Numerous relatively recent texts and articles have used the term “transpulmonary pressure” or “Pl” to describe Palv − Ppl, that is, Pel(L), the elastic recoil pressure of the lung tissue (i.e., without considering the pressure drop down the airway). Among these are John West’s Respiratory Physiology, the Essentials, Weinberger’s Principles of Pulmonary Medicine, Schwartzstein’s Respiratory Physiology, a Clinical Approach, and other current reviews and journal articles (9–13). In these examples, Pl and transpulmonary pressure were used to describe the forces distending lung tissue and/or the mechanical stresses applied to the lung tissue during mechanical ventilation, uses that are conceptually consistent with the traditional approach, albeit using different terms. Is this alternate definition of transpulmonary pressure, then, simply a minor difference in the use of terms without significant consequence, or does it lead to miscommunication and confusion about the interpretation of measurements and the meaning of Pl and Pel(L)?

In the traditional approach, both transpulmonary pressure and elastic recoil pressure are useful concepts that have distinctly different meanings. We believe that the use of “transpulmonary pressure” to denote both Pao − Ppl and Palv − Ppl has resulted in confusion. Here, we first present arguments for using the traditional definitions of “transpulmonary pressure” and “elastic recoil pressure of the lung,” then present published ideas apparently based on misunderstandings of limiting assumptions underlying these definitions.

As noted above, Pl and Pel(L) are both useful concepts that need clear and distinct definitions. Pl is the pressure exerted across the entire lung, including the airways, and depends on both respiratory airflow and resistance, and lung volume and compliance (or 1/elastance). On the other hand, Pel(L) is the pressure exerted across the lung tissue only, and depends on lung volume and elastance (the inverse of compliance) only; it is independent of respiratory airflow and resistance. If we use transpulmonary pressure to denote Pel(L), it is not clear what term should be used for the pressure across the entire lung (Pao − Ppl) or the pressure drop down the airway (Pao − Palv). There is no term commonly in use for the concept of Pl (Pao − Palv) when transpulmonary pressure is assigned to the elastic recoil pressure of the lung, Pel(L), and both terms are needed.

Pl and Pel(L) both vary continuously in time. However, Pel(L) cannot be measured continuously, because it requires determination of Palv, which can only be measured when there is no pressure drop down the airway (i.e., when flow is zero and the intrapulmonary airways are open, assuring that Pao = Palv). If the term transpulmonary pressure was defined as Palv − Ppl there would be no label for Pao − Ppl when it is measured continuously as Pao − Pes.

It is often assumed that when respiratory airflow is zero, Pao = Palv and therefore Pl = Pel(L). However, the measurement of Pl under such static conditions does not guarantee that the pressure measured will be representative of Pel(L). For example, when the intrapulmonary airways are obstructed or closed, as is often the case at very low lung volumes or in severe lung disease, or the alveoli are filled with liquid or foam (Figure 1), Pao differs from the local Palv, and the measurement of static Pl may include a large pressure drop down the (occluded) airway. Under these conditions, Palv and thus Pel(L) of the affected lung regions may not be uniform or measurable.

We agree with the widely held assumption, that the lung tissue cannot substantially resist compressive forces [i.e., that, locally, Palv is never substantially less than Ppl and thus Palv − Ppl = Pel(L) ≥ 0]. However, when this assumption about Pel(L) is incorrectly applied to Pl, estimates of positive pleural pressures or negative transpulmonary pressure (Pao − Ppl) are assumed to be in error. For example, Wikipedia defines transpulmonary pressure as Palv − Ppl and states, “Normally, the pressure within the pleural cavity is slightly less than the atmospheric pressure...” (https://en.wikipedia.org/wiki/Intrapleural_pressure). However, during normal passive exhalation, Ppl is positive at all volumes above the chest wall relaxation volume, and Ppl is often positive during supine expiration, during active expiration such as with high minute ventilation, or when expiratory airway resistance is increased.

Most evidence suggests that in the upright posture the pressure in the distal third of the esophagus approximates average pleural pressure. (Here, the “average” pleural pressure is taken to mean that pressure which, if applied to the whole pleural surface, would result in the same observed lung volume and total flow.) However, it is known that esophageal pressure exceeds average Ppl in the supine posture by 3 to 7 cm H₂O (14), an artifact attributed to the weight of the mediastinal contents that biases Pl estimates toward lower values. This bias is the reason some experts disregard absolute esophageal pressure measured in supine patients with injured lungs (15).

The lung is an elastic network structure, which is deformed by surface tension, gravity, and shape constraints imposed by the thorax. Therefore, pleural pressure (i.e., pressure on the visceral surface of the lung) is nonuniformly distributed and may vary considerably over different regions of the lung. Because the lung parenchyma resists deformation, vigorous respiratory efforts that are associated with large changes in thoracic shape cause different changes in Ppl and Pl in different regions. This causes a partial redistribution of gas between lung regions (pendelluft), which has been recently observed in experimental animals and some patients with respiratory failure (16).

