key: cord-0977367-g02scs7r authors: Busana, Mattia; Giosa, Lorenzo title: The knowns and unknowns of perfusion disturbances in COVID-19 pneumonia date: 2021-09-28 journal: Crit Care DOI: 10.1186/s13054-021-03742-y sha: bc9f3b176bf607153eb8ce239b81e9a1be7bda96 doc_id: 977367 cord_uid: g02scs7r nan Ball et al. [1] recently proposed a thoughtful quantitative analysis of gas and blood distribution in the lungs of patients with severe COVID-19 pneumonia. The paper offers some interesting cues that we consider worth discussing. First, the images of dual-energy CT scan (DECT) were used as surrogates of the ventilation and perfusion distribution in the lung. However, other imaging techniques, like scintigraphy and SPECT, might be more suitable for this task. Indeed, while perfusion is likely accurately depicted by DECT, ventilation is not. Indeed, for at least two reasons, the gas content measured by this technique (notably during tidal breathing and not at constant airway pressure) may not represent a good surrogate of the distribution of ventilation: 1. Airway closure may dissociate the instantaneous gas volume from the actual ventilation, and the phenomenon may be more common than currently thought [2] ; 2. The opening pressures cannot be investigated by a single CT scan [3] . It follows that undetected recruitability (possibly nonlinear along the pressure-volume curve) may contribute to the uncoupling between measured gas volume and alveolar ventilation. This leads, in our opinion, to some inconsistencies. Most importantly, in Fig. 4 , the authors show a plot reasonably inspired by the V A /Q T distributions obtained with the multiple inert gas elimination technique. Unfortunately, such distribution is likely far from what we may expect from COVID-19 pneumonia. Indeed, the displayed distributions are essentially centered on a gas-blood volume ratio ~ 1. However, as we have shown in our theoretical model [4] , such distribution (coupled with the relatively low shunt fraction reported in Table 2 ) is not compatible with the severe hypoxemia observed in these patients. As a side note, considering that the analysis performed by Ball et al. excluded large vessels, one may wonder how the presence of abnormal vasodilation and vascular anastomoses [5] would affect their results. Indeed, nonfunctional vessels in a "gas-rich" region may escape the definition of shunt used by the authors. The impact of perfusion alterations on gas exchange in COVID-19 is far from being understood. Despite the aforementioned limitations, Ball et al. must be congratulated for their effort in reporting compelling data that represent a smart step forward in the understanding and, most importantly, in the quantification of the problem. We would like to thank Dr. Busana and Dr. Giosa for their interest in our study [1] and for giving us the opportunity to expand the discussion regarding disturbances of lung perfusion in COVID-19 patients. They highlighted the differences between ventilation and static aeration as assessed by CT; questioned the interpretation of our gas:blood ratio distribution plots and the definition of shunt, comparing our in vivo results with those of their computational theoretical model [4] . As per local practice during the COVID-19 pandemic, CT scans were acquired during breath-hold and we used lung aeration to define shunt, dead space, and non-aerated/non-perfused areas. Lung aeration may be considered as a potential surrogate of ventilation, especially when protective tidal volume and limited inspiratory plateau pressure are used, as in the case of COVID-19 mechanically ventilated patients [6] . In this context, it seems reasonable to assume that non-aerated areas are also non-ventilated and that aerated regions receive a certain amount of ventilation, especially given the small extent of hyperaeration and poor recruitment with application of PEEP observed in these patients [7] . The pulmonary gas:blood volume ratio was defined as: namely as the ratio of mean-normalized gas fraction to mean-normalized pulmonary blood volume: based on this definition, the values of the ratio are centered around 1 in each patient and cannot be directly translated into V′ A /Q′ T without knowing the cardiac output and alveolar minute ventilation. Such a ratio allows within-patient comparisons of the relative distribution of aeration and perfusion. We excluded only large lung vessels from the analysis: the vasodilation of peripheral vessels affects V′ A /Q′ T at an anatomical scale well below the size of one CT voxel [8] : voxels comprising dilated vessels and patent alveoli would have an intermediate CT attenuation value and a high pulmonary blood volume, corresponding to our "gas:blood ratio < 1" compartment. Despite several technical limitations of the technique we proposed, our overall interpretation of the data is that in the early stages of the disease, hypoxia is explained by low V′ A /Q′ T responsive to higher inspiratory oxygen fractions and non-invasive respiratory assistance, while also by increased true shunt in the late stages requiring invasive mechanical ventilation. In conclusion, we believe that our data should not be interpreted as in contrast with your theoretical model, but rather could represent its confirmation and an attempt of explaining in vivo gas-exchange changes in critically ill patients with severe COVID-19 pneumonia. Gas: Blood volume ratio = f A · PBV PBV · f A Abbreviations DECT: Dual-energy CT scan; SPECT: Single photon emission computed tomography; V A /Q T : Ventilation-perfusion distribution. Lung distribution of gas and blood volume in critically ill COVID-19 patients: a quantitative dual-energy computed tomography study Airway closure, more harmful than atelectasis in intensive care? Opening pressures and atelectrauma in acute respiratory distress syndrome The impact of ventilation-perfusion inequality in COVID-19: a computational model Intrapulmonary bronchopulmonary anastomoses in COVID-19 respiratory failure Effects of positive end-expiratory pressure on regional distribution of tidal volume and recruitment in adult respiratory distress syndrome Computed tomography assessment of PEEP-induced alveolar recruitment in patients with severe COVID-19 pneumonia Structure and composition of pulmonary arteries, capillaries, and veins None. Both authors participated in the writing of the present letter. Availability of data and materials NA. Ethics approval and consent to participate NA. All authors approved the manuscript in the current form. The authors have no conflicts of interest to disclose. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.