2019/03/27 by Randal S. Blank, Ryan E. Lesh · 2 citations
Medicine · #Respiratory Support and Mechanisms #Airway Management and Intubation Techniques #Cardiac Arrest and Resuscitation
paper · doi:10.1213/ane.0000000000004021
Lung protective ventilation has emerged as an important modality to improve clinical outcomes in patients requiring mechanical ventilation, both in the critical care unit and in the surgical arena. Experimental and clinical studies have provided us with important, though incomplete, insight into the mechanisms by which protective ventilation strategies improve postoperative outcomes. Unfortunately, the expanding literature base in this important area has contributed to misconceptions and confusion regarding the definition and importance of what has been termed “low” tidal volume (VT) ventilation and whether or not low VT, per se, are lung protective. A careful review of the existing literature suggests that the term “low” VT has been misused and has been inconsistently applied in both clinical and experimental studies. Moreover, there is little to no evidence supporting the protective nature of “low” VT in the absence of strategies designed to maintain an open lung condition (alveolar recruitment maneuvers and positive end-expiratory pressure [PEEP]). A strong case can be made for establishing and maintaining an open lung condition in ventilated surgical patients and this is quite possibly of greater significance than low VT ventilation per se. While we are in no way advocating a return to the grossly supraphysiologic VT of past practice, we argue that the bulk of evidence supports the greater importance of establishing and maintaining an open lung condition in the ventilated surgical patients, a scenario that is more likely to confer protection and lead to improved clinical outcomes. The first problem a reader of the protective ventilation literature is likely to encounter is one of definition. While there is no accepted standard definition for “low” VT, it seems intuitive that use of the term “low” in this regard should refer to a VT lower than physiologic (or at least not much higher). However, it is used routinely to designate a VT that may simply be lower than “conventional” but still supraphysiologic. Imprecise use of the term has rendered it nearly meaningless. The current literature is replete with examples of studies using “low” VT that are grossly supraphysiologic. In fact, recent meta-analyses of “low” VT ventilation include volumes significantly exceeding the physiologic range by a large margin.1,2 The normal average VT in a healthy subject is approximately 6–6.5 mL/kg body mass.3 In patients with healthy lungs, it may be reasonable to normalize VT to ideal or predicted body weight, and this normalization is the current research, if not clinical, standard. While normalizing VT on the basis of an ideal body weight or predicted body weight is simple and conceptually appealing, the correlation between predicted body weight and total lung capacity is not particularly close.4 Not surprisingly, spirometric measures (eg, forced vital capacity) correlate more closely with total lung capacity.4 Thus, normalization to forced vital capacity may represent a more reasonable standard when such data are available. While inappropriate use of the term “low VT” unnecessarily complicates the literature and may be misleading to practitioners, the unjustified equation of “low” VT with protective ventilation is more concerning. “Low” VT and protective ventilation have emerged as synonyms, despite the fact that the body of evidence does not support the idea that “low,” or for that matter, even physiologic, VT are inherently protective. The advent of protective ventilation in perioperative practice derives from the Acute Respiratory Distress Syndrome Network trial5 in patients with critical lung injury and the more recent observations that similar strategies have the potential to improve clinical outcomes6 and physiologic endpoints7 in surgical patients. Thus, it appears that the perioperative physician may have the capacity to prevent iatrogenic injury and other pathologic states contributing to postoperative pulmonary complications by minimizing lung overdistension (volutrauma) and recruitment/derecruitment phenomena (atelectrauma). Reducing VT as a means of diminishing dynamic lung strain and mechanical power delivery emerged from studies in patients with acute respiratory distress syndrome in whom the reduced functional lung parenchyma (baby lung) and increased lung heterogeneity increase susceptibility to these injurious stimuli. This approach has subsequently carried over into the perioperative arena where, in contradistinction, most patients present with uninjured lungs and less lung heterogeneity. However, both the reduction in total lung capacity associated with chronic disease states and the rapid development of atelectasis and radiographic lung heterogeneity8 after the induction of general anesthesia and mechanical ventilation may predispose this population of patients to lung injury despite normal preoperative lung function. Although the “protective” effects of lung protective ventilation have often been attributed to a lower VT, it seems likely that the studies of lung protective ventilation which have produced improvements in surgical outcomes may have done so largely via factors other than VT. Airway driving pressure rather than VT has emerged as the 1 variable which predicts complications in mechanically ventilated surgical patients.9 Why might this be the case? Airway driving pressure (VT × elastance) reflects the ability of a ventilated lung to accommodate a delivered VT and may thus yield more information than VT per se about the physiologic impact of a given VT on the lung receiving it. Moreover, driving pressure is affected by other aspects of ventilator management, including those designed to prevent atelectasis and maintain an open lung condition. It has been demonstrated that an open lung condition achieved via physiologic titration of PEEP after an alveolar recruitment maneuver lowers driving pressure and improves meaningful postoperative outcomes.10 A review of the evidence emanating from recent and well-designed randomized controlled trials of lung protective ventilation reveals to the open-minded