Effects of Lung Injury and Abdominal Insufflation on Respiratory Mechanics and Lung Volume During Time-Controlled Adaptive Ventilation

Abstract

BACKGROUD:

Lung volume measurements are important for monitoring functional aeration and recruitment and may help guide adjustments in ventilator settings. The expiratory phase of airway pressure release ventilation (APRV) may provide physiologic information about lung volume based on the expiratory flow-time slope, angle, and time to approach a no-flow state (expiratory time [T_E]). We hypothesized that expiratory flow would correlate with estimated lung volume (ELV) as measured using a modified nitrogen washout/washin technique in a large-animal lung injury model.

METHODS:

Eight pigs (35.2 ± 1.0 kg) were mechanically ventilated using an Engström Carescape R860 on the APRV mode. All settings were held constant except the expiratory duration, which was adjusted based on the expiratory flow curve. Abdominal pressure was increased to 15 mm Hg in normal and injured lungs to replicate a combination of pulmonary and extrapulmonary lung injury. ELV was estimated using the Carescape FRC INview tool. The expiratory flow-time slope and T_E were measured from the expiratory flow profile.

RESULTS:

Lung elastance increased with induced lung injury from 29.3 ± 7.3 cm H₂O/L to 39.9 ± 15.1cm H₂O/L, and chest wall elastance increased with increasing intra-abdominal pressures (IAPs) from 15.3 ± 4.1 cm H₂O/L to 25.7 ± 10.0 cm H₂O/L in the normal lung and 15.8 ± 6.0 cm H₂O/L to 33.0 ± 6.2 cm H₂O/L in the injured lung (P = .39). ELV decreased from 1.90 ± 0.83 L in the injured lung to 0.67 ± 0.10 L by increasing IAP to 15 mm Hg. This had a significant correlation with a T_E decrease from 2.3 ± 0.8 s to 1.0 ± 0.1 s in the injured group with increasing insufflation pressures (ρ = 0.95) and with the expiratory flow-time slope, which increased from 0.29 ± 0.06 L/s² to 0.63 ± 0.05 L/s² (ρ = 0.78).

CONCLUSIONS:

Changes in ELV over time, and the T_E and flow-time slope, could be used to demonstrate evolving lung injury during APRV. Using the slope to infer changes in functional lung volume represents a unique, reproducible, real-time, bedside technique that does not interrupt ventilation and may be used for clinical interpretation.

Introduction

ARDS is characterized by a loss of functional lung volume due to air space edema and atelectasis and thus may lead to hypoxemia.¹ The loss of functional residual capacity (FRC) can be decreased by as much as 45% in severe ARDS.² Under normal conditions, FRC is determined at end expiration by the balance of forces between the inward elastic recoil of the lung and outward recoil of the chest wall.³ FRC can be measured as the end-expiratory lung volume (EELV) in spontaneously breathing patients. However, in mechanically ventilated patients, EELV will be equal to FRC only when end-expiratory flow PEEP is zero.⁴ When PEEP is applied to a ventilated patient, EELV will generally be greater than the FRC.

The measurement of EELV during mechanical ventilation at a specified PEEP provides a physiologic assessment of developing lung injury and may be more sensitive than changes in respiratory system compliance.^5-6 Thus, it may be useful for monitoring a disease course, to determine if the lung is being recruited,⁷ and perhaps to guide the setting of optimal PEEP⁴ and tidal volumes (V_T).⁸ EELV has been measured using computed tomography scan volumetry^9-11 and gas dilution techniques^12-16 and has been inferred using static pressure/volume curves and inflection points^17-19 and electrical impedance tomography.^20,21 Despite the value of monitoring EELV, it is not routinely assessed in mechanically ventilated patients since currently available techniques are logistically challenging; require transport; involve radiation; cannot be measured repeatedly; or are costly, inaccurate, or interfere with ventilation.^22-24 More recently, a modified nitrogen washout/washin technique of measuring EELV has been validated using a commercially available ventilator.¹¹

