Supplemental oxygen or something else?

The acute respiratory distress syndrome (ARDS) is characterized by severe hypoxemia and inflammatory damage to the alveolar-capillary barrier, and can be caused by primary damage to the epithelium (pulmonary ARDS) or the endothelium (extrapulmonary ARDS) (1). Once ARDS is recognized, supplemental oxygen is one of the first therapeutic approaches used to rapidly increase alveolar (P_AO₂), arterial (PaO₂), and venous (PvO₂) partial pressures of oxygen. On the other hand, the risks associated with hyperoxia, such as resorption atelectasis and oxygen toxicity, must be considered (2). The negative effects of higher fractions of inspired oxygen (FiO₂) must be recognized independent of higher PaO₂ values. Therefore, in this editorial, we propose to discuss the relationship between unnecessary hyperoxia with high FiO₂ levels, PaO₂ and SpO₂ targets during mechanical ventilation, and optimal cutoffs for PaO₂ or FiO₂ in different clinical conditions.

One longitudinal analysis of previous data from 10 clinical trials, published between 1996 and 2013, sought to determine whether cumulative oxygen exposure that resulted in PaO₂ >80 mmHg (goal) 5 days after enrollment was associated with higher mortality rates at day 90 in ARDS patients (3). The authors found that oxygen exposure resulting in PaO₂ above the protocol goal occurred frequently and was indeed associated with higher 90-day mortality; the data showed that this association may not be explained by reverse causality. In other words, severe patients required more oxygen. However, this cannot explain the increased mortality. The nature of retrospective studies of administrative data precludes any inference of causation. In addition, those patients with high FiO₂ exposure also showed higher plateau pressure, positive end-expiratory pressure (PEEP) levels, and worse hemodynamics, which suggest disease severity regardless of FiO₂ levels. Thus, it is unknown whether oxygen supplementation alone had a significant effect on clinical outcomes.

Aggarwal et al. raise an important question about liberal versus conservative oxygen supplementation. A recent prospective randomized parallel-group trial showed that using a conservative FiO₂ within a range of 0.21–0.80 to achieve the assigned targets of 88–92% SpO₂, compared to 96% SpO₂ for the liberal oxygenation group, did not cause harm (4). In a before-and-after study, a conservative oxygenation strategy (target SpO₂ of 90–92%) was associated with lower incidence of new organ dysfunction (5). It seems clear that FiO₂ should not be fixed, but rather titrated to achieve PaO₂ or SpO₂ targets. There is no reason to keep FiO₂ high when SpO₂ is 100%, or in specific circumstances such as comorbid chronic obstructive pulmonary disease and ARDS. Conversely, fixed FiO₂ levels have been advocated by World Health Organization (WHO) for use in the intraoperative and postoperative periods (6), with a view to reducing the incidence of surgical site infection. This expected beneficial effect must be weighed against the negative effects of supplemental oxygen therapy, such as oxidative stress, which may amplify the inflammatory response. As stated in the British Thoracic Society guidelines for oxygen use (7), a fixed level of FiO₂ should not be the goal in mechanically ventilated patients, instead; oxygen should be prescribed to a target SpO₂ or PaO₂ range.

The Large Observational Study to Understand the Global Impact of Severe Acute Respiratory Failure (LUNG SAFE) trial, conducted during 4 consecutive weeks in the winter of 2014 in a convenience sample of 459 ICUs from 50 countries across 5 continents (8), showed an inverse relationship between FiO₂ and SpO₂, suggesting that clinicians increased FiO₂ to treat hypoxemia, aiming to achieve a target PaO₂ level. In other words, supplemental oxygen therapy was not used liberally. In this line, a recent systematic review and meta-analysis of more than 16,000 acutely ill adults showed that liberal supplemental oxygen was harmful (9). In ARDS, FiO₂ may have consequences on oxygenation index due to the degree of shunt, and therefore may jeopardize early clinical recognition of the syndrome. In this line, LUNG SAFE trial reported that ARDS was found to be underdiagnosed (60.2%). According to the Berlin definition guideline, the PaO₂/FiO₂ ratio should be used at the time of ARDS onset to stratify its severity. Previous observational studies have found that FiO₂ predicts mortality (10). In future clinical trials, standardized ventilator settings could be used instead to better stratify ARDS severity. A previous study (11) showed reduced utility of PaO₂/FiO₂ ratio to stratify ARDS severity at the time of diagnosis due to non-standardized ventilator settings. The authors proposed that ARDS recognition should be a two-step process, whereby initial assessment at ARDS onset (or detection) would be followed by a second assessment 24 h later, under standardized ventilator settings. This would represent a better method for optimizing risk stratification.

