Background
Extracorporeal Membrane Oxygenation (ECMO) is currently an important life support for acute respiratory distress syndrome (ARDS) in adult patients [
1,
2]. According to the 2016 international report of the Extracorporeal Life Support Organization (ELSO) Registry, 58% of the adult patients receiving ECMO for severe respiratory failure can be saved and discharged from hospital [
3]. This report also reveals that about 90% of the 9812 ECMO runs for adult respiratory failure are in venovenous (VV)-associated configurations [
3]. The niche of VV-ECMO in the management of ARDS is to provide a pre-pulmonary blood gas exchange to the venous blood and reduce the patient’s dependence on pulmonary ventilation [
2].While the patient’s dependence on pulmonary ventilation is reduced, the risk and the severity of ventilator-induced lung injury (VILI) can theoretically be mitigated. Although the popularity of adult respiratory ECMO is continuously increasing, the applications of ECMO are still limited in large medical centers and reserved for the most advanced diseases [
1]. The discrepancy in user experience leads to considerable controversies about the timing of respiratory ECMO among experts worldwide. Currently, the timing of respiratory ECMO is mostly determined by the severity of hypoxemia which is represented by the ratio of arterial oxygen tension (PaO
2) to the fraction of inspired oxygen (FiO
2) under mechanical ventilation (MV). In the ELSO Guidelines for Adult Respiratory Failure, the suggested threshold value of PaO
2/FiO
2 (PF) ratio for ECMO institution is 100 mmHg or less [
4]. However, under the inclusion criteria based on PF ratio, patients with a relatively slow-progressive disease may experience a significant escalation in the driving force of MV before their PF ratio can finally meet the threshold value for ECMO [
5]. Since the therapeutic goal of respiratory ECMO is to reduce the negative influence of MV on the success of adult respiratory ECMO, the starting point of respiratory ECMO should also take the determinant ventilator parameters into consideration. Therefore, the study was aimed at identifying the baseline ventilator parameters which were independently associated with hospital mortality in patients receiving VV-ECMO for severe ARDS.
Discussion
This study revealed that the duration of MV before ECMO institution was the only baseline ventilator parameter which was independently associated with hospital mortality in non-trauma patients receiving VV-ECMO for severe ARDS, although the mechanism of how a prolonged duration of MV could jeopardize the survival in these ECMO-treated patients was still unclear. In fact, this duration is a reciprocal measure of the declining rate of PF ratio from the value obtained at the beginning of MV to the given threshold value for ECMO. Clinically, this declining rate of PF ratio is significantly affected by the etiology of ARDS, patient characteristics, and the institutional experience on advanced modes of MV, as the conclusion of CESAR trial [
11,
13]. These uncertainties make the suitable duration of MV before ECMO individualized among institutions. Nevertheless, limitation of this duration should still be important to adult candidates of respiratory ECMO. According to Table
2, regardless of the value of initial PF ratio, most of our patients were found to have severely non-compliant lungs at the beginning of MV. The mean PC
dyn measured at the beginning of MV was 22 cm H
2O in all of the patients, which was only accounted for 10% or less of the normal value [
14]. This finding implied that these patients were very sensitive and vulnerable to the cyclic pulmonary manipulation of MV [
15‐
17]. When the disease progresses and involves more pulmonary segments, clinicians often need to open the collapsed segments with an increased driving pressure of MV to maintain an acceptable blood-gas exchange. This attempt of lung recruitment may considerably increase the amount of dead space ventilation rather than effectively improve the blood-gas exchange, since the local perfusion of the distended segments may drop, as demonstrated by Gattinoni et al. [
18]. The over-distended lungs may also increase the intra-thoracic pressure and compromise the cardiac output [
19]. Therefore, if available and not contraindicated, VV-ECMO combining lung-protective MV is a valuable strategy to reduce pulmonary manipulation and reverse some of the pulmonary segments performing dead space ventilation under high-pressure MV to the segments performing effective blood-gas exchange under a reduced inspiratory pressure. However, the benefit of VV-ECMO is often small in patients with prolonged ventilation (often >7 days) or severe multiple organ dysfunctions [
2]. From this viewpoint, it should be important for ECMO centers to have a practical tool to determine the starting point of respiratory ECMO among candidates with different duration of MV. Technically, there are three common ways to reduce the risk of a prolonged MV before ECMO institution. The first is choosing an arbitrary deadline which is set according to general experiences, such as a 7-day period [
20]. The second is loosening the threshold value of PF ratio for ECMO from 100 mmHg to 150 mmHg, as per the suggestion of ELSO Guidelines. The third is creating a risk assessment model for a multi-axial evaluation, as is our choice here.
