Luciano GattinoniJohn J. MariniAntonio PesentiMichael QuintelJordi ManceboLaurent Brochard

The ‘‘baby lung’’ became an adult

Received: 3 November 2015Accepted: 18 December 2015Published online: 18 January 2016� Springer-Verlag Berlin Heidelberg andESICM 2016

Take-home message: The baby lung wasoriginally defined as the fraction of lungparenchyma that still maintains normalinflation with near normal mechanicalcharacteristics despite at least a part of thewell-aerated baby lung is inflamed. Thebaby lung is not an anatomical but afunctional entity which changes dimensionsand location with prone position and PEEPapplication. Particular attention should bepaid to the general principles regulating thesafety of ventilation during both mechanicalventilation and spontaneous breathing.

L. Gattinoni ()) � A. PesentiDipartimento di Anestesia, Rianimazione edEmergenza Urgenza, Fondazione IRCCS CaGranda—Ospedale Maggiore Policlinico,Milan, Italye-mail: gattinon@policlinico.mi.itTel.: ?39-02-55033232

A. Pesentie-mail: antonio.pesenti@unimi.it

L. Gattinoni � A. PesentiDipartimento di FisiopatologiaMedico-Chirurgica e dei Trapianti,Universita degli Studi di Milano,Via Francesco Sforza 35,20122 Milan, Italy

J. J. MariniDepartment of Medicine, University ofMinnesota, Minneapolis, Saint Paul, MN,USAe-mail: John.J.Marini@HealthPartners.Com

M. QuintelDepartment of Anesthesiology, Emergencyand Intensive Care Medicine, Georg-AugustUniversity of Gottingen, Gottingen,Germanye-mail: mquintel@med.uni-goettingen.de

J. ManceboServicio de Medicina Intensiva, Hospital dela Santa Creu i Sant Pau, Barcelona, Spaine-mail: JMancebo@santpau.cat

L. BrochardKeenan Research Centre, Li Ka ShingKnowledge Insitute, Critical CareDepartment, St. Michael’s Hospital,Toronto, ON, Canadae-mail: BrochardL@smh.ca

L. BrochardInterdepartmental Division of Critical CareMedicine, Toronto, ON, Canada

Abstract The baby lung was origi-nally defined as the fraction of lungparenchyma that, in acute respiratorydistress syndrome (ARDS), stillmaintains normal inflation. Its sizeobviously depends on ARDS severityand relates to the compliance of therespiratory system. CO2 clearanceand blood oxygenation primarilyoccur within the baby lung. While thespecific compliance suggests theintrinsic mechanical characteristics tobe nearly normal, evidence frompositron emission tomography sug-gests that at least a part of the well-aerated baby lung is inflamed. Thebaby lung is more a functional con-cept than an anatomical one; in fact,in the prone position, the baby lung‘‘shifts’’ from the ventral lungregions toward the dorsal lungregions while usually increasing itssize. This change is associated with

Intensive Care Med (2016) 42:663–673DOI 10.1007/s00134-015-4200-8 REVIEW

better gas exchange, more homoge-neously distributed trans-pulmonaryforces, and a survival advantage.Positive end expiratory pressure alsoincreases the baby lung size, bothallowing better inflation of alreadyopen units and adding new pulmonaryunits. Viewed as surrogates of stress

and strain, tidal volume and plateaupressures are better tailored to babylung size than to ideal body weight.Although less information is availablefor the baby lung during spontaneousbreathing efforts, the general princi-ples regulating the safety of

ventilation are also applicable underthese conditions.

Keywords ARDS � Baby lung �Transpulmonary pressure �Prone position � Stress and strain

Introduction

Several diseases may produce the syndrome called acuterespiratory distress syndrome (ARDS). The symptomsinvolved are impaired gas exchange, decreased lungcompliance, increased lung weight and widespreadinvolvement of the lung parenchyma. Whatever the causeof ARDS, the treatment of symptoms is mandatory to buytime for the resolution of the underlying process, such aspneumonia or sepsis. Possible harm from mechanicalventilation is understandable if the ARDS lung is modeledin two regions, one nearly normal and having dimensionssimilar to those of a healthy baby (Fig. 1). The restrictedcapacity of this ‘baby lung’ is primarily responsible forthe mechanical characteristics described in ARDS. Thesecond region, consolidated and collapsed, is primarilyresponsible for the impairment of oxygenation. Thisconcept evolved over several decades: the baby lung,rather than an anatomical reality, became a functionalstatus. Studies relating to prone positioning, positive end-

expiratory pressure (PEEP), stress and strain, and spon-taneous breathing may be considered according to thebaby lung model, as we explore in this paper.

Baby lung: anatomical characteristics

The concept of ‘‘baby lung’’ originated from observationson the first CT scan images obtained in ARDS patients,which showed that densities were preferentially distributedin the dependent lung regionswhile the non-dependent lungregions were relatively spared. This view deeply contrastswith the common belief, derived from antero-posteriorimaging, that ARDS homogeneously involves the entirelung parenchyma. Quantitative analysis of the CT scan (seeFig. 2) [1] showed that the patho-anatomy of the ARDSlung is not homogeneously distributed but comprised ofregions totally deprived of air and regions normally (oralmost normally) aerated. Only a small proportion of thelung is ventilated, and this small (baby) lung has to fulfillthe physiological needs of an adult body.

