). Given these contradictory findings and since no data are available on the effects of hyperoxia on cardiac function during resuscitation from hemorrhagic shock without underlying cardiovascular disease, in the present study we therefore tested the hypothesis that hyperoxia would mitigate cardiac injury and improve cardiac function similar to our aforementioned study in swine with underlying CAD.
After approval by the University of Ulm Animal Care Committee and the Federal Authorities for Animal Research (animal experiment number 1189, date of approval August 15, 2014), experiments were performed in adherence with the National Institute of Health Guidelines on the Use of Laboratory Animals and the European Union “Directive 2010/63/EU on the protection of animals used for scientific purposes.” Twenty-one animals were investigated, of which two animals dropped out because of uncontrollable bleeding during surgery and refractory ventricular fibrillation during insertion of the left ventricular catheter, respectively. Due to these two dropouts, the data presented refer to 19 familial hypercholesterolemia Bretoncelles Meishan (FBM) pigs of either gender (13 females and 6 males) with a median (inter-quartile range) body weight of 77 kg (73; 81) and a median (inter-quartile range) age of 18 months (17; 20). This strain of swine has been used and described in previous studies (10–12) and develops atherosclerosis and CAD provided the animals are fed with a cholesterol-enriched, high-fat diet. To avoid any confounding effect due to a different genetic background, we used the same FBM strain as in our previous study (10), but without the diet mandatory to induce CAD. Accordingly, the animals used showed normal total cholesterol levels [3.1 (2.3;3,5) mmol L−1] as compared with pigs with high-fat diet [12.1 (10.1;13.6) mmol L−1] (10), and sections of the coronary arteries were normal as well (Supplemental Digital Content 1, https://links.lww.com/SHK/A829).
After induction of anesthesia and endotracheal intubation, animals were mechanically ventilated and a central venous catheter was placed similar to our previous publication (10) (Supplemental Digital Content 2, https://links.lww.com/SHK/A830). A balloon-tipped thermodilution pulmonary artery catheter was inserted to measure central venous pressure, mean pulmonary artery pressure, pulmonary artery occlusion pressure (PAOP), and cardiac output (CO) via an 8F catheter sheath in the left carotid artery a pressure-conductance catheter (CD Leycom, Hengelo, The Netherlands). The catheter was advanced into the LV under control of the pressure curve. The pressure conductance catheter technique was used, because in contrast to imaging techniques (echocardiography or magnet resonance) it enables simultaneous assessment of LV volumes and direct pressure measurements. A 4F PiCCO catheter (PULSION Medical Systems SE, Feldkirchen, Germany) for continuous cardiac output measurement and a 10F arterial catheter sheath were placed in the two femoral arteries and used to induce hemorrhagic shock, blood sampling, and blood pressure (mean arterial pressure, MAP) monitoring. The anesthesia and surgical procedures and the experimental protocol are described in detail in the Supplemental Digital Content 2 (see https://links.lww.com/SHK/A830), an overview of the protocol is shown in Figure 1. While the control group received standard resuscitation, the FIO2 was set to 1.0 for the first 24 h of resuscitation in the hyperoxia group, and then modified according to the standard treatment protocol. We used FIO2 of 1.0 rather than other levels of hyperoxia in analogy to our previous study in swine with CAD (10), and because—if existing—any “ideal paO2” and/or “ideal FIO2” are unknown. The duration of hyperoxia was again chosen according to our previous study in swine with CAD (10), and because hyperoxia-induced lung damage in mechanically ventilated patients was demonstrated only after 2 or more days of mechanical ventilation and probably without using lung-protective ventilator settings (13). Data sets were recorded before and at the end of shock as well as every 12 h of resuscitation.
