Changes in the concentrations of mediators in exhaled breath condensate during cardiac valve replacement under cardiopulmonary bypass and their relations with postoperative acute respiratory distress syndrome.

To investigate the changes in the concentrations of interleukin-6 (IL-6), interleukin-8 (IL-8), tumor necrosis factor-α (TNF-α), and 8-iso-prostaglandin F2α (8-isoPGF2α) in exhaled breath condensate (EBC) in patients undergoing cardiac valve replacement under cardiopulmonary bypass (CPB) and its relationship with postoperative acute respiratory distress syndrome (ARDS).A prospective, case-control study was performed on 55 patients undergoing elective cardiac valve replacement under cardiopulmonary bypass, between November 2017 and May 2019. According to the diagnosis of postoperative ARDS, the patients were divided into ARDS group and control group. We compared the clinical characteristics, outcomes, respiratory mechanics, oxygenation parameters, and mediators in the 2 groups immediately after tracheal intubation (T1), at the end of CPB (T2), and 2 hours (T3) and 6 hours (T4) after CPB, and calculated the receiver operating characteristic curve (ROC), sensitivity, and specificity of the corresponding mediators.ARDS occurred in 29 patients after CPB. The ARDS group exhibited prolonged postoperative ventilator support, time to extubation, length of stay in the ICU, and postoperative length of stay. The peak airway pressure (Ppeak) and plat airway pressure (Pplat) at T4 were higher in the ARDS group compared with the control group. The alveolar-arterial oxygen partial pressure [P(A-a)O2] and respiratory index (RI) were higher and PaO2/FiO2 was lower in the ARDS group at T2-4 compared with the control group. The levels of EBC and serum mediators in the ARDS group were significantly higher at T2-4 compared with those in the control group. All the mediators in EBC were correlated significantly with those in the serum in the ARDS group (r = 0.7314, 0.898, 0.8386, 0.792) and control group (r = 0.6093, 0.8524, r = 0.7828, r = 0.6575) (P < .001). Meanwhile, the area under the curve (AUC) of IL-8 in EBC was significantly lower at T2 and the AUC of IL-6 in EBC was significantly higher at T4 than in serum (P < .05). In addition, all of the mediators in EBC had a certain accuracy in diagnose of postoperative ARDS.EBC analysis could be used to predict the high incidence of ARDS after cardiac valve replacement under CPB.

Introduction

Cardiac operation alone with cardiopulmonary bypass (CPB) can cause complex inflammatory reactions and oxidative stress responses, which lead to a series of complications. Acute respiratory distress syndrome (ARDS) is one of the main complications after CPB and has high morbidity and mortality rates.[ Currently, diagnosis of ARDS is performed via arterial blood gas measurements, bronchial lavage fluid analyses, and chest X-ray, which are indirect and lag behind in terms of clinical development.[ Collection of exhaled breath condensate (EBC) is a noninvasive method that can be used to detect biomarkers from the airway. Interleukin [IL]-6, IL-8 are early markers of tissue damage sensitive material and tumor necrosis factor-α (TNF-α) is an inflammatory mediator of the earliest and most important process of inflammation.[ The 3 inflammatory factors are among the most promising biomarkers for predicting the morbidity and mortality of ARDS.[ 8-iso-prostaglandin F2α (8-isopgF2α) is a sensitive indicator that reflects oxidative stress and is accepted as a biomarker of ARDS.[ However, whether the mediators measured in EBC are associated with pulmonary inflammation or oxidative stress remains unclear. Thus, a new method for the early prediction of ARDS induced by CPB is urgently needed.

The purpose of this study was to investigate the relationship between postoperative ARDS and changes in the concentrations of the mediators of inflammatory factors IL-6, IL-8, and TNF-α and oxidative stress 8-isopgF2α in EBC and serum samples of patients undergoing cardiac valve replacement. We hypothesized that increased levels of inflammatory and oxidative mediators in EBC equate to an increased risk of postoperative ARDS after cardiac valve replacement under CPB.

Methods

Patients

The patients undergoing elective cardiac valve replacement under cardiopulmonary bypass from November 2017 to May 2019 were selected as the study subjects. This study was approved by the Ethics Committee of the First People's Hospital of Nantong (the Second Affiliated Hospital of Nantong University) and the Chinese Clinical Trial Register (reference ChiCTR1800014299). Informed consent was obtained from each patient. Fifty five patients (32 males and 23 females) with statuses of II and III under the American Society of Anesthesiologists (ASA) physical status classification system and scheduled for elective cardiac valve replacement under CPB were enrolled in this study. The patients were prospectively and observationally studied in a single center. Patient age ranged from 34 to 80 years. The exclusion criteria included previous ingestion of anti-inflammatory or antioxidant drugs, systemic active infections, liver and kidney dysfunction, chronic obstructive pulmonary disease (COPD), coronary heart disease (CHD), history of neurologic disturbance, CPB failure (and thus, repetition of CPB), and reoperation or repetition of CPB after operation.

Anesthesia and intraoperative management

Patients received routine monitoring in the operating room. Anesthesia was induced with intravenously administered midazolam (0.05 mg/kg), etomidate (0.3 mg/kg), sufentanil (1 ug/kg), and vecuronium bromide (0.15 mg/kg). Endotracheal intubation was performed after the induction of anesthesia. Anesthesia was subsequently maintained with propofol (4–5 mg kg−1·h−1), remifentanil (6–15 μg·kg−1·h−1), and vecuronium bromide (0.12–0.15 mg kg−1·h−1). The ventilator settings consisted of a tidal volume (VT) of 8 to 10 ml/kg, breathing frequency of 12 to 14 breaths/minute, and an inhale:exhale ratio of 1:2. End-tidal CO2 was maintained between 30 and 35 mm Hg, and the bispectral (BIS) index was maintained between 40 and 60. Positive end-expiratory pressure (PEEP) was set at 0 cmH2O. FiO2 0.6 was used during mechanical ventilation and not manipulated until PaO2/FiO2 was below 300 mm Hg or SpO2 was lower than 98%. All patients were transferred to the postsurgical intensive care unit (ICU) after surgery without extubation.

Collection and analysis of EBC and Blood Samples

The EBC collection device was constructed using a portable glass tube connected to the expiratory port of a ventilator. The glass tube was cooled with a mixture of ice and water to generate temperatures of 0°C. The EBC samples were collected at time points T1 (immediately after tracheal intubation), T2 (end of CPB), T3 (2 hours after CPB), and T4 (6 hours after CPB). Each collection lasted for 30 minutes, and 1 to 2 ml of EBC was obtained. The samples were stored immediately at −80°C at the end of each collection. Radial artery blood samples (5 ml) were collected from all patients at T1, T2, T3, and T4 and divided into 2 aliquots. One aliquot (1 ml) was promptly used for blood gas analysis, and the other (4 ml) was centrifuged at 1500 rpm for 10 minutes. The serum was separated and stored at −80°C until analysis.

The levels of IL-6, IL-8, and TNF-α in the EBC and serum samples were measured with an enzyme-linked immunosorbent assay kit (R&D Systems, USA), whereas those of 8-iso-PGF2α were measured with a competitive enzyme immunoassay kit (Abcam, UK). Alveolar-arterial oxygen partial pressure [P(A-a)O2] and the respiratory index (RI) were calculated in accordance with the following formulas: P(A-a)O2 = [FiO2 × (760 − 47) − PaCO2 × 1.25] − PaO2 and RI = P(A-a)O2 / PaO2.

Definitions of outcome variables

The patients were divided into ARDS and control groups in accordance with the diagnosis of postoperative ARDS. All patients received the same treatment before ARDS was diagnosed, and the same empirical treatments after cardiac surgery were symptomatic treatment, oxygen inhalation, component blood transfusion, use of vasoactive drugs, appropriate dieresis, and so on. Standardized treatments were provided to those who developed ARDS and the treatments were positive control of primary disease and use of lung protective ventilation (Vt 6–8 ml/kg predicted body weight, low plateau pressures and positive PEEP) as soon as possible on the basis of the original treatment.

The patients were considered to have ARDS when they met the 2012 Berlin definition of ARDS.[ The ARDS patients were classified into mild [200 mm Hg < PaO2/FiO2 ≤ 300 mm Hg with PEEP or continuous positive airway pressure (CPAP) ≥5 cmH2O], moderate (100 mm Hg < PaO2/FiO2 ≤ 200 mm Hg with PEEP ≥5 cmH2O), and severe (PaO2/FiO2 ≤ 100 mm Hg with PEEP ≥5 cmH2O) in accordance with the oxygenation index.

Pulmonary infections that developed within 3 days after replacement were recorded. Secondary outcome measures, including the time of postoperative ventilator support, time to extubation, length of stay in the ICU, post-ICU ventilator use and postoperative length of stay, were recorded.

Statistical analysis

Statistical analysis was performed with GraphPad Prism version 7.00 (GraphPad Software, USA). The data following normal distribution were presented as mean ± SD, the data with non-normal distribution were presented as medians and interquartile ranges (IQRs), and the counting data were expressed as number or percentage. Independent sample t tests were used for the comparison between 2 non-normally distributed groups, and the Mann–Whitney U test was used to compare the data of the 2 groups for variables that were not normally distributed. Repeated measures of analysis of variance and Tukeys multiple comparison test were adopted to compare repeated measures. The Chi-Squared test or Fisher exact probability method was utilized to compare differences in rates. The relationships between the inflammatory or oxidative mediators in the EBC and serum samples were analyzed using Pearson correlation coefficients. The receiver operating characteristic curve (ROC) and the AUC was used to compare the diagnostic performance of different parameters. MedCalc 19.0.7 software (MedCalc, Ostend, Belgium) was used to compare different diagnostic tests. Results were considered statistically significant when P < .05.

