Perioperative Lung Protection Strategies

in One-lung and Two-lung Ventilation

Peter Slinger, MD, FRCPC
Department of Anesthesia
University of Toronto and
Toronto General Hospital
Toronto, Ontario, Canada

rate most of these injuries. Lung-protective ventilation

Learning Objectives:
strategies using more physiological tidal volumes and
As a result of completing this activity, the participant
appropriate levels of positive end-expiratory pressure
will be able to
(PEEP) can decrease the extent of injury.1 The purpose of
 Identify patients at increased risk for perioperative
this chapter is to look at the effects of mechanical
lung injury
ventilation and its role in ventilator-induced lung injury
 Develop an evidence-based intraoperative ventila-
(VILI) with specific reference to patients with normal lungs
tion strategy to improve postoperative outcomes
and those with chronic obstructive pulmonary disease
 Apply recent knowledge of perioperative fluid
(COPD) during both two- and one-lung ventilation.
management and inflammation to decrease the
incidence of postoperative respiratory complica-
tions in thoracic and nonthoracic surgery CHRONIC OBSTRUCTIVE PULMONARY DISEASE
Author Disclosure Information: The most common chronic respiratory illness in the surgi-
Dr. Slinger has disclosed that he has no financial cal population is COPD, which incorporates three dis-
interests in or significant relationship with any orders: emphysema, peripheral airways disease, and
commercial companies pertaining to this educational chronic bronchitis.
activity.
Respiratory Drive
Many patients with severe COPD have an elevated arterial
atients are at risk for several types of lung injury in carbon dioxide tension (PaCO2) at rest. It is not possible to

P the perioperative period. These injuries include

atelectasis, pneumonia, pneumothorax, broncho-
pleural fistula, acute lung injury (ALI), and acute lung
differentiate these ‘‘CO2-retainers’’ from nonretainers on
the basis of history, physical examination, or spirometric
pulmonary function testing.2 This CO2 retention is likely
injury/acute respiratory distress syndrome (ALI/ARDS). not primarily the result of an alteration of respiratory
Anesthetic management can cause, exacerbate, or amelio- control mechanisms, but rather an inability to maintain the
increased work of respiration required to keep the PaCO2
normal in patients with mechanically inefficient pulmo-
nary function. It was previously thought that chronically
Supplemental Digital Content is available for this article. Direct URL hypercapnic patients relied on a hypoxic stimulus for
citations appear in the printed text and are available in both the HTML
and PDF versions of this article. Links to the digital files are provided in ventilatory drive and became insensitive to PaCO2. This
the HTML and PDF text of this article on the Journal’s Web site explained the clinical observation that COPD patients in
(www.asa-refresher.com). incipient respiratory failure could be put into a hyper-
93
Copyright r 2015 American Society of Anesthesiologists. All rights reserved.
Perioperative Lung Protection Strategies 95

tation, increased work of respiration, and increased airway in an intensive care unit setting for 2 days or longer de-
resistance. The interruption leads to an elevation of the veloped ALI or ARDS. The main risk factors for ALI were
end-expiratory lung volume above the functional residual use of large tidal volumes, restrictive lung disease, and
capacity. This PEEP in the alveoli at rest has been termed blood product transfusion. A prospective study from the
auto-PEEP, or intrinsic PEEP. During spontaneous respi- same group found that tidal volumes higher than 700 mL
ration, the intrapleural pressure will have to be decreased and peak airway pressures above 30 cm H2O were in-
to a level that counteracts auto-PEEP before inspiratory dependently associated with the development of ARDS.16
flow can begin. Thus, COPD patients can have an in- An intraoperative study of patients undergoing esophageal
creased inspiratory load added to their already increased surgery compared the use of tidal volumes of 9 mL/kg
expiratory load. without PEEP during two- and one-lung ventilation versus
Auto-PEEP becomes even more important during 9 mL/kg during two-lung ventilation and 5 mL/kg during
mechanical ventilation. It is directly proportional to tidal one-lung ventilation with PEEP 5 cm H2O throughout.17
volume and inversely proportional to expiratory time. The Significantly lower serum markers of inflammation (the
manometer of standard anesthesia ventilators does not cytokines interleukin 1b [IL-1b], IL-6, and IL-8) were
detect the presence of auto-PEEP. It can be measured by found in the group receiving a lower tidal volume plus
end-expiratory flow interruption, a feature available on PEEP. No major difference in postoperative outcome be-
the newer generation of intensive care ventilators. Auto- tween the two groups was reported; however, the study
PEEP has been found to develop in most COPD patients was not powered to detect this. The study did demonstrate
during anesthesia with one-lung ventilation.10 Paradoxi- better oxygenation in the lower tidal volume group during
cally, a small amount of PEEP (up to 50% of the individual and immediately after one-lung ventilation, but not after
auto-PEEP level, e.g., 5 cm H2O) can decrease hyper- 18 hours. In a study assessing conventional versus pro-
inflation in many ventilated COPD patients.11 This is tective ventilation in critically ill patients without lung in-
thought to be caused by a pneumatic stenting effect of jury, de Oliveira et al.18 randomized patients to ventilation
PEEP on the distal airways, decreasing collapse. with either 10 to 12 mL/kg or 6 to 8 mL/kg predicted body
weight. In both groups, 5 cm H2O of PEEP was applied
and the FiO2 titrated to keep oxygen saturation above
Auto-PEEP becomes even more important 90%. At 12 hours postventilation, inflammatory markers
during mechanical ventilation. in bronchoalveolar lavage fluid (tumor necrosis factor-a
and IL-8) were significantly higher in the larger tidal vol-
ume group. Choi et al.19 compared ventilation with12 mL/
During one-lung ventilation, the interaction between kg without PEEP versus 6 mL/kg with 10 cm PEEP and
auto-PEEP and applied PEEP becomes more complex. Pa- showed procoagulant changes in lavage fluid of the larger
tients with COPD typically develop auto-PEEP during one- tidal volume group after 5 hours of mechanical ventilation.
lung ventilation. Paradoxically, this is often beneficial A randomized controlled trial in 150 critically ill patients
during open-chest surgery, maintaining their end-ex- without ALI compared tidal volumes of 10 versus 6 mL/kg
piratory volume close to their functional residual capacity predicted body weight.20 The conventional tidal volumes
and improving oxygenation. Adding PEEP in these patients were associated with a sustained plasma increase in in-
is often counterproductive, raising end-expiratory vol- flammatory cytokines.
umes, and tends to counteract the redistribution of Work suggesting that noninjurious, or so-called pro-
pulmonary blood flow to the ventilated lung. However, tective, ventilatory settings can induce lung injury in pre-
patients with normal pulmonary function tend to fall be- viously healthy lungs is also important. An animal study
low their functional residual capacity at end-expiration using a very elegant murine ‘‘one-hit’’ VILI model showed
during one-lung ventilation, and adding PEEP is usually that even least injurious lung settings induced biochemical
beneficial (Figure 1).12 and histologic changes consistent with lung injury.21 Work
with rodents undergoing mechanical ventilation showed
significant gene expression (including genes involved in
MECHANICAL VENTILATION
immunity and inflammation) after only 90 minutes of
Historically, anesthesiologists have been taught to ven- protective ventilation.22 Whether this has an impact on
tilate patients with relatively large tidal volumes in the clinical outcome is unknown at this time.
perioperative period. Volumes as high as 15 mL/kg ideal ALI is the most common cause of postoperative respi-
body weight have been suggested to avoid intraoperative ratory failure and is associated with a markedly decreased
atelectasis.13 This far exceeds the normal spontaneous ti- postoperative survival rate.23 A prospective case–control
dal volumes (6 mL/kg) common to most mammals.14 Re- study by Fernandez-Perez and colleagues, assessing intra-
cent studies have identified the use of large tidal volumes as operative ventilator settings and ALI after elective surgery
a major risk factor for development of lung injury in me- in more than 4,000 patients, showed a 3% incidence of
chanically ventilated patients without ALI. Gajic et al.15 ALI in high-risk elective procedures. Compared with con-
reported that 25% of patients with normal lungs ventilated trols, patients with ALI had significantly lower post-

