Oxygen toxicity

Atul Malhotra, MD, FRCPC
David R Schwartz, MD
Richard M Schwartzstein, MD

UpToDate performs a continuous review of over 330 journals and other resources. Updates are added as important new information is published. The literature review for version 13.1 is current through December 2004; this topic was last changed on January 13, 2005. The next version of UpToDate (13.2) will be released in June 2005.

INTRODUCTION The potential toxicity of oxygen was appreciated soon after its discovery by Priestly in the late 18th century [1]. While the administration of supplemental oxygen is invaluable in many clinical situations, excessive or inappropriate oxygen therapy has potentially deleterious consequences. Human and animal data suggest that a spectrum of lung injury, ranging from mild tracheobronchitis to diffuse alveolar damage (DAD), which is histologically indistinguishable from that observed in the acute respiratory distress syndrome (ARDS), may result from high concentrations of inspired oxygen [2-6].

The mechanisms and clinical consequences of oxygen toxicity will be reviewed here. Specific issues related to the administration of oxygen in the setting of hypercapnia and guidelines for long-term oxygen therapy are discussed separately. (See "Use of oxygen in patients with hypercapnia" and see "Long-term supplemental oxygen therapy").

CELLULAR INJURY Hyperoxia appears to produce cellular injury through increased production of reactive oxygen species, such as the superoxide anion, the hydroxyl radical, and hydrogen peroxide [7]. When the production of these reactive species increases and/or the cell's antioxidant defenses are depleted, oxygen radicals can react with and impair the function of essential intracellular macromolecules, resulting in cell death [8].

Oxygen free radicals may also promote a deleterious inflammatory response, leading to secondary tissue damage and/or apoptosis [9-11]. Support for the role of reactive oxygen species in producing cellular injury has come from studies in transgenic mice with altered superoxide dismutase activity. Mice with augmented antioxidant mechanisms have a relative tolerance to hyperoxia, while manganese superoxide dismutase knockout mice die shortly after birth with extensive mitochondrial injury within degenerating neurons and cardiac myocytes [12-14].

Lung tissue is exposed to the highest concentrations of oxygen in the body, placing cells that line the tracheobronchial tree and alveoli at the greatest risk for hyperoxic cytotoxicity [15]. Hyperoxia may also increase susceptibility to mucous plugging, atelectasis, and secondary infection by impairing both mucociliary clearance and the bactericidal capacity of immune cells [16-21].

CLINICAL CONSEQUENCES High fractions of inspired oxygen (FiO2) have been associated with several pulmonary consequences: increased intrapulmonary right-to-left shunt fraction and diminished lung volumes due to absorptive atelectasis; accentuation of hypercapnia; and damage to airways and pulmonary parenchyma. The term "oxygen toxicity" is usually reserved for the last of these consequences, ie, tracheobronchial and pulmonary parenchymal damage.

Absorptive atelectasis High concentrations of supplemental oxygen, cause washout of alveolar nitrogen. This may lead to absorptive atelectasis if absorption of oxygen from the alveoli occurs at a faster rate than it is replenished by inhaled oxygen. Absorptive atelectasis is theoretically more likely in the following circumstances:

   A low regional ventilation-perfusion ratio, which limits replenishment of alveolar oxygen (in general, airway closure is required for this mechanism to be important)

   Qualitative or quantitative surfactant abnormalities that promote alveolar collapse and further reduce the ventilation-perfusion ratio

   A high rate of oxygen uptake, due to an increased in metabolic demand

   An impaired pattern of respiration that fails to correct atelectasis (eg, ventilation at low tidal volumes and/or without intermittent sighs)

Shunting resulting from absorptive atelectasis is minor in younger patients, but can rise to as high as 11 percent in older, otherwise healthy volunteers breathing 100 percent oxygen for 30 minutes [22].

Once well established, absorptive atelectasis is not rapidly reversed by a reduction of FiO2 to maintenance levels, emphasizing the desirability of rapid titration of FiO2 to the lowest fraction necessary to maintain an SaO2 >90 percent [23].

Decrements in vital capacity of up to 20 percent have been noted after hyperoxic exposure in a number of experiments [2,16]. This is presumably due to a combination of absorptive atelectasis and shallow breaths secondary to pleuritic pain from tracheobronchitis.

