Busting the Myths of DLco for Pulmonary Trainees: Isolated Reductions in DLco and the Relationship with VA

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This topic is important.  While I agree with most of the discussion, some improvements needed.

You label both the graph and algorithm as Figure 2 leading to some confusion.

You should discuss that the measured DLCO, the uncorrected predicted DLCO and corrected (for hemoglobin and if available CO and MetHb) predicted DLCO along with %predicted should be reported.    If the reason for low DLCO is anemia, then the corrected %predicted DLCO would be normal. 

Mathematically KCO = DLCO/VA where VA is the volume of distribution of the tracer gas minus estimated dead space.  The discussion of using KCO versus DLCO/VA should reflect that from standard 2017. From standards 2017 “Mathematically, KCO can be calculated as DLCO/VA BTPS. However, KCO should not be reported using the term DLCO/VA, as it may be inferred from this term that DLCO can be corrected or normalised for VA. “  Standards 2017 goes on to say . “ In fact, the relationship between lung volume and carbon monoxide uptake is complex and studies evaluating the effects of reduced VI (and thus VA) show the relationship to be alinear and certainly less than 1:1  (i.e. the fall in DLCO is far less than the fall in lung volume).”

While you cite reference 35 (standards 2017) as showing how DLCO and KCO change with VA, reference 35 does not include that.  The equations used to show how DLCO and KCO change with VA are from the study of Johnson (your reference 32).  These equations closely match what would be expected if the reduction in DLCO at lower lung volumes is primarily due to lower membrane diffusion with pulmonary capillary blood volume remaining relatively constant Those equations were presumably used to generate your Figure 2 graph.  You should cite that reference (32) and generate an identical graph from those equations.   Those equations are discussed in the standards 2005 (your reference 2) as follows:

“In normal subjects with experimental reductions in VI (and, thus, VA), adjustment equations for this effect have been derived and a recent representative example consists of the following:

DLCO(at VAm)=DLCO(at VAp))×[0.58+0.42×(VAm/VAp)]

KCO(at VAm)=Kco(at VAp)×[0.42+0.58/(VAm/VAp)]           

where VAm represents measured VA and VAp represents predicted VA at normal TLC.”

There are some problems with your Figure 2 algorithm. 

1. VA is often normal in subjects with emphysema (see your reference 32) while TLC measured by plethysmography is usually elevated with emphysema.
2. Anemia as cause of low DLCO is assessed by reporting corrected (for Hb) %predicted DLCO which would be normal.
3. DLCO findings with pneumonectomy are much different than with failure to take a deep breath or neuromuscular disease.
4. While patients with low DLCO and normal lung volumes often have pulmonary vascular disease, combined emphysema and interstitial lung disease should be suspected if there is a heavy smoking history.
5. A more inclusive algorithm approach to reduced DLCO would include whether the DLCO is lower than expected for the lung volume by also reporting %predicted DACO – with predicted DLCO also adjusted for the lung volume.

DACO is normal when DLCO is low and KCO high due to low lung volume with normal lung parenchyma to failure to take a deep breath or neuromuscular disease.  However some patients with neuromuscular disease may also have atelectasis which would reduce DACO.  Patients with pneumonectomy have lower DLCO than expected just from the low lung volume so have reduced DACO.

I encourage you to review the approach to DLCO interpretation which includes DACO from the reference below, then adjust your discussion and algorithm.  This is discussed in reference below.

Improving pulmonary function test interpretation

Thomas P. Presti Douglas C. Johnson

European Respiratory Journal 2023 61(1): 2201858; DOI: https://doi.org/10.1183/13993003.01858-2022

Predicted D_LCO should be adjusted for haemoglobin and, if available, other factors including CO-Hb and Met-Hb. Just as anaemia affects D_LCO, lung volume also does. Measurements from the D_LCO manoeuvre (inspired and expired CO, inspired and expired tracer gas, inspired volume, system and patient dead space) directly allow calculation of D_LCO, transfer coefficient of the lung for carbon monoxide (K_CO), and of alveolar volume (V_A) [5]. As V_A becomes lower, D_LCO decreases and K_CO increases as expected due to lower surface area but higher surface area/volume ratio [6].

