North-South Syndrome during Peripheral Veno-Arterial Extracorporeal Membrane Oxygenation Support. A Physiological Approach to Ventilation Optimization.

Background: Mechanical ventilation is required in about 60-70% of patients receiving veno-arterial extracorporeal membrane-oxygenation (VA-ECMO). Optimal ventilatory strategies in this setting are poorly defined. Electrical impedance tomography (EIT) has been described as a tool for PEEP optimization in patients with respiratory distress and might guide ventilatory optimization also in this setting.

Methods: A 50-year-old man with coronary artery disease presented with acute left-sided chest pain and immediately developed ventricular fibrillation (VF) and cardiac arrest. Due to refractory VF, peripheral VA-ECMO was initiated. No unloading device was required as left ventricular pulsatility recovered. Coronary angiography revealed an occluded right coronary artery, which was successfully treated.
By day two, myocardial function had substantially recovered, with left ventricular ejection fraction increasing from < 10% to 30-35% and stable filling pressures. Increased pulsatility was accompanied by differential hypoxemia consistent with north-south syndrome. Right radial arterial blood gas analysis demonstrated severe hypoxemia (PaO₂59 mmHg; PaO₂/FiO₂73) despite high FiO₂. Respiratory mechanics were further assessed using esophageal manometry. Cardiac output and stroke volume were assessed non-invasively by transthoracic echocardiography. Pulmonary perfusion was evaluated by EIT after injection of 7.5% NaCl via a pulmonary artery catheter.

A PEEP titration was performed assessing respiratory mechanics and EIT derived measures at decremental PEEP levels from 20 to 10 cmH₂O (Figure 1). Best PEEP was defined as the PEEP level minimizing lung overdistension and collapse during the trial and was set at 14cmH₂O.

Outcome: During VA-ECMO, ventilation-perfusion (V/Q) mismatch may persist or worsen in the presence of underlying lung injury. Optimizing PEEP aims to balance alveolar recruitment and overdistension, thereby improving respiratory system mechanics and promoting lung recovery.

In this case PEEP optimization using EIT resulted in significant improvement in arterial oxygenation, more homogeneous ventilation distribution, and increased posterior lung recruitment (Figure2). Ventilation-perfusion matching improved, with the EIT-derived shunt fraction decreasing from 10% to 1%. Respiratory mechanics showed improved lung compliance, indicating effective recruitment without relevant overdistension. No major hemodynamic changes were observed, and cardiac output and stroke volume remained stable (Table 1).

After ventilatory optimization, oxygenation remained adequate. VA-ECMO was discontinued after five days. Vasopressors were weaned three days later, and the patient was transferred from the coronary care unit to the medical intensive care unit on day 15. Prolonged respiratory support was required, and a tracheostomy was performed. After 32 days of mechanical ventilation, the patient was successfully weaned to a tracheostomy mask. Seven days later, the patient was transferred to a long-term acute care facility.

Conclusion: Hypoxemia during VA-ECMO may persist despite extracorporeal oxygenation and may become evident with myocardial recovery. This case illustrates that a physiological approach using EIT-guided PEEP optimization can improve gas exchange. Prospective studies are needed to validate ventilation strategies using EIT in VA-ECMO patients.