The elastic pressure-volume (Pel/V) curve of the respiratory system has been suggested as a guide in lung ventilator setting in order to reduce the risks associated with mechanical ventilation in patients with acute adult respiratory distress syndrome (ARDS). The interest has focused on the lower inflection point of the curve and its relation to an optimal level of positive end-expiratory pressure (PEEP) [1-5]. Classically, the lower inflection point has been considered as the pressure required to reopen distal airways that have collapsed in dependent lung zones. Moreover, the reduced slope of the Pel/V curve in its linear part in patients with ARDS has been interpreted as an increased stiffness of the respiratory system .
Recently, the commonly held opinion that the linear segment of the Pel/V curve represents a zone of optimal distensibility, within which tidal ventilation should be confined, has been challenged . The influence of continuing recruitment during the recording of the Pel/V curve has been theoretically analysed [7,8] and animal data have been presented [9,10] suggesting that lower inflection point indicates neither the beginning nor the end of recruitment, which proceeds far beyond lower inflection point. Moreover, the expiratory pressure from which the insufflation starts has been shown to be of importance in patients with ARDS .
The aim of the present study was to evaluate the alveolar closing pressure by quantifying the derecruitment induced by progressive decreases of PEEP levels in patients suffering from ARDS. The hypothesis was that, if PEEP initially was set above the closing pressure of the alveoli, small decrements in PEEP would not result in substantial derecruitment until it had been reduced to below the closing pressure.
A comprehensive exploration of Pel/V curves recorded from different levels of PEEP was performed by modifying the constant-flow insufflation technique [4,10,12,13].
The study was approved by the Ethical Committee of the University of Napoli 'Federico II' and it was conducted according to the Helsinki principles. Informed consent was obtained from the patients' next of kin. Patients requiring mechanical ventilation of the lungs with an inspired oxygen fraction (FiO2) >0.5 for >24 h, and fulfilling the criteria for acute lung injury (ALI) , were considered for the study. Exclusion criteria were age <18 yr, presence of chest tube, known or suspected chest wall abnormalities and contraindications for sedation and paralysis. The lung injury score was computed as described by Murray and colleagues . Seven consecutive patients were studied and their characteristics are shown in Table 1.
During the study the patients were sedated with a continuous infusion of propofol (2 mg kg−1 h−1) and paralysed with pancuronium (0.1 mg kg−1). Monitoring comprised electrocardiography, invasive arterial pressure, central venous pressure and body temperature. The patients were treated according to a standard clinical protocol during the experimental period. All patients were orally intubated with cuffed endotracheal tubes (inner diameter 8-8.5 mm). The cuffs were frequently tested for air leakage. Volume controlled ventilation with a constant inspiratory flow was delivered with a Servo Ventilator® 900 C (Siemens, S-19487 Upplands Väsby, Sweden). The mean minute ventilation was 10.9 ± 1.8 L min−1, respiratory rate 14 breaths min−1. The inspiratory time was 33%, post-inspiratory pause time 5% and FiO2 >0.5. The level of PEEP was set according to a standard protocol (adjusted to maintain the FiO2 <0.6 with an arterial oxygen saturation of 89-93%) and was 8.9 ± 0.9 cmH2O. A heat and moisture exchanger and a connector to the Y-piece of the patient's breathing system were used.
The Servo Ventilator® was linked to a computer via a computer-ventilator interface. The computer was equipped with a card for analogue (A)/digital (D) and D/A conversion, and digital inputs and outputs (PC-30®; Eagle Technology, Cape Town, South Africa). The computer-ventilator interface was used for computer control of the ventilator during recording of the P/V curves [11,12]. Signals from the ventilator transducers representing airway pressure in the expiratory line, inspiratory and expiratory flow in the ventilator were A/D converted at 50 Hz.
The patients were studied (in supine position) when they were stable with respect to haemodynamics, body temperature and other clinical criteria. Some minutes before the recording of the P/V curves the PEEP level was increased to 15 cmH2O. The measurements were performed during constant low flow inflation of the lungs after a 6 s expiration as previously described . The target inspired volume was expected to result in a pressure of 45 cmH2O at zero end-expiratory pressure (ZEEP), based upon calculations during normal tidal volume breathing. For safety reasons, the insufflation was always automatically interrupted at 45 cmH2O. A modification of the method allowed the computer to automatically perform a pre-defined sequence of insufflations at five different levels of PEEP (Fig. 1). The first insufflation was recorded after a prolonged expiration at a PEEP of 15 cmH2O. The subsequent insufflations were recorded after expirations at progressively decreasing PEEP levels in steps of 3.75 cmH2O. Three breaths of normal tidal volume, and at PEEP of 15 cmH2O, separated the five insufflations. Immediately after the recordings, the flow and pressure data were automatically analysed and the corresponding Pel/V curves were plotted on the computer screen at the bedside on a common volume axis as previously described . After the study, a complete analysis was performed. Each of the five Pel/V curves was mathematically defined according to the sigmoid model of Svantesson and colleagues . According to the same model, the lower inflection point and the upper inflection point were mathematically determined from the Pel/V curve recorded from ZEEP.