Quantitatively, the baby lung includes the voxelscomprised in the interval between -900 and -500Hounsfield units (HU) [2], which describe the densities ofthe voxel, where -900 indicates 90 % gas and 10 %tissue and -500 indicates 50 % gas and 50 % tissue. Innormal lungs, the voxels included in this interval (thewell-inflated tissue) correspond to a weight of about600–700 g, while in ARDS they may account for 1/3 oreven 1/5 of this value. Quantitative analysis of the CTscan provided a powerful tool to compare the in vivoanatomy with the physiological variables commonlymeasured in the ARDS definition. The most importantfinding was the discovery that the aerated ARDS lung wasnot stiff but small, with near normal intrinsic mechanicalcharacteristics [2]. The ratio of the respiratory systemcompliance to the residual healthy (‘baby’) lung (ex-pressed as a fraction of the expected healthy lung volume)is roughly 1:1, i.e. 20 mL/cmH2O compliance corre-sponds roughly to 20 % open ventilatable lung, 50 mL/cmH2O to 50 %, and so on. The relationship between the

Fig. 1 A representative CT scan image of an ARDS patientshowing that the ARDS lung can be modeled in one nearly normalregion (having dimensions similar to those of a healthy baby) and agasless region

664

baby lung and gas exchange provided further insight intoflow distribution in ARDS, indicating that the hypoxicvasoconstriction is still very active in the ARDS lung[3].

Baby lung and inflammation

Although the specific compliance of the baby lung isnearly normal, regional quantitative analysis of CT scansshowed that excess tissue mass, at least in severe ARDS,is present both in non-dependent and in dependentregions, suggesting that dependent densities primarilyrepresent atelectasis due to the increased lung weight,which squeezes the gas out of the most dependent pul-monary units. Therefore, the question arises: if theparenchyma of the baby lung is mechanically normal, is itinflamed or simply edematous?

An answer may be provided by positron emissiontomography (PET) which helps to explore the differentproperties and functions of the lung tissue by bothimaging and quantitative approaches [4] (Fig. 3). In fact,PET can assess lung regional vascular permeability bymeasuring the plasma escape rate of 68Gallium-transfer-rin. By this technique, it has been shown thatextravascular lung water is diffusely increased duringARDS, lobar pneumonia, and acute cardiogenic pul-monary edema. One study [5] showed that in ARDS not

just the non-ventilated regions but also the baby lungbears one of the hallmarks of tissue inflammation:increased permeability of the endothelial–epithelial bar-rier. At variance with density and extravascular lungwater, which both increase according to a vertical, grav-itational gradient, increased permeability was evenlydistributed throughout the entire ARDS lung and abnor-mally high in all isogravitational regions [6].

PET also allows investigation of the regional distribu-tion of inflammation, using 18Fluor-deoxy-glucose(18FDG), an analogue of glucose,which is taken up by cellsin proportion to their metabolic activity. During aninflammatory process, several cellular types can effectivelyuptake 18FDG, including macrophages [7], eosinophils [8]and alveolar cells [9], with predominant uptake occurringby neutrophils [10]. In several animal studies, PET with18FDG has been shown to closely mirror the presence ofinflammatory infiltrate [7] caused by various types of lunginjuries [11], including aggressive ventilation [12].

This tool was applied in ARDS patients, who under-went a combined CT/PET scan. In all patients, theinflammatory activity of the lungs was increased (seven-fold, on average) in comparison with controls. In thiscase, too, the baby lung showed a greater activity thancontrols and, in some cases, even a greater activity thanthe collapsed, non-aerated lung [13]. At variance fromprevious studies [12], 18FDG uptake was not normalizedby the fraction of tissue from the corresponding regions,since if the increase in tissue fraction was caused byedema (as likely the case in ARDS), this would lead to anover-correction of 18FDG signal. However, only tissuesof similar density were directly compared, hence avoidingthe need for density-correction. A subsequent studyinvestigated the association between mechanical ventila-tion and inflammatory activity of the baby lung. Theuptake of 18FDG in the normally aerated tissue (the babylung) was tightly correlated both with plateau pressure(with a steep increase for plateau pressure values above26–27 cmH2O) and with the ratio of tidal volume to endexpiratory lung volume of the same regions [14]. Thesedata suggested a causative role for both stress and strainin inflaming the ventilated baby lung. A recent CT-PETstudy, however, indicated that baby lung inflammationwas primarily a function of ARDS severity. Unlike thebaby lung of mild ARDS, no regions were spared theinflammatory reaction in severe ARDS [15].

Increasing the baby lung size

Prone position

At the beginning, the baby lung was considered as if itwere an anatomically well-defined entity, with a definite

Fig. 2 The CT frequency distribution as a function of CT numbercompartments distribution of patients with ARDS (black bars) andnormal subjects (gray bars). The diagonal patterned bars indicatethe baby lung in ARDS patients. The zones of hyperinflated,normally, poorly and not-inflated lung are shown (vertical dashedlines)

665

spatial location (non-dependent) and with normalmechanical intrinsic characteristics. However, when theARDS lung was studied by CT scan in the prone position(PP), it appeared that the densities redistributed from thedorsal to the ventral lung regions, leading to the conceptof ‘‘the sponge model’’. Accordingly, the baby lung wasno longer considered an anatomical entity but instead afunctional one, which may quickly change its locationand/or its dimensions with PP and/or application of PEEP(Fig. 4).

Regional mechanics of the baby lung

The anterior chest wall is more flexible than the posteriorone. Therefore, assuming the PP increases total chest wallelastance, such changes affect trans-pulmonary pressuresand lessen the pressure gradients that determine absolute

lung volume as well as its topographic distribution. Moreuniform distribution of transpulmonary pressure by PPhelps establish and sustain positional recruitment [16, 17](Fig. 5). PP relieves cardiac compression of the support-ing lung that lies beneath the heart in the supine positionand may improve lymphatic drainage as the heart movesventrally to a dependent position [18]. Such changes helpto explain why PP improves shunt and aids resolution ofhydrostatic edema [19, 20].

