Mechanical Ventilation during Extracorporal Support: The Relevance of Vt

To the Editor:

In cases in which pulmonary gas exchange is mainly guaranteed by extracorporeal support, the optimal ventilation strategy to protect the lung remains unclear. It is generally accepted that the ventilator should be set to prevent further ventilator-associated lung injury. Nevertheless, even a lung-protective approach with low Vts may still aggravate lung injury. Thus, an ultraprotective approach with very low Vts (<6 ml/kg) is frequently used in patients undergoing extracorporeal support to facilitate the healing of the injured lung (1). A very interesting concept is the reduction of the Vts to near apneic oxygenation, as done by Araos and colleagues (2). These researchers examined three different ventilation strategies in a swine acute respiratory distress syndrome model over the course of 24 hours, using extracorporeal membrane oxygenation to examine nonprotective, conventional, and near-apneic ventilation. The researchers found that histopathologic lung injury was lower in the conventional and especially the near-apneic group. However, wet-dry lung weight ratio and expression of most genes indicating fibroproliferation were not different between the groups. As remarked in the editorial by Fan (3), there was no comparison of ultraprotective strategies, and the three strategies differed not only in their Vts but also in positive end-expiratory pressure (PEEP) level and respiratory rate. Fan raised the question whether ventilation is needed at all during extracorporeal lung support. This was primarily described by Kolobow in an animal study (4).

Our group also conducted a study using a similar acute respiratory distress syndrome model (5). In the conventional group, protective mechanical ventilation with 6 ml/kg Vt was used. Unlike Araos and colleagues, we used arteriovenous extracorporeal lung assist to reduce Vts to 3 ml/kg body weight, and apneic oxygenation with Vts set to zero in further experimental groups. Moreover, an “open lung concept” was used in all groups by using PEEP levels above the lower inflection point of the lung. This strategy resulted in continuous airway pressure above 20 cm H₂O, even in the apneic group. Mean respiratory rate was similar in the 6 ml/kg and the 3 ml/kg group, with 20 and 17–18 breaths/min, respectively. After 24 hours, a histopathologic examination of the dependent lung showed more inflammation, alveolar exudation, and atelectasis with 3 ml/kg or no Vts. In contrast, alveolar overdistension was reduced with apneic oxygenation in the nondependent lung areas (5).

Hence, our study addressed several of the shortcomings of the data presented by Araos and colleagues and may help to answer the questions raised by Fan (3). Ventilation with protective Vts led to overdistension in the nondependent lung. Nevertheless, despite using high positive airway pressures, the dependent lung in the apneic group showed a worse lung injury score compared with protective Vts. Thus, the combination of both strategies as “near apneic ventilation with low respiratory rates” and higher PEEP levels might be very appealing. This strategy might prevent derecruitment of the dependent lung via repeated recruitment at a low rate set above higher PEEP levels. Overdistension of the nondependent lung may be prevented because of lower peak pressures and minimized shear stress resulting from a low respiratory rate. Another point is that using lower airway, and thus intrathoracic, pressures might reduce hemodynamic compromise. This is enabled by lower respiratory rates and lower Vts. Theoretically, a strategy with sufficient PEEP, low respiratory rates, and very low Vts individually adapted to the size of the residual nonconsolidated lung parts combined with prone positioning might be optimal to protect the lung during extracorporeal lung support.

We strongly agree with Fan that the optimal ventilator strategy during extracorporeal gas exchange should now be addressed in clinical studies.