One potential ventilator mode for which the EELV might be inferred is airway pressure release ventilation (APRV). The APRV expiratory phase duration using the time-controlled adaptive ventilation (TCAV) method may provide physiologic information about functional lung volume, although this has not been validated clinically or experimentally. The rate at which the lung decompresses during expiration should be related to the severity of lung injury. For example, a steep decay in the expiratory flow may indicate increased lung or chest wall elastance and/or decreased functional lung volume. This rate of decay may be represented by the slope and the acuity of the expiratory flow-time angle (φ), as well as the time it takes for the lung to reach a zero-flow state (T_E). This experiment aimed to validate whether these elements of the expiratory flow curve (slope and T_E) in APRV can be used as physiologic markers of lung injury severity and lung volume alterations during mechanical ventilation. The experiment was designed using pulmonary and extrapulmonary models of lung injury, with combinations of detergent instillation and abdominal hypertension. The APRV expiratory flow curve parameters (slope and T_E) were compared against the GE Carescape R860 FRC INview tool for measuring lung volume.^2,11,25,26 The use of slope and/or T_E as a surrogate for lung volume represents a unique, reproducible technique that does not halt ventilation and is available on most mechanical ventilators and thus may be exploited for clinical interpretation.

QUICK LOOK.

Current knowledge

Lung injury leads to a loss of effective lung volume as a result of atelectasis and edema. This loss of lung volume can be prognostic and inform ventilator adjustments. Several methods of estimating effective lung volume have been proposed, but there are limited rapid and cost-effective methods that can be performed at the bedside.

What this paper contributes to our knowledge

Increasing intra-abdominal pressure with insufflation created a combined pulmonary and extrapulmonary lung injury in a porcine model. The slope of the expiratory flow in airway pressure release ventilation set according to the time-controlled adaptive ventilation method correlated with increasing lung injury as well as the estimated lung volume established using a nitrogen washout/washin method. The expiratory flow slope, therefore, represents a method for monitoring loss of effective lung volume.

Methods

All experiments were conducted with approval from the State University of New York Upstate Medical University Institutional Animal Care and Use Committee in accordance with Animal Research: Reporting of In Vivo Experiments guidelines. Eight female and male Yorkshire pigs (35.2 ± 1.0 kg) were anesthetized with an intravenous ketamine (90 mg/kg) and xylazine (10 mg/kg) solution. A 7.5-Fr endotracheal tube (Sheridan, Teleflex, Morrisville, North Carolina) was placed through a tracheostomy and connected to an Engström Carescape R860 mechanical ventilator (GE Healthcare, Madison, Wisconsin). A central venous catheter was placed in the external jugular vein for fluid and medication administration. The right carotid artery was cannulated for arterial blood pressure monitoring and blood gas measurements (cobas b 221, Roche, Basel, Switzerland). An esophageal balloon-tipped catheter (CooperSurgical, Trumbull, Connecticut) was placed into the distal esophagus to measure esophageal pressure to calculate lung and chest wall elastance. A cystostomy was performed for bladder pressure measurement and decompression.

APRV was selected as the mode of use for this study because the inspiratory and expiratory times are adjusted independently without being restricted to a specific inspiratory to expiratory ratio. An expiratory release pressure (P_Low) of 0 cm H₂O, as clinically suggested,²⁷ provides an unobstructed picture of the expiratory flow curve. The expiratory phase can then provide physiologic information about the lung by analyzing the lung collapse time and the acuity of the expiratory flow-time angle, which may be reflective of lung volume and stiffness, just prior to the beginning of exhalation.

Animals were paralyzed with rocuronium (0.01–0.02 mg/kg/min) and placed on the ventilator with the following settings: P_High 25 cm H₂O, P_Low 0 cm H₂O, T_High 4.0 s, and F_IO2 0.30. These settings were held constant throughout the experiment (except during lung injury) to determine the isolated effects of expiratory flow and duration after lung injury and with increased abdominal pressure. The T_Low was adjusted according to the expiratory flow curve as previously described.²⁷ Briefly, the peak expiratory flow (PEF) was measured and used to determine when to terminate expiratory flow (TEF) by multiplying PEF x 75% (PEF x 75% = TEF).

Lung volume was measured using the FRC INview module integrated within the Carescape as described by other authors.^4,25 Briefly, nitrogen washout and washin rates were calculated by measuring end-tidal O₂ and CO₂ during a 10% change in F_IO2. In this system, the measurements are recorded twice, and the average of the washout and washin rates are computed provided the difference between the 2 values is < 20%. Since the expiratory flow does not decay to zero during TCAV, the estimated EELV based on nitrogen washin/washout rates may be biased. Thus, we will hereafter refer to the measured lung volume as the estimated lung volume (ELV) to account for this. Before each ELV measurement, hemodynamic and ventilator parameters were also recorded.

