key: cord-0970650-ucvkbb6s
authors: Protti, Alessandro; Santini, Alessandro; Pennati, Francesca; Chiurazzi, Chiara; Cressoni, Massimo; Ferrari, Michele; Iapichino, Giacomo E.; Carenzo, Luca; Lanza, Ezio; Picardo, Giorgio; Caironi, Pietro; Aliverti, Andrea; Cecconi, Maurizio
title: Lung response to a higher positive end-expiratory pressure in mechanically ventilated patients with COVID-19
date: 2021-10-16
journal: Chest
DOI: 10.1016/j.chest.2021.10.012
sha: d43a173ce97239082dd1246f41eeb550dfac5317
doc_id: 970650
cord_uid: ucvkbb6s

Background International guidelines suggest using a higher (>10 cmH2O) positive end-expiratory pressure (PEEP) in patients with moderate-to-severe acute respiratory distress syndrome (ARDS) due to the novel coronavirus disease (COVID-19). However, even if oxygenation generally improves with a higher PEEP, compliance and arterial carbon dioxide tension (PaCO2) frequently do not, as if recruitment was small. Research question Is the potential for lung recruitment small in patients with early ARDS due to COVID-19? Study design and methods Forty patients with ARDS due to COVID-19 were studied in the supine position within three days of endotracheal intubation. They all underwent a PEEP trial, where oxygenation, compliance, and PaCO2 were measured with 5, 10, and 15 cmH2O of PEEP and all other ventilatory settings unchanged. Twenty underwent a whole-lung static computed tomography at 5 and 45 cmH2O, and the other twenty at 5 and 15 cmH2O of airway pressure. Recruitment and hyperinflation were defined as a decrease in the volume of the non-aerated (density above -100 HU) and an increase in the volume of the over-aerated (density below -900 HU) lung compartments, respectively. Results From 5 to 15 cmH2O, oxygenation improved in thirty-six (90%) patients but compliance only in eleven (28%) and PaCO2 only in fourteen (35%). From 5 to 45 cmH2O, recruitment was 351 (161-462) ml and hyperinflation 465 (220-681) ml. From 5 to 15 cmH2O, recruitment was 168 (110-202) ml and hyperinflation 121 (63-270) ml. Hyperinflation variably developed in all patients and exceeded recruitment in more than half of them. Interpretation Patients with early ARDS due to COVID-19, ventilated in the supine position, present with a large potential for lung recruitment. Even so, their compliance and PaCO2 do not generally improve with a higher PEEP, possibly due to hyperinflation.

International guidelines suggest using a higher (>10 cmH2O) positive end-expiratory pressure (PEEP) in patients with moderate-to-severe acute respiratory distress syndrome (ARDS) due to the novel coronavirus disease . However, even if oxygenation generally improves with a higher PEEP, compliance and arterial carbon dioxide tension (PaCO2) frequently do not, as if recruitment was small.

Is the potential for lung recruitment small in patients with early ARDS due to COVID-19? Hyperinflation variably developed in all patients and exceeded recruitment in more than half of them.

Patients with early ARDS due to COVID-19, ventilated in the supine position, present with a large potential for lung recruitment. Even so, their compliance and PaCO2 do not generally improve with a higher PEEP, possibly due to hyperinflation.