The proposed algorithm to interpret D_LCO has problems. While acknowledging that K_CO increases at lower lung volumes, the standard does not incorporate the predictable relationship of K_CO and D_LCO to lung volume [6, 7], that patients with interstitial lung disease (ILD) may have low, normal, or elevated K_CO, or that patients with low lung volume due to incomplete lung expansion would have normal D_LCO adjusted for lung volume. Predicted D_ACO, the predicted D_LCO for the patient's lung volume, should also be reported [8]. “Measured” D_ACO equals measured D_LCO. Rather than including K_CO in the algorithm to interpret D_LCO, the algorithm should include % predicted D_ACO, which equals % predicted K_ACO and is low in ILD [9].

We interpret D_LCO as follows. D_LCO and D_ACO should both have predicted values adjusted for haemoglobin (and if available CO-Hb and Met-Hb). If haemoglobin is not measured, a low D_LCO could be from anaemia and elevated D_LCO from polycythaemia.

• Normal D_LCO with normal D_ACO and V_A indicates normal diffusion.
• Low D_LCO with normal D_ACO indicates normal diffusion, but with low V_A and D_LCO reduced, and K_CO elevated as expected from the V_A. This indicates low lung volume with normal gas exchange (e.g. incomplete lung expansion from thoracic wall abnormality, pleural effusion, weakness).
• Low D_LCO with low D_ACO and normal V_A indicates abnormal diffusion without restriction. This can occur with pulmonary vascular disease or emphysema (in which V_A is usually normal and underestimates total lung capacity due to incomplete equilibration of the tracer gas throughout the lung). Significant airflow obstruction and elevated plethysmography lung volume are expected with emphysema. Normal spirometry and lung volumes are expected with pulmonary vascular disease.
• Low D_LCO with low V_A, elevated K_CO and mildly low D_ACO is expected with pneumonectomy.
• Low D_LCO with low D_ACO and low V_A indicates abnormal diffusion with D_LCO lower than expected and K_CO not as elevated as expected for the V_A. K_CO may be low, normal, or elevated. This typically occurs with ILD.
• Elevated D_LCO with elevated D_ACO and normal V_A can occur with asthmatics having airway obstruction at the time of testing, likely related to more negative intrathoracic pressure leading to increased pulmonary capillary volume.
• Elevated D_LCO with elevated D_ACO with reduced V_A can occur with alveolar haemorrhage.

Author Response

Comment: You label both the graph and algorithm as Figure 2 leading to some confusion.

Response: Thank you for pointing this out. Change has been made to reflect algorithms as "flowcharts" instead of figures. (Page 5 and 10)

Comment: You should discuss that the measured DLCO, the uncorrected predicted DLCO and corrected (for hemoglobin and if available CO and MetHb) predicted DLCO along with %predicted should be reported. If the reason for low DLCO is anemia, then the corrected %predicted DLCO would be normal. 

Response: We appreciate the comment. It was deliberately not touched because almost all the auto-generated reports regarding diffusion capacity automatically correct for diffusion capacity. Further, anemia (as part of sickle cell) has been briefly discussed on page 4. 

Comment: Mathematically KCO = DLCO/VA where VA is the volume of distribution of the tracer gas minus estimated dead space.  The discussion of using KCO versus DLCO/VA should reflect that from standard 2017. From standards 2017 “Mathematically, KCO can be calculated as DLCO/VA BTPS. However, KCO should not be reported using the term DLCO/VA, as it may be inferred from this term that DLCO can be corrected or normalised for VA. “  Standards 2017 goes on to say . “ In fact, the relationship between lung volume and carbon monoxide uptake is complex and studies evaluating the effects of reduced VI (and thus VA) show the relationship to be alinear and certainly less than 1:1  (i.e. the fall in DLCO is far less than the fall in lung volume).”

