Early Effect of Tidal Volume on Lung Injury Biomarkers in Surgical Patients with Healthy Lungs

Background:The early biological impact of short-term mechanical ventilation on healthy lungs is unknown. The authors aimed to characterize the immediate tidal volume (VT)-related changes on lung injury biomarkers in patients with healthy lungs and low risk of pulmonary complications. Methods:Twenty-eight healthy patients for knee replacement surgery were prospectively randomized to volume-controlled ventilation with VT 6 (VT6) or 10 (VT10) ml/kg predicted body weight. General anesthesia and other ventilatory parameters (positive end-expiratory pressure, 5 cm H2O, FIO2, 0.5, respiratory rate titrated for normocapnia) were managed similarly in the two groups. Exhaled breath condensate and blood samples were collected for nitrite, nitrate, tumor necrosis factor-&agr;, interleukins-1&bgr;, -6, -8, -10, -11, neutrophil elastase, and Clara Cell protein 16 measurements, at the onset of ventilation and 60 min later. Results:No significant differences in biomarkers were detected between the VT groups at any time. The coefficient of variation of exhaled breath condensate nitrite and nitrate decreased in the VT6 but increased in the VT10 group after 60-min ventilation. Sixty-minute ventilation significantly increased plasma neutrophil elastase levels in the VT6 (35.2 ± 30.4 vs. 56.4 ± 51.7 ng/ml, P = 0.008) and Clara Cell protein 16 levels in the VT10 group (16.4 ± 8.8 vs. 18.7 ± 9.5 ng/ml, P = 0.015). Exhaled breath condensate nitrite correlated with plateau pressure (r = 0.27, P = 0.042) and plasma neutrophil elastase (r = 0.44, P = 0.001). Plasma Clara Cell protein 16 correlated with compliance (r = 0.34, P = 0.014). Conclusions:No tidal volume-related changes were observed in the selected lung injury biomarkers of patients with healthy lungs after 60-min ventilation. Plasma neutrophil elastase and plasma Clara Cell protein 16 might indicate atelectrauma and lung distention, respectively.

(EBC), representing the metabolism of nitric oxide in the lung, have been measured frequently for assessing lung injury in patients breathing spontaneously or ventilated in the ICU 2,6,21,22 and after cardiothoracic surgery. [23][24][25] Nitrite concentration in the EBC showed a positive correlation with V T in ICU patients with or without ARDS, 6 and with the degree of lung overdistention in chronic obstructive pulmonary disease (COPD) patients. 26 Increasing nitrite and nitrosylated proteins in the bronchoalveolar lavage may have a prognostic value suggestive of lung injury progression in ARDS. 27 In humans, cytokines such as tumor necrosis factor-α, interleukin (IL)-1β, and IL-6 in plasma and bronchoalveolar lavage were increased in ARDS patients ventilated with greater V T and lower positive end-expiratory pressure (PEEP) compared with those receiving smaller V T and greater PEEP. 28 The levels of tumor necrosis factor-α and IL-8 in bronchoalveolar lavage also increased in ICU patients without ARDS ventilated with V T 10 to 12 ml/kg PBW for 12 h compared with those ventilated with V T 5 to 7 ml/kg PBW and similar PEEP. 3 The antiinflammatory cytokine IL-10 was affected by ventilatory settings and ventilation duration in brain-injured patients 2 and used for functional repair of human donor lungs. 29 IL-11 has a protective role against murine hyperoxia-induced DNA fragmentation and lung injury. 30,31 The plasma concentration of neutrophil elastase (NE) is an indicator of alveolar recruitment 32 and activation of neutrophils during the development of lung injury. 33 Finally, plasma Clara cell protein 16 (CC16), an antiinflammatory protein secreted by the Clara cells of the distal respiratory epithelium, is a marker of acute epithelial lung injury 34,35 and increases in ventilated preterm neonates 36 and after 5-h ventilation during abdominal surgery in adults. 37 We hypothesized that mechanical ventilation induces early V T -related changes in selected lung injury biomarkers in surgical patients with healthy lungs and low risk of pulmonary complications. Healthy patients undergoing knee replacement surgery under general anesthesia were randomized to volume-controlled ventilation with V T 6 (V T 6) or 10 (V T 10) ml/kg PBW. The levels of selected biomarkers in EBC and plasma and their time or tidal volume dependent changes were analyzed.

