Abstract
DLCO reports and interpretation should be standardised and include adjusting predicted DLCO and KCO for lung volume http://ow.ly/ywTA30cOh44
To the Editor:
The American Thoracic Society (ATS) and European Respiratory Society (ERS) should be congratulated on updating standards for diffusing capacity of the lung for carbon monoxide (DLCO) [1]. I agree that “Besides varying with age, sex, height and possible ethnicity, DLCO also changes with Hb, lung volume, COHb, PIO2 …, exercise and body position.” and that “adjustments for these factors be made in the predicted rather than the measured DLCO”. Reporting transfer coefficient of the lung for carbon monoxide (KCO) rather than DLCO/alveolar volume (VA) will help get away from the mistaken notion that DLCO/VA “corrects” DLCO for lung volume [2]. While the new standards describe how to adjust predicted DLCO for haemoglobin (Hb), COHb and inspired oxygen tension (PIO2), it does not discuss how to adjust predicted DLCO and KCO for lung volume.
The following equations [3] were included in the 2005 ATS/ERS DLCO standards [4], and describe how to adjust DLCO and KCO for lung volume. They were developed studying normal subjects with experimental reductions in inspired volume (VI; and thus VA) and fit the model that DLCO and KCO change in a manner expected from having DLCO reduced proportionate to the surface area for gas exchange with the capillary blood component unchanged. Mathematically, they result in DLCO % predicted for lung volume equaling KCO % predicted for lung volume when using the equation KCO(predicted)=DLCO(predicted)/VA(predicted).
DLCO[predicted for lung volume]=DLCO[predicted]×(0.58+0.42×(VAm/VAp))
KCO[predicted for lung volume]=KCO[predicted]×(0.42+0.58/(VAm/VAp))
with VAm/VAp=measured VA/predicted VA.
For example, at VA 50% of predicted, the DLCO predicted for lung volume is 80% and KCO is 160% of that for VA 100% of predicted.
The standards require reporting DLCO and KCO (adjusted, predicted) with specification of the adjustments. Additional reporting requirements should include DLCO (% of adjusted predicted) and VA (% predicted).
Neither the 2005 nor the current standards address how to report DLCO and KCO adjusted for lung volume, or how to interpret DLCO.
In addition to knowing % predicted DLCO and KCO adjusted for all factors except lung volume, it is also very helpful to know % predicted DLCO and KCO when also adjusted for lung volume [2]. Just as adjusting predicted DLCO and KCO for haemoglobin in an anaemic patient yields a better indication of the lung's ability of gas exchange, adjusting DLCO and KCO for lung volume in a patient with low lung volume yields a better indication of the lung's ability of gas exchange.
A shorter nomenclature is needed for DLCO and KCO % predicted also adjusted for lung volume.
I propose DACO and KACO to refer to DLCO and KCO predicted values that have been adjusted for lung volume (the “A” refers to adjusted for lung volume.)
Reporting requirements should include DACO (adjusted, predicted), KACO (adjusted, predicted), as well as DACO (% of adjusted predicted) and KACO (% of adjusted predicted).
The new standards recommend development of a standardised common report form. I propose the following, one when Hb is not measured (box 1) and a second when Hb is measured (box 2), with both including % of FVC for VI if spirometry was done the same day
Diffusing capacity | Predicted range | Actual | % pred | ||
Mean | 95% | ||||
DLCO mL·min−1·mmHg−1 | xx.xx | xx.xx | dd.dd | xx | Predicted not adjusted for Hb |
VA (BTPS) L | x.xx | x.xx | x.xx | xx | |
KCO mL·min−1·mmHg−1·L−1 | xx.xx | xx.xx | x.xx | xx | Predicted not adjusted for Hb |
VI (BTPS) L | x.xx | x.xx | xx.xx | xx | xx% of FVC |
DACO mL·min−1·mmHg−1 | xx.xx | xx.xx | dd.dd | yy | Predicted adjusted for lung volume |
DLCO and KCO are yy% predicted, adjusted for lung volume.
Diffusing capacity | Predicted range | Actual | % pred | Hb xx.x from D-MON-YYYY | |
Mean | 95% | ||||
DLCO mL·min−1·mmHg−1 | xx.xx | xx.xx | dd.dd | xx | Predicted not adjusted for Hb |
DLCO mL·min−1·mmHg−1 | xx.xx | xx.xx | dd.dd | xx | Predicted adjusted for Hb |
VA (BTPS) L | x.xx | x.xx | x.xx | xx | |
KCO mL·min−1·mmHg−1·L−1 | xx.xx | xx.xx | x.xx | xx | Predicted adjusted for Hb |
VI (BTPS) L | x.xx | x.xx | x.xx | xx | xx% of FVC |
DACO mL·min−1·mmHg−1 | xx.xx | xx.xx | dd.dd | yy | Predicted adjusted for lung volume and Hb |
DLCO and KCO are yy% predicted, adjusted for lung volume and Hb.
For both reports, if DLCO and KCO predicted were also adjusted for COHb and/or PIO2, then a line saying “Predicted DLCO and KCO also adjusted for …” should appear at the end, which includes the data used to make the adjustment, such as “COHb of 2.6% and altitude of 2000m.” The 95% values are the lower limit of normal (LLN), with DACO[LLN]=DACO[adjusted,predicted]×DLCO[LLN]/DLCO[predicted].
There is not a clear consensus on interpretation of DLCO. I recommend the following algorithm to interpret DLCO, with DLCO % predicted, adjusted and LLN the lower limit of normal (box 3).
DLCO ≥80% and ≥LLN | DLCO is normal |
DLCO <80% but ≥LLN | DLCO is near lower limit of normal |
DLCO ≥60%, <80%, and <LLN | DLCO is mildly reduced |
DLCO ≥40%, <60%, and <LLN | DLCO is moderately reduced |
DLCO <40% | DLCO is severely reduced |
If DLCO is not normal, and DLCO adjusted for lung volume (DACO) is above the LLN as % predicted, then add phrase “due to low lung volume”.
If DLCO is not normal, and DLCO adjusted for lung volume is below the LLN as % predicted but more than 10% predicted greater than DLCO, then add phrase “in part due to low lung volume”.
As a co-author of the 2005 ERS/ATS DLCO standards, I believe including adjustments of DLCO for lung volume and standardised reports and interpretation would improve the clinical value of DLCO.
Footnotes
Conflict of interest: None declared.
- Received May 8, 2017.
- Accepted May 31, 2017.
- Copyright ©ERS 2017