J o u r n a l P r e -p r o o f pulmonary edema with heavy lungs, acute hypoxemia, and low compliance. 1 Computed tomography (CT) has clarified that hypoxemia depends on a large number of alveoli perfused but not aerated and low compliance on the small dimension of the ventilated lung. 2 A higher positive end-expiratory pressure (PEEP) can be used to reopen the non-aerated alveoli (anatomical recruitment) and relieve hypoxemia. 3, 4 As ventilation gets distributed in more units, compliance will probably increase, and the arterial carbon dioxide tension (PaCO2) will probably decrease. [4] [5] [6] However, in patients with a small non-aerated compartment, recruitment is modest or nil. With a higher PEEP, oxygenation can still improve via other mechanisms, including a decrease in the cardiac output, 6,7 but compliance and PaCO2 will probably not, because of alveolar overdistention. [4] [5] [6] As a general rule, the more severe the hypoxemia, the larger the alveolar collapse, the greater the probability of a positive effect of a higher PEEP on lung morphology (i.e., larger recruitment), lung function (i.e., better gas exchange and mechanics), 4 and possibly survival. 8, 9 In line with this general model and recommendations for treating ARDS of other origins, 10 international guidelines suggest using a higher PEEP (>10 cmH2O) for moderate-to-severe hypoxemia due to However, many patients with this novel disease present with less than expected alveolar collapse, 12, 13 so that their potential for recruitment may be smaller than in other ARDS. Accordingly, compliance or PaCO2 frequently worsen with a higher PEEP. [13] [14] [15] [16] [17] [18] [19] [20] These, and other data, [21] [22] [23] suggest that in COVID-19, hypoxemia is not only due to alveolar collapse and that the primary response to a higher PEEP is not always lung recruitment.

This study aimed to describe the response to a higher PEEP in patients with early ARDS due to COVID-19. We hypothesized that this is generally negative because the potential for lung recruitment is low.

This study was approved by our institutional review board (protocol 465/20). Informed consent was obtained according to local regulations.

Forty patients with laboratory-confirmed COVID-19 underwent a PEEP trial and a lung CT within three days of endotracheal intubation. Inclusion criteria were: (i) admission to our intensive care unit (ICU) with ARDS; 24 (ii) ongoing invasive mechanical ventilation with deep sedation and paralysis; and (iii) one of the authors available for collecting data. Exclusion criteria were: (i) lung CT already taken after intubation; (ii) patient too unstable for transfer to the radiology unit; and (iii) pulmonary air leak. We studied ten non-consecutive patients from 1/3/2020 to 31/5/2020, when we were frequently unavailable due to the exceptional clinical workload, and thirty consecutive ones from 16/10/2020 to 9/12/2020 ( Figure E1 ). Those with a body mass index >35 Kg/m 2 (obese) underwent a slightly different protocol than the others (see below).

Patients were studied in the supine semi-recumbent position. Following a recruitment maneuver, 4 PEEP was set at 15, 10, and 5 cmH2O. If the patient was obese, PEEP was set at 20, 15, and 10 cmH2O. Other settings were kept constant. Gas exchange J o u r n a l P r e -p r o o f and respiratory system mechanics were assessed after twenty minutes at each PEEP level.

Patients were studied in the supine horizontal position. Following a recruitment maneuver, 4 a lung CT was taken at 45 and 5 cmH2O (the first twenty patients) or 15 and 5 cmH2O (the other twenty patients) of airway pressure. If the patient was obese, CTs were taken at 45 and 10 or 20 and 10 cmH2O. The total (tissue and gas) volume, the tissue weight, and the gas volume of the whole lung and its non-aerated (density above -100 HU), poorly-aerated (from -100 to -500 HU), normally-aerated (from -500 to -900 HU), and over-aerated (below -900 HU) compartments were measured as in references 2 and 4. The expected premorbid lung weight was estimated from the subjects' height. 25 Recruitment and hyperinflation induced by any increase in airway pressure were computed as the absolute difference in total volume of the non-aerated or over-aerated compartment between 5 cmH2O (or 10 cmH2O in obese patients) and the higher airway pressure. 4, 26, 27 We used the hyperinflation-to-recruitment ratio to weigh the risks and benefits of higher airway pressure.

To be consistent with other studies on ARDS unrelated to COVID-19, 4,28 we also computed the recruited lung tissue as the difference in the non-aerated tissue weight between 5 cmH2O (or 10 cmH2O in obese patients) and the higher airway pressure and expressed it as a percentage of the lung weight with 5 cmH2O (or 10 cmH2O in obese patients). The tissue remaining non-aerated at 45 cmH2O of airway pressure was considered consolidated.