Materials and Methods
The experimental protocol was approved by the University of Colorado Multiple Institutional Review Board (Aurora, Colorado) before performing the study.

Experimental Protocol
After obtaining informed consent, 30 patients scheduled to receive elective orthopedic surgery for total knee replacement under general anesthesia were prospectively randomized to receive a tidal volume (V T ) of 6 or 10 ml/kg PBW during mechanical ventilation. Exclusion criteria for study patients included the following: American Society of Anesthesiologists class 4; age 70 yr or older; emergency procedure; status postpneumonectomy; diagnosed COPD, emphysema, asthma, pulmonary hypertension, sleep apnea, or any other respiratory disease; oxygen therapy during last month; tobacco use in the last 5 yr; severe obesity (body mass index ≥35 kg/m 2 ); immunosupression within 3 months before the procedure; diagnosed infection; or shock. Preoperatively, all patients received sciatic and femoral regional blocks for postoperative analgesia. The predetermined general anesthesia management included intravenous propofol for induction and maintenance (2 to 2.5 mg/kg and 0.05 to 0.2 mg kg −1 min −1 , respectively), to avoid potential differences of cytokine induction between different anesthetic drugs, 38,39 fentanyl (1 to 2 μg/kg initially, then 0.7 to 10 μg/kg as needed) and rocuronium (0.6 to 1.2 mg/kg for intubation, 0.1 to 0.2 mg/kg as needed when 25% recovery of T1 in train-of-four neuromuscular monitoring). Induction was performed in all patients while breathing FIO 2 1.0. Immediately after confirmation of adequate endotracheal tube placement, mechanical ventilation was started with a volume control ventilation mode with V T of either 6 (V T 6) or 10 (V T 10) ml/kg PBW as randomly selected. To set up the appropriate V T , the PBW was calculated based on the following formulas 7 : PBW-Males = 50 + 0.91 (centimeters of height − 152.4); PBW-Females = 45.5 + 0.91 (centimeters of height − 152.4). The respiratory rate was titrated for eucapnia (end-tidal carbon dioxide partial pressure [E T CO 2 ] 30 to 40 mmHg), and all patients received the same following ventilatory settings: inspiratory:expiratory (I:E) ratio, 1:2; inspiratory pause, 5%; and fresh gas flow, 2 l/min, FIO 2 , 0.5, PEEP, 5 cm H 2 O. Ventilatory parameters, except the respiratory rate, were unchanged during the study. Withdrawal criteria from the study were established as follows: (1) V T needed to be changed after randomization for any clinical or provider-related reason; (2) airway plateau pressure could not be managed to remain less than 30 cm H 2 O; or (3) for any other reason at the discretion of their anesthesia providers.
Immediately after starting mechanical ventilation (0-min time point), before the surgical incision, we initiated the collection of EBC. During the EBC collection, ventilatory and hemodynamic physiology parameters were recorded, and a sample of venous blood for analysis of biomarkers and an arterial blood gas sample were also obtained. At 60 min after the initiation of mechanical ventilation, we repeated the sample and data collection (60-min time point).

Demographics, Physiology, and Outcomes
Age, sex, American Society of Anesthesiologists classification, height, weight, and body mass index were recorded for all patients. Patients' PBW was calculated preoperatively as described earlier for V T calculation.
Physiology parameters were recorded simultaneously with the arterial blood gas sample collections, including hemodynamics (heart rate, respiratory rate, mean blood pressure), temperature, gas exchange (saturation of oxygen Fernandez-Bustamante et al.

PERIOPERATIVE MEDICINE
by pulse-oximetry or Sat p O 2 , and E T CO 2 as described before), and ventilation parameters (exhaled tidal volume or V T , respiratory rate, minute volume ventilation, peak and plateau airway pressures, and respiratory system compliance). Compliance was calculated as exhaled V T /(plateau pressure − PEEP). Measurement of the functional dead space was attempted for calculating the alveolar minute ventilation, but it was impossible because of imperfect capnography calibration in several patients (i.e., equal or nonphysiological E T CO 2 -PaCO 2 differences).
Any complications occurring intraoperatively were recorded (i.e., high airway pressure, hypoxemic event, need for withdrawal). Any postoperative respiratory complications were also recorded, as well as length of hospital stay and any in-hospital complications.

