© 2003 American Thoracic Society
Ventilator-induced Lung Injury Occurs in Rats, but Does It Occur in Humans?Department of Medicine, UCLA School of Medicine, Los Angeles, California Correspondence and requests for reprints should be addressed to Donald F. Tierney, Emeritus Professor of Medicine, UCLA School of Medicine, UCLA, Los Angeles, CA 90095-1690. E-mail: tierney{at}ucla.edu Many research ideas don't live up to our expectations. However, the research ideas Herb Webb and I tested by ventilating rats with high pressures ended up being worth more than we expected (American Review of Respiratory Disease 1974;110:556–565). It was as if we violated a thermodynamic law and got more out of it than we put into it. Initially Herb and I thought that it would be hard to detect any effects in lungs ventilated with such pressures. It turned out that the changes were drastic and lethal in a few minutes. Before we began this project, we were an unlikely team. I had studied pulmonary surfactant for a decade, but my most active research interests were lung metabolism and protection from oxygen toxicity. Herb had done little research and wanted a clinically relevant research project during his fellowship. He had great intellectual curiosity and eagerly applied physiology to patient care. Together we wondered whether ventilating patients with high pressures might injure the lung. Even before working with Herb, I had found it worthwhile and fun to spend time with bright and intellectually curious postdoctoral fellows. For example, my interest in studying the effects of ventilation on the lung probably began shortly after Steve Young arrived at UCLA Harbor General Hospital as a postdoctoral fellow and we spent many hours speculating about what happens to pulmonary surfactant after it had been secreted. We had very little information then about how quickly surface-active material might appear on the alveolar surface, so we decided to label newly synthesized lipids in rat surfactant and determine when the label appeared in alveolar lavage fluid. We also thought about how quickly surfactant might leave the surface and the possibility that the available surfactant might be depleted if secretion lagged behind its loss from the surface. What factors might change the rate of surfactant loss from the surface? The idea that different modes of ventilation might affect the surfactant differently was likely because two studies using excised animal lungs showed changes of the pressure–volume curves after ventilation. So when Herb wanted to evaluate the effects of ventilating rats with high pressures, I was ready. Some of my colleagues thought I had wandered far from my research interests and they asked, "What led you to do that study"? I answered that I enjoyed discussing research ideas with Herb and I wondered about the possible role of surfactant. As a pulmonologist in the Department of Medicine, I was searching for some clinical application of my interest in pulmonary surfactant. Surfactant research had focused on premature infants who developed the respiratory distress syndrome, but I searched for an application that would apply to adults. For me, this project included a possible role for surfactant in diseases of adults. Initially, Herb and I were aware of animal studies using low inspiratory pressures with small tidal volumes that did not cause injury. However, we considered that patients with extensively damaged lungs, such as in those with the adult respiratory distress syndrome (ARDS), required higher pressures, which we thought might damage any normal airspaces. Although such high pressures might cause injury by several mechanisms, we were particularly interested in the possibility that surfactant depletion might occur in the surface film. Lipids, such as those in lung surfactant, form a layer only one molecule thick between air and liquid, and when the area of the surface decreases, these molecules can be forced together until some pop out of the surface. We now know that such "used" lipids don't respread rapidly onto the surface, although newly secreted surfactant will spread. As a consequence, surfactant in the surface may be squeezed out at low lung volumes when the surface area decreases, and then newly secreted surfactant would be required to replace it. During inflation, when gaps between surfactant molecules occur as the surface area increases, newly secreted surfactant fills those gaps. During expiration, surfactant molecules leave the surface but they would normally be replaced during inspiration with most breathing patterns. What would happen if the airspaces went through very large volume and surface area changes for a prolonged period? The greater the decrease of surface area during expiration, the more molecules would leave and need to be replaced by newly secreted surfactant. The combination of rapid deep breaths and low distending pressures, such as occurs in recumbent patients without added end-expiratory pressure, would be the most likely to consume surfactant in this manner. Surfactant molecules might be pumped into and then out of the surface rapidly and exceed the secretion of surfactant onto the surface. At least that was our hypothesis: that this ventilatory pattern might consume more surfactant than could be secreted. If surfactant were depleted, or inactivated by plasma leaking onto the alveolar surface, both edema and airspace collapse would be promoted. Herb anticipated many issues. He carefully controlled the anesthesia, added dead space to the tubing when needed to maintain relatively normal blood gases, fixed the lungs for microscopic examination, and judged the degree of edema without knowledge of the experimental conditions. When he first ventilated anesthetized rats at peak pressures of 45 cm H2O with no added end-expiratory pressures, we could barely believe