Accentuation of hypercapnia Hyperoxic hypercarbia describes the phenomenon of increased PaCO2 associated with increases in FiO2 in individuals with chronic compensated respiratory acidosis. In general, the increased hypercapnia does not lead to CO2 narcosis and respiratory failure, because the relative rise in PaCO2 is small and these patients are acclimatized to their higher baseline level of PaCO2. (See "Use of oxygen in patients with hypercapnia").

Several mechanisms act in concert to produce hyperoxic hypercapnia, including [24-26]:

   The Haldane effect. The rightward displacement of the CO2-hemoglobin dissociation curve in the presence of increased oxygen saturation. This occurs because oxyhemoglobin binds CO2 less avidly than deoxyhemoglobin, thereby increasing the amount of CO2 dissolved in blood, which in turn determines PaCO2.

   An increase in dead space (ie, "wasted") ventilation. This probably reflects worsening ventilation-perfusion matching and redistribution of blood flow from well-ventilated to poorly ventilated alveoli due to a loss of hypoxic pulmonary vasoconstriction. Carbon dioxide-induced bronchodilation may also be important in mediating the increase in dead space [

   A modest decrease in minute ventilation due to decreased stimuli from peripheral chemoreceptors to the central respiratory center. (
See "Control of ventilation").

   The anxiolytic and anti-dyspneic effects of supplemental oxygen can promote sleep, particularly in patients who arrive in the emergency room sleep-deprived. The onset of sleep is associated with loss of the voluntary drive to breath. When the behavioral influences on breathing present during wakefulness are absent, respiration is sustained solely by metabolic control mechanisms. This can result in progressive hypercarbia [

   Hypoventilation results in decreased inspiratory flow demand; this reduces the amount of room air entrained around a face mask or nasal cannula. This may result in an increase in the FIO2 delivered to the patient, even if the rate of flow from the oxygen source is unchanged [

Airway injury Many healthy volunteers experience substernal heaviness, pleuritic chest pain, cough, and dyspnea within 24 hours of breathing pure oxygen; these symptoms are probably due to a combination of tracheobronchitis and absorptive atelectasis [30]. Erythema and edema of large airways can be observed bronchoscopically in most patients treated with a FiO2 of 0.9 for six hours and are thought to reflect hyperoxic bronchitis [31]. In addition, the concentration of reactive oxygen species in exhaled breath increases after only one hour of breathing 28 percent oxygen, regardless of the presence of underlying lung disease [32].

Bronchopulmonary dysplasia (BPD), a disease seen in neonates following recovery from neonatal respiratory distress syndrome, has been attributed to the effects of mechanical ventilation and oxygen toxicity in the immature lung. BPD is characterized by epithelial hyperplasia and squamous metaplasia in the large airways, thickened alveolar walls, and peribronchial and interstitial fibrosis. Infants with BPD generally suffer respiratory distress and require supplemental oxygen for up to six months.

Parenchymal injury Progressive worsening of airspace disease can be observed in patients with ARDS who are sustained on mechanical ventilation, and may be due to progression of the underlying process that produced ARDS, development of ventilator-associated pneumonia, ventilator-induced lung injury secondary to mechanical forces, or DAD from the toxic effects of oxygen. Determining the magnitude of parenchymal injury due solely to oxygen therapy in humans is problematic because confounders are usually present. Many of the limited human data on hyperoxic pulmonary injury are from healthy volunteers, and their relevance to actual clinical settings is debatable.

Several studies involving patients in critical care units have attempted to assess the clinical significance of pulmonary oxygen toxicity but have yielded conflicting results:

   One prospective study randomized 40 patients undergoing cardiac surgery to postoperative treatment with either 100 percent oxygen via an endotracheal tube for a mean of 24 hours, or to the minimum FiO2 required to maintain an arterial oxygen tension of 80 to 120 mmHg [33]. Despite the differences in FiO2 exposure, there was no difference in right-to-left shunt fraction, effective compliance, or dead space to tidal volume ratio between the two groups.

   A second report randomized 10 patients with irreversible brain damage to either 21 percent or 100 percent oxygen for a mean of 52 hours and noted substantial deterioration in the hyperoxic group in terms of decreased PaO2, increased right-to-left shunt fraction, and increased dead space to tidal volume ratio [
34]. However, postmortem examination of the lungs revealed no major histopathologic differences (including no hyaline membranes characteristic of early DAD) among the two treatment groups.

The physiologic deterioration in the hyperoxic patients in the second study may have reflected progressive absorptive atelectasis, which was less prominent in the first series because patients were exposed to hyperoxia for shorter periods of time and had spontaneous respiratory efforts. Alternatively, physiologic changes associated with brain injury or cardiac surgery may explain the apparent differences in behavior of these two populations. (See "Neurogenic pulmonary edema").