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

96 Slinger

Figure 1. A, The inspiratory limb of the lung pressure–volume compliance curve demonstrating the interaction of auto-PEEP and applied PEEP during one-
lung ventilation in a patient with COPD. This patient had 6 cm H2O auto-PEEP. Adding 5 cm H2O PEEP via the ventilator raised the total PEEP to 9 cm H2O,
causing the end-expiratory lung volume to rise above the lower inflection point of the compliance curve (which approximates functional residual capacity) and
causing a decrease in oxygenation as pulmonary blood flow was diverted to the nonventilated lung. B, The interaction of auto-PEEP and applied PEEP in a
patient with normal pulmonary function. This patient had 2 cm H2O auto-PEEP. Adding 5 cm H2O PEEP via the ventilator raised the total PEEP to 7 cm H2O,
which caused the end-expiratory lung volume to rise closer to the lower inflection point of the compliance curve and caused an increase in oxygenation.
COPD ¼ chronic obstructive pulmonary disease; PEEP ¼ positive end-expiratory pressure. Based on data from Slinger et al.12

operative survival and increased length of hospital stay. jury resulting from VILI showed that mechanical ven-
Interestingly, in this study, intraoperative peak airway tilation can lead to epithelial cell apoptosis in the kidney
pressure, but not tidal volume, PEEP, or FiO2, was asso- and small intestine with accompanying biochemical evi-
ciated with ALI. A retrospective cohort study specifically dence of organ dysfunction.26 In mice undergoing in-
assessing intraoperative risk factors for ARDS in critically jurious mechanical ventilation, alveolar stretch was found
ill patients found that the odds of developing ARDS were to induce adhesion molecules not only in the lung, but also
three times greater for those receiving fluid resuscitation at in the liver and kidney. In addition, cytokine and chemo-
a rate above 20 mL/kg/h than if less than 10 mL/kg/h were kine expression in pulmonary, hepatic, and renal tissue
given (odds ratio 3.1, 95% CI ¼ 1.0 to 9.9, P ¼ 0.05).24 after mechanical ventilation was accompanied by en-
Tidal volume and the number of blood products were not hanced recruitment of granulocytes to these organs.27
associated with ARDS in this study. Of interest, the ma- These studies partially explain the remote organ dysfunc-
jority of patients were ventilated with a tidal volume, cor- tion seen with ALI/ARDS and the role optimizing ven-
rected for ideal body weight, of 8 to 10 mL/kg and an tilatory strategies play in ameliorating it.
intraoperative PEEP of 0.

Intraoperative Ventilator-induced Lung Injury

Ventilator-induced Lung Injury Are the lung-protective strategies in ARDS28 applicable to
The phenomenon of VILI is well recognized and can be the intraoperative period, specifically in patients with
particularly significant in surgical procedures that require healthy lungs? A paper examining this question highlights
large transfusions or in cardiopulmonary bypass with as- the lack of randomized controlled trials assessing best in-
sociated lung ischemia–reperfusion injury. The deleterious traoperative tidal volume, PEEP, and use of intraoperative
effects of mechanical ventilation may be mediated by lo- lung recruitment.29 Although outcome studies are lacking,
calized inflammation and the systemic release of in- on the basis of what we know about the effects of
flammatory cytokines (biotrauma). Mechanical stretch mechanical ventilation, it seems reasonable to aim toward
from cyclical alveolar opening and closing sets up an in- protective ventilatory strategies in perioperative practice
flammatory response in the alveolar and vascular endo- (Supplemental Digital Content 1, http://links.lww.com/
thelial cells. Hyperinflation causes nuclear translocation of ASA/A518). Three randomized controlled studies of pa-
nuclear factor kB, a key regulator of the expression of tients undergoing major abdominal surgery have reached
multiple genes involved in the inflammatory response, and conflicting conclusions. A study by Treschan et al.30 found
upregulation of other proinflammatory cytokines. Poly- no difference in respiratory complications with lower ver-
morphonuclear leukocyte recruitment and activation ap- sus higher tidal volumes. Two studies, one by Futier et al.31
pear to be key components of the mechanical stretch- and another by Severgnini et al.,32 found improved out-
induced inflammatory response. The balance between comes with lower tidal volumes (Table 1). The important
apoptosis and necrosis is unfavorably altered by both difference between these studies may be that, in the first,
ischemia–reperfusion and mechanical stretch.25 both low-tidal-volume and high-tidal-volume groups re-
Biotrauma not only aggravates ongoing lung injury, it ceived PEEP, whereas in the latter two studies only the low-
also has important systemic consequences owing to the tidal-volume groups received PEEP. These findings await
spillover of these inflammatory mediators into the sys- confirmation in larger studies.33
temic circulation, inducing remote organ dysfunction. A study of one-lung ventilation for minimally invasive
A study assessing novel mechanisms of remote organ in- esophagectomy also found better pulmonary outcomes