One of the most important ventilator settings to be standardized is the FiO₂ level. Since high FiO₂ levels may affect clinical outcome, unnecessarily high FiO₂ levels should be avoided. It is widely accepted that oxygen toxicity emerges at FiO₂ values higher than 0.5–0.6 (12). However, these values were obtained in humans with normal lungs (13), and are thus not representative of respiratory diseases, in which the threshold for oxygen toxicity in the lungs is lower, precluding any extrapolation of safe oxygen levels in critically ill patients. The decrease in FiO₂ in the presence of high PaO₂ values is not so as prompt as the increase of FiO₂ in the presence of low PaO₂ values (14).

In addition to concentration of oxygen, the duration of oxygen supplementation is also an important parameter to be controlled. In fact, previous studies have shown that increased duration of exposure to excess oxygen was associated with a worsening of oxygenation index (15), although this remains controversial (16). With modern ventilators that enable capture of 24-h windows and visualization of parameter trends, duration of exposure to given oxygen concentrations can be easily tracked.

There is a clear disconnection between offering oxygen to achieve normal PaO₂ levels at bedside and the consequence of high levels of oxygen exposure due to the production of reactive oxygen species (ROS) within the respiratory chain in the mitochondria. In the clinical setting, it is almost impossible to obtain ROS data, whereas PaO₂ is readily measured. Previous studies have shown that a clinically tolerable FiO₂ (0.6) can trigger increased rates of mitochondrial superoxide anion (O₂^–) production and release of hydrogen peroxide (H₂O₂) from lung mitochondria, suggesting that FiO₂ >0.6 exceeds a threshold of mitochondrial antioxidant defenses (17). In ARDS, this consequence can be amplified, as ROS production is already higher due to the lung inflammatory process (18).

In ARDS, hypoxemia can be caused by the alveolar component of the shunt coming from non-ventilated and/or hypoventilated alveoli, the so-called “right-to-left shunt” (19). Non-ventilated areas are unresponsive to higher FiO₂, while hypoventilated areas will increase their P_AO₂ levels and could inhibit hypoxic pulmonary vasoconstriction, further impairing blood-gas exchange. This phenomenon has been evaluated using mathematical models (20) and observed in experimental (21) and clinical studies (22), and is proportional to the degree of shunt. In the setting of high FiO₂ values, oxygen toxicity may emerge as resorption atelectasis (23), impairment of airway smooth muscle cell proliferation (24), induction of bronchopulmonary dysplasia (25), and intense extracellular matrix (ECM) deposition (26).

It is imperative that we move towards large, well-powered and designed, pathology-targeted, randomized controlled trials to better define the role of different FiO₂ and PaO₂ targets on relevant clinical outcomes. The findings of such trials may then be further analyzed in individual-data meta-analyses to investigate additional effects and secondary outcomes.

As pointed out by Aggarwal et al. (14), training physicians to better control oxygen levels with more rigorous titration is feasible, but it will be a slow process. Nevertheless, the recent study by the same authors (3) is a valiant effort to move forward and overcome clinicians’ inertia to changing FiO₂ toward lower levels. This change in practice may affect the outcomes of ARDS patients.

Acknowledgements

Brazilian Council for Scientific and Technological Development (CNPq), Carlos Chagas Filho Rio de Janeiro State Research Foundation (FAPERJ), Department of Science and Technology (DECIT)/Brazilian Ministry of Health, and Coordination for the Improvement of Higher Level Personnel (CAPES).

