Acute respiratory distress syndrome (ARDS) is conceptualized as an acute diffuse inflammatory lung edema, resulting in the loss of aerated lung tissue and increased lung weight (1). ARDS affects 10% of all patients admitted to the ICU and acute mortality varies between 35% and 46%, depending on its severity (2).
The severity of ARDS is inversely related to the size of the ventilatable lung (134). Accordingly, alveolar recruitment (i.e., the opening of previously nonaerated lung units by positive pressure) is a key intervention in ARDS management (5–9). Although no study has clearly demonstrated that lung recruitment (i.e., the open lung approach) is beneficial in severe ARDS, expert opinion suggests higher levels of positive end-expiratory pressure (PEEP) and lung recruitment maneuvers (LRM) in patients with severe ARDS (1011). However, the effect of PEEP and LRM is variable and dependent on the potential for lung recruitment (PLR) (12).
In a seminal study assessing the PLR at 5 and 45 cmH2O in 68 patients (19 patients with acute lung injury and 49 patients with ARDS), Gattinoni et al (13) found that PLR varied between –9.2% and 59.3%, with a median of 9% of total lung weight. The authors found that around one quarter of total lung weight was nonaerated and nonrecruitable at 45 cmH2O. They argued that this proportion represents the ARDS “core disease”—since it was remarkably constant in the population regardless of the disease severity. Patients with higher PLR had more severe ARDS, greater lung weight, poorer respiratory compliance and oxygenation, and a worse prognosis than patients with lower PLR. Accordingly, the authors hypothesized that the PLR reflects the amount of inflammatory lung edema surrounding the “core disease” and PLR is a marker of ARDS severity.
Based on these findings, patients with the most severe ARDS forms requiring venovenous extracorporeal membrane oxygenation (ECMO) should have a high PLR. Currently, quantitative data on PLR in this population are scanty (14) but would be important, given the lack of consensus on how to ventilate patients during ECMO (15) and in order to plan future trials. In this context, our primary aim was to investigate the pattern of lung recruitability of patients with severe ARDS requiring ECMO. Furthermore, we sought to test the hypothesis that, similar to patients with less severe ARDS, patients on ECMO present high PLR and a constant nonrecruitable ARDS “core disease”. Finally, we sought to quantify the relationship between PLR and defined outcome measures (ECMO duration, ICU length of stay, and survival).
This was a single-center, observational cohort study of adult patients admitted to the severe respiratory failure and ECMO service at Guy’s and St Thomas’ Hospital, London, UK, between August 2012 and December 2013. The study had institutional approval (institutional research governance reference number, RJ114/N104). The need for individual informed consent was waived as this was a retrospective analysis of data collected prospectively for usual clinical care, with no breach of privacy or anonymity. The study qualified as a service evaluation as defined by the U.K. NHS Health Research Authority and therefore did not require review by the Research Ethics Committee (http://www.hra.nhs.uk).
Our ICU is one of the five national severe respiratory failure centers serving a population of ~10 million. Referral criteria for ECMO retrieval are a Murray lung injury score of greater than or equal to 3 (which combines degree of lung infiltration on chest x-ray, lung compliance, Pao2/Fio2, and PEEP) (16), or a pH of less than 7.20 due to hypercapnia refractory to conventional ventilation. In our practice, patients are either established on ECMO in the referring hospital (i.e., mobile ECMO) (~80% of patients) or transferred to our center on conventional mechanical ventilation for further multidisciplinary specialist assessment and management. Our retrieval process has been described previously (1718). Patients retrieved to our institution on ECMO routinely undergo a CT of the brain, chest, and abdomen immediately at admission. In addition, they receive a whole-lung “recruiting” CT scan—unless barotrauma or hemodynamic instability is present—in order to estimate the pattern of lung aeration (interstitial vs focal vs diffuse). A recruiting CT consists of the following two whole-lung CT scans: one acquired at airway pressure (CPAP) 5 cmH2O for 20–30 seconds and the second acquired at airway pressure (CPAP) 45 cmH2O for a further 20–30 seconds (13). The high-airway pressure CT scans may aid the diagnosis of the underlying lung disease, which may be concealed at lower pressures
All the patients included in this study were established on ECMO in the referring hospital (mobile ECMO) and underwent a recruitment CT scan at admission (immediately post-ECMO cannulation) or within the first 24 hours of ECMO.
