Clinical Application of Oesophageal & Transpulmonary Pressures in Monitoring Mechanical Ventilation

Personalisation of respiratory support may improve outcomes in patients with acute respiratory failure. Transpulmonary pressure estimate by oesophageal pressure is an essential and useful minimally invasive tool to guide the setup and monitoring of mechanical ventilation by measuring respiratory pathophysiology in critically ill patients, particularly with acute lung injury. However, oesophageal balloon catheter is underused in clinical practice. This article aims to briefly review the most practical aspects of using transpulmonary pressure at a patient’s bedside.

Introduction

Ventilator-induced lung injury (VILI) is an undesirable complication of mechanical ventilation. From the ARMA trial (ARDS Network 2000), the bases of protective lung ventilation were established, and clinicians aim to keep the ventilatory strategy within safe parameters: low tidal volume (4-6 mL/Kg PBW), plateau pressure lower than 30 cm H₂O, airway driving pressure lower than 15 cm H₂O, mechanical power possibly lower than 16 J/min and optimal PEEP (collapse versus overdistension). However, airway pressure (P_aw), defined as the pressure that distends the complete respiratory system (lung and chest wall), is frequently used to calculate respiratory mechanics, but it is not the real force acting on the lung parenchyma, which eventually causes injury. Hence, the net lung distending pressure is the called transpulmonary pressure (P_L) which represents the pressure difference between the inner part of the lung (P_aw) and the lung surface (pleural pressure or P_pl) (Gattinoni et al. 2019). The equation is P_L = P_aw – P_pl. Nevertheless, P_pl is difficult to measure in the clinical setting, and oesophageal pressure (P_oes) is used as a surrogate of P_pl in this scenario; P_L can then be measured by performing oesophageal manometry (only clinically available method) during end-inspiratory or end-expiratory occlusions. Direct measurement of surface P_pl requires a surgical introduction of wafers in the pleural cavity, so the most widely available and used method to provide an indirect determination of P_pl is an air-filled latex balloon sealed over a catheter and placed in the medium oesophagus.

Physiology and Rationale

The respiratory changes in the P_oes are representative of changes in the P_pl applied to the lung surface. There is specific data, particularly from Yoshida et al. (2018a) and Tilmont et al. (2021), showing that P_oes is well correlated with P_pl and confirming the accuracy of their absolute values measured at middle lung regions in ventilated patients. However, it is important to highlight that any particular measure of P_pl does not represent a homogeneous pressure throughout the pleural space in the supine position. Instead, P_pl is regional and changes stepwise from ventral to dorsal and cranial to caudal, so the measurements vary according to the gravitational gradient, lung weight, regional lung inhomogeneities, chest wall elastance, height of the chest wall, distension of the abdomen pushing the diaphragm upwards and the weight of mediastinal organs lying above the oesophageal balloon. Otherwise, in the prone position, there is a reduction of the compressive force of the mediastinum on the dependent lung regions, and the P_oes correlates better with the P_pl throughout different points of the lungs due to a more homogeneous distribution of regional aeration and compliance between the non-dependent and dependent lung compared with supine position, despite the development of a small increase in the P_L (Boesing et al. 2022). It is an important consideration because assuming P_pl measurements are the same at different points on the pleural surface may lead to errors in calculations, clinical interpretations of data, and therapy applications.

Regarding the pressure curves, P_L becomes positive during inspiration when the lungs inflate, either through an increase in the P_aw through the application of positive pressure with mechanical ventilation or from the generation of negative P_pl (transmitted in the P_oes curve) with spontaneous breathing (or a combination of both during assisted and spontaneous modes of ventilation) (Shimatani et al. 2023). Thereby, if the catheter with the oesophageal balloon is in the correct position, end-expiratory P_L derived from P_oes reflects P_L in dependent to middle lung, where atelectasis usually predominates, and end-inspiratory P_L estimated from elastance ratio may indicate the highest level of lung stress in non-dependent lung, where it is vulnerable to VILI or patient-self-inflicted lung injury (P-SILI) and especially in patients with acute respiratory distress syndrome (ARDS).

Beyond the attempts to limit P_L as a method and the objective of protective mechanical ventilation, there is growing evidence that patients with acute respiratory failure (ARF) often have strong spontaneous respiratory effort due to hypoxaemia, hypercapnia and hyperinflammatory conditions, which generates intense negative swings in P_pl (measured with the P_oes) resultant P_L. In the presence of underlying lung injury, this negative pressure also increases the transvascular pressure (P_tv), exacerbates pulmonary oedema, and causes further injury (Shimatani et al. 2023). Finally, these phenomena lead to the development of P-SILI and hence the great value of monitoring respiratory effort continuously.