As a marker of lipid peroxidation, we measured 8-Isoprostane plasma concentrations using a commercially available immunoassay kit (ELISA kit Cayman Chemicals, Ann Arbor, Mich). Plasma nitrate+nitrite levels, the stable metabolite of NO, were analyzed with the chemoluminescence method after the reaction of nitrate+nitrite to NO with vanadium chloride (Sievers NOA 280i analyzer, GE Analytical Instruments, Boulder, Colo). Reaction of the superoxide radical with NO results in formation of peroxynitrite, which is mirrored by the tissue formation of 3-nitrotyrosine (3-NT) (14). For the determination of 3-NT formation, heart specimens underwent post-mortem immune-histochemical analysis. Distinct heart left ventricle cross sections were dissected, fixed in formalin, dehydrated, and embedded in paraffin. Sections of 3 μm thickness were cut and mounted on slides, deparaffinized in xylene, and rehydrated with a graded series of ethanol to water. After heat induced epitope retrieval in 10 mM citrate buffer and mirowaving, the sections were blocked with 10% goat serum before incubating with primary antibody: antinitrotyrosine (polyclonal rabbit; Merck Millipore, Darmstadt, Germany). Primary antibody detection was performed by an alkaline phosphatase-conjugated secondary antirabbit antibody and visualized with an alkaline phosphatase substrate red chromogen (Dako REAL, Dako, Germany) followed by counterstaining with hematoxylin. Two representative 800,000 μm2 sections per slide were graded (with a Zeiss Axio Imager A1 microscope with a ×10 objective). Quantification of the intensity of the red chromogen was performed (using the AxioVision 4.8 software, Zeiss, Jena, Germany). Results are presented as median densitometric red (10). In addition to immune-histochemistry determination of 3- NT formation, western blotting was used to assess the tissue expression of the inducible isoform of the nitric oxide synthase (iNOS), which is well established to be induced by inflammatory cytokines via the NF-κB pathway (15) and heme oxygenase-1 (HO-1), a marker of the tissue adaptive response against oxidative stress (16) that is induced by hyperoxia (17). We assessed the identical parameters of inflammation, apoptosis as well as mitochondrial respiration as in the aforementioned study on swine with CAD (10), and the remaining methods used are described in detail in the Supplemental Digital Content 2 (see https://links.lww.com/SHK/A830).
Data are median and quartiles unless otherwise stated. The Log-rank (Mantel-Cox) test, the Kruskal–Wallis test, the Mann–Whitney test, and Dunn multiple comparisons test were used (P < 0.05 was considered statistically significant). GraphPad Prism 7 for Mac (GraphPad Software, La Jolla, Calif) was used for statistical analysis and graphical presentation. Power analysis (power 0.8; α = 0.05) to determine a clinically relevant (approximately 50%) reduction of norepinephrine requirements needed to achieve MAP targets yielded 10 animals per group. Since interim analysis suggested that the “H0” would not be refuted, the study was terminated prematurely to comply with the “3R” (replacement, reduction, and refinement of animal experiments) principle (18).
In the control group, 5 of 11 animals died before the scheduled end of the experiment at 48 h of resuscitation, whereas two of eight pigs died in the hyperoxia group (P = 0.359; Fig. 2).
Hemodynamics, gas exchange, and acid–base state
Data of hemodynamics, gas exchange, acid-base status and metabolism (Table 1) did not significantly differ between the two groups. The amounts of shed blood were similar as well [control 29 (25;34); hyperoxia 29 (26;34) mL kg−1, P = 1.0]. Despite aggressive fluid resuscitation, which did not significantly differ between the two experimental groups [crystalloid: control 13 (11;15), hyperoxia (11;15) mL kg−1 h−1, P = 0.857], all animals needed norepinephrine infusion to achieve MAP target, the infusion rates [control: 0.99 (0.22;1.60) , hyperoxia: 0.71 (0.43;1.58) μg kg−1 min−1, P = 0.717] being comparable as well. According to the protocol, hemorrhage induced tachycardia, low CO, arterial hypotension, and reduced cardiac filling pressures. MAP and CO normalized in both groups during resuscitation. Hyperoxia was associated with a tendency toward higher CO at 24 and 36 h (P = 0.087 and 0.072, respectively). According to the protocol, the arterial (paO2) and mixed venous oxygen partial pressure (p O2) values were significantly higher in the hyperoxia group at 12 and 24 h of resuscitation (P = 0.001 and 0.002, respectively), resulting in significantly higher (each P = 0.016) paO2 /FIO2 ratios. None of the other parameters of lung mechanics, gas exchange, metabolism, glucose utilization, and acid–base status showed any significant inter-group difference (Table 1).