In the absence of evidence from clinical trials establishing the best ventilation strategy, all patients were ventilated with the same protocol regardless of the degree of lung recruitability, using pressure-controlled ventilation, with the PEEP of 10 cmH2O, a plateau pressure of 20 cmH2O, with the frequency of 10 breaths/min for the first 24–48 hours, and subsequently, according to the treating clinician.
Acute Physiology and Chronic Health Evaluation II (APACHE II) score, Sequential Organ Failure Assessment (SOFA) score, and Murray’s lung injury score (16), as well as baseline variables, were recorded upon referral before ECMO commencement. Quantitative CT scan variables included lung volume and weight obtained at 5 and 45 cmH2O airway pressure. Outcome measures included ICU mortality and ICU length of stay. Data were retrieved from our bespoke database and from the ICU electronic patient record (CareVue; Philips Medical Systems UK Limited, Guildford, Surrey, United Kingdom).
Recruitment CT Scans
The CT scan (Philips Brilliance 40) was performed at the Radiology Department of St Thomas Hospital, London, UK. In view of the labor-intensive nature of quantitative CT, quantitative PLR was only available retrospectively. The distribution of pulmonary opacifications was classified using all CT sections according to Puybasset et al (19) by two independent observers using Osiris software (OsiriX image processing software, http://www.osirixfoundation.com, Geneva, Switzerland).
Quantitative CT analysis was performed on Maluna software (Version 3.17; Maluna, Goettingen, Germany). Analysis was performed using the extrapolation method using 10 CT sections as validated by Reske et al (20). Briefly, only the most cranial and caudal CT sections and a further eight evenly spaced CT sections between them were analyzed. The assumption of linear lung boundaries is made, and the lung volume between two CT sections is approximated to conic section. The total lung volume (Vtotal) and mass (Mtotal), the pulmonary gas volume (Vgas), and the volumes and masses of four differently aerated lung compartments were calculated voxel-by-voxel according to established methods (21). The following Hounsfield units (HU) were used to define different lung compartments: nonaerated, –100 to +100 HU; poorly aerated, –101 to –500 HU; normally aerated, –501 to –900 HU; hyperaerated, –901 to –1,000 HU. All subvolumes and masses, as well as gas volumes, were calculated as percentage of Vtotal and Mtotal, respectively.
Specific lung weight was assumed to be equal to 1, and the total lung weight was calculated from the physical density of the lung expressed in HU. The percentage of potentially recruitable lung was defined as described by Gattinoni et al (13), as the proportion of the total lung weight accounted for by nonaerated lung tissue in which aeration was restored, according to CT, by an airway pressure of 45 cmH2O from an airway pressure of 5 cmH2O. The applied formula was as follows: PLR = (NAWLow − NAWHigh)/total lung weight, where NAW is the nonaerated lung weight, Low is the low PEEP level (5 cmH2O), and High is the high PEEP level (45 cmH2O).
To ascertain difference in pre-ECMO clinical and demographical variables and differences in outcomes, patients were a priori subdivided into two PLR categories with a group with PLR below the median (PLRlow) and a group with PLR above the median (PLRhigh). To quantify the differences in the pattern of lung aeration, patients were subdivided based on PLR quartiles.
Distribution of continuous variables was assessed using the Kolmogorov–Smirnov test. Differences in baseline variables between patients with lower and higher PLR and lung weight were compared using two-tailed t-test or Mann-Whitney U test for continuous data and chi-square test or Fisher test for qualitative data. A p value of less than 0.05 was considered statistically significant except in multiple comparison procedure with adjusted p value. All analyses were carried out with SAS v 9.4 for Windows.
Forty-seven patients 44 ± 15 years old (range 18–74 yr) with severe ARDS requiring ECMO were included in the analysis. The majority of the patients were women (57%), and 85% had pulmonary ARDS (the proportion of patients with influenza was 21.7% in the PLRhigh group and 10% in the PLRlow group; p = 0.2). The APACHE II, SOFA, and Murray’s scores were similar between the two groups (PLRhigh and PLRlow) (Table 1). The median time between commencement of mechanical ventilation and ECMO was similar between the patients with low and high PLR (4.3 ± 3.8 vs 5.2 ± 4.9 d).