Response: Thank you for the comment and insight. You exactly pointed out what we are trying to put forward to the readers. This is a complicated relationship between three variables that remains an enigma even to the expert pulmonary folks. Thus to simplify and explain fully, we first explain the individual components (pages 7&8) and then carry forward this base to explain the relationship between them. Previously described literature on the topic, notably, bombard the reader with too much information in condense form, which makes it hard to understand a complex topic like this. 

Comment: While you cite reference 35 (standards 2017) as showing how DLCO and KCO change with VA, reference 35 does not include that.  The equations used to show how DLCO and KCO change with VA are from the study of Johnson (your reference 32).  These equations closely match what would be expected if the reduction in DLCO at lower lung volumes is primarily due to lower membrane diffusion with pulmonary capillary blood volume remaining relatively constant Those equations were presumably used to generate your Figure 2 graph.

Response: Thank you for pointing that out. This reference was corrected on review.

Comment: European Respiratory Journal 2023 61(1): 2201858;

DOI: https://doi.org/10.1183/13993003.01858-2022

Response: I appreciate that you shared this letter with us. DACO is a different approach that not many America plethysmograph software incorporates in their result report of DLco. We shaped this article to address the reported physiological measures on commonly used PFT reports and DLCO measurements. While DACO incorporation may seem to provide value to the clinician, it is practically hard to adopt a different approach while you are already familiar with one. Apart from the approach that we suggested, there are other approaches to read diffusion capacity of CO but our intention was to keep it simple to make it more adoptable across. Of course, it is with the understanding that actual DLCO interpretation in a complex patient requires incorporation of patient’s clinical, radiographic and physiological data, and mere presence of some numbers is not enough to label that patient with a diagnosis.

Reviewer 2 Report

Comments and Suggestions for Authors

This review article addresses an important and often misunderstood topic in pulmonary medicine—isolated reduction in DLco and its relationship with alveolar volume (VA) and the diffusion coefficient (kCO). The manuscript provides a comprehensive and well-organized physiological and clinical discussion that is valuable for pulmonary trainees and clinicians alike. The use of algorithms and figures to illustrate diagnostic approaches enhances its utility in clinical settings.

However, the manuscript would benefit from improvements in clarity, grammar, and flow.

Additionally, while the manuscript excels in depth, some paragraphs are dense and could benefit from subheadings or bullet points, especially in sections like the pathophysiological mechanisms or factors affecting kCO. This would enhance readability for trainees. Also, the authors might consider rephrasing complex sentences for simplicity and precision—for instance, simplifying the explanations in the “DLco and VA Relationship” and “Integrated Approach” sections could make the content more accessible to the target audience of junior clinicians.

It is also suggested that the authors clarify the term "DLco/VA" early in the manuscript to avoid misinterpretation. The effort to debunk the misconception that DLco/VA equals DLco “corrected” for lung volume is critical and well-placed, but the explanation can be tightened for impact.

In summary, this is a valuable educational article that needs minor revisions to language and format to reach its full potential.

Author Response

Comments: This review article addresses an important and often misunderstood topic in pulmonary medicine—isolated reduction in DLco and its relationship with alveolar volume (VA) and the diffusion coefficient (kCO). The manuscript provides a comprehensive and well-organized physiological and clinical discussion that is valuable for pulmonary trainees and clinicians alike. The use of algorithms and figures to illustrate diagnostic approaches enhances its utility in clinical settings.

However, the manuscript would benefit from improvements in clarity, grammar, and flow. Additionally, while the manuscript excels in depth, some paragraphs are dense and could benefit from subheadings or bullet points, especially in sections like the pathophysiological mechanisms or factors affecting kCO. This would enhance readability for trainees. Also, the authors might consider rephrasing complex sentences for simplicity and precision—for instance, simplifying the explanations in the “DLco and VA Relationship” and “Integrated Approach” sections could make the content more accessible to the target audience of junior clinicians. It is also suggested that the authors clarify the term DLco/VA early in the manuscript to avoid misinterpretation. The effort to debunk the misconception that DLco/VA equals DLco “corrected” for lung volume is critical and well-placed, but the explanation can be tightened for impact.

