To the Editor:
We read with interest the study by Easley et al. 1comparing changes in the extravascular lung water (EVLW), as measured by transpulmonary thermodilution (TPT), with changes in the lung tissue density by computed tomography (CT) in an acute lung injury model before and after endotoxin (lipopolysaccharide) administration and the accompanying editorial by Costa and Vidal Melo.2Although the authors used a reasonable animal model in a well-conducted study, we find significant limitations in data interpretation and a major fault with their conclusions. The study suffers from a small sample size (n = 5), making comparisons between CT-tissue quantification of lung edema and EVLW by TPT (EVLWTPT) difficult. A single EVLWTPToutlier1(fig. 3b, page 1070) seems responsible for most of the differences between the two techniques. However, even when including the outlier, there does not seem to be significant differences in EVLW values as measured by the two methods either after lung lavage or after intravenous lipopolysaccharide. After lung lavage, EVLW by CT was approximately 24 ml/kg versus 23 ml/kg for EVLWTPT(P = 0.1), and after lipopolysaccharide, EVLW by CT was 26 ml/kg versus 29 ml/kg for EVLWTPT(P = 0.2). Furthermore, CT methods for determining EVLW in acute lung injury are very complex and have not been substantiated enough to be considered an accepted standard, as has been pointed out in the editorial.2Moreover, the authors have obtained perfusion images at a single location in the lung base, excluding the upper lung regions where increased perfusion may have resulted in an increase in the microvascular surface area for fluid exchange and could have increased EVLW significantly. Clearly, the study would have been strengthened had gravimetric determination of EVLW been done instead of relying on the CT.
It is well established that lipopolysaccharide causes a rapid increase in capillary permeability and pulmonary recruitment of inflammatory cells, and its administration has been shown to increase EVLW in several animal models. Such an increase was seen by the TPT method but not by the CT. Had the authors controlled for the effects of lipopolysaccharide on EVLW alone, we may have been better able to determine the sensitivity of the two methods for detecting changes in EVLW with changes in V/Q matching and perfusion after lipopolysaccharide administration. As the authors have so eloquently pointed out, understanding the limitations of any device and having as thorough an understanding as possible of the effects changes in physiology have on its accuracy and interpretation are vital for meaningful clinical application. We cannot agree more, and yet, it is doubtful that this study defines the limitations of TPT determinations of EVLW in acute lung injury when pulmonary perfusion is changed. In fact, another equally valid conclusion would be that the TPT method is at least equivalent if not superior to the CT method in this model.
The accompanying editorial appropriately calls into question our current method of introducing medical devices to the market without rigorous scrutiny of efficacy. But TPT has been compared with both the accepted standard gravimetric and dual dilution techniques in a variety of disease states and has performed well.3–5. Furthermore, EVLWTPTis the best pulmonary-specific indice of disease severity and predictor of outcome available to us.6–7Very importantly, EVLWTPT-guided management of hemodynamics has been shown to decrease mortality in acute lung injury.8We believe that the foundation for clinical use of EVLWTPThas been established by these studies. We would, therefore, like to join with the authors of the current study and the accompanying editorial and now call for large prospective interventional investigations to examine the benefit.
*Oregon Health and Sciences University, Portland, Oregon. firstname.lastname@example.org