Potential for Lung Recruitment
Figure 1 shows the frequency distribution of patients according to the PLR. The mean PLR was 27.5% ± 19.9% of the total lung weight, with the median of 24.3% (interquartile range [IQR] = 11.4–37%) and a range of –2% to 76.3%. The PLRlow patients (defined by a PLR of < 24.3%) had a median PLR of 12.1% (IQR = 6.2–20.7%), whereas the PLRhigh patients (defined by a PLR of > 24.3%) had a median PLR of 36% (IQR = 32–54.8%).
PLR and Respiratory Variables Before ECMO Implantation
Patients in the two groups had similar Pao2/Fio2 (PLRhigh 84 ± 27 vs PLRlow 70 ± 23 mm Hg; p = 0.063) (Table 2). However, PLRhigh patients received higher PEEP (15 ± 1.6 vs 11.5 ± 2.6 cmH2O; p < 0.0001) and tidal volumes (7.2 ± 1 vs 6.4 ± 1.3 mL/kg; p = 0.016) but lower plateau (31.3 ± 2.1 vs 35.4 ± 2.8 cmH2O; p < 0.0001) and driving pressures (16.3 ± 2.9 vs 23.8 ± 4.2 cmH2O; p < 0.0001) compared with patients with low PLR. This reflected the significantly greater compliance in patients with PLRhigh (27.9 ± 5.7 vs 16.4 ± 3 mL/cmH2O; p < 0.0001). The median (IQR) compliance on day 1 post-ECMO was 15 mL/cmH2O (7.5–22.2) for the PLRhigh and 14.4 mL/cmH2O (7–21.4) for PLRlow (p = 0.5).
Differences in Lung Aeration by Quartiles of PLR
To quantify the differences in the pattern in lung aeration, patients were divided into quartiles of PLR (as percentage of total lung weight): quartile 1 (PLR = –2% to 11.3%), quartile 2 (PLR = 12.8–24.3%), quartile 3 (PLR = 24.5–37%), and quartile 4 (PLR = 37.3–76.3%). Figure 2 compares CT scans obtained at 5 and 45 cmH2O airway pressure in one representative patient for each PLR quartile. As shown, the appearance at low pressure was fairly similar among the PLR quartiles and could not predict PLR at high pressure. The results of the quantitative lung recruitment CT scan are shown in Figure 3.
Assessment of Lung Aeration with Quantitative CT Scan
We quantified the pattern of lung aeration among quartiles of PLR at 5 and 45 cmH2O airway pressures.
The amount of hyperaerated tissue was small at 5 cmH2O in all quartiles of PLR (ranging from 2.2 ± 6.1 to 0.1 ± 0.1 g from quartile 1 to quartile 4) and did not significantly increase at 45 cmH2O airway pressure (mean difference ranging from 2.9 to 6.7 g).
The quantity of normally aerated lung tissue gradually decreased from 133 ± 171 g in quartile 1 to 62 ± 49 g in quartile 4 at 5 cmH2O airway pressure. We found a statistically significant increase in normally aerated lung tissue between 5 and 45 cmH2O airway pressures in quartile 3 (120 ± 137 vs 248 ± 178 g; p = 0.004) and quartile 4 (62 ± 49 vs 239 ± 145 g; p = 0.007).
Poorly Aerated Tissue.
Similarly, the weight of the poorly aerated lung tissue increased from 140 ± 153 g in quartile 1 to 215 ± 200 g in quartile 4 at 5 cmH2O airway pressure and from 155 ± 125 g in quartile 1 to 568 ± 309 g in quartile 4 at 45 cmH2O airway pressure. The difference in the quantity of poorly aerated lung tissue between 5 and 45 cmH2O airway pressures was 230 ± 179 vs 356 ± 117 g (p = 0.055) for quartile 3 and 215 ± 200 vs 568 ± 309 g (p = 0.04) for quartile 4.