The lung T_E was determined as the length of time required for the lung to deflate to its resting lung volume, as indicated by a flow of 0 L/s (Fig. 1). A prolonged T_E may indicate an obstructive lung process, whereas a shortened T_E may indicate restrictive lung pathology. The expiratory flow-time angle (φ) generated by the PEF and expiratory flow curve is subject to change depending on the aspect ratio of the flow-time curve; therefore, the slope was used for interpretation. The slope was calculated as follows:

$$\text{Slope}=\frac{(\mathrm{P}\mathrm{E}\mathrm{F}-\mathrm{T}\mathrm{E}\mathrm{F})}{{\mathrm{T}}_{\text{Low}}}$$ Equation 1:

Fig. 1.

Expiratory flow in airway pressure release ventilation. After the release of the inspiratory phase, the greatest absolute flow represents the peak expiratory flow (PEF). The expiratory flow is terminated after a time (T_Low) that is adjusted to terminate expiratory flow (TEF) at 75% of PEF. The expiratory flow-time slope and resulting angle φ determine the collapse time of the lung (T_E) where the expiratory flow approaches 0 L/s. In this illustration, PEF is −0.8 L/s and TEF is −0.6 L/s with a T_Low of 0.5 s. The calculated slope is, therefore, 0.4 L/s². A steeper slope or more acute φ leads to a shorter T_E, which requires adjusting the T_Low to preserve estimated lung volume, as compared with a shallower slope or more obtuse φ with a longer T_E, allowing for a longer T_Low.

Figure 1 is an example hypothetical PEF of −0.8 L/s, with TEF of −0.6 L/s and T_Low of 0.5 s. The calculated slope is, therefore, 0.4 L/s².

Lung strain is another marker of progressive lung injury and has been postulated to be a key factor in the development of ventilator-induced lung injury yet cannot be easily measured clinically.⁸ We calculated lung strain by normalizing the V_T of the TCAV breath to ELV:

$$\text{Lung strain}=\frac{{\mathrm{V}}_{\mathrm{T}}}{\mathrm{E}\mathrm{L}\mathrm{V}}$$ Equation 2:

To simulate intra-abdominal hypertension and induce increased chest wall elastance, a 12-mm trocar was placed into the abdomen via the umbilicus. The abdomen was then insufflated with CO₂ to pressures of 6, 8, 10, 12, 14, and 15 mm Hg.

Following these baseline measurements, the animals were transitioned to CPAP at 20 cm H₂O prior to bronchoscopic detergent administration (Sigma-Aldrich, Saint Louis, Missouri). TWEEN leads to surfactant dysfunction, simulating this well-known component of ARDS pathophysiology.²⁸ The airway was visualized with a bronchoscope, which was advanced down the right main bronchus past the right middle lobe bronchus, following which a 3% TWEEN-20 detergent solution (0.375 mL/kg) was instilled, as previously described.^29,30 The bronchoscope was withdrawn to the carina and then advanced down the left main bronchus just past the left cranial lobe bronchus, and another (0.375 mL/kg) TWEEN-20 dose was administered. This model induces a heterogeneous lung injury targeting the diaphragmatic lobes. The animals were then ventilated for 10 min on a volume control mode with V_T 10 mL/kg, PEEP 0 cm H₂O, and F_IO2 0.30. An arterial blood gas was obtained prior to and after detergent instillation to document lung injury by P_aO2/F_IO2. Following lung injury, the abdomen was again insufflated to 6, 8, 10, 12, 14, and 15 mm Hg. A 10-min acclimatization period was implemented following each insufflation pressure change. Since both TWEEN injury and abdominal insufflation may induce systemic hypotension, an intravenous bolus of lactated Ringer’s solution was administered prior to detergent instillation.

Lung, chest wall, and total respiratory system elastance values along with intra-abdominal pressures (IAPs) were recorded at baseline and at each stepwise increase in insufflation pressure before and after lung injury. Each of the expiratory flow parameters (flow-time slope, PEF, TEF, T_E, and T_Low) were also measured. T_E was measured by manually extending the duration of T_Low until the expiratory flow curve approached 0 L/s. With each increase in insufflation pressure, the T_Low was adjusted to maintain TEF at 75% of PEF to generate a time-controlled PEEP.