J o u r n a l P r e -p r o o f 8 The same methods were applied to ten equally spaced vertical levels forming each CT slice from the sternum to the vertebra. The pressure (super)imposed on each level was obtained as in references 2 and 29. In healthy subjects lying supine, the (maximal) superimposed pressure on the most dorsal level is 2.6±0.5 cmH2O. 25 Aiming to describe the response to a higher PEEP, we present all the results as if airway pressure had been increased throughout the study. Moreover, as we included only four obese patients, results of their PEEP trial are reported as obtained with 5, College Station, TX). A two-tailed p-value <0.05 was considered significant.

We studied forty patients with COVID-19 on invasive mechanical ventilation. Their characteristics at ICU admission are reported in Tables E1 and E2 . Thirty-three (82.5%) were males and seven (17.5%) females, with an age of 66 (59-72) years. Three (8%) had a history of chronic obstructive pulmonary disease (COPD), and four (10%) were obese. They were all transferred to the ICU for endotracheal intubation after 2 (1-5) days in the hospital. By that time, thirty-six (90%) had received some form of non-invasive ventilation. Initial C-reactive protein was 14 (8-17) mg/L. Fifteen (38%) died in the ICU.

The study was performed 1 (0-1) day after ICU admission. The lung function and morphology of all forty patients are described in Tables 1, E3 (Table E5) .

Overall, as PEEP was increased from 5 to 15 cmH2O, oxygenation improved in thirtysix (90%) patients whilst compliance only in eleven (28%) and PaCO2 only in fourteen (35%).

The quantitative analysis of lung CTs is shown in Tables 2 and 3 

The main findings of this study can be summarized as follows. In patients with early ARDS due to COVID-19, ventilated in the supine position, the response to a higher PEEP was variable. Arterial oxygenation usually improved but compliance and PaCO2 frequently did not even if lung recruitment was large. This disagreement between changes in lung physiology and anatomy can be at least partly explained by the simultaneous occurrence of hyperinflation and overdistention.

The functional response to a higher PEEP suggested a small potential for recruitment.

The arterial oxygenation quite constantly improved, but the compliance and the PaCO2 did not. An isolated increase in arterial oxygenation does not necessarily signal large recruitment. Other mechanisms can play a role. 6, 7 With a higher PEEP, the mean arterial pressure decreased, and the arteriovenous oxygen content difference increased as if the cardiac output had decreased. Arterial oxygenation may have thus increased because the non-aerated compartment became less perfused, independently of recruitment (see also Figure E5 ). 7 The decrease in compliance with a PEEP >10 cmH2O also suggests little recruitment with net overdistention. [4] [5] [6] Other authors have hypothesized the same based on a very similar response to the PEEP trial in patients with early COVID-19. 13 However, those authors did not study the morphological response to a higher PEEP, so that they could not verify their hypothesis as we did.