The amount of nonaerated, nonrecruitable lung tissue (i.e., “core disease”) increased from 1,686 ± 754 g in quartile 1 to 1,839 ± 488 g in quartile 4 at 5 cmH2O airway pressure, whereas it decreased from 1,587 ± 739 g in quartile 1 to 739 ± 373 g in quartile 4 at 45 cmH2O airway pressure (p < 0.001). There was a statistically significant difference in the nonaerated tissue between 5 and 45 cmH2O airway pressures in all quartiles of PLR. In particular, the weight of nonaerated lung tissue decreased from 1,686 ± 754 vs 1,587 ± 739 g (p = 0.048) in quartile 1, from 1,533 ± 670 vs 1,202 ± 561 g (p < 0.001) in quartile 2, from 1,513 ± 702 vs 1,010 ± 491 g (p < 0.001) in quartile 3, and from 1,839 ± 488 vs 739 ± 373 g (p < 0.001) in quartile 4.
PLR and Outcomes
PLRlow patients had significantly longer ECMO runs (13 vs 8 d; p = 0.013) and ICU length of stay (22 vs 15 d; p = 0.028), whereas difference in ICU mortality rates did not reach statistical significance (46% vs 24%; p = 0.159) (Table 1).
In patients with severe ARDS requiring ECMO, we found that median recruitability was 24.3% (11.4–37%) but with large variations in PLR, ranging from –2% to 76.3% despite similar lung weight. In patients with lower PLR, the disease severity and the small size of aerated lung resulted in very low compliance and high average driving pressures (23.8 ± 4.2 cmH2O). Previous studies have demonstrated that ARDS of pulmonary origin is associated with greater recruitability (13), and our patient cohort was represented mainly by pulmonary ARDS (86%). It is likely that this factor contributed to the relatively high median PLR found in our population (13).
The high median recruitability found in our cohort is in agreement with the seminal study by Gattinoni and coworkers (1314), showing that PLR increases with total lung weight, that is, with ARDS severity (13). However, while in a cohort of patients with a wide spectrum of ARDS severity from mild to severe, the proportion of nonrecruitable lung (i.e., core disease) remained constant (13), in our cohort of very severe ARDS, the nonrecruitable lung progressively decreased as the PLR increased.
In addition, we found that in severe ARDS, a higher PLR may be associated to better prognosis even when recruitment is not attempted and mechanical ventilation is not adapted to the degree of PLR. Indeed, because of the absence of clear data guiding mechanical ventilation during ECMO (22), all patients were ventilated with a standard settings regardless of PLR. Given the observational nature of our study, it is difficult to say whether patients with higher PLR would have benefitted from a more aggressive lung recruitment or higher PEEP levels. It would be tempting to speculate that in patients with severe ARDS on ECMO, a higher amount of recruitable lung may be per se associated with pathophysiologic changes in lung injury that may lead to a quicker resolution and a more favorable outcome, regardless of the ventilatory pattern.
We must acknowledge several study limitations. First, given the retrospective nature of the study and the relatively low number of patients, the association of PLR and outcome must be interpreted with caution. Second, in this—as in previous studies—the PLR was assessed following a short increase in airway pressure in one single time-point and the magnitude of PLR could have been confounded by a difference in lung history pre-ECMO and before the recruitment CT scan. Third, At variance with the seminal paper by Gattinoni et al (13), a recruitment maneuver was not performed before the CT scan. This may introduce a bias in the interpretation of the results due to the lack of volume history standardization. However, given that the rest of the ventilation and the maneuvers are standardized, it is unlikely that modifications in the lung history will be differentially distributed across all the quartiles. Finally, the lack of measurement of transpulmonary pressure makes it difficult to understand the contribution of the chest wall to the overall compliance of the respiratory system. In our opinion, however, the strengths are that patients were referred to our center from a large pool of referring hospitals. Second, we used a highly standardized and protocolized set of investigations and management, which reduces the variability and biases of individual clinicians.
In conclusion, patients with severe ARDS requiring ECMO have a wide range of lung recruitability, and greater recruitability may be associated with a more favorable outcome.
