Mechanical Ventilation with Lower Tidal Volumes and Positive End-expiratory Pressure Prevents Pulmonary Inflammation in Patients without Preexisting Lung Injury

This study represents a part of a large study. Another part has already been published.18 

The study protocol was approved by the Medical Ethics Committee of the University of Amsterdam, and informed consent was obtained from all patients. Adult patients were eligible if scheduled to undergo a surgical procedure of 5 h or longer and all involved physicians (surgeon, anesthesiologist, pulmonologist) consented with the study procedures. Exclusion criteria included a history of any lung disease, use of immunosuppressive medication, recent infections, previous thromboembolic disease, recent ventilatory support, and participation in another clinical trial.

All patients received routine anesthesia according to the local protocol, including intravenous propofol (2–3 mg/kg, thereafter 6–12 mg · kg⁻¹· h⁻¹), fentanyl (2–3 μg/kg, thereafter as required), and rocuronium (as required), and epidural bupivacaine (0.125%)–fentanyl (2.5 μg/ml). The ventilatory protocol consisted of volume-controlled MV, at an inspired oxygen fraction of 0.40, inspiratory-to-expiratory ratio of 1:2, and a respiratory rate adjusted to achieve normocapnia. Randomization was performed by drawing a presealed envelope; patients were randomly assigned to MV with either V^Ts of 12 ml/kg ideal body weight (high V^T[HV^T]) and ZEEP or 6 ml/kg (low V^T[LV^T]) and 10 cm H²O PEEP. The ideal body weight of male patients was calculated as equal to 50 + 0.91 (centimeters of height − 152.4); that of female patients was calculated as 45.5 + 0.91 (centimeters of height − 152.4).3Anesthesiologists were allowed to change the ventilation protocol at any time point upon surgeon’s request or if there was any concern for the patient’s safety. If the surgical procedure exceeded 5 h, anesthesiologists were allowed to change the ventilation strategy after the second sampling (blood and bronchoalveolar lavage).

Bronchoscopy and bronchoalveolar lavage were performed twice on all patients: the first directly after induction of anesthesia and start of MV in the right middle lobe or lingula, and the second performed in the contralateral lung 5 h thereafter, either perioperatively or directly postoperatively. BALF was obtained and processed as previously described.19–21In short, bronchoalveolar lavage was performed by an experienced pulmonologist in a standardized fashion according to the guidelines of the American Thoracic Society, using a flexible fiberoptic video-bronchoscope. Seven successive 20-ml aliquots of prewarmed saline were instilled and aspirated immediately with low suction (recovery, 71 ± 18.4 ml). Arterial blood samples were drawn before both lavages, and hourly blood gas analyses were performed. Cell-free supernatants from BALF and blood were stored at −80°C until analysis. BALF cells were resuspended in ice-cold phosphate-buffered saline. The resuspended cells were partially used for absolute cell counts (using a Bürker-Turk hemocytometer; Emergo, Landsmeer, The Netherlands) and Giemsa-stained cytospin preparation for differential counting.

Myeloperoxidase was determined by enzyme-linked immunosorbent assay.22BALF levels of human neutrophil elastase were assessed with a sandwich-type enzyme-linked immunosorbent assay (Hycult Biotechnology, Uden, The Netherlands). The detection limit of the assay was 4.0 ng/ml. Tumor necrosis factor (TNF)-α, interleukin (IL)-1α, IL-6, IL-8, macrophage inflammatory protein 1α, and macrophage inflammatory protein 1β were measured by enzyme-linked immunosorbent assay (TNF-α, IL-6, IL-8, Sanquin, Amsterdam, The Netherlands; IL-1α, macrophage inflammatory protein 1α, macrophage inflammatory protein 1β, R&D Systems, Minneapolis, MN). Nucleosomes were measured by enzyme-linked immunosorbent assay as described previously with slight modifications.23One unit was arbitrarily set at the amount of nucleosomes released by 100 Jurkat cells. Detection limit of the assay is 0.1 U/ml. Nucleosomes are generated by internucleosomal cleavage of chromatin, during apoptotic cell death. We used the release of nucleosomes as measurement for apoptotic cell death.