In summary, this is a valuable educational article that needs minor revisions to language and format to reach its full potential.

Response: Thank you, reviewer # 2 for in-depth review highlighting the importance of the article, and potential areas of improvement. I have revised the article to reflect on the changes suggested by the reviewer. It was completely intentional to introduce the full term and explanation of “DLco/VA” to the last, since we wanted the readers to understand the individual components properly, before going to their relationship. That way, we have tried to make this incredibly complex relationship as easy as possible. 

Reviewer 3 Report

Comments and Suggestions for Authors

The manuscript is interesting and, in general, fairly well-written. The paper addresses an important topic in pulmonary medicine - the proper interpretation of diffusion capacity (DLco) testing. The authors correctly identify that there is significant misunderstanding among trainees regarding the relationship between DLco, alveolar volume (VA), and the diffusion coefficient (kCO). The paper provides a useful discussion of the physiological basis for DLco testing and clearly explains why the common interpretation of DLco/VA as "DLco corrected for lung volume" is physiologically incorrect. The authors provide a thorough review of various clinical conditions associated with isolated reduction in DLco, which serves as valuable reference material for clinicians. The algorithmic approach presented in Figure 2 is particularly helpful as a clinical decision-making tool for interpreting reduced DLco values based on the relationships between DLco, VA, and kCO.

However, I still have some suggestions to further improve the quality of the manuscript.

I would like to suggest that the authors address these limitations in the article, either by discussing them in the limitations section or, where feasible, by making the appropriate revisions:

1. The authors state that they performed a "comprehensive (but nonsystematic) literature search." A more robust systematic review methodology would strengthen the paper's scientific rigor. Without a systematic approach, there's potential for selection bias in the included studies and references.

2. The paper would benefit from including more quantitative data to support its claims. For example, when discussing the prevalence of reduced DLco in different conditions, more specific statistical information comparing different patient populations would strengthen the arguments.

3. While the paper provides a good theoretical framework, it could offer more practical guidance on how to implement these concepts in daily clinical practice. For instance, specific case examples showing how misinterpretation of DLco relationships could lead to incorrect diagnoses would illustrate the practical importance of the topic.

4. The algorithm in Figure 2 is useful but could be more comprehensive. It doesn't fully address all the clinical scenarios mentioned in the text. Additionally, the relationship between the text explanation and the algorithm is not always clear.

5. There is limited discussion about the variability and reproducibility of DLco measurements, which is crucial for proper test interpretation. The paper doesn't adequately address how test variability might affect the interpretation of isolated DLco reduction.

6. While the paper mentions ATS/ERS standards, it doesn't thoroughly critique current clinical guidelines or provide specific recommendations for improving them based on the physiological principles discussed.

Author Response

Reviewer 3

Comments: The manuscript is interesting and, in general, fairly well-written. The paper addresses an important topic in pulmonary medicine - the proper interpretation of diffusion capacity (DLco) testing. The authors correctly identify that there is significant misunderstanding among trainees regarding the relationship between DLco, alveolar volume (VA), and the diffusion coefficient (kCO). The paper provides a useful discussion of the physiological basis for DLco testing and clearly explains why the common interpretation of DLco/VA as DLco corrected for alveolar volume; is physiologically incorrect. The authors provide a thorough review of various clinical conditions associated with isolated reduction in DLco, which serves as valuable reference material for clinicians. The algorithmic approach presented in Figure 2 is particularly helpful as a clinical decision-making tool for interpreting reduced DLco values based on the relationships between DLco, VA, and kCO.

Response: Thank you for encouraging words and support for the article. This, indeed, is a fairly common, yet misunderstood, topic amongst trainees and clinicians alike. Our effort is to simplify it for the readers to clear the misconception.

Comment: The authors state that they performed a “comprehensive (but nonsystematic) literature search” A more robust systematic review methodology would strengthen the paper’s scientific rigor. Without a systematic approach, there is a potential for selection bias in the included studies and references.

