© 2003 American Thoracic Society
Saving the Breath Condensate ApproachTo the Editor:I read with interest the editorial by Gaston (1) that suggests that breath condensate analysis might be "worth studying, after all." This hopeful viewpoint is based on the study of Corradi and colleagues (2), which showed that condensate concentrations of glutathione decreased in asthma exacerbations, whereas those of malondialdehyde increased. Gaston then questions the need for the "mathematical interpretation" in our earlier publication (3), which emphasized the necessity for a reference indicator of dilution. We would argue that without knowing anything about dilution of respiratory droplets in water vapor, the interpretation of condensate data must remain in serious doubt. That the condensate concentration decreased for glutathione and increased for malondialdehyde only tells us that the ratio of the two mediator concentrations has changed. If the dilution of the droplets is not known, concentrations of both mediators may have increased or decreased in the respiratory droplets. Had Corradi and colleagues (2) been able to measure both reduced and oxidized forms of glutathione, a reference indicator might not have been quite as important. An ideal reference indicator of dilution would have the same concentration in the plasma and respiratory droplets, and the dilution of the droplets would then be calculated from the dilution of this indicator in the condensate. Even if the respiratory and plasma concentrations are not the same but droplet concentrations remain constant, differences in the dilution of respiratory droplets between condensate samples could be detected by differences in the condensate concentrations of the reference indicator. We suggested that the sum of sodium and potassium in the respiratory fluid could be used to detect alterations in the dilution of respiratory droplets. As we indicated (4, 5), correlations between multiple ions are significant over small and large ranges, and contrary to Gaston's assertion (1), the accuracy and specificity of these measurements are much less problematic than those of the mediators or measurements of pH in extremely dilute solutions (comparable to rain water), which have been heavily contaminated by ammonia generated in the mouth (6). If the condensate method is to survive, efforts must be made to develop simple and reliable methods of detecting alterations in droplet dilution. Sodium plus potassium represented a suitable starting point.
Medical College of Wisconsin Milwaukee, Wisconsin REFERENCES
From the Author: Dr. Effros kindly provides an opportunity for me to reiterate several important truths. First, the ratio of one solute to another is not affected by the amount of solvent. Corradi and coauthors have shown that treatment of asthma results in a change in the relation between breath condensate concentrations of glutathione and malondialdehyde (1). This ratio of condensate solute concentrations is informative with regard to disease activity, regardless of the absolute concentration of these solutes in the airway lining fluid (ALF). Second, breath condensate concentration is not a direct measure of ALF concentration. Dr. Effros's program to develop a model by which one can derive ALF concentrations from breath condensate is admirable, if quixotic. May his voyage prosper! In the meantime, breath condensate should be studied for changes both in the absolute concentration of solutes—particularly protons—and in the ratio of solute concentrations: these measurements change with the activity of inflammatory diseases (2). Third, breath condensate acidification is not a reflection of oral ammonia concentration. Definitively, breath condensate pH measured before endotracheal intubation for elective surgery is identical to breath condensate pH subsequently measured through the endotracheal tube. Dr. Effros knows these data, but has chosen to ignore them. It is regrettable that he has brought up again his original condensate electrolyte measurements—and the accompanying pages of speculation—which he and others have shown to be incorrect (3–6). Indeed, Dr. Effros has proposed many hypotheses that are contradicted by a large body of data, including the following: that (1) condensate pH is determined primarily by oral ammonia (note in this regard that it is established that low breath condensate ammonia is necessary, but not sufficient, to explain low breath condensate pH in vivo [6]); (2) condensate electrolyte concentrations should be used to calculate ALF concentrations; (3) ALF must be extremely hypotonic (based exclusively on predictions from condensate urea concentrations); (4) ALF cannot contain ammonia; and (5) nothing of relevance or interest can be made of condensate concentrations unless they are precisely normalized to ALF concentrations. Let us be more inclined to ignore hypotheses that are not supported by data than to ignore data that are not supported by hypotheses.
University of Virginia School of Medicine Charlottesville, Virginia FOOTNOTES Conflict of Interest Statement: Dr. Gaston is a shareholder in Respiratory Research Incorporated. REFERENCES
From the Authors: We thank Dr. Effros for raising the important issue of dilution associated with the analysis of exhaled breath condensate (EBC). The reasoning behind Effros' proposal is that (1) EBC is a solution, (2) electrolytes are released at a constant rate in the vapor stream, and (3) the number of droplets is inversely proportional to the amount of EBC. Unfortunately, none of these assumptions is supported by data. However, we do acknowledge and we did write that an independent method of determining the dilution of airway lining fluid (ALF) by water vapor would be necessary to calculate EBC biomarker concentrations corresponding to those present in ALF (1). The ideal dilution factor should be an ALF component whose concentration (1) is known, (2) does not change in disease, and (3) is released at a constant rate. Unfortunately, none of these conditions is fulfilled by sodium and potassium concentrations, which have been proposed by Effros to normalize EBC droplet content. Moreover, (1) inconsistent ALF electrolyte concentrations have been reported; (2) ALF contains much less sodium and chlorine and much more potassium than extracellular fluid or plasma, but at the same time much less potassium than intracellular fluid; and (3) significant changes in ionic composition of diseased airways are apparent (2). The reason EBC electrolyte concentration is so low is unclear, but neither their intracellular nor their extracellular concentrations are reflected in EBC. A possible explanation is the mechanism of EBC formation, which in our opinion is both a solution and an aerosol. The mechanism leading volatile solutes, such as aldehdyes, to leave the lining ALF following the vapor stream seems to be very selective, resembling the process of water distillation, which is known to give rise to water almost ion-free. On the basis of these observations, the validity of any ion as a dilution marker for EBC is questionable. We do recognize that a way to normalize EBC data is necessary. One possible solution is the ratio between EBC components. Alternatively, EBC data could be normalized for the amount of exhaled air. Before reaching this objective, we believe that it is already possible to compare the same EBC markers in different groups and to study the profile of different substances with similar characteristics within the same individuals. This is especially true when data confirm expected changes in hypothesis-driven studies.
Department of Clinical Medicine Nephrology and Health Sciences Research Center of the National Institute of Occupational Safety and Prevention University of Parma Parma, Italy REFERENCES
This article has been cited by other articles:
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