Baseline characteristics of the randomized patient groups were compared with the Student t  test, Mann–Whitney U test, or chi-square test as appropriate. Linear mixed model analysis was used to detect differences between respiratory variables. This type of analysis takes the association between values for individual patients measured at each time point into account. This implies a maximum of six time points per patient. The fixed effects were hour of MV (0–5) and MV group (LV^T/PEEP or HV^T/ZEEP). Data obtained with linear mixed model analysis are presented as mean and 95% confidence interval (CI). All measured inflammatory mediators were not normally distributed. Differences within groups were analyzed with a Wilcoxon signed-rank test for paired samples comparing t = 5 versus  t = 0 h. The Mann–Whitney U test was used to compare the changes over time between the two randomization groups. We corrected for multiple testing using the Benjamini–Hochberg false discovery rate adjustment.24A P  value of less than 0.05 was considered statistically significant. All statistical analyses were performed with Statistical Package for the Social Sciences 12.0.2 (SPSS, Chicago, IL).

Seventy-four consecutive patients who were scheduled to undergo an elective surgical procedure of 5 h or more were screened (fig. 1). Twenty-eight patients were excluded, leaving 46 patients for randomization. Five patients were randomized but excluded from final analysis, because the initial surgical procedure was converted by the surgeon into another shorter operation (<3 h), and only one bronchoalveolar lavage was performed. One patient was randomized, but no lavages were performed upon the surgeon’s request after induction of anesthesia. In total, 40 patients completed the study protocol. There were no major differences between the two randomization groups with regard to baseline characteristics (table 1). Besides the mechanical ventilator settings (V^T, PEEP, and respiratory rate), there were significant differences in partial pressure of carbon dioxide and pH between the two MV strategies. Partial pressure of carbon dioxide was 5.60 (95% CI, 5.35–5.84) in the LV^T/PEEP group as compared with 4.86 (95% CI, 4.61–5.12) in the HV^T/ZEEP group (P < 0.001). Accordingly, pH was significantly lower in the LV^T/PEEP group (7.36; 95% CI, 7.34–7.38) as compared with the HV^T/ZEEP group (7.40; 95% CI, 7.39–7.42; P < 0.001). Maximum airway pressures were not different between the study groups during 5 h of MV (fig. 2). Perioperative hemodynamic parameters, including number of patients being transfused and the number of transfusions (erythrocytes and plasma) (table 2) were not different between the two ventilation groups.

Ninety-nine percent of the cells from the BALF were macrophages. MV did not alter cell content, and no differences in neutrophil influx were found between groups. Myeloperoxidase and elastase levels in BALF, however, were significantly higher after 5 h of MV with higher V^Ts and ZEEP as compared with baseline levels. Median myeloperoxidase levels increased from 2.80 [interquartile range, 0.0–7.80] to 8.80 [2.35–25.0] ng/ml (P = 0.009) and elastase levels increased from 7.10 [1.60–14.5] to 17.4 [5.70–21.2] ng/ml in the HV^T/ZEEP group (P = 0.013). No increase in myeloperoxidase and elastase levels was observed with the use of lower V^Ts and PEEP (fig. 3). Only for myeloperoxidase was there a statistically significant difference between the two ventilation strategies (P = 0.004).

Mechanical ventilation minimally influenced cytokine and chemokine levels in BALF (fig. 4). BALF levels of TNF-α and IL-8 were influenced by the way patients were ventilated. TNF-α increased in the LV^T/PEEP group (P = 0.028), whereas IL-8 increased in the HV^T/ZEEP group (P = 0.015) after 5 h of MV. Plasma levels of IL-6 and IL-8 did significantly increase during the surgical procedure, but this increase in cytokine generation was similar in both groups (fig. 5).

Mechanical ventilation with higher V^Ts and ZEEP caused an increase in BALF nucleosomes as compared with lower V^Ts and 10 cm H²O PEEP (P = 0.028; fig. 6). There was also a statistically significant difference between the two ventilation strategies (P = 0.043). In plasma, nucleosome levels were equally increased in both groups.

In the postoperative recovery, 28 patients had follow-up chest radiographs. There were no differences in postoperative arterial blood gas analysis (HV^T/ZEEP vs.  LV^T/PEEP): partial pressure of oxygen, 117 ± 42 versus  123 ± 53 mmHg; partial pressure of carbon dioxide, 43 ± 5 versus  42 ± 5 mmHg; and pH, 7.36 ± 0.053 versus  7.34 ± 0.051. There were no differences in the incidence of pulmonary complications (e.g. , acute lung injury, pneumonia) between the two study groups; in each study group, there was one patient requiring prolonged MV for respiratory failure after surgery. One patient ventilated with LV^T/PEEP died postoperatively of multiple organ failure after complicated hemihepatectomy. All other patients were discharged home.

Every measured mediator was tested three times (differences within groups comparing t = 5 vs.  t = 0 and changes between randomization groups). Because this approach serves to inflate type I error, we corrected for multiple testing. As a consequence, three P  values were no longer significant (P > 0.05). There was only a trend for higher levels of BALF nucleosomes in the HV^T/ZEEP group after 5 h of MV (P = 0.084). There was no statistical significant difference between the two MV strategies, regarding nucleosome levels in the BALF (P = 0.12). Also, the level of TNF-α in the LV^T/PEEP group was not significantly increased after 5 h of MV (P = 0.084).