Troponin I plasma levels (Fig. 3) increased significantly after hemorrhagic shock with peak values at 12 h of resuscitation, however without significant inter-group difference (maximum levels control: 17 (11;24); hyperoxia: 9 (6;23) ng mL−1; P = 0.395).
Parameters of LV function are presented in Table 2. Neither end-systolic pressure (ESP), dp/dtmax and dp/dtmin, end-diastolic (EDV), and end-systolic (ESV) volume, ejection fraction (EF) nor the isovolumetric relaxation constant τ significantly differed between the two groups. Despite comparable fluid resuscitation and norepinephrine infusion rates end-diastolic pressures (EDP) were higher in the hyperoxia group during the resuscitation period, and this difference was significant (P = 0.046) at 24 h of resuscitation, i.e., at the end of the hyperoxic ventilation period.
Parameters of systemic inflammation and ROS and RNS formation
The supplemental digital content 3 (see https://links.lww.com/SHK/A831) shows the plasma cytokine, isoprostane, and nitrite+nitrate concentrations. Surprisingly, interleukin-6 levels (IL-6) were significantly higher at baseline in the control animals (P = 0.016), but this difference disappeared during shock and resuscitation. Tumor necrosis factor alpha, isoprostane, and nitrate+nitrite significantly increased during resuscitation as compared with baseline levels, however, without significant inter-group difference. Except for significantly lower IL-1β levels, hyperoxia did not affect plasma cytokine concentrations.
Parameters of cardiac tissue inflammation, oxidative and nitrosative stress, and apoptosis
Immuno-histochemical analysis demonstrated that hyperoxia significantly (P < 0.001) attenuated 3-nitrotyrosine formation (Fig. 4). Supplemental digital content 4 and 5 (see https://links.lww.com/SHK/A832 and https://links.lww.com/SHK/A833) show the results and the quantitative analysis of the tissue western blotting: hyperoxia significantly reduced cardiac tissue expression of the iNOS (P = 0.016), the inhibitor of the nuclear transcription factor kappa B inhibitor alpha (IκBα) (P = 0.006), and the anti-apoptotic B-cell lymphoma-extra-large protein Bcl-xL (P = 0.016), whereas HO-1 expression was not affected (P = 0.274).
Mitochondrial respiration, glucose production, and oxidation
The mitochondrial respiratory activity of heart specimens in the coupled (OxPhos) as well as in the uncoupled state did not significantly differ between the groups (Supplemental Digital Content 6, https://links.lww.com/SHK/A834).
In the present study, we tested the hypothesis whether hyperoxia—similar to our previous study in swine with CAD (10) —would mitigate cardiac injury and improve cardiac function in swine without underlying chronic cardiovascular disease. The main findings were that transient hyperoxia did not beneficially affect myocardial injury, cardiac function, or mitochondrial respiration, despite attenuated tissue nitrosative stress.
In the present study, the time course of troponin levels did not show any significant inter-group difference. This finding is in contrast to previous data on hyperoxia-induced aggravation of regional myocardial ischemia in a porcine model of acute occlusion of the left anterior descending coronary artery (9). However, these authors investigated younger and smaller (mean weight 33 kg) and otherwise healthy pigs, whereas we used adult, human-sized animals. Moreover, our swine underwent 3 h of hemorrhagic shock, i.e., prolonged, whole-body low flow conditions, whereas the duration of the complete study by Guensch et al. was not detailed. Finally, our experiment comprised full-scale intensive care support, whereas explicitly Guensch et al. excluded hemodynamically unstable animals from the analysis. In other words, Guensch et al. investigated the impact of short-term hyperoxia upon acutely induced myocardial ischemia due to isolated iatrogenic stenosis in hemodynamically stable animals, whereas we studied cardiac function in hemodynamically unstable, human-sized pigs that had generalized low-flow conditions due to hemorrhagic shock. We had previously shown in otherwise healthy, adult human-sized swine that 22 h of hyperoxic resuscitation from near-lethal hemorrhagic shock (55% survival only in the control group) was organ-protective and ultimately even improved survival (89% survival in the hyperoxia group under those conditions) (5). In addition, 24 h of hyperoxia followed by 24 h of standard treatment also attenuated organ injury and improved survival (survival 89% in the hyperoxia vs. 50% in the control group) in a porcine model of hemorrhage and resuscitation in animals with pre-existing CAD (10).
