Mechanical ventilation may aggravate preexisting lung injury or even initiate pulmonary damage in patients without lung injury at the start of mechanical ventilation (MV; a phenomenon referred to as ventilator-associated lung injury) (1, 2). The mechanisms underlying ventilator-associated lung injury are beginning to be understood despite the difficulties in distinguishing the effects of MV from those of underlying diseases for which MV was started. Similar to acute lung injury/acute respiratory distress syndrome (ARDS) and pneumonia, ventilator-associated lung injury is associated with local production of proinflammatory mediators. Indeed, during ventilator-induced lung injury, the experimental counterpart of ventilator-associated lung injury, cytokines, are released in the pulmonary compartment (3-5). These ventilation-induced proinflammatory changes have been confirmed in clinical studies in patients with (6-9) and without acute lung injury/ARDS at onset of MV (10).
High-mobility group box (HMGB) 1 is a recently discovered mediator of proinflammatory responses that contributes to acute lung injury (11-13). Local and systemic levels of HMGB-1 were found increased in mice that were instilled with LPS via the airways (11). In addition, acute lung injury induced by either LPS (intranasally or intratracheally), cecal ligation puncture, or pancreatitis was diminished by administration of anti-HMGB-1 antibodies (11, 12, 14, 15). Furthermore, intratracheal administration of HMGB-1 has been found to induce acute lung injury (11, 12). Finally, in rabbits, bronchoalveolar lavage (BAL) fluid (BALF) levels of HMGB-1 were 5-fold higher after 4 h of MV with large tidal volumes as compared with lower tidal volumes. In this ventilator-associated lung injury model, the administration of anti-HMGB-1 antibodies attenuated lung injury (16).
The aim of this study was to investigate the local release of HMGB-1 in the pulmonary compartment during MV and ventilator-associated pneumonia (VAP). For this, we obtained BALF from patients on "short-term" (hours) MV and "long-term" (days) MV without evidence of preexisting lung injury and from patients who developed VAP during the course of MV. Healthy volunteers served as controls.
MATERIALS AND METHODS
We collected BALF samples of patients in one prospective investigation on unilateral VAP (17) and one randomized controlled trial in which patients were randomized to be mechanically ventilated with a lung-protective ventilation strategy or a conventional ventilation strategy (see below) (18). Patients were only eligible for participation in these two studies if they had no history of any lung disease, use of immunosuppressive medication, recent infections, previous thromboembolic disease, and recent ventilatory support. The separate protocols were reviewed and approved by the Medical Ethics Committee of the University of Amsterdam, Amsterdam, The Netherlands. Written informed consent from all patients/closest relatives or volunteers was obtained before inclusion.
Healthy volunteers were all nonsmoking individuals, whereas short-term MV patients were patients who were expected to be intubated and mechanically ventilated for at least 5 h because of elective surgery. All patients received anesthesia according to a local protocol, including intravenous propofol (induction with 2-3 mg kg−1, thereafter 6-12 mg kg−1 h−1), fentanyl (induction with 2-3 µg kg−1, thereafter as required), rocuronium (as required), and epidural bupivacaine (0.125%)-fentanyl (2.5 µg mL−1). 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 to achieve normocapnia. Patients were randomized to a lung-protective MV strategy using lower tidal volumes of 6 mL kg−1 ideal body weight and positive end-expiratory pressure (PEEP) of 10 cm H2O or a conventional strategy with higher tidal volumes of 12 mL kg−1 ideal body weight and no PEEP.
Long-term MV patients were patients ventilated according to a strict local protocol in which optimal PEEP was defined as the lowest level of PEEP with maximum Pao2; in addition, patients were mechanically ventilated with tidal volumes less than 8 mL kg−1 ideal body weight.
Unilateral VAP patients were patients admitted to the intensive care unit for ventilatory support. None of the patients had evidence of acute lung injury/ARDS at the start of MV. Patients were eligible for this study when they fulfilled the following criteria: fever or hypothermia (temperature <35.0°C or >37.7°C), leukocytosis or leucopenia (leukocyte count <3 or >10 × 109 L−1), worsening of arterial oxygen tension (Pao2-Fio2 ratio) and a chest radiograph suspect for a novel unilateral infiltrate. Furthermore, the diagnosis had to be supported by the results of microbiological culture of BALF or a clinical course consistent with VAP. For safety reasons, patients were excluded with Pao2 of 10 kPa or less and Fio2 greater than 0.60. Mechanical ventilation settings were similar to those in long-term ventilation patients.
