Oxygen Toxicity in the ICU
Special Feature
Oxygen Toxicity in the ICU
By David J. Pierson, MD, FACP, FCCP
During rounds in the trauma icu, a resident presents the case of a young man who was admitted the previous night after sustaining multiple injuries in a motorcycle accident. Following an extensive diagnostic evaluation and an exploratory laparotomy, the patient is now hemodynamically stable and being mechanically ventilated as follows: synchronized intermittent mandatory ventilation (SIMV); tidal volume 11 mL/kg; SIMV rate 14 breaths/min; FIO2 0.60; and PEEP 15 cm H2O. Arterial blood gas studies on these settings showed pH 7.42, PCO2 = 38 mmHg, and PO2 96 = mmHg. On the basis of these findings preparations are underway to insert a pulmonary artery catheter and sequentially increase the PEEP in order to increase the PO2, reduce the FIO2, and thus, decrease the patient's risk of developing oxygen toxicity.
The management of patients with acute respiratory failure, and especially with the acute respiratory distress syndrome (ARDS), remains a subject of intense interest and considerable controversy. Much of this controversy relates to the mechanical aspects of ventilatory support and to how the clinician's approach to management relates to barotrauma and other possible forms of ventilator-induced lung injury. Elsewhere in this issue of Critical Care Alert, the findings of one of several recent studies on this topic are summarized and discussed.
However, another area of controversy that actually predates the concept of ventilator-induced lung injury is pulmonary oxygen toxicity. Much less has been written about this topic recently than about alveolar overdistension and mechanical lung injury from high pressures and sheer forces. Yet, as shown by the preceeding clinical vignette, oxygen toxicity remains very much on the minds of clinicians and profoundly influences how they manage patients. This brief review summarizes our current understanding of pulmonary oxygen toxicity, puts it into the context of the patient with hypoxemic respiratory failure, and raises some of the practical issues facing clinicians who manage such patients.
Mechanisms of Oxygen Toxicity
Oxygen is vital to human life but is also a poison. Although most molecular oxygen used by mitochondria is metabolically reduced completely to water during normal cellular respiration, a certain quantity of partially reduced, reactive oxygen metabolites is also generated, proportional to intracellular PO2.1-3 Chief among these toxic metabolites are superoxide anion (O2-), hydrogen peroxide (H2O2), and reactive hydroxyl radicals (OH·). Of these, the last are believed to be the most toxic, producing destructive oxidation and reduction reactions with cell components.
Production of free radicals and H2O2 is a normal consequence of cellular metabolism, which under normal circumstances is countered by the action of superoxide dismutases and other enzymes, such as glutathione reductase and glucose-6-phosphate dehydrogenase, and also by nonenzymatic antioxidants, such as vitamin A and ascorbic acid. Only when excess oxygen is present intracellularly is the balance upset and toxicity produced. An excess of reactive oxygen species leads to decreases in cellular energy, respiratory burst, protein synthesis, and surfactant function, while the amounts of leukotrienes, neutrophil chemoattractants, growth factors, and other substances are increased. These metabolic changes lead to the physiologic alterations mentioned below and also to the formation of edema, hyaline membranes, and fibrosis in the lung parenchyma.1-3
Effects of Oxygen on the Lungs of Normal Animals and Humans
Many studies have shown that exposure to high inspired oxygen pressures can produce both functional and structural changes in the lungs of normal animals, progressing to death if the exposure is sufficiently prolonged. Several early studies in humans documented the physiologic effects of normobaric hyperoxia. Caldwell and associates exposed four normal men to 98% oxygen for periods ranging from 30 to 74 hours.4 All four subjects experienced cough and chest pain. A progressive fall in both vital capacity and diffucing capacity was observed in all four subjects; although these changes disappeared following exposure, it took several weeks for vital capacity to return to normal in one subject.
From a synthesis of published data on pulmonary oxygen toxicity in humans, Jackson provides the following time course.1 Decreased tracheal mucus velocity is detectable after six hours of 100% oxygen breathing, and signs of tracheobronchitis appear after 12 hours. Physiologic changes such as those mentioned earlier occur between 24 and 48 hours, and abnormalities in gas exchange are detectable after 30 hours. Edema, documented in some studies by bronchoalveolar lavage, occurs at 72-96 hours, and fibrosis presumably follows.
All of the studies mentioned above involve exposing normal individuals to high oxygen concentrations. A landmark study of greater relevance to critical care was reported in 1970 by Barber and colleagues.5 These investigators studied 10 patients with irreversible brain damage-five of whom were ventilated with air and five with 100% oxygen for 31-72 hours prior to death or the harvesting of organs for transplantation. Lung function deteriorated in the patients who were ventilated with oxygen, as manifested primarily by a progressive increase in P(A-a)O2. All five of the oxygen-ventilated patients developed radiographic pulmonary consolidation and had heavy, edematous lungs at autopsy; changes in the air-ventilated patients were less pronounced.