In the current study, we demonstrate that short-term MV is associated with significant inflammatory changes in the pulmonary compartment and that a lung-protective strategy attenuates these changes. Based on our findings, it seems that MV is a proinflammatory stimulus in noninjured lungs.

Myeloperoxidase (and also elastase) in the BALF is higher after 5 h of MV with higher V^Ts and ZEEP as compared with baseline levels. No increase in myeloperoxidase and elastase was seen after 5 h of MV with lower V^Ts and PEEP. This implies activation of polymorphonuclear cells, which were recruited to the pulmonary compartment or already present there. Higher concentrations of IL-8 in the BALF of patients ventilated with higher V^Ts and ZEEP support the first idea. However, in the differential cell count, we do not see an increase in neutrophils, which can be explained by the fact that the concentration of IL-8 in the plasma is very high, and thus there is a chemotactic gradient not favoring migration of neutrophils into the lung. Another possibility is that the neutrophils remained in the subepithelium and did not migrate further into the alveoli. Neutrophil count in the BALF is a well-established method to observe neutrophil influx into the lung. However, neutrophils can accumulate in alveolar septa after MV.25A practical limitation was that we did not have reliable methods to obtain and isolate viable lung epithelial cells from our patients, and we could not investigate them in more detail. From a scientific point of view, it would also have been interesting to have obtained lung tissue for specific staining and identification of apoptotic cells. However, we have not performed these assays, because we thought that many patients would not consent to more invasive procedures perioperatively or postoperatively.

For all other measured inflammatory protein levels in BALF, there were no differences between the groups. It should be noted that a period of 5 h is probably too short to detect differences in certain protein levels due to modified transcriptional and translational processes. We hypothesize that most inflammatory mediators measured in BALF were made in alveolar macrophages and lung epithelial cells and released upon stimulation.26,27 

Furthermore, we have shown that there is a trend for higher BALF levels of nucleosomes after 5 h of MV with higher V^Tventilation and ZEEP as compared with baseline levels. During apoptotic cell death, nucleosomes are generated by internucleosomal cleavage of chromatin. The nucleosomes are then packed in apoptotic blebs along with other nuclear components. We used the release of nucleosomes as a measurement for apoptotic cell death. The rapid increase in BALF nucleosomes (i.e. , within hours after initiation of MV) most likely reflects apoptosis of pneumocytes. As far as we know, this is the first study showing an association between MV and alveolar apoptosis in humans. In vitro  experiments have shown that mechanical strain induces proapoptotic changes in human lung epithelial cells.27,28Furthermore, in vivo  animal experiments have shown that impairment of apoptosis pathways limited pulmonary inflammation and lung injury, and also protected against multiple organ failure and death.6,7Therefore, it has been proposed that intraalveolar apoptosis is a potentially harmful process that could be targeted in the treatment of (ventilator-associated) lung injury.29On the other hand, apoptosis may be a pivotal process involved in alveolar repair mechanisms. More research is needed before clinical application of antiapoptotic strategies.

Both surgical stimuli and general anesthesia are associated with increased plasma levels of proinflammatory markers.30,31In the current study, we extended these findings by showing higher concentrations of IL-6 and IL-8 after 5 h of MV in both ventilation strategies. In patients with acute lung injury, systemic cytokine concentrations increase after initiating MV with low PEEP and higher V^T.32We hypothesize, however, that in patients with noninjured lungs, there is no translocation of inflammatory mediators because much higher levels of inflammatory mediators in the systemic compartment were found as compared with the pulmonary compartment.

One limitation of our study is that our study protocol does not allow us to differentiate the effects of lower V^Ts from those by higher PEEP levels. We chose to combine lower V^Ts with PEEP and higher V^Ts with no PEEP, because these settings result in similar maximum airway pressures. Recent studies in open chest rabbits demonstrated that MV with V^Ts of 8–12 ml/kg and ZEEP may cause permanent mechanical alterations and histologic damage to peripheral airways and inflammation in noninjured lungs.25,33Surfactant inactivation or depletion seems to play a major role during ventilation with V^Ts of 10 ml/kg and ZEEP.34Another animal study demonstrated that atelectasis caused increased alveolar–capillary protein leakage and disruption of the vascular endothelium, possibly via  shear stress.35During general anesthesia, atelectasis is potentiated by anesthesia and muscle relaxants altering diaphragmatic position. Also, tidal airway closure can occur and cause peripheral airway injury. This may be a common but unrecognized complication in patients undergoing general anesthesia.36Cyclic opening and closing from ZEEP leads to greater increases in bronchoalveolar lavage cytokines than atelectasis.37Therefore, patients ventilated with ZEEP in our study could have gross atelectasis and peripheral airway injury, caused by tidal airway closure. Of note, no recruitment maneuver was performed in either MV strategies.