In mechanically ventilated subjects, lung compliance was calculated by dividing the tidal volume with end-inspiratory plateau pressure-PEEP difference.
Bronchoalveolar lavage was performed by experienced pulmonologists in a standardized fashion according to the guidelines of the American Thoracic Society by use of a flexible fiber-optic video bronchoscope. Seven successive 20-mL aliquots of prewarmed 0.9% saline were instilled in a subsegment of the lung, and each was aspirated immediately with low suction. In general, 10 to 15 mL of the instilled 20 mL was recovered. There was no difference between the recovered volumes between the different groups.
For healthy volunteers and long-term MV patients, BALF was obtained from the right middle lobe. For short-term ventilation patients, bronchoscopy and BAL were performed twice on all patients: the first just after initiation of ventilation in either the right middle lobe or the lingula and the second performed in the contralateral lung 5 h thereafter either perioperatively or directly postoperatively. For VAP patients, BAL was initiated at the noninfected lung in a subsegment of the middle lobe or lingula, followed by a lavage of a subsegment of the infected lobe, as localized on a chest radiograph.
Healthy volunteers received local application of lidocaine (in throat, near vocal cords) before introduction of the bronchoscope before the lavage. Short-term MV patients were under general anesthesia with propofol and fentanyl-no additive medication was given before the lavage. Ventilator-associated pneumonia patients were sedated with midazolam and morphine, and no additive medication was initiated because of the lavage.
Bronchoalveolar lavage fluid was kept at 4°C until processing, which was performed within 30 min. The first aliquot was discarded; the second and third BALF recoveries from both sides were sent for microbial culture and virus isolation. The remaining BALF was centrifuged, and cell-free supernatants were stored at -80°C until HMGB-1 levels were determined.
Measurements of HMGB1 in BALF were performed by enzyme-linked immunosorbent assay with the use of monoclonal antibodies to HMGB-1 and with standardization to a curve of recombinant human HMGB-1 as described previously (19). Briefly, polystyrene microtiter plates were coated with monoclonal anticalf HMGB-1 antibody. Wells were incubated with bovine serum albumin, washed, and the calibrator and samples were added to the wells. After washing, another antihuman HMGB-1 peroxidase-conjugated monoclonal antibody (a synthetic peptide was used as immunogen) was added to each well. After another washing step, the luminescence reagent was added to the wells. The luminescence was measured using a microplate luminescence reader.
All data are presented as means and standard deviations or number (percentage) except for HMGB-1 levels: these are presented as medians (interquartile range [IQR]). Differences in HMGB-1 levels between groups were analyzed using the Mann-Whitney U test. Wilcoxon signed-rank test was used for paired BALF samples comparing t = 5 vs. t = 0 h and comparing infected versus noninfected lungs. Because no differences were found in pulmonary HMGB-1 levels between patients mechanically ventilated with lung-protective MV and with conventional MV, the data of these patients are taken together in the statistical analysis and presentation of data. Correlations were calculated using Spearman ρ test. A P value ≤ 0.05 was considered statistically significant.
Patient characteristics are given in Table 1. Baseline characteristics and perioperative parameters in short-term MV patients were described in detail previously (18). In short, 74 consecutive patients who were scheduled for an elective surgical procedure of at least 5 h were screened in the period December 2003 to March 2005; 48 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 BAL 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; 21 patients were assigned to lung-protective MV strategy and 19 patients to conventional strategy. There were no major differences between both randomization groups with regard to baseline characteristics. Baseline characteristics and reasons for initiation of MV of long-term ventilation patients have been described before (17-20). In summary, 10 intubated and mechanically ventilated patients without any signs of pulmonary infection or acute lung injury/ARDS and 4 patients with unilateral VAP were recruited. These four VAP patients recovered uneventfully with antibiotic therapy.
Healthy volunteers were neither intubated nor mechanically ventilated. Duration of MV for short-term MV patients was 306 ± 44 min (mean ± SD). Duration of MV in long-term MV and VAP patients at the moment of BALF collection was 7.6 ± 2.6 and 9.3 ± 3.6 days, respectively. Tidal volumes and PEEP levels in the three MV groups were in accordance with the study protocol (short-term MV) or the local MV guideline. Neither long-term MV nor VAP patients developed acute lung injury during the course of MV.