In a second study published at the same time as the Singer paper, patients who had undergone cardiac surgery were randomized to receive either 100% oxygen or an FIO2 sufficient to produce an arterial PO2 of 80-120 mmHg for 24-48 hours.6 The authors of this study were unable to demonstrate any changes in intrapulmonary shunt, dead space to tidal volume ratio, or respiratory system compliance during the study period.
Effects of Drugs
A number of drugs used in the ICU have been shown to affect the toxicity of oxygen.1,3 Epinephrine, norepinephrine, thyroid hormone, and hyperthermia all increase susceptibility to pulmonary oxygen toxicity. Corticosteroids have had variable effects in animal studies. When pretreated with methylprednisolone, animals experience enhanced lung damage from hyperoxia, while administration of the same agent late in the exposure period appears to attenuate the toxic response.
Best known among all drugs that potentiate oxygen toxicity is bleomycin, an anticancer agent. Bleomycin itself causes pulmonary toxicity, and this effect is magnified by the concomitant administration of high concentrations of oxygen. Cyclophosphamide and nitrosourea agents used in chemotherapy also potentiate hyperoxic pulmonary toxicity. Nitrofurantoin and disulfiram also potentiate oxygen toxicity in animal models, although the existence of this effect in humans has not been confirmed.
A number of experimental approaches may have promise as protection of the lung against oxygen toxicity. Although none has been tried in humans, these include the following:3
· Intermittent lowering of FIO2
· Administration of antioxidants (CuZn-SOD; Mn-SOD; catalase)
· Increased gene expression of Mn-SOD in alveolar type II cells
· Antibody neutralization of ICAM-I
· Aerosolized artificial and natural surfactants
· High-dose MgSO4
· Tracheal insufflation of endotoxin
· Previous exposure to sublethal oxygen
Problems with Assessing Oxygen Toxicity in Critically Ill Patients
In preparing this essay, I performed a MEDLINE search on the key words "pulmonary oxygen toxicity" for articles published in English. This search produced 152 citations since 1966-41 of them since 1990. Restricting the search to human studies reduced the total to 56 (14 since 1990), and many of the cited papers dealt mainly with animal data. Except for a few uncontrolled anecdotal reports, there were essentially no studies dealing with actual patients, especially in the setting of acute hypoxemic respiratory failure. The search located several good, recent reviews,7-11 but, in each case, the authors were forced to draw their conclusions from animal data, from studies in normal humans, and from pathologic studies not directly relevant to the question at hand.
This situation creates a problem for the clinician with respect to oxygen toxicity in the ICU. Since the advent of critical care and the emergence of invasive mechanical ventilation more than a generation ago, clinicians have operated on the assumption that oxygen was as toxic for their critically ill patients as it is to normal animals and human volunteers. If this is really so, why do we not see it?
One reason is that the clinical and pathological manifestations of pulmonary oxygen toxicity are essentially the same as those of acute lung injury, making it difficult or impossible to separate the effects of the disease from the putative effects of its treatment. Another is that lung injury has been found to protect animals from the toxic effects of oxygen.12 A third reason is that pulmonary oxygen toxicity appears to be reversible, at least from reported experience with human volunteers.
In several respects, oxygen toxicity is the "sasquatch" of the ICU. Everyone is familiar with the sasquatch, or bigfoot, at least here in the Pacific Northwest. We've all heard frightening stories about it, and we know what it is supposed to look like; in the forest, the prospect of running into it makes us nervous, and we do what we can to stay out of its way. But no one has actually seen it-at least with sufficient documentation to convince others.
Is it possible that clinicians have over-reacted to the threat of oxygen toxicity in the same way? This is an important question, considering the techniques that clinicians commonly use to avoid oxygen toxicity. Primarily among these is PEEP, which in many patients produces an increase in arterial PO2 and permits the FIO2 to be decreased. However, although there may be some disagreement about the clinical risk of oxygen toxicity, there should be no dispute about the potential adverse effects of PEEP. In the clinical vignette with which I began this essay, those managing the patient planned to increase PEEP above the already-established level of 15 cm H2O, despite the need to insert a pulmonary artery catheter in order to detect possible adverse effects on cardiac function. And, in several reported studies of "super-high" PEEP in managing ARDS, the incidence of overt barotrauma has been high.
A Practical Approach to Avoiding Oxygen Toxicity in Patients
Regardless of the conclusions the individual clinician draws from the available literature on oxygen toxicity as it pertains to critically ill patients, several practical steps can be taken to decrease the unnecessary exposure of patients to high inspired oxygen concentrations. (See Table.)