Statistical Analysis

Repeated-measures analysis of variance was used to determine differences between groups before and after lung injury, with increasing IAP for continuous parameters with a post hoc Tukey test if significance was found in the condition × IAP effect. Pearson correlation coefficient (ρ) was used to determine the strength of the relationship between ELV and the flow-time slope and T_E based on the expiratory flow and similarly used to confirm a correlation between the set intra-abdominal insufflation pressure and measured bladder pressure. A t test was used to compare P_aO2/F_IO2 and V_T before and after induced lung injury. A P value of < .05 was considered significant. Data are presented as mean ± SD. Statistical analysis was performed using Prism 9 (GraphPad by Dotmatics, Dotmatics, Boston, Massachusetts).

Results

Baseline P_aO2/F_IO2 in animals prior to intra-abdominal insufflation and Tween lung injury was 489 ± 17.9 mm Hg. TWEEN instillation led to mild-moderate hypoxemia, with a significant decrease in the P_aO2/F_IO2 to 220 ± 11.7 mm Hg (P < .001). Following lung injury, the generated V_T were significantly decreased from 8.3 ± 0.8 mL/kg/breath to 6.3 ± 1.5 mL/kg/breath (P = .02). The normal and injured lungs had a similar total decrease in V_T (normal 75 ± 50 mL and injured 77 ± 50 mL) with increasing IAPs to 15 mm Hg (P = .45) (Table 1). Despite increasing IAP and lung injury, heart rate did not change appreciably (P = .90) (Table 1).

Table 1.

Pulmonary and Hemodynamic Parameters in Normal and Injured Lung With Varying Insufflation Pressures

As IAP increased from 6 mm Hg to 15 mm Hg under baseline conditions, T_Low decreased from 0.54 ± 0.07 s to 0.46 ± 0.04 s such that the TEF/PEF ratio could be maintained at 75% (Table 2). After lung injury, the required T_Low was 0.46 ± 0.07 s and decreased to 0.34 ± 0.05 s with increasing IAP to 15 mm Hg. These adjustments were necessary to maintain a time controlled PEEP, which was 11.9 ± 2.0 cm H₂O in the normal lungs at baseline and remained between 11.9–14.4 cm H₂O with increasing IAP, despite a P_Low set at 0 cm H₂O (Table 1). In the injured lungs, the time controlled PEEP was lower than in the normal lung (P = .004) but remained between 10.4–13.3 cm H₂O with increasing IAP. With a P_High set at 25 cm H₂O, the driving pressure was found to be 12.9 ± 1.8 cm H₂O and 11.6 ± 1.8 cm H₂O before and after lung injury, respectively, with no added abdominal pressure (Table 1). With increasing IAP, the average driving pressure was higher in the injured lung group (P = .001 vs normal lung) but never > 12.9 cm H₂O in the normal lung or > 13.8 cm H₂O in the injured lung group.

Table 2.

Pulmonary Parameters Relevant to Estimated Lung Volume in Normal and Injured Lung With Varying Insufflation Pressures

Bladder pressures were measured to validate the set insufflation pressure and were found to be within 1 cm H₂O of each other at each generated insufflation pressure, with strong positive correlation in both normal (ρ = 0.89) and induced lung injury (ρ = 0.91) (Table 1). With increasing IAPs, chest wall elastance increased from 15.3 ± 4.1 cm H₂O/L to 25.7 ± 10.0 cm H₂O/L under baseline conditions. Following lung injury, chest wall elastance increased from 15.8 ± 6.0 cm H₂O/L to 33.0 ± 6.2 cm H₂O/L (Fig. 2 and Table 1). With increasing IAP to 15 mm Hg in the normal lung, the chest wall elastance increased to account for 53% of the total respiratory system elastance (Fig. 2A). Although there was an increase in chest wall elastance in the injured lung with increasing IAP, the contribution to the total respiratory system elastance reached a maximum of 41%, likely due to the concomitant lung injury favoring an increase in the lung elastance (Fig. 2B). Transpulmonary pressure decreased from 16.1 ± 0.9 cm H₂O to 10.6 ± 2.5 cm H₂O with increasing IAPs in the normal lung and demonstrated a similar decrease from 17.1 ± 0.6 cm H₂O to 13.9 ± 0.8 cm H₂O (P = .88) (Table 1). Since the plateau pressure was held constant, the pleural pressure was found to have a concordant increase in the normal lung from 8.6 ± 0.9 cm H₂O to 13.6 ± 2.4 cm H₂O and from 7.3 ± 0.9 cm H₂O to 10.2 ± 0.7 cm H₂O (P = .92).