Yet, recruitment that occurs after turning neither fullyexplains proning-related improvements in oxygenationnor the reduced propensity to ventilator-induced lunginjury (VILI). Establishing more uniform alveolar dis-tension improves ventilation/perfusion ratios of acutelyinjured lungs [21]. Proning benefits oxygenation in themajority of cases and helps ventilation efficiency in asomewhat lesser proportion of patients [22]. Becauseperfusion distribution is only modestly modified by the

Fig. 3 Representative images of cross-registered computed tomog-raphy (at mean airway pressure) and [18F]-fluoro-2-deoxy-D-glucose (18FDG) positron emission tomography from two patientswith acute respiratory distress syndrome. Positron emission tomo-graphic images represent the 18FDG uptake and the color scale

represents radioactivity concentration. a 18FDG distribution paral-lels that of the opacities detected on CT. b Intense 18FDG uptakecan be observed in normally aerated regions (square 1), whileactivity is lower in the dorsal, ‘‘non-aerated’’ regions of both lungs(square 2). Reproduced from [13]

666

inversion of position [23], change in the distributionalpattern of ventilation appears to be of greater importance.Resolving disparities of regional alveolar distension mayalso help prevent zonal over-distension and inappropriateredirection of blood flows from inflated to collapsed units.The latter phenomenon tends to occur when raising PEEPand mean airway pressure in the supine position. Rightventricular afterload may benefit from reduced over-dis-tension, improved oxygenation, and recruitment ofvascular beds [24, 25].

Proning and ventilator-induced lung injury

Ventilator-induced lung injury arises from repeatedapplication of high mechanical forces that either tearfragile tissue directly or initiate signaling that culminatesin inflammatory change [26]. Stress is augmented at theinterfaces between open and closed (or anatomically non-inflatable) tissues, and these stress-focusing junctions (aswell as VILI) prevail in gravitationally dependent zones[14]. These same units are subjected to surfactant-de-pleting tidal cycles of airspace opening and closure and toshearing forces that exceed those experienced by lungunits that remain continuously open. Recruitment thatoccurs during proning enlarges the baby lung, reduces thenumber of high stress junctions and portends benefit fromrepositioning [22]. By reducing the number of interfacesbetween open and closed units, as well as by moderatingtranspulmonary forces, the regional excursions ofmechanical tension (effective alveolar driving pressures)are lessened, especially in the dorsal zones. Finally, air-way bio-fluids and secretions may directly inhibitsurfactant viability or production, impede the ease ofairspace opening, or predispose to lung infection [27].Airway drainage, regional mechanics, and vascular fillingare each improved by PP. Perhaps in part for this reason,hemorrhagic pulmonary edema and inflammation thatresult from adverse patterns of ventilation are differen-tially affected by positioning. In both healthy canine lungs[28] and those pre-injured by infused oleic acid, proningmay reduce dorsal hemorrhage, edema and inflammation,whereas nondependent zones are relatively spared in bothpositions [28, 29].

Fig. 4 Examples of widening the baby lung by prone position (a, b) and by increasing PEEP from 5 cmH2O (c) to 15 cmH2O (d). Notethat recruitment in general proceeds from dorsal to ventral regions in the prone position and from ventral to dorsal by pressure increase

Fig. 5 The gas:tissue ratio as a function of lung height in supine(rhombus) prone (circles) positions. The open symbols refer tonormal lungs of 14 patients and the filled symbols to lungs of 20ARDS patients. The height = 0 refers to the ventral surface in thesupine position and to the dorsal surface in the prone position. Theheight scale direction is from nondependent (0) to dependent (100).Reproduced from Ref. [74]

667

In addition to attenuating regional transpulmonarydisparities of force, PP confers a secondary benefit simplyby improving the P/F ratio, thereby reducing the need forpotentially iatrogenic interventions to sustain it. Improvedoxygenation by PP may allow reductions of FiO2, meanairway pressure, and pulmonary vascular resistance,thereby lowering the risk of injury to mechanicallystressed membranes [30] and/or right ventricular afterload[31].

Recruitment

The amount of non-aerated and poorly aerated lungmainly determines the gas exchange impairment observedin ARDS. The use of PEEP keeps open newly recruitedand previously non-aerated regions, thus enlarging thebaby lung and improving oxygenation. The effects ofPEEP on alveolar recruitment have been assessed usingpressure–volume curves of the respiratory system and CTscan methods [32–34]. Irrespective of how recruitment ismeasured, an important concept is that the recruitmentprocess occurs throughout inspiration, following the tra-jectory of the pressure–volume relationship of therespiratory system [33, 35, 36]. Therefore, what canremain open with PEEP is the tissue that was opened bythe preceding inspiration. In other words, if the previousend-inspiratory pressure is low, the recruitment obtainedat a given PEEP is less. This is supported by the findingthat de-recruitment that results from decreasing tidalvolume (VT) can be reversed by increasing PEEP [37].

With CT scanning, it has been shown that the amountof alveolar opening–closing decreases when PEEPincreases [38]. At PEEP levels above 10–15 cmH2O andend-inspiratory plateau airway pressures (Pplat) above 30cmH2O, the amount of tissue reopening and collapsingduring the tidal cycle was minimal and the VT wasredistributed more homogeneously between dependentand non-dependent regions.