In contrast, hyperoxia of shorter duration failed to beneficially affect organ (dys) function and outcome in non-lethal hemorrhagic shock (survival 89% and 100% in the control and hyperoxia groups, respectively) (6). Altogether these results agree with our previously introduced concept (6) that hyperoxia during resuscitation from hemorrhagic shock may only be efficient and, consequently, might only mitigate myocardial injury under conditions of critical coronary oxygen transport capacity, e.g., with underlying CAD (10), near-lethal hemorrhagic shock (5), or profound hemodilution associated with myocardial ischemia (19). It is noteworthy in this context that even in the control group (FIO2 = 0.3), median paO2 was 102 mm Hg and 107 mm Hg at 12 and 24 h of resuscitation; in other words, animals never developed arterial hypoxemia (except for one individual animal (44 mm Hg at 12 h) and three individual animals (49 mm Hg to 54 mm Hg at 24 h)) during first 24 h of resuscitation, and thus, we may have missed possibly beneficial effects of hyperoxia. This observation is in good agreement with Takasu et al. (20) who had shown in awake rats undergoing volume-controlled hemorrhagic shock that increasing FIO2 beyond 0.5 did not enhance the survival probability in the absence of hypoxemia.
Hyperoxia is known to induce systemic vasoconstriction in particular in the coronary vascular bed, thereby potentially compromising myocardial metabolism (21). We did not observe any differences in MAP, CO, or stroke volume, and fluid resuscitation and norepinephrine requirements were similar as well. We did not find any inter-group differences for ESP, ESV, EF, dp/dtmax, dp/dtmin, EDV, or the isovolumic relaxation constant τ either. Clearly, EDP values were significantly higher in the hyperoxia group at 24 h of resuscitation, i.e., immediately before resuming standard mechanical ventilation. This finding might suggest a “stiffer” LV at this time point. It should be noted, however, that this was not paralleled by similar effects on τ, dp/dtmin, or PAOP. Hence, this finding might have resulted from a type II statistical error.
It is well established that hyperoxia can cause hyper-inflammation (7). Similar to our previous studies (5, 6, 10) hyperoxia did not further increase plasma cytokine levels beyond the effect of hemorrhage and resuscitation alone, and it even decreased both IL-1β levels and cardiac tissue iNOS expression. In contrast, hyperoxia also significantly reduced the expression of the NF-κB inhibitor IκBα and the anti-apoptotic Bcl-xL. We can only speculate on any discrepancies in the inflammatory response to hyperoxia. However, it is well known that tissue hypoxia, e.g., due to hemorrhagic shock, induces hyperinflammation (4), and hyperoxia may counterbalance this effect due to better maintenance of tissue oxygen tensions (3).
Hyperoxia can reduce NO availability due to excess ROS formation scavenged by NO (22), ultimately causing systemic and coronary vasoconstriction (8). In our experiment, hyperoxia significantly attenuated tissue nitrotyrosine expression, a marker of peroxynitrite formation from the reaction of the superoxide radical with NO (14), whereas it did not affect plasma nitrite+nitrate, the stable metabolite of NO (23), and isoprostane concentrations, well-established markers of ROS-induced lipid peroxidation (24). We (5, 6, 10) and others (25, 26) previously observed such a different response of systemic markers of NO and ROS production and tissue nitrosative stress, respectively. Hyperoxia also reduced cardiac iNOS expression, whereas HO-1 expression, a marker of the tissue adaptive response against oxidative stress (16), was not affected. Altogether, our data suggest that during the early, acute resuscitation phase from hemorrhagic shock, hyperoxia did not aggravate cardiac tissue oxidative or nitrosative stress. In this context, the continuous i.v. norepinephrine infusion may have assumed importance: norepinephrine per se can cause excess formation of the superoxide radical, and thereby induce oxidative stress (27, 28), in particular in the heart (29). The high norepinephrine infusion rates may have blunted any additional hyperoxia-related increase in ROS release.