Table
Avoiding unnecessary exposure to the high oxygen concentrations in patients with acute respiratory failure
· Accept a lower arterial PO2
Reduce FIO2 if PaO2 > 80-90 mmHg
Target PaO2 60-80 mmHg (SpO2 90-95%) for most patients
Target PaO2 50-60 mmHg (SpO2 80-90%) in patients with severe hypoxemic respiratory failure
· Optimize PEEP to increase PaO2
Avoid or reduce PEEP if it impairs cardiac function
Reduce PEEP if compliance falls (suggesting barotrauma risk)
· Consider the other components of tissue oxygen delivery
Hemoglobin: consider transfusion, particularly if patient is anemic, or if high FIO2 and/or or high PEEP is required to support PaO2
Cardiac output: consider fluid administration or dobutamine infusion if blood oxygenation is severely impaired and cardiac output is low
· Use lowest FIO2 required under these circumstances
Particularly when possible oxygen toxicity is a concern, a PaO2 higher than necessary should be avoided. Although normal individuals at sea level maintain PaO2 values of 80-100 mmHg, these levels are unnecessary and often inappropriate in critically ill patients. In the patient example given earlier, the PaO2 of 96 mmHg, if reflective of the patient's steady-state condition, is higher than needed; a value of 60-80 would provide adequate oxyhemoglobin saturation and probably permit a decrease in FIO2 to below 0.50.
Many concepts of "optimum PEEP" or "ideal PEEP" have been put forward. With respect to the present discussion, however, a level of PEEP should be used that permits satisfactory oxygenation of the arterial blood without compromising cardiac function or overdistending the lung. Each clinician will have to decide how to define "satisfactory oxygenation." However, many experienced intensivists consider a PaO2 of 50-60 mmHg ("permissive hypoxemia") to be satisfactory in patients with ARDS whose hemoglobin concentrations and cardiac function are also appropriate. Taking these last two components of tissue oxygen delivery into consideration is particularly important when balancing the potential benefits and hazards of increasing PEEP to permit a lower FIO2.
The opinions of different clinicians vary as to what FIO2 is "safe" under the circumstances set forth in the table. In the institution in which I practice, there is a considerable philosophical spread on this issue. A few colleagues prefer to use high levels of PEEP and, when necessary, to augment cardiac output with fluid and dobutamine in order to decrease FIO2 below 0.40 if it is possible to do this. At the other end of the spectrum are clinicians who feel that the administration of 70-80% oxygen for up to several days is the lesser of available evils when moderate levels of PEEP (for example, 10-15 cm H2O) are already in use. I admit to being closer to this second camp. However, whichever approach is taken, consideration of the principles summarized in the table should assure that exposure of patients to unnecessarily high FIO2 is avoided.
References
1. Jackson RM. Oxygen therapy and toxicity. In: Ayres SM, Grenvik A, Holbrook PR, Shoemaker WC, eds. Textbook of Critical Care. 3rd ed. Philadelphia,PA: WB Saunders; 1995:784-789.
2. Tsan MF. Superoxide dismutase and pulmonary oxygen toxicity. Proc Soc Exp Biol Med 1997;214(2): 107-113.
3. Folz RJ, Oxygen toxicity. In: Crystal RG, West JB, Weibel ER, Barnes PJ, eds. The Lung. Scientific Foundations. 2nd ed. Philadelphia,PA: Lippincott-Raven; 1997:2713-2722.
4. Caldwell P, Changes in lung volume, diffusing capacity, and blood gases in men breathing oxygen. J Appl Physiol 1966;21:1477-1483.
5. Barber RE, Oxygen toxicity in man: A prospective study in patients with irreversible brain damage. N Engl J Med 1970;283(27):1478-1484.
6. Singer MM, Oxygen toxicity in man: A prospective study after open heart surgery. N Engl J Med 1970;283:1473-1478.
7. Durbin CG, Wallace KK. Oxygen toxicity in the critically ill patient. Respir Care 1993;38(7):739-750.
8. Jenkinson SG. Oxygen toxicity. New Horiz 1993;1(4): 504-511.
9. Klein J. Normobaric pulmonary oxygen toxicity. Anesth Analg 1990;70(2):195-207.
10. Lodato RF. Oxygen toxicity. Crit Care Clin 1990;6(3): 749-765.
11. Jenkinson SG. Oxygen toxicity. J Intensive Care Med 1988;3:137-152.
12. de los Santos R, Hyperoxia exposure in mechanically ventilated primates with and without previous lung injury. Exp Lung Res 1985;9:255-275.
Subscribe Now for Access
You have reached your article limit for the month. We hope you found our articles both enjoyable and insightful. For information on new subscriptions, product trials, alternative billing arrangements or group and site discounts please call 800-688-2421. We look forward to having you as a long-term member of the Relias Media community.