EBC and Blood Sample Collection
Exhaled breath condensate samples were collected with an Rtube™ breath condensate vial (Respiratory Research, Inc., Austin, TX) inserted in the expiratory limb of the ventilatory circuit for 20 min. This Rtube™ consists of a sterile polypropylene collection tube with a one-way valve trap that is kept cooled with an outer chilled aluminum sleeve (−80°C) during the collection period to condense the breath at the inner wall of the tube. No humidification was added to the ventilatory circuit, and the heat moisture exchange filter adjacent to the Y connector was removed. After 20 min, EBC samples were immediately placed in regular ice, volume measured, and aliquoted into vials prewashed with deionized water and frozen at −80°C until analysis.
Venous blood were obtained at each time point in EDTA vials and immediately transported in regular ice to the laboratory. Samples were centrifuged at 2,000 rpm for 10 min, plasma volume aliquoted, and frozen at −80°C until analysis.
Arterial blood was obtained from the radial artery at each time point in heparinized syringes and immediately processed for arterial blood gas analysis.

Inflammatory and Oxidative Stress Biomarkers
Nitric Oxide Metabolites. All samples were maintained frozen until analysis to minimize stability changes from freezing/ thawing cycles and ambient contamination. 40 The collection and storage manipulation were strict because of concerns of nitric oxide metabolites instability and contamination from ambient air. 41,42 Concentration of nitrite and nitrate were independently measured in duplicates by a dedicated HPLC system (ENO-20; Eicom, San Diego, CA). 43 Before measurement, plasma samples were mixed 50:50 v/v with methanol, vortexed, and centrifuged to remove fat and protein. The clarified supernatants were used for measurement. EBC samples were run without pretreatment. Nitrite and nitrate concentrations were calculated based on authentic standards. Nitrite and nitrate concentrations were summed to reflect the total nitric oxide (NOx) levels.
Cytokines. Concentrations of cytokine tumor necrosis factor-α, IL-1β, IL-6, IL-8, and IL-10 were measured in EBC and plasma samples with the Fluorokine ® mean arterial pressure Human High Sensitivity Cytokine Base Kit (cat# LHSC000; R&D Systems, Inc., Minneapolis, MN). IL-11 concentration was analyzed in EBC and plasma samples with the Quantikine ® Human IL-11 Immunoassay (cat#D1100; R&D Systems Inc.). Average concentrations from duplicates were used if variation between them was less than 20%, and results excluded from the analysis if variability greater than 20% after analysis repetition. Neutrophil Elastase. Plasma samples were measured in duplicates in 1:50 dilution samples using the Human PMN Elastase Platinum ELISA assay (Cat# ALX-850-265, Enzo ® ; Life Sciences, Inc., Farmingdale, NY). Clara Cell Protein 16. Clara cell protein concentrations were measured in duplicates in 1:25 diluted plasma samples with the Human Clara Cell Protein ELISA assay (Cat# RD191022200; BioVendor LLC, Candler, NC).
All analyses were performed in duplicated samples following the specific manufacturer's instructions.
Correlation analysis was performed between the V Trelated ventilatory parameters, such as plateau pressure or compliance and lung injury biomarkers. Other ventilatory parameters were not included in the correlation analysis if they were intrinsically dependent on our study design (i.e., V T , respiratory rate and administered minute ventilation) or were not available (i.e., transpulmonary pressure or other measurements of lung stretch/strain). Association between the different biomarkers was also examined with correlation analysis to find common factors influencing the observed biomarker changes.

Statistical Analysis
This pilot study was based on the linear correlation between EBC nitrite levels and V T in ventilated ICU patients with no lung injury (R 2 = 0.79) reported by Gessner et al. 6 Based on this finding, our pilot study was designed to detect a correlation coefficient of 0.79, with the null being 0.12 (two-tailed, alpha = 0.05, and power = 0.8). Using these parameters, the estimated sample size was 11, which we increased to 15 in each group to account for any lack of information on non-ICU patients and on the other biomarkers.
All recorded parameters (physiology, biomarkers, and outcomes) were graphically depicted and summarized using means and SDs. All biomarker measurements were examined for normality. If not normally distributed, they were logarithmically transformed for comparison between groups and measuring the association between biomarkers and physiology parameters by Pearson correlation. All continuous variables were statistically compared within the V T groups (same subjects, different time points) with a paired t test. The change scores of continuous variables from 0-to 60-min time points were compared between the two V T groups using an independent t test. Chi-square test was used for comparing categorical variables. Because of the high variability in nonnormally distributed biomarkers in different groups and time points, we used the coefficient of variation, calculated as a ratio of the SD and the mean. All statistical analyses were two-tailed and performed with SPSS version 21 (IBM, New York, NY). Significance was accepted at P value less than 0.05.

