Spontaneous Breathing during Extracorporeal Membrane Oxygenation in Acute Respiratory Failure

Background: We evaluate the clinical feasibility of spontaneous breathing on extracorporeal membrane oxygenation and the interactions between artificial and native lungs in patients bridged to lung transplant or with acute exacerbation of chronic obstructive pulmonary disease (COPD) or acute respiratory distress syndrome.Methods: The clinical course of a total of 48 patients was analyzed. Twenty-three of 48 patients were enrolled in the prospective study (nine bridged to lung transplant, six COPD, and eight acute respiratory distress syndrome). The response to the carbon dioxide removal was evaluated in terms of respiratory rate and esophageal pressure swings by increasing ("relief" threshold) and decreasing ("distress" threshold) the extracorporeal membrane oxygenation gas flow, starting from baseline condition.Results: Considering all 48 patients, spontaneous breathing extracorporeal membrane oxygenation was performed in 100% bridge to lung transplant (9 of 9 extubated), 86% COPD (5 of 6 extubated), but 27% acute respiratory distress syndrome patients (6 of 8 extubated; P < 0.001) and was maintained for 92, 69, and 38% of the extracorporeal membrane oxygenation days (P = 0.021), respectively. In all the 23 patients enrolled in the study, gas flow increase (from 2.3 +/- 2.2 to 9.2 +/- 3.2 l/min) determined a decrease of both respiratory rate (from 29 ; 6 to 8 ; 9 breaths/min) and esophageal pressure swings (from 20 +/- 9 to 4 +/- 4 cm H2O; P < 0.001 for all). All COPD and bridge to lung transplant patients were responders (reached the relief threshold), while 50% of acute respiratory distress syndrome patients were nonresponders.Conclusions: Carbon dioxide removal through extracorporeal membrane oxygenation relieves work of breathing and permits extubation in many patients, mainly bridge to lung transplant and COPD. Only few patients with acute respiratory distress syndrome were able to perform the spontaneous breathing trial, and in about 50% of these, removal of large amount of patient's carbon dioxide production was not sufficient to prevent potentially harmful spontaneous respiratory effort.