Therefore, oesophageal pressure is useful and recommended to personalise mechanical ventilation in ARDS patients as a clinical tool to guide PEEP titration, monitor lung and protective ventilation, detect and treat patient asynchrony, and titrate the support and patient's muscle effort in the setting of weaning of mechanical ventilation (Brochard 2014). The EPVent-1 trial showed that a ventilator strategy using P_oes to estimate the P_L significantly improves oxygenation and compliance compared with the current standard of care, but without reducing mortality (Talmor et al. 2008). Based on the reanalysis of the EPVent-2 trial, independent of multiorgan dysfunction, PEEP titrated to end-expiratory P_L closer to 0 cm H₂O was associated with greater survival than more positive or negative values (Beitler et al. 2019; Sarge et al. 2021). Besides, in terms of survival, there is evidence supporting the measure of ΔP_L as an equivalent predictor of outcome similar to airway driving pressure and suggests maintaining positive end-expiratory P_L in obese patients with improvement of mortality (Chen et al. 2022).

Placement and Technical Considerations

The correct placement and air-filling of the oesophageal balloon catheter is essential to achieve a good correlation with the pressure measured directly in the middle pleural. Thus, the device must be allocated in the distal third of the oesophagus as recommended in most of the physiological studies. If not accounted for, a balloon catheter that is placed too high or too low in the oesophagus will result in falsely low pressures without the presence of cardiac oscillations. In the clinical setting, our group performed three methods to confirm the correct insertion of a balloon catheter in the lower oesophagus third: 1) with the evidence of cardiac pulse artifacts on the oesophageal pressure curve; 2) with the radio-opaques marker of the balloon catheter radiologically visible by chest x-ray (Figure 1); and 3) with an occlusion test performed by gentle external chest compression, particularly in sedated patients. When the airway is occluded at end-expiration pause, changes in P_pl are transmitted to the airway through the lungs. During the occlusion test, the changes of P_oes (ΔP_oes) equal the changes of the P_aw (ΔP_aw), thus their ratio should be 1 (ΔP_oes/ ΔP_aw = 1), assuming that P_pl = P_oes. Hence, a range of up to 20% is normally considered acceptable, corresponding to ΔP_oes/ΔP_aw from 0.8 to 1.2 (Ball et al. 2024).

Otherwise, inflating the balloon catheter with the correct amount of air is important to the proper interpretations of the P_oes. In fact, balloon under-filling would result in overestimation of the P_pl, thus underestimation of the P_L, while balloon over-filling would lead to underestimation of the P_pl, thus overestimation of the P_L, especially end-expiratory P_L in both situations. There is still discussion about the appropriate methods suggested to calculate the optimal volume of balloon air-filling, but it is often difficult to translate into clinical practice. Thereby, several authors suggest titrating volume inflation individually, and most oesophageal probe manufacturers suggest inflating the balloon with a fixed amount of air, in a range from 0.5 to 4.0 mL, according to the size and elastic properties of the device. However, balloon filling involves complete balloon deflation, equilibration at atmospheric pressure via brief disconnection, injection of maximum air volume to homogenously stretch its walls and then deflation to optimum filling volume (V_best). In theory, V_best is the minimum volume at which tidal P_oes swings are maximum (Figure 2) (Jonkman et al. 2023).

In practice, two main methods exist to calculate P_L from P_oes: direct method and elastance–derived method. In the direct method, P_L is calculated as the absolute difference between P_aw and P_oes (P_L = P_aw – P_oes) at end-expiration or end-inspiration. On the other hand, the elastance-derived method uses the tidal change in P_oes (measured with circuit occlusions) to calculate the ratio between lung elastance (E_l) and respiratory system elastance (E_rs) (end-inspiratory P_L = P_plat x E_l/E_rs). Likewise, the direct method is often used to guide positive end-expiratory pressure (PEEP) titration by end-expiratory P_L in dependent lung regions, while it is believed that the elastance-derived method is the most appropriate way to estimateend-inspiratory P_L reflecting lung distending pressure of non-dependent regions.