Dynamic compliance (CRSdyn) was measured as the slope of the Pel/V curves at a Pel of 16 cmH2O recorded from a PEEP of 15 to 0 cmH2O. The volume recruited by different levels of PEEP was estimated measuring the volume difference between the Pel/V curve that was recorded from ZEEP and each of the other Pel/V curves (ΔVPEEP 3.75-ΔVPEEP 15). These measurements were made at 16 cmH2O (Fig. 2).
Data are expressed as mean ± standard deviation (SD). Linear regression analysis was employed.
Seven consecutive patients were studied (Table 1). All of them had a PaO2/FiO2 <27 kPa (mean = 16, SD = 4.1 kPa). The mean value of quasi-static compliance of the respiratory system was 42.6 ± 11.6 mL cmH2O−1. All patients remained haemodynamically stable through the experimental period.
In each patient the multiple Pel/V curves showed a shift towards smaller volumes when PEEP was lowered stepwise (Fig. 2). The relation between the reduction in PEEP and the volume loss observed (ΔVPEEP) was almost linear (r = 0.99; P < 0.01) (Fig. 3). Dynamic compliance, determined as the slope of the Pel/V curve at 16 cmH2O, was always higher for Pel/V curves recorded at lower PEEP values. A high correlation factor between dynamic compliance and PEEP was observed (P < 0.01) (Fig. 4).
The transition from the lower sigmoid segment to the linear part of the Pel/V curves defining the lower inflection point was in general smooth to the eye (Fig. 2). In six patients, the mathematically identified lower inflection point varied from 3.8 to 14.5 cmH2O (mean 9.2 cmH2O). In one patient, no lower inflection point was identified. No relationship was found between the lower inflection point and ΔVPEEP. When the lower inflection point was correlated to PEEP level, r = 0.88 (lower inflection point = −1.383 PEEP + 14.289).
The pressure level of the upper inflection point ranged from 20 to 26 cmH2O (mean 23.8 cmH2O). The upper inflection point increased linearly with the decreasing PEEP level. A high correlation factor between PEEP and the upper inflection point was observed (P < 0.001) (Fig. 4). The time required for the recording was on average 5 min.
The methods employed in this study for the recording of the multiple Pel/V curves were based upon computer control of a standard ventilator that is the result of a stepwise development [4,5,12,13,16]. The precise control of all events performed by the ventilator and of the measurement procedure followed a strictly defined protocol. The multiple Pel/V curves were easy to perform. There was no disconnection of the patient from the ventilator and the time required for a complete recording was approximately 5 min. The rough bedside visual analysis of the Pel/V curves made it possible for the clinician to titrate ventilation in real time.
The main results of our study were that multiple Pel/V curves recorded at gradual stepwise lowering of end-expiratory pressures showed a progressive volume loss, while compliance increased progressively. Our observations are in line with theoretical analyses [7,8], and support previous observations in animal models [9,10], normal subjects  and patients with ALI . Our finding is in fact the mirror picture of a complex phenomenon that reflects continuing recruitment that starts below the lower inflection point and proceeds above this point. The results also support observations based on computed tomography of the lungs by Gattinoni and colleagues .
At first sight, higher compliance at lower lung volumes observed in the Pel/V curves recorded at low PEEP or ZEEP may seem contradictory. A discussion of the physical and biological significance of compliance is therefore warranted. By definition compliance is volume change over pressure change. During insufflation, volume change may represent two physical phenomena: (a) distension of previously opened lung units and (b) recruitment of previously collapsed or closed units. Both affect the shape of P/V curves and thus influence compliance. In theory, when a lung unit 'pops open' at a certain pressure it has an infinitely high compliance. Accordingly, as long as units are recruited during an insufflation, they will contribute to a higher compliance than that observed at the same airway pressure in the same lung when fully recruited. Our observation that compliance increased with decreasing PEEP may thus indicate that more recruitment occurred during insufflation from lower PEEP levels compared to the higher levels.
The linear relationship between PEEP and ΔVPEEP(Fig. 3) indicates a wide range of pressures at which lung units closed in our patients. Alveolar closure begins already at an end-expiratory pressure of 15 cmH2O. Clinical protocols in our department limited PEEP to 15 cmH2O. However, other authors have observed a similar finding up to pressures as high as 20 cmH2O .