In a rat model of cardiac arrest, mitochondrial respiration in cardiac myocytes isolated 60 min after restoration of spontaneous circulation (ROSC) was lower after hyperoxia yielding mean paO2 ≈ 280 mm Hg when compared with animals with post-ROSC paO2 ≈ 105 mm Hg (30). A hyperoxia-related reduction of myocardial O2 uptake was already demonstrated in patients with stable CAD as well as in healthy controls (31). This drop in myocardial O2 uptake coincided with increased lactate extraction and even reversed lactate production to extraction, indicating improved myocardial metabolism (32). In the present study, immediate post mortem tissue mitochondrial respiration did not significantly differ between the two groups, thereby confirming our previous findings in swine with CAD (10). Hence, together with the above-mentioned comparable plasma troponin levels, our data suggest that exposure to hyperoxia during the acute phase of resuscitation from hemorrhagic shock was not associated with impaired myocardial energy metabolism, however, without any beneficial effect either.
Hyperoxia can induce formation of atelectasis and thereby increase intrapulmonary right-to-left shunt in alveoli with low ventilation/perfusion ratios (33). In our study, the paO2/FIO2 ratio was significantly higher during the hyperoxia period despite similar lung mechanics as assessed by airway pressures. It is unlikely, that the higher paO2/FIO2 ratios indicated improved lung function: the relationship between paO2/FIO2 ratio and FIO2 is non-linear and depends on the underlying intrapulmonary shunt fraction. For moderately increased shunt fractions (between 10% and 30% of cardiac output) as in our experiment, the paO2/FIO2 ratio in fact increases with FIO2 > 60% (34, 35).
Hyperoxia was initiated only with the start of resuscitation phase rather than already during the hemorrhagic shock phase. Clearly, we may have thus missed potentially beneficial effects of hyperoxia. However, this approach was chosen to replicate the design of our previous studies (5, 6, 10) to allow for comparison between these different studies.
Our data may be questioned since we used the same FBM strain used as in the previous study in swine with pre-existing CAD (10). However, no atherogenic diet was applied, and subsequently cholesterol levels were normal, and CAD was absent upon histological analysis (Supplemental Digital Content 1, https://links.lww.com/SHK/A829). Moreover, using this strain avoided any confounding effect of different genetic background. Clearly, the large data variation and the limited number of experiments resulting from premature study termination to comply with the “3R” principle (18), as well as dropouts and premature deaths should caution any firm conclusion due to possible statistical type 2 error.
In conclusion, we could not confirm our original hypothesis that hyperoxia (FIO2 of 1.0)—similar to our aforementioned study in swine with underlying CAD—would also mitigate cardiac injury and improve cardiac function during resuscitation from hemorrhagic shock without underlying cardiovascular disease, when compared with a control group with FIO2 of 0.3. Nevertheless, hyperoxia was devoid of deleterious side effects. Hence, under these conditions hyperoxia most likely represents a futile therapeutic approach under these conditions.
1. Rixen D, Raum M, Holzgraefe B, Sauerland S, Nagelschmidt M, Neugebauer EA. A pig hemorrhagic shock model: oxygen debt and metabolic acidemia as indicators of severity. Shock
16 3:239–244, 2001.
2. Barbee RW, Reynolds PS, Ward KR. Assessing shock resuscitation strategies by oxygen debt repayment. Shock
33 2:113–122, 2010.