Response: Yes, we acknowledge that as a limitation of this paper.

Comment: While the paper provides a good theoretical framework, it could offer more practical guidance on how to implement these concepts in daily clinical practice. For instance, specific case examples showing how misinterpretation of DLco relationships could lead to incorrect diagnoses would illustrate the practical importance of the topic.

Response: While this is an excellent suggestion, it would make our already lengthy article even more wordy and laborious to read. Exemplifying the narrative will surely make it more understandable and is something that can be put to writing in future.

Comment: There is limited discussion about the variability and reproducibility of DLco measurements, which is crucial for proper test interpretation. The paper doesn’t adequately address how test variability might affect the interpretation of isolated DLco reduction.

Response: Detailed discussion about test variability and reproducibility is beyond the scope of this review.

Comment: While the paper mentions ATS/ERS standards, it doesn’t thoroughly critique current clinical guidelines or provide specific recommendations for improving them based on the physiological principles discussed.

Response: Again, detailed discussion about acceptability, reproducibility and variability of DLco testing is something that is beyond the scope of this article.

Reviewer 4 Report

Comments and Suggestions for Authors

Please find the attached file.

Comments for author File: Comments.pdf

Author Response

Comment: The emphasis may be more placed on the Hb correction in the interpretation of DLco or Kco because suggested diseases in the review may show anemia. To differentiate the precise degree of alveolar-capillary units requires Hb correction.

Response: We acknowledge the comment. It is briefly mentioned in the review and this adjustment is also a part of algorithm as shown. Detailed discussion on this topic is beyond the scope of this article. Mostly DLco is auto adjusted for anemia before it is reported to the clinicians.

Comment: Please note that congestive heart failure may increase diffusing capacity.

Response: Acknowledged

Comment: In the early sections, V♙ is explained as lung volume. V♙ is alveolar volume assessed in the single breath methods, which the authors precisely pointed out in the latter parts. The explanation of the term may be unified throughout the review. The readers might be confused by lung volume assessed using He dilution methods.

Response: Thank you for pointing that out. Partly the term “lung volume” was retained to refer to the misconception itself and was later corrected in the article to reflect just as you pointed out. We proofread the article again to maintain consistency.

Comment: For the clinical practice, the significance of prediction equation may be provided in the review because the Global equations are suggested while the discussion on the universality across the ethnicities remains undetermined.

Response: In its current state, this is something that is important but is beyond the scope of this article.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

There are still some revisions needed.

In section 5.1  VA is not the number of contributing units.  Instead VA is the lung volume in which the inspired CO and tracer gas distributes.  You could describe KCO as efficiency per lung volume.  Alternatively you could describe KCO = DLCO/VA where VA is the volume of distribution of the inhaled CO and tracer gas.

 

In section 5.2 you say "In other words, DLco is not the measurement of an actual physical variable. Instead, it is a calculation of what would be the flux of CO from the alveoli to the blood in the hypothetical condition of the subject’s lungs being filled with 100% CO."  That statement is incorrect.  The calculation of DLCO does not assume the lungs would be filled with 100% CO (which would kill the patient).  Instead it measures (or calculates) the rate of uptake of CO of the entire lung in which CO and tracer gas were distributed per pressure gradient of CO.  It uses the tracer gas to determine the lung volume in which gas was distributed (the VA) ( from the inspired and expired tracer gas and volume inspired ) and the initial CO concentration (prior to diffusion of CO, the initial CO is diluted the same as the tracer gas). 

For patients without obstruction the VA plus estimated dead space closely matches TLC by plethysmography, while patients with emphysema can have VA much lower than TLC.   Emphysema patients usually have elevated TLC by plethysmography and normal VA – leading to another change needed in the algorithm.

Your figure 2 is confusing and I think is better understood having a single y axis and having alveolar volume go from 40 to 100 - not 100 to 40.  The figure was based equations from the study (your reference 32).  Those equations could be used by you to generate graph with single y-axis.