   A retrospective study of 16 previously healthy, nonsmoking survivors of ARDS found that treatment with an FiO2 >0.6 for more than 24 hours was both a sensitive and specific predictor of a reduced diffusing capacity (DLCO) at one year [35]. Although there were potential biases in this study based on severity of illness (patients with more biologically severe ARDS presumably received higher concentrations of oxygen), the duration of FiO2 >0.6 was the only variable which predicted the degree of diffusing abnormality.

  Potentiation by bleomycin Patients who receive bleomycin appear to be more susceptible to diffuse alveolar damage following oxygen exposure, based upon in vitro data and uncontrolled clinical experience [36]. The typical presentation of combined bleomycin/oxygen toxicity involves a patient with testicular cancer or Hodgkin's disease who, after receiving bleomycin, requires supplemental oxygen due to aspiration, pneumonia, or general anesthesia. Following oxygen administration, the patient develops subacute worsening of bilateral alveolar infiltrates with increasing dyspnea, nonproductive cough, and decreasing lung compliance.

The diagnosis of this syndrome is established by the exclusion of other processes such as congestive heart failure, worsening pneumonia, ongoing aspiration, or alveolar hemorrhage. Lung injury associated with amiodarone or external beam radiation may involve oxygen radicals in its pathogenesis, and also may predispose individuals to hyperoxic complications.

Extrapulmonary toxicity The retinopathy of prematurity (previously called retrolental fibroplasia) has been attributed to the toxic effects of oxygen. One cohort study of 101 infants demonstrated a significant association between the duration of transcutaneously measured PaO2 >80 mmHg and the incidence and severity of retinopathy [37]. Central nervous system symptoms, including generalized tonic-clonic seizures, have been reported secondary to hyperoxia but are unusual in the absence of hyperbaric therapy.

Hyperoxia may also alter cardiovascular function [38,39]. Increased oxygen tension can lead to local coronary vasoconstriction, and microscopic foci of myocardial necrosis have been observed in animal models [40]. Reductions in stroke volume and cardiac output, relative bradycardia, and an increase in systemic vascular resistance may ensue, but the clinical relevance of hyperoxia-related hemodynamic effects remains unclear [41].

PREVENTION There is no single threshold of FiO2 defining a safe upper limit for prevention of oxygen toxicity. The relative importance of the duration and magnitude of hyperoxic exposure also has not been clearly defined, although it is most likely that the area under the FiO2 versus time curve is the best predictive variable. Factors that are difficult to quantify, such as the adequacy of a given patient's antioxidant defenses, probably also play a role in determining individual susceptibility.

Reduction of FiO2 Reducing the FiO2 to the lowest tolerable limit is a good principle for all patients, in particular those likely to be at risk of hyperoxia-induced lung injury because of a prolonged duration of oxygen therapy or prior therapy with bleomycin. In practical terms, oxygen should be administered to achieve a PaO2 of 60 to 65 mmHg (SaO2 approximately 90 percent).

A number of therapeutic strategies can be employed in patients with ARDS to minimize the need for a high FiO2 and thereby reduce the risk of oxygen toxicity. However, despite their dramatic effect on FiO2 in some cases, none of these therapies has been unequivocally demonstrated to improve overall mortality in ARDS. Interventions of this type include:

   Administration of PEEP to prevent alveolar derecruitment and lower the right-to-left shunt fraction.
   Titration of an "ideal PEEP" in conjunction with a protective ventilation strategy, which has reduced mortality in one study, but has been somewhat controversial [

   Performance of additional alveolar recruitment maneuvers (eg, a sustained delivery of continuous positive airway pressure of 40 cmH2O for 60 seconds) [

   Prone positioning [

   Inverse-ratio or other alternative modes of ventilation. (
See "Alternate modes of mechanical ventilation").

nitric oxide [45,46].

   Extracorporeal membrane oxygenation [

   Liquid ventilation.

Augmentation of antioxidants Because oxygen radicals lead to pulmonary toxicity, enhanced defenses against these molecules theoretically should minimize or prevent lung damage [48]. The potential importance of augmented antioxidant mechanisms is illustrated by the relative tolerance to hyperoxia of transgenic mice with increased superoxide dismutase activity [12].