Footnote

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. Rocco PR, Pelosi P. Pulmonary and extrapulmonary acute respiratory distress syndrome: myth or reality? Curr Opin Crit Care 2008;14:50-5. [Crossref] [PubMed]
2. 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 2017;196:964-84. [Crossref] [PubMed]
3. Aggarwal NR, Brower RG, Hager DN, et al. Oxygen Exposure Resulting in Arterial Oxygen Tensions Above the Protocol Goal Was Associated With Worse Clinical Outcomes in Acute Respiratory Distress Syndrome. Crit Care Med 2018;46:517-24. [Crossref] [PubMed]
4. Panwar R, Hardie M, Bellomo R, et al. Conservative versus Liberal Oxygenation Targets for Mechanically Ventilated Patients. A Pilot Multicenter Randomized Controlled Trial. Am J Respir Crit Care Med 2016;193:43-51. [Crossref] [PubMed]
5. Suzuki S, Eastwood GM, Glassford NJ, et al. Conservative oxygen therapy in mechanically ventilated patients: a pilot before-and-after trial. Crit Care Med 2014;42:1414-22. [Crossref] [PubMed]
6. Allegranzi B, Zayed B, Bischoff P, et al. New WHO recommendations on intraoperative and postoperative measures for surgical site infection prevention: an evidence-based global perspective. Lancet Infect Dis 2016;16:e288-303. [Crossref] [PubMed]
7. O'Driscoll BR, Howard LS, Earis J, et al. BTS guideline for oxygen use in adults in healthcare and emergency settings. Thorax 2017;72:ii1-90. [Crossref] [PubMed]
8. Bellani G, Laffey JG, Pham T, et al. Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. JAMA 2016;315:788-800. [Crossref] [PubMed]
9. Chu DK, Kim LH, Young PJ, et al. Mortality and morbidity in acutely ill adults treated with liberal versus conservative oxygen therapy (IOTA): a systematic review and meta-analysis. Lancet 2018;391:1693-705. [Crossref] [PubMed]
10. Britos M, Smoot E, Liu KD, et al. The value of positive end-expiratory pressure and Fio(2) criteria in the definition of the acute respiratory distress syndrome. Crit Care Med 2011;39:2025-30. [Crossref] [PubMed]
11. Villar J, Blanco J, del Campo R, et al. Assessment of PaO(2)/FiO(2) for stratification of patients with moderate and severe acute respiratory distress syndrome. BMJ open 2015;5. [Crossref] [PubMed]
12. 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;101:392-7. [Crossref] [PubMed]
13. Comroe JH, Dripps RD, Dumke PR, et al. Oxygen toxicity—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 feet. JAMA 1945;128:710-7. [Crossref]
14. Aggarwal NR, Brower RG. Targeting normoxemia in acute respiratory distress syndrome may cause worse short-term outcomes because of oxygen toxicity. Ann Am Thorac Soc 2014;11:1449-53. [Crossref] [PubMed]
15. de Jonge E, Peelen L, Keijzers PJ, et al. Association between administered oxygen, arterial partial oxygen pressure and mortality in mechanically ventilated intensive care unit patients. Crit Care 2008;12:R156. [Crossref] [PubMed]
16. Eastwood G, Bellomo R, Bailey M, et al. Arterial oxygen tension and mortality in mechanically ventilated patients. Intensive Care Med 2012;38:91-8. [Crossref] [PubMed]
17. Turrens JF. Mitochondrial formation of reactive oxygen species. J Physiol 2003;552:335-44. [Crossref] [PubMed]
18. Chabot F, Mitchell JA, Gutteridge JM, et al. Reactive oxygen species in acute lung injury. Eur Respir J 1998;11:745-57. [PubMed]
19. Riley RL, Cournand A. Ideal alveolar air and the analysis of ventilation-perfusion relationships in the lungs. J Appl Physiol 1949;1:825-47. [Crossref] [PubMed]
20. Kjaergaard S, Rees SE, Nielsen JA, et al. Modelling of hypoxaemia after gynaecological laparotomy. Acta Anaesthesiol Scand 2001;45:349-56. [Crossref] [PubMed]
21. Okamoto T, Wheeler D, Liu Q, et al. Variability in Pressure of Arterial Oxygen to Fractional Inspired Oxygen Concentration Ratio During Cellular Ex Vivo Lung Perfusion: Implication for Decision Making. Transplantation 2015;99:2504-13. [Crossref] [PubMed]
22. Karbing DS, Kjaergaard S, Smith BW, et al. Variation in the PaO2/FiO2 ratio with FiO2: mathematical and experimental description, and clinical relevance. Crit Care 2007;11:R118. [Crossref] [PubMed]
23. Hedenstierna G, Edmark L. Mechanisms of atelectasis in the perioperative period. Best Pract Res Clin Anaesthesiol 2010;24:157-69. [Crossref] [PubMed]
24. Hartman WR, Smelter DF, Sathish V, et al. Oxygen dose responsiveness of human fetal airway smooth muscle cells. Am J Physiol Lung Cell Mol Physiol 2012;303:L711-9. [Crossref] [PubMed]
25. Shahzad T, Radajewski S, Chao CM, et al. Pathogenesis of bronchopulmonary dysplasia: when inflammation meets organ development. Mol Cell Pediatr 2016;3:23. [Crossref] [PubMed]
26. Vogel ER, Britt RD Jr, Faksh A, et al. Moderate hyperoxia induces extracellular matrix remodeling by human fetal airway smooth muscle cells. Pediatr Res 2017;81:376-83. [Crossref] [PubMed]