Clinical Applications

These can be divided into two groups: patients with spontaneous respiratory effort and those without.

a. Passive Patient Without Respiratory Effort

Differentiating lung compliance from chest wall compliance

Lung compliance (C_L) is defined as (Shimatani et al. 2023):

(C_L) = TV/ΔP_L = TV/([P_plat – P_{oes end-insp}] – [total PEEP – P_{oes end-exp}])

Chest wall compliance (C_cw) is defined as (Shimatani et al. 2023):

C_cw = TV/ΔP_cw = TV/ΔP_oes = TV/(P_{oes end-insp} – P_{oes end-exp})

PEEP Adjustment

A negative transpulmonary pressure (P_L) at end-expiration leads to the collapse of lung tissue, resulting in atelectasis, increased lung heterogeneity, intrapulmonary shunting, and reduced end-expiratory volumes. To maintain alveolar patency, a positive P_L at end-expiration (PEEP_tot – P_{oes end-exp}) is required. Increasing PEEP is essential to achieve this. This strategy can be particularly beneficial in patients with ARDS and elevated P_pl due to other causes (obesity, chest wall abnormalities, pleural effusion, abdominal hypertension, ascites). However, the potential for overdistension of non-dependent regions due to spatial differences in P_pl must be considered (Jonkman et al. 2023).

The optimal benefit may be achieved by maintaining a P_L at end-expiration between 0±2 cmH₂O, compared to higher or lower values (Figure 3) (Sarge et al. 2021; Chen et al. 2022). The risk of hyperinflation of the baby lung is the most significant injury mechanism (Güldner et al. 2016). These strategies optimise the balance between lung collapse and overdistension, even in cases of unilateral lung injury (Chiumello et al. 2008).

Limiting Lung Damage

The stress applied to the lung during mechanical ventilation has traditionally been limited by P_plat and ΔP, but this does not accurately reflect the actual damage inflicted on the lung parenchyma (Chiumello et al. 2008). A more appropriate method to limit the damage caused by tidal volume and thus avoid VILI would be through the titration of P_L at end-inspiration and ΔP_L (Baedorf-Kassis et al. 2016; Grasso et al. 2012). This is particularly important in patients with elevated P_pl and ARDS.

Regional pleural variability is an important consideration due to differing levels of hydrostatic pressure from the weight of the mediastinum. An anteroposterior gradient exists in the supine position, such that in healthy patients, there is a variability of ±1.5 cmH₂O in P_es, which may be clinically insignificant in most cases. However, in patients with ARDS, this value can increase to 10 cmH₂O on average, with P_{pl, dorsal} ≈ P_oes + 5 cmH₂O and P_{pl, ventral} ≈ P_oes − 5 cmH₂O. Therefore, if mechanical ventilation parameters, including PEEP, are titrated using P_oes, the mid-zone may be targeted, with the risk of collapse in the dorsal region and overdistension in the ventral region (Ball et al. 2024).

Transpulmonary pressure (P_L) at end-inspiration (Figure 4) can be calculated by various methods, as described above: directly or using the elastance-derived method, each providing different insights.

• The direct method is calculated by the absolute difference between airway and oesophageal pressures at end-inspiration or expiration (Grasso et al. 2012). P_L at end-inspiration is calculated based on values obtained from the area where the balloon is situated, a region that typically collapses (dependent zones), with a high risk of atelectrauma and a lower risk of hyperdistension.

P_{L end-expiratory} = PEEP − (P_oes − PEEP) using an expiratory pause to account for PEEP_i

P_{L end-inspiratory} = P_plat − (P_oes − P_plat) using an inspiratory pause

• The elastance-derived method, proposed by Gattinoni et al. (2004), focuses on regions more prone to overdistension, helping to avoid it (Yoshida et al. 2018a; Grieco et al. 2017a). In passive conditions, the range would be 0.5-0.9 in critically ill patients with ARDS (Grasso et al. 2012).

P_{L end-inspiratory} = P_plat * E_L/E_rs.

cdq_L/E_rs = [(P_plat − PEEP) − (P_oes − P_plat − P_oes − PEEP)] / (P_plat − PEEP).

Elastance-derived P_L end-inspiration determines lung stress and is mathematically coupled to ΔPL, serving as a surrogate for lung strain (Chiumello et al. 2016; Grieco et al. 2017b). Monitoring P_L at end-inspiration can be precisely determined using the elastance-derived method or by applying a correction factor (+5 cmH₂O). This is crucial when assessing the risk of VILI, although applying a fixed value is simplistic and inaccurate in all scenarios (Ball et al. 2024).

It has been proposed that the direct method can be used to adjust PEEP and recruit atelectatic alveolar units in dependent zones, while the elastance-derived method can detect maximum stress at the set inspiratory pressure (Gattinoni et al. 2004). In healthy volunteers, direct measurement has proposed targets of ΔP_L<15-20 cmH₂O, <10-12 cmH₂O in ARDS, and P_{L end-insp} <20 cmH₂O. It has been suggested that this data can be used to facilitate targeted tidal volume reduction, reducing these values to safe limits (Mauri et al. 2016; Baedorf-Kassis and Talmor 2021; Pelosi et al. 2021).