Demographics, Physiology, and Outcomes
Thirty patients who fulfilled the inclusion and exclusion criteria were enrolled and consented for the study. One patient from each V T group was removed from the study because of a nondisclosed steroid course within 10 days before surgery and previously undiagnosed sleep apnea symptoms. Only the remaining 28 patients (14 per group) were included in the final analyses. No significant differences were found between the two groups in terms of age, sex, comorbidities, American Society of Anesthesiologists classification, height, weight, body mass index, PBW, American Society of Anesthesiologists classification and comorbidities. No respiratory complications were detected, and there was no significant difference in hospital length of stay between the two groups (table 1).
Hemodynamic parameters, temperature, and oxygenation were comparable at the 0-min time point in both V T groups (table 2). E T CO 2 was within the targeted range (30 to 40 mmHg) in both groups. At the 0-min time point, as it was expected from the study design, the V T 10 group had lower respiratory rate and greater exhaled V T , minute volume ventilation, peak and plateau airway pressures and compliance. The arterial blood gas analysis showed a lower PaCO 2 and higher pH in the V T 10 compared with the V T 6 group.
From 0-min to 60-min time points, both groups experienced a small but statistically significant decrease in oxygenation in terms of Sat p O 2 (P = 0.045 for both). Only in the V T 10 group, there was a reduced but statistically significant decrease from 0 to 60 min in heart rate (P = 0.030), respiratory rate, and minute volume ventilation (P ≤ 0.001 for both), an increase in plateau pressure (P = 0.010) and a reduction in compliance (P = 0.023). The decrease of compliance in the V T 10 group was related to a small but significant increase in plateau pressure from 17.8 ± 3.1 to 18.6 ± 2.9 cm H 2 O, whereas tidal volumes and PEEP were unchanged. The observed decrease in PaO 2 in both groups only reached statistical significance in the V T 10 group (P = 0.013). Similarly, the decrease in Sat p O 2 observed in both groups only reached statistical significance in the V T 6 group (P = 0.040).
After 60 min of mechanical ventilation, the E T CO 2 remained within the target range in both groups. From 0 min to 60 min, there was a similar decrease in oxygenation in both groups. The decrease over time observed in minute ventilation, bicarbonate concentrations, and base excess Physiology, ventilator, and blood gas data from both tidal volume (V T ) groups and time points expressed as mean ± SD. Comparison within each group from different time points was performed with a paired t test. The change scores of each variable from 0 to 60 min time points were compared between the two V T groups using an independent t test. Significant P value was accepted as <0.05. CO 3 H − = bicarbonate concentration; E T CO 2 = end-tidal carbon dioxide; PaCO 2 = arterial carbon dioxide partial pressure; PaO 2 = arterial oxygen partial pressure; PBW = predicted body weight; Sat a O 2 = arterial saturation of oxygen; Sat p O 2 = peripheral saturation of oxygen by pulse oximetry; V T = tidal volume. Fernandez-Bustamante et al.

Early Effect of Tidal Volume on Healthy Lungs
was significantly greater in the V T 10 group than in the V T 6 group. There were no complications or withdrawals related to the study. No respiratory complications (respiratory failure for any reason requiring reintubation, noninvasive ventilation or oxygen therapy needed at hospital discharge) were observed in the early postoperative period and during the rest of the hospital stay. The hospital length of stay was not different between the two groups (2.9 ± 0.9 days in the V T 6 group vs. 2.6 ± 0.5 days in the V T 10 group, P = 0.445).

Effect of Mechanical Ventilation with Different Tidal Volumes on Measured Biomarkers
Nitric Oxide Metabolites. Nitrite and nitrate concentrations were detectable in EBC and plasma samples from all ventilated patients. At 0 min, the average EBC concentrations of nitrite and nitrate were similar in patients from both groups. There was a slight decrease from 0 to 60 min of EBC nitrite in the V T 6 group and an increase of EBC nitrate in the VT10 group, but no changes reached statistical significance in the EBC or plasma concentrations of nitric oxide metabolites between time points or between the V T groups ( fig. 1). The variability of EBC nitrite, nitrate, and total NOx (nitrite + nitrate) concentrations, measured with the coefficient of variation was comparable in both groups at 0 min ( fig. 2). At 60 min, a few individuals in the V T 10 group doubled their respective baseline levels of EBC nitrite and EBC nitrate. No individuals in the V T 6 group showed this magnitude of increase in EBC nitrite or nitrate or total NOx. Thus, the coefficient of variation of EBC nitrite and nitrate, and therefore of total NOx, decreased in the V T 6 group (by 47.1, 17.1, and 32.6%, respectively) but increased in the V T 10 patients (by 4.2, 75.1, and 61.9%, respectively) ( fig. 2). Cytokines. Only a few EBC samples (out of the total 56) showed detectable levels of tumor necrosis factor-α (n = 8), IL-1β (n = 21), and IL-10 (n = 19). Most of the measured concentrations were lower than the detection limit and not reliable for quantitative comparison. No significant differences were observed in plasma cytokine concentrations from 0-to 60-min time points, or between the change scores of the V T groups ( fig. 3A). Neutrophil Elastase. Neutrophil elastase was not detectable in the EBC samples. Plasma NE levels from all patients combined did not significantly increase from 0-to 60-min time points. The mean concentration of NE from the V T 6 patients significantly increased from 0 to 60 min (P = 0.008), A B C Fig. 1. Nitrite, nitrate, and total nitric oxide (total NOx = nitrite + nitrate) concentrations in exhaled breath condensate (EBC) and plasma. No statistical differences were found between the tidal volume (V T ) groups at any time point in either EBC or plasma concentrations of nitrite (A), nitrate (B), or total NOx (C). Note the similar scale of nitrite in EBC and plasma and the approximately 20-fold concentrations of nitrate and total NOx in plasma compared with EBC concentrations. Graphs represent mean and SD. Values from different time points within each group were compared with a paired t test. The change scores of continuous variables from 0 to 60 min time points were compared between the two V T groups using an independent t test. Significant P value was accepted as P < 0.05. Fernandez-Bustamante et al.

PERIOPERATIVE MEDICINE
but not in the V T 10 group. There were no significant differences between the change scores of the V T groups ( fig. 3B). Clara Cell Protein 16. Clara cell protein 16 was not detectable in the EBC samples. CC16 concentrations from all patients significantly increased from 0-to 60-min time points. It significantly increased in the V T 10 group (P = 0.015) but not in the V T 6 group (P = 0.081). There were no significant differences between the change scores of the V T groups ( fig. 3C).

Correlation between Biomarkers and Ventilatory Parameters
Nitric Oxide Metabolites. Measurements of EBC nitrite concentration, but not of EBC nitrate or total NOx, logarithmically correlated with the plateau pressure measurements in the pooled samples from all patients and time points (r = 0.27, P = 0.042) (fig. 4A).
Exhaled breath condensate nitrite concentrations showed a significant positive logarithmic association with plasma NE when all measurements were pooled for analysis (r = 0.44, P = 0.001) (fig. 4B). Levels of EBC nitrate or total NOx did not present any significant association with NE measurements.
Cytokines. Plasma levels of these parameters did not correlate with any ventilatory parameters. Neutrophil Elastase. Plasma concentrations of NE did not correlate with any ventilatory physiology parameters. Plasma concentrations of NE, as mentioned earlier, showed a positive logarithmic correlation with EBC nitrite levels. Clara Cell Protein 16. Plasma levels of CC16 showed a significant positive correlation with compliance (r = 0.34, P = 0.014) (fig. 4C).