Animal data suggest potentially beneficial roles for strategies that either inhibit oxidant generation (eg, deferoxamine, allopurinol, tungsten) or provide supplemental antioxidants (eg, alpha tocopherol, beta carotene, N-acetylcysteine, dimethylthiourea) [15,49-53]. As an example, one prospective randomized baboon study demonstrated decreased hyperoxic pulmonary injury in animals randomized to aerosolized manganese superoxide dismutase [54]. A different study induced protection against hyperoxic lung injury in rats by transferring the heme oxygenase-1 gene via a recombinant adenovirus vector [55]. Data on the efficacy of these potential therapies in humans are lacking.

Immune modulators Manipulation of the cytokine milieu may permit modification of the deleterious inflammatory response provoked by hyperoxia. As an example, one study of transgenic mice with increased lung expression of interleukin (IL)-11 found reduced hyperoxic lung injury, possibly mediated through diminished hyperoxia-induced expression of IL-1 and tumor necrosis factor [9]. The improved tolerance to hyperoxia occurred despite only small changes in lung antioxidant concentrations.

RECOMMENDATIONS Although the precise clinical consequences of oxygen toxicity remain incompletely understood, the potential risks should be balanced against the benefits of an increase in arterial oxygen content. The FiO2 and duration of oxygen exposure at which clinically significant oxygen toxicity occurs cannot be predicted with certainty in a given individual, but adherence to the general therapeutic principle of "less is better" appears prudent.

The following measures should be undertaken to minimize the risk of oxygen toxicity:

   Arterial oxygenation should be closely monitored with pulse oximetry and arterial blood gas analysis, and FiO2 should be titrated to the lowest concentration required to meet oxygenation goals.

   We generally aim to maintain an SaO2 of about 90 percent in patients receiving supplemental oxygen. However, clinical judgment should be used in selecting a target oxygen saturation for an individual patient; eg, those with coronary artery disease or pulmonary hypertension may tolerate hypoxemia poorly.

   Patients with any previous
bleomycin exposure should not receive an FiO2 above 0.35 without careful consideration. During the perioperative period, careful monitoring with pulse oximetry and discussions with all involved health care personnel can help avoid the tendency to increase FiO2 for minor and/or brief desaturations.

   Additional strategies to maximize oxygenation and reduce FiO2 (eg, diuresis, bronchopulmonary hygiene, prone positioning, inhaled
nitric oxide, and optimal PEEP) should be undertaken when FiO2 exceeds 0.6 for more than six hours.

   The use of antioxidants to prevent oxygen toxicity remains experimental.

Use of UpToDate is subject to the Subscription and License Agreement.


1.  Gilbert, DL. Oxygen: An overall biological view. In: Oxygen and Living Processes, Gilbert, DL (Ed), Springer-Verlag, New York, 1981, p. 376.

2.  Jenkinson, SG. Oxygen toxicity. New Horiz 1993; 1:504.

3.  Deneke, SM, Fanburg, BL. Normobaric oxygen toxicity of the lung. N Engl J Med 1980; 303:76.

4.  Hedley-Whyte, J. Causes of pulmonary oxygen toxicity. N Engl J Med 1970; 283:1518.

5.  Jackson, RM. Pulmonary oxygen toxicity. Chest 1985; 88:900.

6.  Davis, WB, Rennard, SI, Bitterman, PB, Crystal, RG. Pulmonary oxygen toxicity. Early reversible changes in human alveolar structures induced by hyperoxia. N Engl J Med 1983; 309:878.

7.  Freeman, BA, Crapo, JD. Hyperoxia increases oxygen radical production in rat lungs and lung mitochondria. J Biol Chem 1981; 256:10986.

8.  Fridovich, I. Oxygen toxicity: A radical explanation. J Exp Biol 1998; 201:1203.

9.  Waxman, AB, Einarsson, O, Seres, T, et al. Targeted lung expression of interleukin-11 enhances murine tolerance of 100 percent oxygen and diminishes hyperoxia-induced DNA fragmentation. J Clin Invest 1998; 101:1970.

10.  Barazzone, C, Horowitz, S, Donati, YR, et al. Oxygen toxicity in mouse lung: Pathways to cell death. Am J Respir Cell Mol Biol 1998; 19:573.

11.  Mantell, LL, Lee, PJ. Signal transduction pathways in hyperoxia-induced lung cell death. Mol Genet Metab 2000; 71:359.

12.  White, CW, Avraham, KB, Shanley, PF, Groner, Y. Transgenic mice with expression of elevated levels of copper-zinc superoxide dismutase in the lungs are resistant to pulmonary oxygen toxicity. J Clin Invest 1991; 87:2162.