Discovering with lung CT that patients with early COVID-19 have a very large potential for recruitment came as a surprise. In most, 14, 18, 30 but not all, 31 other studies on COVID-19, the potential for lung recruitment was small. Herein it was 24 (14-35)%, probably larger than reported in other pulmonary ARDS (16 [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] %) 28 (see also Figure E6 ). The reasons why our findings differ from previous ones may include our use of CT, the performance of a recruitment maneuver at the beginning of the study, enrollment of patients soon after their ICU admission, before any later decrease in lung recruitability. 14, 32, 33 In our study population, the alveolar collapse was almost fully reversible (see also Figure E7 ), and the residual consolidated tissue only 16 (9-22)% of the lung weight (in other pulmonary ARDS it is 28 %). 28 J o u r n a l P r e -p r o o f PEEP can be at least partly explained by simultaneous alveolar overdistention. 26, 27 The net effect of PEEP depends on two opposite phenomena: non-aerated units regaining aeration versus already aerated units receiving more gas, up to the point of becoming over-stretched. 4, 34, 35 As PEEP was increased from 5 to 10 cmH2O, the predominant response seemed to be dorsal recruitment, with less non-aerated tissue, better arterial oxygenation, and better compliance. When PEEP was increased to 15 cmH2O, overdistention of the non-dependent lung regions possibly prevailed over any additional dorsal recruitment, with ventral overdistention at the CT and a sharp decline in compliance. Three aspects of our findings should be noted. First, CT is not ideal for measuring overdistention for the following reasons: hyperinflation can occur without overdistention, as in emphysema; 36 overdistention may develop without hyperinflation, at the interface between non-aerated and aerated units; 37 with ARDS, the decrease in CT density due to excessive inflation can be masked by the increased tissue mass. 2 With all these limitations, the decrease in compliance of the whole respiratory system (Table E5 ) and of the ventral lung levels (see Figure E2 ) suggest that overdistention developed in most patients. Second, end-expiratory lung CT underestimates endinspiratory hyperinflation. Third, all of these changes occurred with seemingly protective ventilation. In all patients but one, including those with the largest PEEPinduced hyperinflation, driving and plateau airway pressures did not exceed 15 and 30 cmH2O, not even with 15 cmH2O of PEEP. 38 Other factors may have contributed to the poor functional response to a higher PEEP in the face of large anatomical recruitment. On one side, the modest improvement in gas exchange could have been due to an abnormal distribution of pulmonary blood flow. 21-23 Arterial oxygenation will not increase much if the recruited alveoli are not well J o u r n a l P r e -p r o o f perfused. On the other side, compliance measured during tidal ventilation may not have increased with a higher PEEP because of cyclic recruitment (which increases compliance per se) with a lower PEEP. 39, 40 The relationship between changes in lung aeration and compliance is complex; recruitment should not be estimated only from the latter. 41, 42 The superimposed pressure can be defined as the hydrostatic pressure acting on each lung level. With ARDS, it increases and contributes to the alveolar collapse. 2, 29, 43 PEEP restores aeration by counteracting the superimposed pressure. [43] [44] [45] Considering that in early ARDS due to COVID-19 the lung weight gain is modest (around 250 g), the airway pressure needed to recruit the lung (the opening pressure) and keep it open (PEEP) may be quite low. If so, a PEEP >10 cmH2O will induce significant overdistention. 46 Net hyperinflation was associated with C-reactive protein and compliance. With less inflammation, there will be less pulmonary edema, lower superimposed pressure, less alveolar collapse, larger lung gas volume, and higher compliance. 5 Changes in pulmonary perfusion will play a major role in the pathogenesis of hypoxemia. [21] [22] [23] Possibly, a lower PEEP will be appropriate. By contrast, with more inflammation and lower compliance, the superimposed pressure should be higher and the balance between hyperinflation and recruitment more favorable. A higher PEEP will be more indicated.

Some of the limitations of this study deserve a comment. First, the sample size was based on feasibility limitations due to the ongoing pandemic. Some subgroup analyses were probably underpowered. Second, during the first wave of the pandemic, we could not enroll all consecutive eligible patients, which may have been a source of bias.

