Abnormal carbon monoxide diffusion capacity in COVID-19 patients at time of hospital discharge

To the Editor:

I have read with great interest the article by Mo et al. [1] entitled “Abnormal pulmonary function in COVID-19 patients at time of hospital discharge” recently published in the European Respiratory Journal. In this study, the authors describe pulmonary function tests in patients who suffered from coronavirus disease 2019 (COVID-19), which were performed on discharge from hospital. Patients were classified into three groups of severity. In the patients who suffered more severely, measured carbon monoxide diffusion capacity (D_LCO) was lower. However, when looking at D_LCO/V_A, which represents the transfer factor of carbon monoxide, the picture is different. Patients who had severe pneumonia had an average D_LCO/V_A of 82% of predicted while patients in groups classified as having mild disease or pneumonia had average values above 90%. Notably, all these averages have relatively high standard deviation values (e.g. 13.9% in severe pneumonia), meaning that some patients in the post-severe pneumonia group had a D_LCO/V_A >90% of predicted.

The authors present timely and important data as questions are now being raised concerning the long-term effects of COVID-19 [2]. However, I disagree with the authors concerning their interpretation of both the diffusion data and also what their findings mean clinically. In presenting their results the authors use the phrase “D_LCO corrected for alveolar volume (D_LCO/V_A)” which is not a correct expression and the notion of “D_LCO corrected for alveolar volume” is not acceptable these days [3]. The authors explain further that “D_LCO/V_A was still within the normal range, which might indicate that D_LCO decrease was more than the D_LCO/V_A in recovered subjects”, which is actually restating a finding without drawing a conclusion.

It is important to give emphasis here that by using carbon monoxide, D_LCO represents the gas diffusion capacity across the alveolar-capillary barrier and is in fact a mathematical product of the rate constant representing alveolar carbon monoxide uptake efficiency (D_LCO/V_A) multiplied by alveolar volume (V_A) [3]. Thus, when D_LCO/V_A is normal or near normal, D_LCO may still be low because of reduced alveolar volume, which means that complete acini with alveoli and blood vessels surrounding them are affected [3, 4]. In the remaining functionally available alveolar volume, the actual carbon monoxide uptake is normal as the diffusion through the alveolar-capillary barrier is not impaired and the blood volume within the pulmonary capillary bed is not reduced. Interstitial abnormalities or pulmonary vascular abnormalities are associated with reduced D_LCO/V_A which accompanies reduced D_LCO [4].

In the discussion of their findings, the authors cite previous studies describing follow-up of severe acute respiratory syndrome (SARS) patients having impaired D_LCO as the most common abnormality, with affected proportion of patients ranging from 15.5% to 43.6%. It would be important to describe how impaired D_LCO was defined in those studies, citing values as % of predicted and comparing that with the findings of the present study.

The conclusion which can be obtained from this study is that low D_LCO is caused mainly by reduced alveolar volume, and not residual interstitial abnormalities or pulmonary vascular abnormalities caused by COVID-19. The latter is important, since the proposed pathogenic mechanism describes an initial abnormality induced by COVID-19 similar to the effect of SARS and consists of a microvascular injury with some interstitial thickening with clear lungs on imaging and physical examination accompanied by profound hypoxaemia [5, 6], followed by development of alveolar abnormalities with consequent gradual loss of the alveolar spaces [6]. The filling of the alveoli with exudate at a later stage causes the loss of lung volume and reduced pulmonary compliance [7]. The finding of decreased alveolar volume on discharge may be explained by transient changes in mechanical properties of the chest wall and respiratory muscles after critical illness and may well address the concerns about a possible long lasting pulmonary parenchymal dysfunction post-COVID-19 [2].

In conclusion, the authors presented very useful data, but it is important to draw attention to the up-to-date manner in which we look at the carbon monoxide diffusion test and the actual meaning of the findings when considering the results of D_LCO and D_LCO/V_A together.

Shareable PDF