In a series of studies on ARDS patients, Puybassetand coworkers performed extensive measurements onlung CT scans, gas exchange, hemodynamics andmechanics at zero end-expiratory pressure and PEEP 10cmH2O [39–41]. According to CT scan distribution ofdensities, they described three types of patterns: lobar,diffuse and patchy. Those with a diffuse, as compared tothe lobar pattern, had a significantly lower functionalresidual capacity and a significantly higher amount ofnon-aerated lung volume. A diffuse versus lobar densitypattern, a lower PaO2/FiO2, and less proportion of aer-ated lung, identified patients with higher mortality risk.The decrease in non-aerated lung with PEEP 10 cmH2Owas almost negligible (-82 mL) in patients with lobarpattern, whereas the highest decrease in non-aeratedlung (-383 mL) was observed in those with a diffusepattern.

The overall effects of PEEP on recruitability, how-ever, are complex. The percentage of potentiallyrecruitable lung, as analyzed with CT scans performed upto 45 cmH2O airway pressure, was highly variable inacute lung injury/ARDS patients [42]. About 24 % of thelung could not be recruited even at 45 cmH2O, and lungrecruitment occurring from PEEP 5–15 cmH2O correlatedwell with the percentage of potentially recruitable lung.Interestingly, patients with a high percentage of poten-tially recruitable lung had higher mortality rates thanthose with a low percentage. As recruitability increasedwith severity, it follows that the smaller the baby lung atlow pressure, the greater would be its relative increasewhen PEEP is increased. A further analysis by Caironiand coworkers [43], showed that, in patients with a higherpercentage of potentially recruitable lung, raising PEEPfrom 5 to 15 markedly decreased the amount of opening–closing lung tissue, but PEEP did not produce any effectin patients with a lower percentage of potentiallyrecruitable lung. Analysis of the association betweenmortality and determinants of VILI showed that theamount of opening–closing was independently associatedwith an increased risk of death. A recent study [44] hasshown that, among different bedside PEEP selectionmethods in ARDS patients, the one based on oxygenationcriteria was most closely related to lung recruitability.Taken together, these data [42–44] suggest high PEEPlevels should be used in more severe ARDS but not inpatients with a lower percentage of potentiallyrecruitable lung. In other words, the smaller the baby lungat low pressure the greater the PEEP to be applied.

Recently, Cressoni et al. [45] showed that the greaterthe ARDS severity, the greater the lung inhomogeneity.Such inhomogeneities were mainly distributed at theinterface between lung regions with high and normal CTdensity, i.e. at the ‘‘boundary zones’’ of the baby lung,which were more numerous in non-survivors than insurvivors and were always ‘‘inflamed’’, as shown by PET[15]. The authors proposed that lung inhomogeneity actsas a stress raiser, and they calculated that these regions areexposed to a stress which almost doubles the transpul-monary pressure applied. Such a mechanism (stressraising) can further magnify VILI. Interestingly, theextent of inhomogeneity was associated with the fractionof poorly aerated lung tissue, and it decreased when PEEPwas increased.

Are there clinical data to suggest which PEEP level ismost appropriate in ARDS? In the ExPress study [46], theuse of low VT with high PEEP (about 15 cmH2O, indi-vidually titrated to reach a Pplat of 28–30 cmH2O) tendedto increase mortality in acute lung injury patients whencompared to low VT and moderate PEEP. Conversely, thelow VT/high PEEP strategy decreased mortality in sickerARDS patients. A secondary post hoc analysis of ran-domized PEEP trials further supports the selective use ofhigh PEEP levels in severe ARDS patients, i.e. those with

668

smaller baby lungs [47]. Our understanding of the effectsof PEEP in the baby lung is maturing. The new clinicalchallenge is to determine whether a PEEP titration strat-egy based on sound physiological and morphologicalprinciples, and which can be easily used at the bedside(using measurements such as dead space or drivingpressures), improves outcomes in ARDS.

Baby lung and mechanical ventilation

Tidal volume and plateau pressure

In clinical practice, tidal volume is scaled to patient idealbody weight (IBW), predicted from the patient height(VT/IBW) to avoid excessive strain of the lung par-enchyma. Relatively high tidal volumes, quantified as12 mL/kg IBW or greater, cause un-physiological strainin the size-reduced ARDS lung. In turn, excessive straincauses, through mechanical transduction, increased localinflammation and inflammatory cytokines which likelycontribute to increased mortality. Halving the harmfultidal volume may decrease mortality but predisposes toCO2 retention (permissive hypercapnia) which, whenmoderate, is generally well tolerated and possible bene-ficial. Several clinical studies have analyzed the effects oflower versus higher values of VT/IBW. Among theseoutcome studies testing different tidal volumes, only theARDS Network study, which compared the two extremevalues (6 vs. 12 mL/kg) in 861 patients, showed a sig-nificant mortality reduction in the lower VT group (31 vs.40 %) [48]. Although other studies did not find any sig-nificant mortality reduction [49–51] when testing VTvalues intermediate within the 6–12 mL/kg interval, lowVT ventilation (6 ml/k IBW ventilation) has been widelyaccepted by the scientific community. Over-distension,however, has been described in the non-dependent lungregions even at 6 ml/kg IBW [52]. Although the ideal (orpredicted) body weight was introduced as a better surro-gate than measured weight to adjust for variations in lungsize, it is evident that this approach does not performoptimally well in ARDS. In a man with an ideal bodyweight of 70 kg, the aeratable lung volume, i.e. the babylung, may range from 200 to 300 ml to more than1000 ml, depending on the ARDS severity. Referring tothe gas volume of the baby lung instead of to the bodyweight is more physiologically adequate as it betterreflects its distortion (strain) during ventilation. In fact,strain is defined as the ratio between tidal volume and thebaseline gas volume of the baby lung. In the exampleabove, the estimated strain induced by a tidal volume of6 ml/kg IBW would range between 2.1 and 0.42—valuesfound in animal studies to be lethal (i.e. leading to death)and innocent (without any harm), respectively. Therefore,a safer mechanical ventilation strategy should consider

the size of the baby lung instead of body weight [53].More recently, a post hoc analysis of 3562 patients withARDS enrolled in nine previously reported randomizedtrials [54], compared the relative predictive power of tidalvolume normalized to estimated lung compliance (drivingpressure) with tidal volume normalized to ideal bodyweight. The former index performed considerably better,suggesting that compliance—an indicator of baby lungsize—is a better surrogate than ideal body weight whentailoring tidal volume to prevent excessive strain.