Third, we did not include a control group to compare patients with COVID-19 with those with other ARDS, especially for the frequency and severity of overdistention. It may be worth noting that in our study, increasing PEEP from 5 to 15 cmH2O enlarged the over-aerated compartment in all patients, including nineteen with no history of COPD (on average by 118 [53-253] ml). By contrast, in a previous study on thirty-two patients with ARDS of other origins, increasing PEEP from 0 to 15 cmH2O did the same in only fourteen (44%), and only in eight (31%) of those with no history of COPD (on average by 25 [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] ml). 27 Fourth, lung CTs were not taken at 10 cmH2O of PEEP. Our model, with predominant recruitment below that threshold and hyperinflation above it, has to be validated. Fifth, the lung phenotype in COVID-19 changes over time, 47 so that our findings may not be valid for later stages of the disease. 14, 32, 33 Clinical implications International guidelines suggest using a higher PEEP to relieve moderate-to-severe hypoxemia due to COVID-19. 11 Accordingly, among 3988 patients admitted to an Intensive Care Unit in our region (Lombardy, Italy), half were ventilated with a PEEP >12 cmH2O, and one fourth with a PEEP >15 cmH2O. 48 In retrospect, PEEP on day 1 of ICU admission was an independent risk factor of death: for any 1-cmH2O increase, mortality increased by 4%. 48 We did not study the impact of a higher or lower PEEP on clinically relevant outcomes, such as survival or duration of mechanical ventilation. Therefore, our findings do not provide any evidence on how to set the ventilation in patients with COVID-19. Even so, they suggest that the response to a higher PEEP can be hardly predictable; and that some patients might benefit from a lower PEEP, even if their ARDS is moderate J o u r n a l P r e -p r o o f lung aeration (measured with the CT or other technique) as a whole may help the clinicians to set PEEP according to the characteristics of every single patient.

In conclusion, in this group of patients with early COVID-19, ventilated in the supine position, the response to a higher PEEP was variable and usually less favorable than expected for the severity of hypoxemia and the potential for lung recruitment. Signs of hyperinflation and overdistention were common. 

Study Question: What is the response to a higher PEEP in mechanically ventilated patients with early ARDS due to COVID-19?

Results: When PEEP is increased from 5 to 15 cmH<sub>2<sub>O, oxygenation usually improves but compliance and the arterial carbon dioxide tension do not. Lung computed tomography shows that when the airway pressure is increased from 5 up to 45 cmH<sub>2<sub>O, recruitment is large but hyperinflation can be even larger.

Interpretation: In patients with early ARDS due to COVID-19, a higher PEEP can induce net hyperinflation with overdistention. Hounsfield units (HU). With a higher pressure, volumes with density above -100 HU (non-aerated) decreased, as for alveolar recruitment, while those with density from -500 to -900 (normally-aerated) increased, as for better aeration. Volumes with a density below -900 HU (over-aerated) simultaneously increased, as for hyperinflation.

Volumes with a density from -800 to -900 HU, which can become over-aerated after tidal inflation, 26 increased as well. The over-aerated compartment in some patients at J o u r n a l P r e -p r o o f Twenty patients underwent a lung CT at 5 and 45 cmH2O of airway pressure. Herein we compare the distribution of tissue and gas in their whole lungs and in their four compartments at these two airway pressures. Data are reported as median (Q1-Q3).

P-value refers to the Wilcoxon signed rank-sum test. If the non-aerated compartment had a density >0 HU (i.e., higher than the density of water), the gas volume was considered zero. Twenty patients underwent a lung CT at 5 and 15 cmH2O of airway pressure. Herein we compare the distribution of tissue and gas in their whole lungs and in their four compartments at these two airway pressures. Data are reported as median (Q1-Q3).

P-value refers to the Wilcoxon signed rank-sum test. If the non-aerated compartment had a density >0 HU (i.e., higher than the density of water), the gas volume was considered zero.

J o u r n a l P r e -p r o o f 

Acute respiratory distress in adults

What has computed tomography taught us about the acute respiratory distress syndrome?