13.  Lebovitz, RM, Zhang, H, Vogel, H, et al. Neurodegeneration, myocardial injury, and perinatal death in mitochondrial superoxide dismutase-deficient mice. Proc Natl Acad Sci U S A 1996; 93:9782.

14.  Folz, RJ, Abushamaa, AA, Suliman, HB. Extracellular superoxide dismutase in the airways of transgenic mice reduces inflammation and attenuates lung toxicity following hyperoxia. J Clin Invest 1999; 103:1055.

15.  Heffner, JE, Repine, JE. Pulmonary strategies of antioxidant defense. Am Rev Respir Dis 1989; 140:531.

16.  Carvalho, CR, de Paula, Pinto Schettino G, Maranhao, B, Bethlem, EP. Hyperoxia and lung disease. Curr Opin Pulm Med 1998; 4:300.

17.  Huber, GL, Porter, SL, Burley, SW, et al. The effect of oxygen toxicity on the inactivation of bacteria by the lung. Chest 1972; 61:Suppl:66S.

18.  Suttorp, N, Simon, LM. Decreased bactericidal function and impaired respiratory burst in lung macrophages after sustained in vitro hyperoxia. Am Rev Respir Dis 1983; 128:486.

19.  Sackner, MA, Hirsch, JA, Epstein, S, Rywlin, AM. Effect of oxygen in graded concentrations upon tracheal mucous velocity. A study in anesthetized dogs. Chest 1976; 69:164.

20.  Vlessis, AA, Bartos, D, Muller, P, Trunkey, DD. Role of reactive O2 in phagocyte-induced hypermetabolism and pulmonary injury. J Appl Physiol 1995; 78:112.

21.  Griffith, DE, Garcia, JG, James, HL, et al. Hyperoxic exposure in humans. Effects of 50 percent oxygen on alveolar macrophage leukotriene B4 synthesis. Chest 1992; 101:392.

22.  Wagner, PD, Laravuso, RB, Uhl, RR, West, JB. Continuous distributions of ventilation-perfusion ratios in normal subjects breathing air and 100 per cent O2. J Clin Invest 1974; 54:54.

23.  Santos, C, Ferrer, M, Roca, J, et al. Pulmonary gas exchange response to oxygen breathing in acute lung injury. Am J Respir Crit Care Med 2000; 161:26.

24.  Dunn, WF, Nelson, SB, Hubmayr, RD. Oxygen-induced hypercarbia in obstructive pulmonary disease. Am Rev Respir Dis 1991; 144:526.

25.  Aubier, M, Murciano, D, Milic-Emili, J, et al. Effects of the administration of O2 on ventilation and blood gases in patients with chronic obstructive pulmonary disease during acute respiratory failure. Am Rev Respir Dis 1980; 122:747.

26.  Sassoon, CS, Hassell, KT, Mahutte, CK. Hyperoxic-induced hypercapnia in stable chronic obstructive pulmonary disease. Am Rev Respir Dis 1987; 135:907.

27.  Robinson, TD, Freiberg, DB, Regnis, JA, Young, IH. The role of hypoventilation and ventilation-perfusion redistribution in oxygen-induced hypercapnia during acute exacerbations of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2000; 161:1524.

28.  Malhotra, A, Schwartz, DR, Ayas, N, et al. Treatment of oxygen-induced hypercapnia. Lancet 2001; 357:884.

29.  Orem, J. The nature of the wakefulness stimulus for breathing. Prog Clin Biol Res 1990; 345:23.

30.  Comroe, JH, Dripps, RD, Dumke, PR, Deming, M. The effect of inhalation of high concentrations of oxygen for 24 hours on normal men at sea level and at a simulated altitude of 18,000. JAMA 1945; 128:710.

31.  Sackner, MA, Landa, J, Hirsch, J, Zapata, A. Pulmonary effects of oxygen breathing; A 6-hour study in normal men. Ann Intern Med 1975; 82:40.

32.  Carpagnano, GE, Kharitonov, SA, Foschino-Barbaro, MP, et al. Supplementary oxygen in healthy subjects and those with COPD increases oxidative stress and airway inflammation. Thorax 2004; 59:1016.

33.  Singer, MM, Wright, F, Stanley, LK, et al. Oxygen toxicity in man. A prospective study in patients after open-heart surgery. N Engl J Med 1970; 283:1473.

