This Editorial View accompanies the following article: Nielsen VG, Baird MS, McAdams ML, Freeman BA: Desflurane increases pulmonary alveolar‐capillary membrane permeability after aortic occlusion‐reperfusion in rabbits: Evidence of oxidant‐mediated lung injury. Anesthesiology 1998;88:1524–34.
THE halogenated anesthetics affect lung function at a physiologic level as well as on a cellular level. It appears that at least some of the effects of the halogenated agents may be modified by underlying disease processes affecting the lung. An example of this phenomenon is that the bronchodilatory effects of the halogenated anesthetics are modulated by the state of the lung epithelium. The bronchodilatory effects of halothane appear to be enhanced when the epithelium of the airways is intact (not denuded). 
More recently, the bronchodilatory effects of sevoflurane and desflurane have been shown to be partially dependent on an intact airway epithelium, 
suggesting that their use in diseases where the epithelium is abnormal (i.e., asthma) might not lead to maximal bronchodilation. Both these studies suggest that production of nitric oxide (NO) by intact epithelium is required for inhaled anesthetics to be effective bronchodilators. In the investigation by Nielsen et al., 
the protein permeability of the alveolar capillary membrane was shown to be increased by desflurane administration to animals that had undergone aortic occlusion and reperfusion (oxidant‐stress injury). This increased permeability to protein did not occur in normal, noninjured animals and occurred to a much smaller degree in oxidant‐stress ‐ injured animals anesthetized with fentanyl and droperidol.
What is the importance of this investigation for clinicians? Should we now avoid the administration of volatile anesthetics to patients who undergo oxidant‐stress injuries, including patients who undergo solid organ transplantations, corrections of vascular occlusion or vascular repairs, or cardiac bypass? That cannot be concluded from the results of this investigation. Acute lung injury is defined clinically as bilateral infiltrates on chest radiograph, defective oxygenation measured as a PaO (2/FI)O2 ratio less than 300 and no evidence of congestive heart failure causing hydrostatic edema. The presence of increased protein permeability alone does not define lung injury. However, the etiology of pulmonary edema (hydrostatic vs. permeability) can be defined by measuring concentration of protein in the edema fluid relative to plasma total protein; a ratio greater than 0.75 suggests exudation and increased permeability. The investigators demonstrated that despite the increased protein permeability of the alveolar capillary membrane, other measurements commonly used to demonstrate lung injury were not increased by desflurane. For example, the extravascular lung water or edema of the oxidant‐injured lungs was not different in the animals exposed to desflurane compared with the edema in the animals given fentanyl and droperidol. Further, the arterial oxygen levels were not different in the two groups of animals, suggesting that the quantity of alveolar edema was not different in the two groups of animals. To summarize these findings, desflurane appeared to selectively increase one commonly used parameter of lung injury: the measurement of protein permeability of the alveolar‐capillary membrane.
What is the scientific basis for these findings? Normally, Na/K/ATPases in the alveolar epithelium keep the airspaces from becoming edematous by the active transport of sodium across the epithelium and into the lung interstitium, with water following passively. Once water passes into the interstitium, it follows a hydrostatic pressure gradient toward the lung hilum, where it is removed by lymphatics. Alveolar epithelial fluid clearance may be measured in vivo to assess the functional integrity of the alveolar epithelium. 
Volatile anesthetics have recently been shown to decrease the normal alveolar epithelial fluid clearance in experimental animals, 
and the ability to clear excess alveolar epithelial fluid has been shown to correlate with survival in humans. 
One could imagine that the administration of a volatile anesthetic to a patient who already had decreased alveolar epithelial fluid clearance could lead to a decrement in this lung function, as well as perhaps to an increase in the alveolar epithelial protein permeability. Another possibility is that the lungs exposed to desflurane had increased permeability to both protein and water, but active transport of sodium across the epithelium was stimulated by the sympathomimetic action of desflurane, resulting in removal of water faster than protein and concentration of the remaining protein in the alveolus. Anesthesia with fentanyl and droperidol is sympatholytic and would not stimulate transport. Alternatively, limitation of antioxidant defenses by halogenated agents, as demonstrated in this study, might result in increased lipid peroxidation and subsequent increased permeability to protein without significant increases in edema.
The implication of this investigation is that the administration of desflurane to oxidant‐stress ‐ injured lungs caused mild lung injury, whereas the administration of fentanyl and droperidol caused much less. There is no reason to believe desflurane is a unique volatile anesthetic in causing this injury; all the halogenated anesthetics may cause similar injury in oxidant‐injured lungs. There is clearly a need for further investigations; the administration of volatile anesthetics might cause more extensive permeability changes or further lung injury if the preexisting oxidant injury had been more extensive or in other kinds of disease processes affecting the lung.
There is insufficient experimental or clinical data at this time to justify avoiding halogenated anesthetics in patients at risk for ischemia‐reperfusion injury. However, it is clear that halogenated anesthetics may interfere with the normal lung functions that play a critical role in pulmonary edema clearance 
and that in oxidant‐injured lungs, the halogenated anesthetics may increase the alveolar barrier permeability. 
The importance of these findings is that we may have to change our clinical practice as new information becomes available and that we may be able to define certain coexisting diseases where the administration of a volatile anesthetic is not the best choice.
Jeanine P. Wiener‐Kronish, M.D.
Professor of Anesthesia and Medicine; Vice‐Chairman, Department of Anesthesia
Michael A. Gropper, M.D., Ph.D.
Assistant Professor of Anesthesia and Physiology; University of California, San Francisco; San Francisco, California; firstname.lastname@example.org; email@example.com
1. Park KW, Dai HB, Lowenstein E, et al: Isoflurane and halothane-mediated dilatation of distal bronchi in the rat depends on the epithelium. Anesthesiology 1997; 86:1078-87
2. Park KW, Dai HB, Lowenstein E, et al: Epithelial dependence of the bronchodilatory effect of sevoflurane and desflurane in rat distal bronchi. Anesth Analg 1998; 86:646-51
3. Nielsen VG, Baird MS, McAdams ML, et al: Desflurane increases pulmonary alveolar-capillary membrane permeability after aortic occlusion-reperfusion in rabbits: Evidence of oxidant-mediated lung injury. Anesthesiology 1998; 88:1524-34
4. Rezaiguia-Delclaux S, Jayr C, Luo DF, et al: Halothane and isoflurane decrease alveolar epithelial fluid clearance in rats. Anesthesiology 1998; 88:in press.
5. Wiener-Kronish JP, Albertine KH, Matthay MA: Differential effects of E. coli endotoxin on the lung endothelial and epithelial barriers of the lung. J Clin Invest 1991; 88:864-75
6. Matthay MA, Wiener-Kronish JP: Intact epithelial barrier function is critical for the resolution of alveolar edema in man. Am Rev Resp Dis 1990; 142:1250-7
© 1998 American Society of Anesthesiologists, Inc.