The exact significance of the presence or absence, and the nature of lower inflection point is still unknown. Computed tomography of diseased lungs has taught us that non-aerated lung areas may result from collapse of small airways or true alveolar atelectasis [18,20]. It is considered that the pressure necessary to reopen small airways is lower than that required to reopen true alveolar atelectasis. This is mainly due to the presence of a certain volume in zones distal to the collapsed small airways that reduce the opening pressure . In our patients, the transition from the lower to the linear segment of the Pel/V curve, i.e. the lower inflection point, was in general smooth to the eye. This finding, associated with the progressive volume loss at decreasing PEEP levels, is in accord with the new view on the lower inflection point that seems to be poorly correlated with the alveolar closure [9,10] and should not be considered the best parameter to set PEEP. Interestingly, Vieira and colleagues  recently observed that the presence of lower inflection point does not discriminate between a more homogeneously diseased lung and/or the need for recruitment. Moreover, the lower inflection point can be partly affected by chest wall elastic characteristics at low lung volume as observed both in pigs  and human beings [22,23], and influenced by the abdomen .
Taking all these observations into account, it seems that lower inflection point is a poor indicator of the PEEP level required to prevent alveolar collapse. The quantification of the derecruitment, which could be prevented, with the application of a certain PEEP level is a more appropriate way to set PEEP. Accordingly, PEEP should be set above the level at which a clear derecruitment is observed from a loss of volume compared to curves recorded from higher levels of PEEP. Contrary to our expectations, we found that ΔVPEEP from multiple Pel/V curves did not allow a definition of a clear expiratory pressure at which lung collapse started. Maggiore and colleagues  observed a substantial derecruitment already with the first PEEP decrement from their baseline PEEP of 20 cmH2O. Thus, it is possible that our findings are hindered by a 'too low' basal PEEP level, that prevented us from seeing the pressure level at which a clear derecruitment starts. In any case, although our data seem to indicate that derecruitment is a continuous process with no special reference to the lower inflection point, more studies are needed to clarify this pivotal point.
Some methodological problems could have affected our results and merit some discussion. Since we used the Pel/V curves as a bedside clinical tool to estimate mechanics and titrate ventilation, we did not measure oesophageal pressure that is difficult to perform reliably in the clinical setting. In order to minimize any possible bias from chest wall mechanics, patients with known or suspected chest wall abnormalities were excluded and all patients were paralysed during the study. Six of our seven patients had medical conditions leading to ARDS. However, one surgical patient was also included and the recordings in this patient did not differ from the others. This is supported by the recent study by Maggiore and colleagues . They recorded chest wall Pel/V curves in five patients with a methodology similar to ours and found that the chest wall had little or no influence on the Pel/V curves of the respiratory system.
It is also important to point out the fact that the Pel/V curve from ZEEP was the last in the sequence. This means that the preceding prolonged insufflations can have behaved as recruitment manoeuvres [9,25]. Accordingly, the exact value of the lower inflection point may have been affected. It is possible that the three normal breaths interposed between the insufflations from the stepwise lower PEEP levels were not sufficient to recruit the lung to the value at baseline with the highest PEEP (+15 cmH2O). Maggiore and colleagues observed that the end-expiratory volume was not completely restored even after 10 breaths . Although the methodology can induce some bias in the quantification of the end-expiratory volume, we think that the main results of the study were not affected.
The upper inflection point, expressed as pressure, was significantly lower when PEEP was higher. This is in line with the previous observations by Mergoni and colleagues  who found very low levels of the upper inflection point in ARDS patients when P/V curves were traced from a PEEP of 15 cmH2O. The behaviour of the upper inflection point in our patients seems to confirm the hypothesis that their lungs needed to be recruited. PEEP in fact, by reducing end-expiratory collapse, displaced the Pel/V curve upwards, thereby starting from a higher elastic recoil position. Our results are in line with Servillo and colleagues  who recently observed a lower level of upper inflection point when ARDS patients were ventilated with a tidal volume of 10-12 mL compared to 6 mL. In the present study, we found that the upper inflection point, expressed as volume, remained constant with the stepwise decreases in PEEP levels. This observation, together with the behaviour of the upper inflection point expressed as pressure, conforms with the notion of the upper inflection point as essentially a marker of the end of recruitment [7-11,26].
In conclusion, although limited by the small patients' sample, our data indicate that alveolar closure in ARDS patients occurs over a wide range of pressures and that the lower inflection point is a poor indicator of alveolar closure. Multiple Pel/V curves are useful to evaluate lung collapse and to quantify alveolar recruitment making possible an individual titration of PEEP and tidal ventilation. A simple bedside tool to measure Pel/V curves would be desirable in the future.