34.  Barber, RE, Hamilton, WK. Oxygen toxicity in man. A A/Barber RE, Lee J, Hamilton WK: Oxygen toxicity in man. A prospective study in patients with irreversible brain damage. N Engl J Med 1970; 283:1478.

35.  Elliott, CG, Rasmusson, BY, Crapo, RO, et al. Prediction of pulmonary function abnormalities after adult respiratory distress syndrome (ARDS). Am Rev Respir Dis 1987; 135:634.

36.  Berend, N. Protective effect of hypoxia on bleomycin lung toxicity in the rat. Am Rev Respir Dis 1984; 130:307.

37.  Flynn, JT, Bancalari, E, Snyder, ES, et al. A cohort study of transcutaneous oxygen tension and the incidence and severity of retinopathy of prematurity. N Engl J Med 1992; 326:1050.

38.  Lodato, RF. Effects of normobaric hyperoxia on hemodynamics and O2 utilization in conscious dogs. Adv Exp Med Biol 1990; 277:807.

39.  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 1972; 45:763.

40.  Busing, CM, Kreinsen, U, Buhler, F, Bleyl, U. Light and electron microscopic examinations of experimentally produced heart muscle necroses following normobaric hyperoxia. Virchows Arch A Pathol Anat Histol 1975; 366:137.

41.  Lodato, RF. Decreased O2 consumption and cardiac output during normobaric hyperoxia in conscious dogs. J Appl Physiol 1989; 67:1551.

42.  Amato, MB, Barbas, CS, Medeiros, DM, et al. Effect of a protective-ventilation strategy on mortality in the acute respiratory distress syndrome. N Engl J Med 1998; 338:347.

43.  Rothen, HU, Sporre, B, Engberg, G, et al. Reexpansion of atelectasis during general anaesthesia: a computed tomography study. Br J Anaesth 1993; 71:788.

44.  Albert, RK. The prone position in acute respiratory distress syndrome: Where we are, and where do we go from here. Crit Care Med 1997; 25:1453.

45.  Dellinger, RP, Zimmerman, JL, Taylor, RW, Straube, RC. Placebo and inhaled nitric oxide mortality the same in ARDS clinical trial. Crit Care Med 1998; 26:619.

46.  Dellinger, RP, Zimmerman, JL, Taylor, RW, et al. Effects of inhaled nitric oxide in patients with acute respiratory distress syndrome: Results of a randomized phase II trial. Inhaled Nitric Oxide in ARDS Study Group. Crit Care Med 1998; 26:15.

47.  Lewandowski, K, Rossaint, R, Pappert, D, et al. High survival rate in 122 ARDS patients managed according to a clinical algorithm including extracorporeal membrane oxygenation. Intensive Care Med 1997; 23:819.

48.  Comhair, SA, Erzurum, SC. Antioxidant responses to oxidant-mediated lung diseases. Am J Physiol Lung Cell Mol Physiol 2002; 283:L246.

49.  Kolleck, I, Sinha, P, Rustow, B. Vitamin E as an antioxidant of the lung: mechanisms of vitamin E delivery to alveolar type II cells. Am J Respir Crit Care Med 2002; 166:S62.

50.  Bitterman, M, Melamed, Y, Ben-Amotz, A. Beta-carotene and CNS oxygen toxicity in rats. J Appl Physiol 1994; 76:1073.

51.  Tang, G, White, JE, Gordon, RJ, et al. Polyethylene glycol-conjugated superoxide dismutase protects rats against oxygen toxicity. J Appl Physiol 1993; 74:1425.

52.  Turrens, JF, Crapo, JD, Freeman, BA. Protection against oxygen toxicity by intravenous injection of liposome- entrapped catalase and superoxide dismutase. J Clin Invest 1984; 73:87.

53.  Dennery, PA, Spitz, DR, Yang, G, et al. Oxygen toxicity and iron accumulation in the lungs of mice lacking heme oxygenase-2. J Clin Invest 1998; 101:1001.

54.  Welty-Wolf, KE, Simonson, SG, Huang, YCT, et al. Aerosolized manganese SOD decreases hyperoxic pulmonary injury in primates. II. Morphometric analysis. J Appl Physiol 1997; 83:559.

55.  Otterbein, LE, Kolls, JK, Mantell, LL, et al. Exogenous administration of heme oxygenase-1 by gene transfer provides protection against hyperoxia-induced lung injury. J Clin Invest 1999; 103:1047.