Fig. 2.

Partitioning of the respiratory system elastance into lung and chest wall elastance. The contribution of lung and chest wall elastance to the total respiratory system elastance with increasing intra-abdominal pressure in A: normal lung and B: after induced lung injury. Error bars represent SD.

The average baseline lung elastance was 29.3 ± 7.3 cm H₂O/L and increased to 39.9 ± 15.1 cm H₂O/L after bronchoscopic TWEEN instillation (Fig. 2, Table 1). There was an initial decrease in lung and respiratory system elastance in both the normal and injured lungs, with increasing insufflation pressures ≤ 8 mm Hg (Fig. 2, Table 1). Thereafter, with IAPs of ≥ 10 mm Hg, both lung and chest wall elastance values increased.

PEF did not change appreciably with increasing IAP (range 0.70–0.73 L/s) nor with lung injury plus increasing abdominal pressure (range 0.63–0.66 L/s) (Table 2). The flow-time slope increased (the φ flow-time angle became more acute) with increasing abdominal pressures from 6 mm Hg to 15 mm Hg: Slope increased from 0.29 ± 0.07 L/s² to 0.58 ± 0.10 L/s² in the healthy lungs and from 0.29 ± 0.06 L/s² to 0.63 ± 0.05 L/s² in the injured lungs (Fig. 3, Table 2). This change in slope is supported by the increased chest wall elastance associated with increasing IAP, leading to a higher velocity exhalation of gas from the lung. Given the similar PEF but more acutely angled slope, there was an expected decrease in T_E as the lung deflated at a higher rate to its end-expiratory volume. T_E decreased from 2.6 s to 1.2 s in the healthy lung, with IAP increasing from 0 mm Hg to 15 mm Hg and decreased from 2.3 s to 1.0 s in the injured lung (Fig. 3, Table 2).

Fig. 3.

Changes in the expiratory time and flow. The slope (red) increased with increasing intra-abdominal pressure (IAP) in both A: normal lungs and B: after induced lung injury, leading to a more visually acute expiratory flow angle φ. The time at expiratory pressure (T_Low) was adjusted (shortened) with increasing IAP to accommodate the changing slope to maintain a termination expiratory flow/peak expiratory flow ratio of 75%. The expiratory time similarly decreased with increasing IAP. Representations of the expiratory φ flow-time angle and shortened T_Low are depicted in the green-dashed boxes. Error bars represent SD.

With lung injury, the ELV decreased and was reflected by a decrease in T_E (Fig. 4A) and increase in slope (Fig. 4B); however, the relationship with T_E was found to be linear, whereas it was curvilinear with slope. Similarly, in both the normal and injured lungs, as the IAP increased the flow-time slope became steeper (or the φ flow-time angle more acute) with a shorter T_E and concomitant decrease in ELV. There was a strong correlation between ELV and T_E in the normal lung (ρ = 0.88) and injured lung (ρ = 0.95). There was a moderate correlation between ELV and slope in the normal lung (ρ = 0.78) and the injured lung (ρ = 0.78). Lung strain increased with and without lung injury as IAP increased (P = .004) (Table 2).

Fig. 4.

Change in expiratory time (T_E) and the expiratory slope relative to estimated lung volume (ELV). Changes in A: T_E and B: expiratory slope relative to ELV with increasing intra-abdominal pressures from 0 mm Hg to 15 mm Hg in normal and injured lungs. Error bars represent SD. ELV error bars were removed for figure clarity (values reported in Table 2).

Discussion

This was a physiologic study of the relative impact of increasing chest wall elastance with increasing IAP in normal lungs and mild ARDS on ELV. Additionally, the association between this ELV and the expiratory flow parameters of the APRV waveform was studied (Fig. 1). Whereas lung injury increased lung elastance, increasing IAP in normal and injured lungs increased chest wall elastance with an associated increase in pleural pressure (Fig. 2). The expiratory flow slope and time to achieve a no-flow state were found to correlate with this ELV (Fig. 4), suggesting that they may be useful to monitor the progression of lung injury and signal the need for an adjustment in T_Low (Fig. 3).

