Spontaneous breathing during veno-venous extracorporeal membrane oxygenation

Abstract

Veno-venous extracorporeal membrane oxygenation (VV ECMO) has started to be applied in awake spontaneously breathing patients as an alternative to invasive mechanical ventilation. As the physiologic cardiorespiratory variability is increased in this condition, the dynamic interaction between patient respiratory activity and extracorporeal system function affects the clinical management. The effect of extracorporeal CO₂ removal on patient respiratory drive is variable and not always predictable, with some patients responding to CO₂ removal with a decrease in respiratory rate and effort and other patients demonstrating a persistently high work of breathing independent on CO₂ unload. While the pathophysiological mechanisms of this different interactions are still to be clarified, improved monitoring ability is needed both to titrate the support in responders and to avoid the risk of ventilation injury in non-responders. Acute changes in patient respiratory patterns may also occur during spontaneous breathing, making it difficult to maintain constant levels of extracorporeal respiratory support, also because changes in the distribution of venous blood volume due to lung-heart interactions affect extracorporeal blood flow. Assessment of native lung function and of its evolution over time is challenging while respiratory gas exchanges are provided by the extracorporeal system, since both oxygenation and decarboxylation capabilities can be fully evaluated only when alveolar ventilation is restored reducing extracorporeal CO₂ removal. The rationale for using "awake ECMO" varies across different types of acute respiratory failure: the pathophysiological mechanisms of the underlying disease affect the patient-ECMO interaction and the goal of support. In this review we discuss the pathophysiology, technical challenges and monitoring issues of the use of ECMO in awake spontaneously breathing patients with acute respiratory failure of different etiologies.

Keywords: Awake; acute respiratory distress syndrome (ARDS); bridge to lung transplant; chronic obstructive pulmonary disease (COPD); extracorporeal membrane oxygenation (ECMO); spontaneous breathing; veno-venous (VV).

Figure 1

Extracorporeal CO₂ removal for liter of sweep gas flow as a function of partial pressure of CO₂ entering the membrane lung (pCO₂ pre-ML). Clinically, the lower the pCO₂ pre-ML, the higher gas flows required to remove the same amount of CO₂. In ARDS patients with low level of pCO₂ pre-ML the amount of CO₂ removed at maximum gas flow was not sufficient to decrease respiratory distress. ARDS, acute respiratory distress syndrome; COPD, chronic obstructive pulmonary disease. Reprinted with permission from reference (24).

Figure 2

Systemic and pulmonary systolic arterial pressures traces before and after the connection to femoro-femoral VV ECMO in an awake patient bridged to lung transplantation. The patient was spontaneously breathing and severely hypoxemic. Carbon dioxide arterial tension was normal (around 39 mmHg) both before and after bypass connection. As shown, pulmonary arterial systolic pressure progressively improved from around 100 to 60 mmHg after VV ECMO start. VV ECMO, veno-venous extracorporeal membrane oxygenation.

Figure 3

Respiratory rate responses to change in sweep gas flow for individual patients of the three groups. All the bridge to lung transplant (n=9) and the COPD (n=6) patients decreased their respiratory rate below the threshold level of 10, while only 50% of ARDS (n=8) did so, despite maximum level of gas flow. ARDS, acute respiratory distress syndrome; COPD, chronic obstructive pulmonary disease. Reprinted with permission from reference (24).