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

Perioperative Lung Protection Strategies 97

Table 1. Studies of Lung-protective Ventilation in Abdominal Surgery

Ventilation in Study Ventilation in Control Primary
References Group Group Outcomes Results

Treschan et al.30 VT 6 mL/kg þ 5 PEEP VT 12 mL/kg þ 5 PEEP PFTs, LOS, respiratory Not significant
complications
Severgnini et al.32 VT 7 mL/kg þ 10 VT 9 mL/kg þ 0 PEEP PFTs, chest imaging, All except LOS significantly
PEEP þ recruitment blood gases, LOS decreased in study group
Futier et al.31 VT 6–8 mL/kg þ 6–8 VT 10–12 mL/kg þ 0 Respiratory failure, LOS All significantly decreased in
PEEP þ recruitment PEEP study group
LOS ¼ length of stay; PEEP ¼ positive end-expiratory pressure (in cm H2O); PFTs ¼ pulmonary function tests; VT ¼ tidal volume.

with lower tidal volumes and PEEP.34 One-lung ventilation PERIOPERATIVE MANAGEMENT
itself may be injurious to both the ventilated and the non-
ventilated lung (Figure 2; Supplemental Digital Content 2, Surgical Environmental Factors
http://links.lww.com/ASA/A519),35 and this injury de- Numerous factors in the surgical environment can con-
pends on the duration of one-lung ventilation.36 It may be tribute to lung injury, the most obvious being the surgical
best to avoid traditional one-lung ventilation whenever approach. Site of operation is an important predictor of
possible by applying continuous positive airway pressure pulmonary complications, with upper abdominal and
to the nonventilated lung (Supplemental Digital Content 3, thoracic incisions (any surgery approaching the dia-
http://links.lww.com/ASA/A520).37 This is a particularly phragm) being the most important.40 A decrease in respi-
attractive option during minimally invasive intrathoracic ratory complications has been documented when major
surgery that does not involve the lungs (i.e., cardiac, vas- cavity procedures are performed with minimally invasive
cular, or esophageal surgery). compared with open techniques.41,42 Atelectasis, a patho-
ALI after pneumonectomy is a well-known complica- logic state that can contribute to lung injury, occurs fre-
tion with a high mortality rate. Traditionally, the compli- quently after open surgical procedures and in up to 90% of
cation has often been blamed on the anesthesiologist’s patients undergoing general anesthesia.43 Thus, anes-
administering of excess fluids during surgery. However, thesiologists must be aware of techniques to avoid or treat
there is now evidence that this ALI is related more to the it.44 Although open to debate, retrospective45,46 and pro-
use of excessively large tidal volumes during one-lung spective47 studies have shown that appropriate thoracic
ventilation than to fluids (Supplemental Digital Content 4, epidural analgesia reduces the incidence of respiratory
http://links.lww.com/ASA/A521).38 Although there has complications (atelectasis, pneumonia, and respiratory
not been a convincing human prospective study on the use failure) after major abdominal and thoracic surgery. The
of small versus large tidal volumes during one-lung ven- benefits of epidural analgesia appear to be directly pro-
tilation, one large animal study clearly showed that the use portional to the severity of the patients’ underlying lung
of large versus small (12 vs. 6 mL/kg) tidal volumes, with disease. Patients with COPD seem to derive the most
the addition of PEEP 5 cm H2O in the small-volume group, benefit from epidural analgesia.48 Although it has not been
resulted in a significant increase in lung water after pneu- specifically studied in high-risk patients, reviews compar-
monectomy in the large-tidal-volume group (Figure 3).39 ing paravertebral block with epidural analgesia in patients

Figure 2. A, Comparison of histologic lung injury scores from the dependent (D) lung of pigs after either one-lung ventilation (OLV) or two-lung ventilation
(2LV). B, Comparison of histologic lung injury scores from the nondependent (ND) lung of pigs after either one-lung ventilation (OLV) or two-lung ventilation
(2LV). Tidal volumes were 10 mL/kg with PEEP 5 cm H2O for both OLV and 2LV. HR ¼ alveolar hemorrhage; NI ¼ neutrophil infiltration. *Po 0.05.

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

98 Slinger

proinflammatory response in the lungs to a host of insults,

whether this is before, during, or after the insult.

Volatile anesthetic agents have

immune-modulatory effects.

Ultraprotective Lung Ventilation

Following along the continuum of lung-protective ven-
tilation in ALI/ARDS is the concept of ultraprotective
Figure 3. Comparison of the extravascular lung water from the nonoper-
ated lung after pneumonectomy in sheep. Control animals were ventilated ventilation. This concept utilizes extracorporeal lung
with a tidal volume (VT) of 12 mL/kg during two-lung ventilation for 4 hours support devices and near-static ventilation.57 ARDSNet
and did not undergo a pneumonectomy nor did they receive one-lung and animal data demonstrate that lower tidal volumes
ventilation. The lung water was significantly higher after 4 hours of one-lung
ventilation with VT 12 mL/kg in the pneumonectomy animals than in the (3 mL/kg compared with 6 to 12 mL/kg) significantly re-
pneumonectomy animals ventilated with VT 6 mL/kg plus 5 cm H2O PEEP. duces endothelial and epithelial injury.58,59 In other words,
Based on data from Kuzkov et al.39 ‘‘protective’’ tidal volumes can still induce VILI. However,
CO2 clearance and oxygenation become an issue at these
undergoing thoracic surgery showed equivalent analgesia lower minute volumes. The Novalungs, a pumpless de-
efficacy but a better side effect profile and a lower com- vice, allows for this marked reduction in minute ven-
plication rate with paravertebral block.49,50 In patients tilation and the simultaneous correction of PaCO2 and pH.
who develop early desaturation after major abdominal An animal model of postpneumonectomy ARDS using the
surgery, aggressive physiotherapy with continuous positive Novalungs with tidal volumes of 2.2 mL/kg and a respi-
airway pressure in the postoperative period leads to lower ratory rate of 6 breaths/min showed significantly better
rates of major respiratory complications.51 outcomes compared with conventional lung-protective
strategies.60 Case reports in human patients in a variety of
clinical scenarios have been encouraging.61 Tidal volumes
Role of Volatile Anesthetic Agents in Lung Protection at or below 3 mL/kg, low inspiratory plateau pressure,
Volatile anesthetic agents have immune-modulatory effects. high PEEP, and low respiratory rates are all possible with
Much work has been done, especially in the cardiac setting, the Novalungs, causing less VILI and less subsequent re-
on the role of volatile agents in ischemia–reperfusion injury mote secondary organ failure. In a randomized trial in
and in preconditioning and postconditioning. Recent stud- patients with severe ARDS, the use of extracorporeal
ies in models of ALI during one-lung ventilation and in cases membrane oxygenation with a pump in combination with
of lung ischemia–reperfusion52 suggest that volatile agents protective ventilation significantly increased the survival
may act as preconditioning and postconditioning agents, rate to 63% versus 47% with conventional ventilation
inducing lung protection by inhibiting the expression of strategies.62
proinflammatory mediators. Isoflurane pretreatment in an
endotoxin-mediated animal model of lung injury exerted Fluids, Inflammation, and the Glycocalyx
protective effects as evidenced by reduction in recruitment It has long been a concern that excess amounts of intra-
of polymorphonuclear leukocytes and microvascular pro- venous fluids predispose patients to develop ALI. How-
tein leakage.53 Postconditioning with sevoflurane attenu- ever, a conflicting concern for anesthesiologists is that fluid
ated lung damage and preserved lung function in an in vivo restriction in thoracic surgery may contribute to post-
rat ALI model.54 In a prospective study, patients undergoing operative renal dysfunction, which previously was re-
thoracic surgery with one-lung ventilation were randomized ported to be associated with a very high (19%) death
to either propofol or sevoflurane anesthesia.55 Assessing rate.63 A more recent retrospective study assessing all
inflammatory markers in the nonventilated lung, the au- pulmonary resection patients found that acute kidney in-
thors showed an attenuated inflammatory reaction with jury, as defined by Acute Kidney Injury Network criteria,
sevoflurane. Notably, the sevoflurane group had improved occurred in 67 of 1,129 patients (6%) and was not asso-
outcome and significantly lower adverse events overall. ciated with a statistically significant increase in mortality
A study comparing one-lung ventilation with desflurane compared with patients who did not experience acute
versus propofol anesthesia examined the inflammatory re- kidney injury (3% vs. 1%).64
sponse in the ventilated lung.56 The inflammatory markers Fluid requirements vary widely between patients and
IL-8, IL-10, PMN elastase, and tumor necrosis factor-a procedures, and ultimately represent the sum of pre-
were significantly lower in the desflurane group (Figure 4). operative deficits, maintenance requirements, and ongoing
Although much remains to be done, this exciting work does losses. Fluid management for major esophageal surgery
point toward a role for volatile agents in attenuating the is especially challenging.65 Particularly if fluid intake

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

Perioperative Lung Protection Strategies 99

Figure 4. Comparison of the increase in inflammatory cytokine concentrations from bronchoalveolar lavage (BAL) samples before (1) and after (2) one-lung
ventilation (OLV) with intravenous versus volatile anesthesia. Graphs comparing (A) propofol versus desflurane and (B) the nonventilated lung with propofol
versus sevoflurane (*Po 0.05). Based on data from De Conno et al.55 and Schilling et al.56

has been limited by esophageal obstruction or dysphagia, responsiveness or correlate with circulating blood volume
patients undergoing esophageal procedures may be rela- after transthoracic esophagectomy.73,74 However, fluid
tively hypovolemic after long preoperative fasts, which can responsiveness remains an elusive goal for managing pa-
complicate fluid management. Perioperative losses occur tients (Figure 5). As patients approach the upper inflection
via a number of mechanisms including urinary, gastro- point of the Frank–Starling curve, small increases in car-
intestinal, and evaporative losses, bleeding, and interstitial diac output create large increases in lung water, and this
fluid shifting. This shift of fluid from the vascular com- effect is exacerbated in a situation of increased capillary
partment into the interstitial space accompanies surgical permeability such as sepsis.76 In addition to the potential
trauma and is likely to reflect vascular injury and loss of importance of the amount and timing of fluid admin-
endothelial integrity. So-called ‘‘third space’’ losses de- istration, there is some clinical evidence that the choice of
scribe fluid loss into noninterstitial extracellular spaces fluid type may be important in affecting clinical out-
that are not in equilibrium with the vascular compartment, comes.77 Intravascular colloid retention during treatment
and thus are considered to be a ‘‘nonfunctional’’ ex- of hypovolemia may approach 90%, versus 40% when
tracellular fluid compartment. However, it is very possible administered during normovolemia.66
that the ‘‘third space’’ does not exist, and was described as a The relationship of hydrostatic and oncotic pressure to
result of measurement errors in early studies of the fluid determine fluid flux across a semipermeable membrane,
compartments in the body (Supplemental Digital Content 5,
http://links.lww.com/ASA/A522).66
One of the factors complicating fluid management for
esophageal resection is that thoracic epidural analgesia has
been shown to improve outcome for these patients,67 but
its use tends to contribute to hypotension. Hypotension is
well known to contribute to ischemia of the gut anasto-
mosis,68 and treatment with excessive fluids is likely to
exacerbate the problem.69 Many surgeons are concerned
about the effects of vasopressors on anastomotic gut blood
flow.70 However, several animal studies suggest that
treatment of intraoperative hypotension with norepi-
nephrine does not cause any reduction of gut blood flow in
the presence of normovolemia.71,72
An ideal fluid regimen for major procedures, including
esophageal surgery, is individualized and optimizes cardiac
output and oxygen delivery while avoiding excessive fluid
administration. There is some evidence that fluid therapies
designed to achieve individualized and specific flow-re- Figure 5. Superimposition of the Frank–Starling (F–S) and Marik–Phillips
(M–P) curves demonstrating the effects of increasing preload on cardiac
lated hemodynamic endpoints such as stroke volume or stroke volume (SV) and extravascular lung water (EVLW). Note that as the
cardiac index (collectively referred to as goal-directed fluid patient approaches the upper inflection point of the Frank–Starling curve,
therapy) may provide a superior alternative to fixed regi- small increases in cardiac output (CO, arrow ‘‘b’’) result in large increases in
EVLW. This effect is exacerbated in sepsis. From Marik PE, Lemson J. Fluid
mens or those based on static measures of cardiac filling, responsiveness: An evolution of our understanding. Br J Anaesth 2014;
such as central venous pressure, which do not predict fluid 112(4):617–20, by permission of Oxford University Press.75

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

100 Slinger

such as the lung capillary endothelium, was described in a

classic equation developed in 1896 by Starling.78 However,
several subsequent clinical observations are not explained
by the Starling formula, such as the intact organism’s rel-
ative resistance to developing edema and the inability of
therapy with hyperoncotic agents to draw fluid from the
pulmonary interstitium into the vascular compartment.79
This discrepancy is now attributed to the glycocalyx,
a microcilial layer that lines the endothelium and acts
as a molecular sieve (Figure 6).80 This layer tends to in-
crease the oncotic pressure on the inner surface of the
endothelium and decrease leukocyte and platelet adhesion
to the endothelium. The glycocalyx deteriorates during
ischemia–reperfusion injury and in the presence of a wide
variety of inflammatory mediators such as cytokines,
which probably contributes to the increased vascular per-
meability seen in these situations (Supplemental Digital
Content 6, http://links.lww.com/ASA/A523). In addition,
the glycocalyx deteriorates in the presence of atrial na-
triuretic peptide, which may explain the increase in plasma
protein filtration that has been seen with colloid boluses.
Protecting the glycocalyx may be among the anesthesiol-
ogist’s most important duties in the perioperative period.
Volatile anesthetics may have a protective effect on the
glycocalyx (Figure 7).81,82

Other Therapies for Lung Protection

Several additional therapies may play a future role in lung
protection. The place of permissive hypercapnia in pro-
tective ventilation has been alluded to previously and, as
found in the original ARDSNet data, may be protective in
the presence of higher tidal volumes.83 Hypercapnic
acidosis is protective in a variety of models of ALI. Bene-

Figure 7. Sevoflurane reduces endothelial glycocalyx shedding in response

to ischemia–reperfusion. Isolated guinea pig hearts were exposed to
20 minutes of warm ischemia and 10 minutes of reperfusion. Electron
micrographs show: A, intact endothelial glycocalyx in the nonischemic
perfusion group; B, marked endothelial glycocalyx shedding after ischemia–
Figure 6. The glycocalyx model for fluid exchange between the intra- reperfusion; C, the endothelial glycocalyx is relatively preserved from
vascular and interstitial space. The various components of the glycocalyx ischemia–reperfusion after pretreatment with 1 minimum alveolar concen-
model and revised Starling forces are shown. In steady state, net filtration tration of sevoflurane. From Chappell D, Heindl B, Jacob M, et al.:
into the interstitium occurs and is subsequently removed by the lymphatic Sevoflurane reduces leukocyte and platelet adhesion after ischemia-
system. EC ¼ endothelial cell; EGL ¼ endothelial glycocalyx layer; pc ¼ capil- reperfusion by protecting the endothelial glycocalyx. Anesthesiology 2011;
lary oncotic pressure; pg ¼ oncotic pressure in subglycocalyx space; 15:483–91. Copyright r 2011, the American Society of Anesthesiologists
pi ¼ interstitial oncotic pressure; Pc ¼ capillary hydrostatic pressure; Inc., Lippincott Williams & Wilkins.82
Pi ¼ interstitial hydrostatic pressure.