reader that we should not equate a ventilation strategy consisting solely of “low” or even normal VT with that of a lung protective regimen. It may also be that the most important components of an lung protective ventilation strategy in perioperative patients with uninjured lungs are those related to the open lung strategy rather than lower VT per se: First, with very few exceptions, studies of lung protective ventilation in surgical patients have compared ventilation “bundles”—approaches to ventilation consisting of multiple variables. Typically, experimental and control groups differ with regard to the manipulation of VT, alveolar recruitment maneuvers, and PEEP.6,7 The largest differences between lung protective ventilation and conventional groups, both arithmetically and physiologically, appear to be levels of PEEP.11 What we have learned from these studies is that a strategy utilizing a lower than conventional VT combined with an open lung strategy (alveolar recruitment maneuvers plus PEEP) is superior to that of a higher VT (typically grossly supraphysiologic) without an open lung strategy (without alveolar recruitment maneuvers or PEEP). While important, these studies are not designed to, and cannot be used to, define the optimal lung protective ventilation regimen. More importantly, they should not be used to conclude that the lower VT used in the experimental group are protective, because additional uncontrolled and potentially important variables (PEEP and alveolar recruitment maneuvers) differ between groups. Second, while combining an open lung strategy with lower VT improves postoperative physiologic endpoints7 and clinical outcomes,10 an isolated decrease in VT does not.12,13 Third, if lower VT were intrinsically protective, large well-designed cohort studies would likely be able to detect outcome improvements associated with the delivery of lower VT. The converse appears to be true. Delivery of lower VT is associated with increased mortality in patients receiving 2-lung ventilation14 and higher postoperative complication rates in patients receiving 1-lung ventilation.15 Taken together with results of clinical trials, we conclude that (1) there is a paucity of evidence with regard to the effect of truly low (physiologic or subphysiologic) VT and (2) that the use of lower VT is not intrinsically lung protective. How can we explain the lack of a direct relation or even an inverse relation between VT and the incidence of adverse outcomes? While the administration of a lower VT to a fixed amount of lung parenchyma should result in a lower degree of dynamic strain and reduced mechanical energy delivery (per tidal breath), any potential benefit gained by reducing overdistension and dynamic lung strain may be offset by other more deleterious effects if a lower VT is applied in the context of an incompletely recruited lung (no alveolar recruitment maneuver and/or inadequate levels of PEEP). This possibility is particularly important in the perioperative context, because atelectasis is both very common during anesthesia, is established shortly after induction of general anesthesia, and is thought to contribute to lung injury and other complications through effects on inflammation and enhancement of periatelectatic strain mechanisms.16 Historically, the observation that atelectasis occurred in the surgical patient undergoing general anesthesia provided the initial logical, though perhaps misguided, recommendation for the use of high (grossly supraphysiologic) VT to promote an open lung state.17 In addition to promoting atelectasis when used without sufficient alveolar recruitment maneuvers and PEEP, reducing VT may not actually reduce lung stress and strain if its use results in a severely atelectatic lung. If a significant amount of lung parenchyma is poorly aerated or atelectatic, even “low VT” may result in injurious conditions. This concept is supported by studies which demonstrate that (1) a mildly supraphysiologic VT regimen generally thought of as “protective” reduces both driving pressure18 and postoperative pulmonary complications10 only when combined with strategies designed to maintain an open lung condition (alveolar recruitment maneuvers and titrated individualized PEEP), (2) when compared to that of high VT, the use of VT well below the established strain threshold in an experimental animal model is also injurious when applied at a higher respiratory rate (ie, high mechanical power),19 and (3) experimental observations confirm that ventilation of periatelectatic regions results in profound injury, likely through a stress amplification process.16 Finally, cyclic recruitment and derecruitment phenomena may also contribute to injury via atelectrauma, a process enhanced in an incompletely recruited lung. Although the use of physiologic VT in the mechanical ventilation of the surgical patient is conceptually appealing, direct evidence for outcome improvements resulting from this approach is lacking. Meta-analyses that purport to evaluate “low” VT” should be read and evaluated with this caveat in mind. Clinical trials of protective ventilation in this patient population have primarily shown that lower (but generally still supraphysiologic) VT along with at least some elements of an open lung strategy (alveolar recruitment maneuvers, PEEP) are superior to regimens utilizing grossly supraphysiologic VT lacking all components of an open lung strategy. Evidence from a recent multicenter trial demonstrating that an open lung strategy alone can improve physiologic end points and clinical outcomes10 along with cohort studies demonstrating an inverse relation between the VT and complication rates after surgery14,15 lead us to hypothesize that maintaining an open lung state may be more important than the selection of an absolute VT. Whether or not this is true, the perioperative physician should be advised that the use of the term “low VT” in the literature is largely imaginative and that there is a paucity of evidence linking “low” or even physiologic VT to improved outcomes. DISCLOSURES Name: Randal S. Blank, MD, PhD. Contribution: This author helped write and edit the manuscript. Name: Ryan E. Lesh, MD. Contribution: This author helped write and edit the manuscript. This manuscript was helped by: Richard C. Prielipp, MD.