There were no reported adverse events related to the BAL in any study group.
HMGB-1 levels during MV
High-mobility group box 1 levels in healthy volunteers were 1.6 [IQR, 0.7 - 3.7] ng mL−1 (Fig. 1). Patients who were intubated and mechanically ventilated for a short period had increased HMGB-1 levels directly after intubation and start of MV, although differences were not statistically significant (4.8 [IQR, 0.7 - 15.1] ng mL−1; P = 0.147 vs. healthy volunteers). In addition, HMGB-1 levels did not change during 5 h of MV (1.7 [IQR, 0.8 - 8.5] ng mL−1; P = 0.493 vs. healthy volunteers; P = 0.250 vs. start of MV). There was no correlation between HMGB-1 levels and lung compliance (data not shown). In contrast, long-term MV was associated with elevated HMGB-1 levels (11.7 [IQR, 8.7 - 37.0] ng mL−1; P < 0.0001 vs. healthy volunteers).
HMGB-1 levels in BALF with VAP
Although increased HMGB-1 levels were found with VAP (17.4 [IQR, 8.5 - 23.2] ng mL−1; P = 0.014 vs. healthy volunteers), no differences existed between BALF levels from the infected and the contralateral, noninfected site (6.1 [IQR, 5.8 - 13.2] ng mL−1; P = 0.625 vs. infected site; Fig. 2). Moreover, HMGB-1 levels in VAP patients did not differ from those in patients on long-term MV without pulmonary infection (P = 0.839).
The main finding of our study is that MV is associated with increased HMGB-1 levels in BALF in patients without preexistent lung injury. Second, we report here that VAP does not result in different HMGB-1 levels compared with MV alone, whereas similar MV settings were used for a comparable period.
Our results extend, at least in part, previous experimental and clinical investigations in finding increased HMGB-1 levels in the bronchoalveolar space during lung injury. First, HMGB-1 in BALF is increased in murine models of LPS or hemorrhage-induced lung injury (11, 13) and in a mouse model of ventilator-induced lung injury (16). Second, HMGB-1 in pulmonary epithelial lining fluid is elevated in septic patients with acute lung injury/ARDS (11). One limitation of this last study is that it leaves unanswered to what extent the observed elevated HMGB-1 levels in patients with acute lung injury/ARDS are induced by MV. We report here for the first time that MV itself is associated with elevated HMGB-1 levels in patients without lung injury at onset of MV.
Little is known about the role of extracellular HMGB-1 in the alveolar space in healthy and diseased lungs. Two earlier published studies showed that intratracheally or intranasally administered HMGB-1 itself causes acute lung injury in mice, manifested by neutrophil accumulation, interstitial edema, and hemorrhages in the lungs (11, 12). Moreover, Ogawa et al. (16) showed that blocking endogenous HMGB-1 in a rabbit model of ventilator-induced lung injury improved oxygenation and limited microvascular permeability and neutrophil influx into the alveolar lumen, implicating HMGB-1 as a deteriorating factor during the development of ventilator-induced lung injury and as an appropriate therapeutic target in ventilator-induced lung injury.
Besides its potential harmful role, HMGB-1 has been suggested to have a physiological role in the host defense. Zetterstrom et al. (21) suggested that HMGB-1 may contribute to the local antibacterial barrier system in the upper airways. Furthermore, healthy subjects have 1,000-fold higher HMGB-1 levels in their pulmonary epithelial lining fluid than in their plasma (11). These HMGB-1 epithelial lining fluid concentrations of healthy subjects were comparable with those of patients during the acute phase of acute lung injury/ARDS. Remarkably, this phenomenon was not found in plasma: HMGB-1 plasma levels were not or hardly detectable in healthy subjects, whereas patients with acute lung injury/ARDS with sepsis had clearly elevated HMGB-1 plasma levels. Therefore, whereas HMGB-1 is not detectable in plasma from healthy subjects and is increased in plasma form septic patients, HMGB-1 is ("abundantly") present in the bronchoalveolar space from healthy subjects as much as in patients during their acute phase of acute lung injury/ARDS. A function of HMGB-1 in the bronchoalveolar space in healthy lungs can underlie this phenomenon. Further investigations are warranted to investigate possible functions of extracellular pulmonary HMGB-1 in healthy and diseased lungs.