3. Dyson A, Simon F, Seifritz A, Zimmerling O, Matallo J, Calzia E, Radermacher P, Singer M. Bladder tissue oxygen tension monitoring in pigs subjected to a range of cardiorespiratory and pharmacological challenges. Intensive Care Med
38 11:1868–1876, 2012.
4. Eltzschig HK, Carmeliet P. Hypoxia and inflammation. N Engl J Med
364 7:656–665, 2011.
5. Knöller E, Stenzel T, Broeskamp F, Hornung R, Scheuerle A, McCook O, Wachter U, Vogt JA, Matallo J, Wepler M, et al. Effects of hyperoxia and Mild therapeutic hypothermia during resuscitation from porcine hemorrhagic shock. Crit Care Med
44 5:e264–e277, 2016.
6. Nussbaum BL, Stenzel T, Merz T, Scheuerle A, McCook O, Wachter U, Vogt JA, Matallo J, Gässler H, Gröger M, et al. Hyperoxia or therapeutic hypothermia during resuscitation from non-lethal hemorrhagic shock in swine. Shock
48 5:564–570, 2017.
7. Hafner S, Beloncle F, Koch A, Radermacher P, Asfar P. Hyperoxia in intensive care, emergency, and peri-operative medicine: Dr. Jekyll or Mr. Hyde? A 2015 update. Ann Intensive Care
5 1:42, 2015.
8. McNulty PH, King N, Scott S, Hartman G, McCann J, Kozak M, Chambers CE, Demers LM, Sinoway LI. Effects of supplemental oxygen administration on coronary blood flow in patients undergoing cardiac catheterization. Am J Physiol Heart Circ Physiol
288 3:H1057–H1062, 2005.
9. Guensch DP, Fischer K, Shie N, Lebel J, Friedrich MG. Hyperoxia exacerbates myocardial ischemia in the presence of acute coronary artery stenosis in swine. Circ Cardiovasc Interv
8 10:e002928, 2015.
10. Hartmann C, Loconte M, Antonucci E, Holzhauser M, Hölle T, Katzsch D, Merz T, McCook O, Wachter U, Vogt JA, et al. Effects of hyperoxia during resuscitation from hemorrhagic shock in swine with preexisting coronary artery disease. Crit Care Med
45 12:e1270–e1279, 2017.
11. Wepler M, Hafner S, Scheuerle A, Reize M, Gröger M, Wagner F, Simon F, Matallo J, Gottschalch F, Seifritz A, et al. Effects of the PPAR-beta/delta agonist GW0742 during resuscitated porcine septic shock. Intensive Care Med Exp
1 1:28, 2013.
12. Matějková Š, Scheuerle A, Wagner F, McCook O, Matallo J, Gröger M, Seifritz A, Stahl B, Vcelar B, Calzia E, et al. Carbamylated erythropoietin-FC fusion protein and recombinant human erythropoietin during porcine kidney ischemia/reperfusion injury. Intensive Care Med
39 3:497–510, 2013.
13. Kapanci Y, Tosco R, Eggermann J, Gould VE. Oxygen pneumonitis in man, light- and electron-microscopic morphometric studies. Chest
62 2:162–169, 1972.
14. Beckman JS, Koppenol WH. Nitric oxide, superoxide, and peroxynitrite: the good, the bad, and ugly. Am J Physiol
271 (5 pt 1):C1424–C1437, 1996.
15. Lind M, Hayes A, Caprnda M, Petrovic D, Rodrigo L, Kruzliak P, Zulli A. Inducible nitric oxide synthase: Good or bad? Biomed Pharmacother
16. Sharma HS, Das DK, Verdouw PD. Enhanced expression and localization of heme oxygenase-1 during recovery phase of porcine stunned myocardium. Mol Cell Biochem
196 (1–2):133–139, 1999.
17. Lee PJ, Choi AM. Pathways of cell signaling in hyperoxia. Free Radic Biol Med
35 4:341–350, 2003.