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

Perioperative Lung Protection Strategies 101

ficial effects include attenuation of lung neutrophil 4. Simpson SQ: Oxygen-induced acute hypercapnia in chronic obstruc-
recruitment, pulmonary and systemic cytokine concen- tive pulmonary disease: What’s the problem? Crit Care Med 2002;
30:258–60.
trations, cell apoptosis, and free radical injury.84 Inhaled 5. Hanson CW. III, Marshall BE, Frasch HF, Marshall C: Causes of
hydrogen sulfide, by inhibiting inflammatory and apop- hypercarbia in patients with chronic obstructive pulmonary disease.
totic responses, showed beneficial effects in a model of Crit Care Med 1996; 24:23–8.
6. Morgan MDL, Edwards CW, Morris J, Matthews HR: Origin and
VILI independent of its effects on body temperature.85 In- behavior of emphysematous bullae. Thorax 1989; 44:533–8.
haled aerosolized activated protein C in a sheep model of 7. O’Donnell DE, Sanii R, Anthonisen NR, Younes M: Effect of
dynamic airway compression on breathing pattern and respiratory
ALI demonstrated improved oxygenation as well as lung
sensation in severe chronic obstructive pulmonary disease. Am Rev
aeration (as assessed by computed tomography scan).86 Respir Dis 1987; 135:912–8.
b-Adrenergic agonists have potential benefits in that they 8. Slinger P: Anesthesia for patients with COPD. Curr Rev Clin Anesth
accelerate the rate of alveolar fluid clearance by increasing 1995; 15:169–80.
9. Ben-David B, Stonebraker VC, Hershman R, Frost CL, Williams HK:
cellular cyclic adenosine monophosphate, and they have Survival after failed intraoperative resuscitation: A case of ‘‘Lazarus
antiinflammatory properties.87 A randomized controlled syndrome’’. Anesth Analg 2001; 92:690–4.
trial in 40 patients with ALI showed a decrease in ex- 10. Slinger P, Hickey D: The interaction between applied PEEP and auto-
PEEP during one-lung ventilation. J Cardiothorac Vasc Anesth 1998;
travascular lung water and plateau airway pressure with 12:133–7.
intravenous salbutamol, although there was no difference 11. Caramez MP, Borges JB, Tucci MR, et al.: Paradoxical responses to
positive end-expiratory pressure in patients with airway obstruction
in outcome.88 Randomized placebo-controlled trials of during controlled ventilation. Crit Care Med 2005; 33:1519–28.
different therapies including surfactant, prone positioning, 12. Slinger P, Kruger M, McRae K, Winton T: The relation of the
inhaled nitric oxide, and antiinflammatory agents have not static compliance curve and positive end-expiratory pressure to
shown significant clinical benefits in patients with estab- oxygenation during one-lung ventilation. Anesthesiology 2002; 95:
1096–102.
lished ALI.89 Although it is unreasonable to expect there to 13. Bendixen HH, Hedley-White J, Laver MB: Impaired oxygenation in
be a single therapy (or ‘‘magic bullet’’) that will prevent surgical patients during general anesthesia with controlled ventila-
tion: A concept of atelectasis. N Engl J Med 1963; 96:156–66.
ALI, this exciting research does hold promise in furthering
14. Tenny SM, Remmers JE: Comparative quantitative morphology of
our understanding and management of injured or at-risk the mammalian lung: Diffusing area. Nature 1963; 196:54–6.
lungs. 15. Gajic O, Dara SI, Mendez JL, et al.: Ventilator-associated lung injury
in patients without acute lung injury at the onset of mechanical
ventilation. Crit Care Med 2004; 32:1817–24.
SUMMARY 16. Gajic O, Frutos-Vivar F, Esteban A, Hubmayr RD, Anzueto A:
Ventilator settings as a risk factor for acute respiratory distress
Anesthesiologists manage a heterogenous group of pa- syndrome in mechanically ventilated patients. Intensive Care Med
2005; 31:922–26.
tients in the perioperative period, from those with healthy 17. Michelet P, D’Journo X-B, Roch A, et al.: Protective ventilation
lungs or ‘‘at-risk’’ lungs to patients with established ALI/ influences systemic inflammation after esophagectomy: A random-
ARDS. More patients are at risk for ALI during and after ized controlled study. Anesthesiology 2006; 105:911–19.
18. Pinheiro de Oliveira R, Hetzel MP, Silva M, Dallegrave D,
surgery than previously thought. Appropriate perioper- Friedman G: Mechanical ventilation with high tidal volume induces
ative management may prevent or ameliorate this lung inflammation in patients without lung disease. Crit Care 2010;
injury. Are the proven lung-protective strategies in ARDS 14:R39.
19. Choi G, Wolthuis EK, Bresser P, et al.: Mechanical ventilation with
applicable to the perioperative environment, specifically in lower tidal volumes and positive end-expiratory pressure prevents
patients with healthy lungs? Although evidence from alveolar coagulation in patients without lung injury. Anesthesiology
randomized controlled trials is lacking, applying pro- 2006; 105:689–95.
20. Determann R, Royakkers A, Wolthuis EK, et al.: Ventilation with
tective ventilatory strategies in the intraoperative period lower tidal volumes as compared with conventional tidal volumes for
seems reasonable on the basis of our current understanding patients without acute lung injury: A preventive randomized
of mechanical ventilation and lung injury. Apart from controlled trial. Crit Care 2010; 14:R1.
21. Wolthuis EK, Vlaar APJ, Choi G, et al.: Mechanical ventilation using
ventilation strategies, the anesthesiologist also needs to be non-injurious ventilation settings causes lung injury in the absence of
concerned about fluid management and prevention of pre-existing lung injury in healthy mice. Crit Care 2009; 13:R1.
systemic inflammation as vital aspects of lung protection 22. Ng CSH, Song Wan Ho AMH, Underwood MJ: Gene expression
changes with a ‘‘non-injurious’’ ventilation strategy. Crit Care 2009;
(Supplemental Digital Content 7, http://links.lww.com/ 13:403–10.
ASA/A524). 23. Fernandez-Perez ER, Sprung J, Alessa B, et al.: Intraoperative
ventilator settings and acute lung injury after elective surgery: A
nested case control study. Thorax 2009; 64:121–27.
24. Hughes C, Weavind L, Banerjee A, et al.: Intraoperative risk factors
for acute respiratory distress syndrome in critically ill patients.
REFERENCES Anesth Analg 2010; 111:464–67.
25. Lionetti V, Recchia FA, Ranieri VM: Overview of ventilator-induced
1. Slinger P, Kilpatrick B: Perioperative lung protective strategies in
lung injury mechanisms. Curr Opin Crit Care 2005; 11:82–6.
cardiothoracic anesthesia. Anesth Clin 2012; 30:607–28.
2. Parot S, Saunier C, Gauthier H, Milic-Emile J, Sadoul P: Breathing 26. Imai Y, Parodo J, Kajikawa O, et al.: Injurious mechanical ventilation
pattern and hypercapnia in patients with obstructive pulmonary and end-organ epithelial cell apoptosis and organ dysfunction in an
disease. Am Rev Respir Dis 1980; 121:985–91. experimental model of acute respiratory distress syndrome. JAMA
3. Aubier M, Murciano D, Milic-Emili J, et al.: Effects of the 2003; 280:2104–112.
27. Hegeman MA, Henmus MP, Heijnen CJ, et al.: Ventilator-induced
administration of O2 on ventilation and blood gases in patients with
endothelial activation and inflammation in the lung and
chronic obstructive pulmonary disease during acute respiratory
distal organs. Crit Care 2009; 13:R182.
failure. Am Rev Respir Dis 1980; 122:747–54.