Although HMGB-1 levels were clearly elevated in our study several days after the start of MV (long term), a shorter duration of MV (short term) did not lead to statistically significant higher HMGB-1 levels compared with levels from either healthy volunteers or directly after intubation. This finding is in line with a recent study on HMGB-1 involvement during ventilator-induced lung injury in rabbits (16) in which ventilation during 4 h with tidal volumes of 8 mL kg−1 in rabbits did not result in elevated BALF levels of HMGB-1 relative to control rabbits. Notwithstanding, BALF HMGB-1 levels were 5-fold higher after 4 h of MV with large tidal volumes (i.e., 30 mL kg−1) as compared with the lower tidal volumes (8 mL kg−1) in this rabbit study. In the clinical setting, tidal volumes are never as high as this; moreover, the tidal volumes in patients with acute lung injury/ARDS are advised not to exceed 6 mL kg−1 (22). Whether suffering from acute lung injury/ARDS or not, we extended these findings to all patients in the intensive care unit. Therefore, all individuals included in this analysis, except for those on conventional ventilation settings during short-term ventilation, were ventilated with "lower" tidal volumes. The results from our patient study with short-term MV together with the observation from the rabbit study suggest that, in contrast to ventilation for a longer period (i.e., days), HMGB-1 does not play a major role during ventilation for a short duration (i.e., hours) with tidal volumes of 6 to 12 mL kg−1.
In our study, VAP patients had higher HMGB-1 BALF levels compared with healthy volunteers, similar to the long-term MV patients who did not develop VAP. In addition, HMGB-1 in BALF from the infected site from VAP patients was not altered compared with the contralateral, noninfected site; neither was it compared with levels of long-term MV patients who did not develop VAP. However, it remains inconclusive from these data whether there is an additional or synergistic effect of VAP on MV-induced HMGB1 levels because we did not measure HMGB-1 in the VAP patients before they developed VAP. In addition, a possible role of HMGB-1 as an antibacterial factor (21) can implicate lower HMGB-1 levels before or during the development of VAP. We did not find decreased levels in patients with already-established VAP here. It remains to be elucidated whether lower BALF levels of HMGB-1 are associated with an increased risk of developing VAP.
Because MV itself can induce lung injury, including histopathological changes and the production of proinflammatory cytokines and chemokines such as TNF-α, IL-6, and IL-8, it is of interest to compare these inflammatory changes with local HMGB-1 levels. The design of our studies did not allow us to obtain lung tissue for histopathological evaluation. We measured, however, levels of various cytokines, chemokines, markers of neutrophil activation, and proteins involved with fibrin turnover. We reported on these values before (17, 18, 23). We found neither a correlation between HMGB-1 levels in BALF and levels of various cytokines, chemokines, and markers of neutrophil activation nor a correlation with proteins involved in fibrin turnover.
In conclusion, MV itself is associated with increased HMGB-1 levels in the alveolar space in patients without preexistent lung injury. Furthermore, bronchoalveolar HMGB-1 concentrations during VAP were elevated compared with healthy volunteers but not when compared with intubated and mechanically ventilated patients who did not develop VAP. Further studies are needed to investigate the role of HMGB1 during MV, VAP, and acute lung injury/ARDS and its potential as a therapeutic target herein.