Discussion
Our prospective randomized pilot study examined the early changes of selected lung injury biomarkers induced by 60-min mechanical ventilation with either 6 or 10 ml/kg PBW V T in surgical patients with healthy lungs and low risk of pulmonary complications.
Nitrite or nitrate concentrations in EBC did not significantly change after 60-min ventilation in our patients, and changes were not significantly different between the V T groups. Nitrite and nitrate levels in EBC, individually or combined, are often used as biomarkers of lung nitrosative stress. Increased nitric oxide metabolites reflect activation , and total nitric oxide (total NOx = nitrite + nitrate) (C) was comparable in both groups. At the 60-min time point, the variability of nitrite and nitrate concentrations, and therefore of total NOx, measured as the coefficient of variation (CV) (SD/mean) was greater in the tidal volume 10 (V T 10) group compared with that of the V T 6 group. The increased variability in total NOx in the V T 10 group was because of the increase in three patients that presented a raise predominantly in nitrate concentration, with one of these patients showing a simultaneous 2.5-fold increase in nitrite concentration (note the different scales for nitrite and nitrate concentrations). Fernandez-Bustamante et al.
Early Effect of Tidal Volume on Healthy Lungs of macrophages, neutrophils, or other various lung cell types 27,44,45 through the inducible or endothelial nitric oxide synthase. [46][47][48] Nitrite in EBC and nitric oxide in exhaled air are accepted by the American Thoracic Society as complementary but not equivalent measurements of nitrosative stress. 42,49,50 Nitrite and nitrate are relatively stable metabolites of nitric oxide in aqueous EBC samples. 51 In physiological conditions, nitric oxide can be oxidized to nitrite by lung epithelial cells, making nitrite measurable in the EBC of healthy individuals. 52 Nitrite can be reduced to nitric oxide, if decreased availability of oxygen, or further oxidized to nitrate. 53 Human cells are not able to reduce nitrate to nitrite; thus nitrate is either exhaled or diffuses into the circulation. 53 Multiple questions remain about the biological role of nitric oxide metabolites in pulmonary function. For example, EBC nitrite significantly increases in healthy recreational runners in a time-dependent manner without change in lipid peroxidation, 54 suggesting an increased nitric oxide production independent of significant lung injury. Nitrite level increases with lung distention in the EBC of COPD or ventilated patients with or without ARDS 6,26 as well as with high V T ventilation of isolated rabbit lungs. 55 EBC nitrate reflects the severity of asthma better than nitrite or exhaled nitric oxide. 56 Our observed nitrite and nitrate concentrations in EBC were comparable to those found in healthy humans, 23,54,57 suggesting healthy lung conditions during our study period. An insufficient time or intensity of the mechanical insult may explain the absence of significant changes in EBC nitrite or nitrate in our study patients.
The coefficients of variation of EBC nitrite and nitrate were noticeably different in the V T groups after 60 min of ventilation, decreasing in the V T 6 group while increasing in the V T 10 group. This difference reflected a few patients in the V T 10 group who had at least a two-fold increase of their initial EBC nitrite or nitrate levels. These patients did not differ from the rest of the group in physiology parameters or other biomarkers. Low V T ventilation attenuates the increases in the concentration of other biomarkers (cytokines, CC16, and procoagulants). [58][59][60][61][62] However, the interpretation of the increased variability of EBC nitrite or nitrate as an early sign of individual susceptibility to ventilator-induced lung injury deserves caution. The additional alveolar space ventilated with larger V T could contribute to the increased nitric oxide metabolites without any associated lung injury. Furthermore, we cannot exclude a role of ambient contamination in this variability because our study lacked controlling for inspiratory nitric oxide metabolites. 41 EBC nitrite levels positively correlated with ventilatory plateau pressures in our study. Gessner et al. 6 reported a positive correlation between EBC nitrite concentrations and lung distention parameters (V T , mean airway pressure, and PEEP) in ICU patients, ventilated for at least

PERIOPERATIVE MEDICINE
24 h for pneumonia or COPD exacerbation. The relationship between EBC nitrite and V T had a steeper correlation coefficient in ARDS patients than in patients with mild, or without, lung injury. The same authors described that EBC nitrite levels correlated logarithmically with residual volume, total lung capacity, and intrathoracic gas volume, but not with parameters of expiratory flow or EBC cytokine levels, in spontaneously breathing COPD patients. 26 Our findings suggest that plateau pressure, rather than volume parameters, might influence nitrite changes in EBC of healthy ventilated patients. The time and pressure dependence of nitrite generation in healthy lungs during mechanical ventilation needs further investigation. Exhaled breath condensate nitrite concentrations correlated with plasma NE levels in our study patients, which may suggest a shared process between pulmonary nitrite production and plasma NE. Concurrent increases in exhaled nitric oxide and NE activity have been observed in a murine oleic acid-induced lung injury model. 63 Release of NE by activated neutrophils may reflect ongoing alveolar neutrophil migration 32 during mechanical ventilation. Plasma NE has been evaluated as a marker and predictor of ARDS development. 33,64 In our study, plasma NE levels significantly increased after 60-min ventilation in the V T 6, but not in the V T 10 group. A greater atelectrauma in the V T 6 compared with the V T 10 group, despite the same PEEP levels, might contribute to this finding. 65,66 One patient from the V T 10 group had greater than 200 ng/ml NE values (level usually measured in ARDS) 33 at 0 and 60 min without any sign of lung injury, suggesting that the role of high NE levels in healthy individuals needs clarification.
Cytokines have been measured in EBC samples. 2,21,26 Our low detection rate is similar to findings in other studies in healthy patients ventilated during surgery 2 and reinforces the normal pulmonary status of our study patients. Plasma cytokine concentrations did not significantly change in relation to surgical trauma or greater V T. Possible contributing factors to this finding were the short study duration, lack of lung injury, or preoperative regional anesthesia. 67, 68 However, the inhibitory effect of regional anesthesia on systemic inflammatory response after orthopedic surgery is still unproven. 69,70 Plasma CC16 significantly increased after 60 min of mechanical ventilation in all patients and in the V T 10 group, but not in the V T 6 group. CC16 is a small protein secreted by epithelial Clara cells and an accepted biomarker of alveolar-capillary permeability. 34 Leakage of CC16 into the circulation is observed after lipopolysaccharide inhalation in healthy humans, 71 lung contusion, 72 and mechanical ventilation in preterm neonates 36 or animal models. 73 Plasma CC16 concentrations are increased in ventilated ARDS patients 58 and predict poor outcomes such as greater mortality and fewer ventilator-free days. 35 Increased plasma CC16 has been observed in surgical patients after 5 h of mechanical ventilation. 37 The observed increase in plasma CC16 levels in our patients could be related to passive leak because of positive pressure ventilation or upregulation of protein synthesis. Plasma CC16 concentrations from all patients correlated with lung compliance. The V T 10 patients received almost double insufflating lung volumes (≈620 ml in V T 10 compared with ≈350 ml in V T 6) with a small difference (≈3 cm H 2 O) in drive pressure, and therefore they had greater measured compliance. Thus, compliance was more affected by volume (exhaled V T ) than drive pressure (plateau pressure − PEEP) change. Greater passive leakage of CC16 might reflect greater lung end-inspiratory volume and possible increased lung strain. 74 Plasma CC16 levels, however, were not significantly different between the V T groups after 1 h in our study or after 5 h in Determann et al.'s study. 37 Our findings could therefore represent an immediate response of plasma CC16 to short-term positive pressure ventilation that might be influenced by lung distention with unknown clinical relevance.
The impact of the selected and often used V T 1,12 on respiratory physiology or clinical outcomes of our healthy patients was minimal after 60-min ventilation. The clinically insignificant worsening of oxygenation and compliance observed in our patients may reflect an under-recruitment phenomenon from ventilation within the lower portion of the pressure-volume curve. 65,66 This explanation, however, cannot be confirmed with our study design.
Our relatively low number of patients and the short duration of ventilation constitute the major limitations of our study. However, our study was designed primarily to characterize the early effect of different V T on biomarkers in patients with healthy lungs. Accordingly, we excluded patients with any respiratory or immune disease, or with a body mass index 35 kg/m 2 or greater because of the increased likelihood of obesity-related respiratory impairment. The effect of obesity or age on the lung production of nitrite and other biomarkers is unknown. We also excluded any surgery involving a nonsupine body position or restricted lung excursion to avoid confounders as positioning or external thoracic restraints. Finally, the understanding of the influence of minute ventilation, alveolar ventilation, or other ventilatory settings on the studied lung injury biomarkers during intraoperative ventilation is limited. Although we attempted to control for as many potential variables (PEEP, FIO 2 ) to focus on the V T , there are still several factors that may have affected our findings. For practical reasons, the effects of minute ventilation or respiratory rate need to be addressed separately.
In conclusion, we studied the early changes of selected lung injury biomarkers after 60-min ventilation with V T 6 or 10 ml/kg PBW in patients with healthy lungs. No significant changes in EBC nitrite or nitrate levels were observed. The variability of EBC nitrite and nitrate decreased in the V T 6 but increased in the V T 10 group after 60-min ventilation. We observed a significant increase in plasma levels of NE in the V T 6 group and CC16 in the V T 10 group, which may represent the effect of atelectrauma and increased alveolar distention, respectively. Future studies in patients with higher risk for postoperative pulmonary complications may confirm whether EBC nitric oxide metabolites, plasma NE, or CC16 constitute early diagnostic or predictive biomarkers of lung inflammation.