Relationships between lung computed tomographic density, gas exchange, and PEEP in acute respiratory failure

Lung recruitment in patients with the acute respiratory distress syndrome

Pressure-volume curve of total respiratory system in acute respiratory failure. Computed tomographic scan study

How to monitor lung recruitment in patients with acute lung injury

Depression of cardiac output is a mechanism of shunt reduction in the therapy of acute respiratory failure

Higher vs lower positive end-expiratory pressure in patients with acute lung injury and acute respiratory distress syndrome: systematic review and meta-analysis

Driving pressure and survival in the acute respiratory distress syndrome

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

Surviving Sepsis Campaign: guidelines on the management of critically ill adults with Coronavirus Disease 2019 (COVID-19)

Quantitative-analysis of computed tomography in COVID-19 and non COVID-19 ARDS patients: A case-control study

Physiological and quantitative CT-scan characterization of COVID-19 and typical ARDS: a matched cohort study

Computed tomography assessment of PEEPinduced alveolar recruitment in patients with severe COVID-19 pneumonia

Recruitability and effect of PEEP in SARS-Cov-2-associated acute respiratory distress syndrome

Respiratory physiology of COVID-19-induced respiratory failure compared to ARDS of other etiologies

Potential for lung recruitment and ventilationperfusion mismatch in patients with the acute respiratory distress syndrome from Coronavirus Disease

Lung Recruitability in COVID-19-associated acute respiratory distress syndrome: a single-center observational study

Effect of positive end-expiratory pressure and proning on ventilation and perfusion in COVID-19 acute respiratory distress syndrome

Effects of positive end-expiratory pressure in "high compliance" severe acute respiratory syndrome coronavirus 2 acute respiratory distress syndrome

Pathophysiology of COVID-19-associated acute respiratory distress syndrome: a multicentre prospective observational study

Modeling lung perfusion abnormalities to explain early COVID-19 hypoxemia

Pulmonary angiopathy in severe COVID-19: physiologic, imaging, and hematologic observations

Acute respiratory distress syndrome: the Berlin Definition

Limits of normality of quantitative thoracic CT analysis

Regional distribution of gas and tissue in acute respiratory distress syndrome. III. Consequences for the effects of positive end-expiratory pressure. CT Scan ARDS Study Group. Adult Respiratory Distress Syndrome

Incidence and regional distribution of lung overinflation during mechanical ventilation with positive end-expiratory pressure

Respiratory mechanics, lung recruitability, and gas exchange in pulmonary and extrapulmonary acute respiratory distress syndrome

Vertical gradient of regional lung inflation in adult respiratory distress syndrome

Respiratory mechanics of COVID-19-versus non-COVID-19-associated acute respiratory distress syndrome

Assessment of lung reaeration at 2 levels of positive end-expiratory pressure in patients with early and late COVID-19-related acute respiratory distress syndrome

Paradoxically improved respiratory compliance with abdominal compression in COVID-19 ARDS

Paradoxical effect of chest wall compression on respiratory system compliance: a multicenter case series of patients with ARDS, with multimodal assessment

Effects of positive end-expiratory pressure and different tidal volumes on alveolar recruitment and hyperinflation

Selecting the right level of positive end-expiratory pressure in patients with acute respiratory distress syndrome

Recruitment and derecruitment during acute respiratory failure: a clinical study

Lung inhomogeneity in patients with acute respiratory distress syndrome

Tidal hyperinflation during low tidal volume ventilation in acute respiratory distress syndrome

Pressurevolume curves and compliance in acute lung injury: evidence of recruitment above the lower inflection point

Potential for lung recruitment estimated by the recruitment-to-inflation ratio in acute respiratory distress syndrome. A clinical trial

Lung recruitment assessed by respiratory mechanics and computed tomography in patients with acute respiratory distress syndrome. What is the relationship?

The recruitability paradox

Compressive forces and computed tomography-derived positive end-expiratory pressure in acute respiratory distress syndrome

Regional effects and mechanism of positive end-expiratory pressure in early adult respiratory distress syndrome

Effects of positive end-expiratory pressure on regional distribution of tidal volume and recruitment in adult respiratory distress syndrome

Barotrauma in mechanically ventilated patients with Coronavirus disease 2019: a survey of 38 hospitals in Lombardy

Radiological findings from 81 patients with COVID-19 pneumonia in Wuhan, China: a descriptive study

Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy