Can 'permissive' hypercapnia modulate the severity of sepsis-induced ALI/ARDS?

Ventilatory strategies that reduce lung stretch by reducing tidal and minute ventilation, which results in a ‘permissive’ hypercapnic acidosis, improve outcome in patients with acute lung injury/acute respiratory distress syndrome (ALI/ARDS) [1], [2]. Reassuringly, evidence from clinical studies attests to the safety and lack of detrimental effects of hypercapnic acidosis [2]. Of particular importance, a secondary analysis of data from the ARDSnet tidal volume study [1] demonstrated that the presence of hypercapnic acidosis at the time of randomization was associated with improved patient survival in patients who received high tidal volume ventilation [3]. These findings have resulted in a shift in paradigms regarding hypercapnia — from avoidance to tolerance — with hypercapnia increasingly permitted in order to realize the benefits of low lung stretch. Consequently, low tidal and minute volume ventilation and the accompanying ‘permissive’ hypercapnia are now the standard of care for patients with ALI/ARDS, and are increasingly used in the ventilatory management of a diverse range of diseases leading to acute severe respiratory failure, including asthma and chronic obstructive pulmonary disease.


Introduction
Ventilatory strategies that reduce lung stretch by reducing tidal and minute ventilation, which results in a 'permissive' hypercapnic acidosis, improve outcome in patients with acute lung injury/acute respiratory distress syndrome (ALI/ARDS) [1,2]. Reassuringly, evidence from clinical studies attests to the safety and lack of detrimental eff ects of hypercapnic acidosis [2]. Of parti cu lar importance, a secondary analysis of data from the ARDSnet tidal volume study [1] demonstrated that the presence of hypercapnic acidosis at the time of randomization was associated with improved patient survival in patients who received high tidal volume ventilation [3]. Th ese fi ndings have resulted in a shift in paradigms regarding hypercapnia -from avoidance to tolerancewith hypercapnia increasingly permitted in order to realize the benefi ts of low lung stretch. Conse quently, low tidal and minute volume ventilation and the accompanying `permissive' hypercapnia are now the standard of care for patients with ALI/ARDS, and are increasingly used in the ventilatory management of a diverse range of diseases leading to acute severe respira tory failure, includ ing asthma and chronic obstructive pulmonary disease.
Th e infl ammatory response plays a central role in the pathogenesis of injury and in the repair process in ALI/ ARDS [4]. Infl ammation is a highly conserved process in evolution, which is essential for survival. It functions to resolve the injurious process, facilitate repair, and return the host to a state of homeostasis. Th e ideal infl ammatory process is rapid, causes focused destruction of pathogens, yet is specifi c and ultimately self-limiting [5]. In contrast, when the infl ammatory response is dysregulated or persistent, this can lead to excessive host damage, and contribute to the pathogenesis of lung and systemic organ injury, leading to multiple organ failure and death. Th e potential for hypercapnia and/or its associated acidosis to potently inhibit the immune response is increasingly recognized [6,7]. Where the host immune response is a major contri bu tor to injury, such as in non-septic ALI/ ARDS, these eff ects would be expected to result in potential benefi t. Th is has been demonstrated clearly in relevant pre-clinical ALI/ARDS models, where hypercapnic acidosis has been demonstrated to attenuate ALI induced by free radicals [8], pulmonary [9] and systemic ischemia-reperfusion [10], pulmonary endo toxin instillation [11], and excessive lung stretch [12]. Th e protective eff ects of hypercapnic acidosis in these models appear due, at least in part, to its anti-infl ammatory eff ects.
Th e eff ects of hypercapnia in sepsis-induced lung injury, where a robust immune response to microbial infec tion is central to bacterial clearance and recovery, is less clear. Of concern, severe sepsis-induced organ failure, whether pulmonary or systemic in origin, is the leading cause of death in critically ill adults and children [13]. Sepsis-induced ARDS is associated with the highest mortality rates. Evidence suggests that approximately 40% of patients with severe sepsis develop ARDS [13]. Furthermore, infection frequently complicates critical illness due to other causes, with an infection prevalence of over 44% reported in this population [14]. Th ese issues underline the importance of understanding the eff ects of hypercapnia on the immune response, and the implications of these eff ects in the setting of sepsis.