Lung injury, and specifically ARDS, is associated with marked reductions in aerated lung volume.³¹ The FRC of patients with moderate-to-severe ARDS may be < 1 L¹² as compared with ∼3 L in healthy patients.³² This reduction in functional lung serves as a marker of progressive lung injury and has important prognostic implications that can be used to direct recruitment efforts to prevent further lung injury and reestablish lung volumes.²⁶ It follows that EELV is a primary determinant of lung oxygen capacity since oxygen and CO₂ exchange is a function of aeratable lung volume.²⁶ EELV has not been standardly incorporated into the clinical assessment of patients due to the lack of readily available, accurate, and cost-effective measurement techniques.³³ Thus, there is an appeal to finding a marker of EELV changes using a standard mechanical ventilator, which served as the premise for this study.

Abdominal insufflation ≤ 15 mm Hg is common in laparoscopic procedures^34,35 and is known to lead to increases in respiratory system and chest wall elastance, as well as respiratory system resistance.^36,37 Increased abdominal pressure can also affect expiration due to diaphragm elevation and inspiration by pushing the rib cage outward at the zone of apposition.³⁸ The pneumoperitoneum and additional patient factors such as position, height, and weight may have considerable effects on intraoperative respiratory system mechanics.^36,39 Thus, our findings are relevant to surgical patients with or without lung injury, as well as critically ill patients with extrapulmonary ARDS in whom intra-abdominal hypertension may lead to altered respiratory mechanics.^40,41

Measured bladder pressure correlated well with insufflation pressure, confirming that targeted pressures were achieved. In the direct lung injury group, there was an expected increase in lung elastance compared to uninjured lungs. Both groups demonstrated increasing chest wall elastance with increasing IAP, but this effect was more pronounced in the injured group. Lung and respiratory system elastances initially decreased with increasing insufflation pressure, which was consistent with the initial increase in exhaled V_T. This was observed until insufflation pressures of 8 mm Hg, suggesting that this is the point at which competing alveolar de-recruitment and subsequent reduction in V_T outweigh this effect. Despite increasing abdominal insufflation pressures, the PEF remained unchanged. This suggests that the increased flow expected with rising lung elastance was balanced by the overall reduction in aeratable lung and resulting V_T.

ELV can be used to estimate lung strain, which is another important factor to consider in terms of prognostication and to herald the need for ventilator adjustments where high levels of dynamic strain have been associated with mortality and lung edema.⁴² The idea of using ELV and elastance to guide PEEP has been well established to optimize recruitment and reduce strain,^43,44 particularly with increased abdominal pressure, such as during laparoscopy.⁴⁵ Decreasing PEEP in patients undergoing laparoscopy leads to decreased oxygenation and compliance and increased driving pressure and atelectasis, but there is no method of determining optimal PEEP.⁴⁶ The calculated lung strain revealed a continued increase in strain concomitant with increasing IAP. Since the plateau pressure was held constant throughout the experiment, increasing insufflation pressures led to an accompanying decrease in V_T and lower transpulmonary pressure. This combined with the correlate increase in pleural pressure led to a reduction in V_T relative to the smaller functional lung, preventing overdistention of the fewer alveoli available to participate in gas exchange.⁴⁷

Different methods have been proposed to adjust ventilator settings based on lung and/or respiratory system elastance values⁴¹ and intratidal variations in global lung mechanics.⁴⁸ These mechanical properties have been particularly relevant for determining PEEP responsiveness and so-called optimal PEEP.⁴¹ Our study found that both the expiratory flow-time slope and T_E correlate with ELV and can detect early changes in respiratory physiology. Thus, changes in the slope might inform the clinician about lung recruitability. For example, a steeper expiratory slope suggests a reduction in ELV and thus an opportunity to recruit. Furthermore, this would prompt an adjustment to T_Low to maintain a time controlled PEEP sufficient to prevent de-recruitment. In summary, we propose that the expiratory flow pattern in APRV may be added to the repertoire of physiologic information that guides adjustments to ventilator settings in real time.