The same line of reasoning may be applied whenconsidering the stress, i.e. the pressure, which developswithin lung structures to counteract the applied transpul-monary pressure. Therefore the stress equals thetranspulmonary pressure (with opposite sign). The widelyaccepted surrogate for stress is the airway plateau pres-sure, and 30 cmH2O is the suggested limit [55]. A simpleanalysis of the relationship of transpulmonary pressure(PL) and airway pressure (PAW) shows how this sugges-tion is questionable. In fact, the same PAW is associatedwith a wide range of PL due to different ratios of chestwall elastance (EW) to respiratory system elastance (ERS)

PL ¼ PAW � EW

ERS

A post hoc analysis by the ARDS network could notidentify a safe upper limit for plateau pressures in patientswith ALI/ARDS, as values even lower than 30 cmH2Omight cause lung damage [56]. These data indicate that at30 cmH2O the baby lung may be very close to its totallung capacity or very far from it. Depending on chest wallelastance, transpulmonary pressure associated with 30cmH2O can be estimated to vary between 28 and 8cmH2O [53]. In animal studies, these levels are lethal orinnocent, respectively [57].

Baby lung and spontaneous breathing

Although awareness of the baby lung concept hasencouraged significant improvement in ventilator man-agement, the same reasoning has not always beenfollowed when the patient makes breathing efforts, asduring noninvasive ventilatory support and when allow-ing some spontaneous breathing activity during the earlyphase of ARDS.

In theory, baby lung size should be the same at a giventranspulmonary pressure, whether it is generated by apositive pressure, by a negative pressure (and thereforethe respiratory muscles) or by a combination of the two.The validity of this principle was illustrated in an earlyanimal model of VILI [58]. During spontaneous breath-ing, however, negative pleural pressure swings will havedifferent vascular effects than during controlledmechanical ventilation, resulting in higher filling

669

pressures of the heart and in the pulmonary circulation[59]. This vascular congestion may increase the risk ofVILI [60, 61]. In addition, the measurement of esophagealpressure may poorly reflect the distribution of regionalpleural and transpulmonary pressures. Under the action ofthe diaphragm and of other accessory muscles, theregional pleural pressures may have a wider distributionthan during controlled mechanical ventilation, where thegradient of pleural pressure is mostly dictated by gravity.Local variations of distending pressure may help explainthe observation of pendelluft (i.e., regional redistributionof ventilation during the tidal cycle), among injured lungregions during spontaneous breathing [62]. In such cases,the risk of VILI may be poorly reflected by the globaltranspulmonary pressure.

Experimental and clinical data

Sustained hyperventilation induced by severe acidosis maylead to acute lung injury in animals [63]. In experimentalmodels, however, allowing spontaneous breathing seems tohave different consequences, depending on the severity oflung injury [64]. In the case of severe lung injury, sponta-neous breathing appears to worsen damage while theopposite may happen in those with mild injury. These dif-ferencesmay reflect an increase in transpulmonary pressureand tidal volume during spontaneous ventilation, associ-ated with high regional distending pressures and increasesof trans-vascular pressure. A series of studies conducted inpatients, confirmed by a multicenter trial, strongly sug-gested that the systematic administration of neuromuscularblockers can reduce biomarkers of inflammation andimprove survival in patients with ARDS [65, 66]. Neuro-muscular blockade was also the first treatment to show asignificant reduction in the rate of pneumothorax (11.7 vs.4.0 %).Of note, even the seminal study on low tidal volumedid not show any reduction of barotrauma [48]. Theseresults seem to have been determined primarily by datafrom patients with PaO2/FiO2 at or below 120 mmHg [65].Although these results are not fully understood, they sug-gest a possible negative effect of spontaneous breathing onlocal distending pressures, as discussed above, by highventilatory drive-caused asynchronies such as breathstacking and double cycling, or by reverse triggering. [67,68]. In this situation, clinicians should carefullymonitor theresults of patient–ventilator synchronization in terms of theresulting tidal volume and transpulmonary pressure. Clin-icians must be aware of the physiological differencesbetween volume-controlled and pressure controlled venti-lation on both trans-pulmonary pressure and patientcomfort [69]. The use of non- synchronized pressure-tar-geted modes in this situation may help to maintain somespontaneous breathing activity withminimal risks, offeringan interesting compromise [70].

Noninvasive ventilatory support

In non-intubated patients, gauging the severity of lunginjury presents problems because respiratory mechanicsare difficult to assess. Noninvasive ventilation has beenfrequently attempted in lung-injured patients as a way toavoid intubation, with various degrees of success [71, 72].Recently, another technique of noninvasive ventilatorysupport in the form of high-flow heated and humidifiedoxygen therapy has been tested in patients with acutehypoxemic respiratory failure mostly resulting from sev-ere infectious pneumonia [73]. Patients with PaO2/FiO2 ator below 200 mmHg were less likely to need intubationthan those treated by conventional masks. Overall,patients treated by high-flow nasal oxygen had a lowerhospital and 90-day mortality. Such surprising resultsclearly need confirmation. However, one possibility toexplain the benefit in survival with this simple and well-tolerated technique may be a reduction in ventilatoryneeds due to a washout of anatomic dead space. Althoughnot clearly demonstrated, it is conceivable that a reductionin tidal volume and/or in minute ventilation imposed onthe baby lung (due to dead space wash-out or improvedcomfort) could explain this observed decrease inmortality.