Function of the innate immune response
Th e immune system can be viewed as having two interconnected branches, namely the innate and adaptive immune responses [5]. Th e innate immune system is an ancient, highly conserved response, being present in Can 'permissive' hypercapnia modulate the severity of sepsis-induced ALI/ARDS? Gerard Curley, Mairead Hayes, and John G Laff ey* This article is one of eleven reviews selected from the Annual Update in Intensive Care and Emergency Medicine 2011 (Springer Verlag) and co-published as a series in Critical Care. Other articles in the series can be found online at http://ccforum.com/series/annual. Further information about the Annual Update in Intensive Care and Emergency Medicine is available from http://www.springer.com/series/8901 some form in all metazoan organisms. Th is response is activated by components of the wall of invading microorganisms, such as lipopolysaccharide (LPS) or peptidoglycan, following the binding of these pathogen-associated molecular patterns to pattern recognition receptors, such as the Toll-like receptors (TLRs) on tissue macrophages. Th e innate immune response is also activated by endogenous `danger' signals, such as mitochondrial components [15], providing an elegant explanation for why non-septic insults can also lead to organ injury and dysfunction. An infl ammatory cascade is then initiated, involving cytokine signaling activation of phago cytes that kill bacteria, as is activation of the (later) adaptive immune response.

Activation of the innate immune response
Hypercapnic acidosis has been demonstrated to inhibit multiple components of the host innate immune responses. Activation of the innate immune response initiates a conserved signaling cascade that culminates in the activation of transcription factors, such as nuclear factor kappa-B (NF-κB) [5]. Th ese transcription factors drive the expression of multiple genes that activate and regulate the pro-infl ammatory and repair processes. Increasing evidence suggests that hypercapnic acidosis directly inhibits the activation of NF-κB [16]. Intriguingly, this eff ect of hypercapnic acidosis may be a property of the CO 2 rather than its associated acidosis [17][18][19]. If confi rmed, this fi nding suggests the presence of a molecular CO 2 sensor in mammalian cells. Th is mechanism of action of hypercapnic acidosis has been demonstrated to underlie some of the anti-infl ammatory eff ects of hypercapnia [16], and to be a key mechanism by which hypercapnia -whether buff ered or not -reduces pulmonary epithelial wound healing [18].

Coordination of the innate immune response
Hypercapnic acidosis also interferes with coordination of the innate immune response by reducing cytokine signaling between immune eff ector cells. Hypercapnic acidosis reduces neutrophil [20] and macrophage [21] production of pro-infl ammatory cytokines such as tumor necrosis factor (TNF)-α, interleukin (IL)-1β, IL-8 and IL-6. Hypercapnic acidosis reduced endotoxin stimulated macrophage release of TNFα and IL-1β in vitro [21]. Peritoneal macrophages incubated under hypercapnic conditions demonstrated a prolonged reduction in endotoxinstimulated TNF-α and IL-1β release [22]. In contrast, a recent study reported rapid onset and rapid reversibility of IL-6 inhibition by hypercapnia in mature macrophage stimulated with LPS [19]. Th e mechanism underlying hypercapnic acidosis-mediated inhibition of cytokine and chemokine production appears to be mediated at least in part via inhibition of activation of NF-κB.