Obese Patients

Patients with morbid obesity represent a unique subset, as they may be associated with increased P_pl due to the excessive load from the weight of the chest wall, often leading to reduced lung volumes (De Jong et al. 2020). These patients may have airway and plateau pressures considered unsafe, even when P_L is maintained within protective limits. Therefore, monitoring of these patients should be conducted using oesophageal pressure. P_L-guided ventilation in these patients, aiming for a positive P_L at end-expiration, resulted in higher PEEP, improved end-expiratory volumes, lung elastance, and oxygenation, while avoiding pulmonary overdistension (Pirrone et al. 2016; Fumagalli et al. 2017; De Santis et al. 2021; Liou et al. 2022; Rowley et al. 2021). Additionally, a reduction in mortality was observed in patients with BMI >30 kg/m² (Chen et al. 2022).

b. Patients With Respiratory Effort

Monitoring Respiratory Effort

Monitoring dynamic effort using the swing of P_oes (ΔP_oes) is straightforward and accessible. Although this parameter underestimates P_mus, which encompasses the effort required to mobilise both the thoracic cage and pulmonary parenchyma, it effectively measures the pressure generated by all inspiratory muscles. A physiological effort during assisted ventilation is typically reflected by a ΔP_oes of 2–12 cmH₂O and a ΔP_mus of 3–15 cmH₂O (Figure 5) (Goligher et al. 2020a; Goligher et al. 2020b). While the exact limits, particularly the upper thresholds, are not fully established, very low values may indicate excessive ventilatory assistance and, consequently, a risk of muscle atrophy. On the other hand, high values indicate excessive respiratory effort in both assisted mechanical ventilation and spontaneous breathing leading to VILI or P-SILI during weaning from mechanical ventilation (Levine et al. 2008; Goligher et al. 2015). Both pressures may be altered in the presence of expiratory muscle activity, as this can affect transpulmonary pressure (Doorduin et al. 2018; Yoshida et al. 2018b).

In cases of dynamic hyperinflation, an inspiratory effort is required to overcome PEEP_i before any volume displacement occurs. Quantification of PEEP_i requires measurement of the drop in Pes before inspiratory flow begins (Doorduin et al. 2018).

The feasibility of weaning can be assessed by monitoring respiratory effort during a weaning trial, as this allows for the evaluation of excessive inspiratory and expiratory effort. This can be achieved by measuring P_oes (Doorduin et al. 2018).

Estimating Pulmonary Stress

Dynamic pulmonary stress can be estimated through changes in dynamic inspiratory transpulmonary pressure. During an inspiratory pause at the plateau phase, dynamic P_L provides the best representation of stress in non-dependent lung regions (Yoshida et al. 2018b). The upper limits of dynamic P_L are not fully established, but a threshold of <15–20 cmH₂O has been suggested (Goligher et al. 2020b).

Evaluating Patient-Ventilator Interaction

Detecting patient-ventilator asynchronies is complex. Auto-triggered breaths, trigger delays, ineffective efforts around the cycling-off phase, reverse triggering, and early or delayed cycling-off are challenging to identify without monitoring patient effort. Additionally, triggering resulting from expiratory muscle relaxation is also difficult to detect without such monitoring (Jonkman et al. 2022).

Conclusion

Monitoring using oesophageal pressure is a valuable tool, supported by strong evidence that enables personalised ventilation with greater precision in assessing protective ventilation. This approach is particularly beneficial for patients such as those with obesity or other conditions that increase pleural pressure. Additionally, it allows for the precise quantification of respiratory effort, offering significant advantages in clinical management. Finally, to prevent atelectasis due to collapse, it has been proposed to adjust PEEP such that P_{L end-exp} is slightly positive, and this is assumed to ensure that the lung (if recruitable) is maintained open, particularly in obese patients; to prevent injury from inspiratory stretch (VILI or P-SILI), attempts are also made to limit P_{L end-insp} andΔP_L to reduce lung stress in controlled or assisted mechanical ventilation; and to prevent lung injury from excessive respiratory effort, it is suggested and recommended to guide the weaning from mechanical ventilation in assisted mechanical ventilation or spontaneous breathing using ΔP_oes.

Conflict of Interest

None.

Abbreviations

P_L: transpulmonary pressure; P_{L end-insp}: P_L at end inspiration; P_{L end-exp}: P_oes at end inspiration; ΔP_L: transpulmonary driving pressure; TV: tidal volume; P_aw: airway pressure; P_plat: plateau pressure; P_oes: oesophageal pressure; P_{oes end-insp}: P_oes at end inspiration; P_oesend-exp: P_oes at end expiration; ΔP_cw: chest wall driving pressure; ΔP_oes: oesofageal driving pressure; E_l: lung elastance; E_rs: respiratory system elastance; P_plat: plateau pressure