Pulmonary Vascular Changes in Acute Respiratory Distress Syndrome due to COVID-19

To the Editor:

This letter is in response to an article published in a recent issue of the Journal by Patel and colleagues (1). The authors’ observation is consistent with a previous report, suggesting varying grades of pulmonary thromboembolism, pulmonary vascular microthrombosis, and pulmonary vascular dilatation in an advanced stage of acute hypoxemic failure owing to coronavirus disease (COVID-19) (1, 2). In the current study, radiologic findings were obtained when nearly 50% of patients were on extracorporeal membrane oxygenation (ECMO). Therefore, interpretation and generalization of the findings becomes somewhat more intriguing because of complexities arising from hemodynamic, oxygenation, and hematologic alterations induced by ECMO (3). Venoarterial ECMO is known to increase afterload, left ventricular (LV) end diastolic pressure, left atrial pressure, and postcapillary venous dilatation. Furthermore, femoral arterial oxygenated flow may not reach the coronary circulation because of the watershed effect (north–south syndrome) and may induce LV ischemia and aggravate LV dysfunction (4). In addition, the venous return diversion to the ECMO circuit may induce stagnation in the pulmonary circulation, which may get further aggravated by an increase in pulmonary vascular resistance because of positive end-expiratory pressure. However, total lung–blood volume may get reduced and there is a lesser hydrostatic pressure gradient for pulmonary edema formation. Therefore, a reduction in dynamic compliance may be expected during venoarterial ECMO because of a fall in total pulmonary fluid volume. Undoubtedly, venovenous ECMO maintains better cardiorespiratory interaction than venoarterial ECMO, and it does not increase LV end-diastolic pressure, left atrial pressure, and pulmonary venous pressure. A rise in mixed venous oxygen saturation diminishes hypoxic pulmonary vasoconstriction response to increase pulmonary vascular dilatation in diseased lobes. Furthermore, a higher minute ventilation because of hypoxic and hypercarbic drive may subside during ECMO, and the ventilatory ratio may also come down substantially. Admittedly, physiological dead space, shunt fraction, and ventilation perfusion matching would be entirely different in patients with acute hypoxemic failure on ECMO compared with patients not on ECMO. Furthermore, clinicians may be tempted to change the ventilator settings during ECMO run. Peripheral pulmonary vascular dilatation gets obscured by opacification of the lung, and the presence of such dilatation in a healthy aerated lung region suggests extraalveolar vessel dilatation as a result of intraalveolar vessel compression owing to positive end-expiratory pressure (5). Demarcation of pulmonary vascular dilatation between precapillary and postcapillary component could have helped in assessing the role of pulmonary vascular resistance and LV dysfunction. Even though the sample size was small, stratification and comparison of oxygenation parameters and the computed tomographic scan findings in patients on ECMO and in patients not on ECMO would have provided an additional pathophysiological insight. In addition, comparisons among patients on different types of ECMO would have been worthwhile. The study did not provide sufficient data to suggest whether the correlation between different ventilatory and oxygenation parameters were assessed before ECMO or after ECMO. Furthermore, the author’s interpretation that there was no correlation between the computed tomographic scan findings and Pa_O2/Fi_O2 ratio or the ventilatory ratio seems to suggest an overwhelming role of ECMO in improving Pa_O2/Fi_O2 ratio and ventilatory ratio in nearly 50% of patients. An increase or decrease in the incidence of pulmonary vessel dilatation, microthrombosis, and pulmonary thromboembolism during ECMO may indicate the qualitative and quantitative change required in anticoagulation therapy. A persistent increase in the incidence of pulmonary thromboembolism while on ECMO despite adequate activated clotting time could be worrying. In addition, establishing a correlation between change in prevalence of pulmonary vascular dilatation with increasing duration of ECMO run would also have been meaningful.