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

102 Slinger

28. The Acute Respiratory Distress Syndrome Network. Ventilation with 52. Fujinaga T, Nakamura T, Fukuse T, et al.: Isoflurane inhalation after
lower tidal volumes as compared with traditional tidal volumes for circulatory arrest protects against warm ischemia reperfusion injury
acute lung injury and the acute respiratory distress syndrome. N Engl of the lungs. Transplantation 2006; 82:1168–74.
J Med 2000; 342:1301–8. 53. Reutershan J, Chang D, Hayes JK, Ley K: Protective effects of
29. Beck-Schimmer B, Schimmer RC: Perioperative tidal volume and isoflurane pretreatment in endotoxin-induced lung injury. Anesthesi-
intraoperative open lung strategy in healthy lungs: Where are ology 2006; 104:511–7.
we going? Best Pract Res Clin Anaesthesiol 2010; 24:199–210. 54. Voigtsberger S, Lachmann RA, Leutert AC, et al.: Sevoflurane
30. Treschan TA, Kaisers W, Schafer MS, et al.: Ventilation with low tidal ameliorates gas exchange and attenuates lung damage in experimen-
volumes during upper abdominal surgery does not improve post- tal lipopolysaccharide-induced lung injury. Anesthesiology 2009;
operative lung function. Br J Anaesth 2012; 109:263–71. 111:1238–48.
31. Futier E, Constantin J-M, Paugam-Burtz C, et al.: A trial of 55. De Conno E, Steurer MP, Wittlinger M, et al.: Anesthetic-induced
intraoperative low tidal-volume in abdominal surgery. N Engl J improvement of the inflammatory response to one-lung ventilation.
Med 2013; 369:428–36. Anesthesiology 2009; 110:1316–26.
32. Severgnini P, Selmo G, Lanza C, et al.: Protective mechanical 56. Schilling T, Kozian A, Kretzschmar M, et al.: Effects of propofol and
ventilation during general anesthesia for open abdominal surgery desflurane anaesthesia on the alveolar inflammatory response to one-
improves postoperative pulmonary function. Anesthesiology 2013; lung ventilation. Br J Anaesth 2007; 99:368–75.
118:1254–7. 57. The Cardiothoracic Surgery Network website. Available at:
33. Goldenberg NM, Steinberg BE, Lee WL, Wijeysundera DN, www.CTSNet.org. Accessed January 30, 2015.
Kavanagh BP: Lung-protective ventilation in the operating room. 58. Hager DN, Krishnan JA, Hayden DL, Brower RG. ARDS Clinical
Time to implement? Anesthesiology 2014; 121:184–8. Trials Network. Tidal volume reduction in patients with acute lung
34. Shen Y, Zhong M, Wu W, et al.: The impact of tidal volume on injury when plateau pressures are not high. Am J Respir Crit Care
pulmonary complications following minimally invasive esophagec- Med 2005; 10:1241–5.
tomy. J Thorac Cardiovasc Surg 2013; 146:1267–73. 59. Frank JA, Gutierrez JA, Jones KD, et al.: Low tidal volume reduces
35. Kozian A, Schilling T, Freden F, et al.: One-lung ventilation induces epithelial and endothelial injury in acid-injured rat lungs. Am J Respir
hyperperfusion and alveolar damage in the ventilated lung. Br J Crit Care Med 2002; 165:242–9.
Anaesth 2008; 100:549–59. 60. Iglesias M, Jungebluth P, Petit C, et al.: Extracorporeal lung
36. Tekinbas C, Ulusoy H, Yulug E, et al.: One-lung ventilation: For membrane provides better lung protection than conventional treat-
how long? J Cardiothorac Vasc Surg 2007; 134:405–10. ment for severe postpneumonectomy noncardiogenic acute respira-
37. Verhage RJ, Boone J, Rijkers GT, et al.: Reduced local immune tory distress syndrome. J Thorac Cardiovasc Surg 2008; 6:1362–71.
response with continuous positive airway pressure during 61. Mallick A, Elliot S, McKinlay J, Bodenham A: Extracorporeal carbon
one-lung ventilation for oesophagectomy. Br J Anaesth 2014; 112: dioxide removal using the Novalung in a patient with intracranial
920–8. bleeding. Anaesthesia 2007; 62:72–4.
38. Slinger P: Postpneumonectomy pulmonary edema: Good news, 62. Peek GJ, Mugford M, Tiruvoipati R, et al.: Efficacy and economic
bad news. Anesthesiology 2006; 105:2–5. assessment of conventional ventilatory support versus extracorporeal
39. Kuzkov V, Subarov E, Kirov M, et al.: Extravascular lung water after membrane oxygenation for severe adult respiratory failure (CESAR): A
pneumonectomy and one-lung ventilation in sheep. Crit Care Med multicentre randomised controlled trial. Lancet 2009; 374:1351–63.
2007; 35:1550–9. 63. Gollege G, Goldstraw P: Renal impairment after thoracotomy:
40. Smetana GW: Postoperative pulmonary complications: An update Incidence, risk factors and significance. Ann Thorac Surg 1994;
on risk assessment and reduction. Cleve Clin J Med 2009; 76: 58:524–8.
S60–5. 64. Ishikawa S, Greisdale DEG, Lohser J: Acute kidney injury after lung
41. Weller WE, Rosati C: Comparing outcomes of laparoscopic versus resection surgery: Incidence and perioperative risk factors. Anesth
open bariatric surgery. Ann Surg 2008; 248:10–15. Analg 2012; 114:1256–62.
42. Ramivohan SM, Kaman L, Jindal R, Singh R, Jindal SK: Post- 65. Blank RS, Huffmyer JL, Jaeger JM: Anesthesia for esophageal surgery
operative pulmonary function in laparoscopic versus open cholecys- (chapter 30). In: Slinger P, ed. Principles and Practice of Anesthesia
tectomy: Prospective, comparative study. Indian J Gastroenterol for Thoracic Surgery. New York: Springer; 2011.
2005; 24:6–8. 66. Chappell D, Jacob M, Hofmann-Kiefer K, Conzen P, Rehm M: A
43. Duggan M, Kavanagh B: Pulmonary atelectasis: A pathogenic rational approach to perioperative fluid management. Anesthesi-
perioperative entity. Anesthesiology 2005; 102:834–54. ology 2008; 109:723–40.
44. Tusman G, Bohm SH, Suarez-Shipman F: Alveolar recruitment 67. Cense HA, Lagarde SM, de Jong K, et al.: Association of no epidural
improves ventilatory efficiency of the lungs during anesthesia. Can J analgesia with post-operative morbidity and mortality after trans-
Anaesth 2004; 51:723–7. thoracic esophageal cancer resection. J Am Coll Surg 2006;
45. Ballantyne JC, Carr DB, de Ferranti S: The comparative effects of 202:395–400.
postoperative analgesic therapies on pulmonary outcome: Cumula- 68. Pathak D, Pennefather SH, Russell GN, et al.: Phenylephrine infusion
tive meta-analysis of randomized, controlled trials. Anesth Analg improves blood flow to the stomach during oesophagectomy in the
1998; 86:598–612. presence of a thoracic epidural analgesia. Eur J Cardiothorac Surg
46. Liu SS, Wu CL: Effect of postoperative analgesia on major post- 2013; 44:130–3.
operative complications: A systematic update of the evidence. Anesth 69. Holte K, Sharrock NE, Kehlet H: Pathophysiology and clinical
Analg 2007; 3:689–702. implications of perioperative fluid excess. Br J Anaesth 2002;
47. Rigg J, Jamrozik K, Myles P, et al.: Epidural anaesthesia and analgesia 89:622–32.
and outcome after major surgery: A randomized trial. Lancet 2002; 70. Theodorou D, Drimousis PG, Larentzakis A, et al.: The effects of
359:1276–82. vasopressors on perfusion of gastric graft after esophagectomy.
48. Licker MJ, Widikker I, Robert J, et al.: Operative mortality and J Gastrointest Surg 2008; 12:1497–501.
respiratory complications after lung resection for cancer: Impact of 71. Klijn E, Niehof S, de Jong J, et al.: The effect of perfusion pressure on
chronic obstructive pulmonary disease and time trends. Ann Thorac gastric tissue blood flow in an experimental gastric tube model.
Surg 2006; 81:1830–8. Anesth Analg 2010; 110:541–6.
49. Scarci M, Joshi A, Attia R: In patients undergoing thoracic surgery is 72. Hiltebrand LB, Koepfli E, Kimberger O, Sigurdsson GH, Brandt S:
paravertebral block as effective as epidural analgesia for pain Hypotension during fluid restricted abdominal surgery. Anesthesi-
management. Interact Cardiovasc Thorac Surg 2010; 10:92–6. ology 2011; 114:557–64.
50. Davies RG, Myles PS, Graham JM: A comparison of the analgesic 73. Oohashi S, Endoh H: Does central venous pressure or pulmonary
efficacy and side effects of paravertebral vs. epidural blockade for capillary wedge pressure reflect the status of circulating blood volume
thoracotomy—A systematic review and meta-analysis of randomized in patients after extended transthoracic esophagectomy? J Anesth
trials. Br J Anaesth 2006; 96:418–26. 2005; 19:21–5.
51. Squadrone V, Coha M, Cerutti E, et al.: Continuous positive airway 74. Kobayashi M, Ko M, Kimura T, et al.: Perioperative monitoring of
pressure for the treatment of postoperative hypoxemia: A random- fluid responsiveness after esophageal surgery using stroke volume
ized controlled trial. JAMA 2005; 293:589–95. variation. Expert Rev Med Devices, 52008:311–6.

