Electrical Impedance Tomography to Set Positive End Expiratory Pressure During Pediatric Extracorporeal Membrane Oxygenation for Respiratory Failure... Is it Feasible?

Abstract

Dear Editor, Setting positive end expiratory pressure (PEEP) during extracorporeal membrane oxygenation (ECMO) for severe pediatric acute respiratory distress syndrome (PARDS) is challenging, and no robust data exist to guide decision-making.1 Although PARDS shares common features with ARDS¸ data from adults cannot be directly translated to children because pediatric lungs have a higher elastin/collagen ratio and a higher compliance of the chest wall than adults.2 Thus, appropriate PEEP selection methods should be used to set PEEP in children with PARDS to prevent both atelectrauma and overdistension. Prospective data on optimal PEEP selection in PARDS are scarce and variable among centers.2 Recently, Khemani et al.2 suggested that in children with PARDS, the use of PEEP levels lower than that recommended by the ARDSNet Table3 was associated with an increased risk of death. In general, when ECMO is applied to manage PARDS, native lung FiO2, respiratory rate and tidal volume (Vt) are lowered to reduce the risk of ventilator-induced lung injury.1 Reduction of Vt may cause de-recruitment, thus, an adequate PEEP level should be used.1 In this context, recent data obtained in the adult population supported with ECMO for ARDS suggest that a high PEEP on conventional ventilation during the early ECMO period may be of benefit to prevent atelectrauma and extensive lung collapse, which may in turn prolong ECMO duration and increase the risk of ECMO-related complications.4 On the same time, setting an inappropriate high PEEP level may cause overdistention and irreversible lung damage. Setting PEEP according to the lower PEEP/FiO2 ARDSNet table is impractical in PARDS supported with ECMO, thus, alternative physiologic-based approaches should be used. In this case report, we will describe the contemporary use of three different techniques (Figure 1) to set PEEP in a 14-year-old child supported with veno-venous ECMO for a chest trauma. These techniques, based on direct imaging - electrical impedance tomography5,6—or, on respiratory system mechanics—best respiratory system compliance (CRS)7 and the dynamic airway pressure-time curve profile (Stress Index)8—are commonly applied during a decremental PEEP trial (Figure 2) maintaining constant both the ECMO settings (blood flow, FiO2, and sweep-gas) and the ventilator settings (FiO2, Vt, and respiratory rate). Informed consent for all bedside monitoring and data treatment was obtained by the parents.Figure 1.: Positive end expiratory pressure (PEEP) titration according to the best dynamic compliance of the respiratory system, the stress index (SI), and the electrical impedance tomography (EIT). ΔP, driving pressure; Cdyn, dynamic compliance of the respiratory system.Figure 2.: Trend of collapse and overdistension during a decremental positive end expiratory pressure (PEEP) titration with electrical impedance tomography (EIT).PEEP setting guided by the best CRS is the most common technique used in ARDS trials3 and consists in the selection of the PEEP value that results in the highest CRS during a decremental PEEP trial. EIT is a noninvasive, real time, radiation-free tomographic imaging technique providing global and regional analysis of both lungs.5,6 Currently, only few studies evaluated PEEP titration guided by EIT in adult patients receiving ECMO,9,10 and no published study evaluated EIT in children undergoing ECMO. Stress index (SI) is a technique used to study tidal recruitment/overdistension at bedside with a dedicated software.8 It requires volume control ventilation (VCV) with a constant flow, so that alveolar volume and airway pressure increase at a constant rate. The slope of the airways pressure rise reflects changes in CRS over time. An upward concavity of the pressure-time curve (SI > 1, decrease CRS) indicates overdistention, a downward concavity of the pressure time curve (SI < 1, increase CRS) indicate tidal recruitment. A linear shape of the pressure-time curve (SI = 1, constant CRS) indicate protective ventilation. No published study evaluated the use PEEP titration guided by SI in children undergoing ECMO. Our patient, before the decremental PEEP trial, was ventilated in VCV using a Servo-U ventilator (Maquet, Getinge Group, Germany) with a tidal volume of 6 ml/kg (480 ml) and a plateau pressure 20 cm H2O. PEEP was set at 8 cm H2O, respiratory rate at 12 breaths/min, I:E ratio at 1:1.5, and FiO2 at 45%. Deep sedation with continuous infusion of both midazolam 0.3 mg/kg/h and fentanyl 6 mcg/kg/h was provided to avoid any attempt of spontaneous breathing. An EIT belt was placed around the patient’s thorax between the fourth and fifth intercostal space and connected to the EIT monitor (Enlight 2100, Timpel, Sao Paulo, Brazil).5,6 This EIT monitor is currently licensed for neonatal and pediatric use only in Europe and Brazil, while is still limited licensed (research purposes) in USA, Canada, and Australia.5 Before beginning the decremental PEEP trial, PEEP was arbitrarily set at 18 cm H2O for 5 minutes as steady state baseline trying to maintain a plateau pressure ≤30 cm H2O. Then, PEEP was lowered of 2 cm H2O decrements every 120 seconds. The trial was stopped in case of hemodynamic instability (>20% reduction of systolic blood pressure). During the decremental PEEP trial, dynamic CRS progressively increased with a peak at 9 cm H2O of PEEP and then decreased. SI progressively decreased during the PEEP trial and identified a wide range of PEEP (13–15 cm H2O) where the amount of overdistension and tidal recruitment were minimal. EIT, instead, identified 12.5 cm H2O as an optimal PEEP level, which corresponded to the point where both overdistension and alveolar collapse were minimal. In this case report, we contemporary evaluated three different techniques to set PEEP. EIT identified the optimal PEEP level where both alveolar collapse and overdistension were at minimal without using radiation (e.g., computerized tomography). SI was less accurate and did not identify a single level of PEEP. Consistent with previous findings,7 setting of PEEP based on the best CRS has been shown to be unreliable since global CRS primarily reflects changes in the mechanical properties of the lung including tissue already opened for ventilation. Multiple pressure-volume (PV) curves and dead space monitoring represent other methods to set PEEP at bedside; however, they present some limitations. PV curves are relatively complex to perform at bedside due to the need of merging two or more curves starting from different lung volumes. Dead space monitoring, instead, is impractical in ECMO considering the amount of CO2 removed by the oxygenator. According to these findings, EIT seems an interesting noninvasive and radiation-free bedside tool to individualize PEEP selection in children undergoing ECMO. Further investigations are needed to evaluate its benefit to further decrease ventilator-induced lung injury during ECMO. Acknowledgement The authors thank M.A. Barbieri for the English editing of the paper.