Conclusions

The size of the baby lung determines tissue complianceand directly dictates the mechanical properties of theARDS lung. For a given tidal volume, therefore, stressand strain relate to baby lung size. Physical characteristicssuch as mechanical homogeneity and dimension of thebaby lung may be deeply affected by PP and PEEP. Thepositive effects of these two maneuvers appear inverselyrelated to the baby lung size. Although less well docu-mented, these same rules are likely to apply whenspontaneous breathing efforts are made. The baby lungconcept may help to understand and implement safemechanical or spontaneous ventilation during the differ-ent stages of ARDS.

Acknowledgments We thank Prof. Giacomo Bellani (Diparti-mento di Medicina e Chirurgia, Universita degli Studi di MilanoBicocca, Milan, Italy) and Eleonora Carlesso (Dipartimento diFisiopatologia Medico-Chirurgica e dei Trapianti, FondazioneIRCCS Ca’ Granda-Ospedale Maggiore Policlinico, Universitadegli Studi di Milano, Milan, Italy) for their invaluable support inthe preparation of this manuscript.

Compliance with ethical standards

Conflicts of interest None.

670

References

1. Gattinoni L, Mascheroni D, Torresin Aet al (1986) Morphological response topositive end expiratory pressure inacute respiratory failure. Computerizedtomography study. Intensive Care Med12:137–142

2. Gattinoni L, Pesenti A, Avalli L et al(1987) Pressure-volume curve of totalrespiratory system in acute respiratoryfailure. Computed tomographic scanstudy. Am Rev Respir Dis 136:730–736

3. Cressoni M, Caironi P, Polli F et al(2008) Anatomical and functionalintrapulmonary shunt in acuterespiratory distress syndrome. Crit CareMed 36:669–675

4. Bellani G, Amigoni M, Pesenti A(2011) Positron emission tomographyin ARDS: a new look at an oldsyndrome. Minerva Anestesiol77:439–447

5. Kaplan JD, Calandrino FS, Schuster DP(1991) A positron emissiontomographic comparison of pulmonaryvascular permeability during the adultrespiratory distress syndrome andpneumonia. Am Rev Respir Dis143:150–154

6. Sandiford P, Province MA, Schuster DP(1995) Distribution of regional densityand vascular permeability in the adultrespiratory distress syndrome. Am JRespir Crit Care Med 151:737–742

7. Zambelli V, Di Grigoli G, Scanziani Met al (2012) Time course of metabolicactivity and cellular infiltration in amurine model of acid-induced lunginjury. Intensive Care Med 38:694–701

8. Harris RS, Venegas JG,Wongviriyawong C et al (2011) 18F-FDG uptake rate is a biomarker ofeosinophilic inflammation and airwayresponse in asthma. J Nucl Med52:1713–1720

9. Saha D, Takahashi K, de Prost N et al(2013) Micro-autoradiographicassessment of cell types contributing to2-deoxy-2-[(18)F]fluoro-D-glucoseuptake during ventilator-induced andendotoxemic lung injury. Mol ImagingBiol 15:19–27

10. Musch G (2011) Positron emissiontomography: a tool for betterunderstanding of ventilator-induced andacute lung injury. Curr Opin Crit Care17:7–12

11. de Prost N, Feng Y, Wellman T et al(2014) 18F-FDG kinetics parametersdepend on the mechanism of injury inearly experimental acute respiratorydistress syndrome. J Nucl Med55:1871–1877

12. Musch G, Venegas JG, Bellani G et al(2007) Regional gas exchange andcellular metabolic activity in ventilator-induced lung injury. Anesthesiology106:723–735

13. Bellani G, Messa C, Guerra L et al(2009) Lungs of patients with acuterespiratory distress syndrome showdiffuse inflammation in normallyaerated regions: a [18F]-fluoro-2-deoxy-D-glucose PET/CT study. CritCare Med 37:2216–2222

14. Bellani G, Guerra L, Musch G et al(2011) Lung regional metabolic activityand gas volume changes induced bytidal ventilation in patients with acutelung injury. Am J Respir Crit Care Med183:1193–1199

15. Cressoni M, Chiumello D, Chiurazzi Cet al (2016) Lung inhomogeneities,inflation and [18F]2-fluoro-2-deoxy-D-glucose uptake rate in acute respiratorydistress syndrome. Eur Respir J47(1):233–242

16. Cakar N, der Kloot TV, Youngblood Met al (2000) Oxygenation response to arecruitment maneuver during supineand prone positions in an oleic acid-induced lung injury model. Am J RespirCrit Care Med 161:1949–1956

17. Guerin C, Baboi L, Richard JC (2014)Mechanisms of the effects of pronepositioning in acute respiratory distresssyndrome. Intensive Care Med40:1634–1642

18. Albert RK, Hubmayr RD (2000) Theprone position eliminates compressionof the lungs by the heart. Am J RespirCrit Care Med 161:1660–1665

19. Chatte G, Sab JM, Dubois JM et al(1997) Prone position in mechanicallyventilated patients with severe acuterespiratory failure. Am J Respir CritCare Med 155:473–478

20. Nakos G, Tsangaris I, Kostanti E et al(2000) Effect of the prone position onpatients with hydrostatic pulmonaryedema compared with patients withacute respiratory distress syndrome andpulmonary fibrosis. Am J Respir CritCare Med 161:360–368

21. Lamm WJ, Graham MM, Albert RK(1994) Mechanism by which the proneposition improves oxygenation in acutelung injury. Am J Respir Crit Care Med150:184–193