APRV is available on most ventilators; therefore, finding an ELV corollary that can be readily assessed without the requirement of O₂ changes or a prolonged washout maneuver is attractive. The TCAV methodology advocates for an extended T_High to maximize mean airway pressure and alveolar recruitment without requiring a potentially injurious plateau pressure and a brief T_Low to prevent alveolar de-recruitment at expiration. Adjusting the T_Low requires a cognizant balance between allowing sufficient time for CO₂ removal while limiting alveolar collapse time. P_Low is generally set at 0 cm H₂O to allow maximal CO₂ removal, where a P_Low > 0 cm H₂O would act as a resistor for optimal ventilation. The flow-time slope of the expiratory flow curve, therefore, provides important physiologic information regarding lung elastance, and the T_Low may be adjusted using this curve. Clinically, the T_Low must be set < T_E to ensure the lung does not collapse to its FRC, leading to atelectasis and subsequent injury from recruitment/de-recruitment. Adjusting the T_Low such that the TEF is 75% of the PEF has been shown to be the sweet spot in terms of minimizing alveolar strain and heterogeneity^30,49 and maximizing ventilation.^50,51

Despite lung injury and increasing IAP with a P_Low set at 0 cm H₂O, the time controlled PEEP was 12–14 cm H₂O in the normal lung and 10–13 cm H₂O in the injured lung. This is an important finding since the plateau pressure was held constant; the driving pressure was maintained by adjusting the T_Low according to the acuity of the flow-time slope. The T_E changed on the order of seconds with increasing insufflation pressure, whereas the T_Low was adjusted at the level of the 1/100th of a second (centisecond) with increasing insufflation pressures, illustrating the precision that is required to optimize the time controlled PEEP. Maintaining PEEP in the setting of lung injury and increasing chest wall elastance is vital to prevent atelectasis and subsequent recruitment/de-recruitment injury; however, the optimal method of adjusting PEEP using conventional ventilation is unknown.^52-55 A patient-directed methodology that factors in the collapse time of the lung subunits allows for adaptive ventilator adjustments according to lung injury severity and is, therefore, ideal.

Study Limitations

The Carescape INview tool has been previously validated but not with APRV.^11,25,26 The ELV calculations are predicated on the assumption that the homogeneity of alveolar gas distribution and gas exchange/cellular metabolism is constant throughout the measurement.⁵⁶ ELV measurements may also be influenced by hemodynamic instability, variable V_T,¹⁴ and alteration in muscle tone during anesthesia and paralysis.¹² One of the advantages of using the expiratory flow waveform is that patient hemodynamics should not influence ELV calculations during mechanical ventilation in contrast to spontaneous breathing. Thus, our animals were paralyzed during the protocol to limit any artifact of spontaneous breathing on measurements. ELV was not validated in multiple settings, such as higher/lower P_High. Since the P_High influences the V_T, and the objective of the study was to quantify ELV based on the expiratory flow slope, we did not want to obscure the findings by altering multiple settings, particularly those that impact delivered V_T. Additionally, ELV and the expiratory flow slope were studied in a model of mild ARDS. Though it is expected that the results of this study could be extrapolated to moderate and severe ARDS, this increasing injury severity was not specifically studied.

Our study was a physiologic investigation and thus does not reflect clinical outcomes or how ELV may vary with progressive lung injury over time. Animal models with defined lung injury represent a controlled environment that does not fully replicate patients with heterogeneous ARDS, whereby variable respiratory resistance (eg, bronchial tone, PEF), patient physiology (eg, V_T, fluid status, body mass index), and disease severity may lead to altered ELV measurements.⁵⁷ This study used sequentially increasing IAPs to mimic the increasing chest wall elastance that develops in critically ill patients with progressive chest wall edema and intra-abdominal hypertension, which may have influenced measurements as compared with a random sequence of applied IAPs. Though caution must be given to interpreting the data, the animal model used in this study was designed to evaluate a combination of pulmonary and extrapulmonary lung injury to allow for generalization across various lung injury phenotypes.

Conclusions

Establishing a surrogate for ELV may allow clinicians to trend functional volume loss, establish earlier diagnoses of lung injury, and monitor evolving respiratory pathophysiology. The most important conclusion of this study is that multiple parameters can be assessed on a breath-to-breath basis during TCAV, including T_Low, T_E, the expiratory flow-time slope, and visual interpretation of the φ flow-time angle, to demonstrate worsening lung injury, increased lung and chest wall elastance, and predict decreasing ELV. This was a physiologic study, but the results may be extrapolated to patients in whom increasing acuity of the flow-time slope and shortening of the T_Low may indicate a need to recruit the lost functional lung volume through adjustments in APRV parameters. Since the expiratory flow-time slope informs T_Low adjustments, our findings demonstrate the ability of TCAV to adapt to individual patient lung pathophysiology.