© 2003 American Thoracic Society
Lung Recruitment in Real TimeLearning Was Never So EasyDepartment of Critical Care and the Lung Biology Program Hospital for Sick Children and the Interdepartmental Division of Critical Care Medicine University of Toronto Toronto, Ontario, Canada In the current issue of AJRCCM (pp. 1620–1626), we read a paper by Halter and colleagues (1) providing video evidence that after surfactant inactivation, alveoli collapse in the absence of positive end-expiratory pressure, and remain collapsed—a state prevented by applying end-expiratory pressure. This simple, frankly obvious concept has been studied several times over and in the minds of many might be so evident as to not need demonstration. Why then is this manuscript noteworthy? I believe there are several lessons to learn. The first area to benefit is the study of atelectasis. The authors are correct when they point out that atelectasis at the alveolar level has not previously been demonstrated; this is because it was not directly possible. Despite brilliant theoretic modeling of dynamic lung structure (2), previous inferences about atelectasis were just that; they were inferences, not demonstrations. Therefore, in any given situation, one could not be sure about the exact nature of alveolar collapse (or its prevention) or whether the alveoli really were collapsed. In the context of evolving knowledge, an accurate picture is important, literally as well as metaphorically. The current demonstration is a reminder about what atelectasis and lung recruitment really are. This is important in the current era of mixed results from human studies addressing lung recruitment in acute respiratory distress syndrome (ARDS), given that its microanatomic basis is the subject of contemporary controversy (3). After all, future interventional strategies that are based on sound understanding of the mechanisms will in most circumstances provide the best approach to ultimately successful therapy. As in any area of life, this is easier when investigators can see what they are dealing with. The second area to gain from the work of Halter and coworkers (1) is enhanced integration of lung imaging as an accepted, and necessary, element of physiologic research. Medical imaging has come a long way since the introduction of the radiograph, and we have learned a great deal from benchmark imaging studies of lung volume alterations. For example, almost 30 years ago, Froese and Bryan used plain lateral radiographs to visualize the cephalad diaphragmatic shift after administration of neuromuscular blocking medications (4), thereby providing an explanation for hypoxemia and atelectasis associated with general anesthesia and surgery. Over a decade later, Gattinoni and colleagues used computed tomography scans to demonstrate that volume loss in ARDS generally occurred in the dependent lung regions and that application of positive end-expiratory pressure recruited—–in part—such areas (5, 6). Currently, xenon contrast–enhanced computed tomography can demonstrate regional ventilation (7), and magnetic resonance imaging is changing our understanding of atelectasis in the lungs of preterm infants (8). The approach used in the current study, intravital microscopy, is quite different and has provided important biological insights in the pulmonary circulation (9). Halter and coworkers applied the microscope lens to the surface of the pleura and essentially recorded a movie of the ongoing microscopic events—quite literally—to "focus" on the alveoli (1). These days of course, "functional imaging" (meaning that one can see what is happening, as it happens) is in vogue—and for very good reasons. An important review of positron emission tomography makes the critical point that "diseases are biological processes" (10). We would all do well to remember that and open our minds (and our journals) to acceptance of studies that show us the biological events directly. When applicable, this must be a preferred option over "representative" still photographs, or worse, the condensation of exciting biological events and processes into comparatively dull black and white two-dimensional graphs. Aside from better representation, the cost of printing individual color plates in paper manuscripts is extremely high; in contrast, posting a video on the website is more informative and is free of charge! So much for the contributions—are there important design or methodologic concerns? Yes, the paper by Halter and coworkers (1) is not perfect (what paper is?). The number of laboratory animals is very small, and there is definite species specificity; the interventions were extremely short-term, and surfactant inactivation alone—not histologically demonstrated injury—was established. The residual oxygenation deficit, despite "visible" recruitment of alveolar volume, advances the authors' suggestion that the images may not be representative of all lung regions. In addition, the experimental details (ironically, for a paper making extensive use of the web-based supplement for pictorial reference) are brief, whereas they had the capacity to be virtually unlimited. For example, after the recruitment maneuver, was the distending pressure removed and the lungs allowed to collapse before application of end-expiratory pressure (effectively rerecruitment), or was the distending pressure lowered to the chosen level of end-expiratory pressure (maintenance of recruitment)? Whereas "readers"' of the paper will have some difficulty figuring this out, "viewers" will observe that the latter approach was taken. The point illustrates two issues. First, details are important when describing lung recruitment: maintenance is easier than regaining lost ground once the lung has collapsed (11, 12). Second, a picture (especially a movie!) may provide far more useful details than conventional scientific text. In its presentation of real-time images, the paper by Halter and coworkers (1) is similar to recent publications in this (13) and other (14) journals, where online real-time images are available to the reader, enhancing the reading experience and offering major advantages over standard text. We need to recognize what others are rapidly learning. Computer-based video systems improve performance in many areas, from the surgeon's dissection (15) to the golfer's swing (16). Does this development herald the onset of a publication fad characterized by "pretty pictures but poor science?" No it does not; high-impact journals will never accept pictorial images as a substitute for methodologic and analytic rigor; however, they will encourage investigators to submit truly representative images—without limitations of space or cost—that are real time, obvious, and convincing to all. In the case of Halter and coworkers (1), their data are refreshingly clear: we do not have to second guess–derived graphs, statistics, or meta-analyses—we just have to look! Acknowledgments The author thanks Dr. John Boylan for helpful discussions. REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