22. Gattinoni L, Vagginelli F, Carlesso Eet al (2003) Decrease in PaCO2 withprone position is predictive of improvedoutcome in acute respiratory distresssyndrome. Crit Care Med31:2727–2733

23. Wiener CM, Kirk W (1985) Albert RK(1990) Prone position reversesgravitational distribution of perfusion indog lungs with oleic acid-inducedinjury. J Appl Physiol Bethesda Md68:1386–1392

24. Jozwiak M, Teboul J-L, Anguel N et al(2013) Beneficial hemodynamic effectsof prone positioning in patients withacute respiratory distress syndrome. AmJ Respir Crit Care Med 188:1428–1433

25. Richard J-C, Bregeon F, Costes N et al(2008) Effects of prone position andpositive end-expiratory pressure onlung perfusion and ventilation. CritCare Med 36:2373–2380

26. Dreyfuss D, Saumon G (1998)Ventilator-induced lung injury: lessonsfrom experimental studies. Am J RespirCrit Care Med 157:294–323

27. Marini JJ, Gattinoni L (2008)Propagation prevention: acomplementary mechanism for ‘‘lungprotective’’ ventilation in acuterespiratory distress syndrome. Crit CareMed 36:3252–3258

28. Broccard AF, Shapiro RS, Schmitz LLet al (1997) Influence of prone positionon the extent and distribution of lunginjury in a high tidal volume oleic acidmodel of acute respiratory distresssyndrome. Crit Care Med 25:16–27

29. Broccard A, Shapiro RS, Schmitz LLet al (2000) Prone positioningattenuates and redistributes ventilator-induced lung injury in dogs. Crit CareMed 28:295–303

30. Marini JJ, Hotchkiss JR, Broccard AF(2003) Bench-to-bedside review:microvascular and airspace linkage inventilator-induced lung injury. CritCare Lond Engl 7:435–444

31. Repesse X, Charron C, Vieillard-BaronA (2015) Acute cor pulmonale inARDS: rationale for protecting the rightventricle. Chest 147:259–265

32. Gattinoni L, Pesenti A (2005) Theconcept of ‘‘baby lung’’. Intensive CareMed 31:776–784

33. Gattinoni L, Caironi P, Pelosi P,Goodman LR (2001) What hascomputed tomography taught us aboutthe acute respiratory distress syndrome?Am J Respir Crit Care Med164:1701–1711

34. Malbouisson LM, Muller JC,Constantin JM et al (2001) Computedtomography assessment of positive end-expiratory pressure-induced alveolarrecruitment in patients with acuterespiratory distress syndrome. Am JRespir Crit Care Med 163:1444–1450

671

35. Jonson B, Richard JC, Straus C et al(1999) Pressure-volume curves andcompliance in acute lung injury:evidence of recruitment above the lowerinflection point. Am J Respir Crit CareMed 159:1172–1178

36. Crotti S, Mascheroni D, Caironi P et al(2001) Recruitment and derecruitmentduring acute respiratory failure: aclinical study. Am J Respir Crit CareMed 164:131–140

37. Richard JC, Maggiore SM, Jonson Bet al (2001) Influence of tidal volumeon alveolar recruitment. Respective roleof PEEP and a recruitment maneuver.Am J Respir Crit Care Med163:1609–1613

38. Gattinoni L, Pelosi P, Crotti S, ValenzaF (1995) Effects of positive end-expiratory pressure on regionaldistribution of tidal volume andrecruitment in adult respiratory distresssyndrome. Am J Respir Crit Care Med151:1807–1814

39. Puybasset L, Cluzel P, Gusman P et al(2000) Regional distribution of gas andtissue in acute respiratory distresssyndrome. I. Consequences for lungmorphology. CT Scan ARDS StudyGroup. Intensive Care Med 26:857–869

40. Rouby JJ, Puybasset L, Cluzel P et al(2000) Regional distribution of gas andtissue in acute respiratory distresssyndrome. II. Physiological correlationsand definition of an ARDS SeverityScore. CT Scan ARDS Study Group.Intensive Care Med 26:1046–1056

41. Puybasset L, Gusman P, Muller JC et al(2000) Regional distribution of gas andtissue in acute respiratory distresssyndrome. III. Consequences for theeffects of positive end-expiratorypressure. CT Scan ARDS Study Group.Adult respiratory distress syndrome.Intensive Care Med 26:1215–1227

42. Gattinoni L, Caironi P, Cressoni M et al(2006) Lung recruitment in patientswith the acute respiratory distresssyndrome. N Engl J Med354:1775–1786

43. Caironi P, Cressoni M, Chiumello Det al (2010) Lung opening and closingduring ventilation of acute respiratorydistress syndrome. Am J Respir CritCare Med 181:578–586

44. Chiumello D, Cressoni M, Carlesso Eet al (2014) Bedside selection ofpositive end-expiratory pressure inmild, moderate, and severe acuterespiratory distress syndrome. Crit CareMed 42:252–264

45. Cressoni M, Cadringher P, Chiurazzi Cet al (2014) Lung inhomogeneity inpatients with acute respiratory distresssyndrome. Am J Respir Crit Care Med189:149–158

46. Mercat A, Richard JCM, Vielle B et al(2008) Positive end-expiratory pressuresetting in adults with acute lung injuryand acute respiratory distresssyndrome—A randomized controlledtrial. JAMA 299:646–655

47. Goligher EC, Kavanagh BP, RubenfeldGD et al (2014) Oxygenation responseto positive end-expiratory pressurepredicts mortality in acute respiratorydistress syndrome. A secondaryanalysis of the LOVS and ExPresstrials. Am J Respir Crit Care Med190:70–76