the results! Even during our most exaggerated discussions of "what if" we hadn't anticipated such dramatic changes. Within minutes the rats were cyanotic and appeared moribund; their lungs were very heavy and edematous, with foam in the airways. We were surprised because we knew that humans ventilated with similar pressures did not develop overwhelming changes so quickly. It was like a mystery story when the sleuth suspects a culprit who seems to have a powerful alibi! We knew that human lungs weren't injured to the extent rat lungs were, but were human lungs likely to be injured at all from those pressures? Herb then ventilated rats with the same high pressure of 45 cm H2O but added an end-expiratory pressure of 10 cm H2O, as might be done to increase oxygenation of patients with ARDS. Once again, the results were striking but this time because there was much less lung edema. So we realized that lung injury with 45 cm H2O was not solely due to high inflation pressures. More edema occurred when the lung was smallest at the end of expiration. How could increased end-expiratory pressure attenuate the effect of high inspiratory pressure? Our best explanation remained that pulmonary surfactant might be lost rapidly from the surface while the volume was lowest, which would be when no positive end-expiratory pressure was applied. The idea that surfactant can be depleted by the combination of high inspiratory and low expiratory pressures was consistent with some earlier studies done with excised lungs. In those studies, lung compliance progressively decreased with ventilation and the lower the end-expiratory pressure, the greater the effect. Furthermore, the compliance recovered if the lungs were not ventilated for a while, suggesting that surfactant secretion lagged behind demand. Despite all these considerations, we were still faced with the facts that many humans had been ventilated using high pressures and that their lungs were not injured to the degree that we saw in these rats. We considered some quantitative differences between species that might indicate greater sensitivity of rat lungs. To the extent rat lungs have smaller airspaces than humans at the end of expiration, they would be expected to lose surfactant from the surface more rapidly because the surface tension would need to be lower. Rats do have smaller airspaces and their chest walls are highly compliant, which may allow their lungs to collapse relatively more at the end of expiration. Therefore, if these concepts of surfactant depletion apply to the human lung, it too could be injured, but less rapidly than the small rat lung. It took a decade or two for others to conclude that human lungs could be injured by such ventilation. When is it reasonable to treat humans using results obtained with animals? Animal studies are used to determine the safety of some new procedures. Using rats, we had found a harmful effect of a procedure and decided that, although we would be cautious in extrapolating to humans, we would try to avoid unnecessary use of similar pressures in humans. These studies influenced our management of patients in a conservative manner. Rat lungs developed alveolar edema when ventilated at high inspiratory and low expiratory pressures and it seemed reasonable to avoid that combination in patients when feasible. Our final paragraph nearly 30 years ago suggested management similar to current recommendations using protective ventilation or low tidal volumes. Consistent with our initial hypothesis about loss of surfactant, we were most wary of conditions requiring high pressures on inspiration but associated with low lung volumes at end expiration, as in patients with the adult respiratory distress syndrome. In that setting, we tried to minimize high pressures if possible and to use positive end-expiratory pressures. That would result in low tidal volumes and be similar to recommendations made subsequently based upon human trials. We like to think that our work helped stimulate other investigators to do additional animal studies and eventually to study patients. However, this article seemed to interest few clinicians or investigators for a decade or more, perhaps because a similar degree of injury in patients was not apparent. Documenting such an effect in humans would take years of work and we did not follow this publication with additional research or presentations of our own. In retrospect, it seems almost irresponsible that we didn't publicize our concerns that such ventilatory patterns might be harmful to humans, but we were hesitant to extrapolate our studies too far. There is some irony and perhaps serendipity here. Although we probably would have done this study anyway, the postulate of surfactant depletion captured my attention and spurred me onward. But subsequently, few articles have discussed it. How important is the concept that surfactant might be inactivated or depleted? It is now clear that other mechanisms can injure the lung during ventilation with these pressures, but the possibility that surfactant depletion from the surface might be an initiating event could now be tested by instilling surfactant into animal lungs before ventilating. It might be worthwhile knowing if surfactant protected lungs from ventilator injury. For me, this project was a prime example of why I enjoy research; few activities in life are as exciting as testing an idea or finding something unexpected and pursuing it. The excitement of discovery and the collegiality of working with Herb, whose meticulous attention to detail and pursuit of physiological mechanisms were essential for success, more than justified the time and effort. It's particularly satisfying nearly 30 years later, when we don't recall false starts and blind alleys but only good fortune! This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