In characterizing the lung and chest wall, it is important to distinguish between elastance (reciprocal of the slope of P–V curve) and the recoil pressure (the position of P–V curve) at a specified volume (Figure 2). For example, in the clinical literature it is often implied that recumbent obese patients have a high chest wall elastance (low chest wall compliance) simply because they have greater-than-expected end-inspiratory plateau airway pressures during mechanical ventilation. The high end-inspiratory plateau pressures of the relaxed respiratory system are the sum of the elastic recoil pressures of the lung [Pel(L)] and chest wall [Pel(cw)]. Because in recumbent obese patients the chest wall is loaded by the increased mass of the abdomen, abdominal and pleural pressures are elevated, while elastic recoil pressure of the lung [Pel(L)] is often normal. It takes added pressure to displace the diaphragm/abdomen during inflation, because the chest wall P–V curve is shifted to higher pressures, not because the slope is decreased. As a result, in recumbent patients with increased abdominal mass, absolute lung volumes and Pel(L) at end-inspiration and at end-expiration are both lower than in normal individuals, irrespective of chest wall elastance, which is often normal even in severely obese subjects (18–20). In this case, the position of the P–V curve is shifted to higher pressures, but its slope may be unchanged and normal (Figure 2). Importantly, the application of PEEP to counterbalance the high intrinsic PEEP (21) may benefit obese patients by reducing their work of breathing and preventing atelectasis and atelectrauma. Therefore, it would be inappropriate to rely simply on elastance measurements, be it lung, chest wall, or respiratory system, when individualizing PEEP management. A further example is that in patients with ARDS, there is no correlation between the end-expiratory esophageal pressure, which is often very high, and chest wall elastance, which is often normal ([17] and unpublished data).

Figure 2. Pressure–volume plots connecting static end-expiratory volume (FRC) and end-inspiratory volumes in a patient with acute respiratory distress syndrome. Pao reflects the pressure across the entire respiratory system; Pes is the measured pressure across the chest wall, and Ppl-elast is the elastance-derived estimate of the pressure across the chest wall based on the assumption that Ppl = 0 when Pao = 0 (see text). The slopes of the lines represent the compliances of the respective structures. Note that the compliance is not sufficient to specify the pressure at any volume. The measured and elastance-derived estimates of pleural pressure differ by 11 cm H₂O (arrow).

Download Figure | Download Powerpoint

[More] [Minimize]

Perhaps the most important controversy involving Pl and Pel(L) is the emergence of a practice of estimating Pl or Ppl from airway pressure and the ratio of chest wall (or lung) elastance to total respiratory elastance. This practice was introduced by Gattinoni and colleagues to estimate transpulmonary pressure in critically ill patients (12) to account for restriction of the lungs by the chest wall. Gattinoni and colleagues write,

The distending force of the lung per unit area, i.e., the pressure, is that applied to the visceral pleura. This is the transpulmonary pressure (PL), which is the difference between the pressure inside the alveoli and the pleural pressure (Ppl). Unfortunately, in normal clinical practice, it is usual to consider the plateau or airway pressure (Paw) as the distending force of the lung. Under static conditions, Paw closely reflects the intra-alveolar pressure, which, in part, is spent to inflate the lung, and, in part, to inflate the chest wall. (12)

In the second sentence (our italics) they define “transpulmonary pressure” as Palv − Ppl [i.e., what is traditionally Pel(L)]. The following sentences state that airway pressure (Paw or Pao) measured statically is equivalent to alveolar pressure, implying that without airflow, Pl = Pel(L). This ignores the possibility that Pao can differ from Palv even statically when the small airways of the lung are closed or flooded. Gattinoni and colleagues (12) continue,

The elastance of the respiratory system (Ers) is the Paw required to inflate the respiratory system to 1 L above its resting position under static conditions. Indeed, Paw equals the sum of the pressure used to inflate the lung (PL) and the one used to inflate the chest wall (Ppl):

$\text{Paw}=\text{PL}+\text{Ppl,}$ (1)

and

$\text{Ers}=\text{EL}+\text{Ew,}$ (2)

where EL and Ew are the elastances of the lung and chest wall.

Accordingly,

$\text{PL}=\text{Paw}\cdot \text{EL}/\text{Ers}\text{.}$ (3) (12)

This conclusion is inconsistent with respiratory mechanics. Elastance is the reciprocal of the slope of the P–V relationship; therefore, it cannot by itself specify the elastic recoil pressure at any volume (Figure 2). However, Equation 3 states that the value of Pl can be determined without knowing the value of Pes at any volume merely by measuring Pao and the changes in pressures with tidal volume. The equations erroneously imply that when Pao = 0, Pl = 0, and by extension, when airway pressure is atmospheric, pleural pressure must also be atmospheric (22, 23).

Since its introduction, the elastance-derived estimation of Pl and Ppl has been applied in numerous studies (12, 15, 24–31). Its appeal seems to be that it avoids having to make sense of the substantially positive baseline values of Pes (and negative values of Pl) that are common in the supine position, especially in critical illness (15). The elastance-based estimation of Ppl leaves unexamined the possibility that in a large part of the lung, both Palv and Ppl are substantially greater than atmospheric pressure because the alveoli do not contain air in equilibrium with air in the central airway. This might occur with small airway closure due to severe obesity (18) or alveolar flooding in ARDS (32). In patients with ARDS, the elastance-derived values of Pl are substantially different from directly measured values, leading to different recommendations for appropriate PEEP settings (22, 26).

In a recent post hoc analysis of several clinical trials in ARDS, driving pressure, which was defined as the tidal excursion in airway pressure measured at points without airflow, was shown to predict mortality (33). The rationale for this study was that driving pressure of the respiratory system is an easily measured surrogate for cyclic stress applied to the lung. We suggest that the transpulmonary driving pressure [i.e., the tidal excursions in Pel(L) measured during end-expiratory and end-inspiratory airway occlusions] would be a better surrogate to assess the stress applied to the lung tissue, as it would exclude any contribution from the chest wall.

In summary, we urge specificity when using the terms transpulmonary pressure and elastic recoil pressure to define the physical state of the lungs, the chest wall, and the integrated respiratory system. We have provided several examples that underscore the potential for erroneous conclusions when terms such as stress, strain, and transpulmonary pressure are used in an ambiguous manner. We contend that our concern does not simply address a semantic nuance in a debate among physiologists but that it is central to the care of patients with respiratory failure.