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.

Perioperative Lung Protection Strategies 103

75. Marik PE, Lemson J: Fluid responsiveness: An evolution of our 82. Chappell D, Heindl B, Jacob M, et al.: Sevoflurane reduces leukocyte
understanding. Br J Anaesth 2014; 112:617–20. and platelet adhesion after ischemia-reperfusion by protecting the
76. Marik PE, Lemson J: Fluid responsiveness: An evolution of our endothelial glycocalyx. Anesthesiology 2011; 15:483–91.
understanding. Br J Anaesth, 2104112:681–5. 83. Kregenow DA, Rubenfeld GD, Hudson LD, Swenson ER: Hyper-
77. Wei S, Tian J, Song X, Chen Y: Association of perioperative fluid capnic acidosis and mortality in acute lung injury. Crit Care Med
balance and adverse surgical outcomes in esophageal cancer and 2006; 34:1–7.
esophagogastric junction cancer. Ann Thorac Surg 2008; 86: 84. Curley G, Laffey JG, Kavanagh BP: Bench-to-bedside: Carbon
266–72. dioxide. Crit Care 2010; 14:220.
78. Starling EH: On the absorption of fluids from the connective 85. Faller S, Ryter SW, Choi AMK, et al.: Inhaled hydrogen sulfide protects
tissue spaces. J Physiol 1896; 19:312–26. against ventilator-induced lung injury. Anesthesiology 2010; 113:104–15.
79. Woodcock TE, Woodcock TM: Revised Starling equation and the 86. Waerhaug K, Kuzkov VV, Kuklin VN, et al.: Inhaled aerosolised
glycocalyx model of transvascular fluid exchange. Br J Anaesth 2012; recombinant human activated protein C ameliorates endotoxin-
108:384–94. induced lung injury in anesthetised sheep. Crit Care 2009; 13:R51.
80. Chau EH, Slinger P: Perioperative fluid management for pulmonary 87. Matthay M: b-Adrenergic agonist therapy as a potential treatment
resection surgery and esophagectomy. Semin Cardiothorac Vasc for acute lung injury. Am J Respir Crit Care Med 2006; 173:254–5.
Anesth 2014; 18:36–44. 88. Perkins GD, McAuley DF, Thickett DR, Gao F: The b-agonist lung
81. Annecke T, Rehm M, Bruegger D: Ischemia-reperfusion-induced injury trial. Am J Respir Crit Care Med 2006; 173:281–7.
unmeasured anion generation and glycocalyx shedding: Sevoflurane 89. Bernard GR: Acute respiratory distress syndrome: A historical
versus propofol anesthesia. J Invest Surg 2012; 25:162–8. perspective. Am J Respir Crit Care Med 2005; 172:798–806.

Copyright r 2015 American Society of Anesthesiologists. All rights reserved.