Since the inception of the journal Inhalation Toxicology, the editors have attracted some of the best basic science available in the field of respiratory response to airborne substances. It is important that articles undergo peer review. In our peer-review process we must judge whether a manuscript will be a useful contribution to solving important scientific problems. In some cases the expert reviewers and the authors cannot reach a consensus on the interpretation, conclusions, and contribution of certain papers. Often these papers must be rejected, but in most cases the authors and reviewers reach a compromise and appropriate revisions are made. As stated in a British medical journal, “We are in the debate business.” We must judge whether a paper will be a useful contribution to the scientific community. It is possible that reviewers and authors may disagree. That is science. The journal has an obligation to report new data that warrant an open review by the scientific community.
In the preceding article the expert reviewers and authors agreed on many revisions improving the quality and clarity of the manuscript, but there were two specific areas in the manuscript where there was substantial disagreement, and since, in my view, this article may have significant interest for our readers, I have decided to publish the article, but also to provide the readers with the major concerns and questions raised by the expert reviewers assigned to this article. I have asked the author to respond to these questions so that the scientific community can judge the paper for itself.
REVIEWER'S QUESTION
“The use of CFD in the Weibel lung model is not justified because this geometry is highly idealized. CFD analysis in the Weibel model maybe used to examine mechanisms but certainly not for prediction purposes. Lung compliance can not be ruled out as it is the sole mechanism for the airflow in the lung. A noncompliant lung means rigid airways and hence no airflow through the lung making this study irrelevant. For a symmetric geometry one can assume flow partitioning based on cross-sectional area (i.e., rigid airway model); however, the model here introduces a leak at each generation. The author needs to examine this effect more closely. My point is precisely about the flow partitioning and not central wall rigidness. Lung compliance determines what faction of flow go to each airway. Realistic inlet and outlet boundary conditions are required before any CFD attempt is justified. The solution of airflow in the idealized geometry may be numerically correct but physiologically unrealistic.
AUTHOR'S RESPONSE TO REVIEWER'S QUESTION
In the region of central airways, the walls of the lung do not expand (in both length and diameter; Calay et al. (2002) during the breathing cycle because of the larger diameter of the airways. Hence, the lungs are not compliant in this region. In this article, the geometry from central airways is considered, and hence, flow partitioning based on cross-sectional area is justified and physiologically realistic.
REVIEWER'S QUESTION
The original submission implied the values for mass transfer coefficients were applicable to formaldehyde. When challenged on the choice of values for the mass transfer coefficients compared to the values used by Kimbell et al. (Toxicol. Sci. 64:111–121, 2001) that were derived to match experimentally determined uptake of formaldehyde, the author contended that (1) the Kimbell study was not relevant since that study was for steady-state inspiratory flow and the author's work is for steady-state expiratory flow, and (2) Kimbell studied the upper respiratory tract and the author is studying the lower respiratory tract. The author did not respond to the reviewer pointing out that the respiratory epithelium of the nose and the epithelium lining the conducting airways are very comparable relative to types of cells present although the percentage of cell types in an airway can vary (see Mercer et al. in Am. J. Respir. Cell Mol. Biol. 10:613–624, 1994). Comments about formaldehyde were removed from the revised article but there was no change in the choice of values for mass transfer coefficients.
The percentages of the cells in a given airway change slightly but certainly not enough to warrant the mass transfer coefficients used by the author that are more than two orders of magnitude lower than those derived by Kimbell et al. based on experimental data.
Can the author provide a reference to support that mass transfer coefficients (both overall and at the wall) vary by more than two orders of magnitude for the kinds of flows used by Kimbell and those used by the author? Alternatively, can the author explain that the approach (and correspondingly some coefficients and other variables) used by Kimbell differs from the approach taken by the author?
The liquid side resistance should be the same whether inspiration or expiration is occurring. Any difference in mass transfer coefficient values between inspiration and expiration must be due to the gas-phase kinematics, which would be influenced by the gas flow patterns. Thus, for the author to defend the mass transfer coefficients that were used over the coefficients developed by Kimbell et al., a dramatic difference in the gas-phase coefficients must be shown between inhalation and exhalation. This raises the question of whether the author is reluctant to address this because it would involve more computations and more effort.
For a modeling article, the author has to defend the choice of values for all variables and parameters. This has not been adequately done for the choice of mass transfer coefficients.
If the author does additional work using more defensible mass transfer coefficients and addresses the sensitivity of the results to the choice of mass transfer coefficients, the article would represent a valuable contribution to the literature on models for reactive gas transport in the lungs. Just showing ASPM can be used instead of CFDM is not sufficient.
AUTHOR'S RESPONSE TO REVIEWER'S QUESTION
This study is not really about a particular chemical but to show a trend and demonstrate the usefulness of ASPM. Hence, the values of the wall mass transfer coefficients were retained to demonstrate the trend. I agree with the reviewer that the respiratory epithelium of the nose and the epithelium lining the conducting airways are very comparable. But the purpose of the article is to demonstrate the trend and not to do any quantitative comparisons for a particular chemical. Using the values of mass transfer coefficient from Kimbell et al. (Toxicol. Sci. 64:111–121, 2001) does not change the conclusions of the paper. Moreover, Kimbell's study was for the upper respiratory tract (URT) and the present study is for the lower respiratory tract (LRT). I am not referring to formaldehyde used by Kimbell et al. and hence, quantitative comparison is irrelevant. The flow patterns will be very different in the URT (Kimbell et al.) and LRT (present study) and between inhalation (Kimbell et al.) and exhalation (present study) so I cannot compare the overall mass transfer coefficients even if I use the same wall mass transfer coefficients as Kimbell et al.; also, this involves more computations and effort for no new information. I defended the geometry and flow parameters used in Sections 2 and 3. For the choice of mass transfer coefficient, I am only presenting a trend without any particular reference to any chemical. Using more defensible mass transfer coefficients does not change the conclusions of the paper. Sensitivity of the results to the choice of mass transfer coefficients is already shown in Figures 6 and 7.