18. Lilley E, Armstrong R, Clark N, Gray P, Hawkins P, Mason K, Lopez-Salesansky N, Stark AK, Jackson SK, Thiemermann C, et al. Refinement of animal models of sepsis and septic shock. Shock
43 4:304–316, 2015.
19. Kemming GI, Meisner FG, Meier J, Tillmanns J, Thein E, Eriskat J, Habler OP. Hyperoxic ventilation at the critical hematocrit: effects on myocardial perfusion and function. Acta Anaesthesiol Scand
48 8:951–959, 2004.
20. Takasu A, Prueckner S, Tisherman SA, Stezoski SW, Stezoski J, Safar P. Effects of increased oxygen breathing in a volume controlled hemorrhagic shock outcome model in rats. Resuscitation
45 3:209–220, 2000.
21. Farquhar H, Weatherall M, Wijesinghe M, Perrin K, Ranchord A, Simmonds M, Beasley R. Systematic review of studies of the effect of hyperoxia on coronary blood flow. Am Heart J
158 3:371–377, 2009.
22. Cabrales P, Tsai AG, Intaglietta M. Nitric oxide regulation of microvascular oxygen exchange during hypoxia and hyperoxia. J Appl Physiol (1985)
100 4:1181–1187, 2006.
23. Nüssler AK, Bruckner UB, Vogt J, Radermacher P. Measuring end products of nitric oxide in vivo. Methods Enzymol
24. Basu S. Metabolism of 8-iso-prostaglandin F2alpha. FEBS Lett
428 (1–2):32–36, 1998.
25. Beckman JA, Goldfine AB, Leopold JA, Creager MA. Ebselen does not improve oxidative stress and vascular function in patients with diabetes: a randomized, crossover trial. Am J Physiol Heart Circ Physiol
311 6:H1431–H1436, 2016.
26. Fischer UM, Klass O, Stock U, Easo J, Geissler HJ, Fischer JH, Bloch W, Mehlhorn U. Cardioplegic arrest induces apoptosis signal-pathway in myocardial endothelial cells and cardiac myocytes. Eur J Cardiothorac Surg
23 6:984–990, 2003.
27. Rump AF, Rosen R, Klaus W. Cardioprotection by superoxide dismutase: a catecholamine-dependent process? Anesth Analg
76 2:239–246, 1993.
28. Rump AF, Klaus W. Evidence for norepinephrine cardiotoxicity mediated by superoxide anion radicals in isolated rabbit hearts. Naunyn Schmiedebergs Arch Pharmacol
349 3:295–300, 1994.
29. Neri M, Cerretani D, Fiaschi AI, Laghi PF, Lazzerini PE, Maffione AB, Micheli L, Bruni G, Nencini C, Giorgi G, et al. Correlation between cardiac oxidative stress and myocardial pathology due to acute and chronic norepinephrine administration in rats. J Cell Mol Med
11 1:156–170, 2007.
30. Angelos MG, Yeh ST, Aune SE. Post-cardiac arrest hyperoxia and mitochondrial function. Resuscitation
82 (suppl 2):S48–51, 2011.
31. Ganz W, Donoso R, Marcus H, Swan HJ. Coronary hemodynamics and myocardial oxygen metabolism during oxygen breathing in patients with and without coronary artery disease. Circulation
45 4:763–768, 1972.
32. Fedele FA, Gewirtz H, Capone RJ, Sharaf B, Most AS. Metabolic response to prolonged reduction of myocardial blood flow distal to a severe coronary artery stenosis. Circulation
78 3:729–735, 1988.
33. Radermacher P, Maggiore SM, Mercat A. Fifty Years of Research in ARDS. Gas exchange in acute respiratory distress syndrome. Am J Respir Crit Care Med
196 8:964–984, 2017.
34. Feiner JR, Weiskopf RB. Evaluating pulmonary function: an assessment of PaO2/FIO2. Crit Care Med
45 1:e40–e48, 2017.
35. Aboab J, Louis B, Jonson B, Brochard L. Relation between PaO2/FIO2 ratio and FIO2: a mathematical description. Intensive Care Med
32 10:1494–1497, 2006.