What interested us was that the baseline PIP could not be identified as a prognostic predictor of adult respiratory ECMO in the current study, which was different from the suggestion of RESP score. The RESP score is derived from a retrospective analysis of 2355 adult patients in ELSO’s registry and reveals that the baseline PIP, with a threshold value of 42 cmH
2O, is a predictive factor for hospital mortality in adult patients receiving respiratory ECMO [
11]. However, in the current study including 106 patients, the patients with a baseline PIP > 42 cm H
2O (
n = 14) showed a significantly lower hospital mortality rate than the patients with a baseline PIP ≤ 42 cm H
2O (21% vs. 58%,
p = 0.02). We thought that the discrepancy in sample size between the two studies should have some connection to this unexplained result. In the original study of RESP score [
11], the odds ratio of baseline PIP for hospital survival is close to 1.0 (0.992). Furthermore, the medians of baseline PIP were surprisingly both the same (36 cmH
2O) in the survivors (
n = 1338) and non-survivors (
n = 1017). Therefore, the baseline PIP might not be a very suitable criterion to initiate respiratory ECMO. Some researchers suggest that the baseline plateau pressure (P
plat) is also a valuable indicator used for this purpose [
21]. Although P
plat is a better airway pressure than PIP for measuring the pressure applied to lung itself during MV, we were unable to reproduce the above-mentioned result because the data of P
plat were severely incomplete in this retrospective study due to unknown reason.
Although the impact of the baseline pressure settings of MV remains equivocal on the outcomes of adult respiratory ECMO, an unreduced static or dynamic driving pressure of MV during the first three days of adult respiratory ECMO is recently reported to be a predictor of hospital mortality in these ECMO-treated patients [
22,
23]. When the results of these updated investigations on adult respiratory ECMO are reviewed together, researchers may find that there seem to be some links among the duration of MV before ECMO institution, the driving pressure of MV during ECMO, and the outcomes of ECMO. According to Tables
2 and
3, all of the patients showed an improvement in arterial oxygenation and a downgrade of driving pressure of MV during the first 24 h of ECMO institution. Since the collapsed alveoli should be difficult to be re-opened by a decreased driving pressure, we could assume that the improvement in arterial oxygenation should depend on the oxygenator and the non-collapsed alveoli. However, as shown in fig.
3, only patients maintaining the trend of improvement in arterial oxygenation throughout the ECMO support could eventually wean ECMO and survive. It is notably that the pulmonary compliance remained impaired in the survivors after a successful rescue of ECMO. These findings implied that the size of the non-collapsed pulmonary segments before ECMO should be one of the key contributors to the patient’s progression on ECMO and the outcome. To keep the non-collapsed pulmonary segments opened under a reduced driving pressure during ECMO, the decrease in the driving pressure during the early support of ECMO was achieved by a reduction in PIP rather than in PEEP. Actually, choosing a moderate high PEEP is also an important issue in patients treated with lung-protective MV [
17]. Despite not used in this study, an esophageal balloon is considered to be a useful tool to measure the serial changes of transpulmonary pressure and suggested an appropriate level of PEEP and driving pressure in this scenario [
24]. In general, ventilating the patients with an unreduced driving pressure is not an effective solution to improve hypoxemia on VV-ECMO, since this behavior means that the clinicians still want to recruit the collapsed segments. In cases with a refractory hypoxemia during VV-ECMO, the circuit of ECMO should be rearranged to a hybrid configuration which may promptly correct the arterial hypoxemia and allow a decrease in driving pressure [
25]. However, significant hypoxemia may recur during the weaning process of ECMO in any configuration if the pulmonary parenchyma is extensively damaged and the residual reservoir is below the requirement for surviving without ECMO.
The major limitations of this study are its retrospective design and moderate sample size. This study did not provide a comprehensive discussion of adult respiratory ECMO since only non-trauma patients treated with pressure-controlled ventilation and VV-ECMO in a single center were included. Further collaborative and prospective studies that adopt new techniques to obtain novel respiratory parameters in a large cohort of ECMO-treated ARDS patients are necessary to validate our hypothesis and determine the influences of baseline ventilator parameters on outcomes of adult respiratory ECMO.