48. The Acute Respiratory DistressSyndrome Network (2000) Ventilationwith lower tidal volumes as comparedwith traditional tidal volumes for acutelung injury and the acute respiratorydistress syndrome. N Engl J Med342(18):1301–1308

49. Brochard L, Roudot-Thoraval F,Roupie E et al (1998) Tidal volumereduction for prevention of ventilator-induced lung injury in acute respiratorydistress syndrome. The MulticenterTrail Group on Tidal Volume reductionin ARDS. Am J Respir Crit Care Med158:1831–1838

50. Brower RG, Lanken PN, MacIntyre Net al (2004) Higher versus lowerpositive end-expiratory pressures inpatients with the acute respiratorydistress syndrome. N Engl J Med351:327–336

51. Stewart TE, Meade MO, Cook DJ et al(1998) Evaluation of a ventilationstrategy to prevent barotrauma inpatients at high risk for acuterespiratory distress syndrome. Pressure-and Volume-Limited VentilationStrategy Group. N Engl J Med338:355–361

52. Terragni PP, Rosboch G, Tealdi A et al(2007) Tidal hyperinflation during lowtidal volume ventilation in acuterespiratory distress syndrome. Am JRespir Crit Care Med 175:160–166

53. Chiumello D, Carlesso E, Cadringher Pet al (2008) Lung stress and strainduring mechanical ventilation for acuterespiratory distress syndrome. Am JRespir Crit Care Med 178:346–355

54. Amato MBP, Meade MO, Slutsky ASet al (2015) Driving pressure andsurvival in the acute respiratory distresssyndrome. N Engl J Med 372:747–755

55. Tobin MJ (2000) Culmination of an erain research on the acute respiratorydistress syndrome. N Engl J Med342:1360–1361

56. Hager DN, Krishnan JA, Hayden DL,Brower RG (2005) Tidal volumereduction in patients with acute lunginjury when plateau pressures are nothigh. Am J Respir Crit Care Med172:1241–1245

57. Protti A, Andreis DT, Monti M et al(2013) Lung stress and strain duringmechanical ventilation: any differencebetween statics and dynamics? CritCare Med 41:1046–1055

58. Dreyfuss D, Soler P, Basset G, SaumonG (1988) High inflation pressurepulmonary edema. Respective effects ofhigh airway pressure, high tidal volume,and positive end-expiratory pressure.Am Rev Respir Dis 137:1159–1164

59. Magder S, Verscheure S (2014) Properreading of pulmonary artery vascularpressure tracings. Am J Respir CritCare Med 190:1196–1198

60. Hotchkiss JR, Blanch L, Naveira A et al(2001) Relative roles of vascular andairspace pressures in ventilator-inducedlung injury. Crit Care Med29:1593–1598

61. Broccard AF, Hotchkiss JR, KuwayamaN et al (1998) Consequences ofvascular flow on lung injury induced bymechanical ventilation. Am J RespirCrit Care Med 157:1935–1942

62. Yoshida T, Torsani V, Gomes S et al(2013) Spontaneous effort causes occultpendelluft during mechanicalventilation. Am J Respir Crit Care Med188:1420–1427

63. Mascheroni D, Kolobow T, FumagalliR et al (1988) Acute respiratory failurefollowing pharmacologically inducedhyperventilation: an experimentalanimal study. Intensive Care Med15:8–14

64. Yoshida T, Uchiyama A, Matsuura Net al (2012) Spontaneous breathingduring lung-protective ventilation in anexperimental acute lung injury model:high transpulmonary pressureassociated with strong spontaneousbreathing effort may worsen lunginjury. Crit Care Med 40:1578–1585

65. Papazian L, Forel JM, Gacouin A et al(2010) Neuromuscular blockers in earlyacute respiratory distress syndrome.N Engl J Med 363:1107–1116

66. Forel J-M, Roch A, Marin V et al(2006) Neuromuscular blocking agentsdecrease inflammatory response inpatients presenting with acuterespiratory distress syndrome. Crit CareMed 34:2749–2757

67. Slutsky AS (2010) Neuromuscularblocking agents in ARDS. N Engl JMed 363:1176–1180

68. Akoumianaki E, Lyazidi A, Rey N et al(2013) Mechanical ventilation-inducedreverse-triggered breaths: a frequentlyunrecognized form of neuromechanicalcoupling. Chest 143:927–938

69. Rittayamai N, Katsios CM, Beloncle Fet al (2015) Pressure-controlled vsvolume-controlled ventilation in acuterespiratory failure: a physiology-basednarrative and systematic review. Chest148:340–355

672

70. Richard JCM, Lyazidi A, AkoumianakiE et al (2013) Potentially harmfuleffects of inspiratory synchronizationduring pressure preset ventilation.Intensive Care Med 39:2003–2010

71. Antonelli M, Conti G, Rocco M et al(1998) A comparison of noninvasivepositive-pressure ventilation andconventional mechanical ventilation inpatients with acute respiratory failure.N Engl J Med 339:429–435

72. Brochard L, Lefebvre J-C, Cordioli RLet al (2014) Noninvasive ventilation forpatients with hypoxemic acuterespiratory failure. Semin Respir CritCare Med 35:492–500

73. Frat J-P, Thille AW, Mercat A et al(2015) High-flow oxygen through nasalcannula in acute hypoxemic respiratoryfailure. N Engl J Med 372:2185–2196

74. Gattinoni L, Taccone P, Mascheroni Det al (2013) Prone positioning in acuterespiratory failure. In: Tobin MJ (ed)Principles and practice of mechanicalventilation, 3rd edn. McGraw Hill, NewYork, pp 1169–1181

673