1. Matamis D, Lemaire F, Harf A, Brun-Buisson C, Ansquer JC, Atlan G. Total respiratory pressure-volume curves in the adult respiratory distress syndrome. Chest
2. Ranieri VM, Giuliani R, Fiore T, Dambrosio M, Milic-Emili J. Volume-pressure curve of the respiratory system predicts effects of PEEP in ARDS: 'occlusion' versus 'constant flow' technique. Am J Respir Crit Care Med
3. Roupie E, Dambrosio M, Servillo G, et al.
Titration of tidal volume and induced hypercapnia in acute respiratory distress syndrome. Am J Respir Crit Care Med
4. Servillo G, Svantesson C, Beydon L, et al.
Pressure-volume curves in acute respiratory failure. Automated low flow inflation versus occlusion. Am J Respir Crit Care Med
5. Servillo G, De Robertis E, Coppola M, Blasi F, Rossano F, Tufano R. Application of a computerised method to measure static pressure volume curve in acute respiratory distress syndrome. Intens Care Med
6. Lemaire F. ARDS and PV curves: the inseparable duet? Intens Care Med
7. Jonson B, Svantesson C. Elastic pressure-volume curve: what information do they convey? Thorax
8. Hickling KG. The pressure-volume curve is greatly modified by recruitment. A mathematical model of ARDS lungs. Am J Respir Crit Care Med
9. Liu JM, De Robertis E, Blomquist S, Dahm PL, Svantesson C, Jonson B. Elastic pressure-volume curves of the respiratory system reveal a high tendency to lung collapse in young pigs. Intens Care Med
10. De Robertis E, Liu JM, Blomquist S, Dahm PL, Thorne J, Jonson B. Elastic properties of the lung and chest wall in young and adult healthy pigs. Eur Respir J
11. Jonson B, Richard JC, Straus C, Mancebo J, Lemaire F, Brochard L. Pressure-volume curves in acute lung injury. Evidence of recruitment above the lower inflection point. Am J Respir Crit Care Med
12. Svantesson C, Drefeldt B, Sigurdsson S, Larsson A, Brochard L, Jonson B. A single computer-controlled mechanical insufflation allows determination of the pressure-volume relationship of the respiratory system. J Clin Monit
13. Sigurdson S, Svantesson C, Larsson A, Jonson B. Elastic pressure-volume curves indicate derecruitment after a single deep expiration in anaesthetised and muscle-relaxed healthy man. Acta Anaesthesiol Scand
14. Bernard GR, Artigas A, Brigham KL, et al.
The American-European consensus conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical trial coordination. Am J Respir Crit Care Med
15. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded definition of the adult respiratory distress syndrome. Am Rev Respir Dis
16. Servillo G, Coppola M, Blasi F, Tufano R. The measurement of the pressure-volume curves with computerized methods. Minerva Anesthesiol
17. Svantesson C, Sigurdsson S, Larsson A, Jonson B. Effects of recruitment of collapsed lung units on the elastic pressure-volume relationship in anaesthetised healthy adults. Acta Anaesthesiol Scand
18. Gattinoni L, Pesenti A, Bombino M, et al.
Relationships between lung computed tomographic density, gas exchange, and PEEP in acute respiratory failure. Anesthesiology
19. Maggiore SM, Jonson B, Richard JC, Jaber S, Lemaire F, Brochard L. Alveolar derecruitment at decremental positive end-expiratory pressure levels in acute lung injury. Comparison with the lower inflection point, oxygenation and compliance. Am J Respir Crit Care Med
20. Vieira SR, Puybasset L, Lu Q, et al.
A scanographic assessment of pulmonary morphology in acute lung injury. Significance of the lower inflection point detected on the lung pressure-volume curve. Am J Respir Crit Care Med
21. Lachmann B. Open up the lung and keep the lung open. Intens Care Med
22. Mergoni M, Martelli A, Volpi A, Primavera S, Zuccoli P, Rossi A. Impact of positive end-expiratory pressure on chest-wall and lung pressure-volume curve in acute respiratory failure. Am J Respir Crit Care Med
23. Svantesson C, Sigurdsson S, Larson A, Jonson B. The contribution of the chest wall to the elastic pressure-volume curve of the total respiratory system. Intens Care Med
24. Ranieri VM, Brienza N, Santostasi S, et al.
Impairment of lung and chest wall mechanics in patients with acute respiratory distress syndrome: role of abdominal distension. Am J Respir Crit Care Med
25. Richard JC, Maggiore SM, Jonson B, Mancebo J, Lemaire F, Brochard L. Influence of tidal volume on alveolar recruitment. Respective role of PEEP and recruitment maneuvre. Am J Respir Crit Care Med
26. Servillo G, De Robertis E, Maggiore S, Lemaire F, Brochard L, Tufano R. The upper inflection point of the pressure-volume curve. Influence of methodology and of different modes of ventilation. Intens Care Med