1. Ranieri VM, Rubenfeld GD, Thompson BT, et al.; ARDS Definition Task Force: Acute respiratory distress syndrome: The Berlin Definition. JAMA 2012; 307:2526–2533
2. Bellani G, Laffey JG, Pham T, et al. Epidemiology, patterns of care, and mortality for patients with acute respiratory distress syndrome in intensive care units in 50 countries. JAMA 2016; 315:788–800
3. Ferguson ND, Fan E, Camporota L, et al. The Berlin definition of ARDS: An expanded rationale, justification, and supplementary material. Intensive Care Med 2012; 38:1573–1582
4. Gattinoni L, Pesenti A. The concept of “baby lung”. Intensive Care Med 2005; 31:776–784
5. Dreyfuss D, Saumon G. Role of tidal volume, FRC, and end-inspiratory volume in the development of pulmonary edema following mechanical ventilation
. Am Rev Respir Dis 1993; 148:1194–1203
6. Dreyfuss D, Soler P, Basset G, et al. High inflation pressure pulmonary edema. Respective effects of high airway pressure, high tidal volume, and positive end-expiratory pressure. Am Rev Respir Dis 1988;137:1159–1164
7. Muscedere JG, Mullen JB, Gan K, et al. Tidal ventilation at low airway pressures can augment lung injury. Am J Respir Crit Care Med 1994; 149:1327–1334
8. Slutsky AS, Ranieri VM. Ventilator-induced lung injury. N Engl J Med 2013; 369:2126–2136
9. Webb HH, Tierney DF. Experimental pulmonary edema due to intermittent positive pressure ventilation with high inflation pressures. Protection by positive end-expiratory pressure. Am Rev Respir Dis 1974; 110:556–565
10. Fan E, Del Sorbo L, Goligher EC, et al. An official American Thoracic Society/European Society of Intensive Care Medicine/Society of Critical Care Medicine Clinical Practice Guideline: Mechanical ventilation
in adult patients with acute respiratory distress syndrome. Am J Respir Crit Care Med 2017; 195:1253–1263
11. Gattinoni L, Carlesso E, Cressoni M. Selecting the ‘right’ positive end-expiratory pressure level. Curr Opin Crit Care 2015; 21:50–57
12. Gattinoni L, Collino F, Maiolo G, et al. Positive end-expiratory pressure: How to set it at the individual level. Ann Transl Med 2017; 5:288–298
13. Gattinoni L, Caironi P, Cressoni M, et al. Lung recruitment
in patients with the acute respiratory distress syndrome. N Engl J Med 2006; 354:1775–1786
14. Caironi P, Carlesso E, Cressoni M, et al. Lung recruitability is better estimated according to the Berlin definition of acute respiratory distress syndrome at standard 5 cm H2
O rather than higher positive end-expiratory pressure: A retrospective cohort study. Crit Care Med 2015; 43:781–790
15. Camporota L, Nicoletti E, Malafronte M, et al. International survey on the management of mechanical ventilation
during ECMO in adults with severe respiratory failure. Minerva Anestesiol 2015; 81:1170–1183
16. Murray JF, Matthay MA, Luce JM, et al. An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis 1988; 138:720–723
17. Burns J, Cooper E, Salt G, et al. Retrospective observational review of percutaneous cannulation for extracorporeal membrane oxygenation
. ASAIO J 2016; 62:325–328
18. Sherren PB, Shepherd SJ, Glover GW, et al. Capabilities of a mobile extracorporeal membrane oxygenation
service for severe respiratory failure delivered by intensive care specialists. Anaesthesia 2015; 70:707–714
19. Puybasset L, Cluzel P, Gusman P, et al. Regional distribution of gas and tissue in acute respiratory distress syndrome. I. Consequences for lung morphology. CT Scan ARDS Study Group. Intensive Care Med 2000; 26:857–869
20. Reske AW, Reske AP, Gast HA, et al. Extrapolation from ten sections can make CT-based quantification of lung aeration more practicable. Intensive Care Med 2010; 36:1836–1844
21. Simon BA. Non-invasive imaging of regional lung function using x-ray computed tomography. J Clin Monit Comput 2000; 16:433–442
22. Gattinoni L, Tonetti T, Quintel M. How best to set the ventilator on extracorporeal membrane lung oxygenation. Curr Opin Crit Care 2017; 23:66–72
Keywords:Copyright © 2019 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.
adult acute respiratory distress syndrome; extracorporeal membrane oxygenation; mechanical ventilation; recruitment