Our data are different from those from previous studies in which MV strategies were investigated in patients with noninjured lungs undergoing surgery. Indeed, Wrigge et al.  38demonstrated that MV with V^Tof 15 ml/kg ideal body weight and ZEEP for 1 h caused no consistent changes in plasma levels of measured cytokines. In a study of patients undergoing thoracic or abdominal surgery, no differences in inflammatory responses were found between two ventilation strategies similar to the ones used in our study after MV for 3 h.39These studies, however, looked at inflammatory mediators only after 1 and 3 h of MV, respectively. In other studies in which MV during or after cardiopulmonary bypass surgery was investigated, increased levels of proinflammatory mediators were reported, but not consistently.40–43Wrigge et al.  40showed that ventilation for 6 h with lower V^T(6 ml/kg ideal body weight) had no or only minor effect on systemic and pulmonary inflammatory responses in patients after cardiopulmonary bypass surgery as compared with higher V^T(12 ml/kg). Only TNF-α levels in the BALF were significantly higher in the high V^Tgroup than in the low V^Tgroup. Koner et al.  43investigated different ventilation strategies during cardiopulmonary bypass and did not find any changes in systemic cytokine levels, postoperative pulmonary function, or duration of hospitalization with either MV strategy. Unfortunately, no pulmonary cytokine levels were measured in that study. In contrast, two other studies did find a difference between different ventilation strategies in patients undergoing cardiopulmonary bypass.41,42 

Considering the minor differences in pulmonary inflammatory mediators caused by the two different ventilation strategies in patients during general anesthesia, it seems that the inflammatory response plays a minor role. From experimental studies, it is known that the inflammatory response occurs after 4–6 h or the damage being mainly mechanical without any relevant inflammatory response.8,11,44MV with moderate V^Tand ZEEP can cause mechanical injury with alveolar–bronchiolar uncoupling.25Therefore, in our patient group, there may be lung injury in the absence of a relevant inflammatory response.

The inflammatory changes observed in healthy lungs are mere physiologic adaptations to the artificial process of MV. However, we propose that lung injury is induced by a “multiple-hit” model, whereby predisposing conditions, such as injurious MV or major surgery, may result in (weak) pulmonary inflammation. Possible second hits, such as transfusion of blood products which may cause transfusion-related acute lung injury, prolonged (injurious) MV, aspiration, shock, sepsis, and pulmonary infection, may all cause additional lung injury, finally resulting in full-blown ARDS with high morbidity and mortality. There is indeed clinical evidence supporting this multiple-hit hypothesis. High V^Tventilation was independently associated with development of ARDS in patients who did not have ARDS at the onset of MV in the intensive care unit.13,14During MV of pneumonectomy patients, higher intraoperative V^Ts were identified as a risk factor of postoperative respiratory failure.15Furthermore, postoperative patients who were ventilated with a lower V^Tstrategy had a lower risk of pulmonary infection, and duration of intubation and duration of stay tended to be shorter.17Therefore, we would like to encourage the use of lower V^Ts and PEEP according to the principle of primum non nocere : Ventilator-associated lung injury can be limited. However, our results do not imply that these two different ventilation strategies can lead to different postoperative complications.

Of course, the aforementioned studies, including ours, have investigated patients who underwent major surgery. Inflammatory effects of the surgical procedure itself could not be excluded but are equal in both groups. However, investigating the effects of MV in healthy humans would lack any clinical significance. Similar results are probably not reproducible if the duration of MV was less than 5 h. Also, the type of surgery could have affected the variables investigated. We do realize that further studies are needed to elucidate the effects of prolonged MV.

In conclusion, MV for 5 h with lower V^Ts and PEEP may limit pulmonary proinflammatory changes in patients with noninjured lungs during major surgery. Even during a relatively brief period of MV, patients will most likely benefit from lower V^Ts and PEEP. The specific contribution of both lower V^Ts and PEEP on the protective effects of the lung should be further investigated.

The authors thank Lucien Aarden, Ph.D. (Academic Staff), and Shabnam Solati (Technical Staff) at Sanquin Research, Department of Immunopathology, Amsterdam, The Netherlands, for their technical expertise on the measurements of nucleosome concentrations.