1. Pinhu L, Whitehead T, Evants T, Griffiths M: Ventilator-associated lung injury. Lancet
2. dos Santos CC, Slutsky AS: Mechanotransduction, ventilator-induced lung injury and multiple organ dysfunction syndrome. Intensive Care Med
3. Wilson MR, Choudhury S, Goddard ME, O'Dea KP, Nicholson AG, Takata M: High tidal volume upregulates intrapulmonary cytokines in an in vivo mouse model of ventilator-induced lung injury. J Appl Physiol
4. Belperio JA, Keane MP, Burdick MD, Londhe V, Xue YY, Li K, Phillips RJ, Strieter RM: Critical role for CXCR2 and CXCR2 ligands during the pathogenesis of ventilator-induced lung injury. J Clin Invest
5. Imai Y, Kawano T, Iwamoto S, Nakagawa S, Takata M, Miyasaka K: Intratracheal anti-tumor necrosis factor-alpha antibody attenuates ventilator-induced lung injury in rabbits. J Appl Physiol
6. Ranieri VM, Suter PM, Tortorella C, De Tullio R, Dayer JM, Brienza A, Bruno F, Slutsky AS: Effect of mechanical ventilation on inflammatory mediators in patients with acute respiratory distress syndrome: a randomized controlled trial. JAMA
7. Ranieri VM, Giunta F, Suter PM, Slutsky AS: Mechanical ventilation as a mediator of multisystem organ failure in acute respiratory distress syndrome. JAMA
8. Lista G, Colnaghi M, Castoldi F, Condo V, Reali R, Compagnoni G, Mosca F: Impact of targeted-volume ventilation on lung inflammatory response in preterm infants with respiratory distress syndrome (RDS). Pediatr Pulmonol
9. Parsons PE, Eisner MD, Thompson BT, Matthay MA, Ancukiewicz M, Bernard GR, Wheeler AP, NHLBI Acute Respiratory Distress Syndrome Clinical Trials Network: Lower tidal volume ventilation and plasma cytokine markers of inflammation in patients with acute lung injury. Crit Care Med
10. Schultz MJ, Haitsma JJ, Slutsky AS, Gajic O: What tidal volumes should be used in patients without acute lung injury? Anesthesiology
11. Ueno H, Matsuda T, Hashimoto S, Amaya F, Kitamura Y, Tanaka M, Kobayashi A, Maruyama I, Yamada S, Hasegawa N, et al.: Contributions of high mobility group box protein in experimental and clinical acute lung injury. Am J Respir Cri. Care Med
12. Abraham E, Arcaroli J, Carmody A, Wang H, Tracey KJ: HMG-1 as a mediator of acute lung inflammation. J Immunol
13. Kim JY, Park JS, Strassheim D, Douglas I, Diaz del Valle F, Asehnoune K, Mitra S, Kwak SH, Yamada S, Maruyama I, et al.: HMGB1 contributes to the development of acute lung injury after hemorrhage. Am J Physiol Lung Cell Mol Physiol
14. Suda K, Kitagawa Y, Ozawa S, Saikawa Y, Ueda M, Ebina M, Yamada S, Hashimoto S, Fukata S, Abraham E, et al.: Anti-high-mobility group box chromosomal protein 1 antibodies improve survival of rats with sepsis. World J Surg
15. Sawa H, Ueda T, Takeyama Y, Yasuda T, Shinzeki M, Nakajima T, Kuroda Y: Blockade of high mobility group box-1 protein attenuates experimental severe acute pancreatitis. World J Gastroenterol
16. Ogawa EN, Ishizaka A, Tasaka S, Koh H, Ueno H, Amaya F, Ebina M, Yamada S, Funakoshi Y, Soejima J, et al.: Contribution of high mobility group box-1 to the development of ventilator-induced jung injury. Am J Respir Crit Care Med
17. Choi G, Schultz MJ, van Till JW, Bresser P, van der Zee JS, Boermeester MA, Levi M, van der Poll T: Disturbed alveolar fibrin turnover during pneumonia is restricted to the site of infection. Eur Respir J
18. Choi G, Wolthuis EK, Bresser P, Levi M, van der Poll T, Dzoljic M, Vroom MB, Schultz MJ: Mechanical ventilation with lower tidal volumes and positive end-expiratory pressure prevents alveolar coagulation in patients without lung injury. Anesthesiology
19. Yamada S, Inoue K, Yakabe K, Imaizumi H, Maruyama I: High mobility group protein 1 (HMGB1) quantified by ELISA with a monoclonal antibody that does not cross-react with HMGB2. Clin Chem
20. van Till JW, Levi M, Bresser P, Schultz MJ, Gouma DJ, Boermeester MA: Early procoagulant shift in the bronchoalveolar compartment of patients with secondary peritonitis. J Infect Dis
21. Zetterstrom CK, Bergman T, Rynnel-Dagoo B, Erlandsson Harris H, Soder O, Andersson U, Boman HG: High mobility group box chromosomal protein 1 (HMGB1) is an antibacterial factor produced by the human adenoid. Pediatr Res
22. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med
23. Wolthuis EK, Choi G, Dessing MC, Bresser P, Lutter R, Dzoljic M, van der Poll T, Vroom MB, Hollman M, Schultz MJ: Mechanical ventilation with lower tidal volumes and positive end-expiratory pressure prevents pulmonary inflammation in patients without preexisting lung injury. Anesthesiology