Several changes are needed to your algorithm.  Checks for anemia should be done prior to doing chest CT.  If possible predicted DLCO should be adjusted for hemoglobin.  The approach to interpreting DLCO should emphasize that it is important to evaluate whether the DLCO is lower than expected for the patient's lung volume  -- whether the DACO is low as well.  If the DLCO is what is expected for the lung volume (if DACO is normal), that is consistent with failure to take deep breath which could be from neuromuscular disease or extra-parenchymal disease such as kyphoscoliosis or pleural effusion.  Since patients with ILD and DLCO below 80% can have KCO low, normal or above 100% you need adjust the algorithm to include that.

The algorithm approach to reduced DLCO would be improved by including DACO - or DLCO with predicted adjusted for lung volume.  Patients with emphysema typically have a normal VA (much lower than their TLC by plethysmography) so should also be under "Normal VA").  Some patients with ILD have KCO above 100%, but not as high as expected from their low VA -so ILD should show up with KCO low, normal, or high.  Patients with ILD and pneumonectomy will have reduced DACO.  Patients with normal lung parenchymal will have normal DACO.

Attached is file with these comments and a proposed modified algorithm.

 

Comments for author File: Comments.pdf

Author Response

Comment: 

In section 5.2 you say "In other words, DLco is not the measurement of an actual physical
variable. Instead, it is a calculation of what would be the flux of CO from the alveoli to the
blood in the hypothetical condition of the subject’s lungs being filled with 100% CO." That
statement is incorrect. The calculation of DLCO does not assume the lungs would be filled
with 100% CO (which would kill the patient). Instead, it measures (or calculates) the rate of
uptake of CO of the entire lung in which CO and tracer gas were distributed per pressure
gradient of CO. It uses the tracer gas to determine the lung volume in which gas was
distributed (the VA) ( from the inspired and expired tracer gas and volume inspired ) and the
initial CO concentration (prior to dilusion of CO, the initial CO is diluted the same as the
tracer gas).

Response: 

Thank you for pointing that out, correction has been made and will reflect on revision. This incorrect notion is now deleted with the reference. While it was not intended to mislead the readers and estray them From basic Physiology, we would definitely like to avoid any confusion or misleading information given the delicacy of the subject. In this regard, manuscript is revised and the following lines will be deleted from manuscript, along with its reference. (Lines 177-180).

Comment: 

Several changes are needed to your algorithm. Checks for anemia should be done prior to
doing chest CT. If possible predicted DLCO should be adjusted for hemoglobin. The
approach to interpreting DLCO should emphasize that it is important to evaluate whether
the DLCO is lower than expected for the patient's lung volume -- whether the DACO is low
as well. If the DLCO is what is expected for the lung volume (if DACO is normal), that is
consistent with failure to take deep breath which could be from neuromuscular disease or
extra-parenchymal disease such as kyphoscoliosis or pleural e?usion. Since patients with
ILD and DLCO below 80% can have KCO low, normal or above 100% you need adjust the
algorithm to include that.

Response:

This was one of the major changes to the manuscript. We have accepted the altered proposed algorithm by a reviewer and have incorporated concept of DACO which physiologically should be helpful in interpretation of diffusion capacity. Initially this concept was foregone, intentionally to keep the algorithm simple for the readers and not to complicate this already complicated subject. We do agree that full understanding and knowledge of the fuel capacity of the lung, particularly how it is tested in the lab is far more complicated for any one review article to encompass.

Please see lines 244&245 stating the updated algorithm.

Further, to help the readers understand the new algorithm and added subject of DACO, we have added a paragraph to explain it further in text. This addition can be seen starting from line 251 to 257.

Reviewer 4 Report

Comments and Suggestions for Authors

Am J Respir Crit Care Med2012 Jul 15;186(2):132-9.

may be appropriate to be cited to enhance the understanding of the readership for this review.

Author Response

Comment: 

Am J Respir Crit Care Med2012 Jul 15;186(2):132-9.

may be appropriate to be cited to enhance the understanding of the readership for this review.

 

Response:

Reference is now added in the manuscript, as advised by the reviewer. Please see line 202 and 383/384.