The cellular innate immune response
Neutrophils and macrophages are important eff ectors of the innate immune response in the setting of bacterial infection. Neutrophils rapidly migrate from the bloodstream to areas of infection, and rapidly phagocytose invading microorganisms. Tissue macrophages and their blood borne monocyte counterparts are activated by bacterial products such as endotoxin, and coordinate the activation of the adaptive immune response in the setting of infection by presenting foreign antigen to lymphocytes and secreting chemokines. Both monocytes and macrophages phagocytose and kill pathogens by similar mechanisms but at a slower rate than neutrophils.
Hypercapnic acidosis may impact on the cellular immune response via both direct and indirect mechanisms. Hypercapnic acidosis inhibits neutrophil expression of the chemokines, selectins and intercellular adhesion molecules [16,20], which facilitate neutrophil binding to the endothelium and migration out of the vascular system. Th e potential for hypercapnic acidosis to inhibit neutrophil chemotaxis and migration to the site of injury has been confi rmed in vivo, where hypercapnic acidosis inhibits pulmonary neutrophil infi ltration in response to endotoxin instillation [11]. Hypercapnic acidosis directly impairs neutrophil phagocytosis in vitro [23]. Th is inhibitory eff ect appears to be a function of the acidosis per se, with buff ering restoring neutrophil phagocytosis [24]. Hypercapnic acidosis also inhibits phagocytosis of opsonized polystyrene beads by human alveolar macrophages, although the levels of CO 2 utilized to demonstrate this eff ect were well beyond the range encountered clinically [19].
Neutrophils and macrophages kill ingested bacteria by producing free radicals such as superoxide, hydrogen peroxide, and hypochlorous acid, and releasing these into the phagosome. Th is is a pH-dependent process, with free radical production decreased at low pH [25]. Hypercapnic acidosis inhibits the generation of oxidants such as superoxide by unstimulated neutrophils and by neutrophils stimulated with opsonized Escherichia coli or with phorbol esters [20]. In contrast, hypocapnic alkalosis stimulates neutrophil oxidant generation [20]. Inhibition of the intracellular pH changes with acetazolamide attenuated these eff ects. More recently, hypercapnic acidosis has been demonstrated to reduce oxidative reactions in the endotoxin injured lung by a mechanism involving inhibition of myeloperoxidase-dependent oxida tion [26]. Th e potential for hypercapnic acidosis to reduce free radical formation, while benefi cial where host oxidative injury is a major component of the injury process, may be disadvantageous in sepsis, where free radicals are necessary to cause bacterial injury and death.
Neutrophil apoptosis following phagocytic activity generally occurs within 48 hours of release into the circulation. Conversely, neutrophil death via necrosis causes release of intracellular contents, including harmful enzymes, which can cause tissue destruction. Neutrophils appear to have an increased probability of dying by necrosis following intracellular acidifi cation during phagocytosis [27]. Hypercapnic acidosis may, therefore, increase the probability of neutrophil cell death occurring via necrosis rather than apoptosis.

Hypercapnia and the adaptive immune response
Th e adaptive immune system is activated by the innate response following activation of pattern recognition receptors that detect molecular signatures from microbial pathogens. Specifi c major histocompatibility complex molecules on T and B lymphocytes also bind microbial components. Th ese activation events lead to the generation of T and B lymphocyte-mediated immune responses over a period of several days.
Much of the focus to date regarding the eff ects of hypercapnic acidosis on immune response to injury and/ or infection has been on the innate immune response. Less is known about the eff ects of hypercapnic acidosis on adaptive or acquired immunity. However, important clues as to the potential for hypercapnic acidosis to modulate the adaptive response come from the cancer literature. Th e tumor microenvironment is characterized by poor vascularization, resulting in tissue hypoxia and acidosis. In a situation analogous to sepsis, acidosis in this setting may hamper the host immune response to tumor cells, potentially leading to increased tumor growth and spread. Th e cytotoxic activity of human lympho kine activated killer cells and natural killer cells is diminished at acidic pH [28]. Metabolic acidosis reduces lysis of various tumor cell lines by cytotoxic T-lymphocytes [29]. In contrast, the motility of IL-2-stimulated lymphocytes appears to be stimulated in the presence of an acidifi ed extracellular matrix and severe extracellular acidosis (pH 6.5) also appears to enhance the antigen presenting capacity of dendritic cells [30]. Th e net eff ect of these contrasting actions of metabolic acidosis on the adaptive immune response is unclear. However, the demonstration that hypercapnic acidosis enhanced systemic tumor spread in a murine model [31] raises clear concerns regarding the potential for hypercapnic acidosis to suppress cell-mediated immunity.

Hypercapnia and acidosis modulate bacterial proliferation
Carbon dioxide has broadly similar eff ects within the various families of microorganisms, but the sensitivity to CO 2 varies across the families, e.g., yeasts are quite resistant to the inhibitory eff ects of CO 2 , Gram-positive organisms are somewhat less resistant, and Gramnegative organisms are the most vulnerable [32]. Optimal anaerobic E. coli growth occurs at a CO 2 tension (PCO 2 ) of 0.05 atmospheres, which is similar to the PCO 2 in the gut. Th e aerobic growth rate of E. coli was not inhibited by a PCO 2 of 0.2 atmospheres but was inhibited at partial pressures above 0.6 atmospheres [33]. It is important to remember that these levels are extremely high in the context of human physiology.
Of concern, however, is the demonstration by Pugin et al. that more clinically relevant degrees of metabolic acidosis can directly enhance bacterial proliferation in vitro [34]. Cultured lung epithelial cells exposed to cyclic stretch similar to that seen with mechanical ventilation produced a lactic acidosis that markedly enhanced the growth of E. coli [34]. Th is was a direct eff ect of hydrogen ions, as direct acidifi cation of the culture medium to a pH of 7.2 with hydrochloric acid enhanced E. coli growth. In contrast, alkalinizing the pH of conditioned media from stretched lung cells abolished the enhancement of E. coli growth. A range of Gram-positive and Gramnegative bacteria (including E. coli, Proteus mirabilis, Serratia rubidaea, Klebsiella pneumoniae, Enterococcus faecalis, and Pseudomonas aeruginosa) isolated from patients with ventilator-associated pneumonia (VAP), grew better in acidifi ed media (Fig. 1). Interestingly, this eff ect was not seen with a methicillin resistant Staphylococcus aureus (MRSA) strain, which appeared to grow best at an alkaline pH [34].
Th e eff ects of hypercapnic acidosis on bacterial proliferation at levels encountered in the context of permissive hypercapnia are unclear. Th e net eff ect is likely to be a combination of the eff ects of the acidosis and of the hypercapnia. Nevertheless, the demonstration that clinically relevant levels of metabolic acidosis enhance bacterial growth is of concern.

Implications for hypercapnia in sepsis
Immunocompetence is essential to an eff ective host response to microbial infection. Hypercapnia and/or acidosis may modulate the interaction between host and bacterial pathogen via several mechanisms, resulting in a broad based suppression of the infl ammatory response.

Hypercapnia, acidosis and the host response
Th e initial host response to invading pathogens is dominated by neutrophil activation, migration to the infective site, and phagocytosis and killing of bacteria. Compartmentalized release of neutrophil proteolytic enzymes and myeloperoxidase-dependent oxygen radicals results in eff ective pathogen destruction. However, excessive release of these potent mediators into the extracellular space results in damage to host tissue and worsening ALI. Consistent with this is the fi nding that recovery of neutrophil count in neutropenic patients worsens the severity of ALI [35]. Hypercapnic acidosis may reduce the potential for damage to host tissue during the response to infection, by reducing lung neutrophil recruit ment [10], adherence [16], intracellular pH regulation [12], oxidant generation [8], and phagocytosis [23]. Th ese mechanisms are considered to underlie some of the protective eff ects of hypercapnic acidosis in nonsepsis induced ALI [7]. However, these eff ects of hypercapnic acidosis may be detrimental in sepsis, given the central role of neutrophil mediated phagocytosis of microbial pathogens and activation of the cytokine cascade to the host response to infection. In this context, defects in neutrophil function are associated with increased sepsis severity and worse outcome [36].

Early versus late bacterial infection
Th e eff ects of this hypercapnic acidosis-induced immune modulation may vary depending upon the stage of the infective process. Th e anti-infl ammatory properties of hypercapnic acidosis may reduce the intensity of the initial host response to infection, thus attenuating tissue damage (Fig. 2). However, the mechanisms whereby bacteria mediate tissue injury are complex and not limited to the contribution from an excessive host response. In late or prolonged pneumonia, in which tissue injury from direct bacterial spread and invasion makes a signifi cant contribution, hypercapnic acidosis might impair bactericidal host responses. In the absence of eff ective antibiotic therapy, this may lead to enhanced bacterial spread and replication leading to more severe tissue destruction and lung and systemic organ injury (Fig. 2).

Impact on repair following injury
Hypercapnic acidosis has been demonstrated to retard the repair process following lung cell and tissue injury. Hypercapnia slowed resealing of stretch-induced cell membrane injuries [37] and inhibited the repair of pulmo nary epithelial wounds [18] by a mechanism involv ing inhibition of the NF-κB pathway. Th ese fi ndings raise the potential that hypercapnic acidosis could lead to increased bacterial translocation through defects in the pulmonary epithelium, while also delaying the recovery process following a septic insult.
Recent studies in relevant preclinical models have signifi cantly advanced our understanding of the eff ects of hypercapnic acidosis in both pulmonary and systemic sepsis-induced ALI/ARDS. Th ese studies reveal the importance of severity, site, and stage of the infective process, the need for antibiotic therapy, and the utility of buff ering the hypercapnic acidosis in this setting.

Early lung infection
Th e eff ect of hypercapnic acidosis on pneumonia-induced ALI appears to depend on the stage and severity of the infection. In an acute severe bacterial pneumoniainduced lung injury, hypercapnic acidosis improved physio logical indices of injury [38]. Intriguingly, these protective eff ects were mediated by a mechanism indepen dent of neutrophil function. In contrast, hypercapnic acidosis did not alter the magnitude of lung injury in a less severe acute bacterial pneumonia [39]. Importantly these in vivo studies showed no increase in bacterial count in animals exposed to hypercapnic acidosis, a reassuring fi nding given concerns regarding retardation of the host bactericidal response and potential bacterial proliferation.
In the clinical setting, many critically ill patients will have established infection at the time of presentation. Th us animal models of established bacterial pneumonia, in which hypercapnic acidosis was introduced several hours following induction of infection with E. coli, more closely resemble the clinical setting. In an established pneumonia model, hypercapnic acidosis induced after the development of a signifi cant pneumonia-induced lung injury reduced physiological indices of lung injury [40]. Of importance, these protective eff ects of hyper capnic acidosis were enhanced in the presence of appropriate antibiotic therapy [40]. Again, reassuringly, lung bacterial loads were similar in the hypercapnic acidosis and normo capnia groups [40].

Prolonged lung infection
In an animal model of prolonged untreated pneumonia, sustained hypercapnic acidosis worsened histological and physiological indices of lung injury, including compliance, arterial oxygenation, alveolar wall swelling and neutrophil infi ltration [23]. Of particular concern to the clinical setting, hypercapnic acidosis was associated with a higher lung bacterial count. Th e mechanism underlying this eff ect appeared to be inhibition of neutrophil function, as evidenced by impaired phagocytotic ability in neutrophils isolated from hypercapnic rats [23]. Of importance to the clinical context, the use of appropriate antibiotic therapy abolished these deleterious eff ects of hypercapnia, reducing lung damage and lung bacterial load to levels comparable to those seen with normocapnia.
Th ese fi ndings have been confi rmed and considerably expanded in a recent study of hypercapnia in the fruit fl y [41]. Helenius et al., in a series of elegant in vivo studies, found that prolonged hypercapnia decreased expression of specifi c anti-microbial peptides in Drosophilia melano gaster [41]. Hypercapnia decreased bacterial resis tance in adult fl ies exposed to pathogens as evidenced by increased bacterial loads and increased mortality in fl ies inoculated with E. faecalis, A. tumefaciens, or S. aureus [41]. Th e previously demon strated suppressive eff ects of hypercapnic acidosis on the NF-κB pathway appeared to underlie the decreased resistance to infection [41]. Th ese fi ndings raise signifi cant concerns regarding the safety of hypercapnia in the setting of prolonged pneumonia, particularly in the absence of eff ective antibiotic therapy.

Hypercapnia in systemic sepsis
A growing body of evidence attests to a benefi cial role of hypercapnia in the setting of systemic sepsis. Improvements in hemodynamic parameters and lung injury have been demonstrated in evolving, established, and prolonged systemic sepsis in animal models. Th is is in contrast to the detrimental eff ects of hypercapnic acidosis seen in prolonged pulmonary sepsis, suggesting that the eff ects of hypercapnic acidosis depend not only on the stage of the infective process, but also on the site of the primary infection.

Early systemic sepsis
Hypercapnic acidosis reduces the severity of early septic shock and lung injury induced by systemic sepsis. In a rodent model of peritoneal sepsis induced by cecal ligation and puncture, hypercapnic acidosis slowed the development of hypotension, preserved central venous oxygen saturation, and attenuated the rise in serum lactate compared to control conditions, in the fi rst 3 hours post injury [42]. Th e severity of early lung injury was reduced as evidenced by a decrease in the alveolararterial oxygen gradient, and reduced lung permeability, compared to normocapnia. Alveolar neutrophil concentration was reduced by hypercapnic acidosis but IL-6 and TNF-α were unchanged [42]. Of importance, there were no diff erences in bacterial loads in the lung, blood, or peritoneum in the hypercapnia group.

Prolonged systemic sepsis
Using an ovine model of fecal peritonitis, Wang et al compared the eff ects of hypercapnic acidosis with those of dobutamine [43]. Over an 18-hour study period, hyper capnic acidosis resulted in improved hemodynamics of a magnitude comparable to that of dobutamine. Compared with normocapnia, both hypercapnic acidosis and dobutamine raised cardiac index and systemic oxygen delivery and reduced lactate levels. In addition, hyper capnic acidosis attenuated indices of lung injury, including lung edema, alveolar-arterial oxygen partial pressure diff erence and shunt fraction. Hypercapnic acidosis did not decrease survival time compared to normo capnia in this setting [43]. In a more prolonged systemic sepsis model, Costello et al. demonstrated that sustained hypercapnic acidosis reduced histological indices of lung injury compared with normocapnia in rodents following cecal ligation and puncture [42]. Reassuringly there was no evidence of an increased bacterial load in the lung, blood, or peritoneum of animals exposed to hypercapnia.

Intraperitoneal hypercapnia
Direct intra-abdominal administration of CO 2 -by means of a pneumoperitoneum -reduces the severity of abdominal sepsis-induced lung and systemic organ injury. Insuffl ation of CO 2 into the peritoneal cavity prior to laparotomy for endotoxin contamination increased animal survival [44]. Most recently, CO 2 pneumo peritoneum has been demonstrated to increase survival in mice with polymicrobial peritonitis induced by cecal ligation and puncture (Fig. 3) [31]. Th ese protective eff ects of intraperitoneal carbon dioxide insuffl ation appear be due to the immunomodulatory eff ects of hyper capnic acidosis, which include an IL-10 mediated downregulation of TNF-α [44]. Importantly, these eff ects appear to be mediated by the localized peritoneal acidosis, rather than by any systemic eff ect.

Buff ering hypercapnic acidosis in sepsis
Th e immunomodulatory eff ects of hypercapnic acidosis in sepsis may occur as a function of either hypercapnia or acidosis. As discussed, evidence suggests that hypercapnic acidosis exerts certain eff ects via its associated acidosis [24], while other eff ects appear be a function of the hypercapnia per se [17]. Buff ered hypercapnia, i.e., hypercapnia in the presence of normal pH, may be seen in ALI/ARDS patients as a renal compensatory measure, or as a result of the administration of bicarbonate, a common clinical practice in the ICU, and one that was Four hours later, animals underwent a laparotomy and induction of a CO 2 pneumoperitoneum (laparotomy + CO 2 ), laparotomy alone, or no laparotomy; survival was determined over the following 8 days. Modifi ed from [31] with permission.
permitted in the ARDSnet tidal volume study [1]. Aside from well established concerns regarding the use of sodium bicarbonate, there is evidence from animal models of lung and systemic sepsis that the anti-infl am matory and protective eff ects of hypercapnic acidosis are lost with buff ering. Th is has signifi cant implications in clinical scenarios where the buff ering of hypercapnia resulting from protective ventilator strategies is considered.

Pulmonary sepsis
In rodent models of acute pneumonia induced by intratracheal E. coli and by endotoxin, buff ered hypercapnia worsened lung injury [24]. Compared with normocapnic controls, buff ered hypercapnia increased multiple indices of lung injury including arterial oxygenation, lung compliance, pro-infl ammatory pulmonary cytokine concentrations, and measurements of structural lung damage. In these experiments, buff ered hypercapnia was established in the animals by exposure to hypercapnic conditions until renal buff ering to normal pH had occurred, thus avoiding the confounding eff ects of exogenous acid or alkali administration. Th is contrasts with the protective eff ects of hypercapnic acidosis in similar models [11,38]. Of note, buff ered hypercapnia did not reduce the phagocytic capacity of neutrophils, and did not increase lung bacterial load in these studies [24].

Systemic sepsis
In a study designed to assess the contribution of acidosis versus hypercapnia to the eff ects of hypercapnic acidosis on the lung and hemodynamic profi le in systemic sepsis, Higgins et al. exposed rats to environmental hypercapnia until renal buff ering had restored pH to the normal range [45]. Both buff ered hypercapnia and hypercapnic acidosis reduced the severity of early shock and attenuated the increase in serum lactate compared with normocapnia. In contrast, buff ered hypercapnia did not attenuate physiologic or histologic indices of lung injury in these studies [45]. Reassuringly, there was no evidence to suggest that buff ered hypercapnia worsened the degree of lung injury compared to normocapnia, and buff ered hypercapnia did not increase the bacterial load in the lungs or the bloodstream [45].

Hypercapnia and sepsis: where are we now?
Th e generally benefi cial eff ects of hypercapnic acidosis in the setting of experimental non-septic infl ammatory injury contrast with a more complex spectrum of eff ects in the setting of live bacterial infection. Hypercapnia and/ or acidosis exert diverse -and potentially confl ictingeff ects on the innate and adaptive immune responses. Overall, hypercapnic acidosis appears to suppress the immune response, although the net eff ect of its multiple actions appears to vary depending on the site of infection and also on whether the acidosis produced by the hypercapnia is buff ered or not. Hypercapnic acidosis appears to protect the lung from injury induced by evolving or more established lung and systemic bacterial sepsis in relevant pre-clinical models. In contrast, the eff ects of hypercapnic acidosis in prolonged untreated bacterial sepsis appear to diff er depending on the source of the infection, with the immunosuppressive eff ects of hypercapnic acidosis worsening lung injury in the setting of prolonged pneumonia. Th is deleterious eff ect is abrogated by eff ective antibiotic therapy. In contrast, hyper capnic acidosis reduced lung damage caused by pro longed systemic sepsis, again highlighting the potential importance of the source of infection. Finally, buff ering of the acidosis induced by hypercapnia does not confer signifi cant benefi t in the setting of lung or systemic sepsis, and may actually worsen lung injury in the setting of pneumonia.
Taken together, recent experimental fi ndings in relevant pre-clinical models provide some reassurance regarding the safety of hypercapnia in sepsis, particularly in early pneumonia, and in the setting of abdominal sepsis. However, in the setting of prolonged pneumonia, the immunosuppressive eff ects of hypercapnia remain a concern. While the use of ventilation strategies resulting in hypercapnia is clearly justifi ed in patient with ALI/ ARDS, care is warranted in the setting of sepsis. Th e fi nding that deleterious eff ects of hypercapnia in the setting of prolonged pneumonia are abrogated by appropriate antibiotic therapy is of importance.
Clinicians should carefully consider the use of early empiric antibiotic therapy in hypercapnic ALI/ARDS patients in whom sepsis is suspected or confi rmed. However, concerns persist, particularly where antibiotic cover may be suboptimal, or the bacteria are more resistant to antibiotic therapy. Th e fi ndings that hypercapnia may increase septic lung injury in the setting of prolonged pneumonia is also of relevance to other patient groups, such as patients with infective exacerbations of chronic obstructive airways disease.

Conclusion
Hypercapnia is an integral component of protective lung ventilatory strategies in patients with severe respiratory failure. Th e potential for hypercapnia to modulate the immune response, and the mechanisms underlying these eff ects are increasingly well understood. Th e fi ndings that hypercapnic acidosis is protective in systemic sepsis, and in the earlier phases of pneumonia-induced sepsis, provide reassurance regarding the safety of hypercapnia in the clinical setting. However, the potential for hypercapnic acidosis to worsen injury in the setting of prolonged lung sepsis must be recognized. Additional studies are needed to further elucidate the mechanisms underlying the eff ects of hypercapnia and acidosis in the setting of sepsis-induced lung injury.