Results

Clinical characteristics and outcomes of patients

Fifty five patients participated in this study, and 29 of them suffered from ARDS after cardiac valve replacement under CPB. Among the 29 patients, 22 suffered mild ARDS, 6 had moderate ARDS, and 1 had with severe ARDS. The 2 groups did not differ in terms of gender, age, height, weight, ASA statuses, New York Heart Association (NYHA) statuses, surgery type ratio, operation duration, CPB duration, block duration, bleeding amount, transfusion volume, pulmonary infections, and mortality (P > .05). The time of postoperative ventilator support, time to extubation, length of stay in the ICU, post-ICU ventilator use and postoperative length of stay in the ARDS group were extended compared with those in the control group (P < .05). One patient selected in the ARDS group was died at 4 d after CPB, so the postoperative length of stay of that patient was not calculated. The basic clinical characteristics and outcomes of all patients in this study are shown in Table 1.

Table 1

Clinical characteristics and outcomes of patients in the 2 groups.

Mean arterial pressure and heart rate of patients

The variables of the mean arterial pressure and heart rate of patients in the 2 groups at different time points did not significantly vary between the 2 groups (P > .05) (Fig. 1 A and B).

Figure 1

Comparison of mean arterial pressure and heart rate. A, Comparison of mean arterial pressure between the ARDS group and control group. B, Comparison of heart rate between the ARDS group and control group. ARDS = acute respiratory distress syndrome, HR = heart rate, MAP = mean arterial pressure.

Hematologic and biochemistry indicators

The hematologic and biochemistry indicators include white blood cell (WBC), hemoglobin (Hb), platelet (PLT), blood urea nitrogen (BUN), creatinine (Cr), natrium (Na), kalium (K), and albumin (ALB). There were no significant differences in all of them between ARDS and control group before the operation (P > .05). There were no significant differences in Hb and K between ARDS and control group at 24 hours after operation (P > .05). Compared with the control group, the WBC, BUN, Cr, and Na of ARDS group were significantly higher while the Hb and ALB of ARDS group were significantly lower at 24 hours after operation (P < .05) (Table 2).

Table 2

Comparison of hematologic and biochemistry indicators in the 2 groups.

Respiratory mechanics and oxygenation parameters

The respiratory mechanics and oxygenation parameters include tidal volume (VT), peak airway pressure (Ppeak), plat airway pressure (Pplat), PaO2/FiO2, P(A-a)O2, and RI. There was no significant difference in VT between ARDS and control group (P > .05). Compared with the control group, the Ppeak and Pplat of ARDS group were significantly higher at T4 (P < .001), the PaO2/FiO2 of ARDS group was significantly lower at T2–4 (P < .001), the P(A-a)O2 of ARDS group was significantly higher at T2–3 (P < .001) and T4 (P = .003), and the RI of ARDS group was significantly higher at T2 (P = .012), T3–4 (P < .001)) (Fig. 2A-E).

Figure 2

Comparison of mean respiratory mechanics and oxygenation parameters (VT, Ppeak, Pplat, P(A-a)O2, PaO2/FiO2 and RI). A, Comparison of VT between the ARDS group and control group. B, Comparison of Ppeak between the ARDS group and control group. C, Comparison of Pplat between the ARDS group and control group. D, Comparison of P(A-a)O2 between the ARDS group and control group. E, Comparison of PaO2/FiO2 between the ARDS group and control group. F, Comparison of RI between the ARDS group and control group. ARDS = acute respiratory distress syndrome, Ppeak = peak airway pressure, Pplat = plat airway pressure, P(A-a)O2 = alveolar-arterial oxygen partial pressure, RI = respiratory index, VT = tidal volume.

Concentration of mediators in serum and EBC samples

The time course of the levels of mediators, including IL-6, IL-8, TNF-α, and 8-isopgF2α, in the serum and EBC samples (Fig. 3A-H) showed that the mediators in serum and EBC were significantly higher in the ARDS group at T2–4 compared with the control group (P < .05), and the levels of all mediators in EBC and serum were increased at T2–3 relative to T1, and decreased at T4 relative to T3 (P < .05).

Figure 3

Comparison of serum and EBC levels of IL-6, IL-8, TNF-α and 8-isopgF2α. A, Comparison of serum level of IL-6 between the ARDS group and control group. B, Comparison of EBC level of IL-6 between the ARDS group and control group. C, Comparison of serum level of IL-8 between the ARDS group and control group. D, Comparison of EBC level of IL-8 between the ARDS group and control group. E, Comparison of serum level of TNF-α between the ARDS group and control group. F, Comparison of EBC level of TNF-α between the ARDS group and control group. G, Comparison of serum level of 8-isopgF2α between the ARDS group and control group. H, Comparison of EBC level of 8-isopgF2α between the ARDS group and control group. 8-isoPGF2α = 8-iso-prostaglandin F2α, ARDS = acute respiratory distress syndrome, EBC = exhaled breath condensate, IL-6 = interleukin-6, IL-8 = interleukin-8, TNF-α = tumor necrosis factor-α.

Correlation analysis of mediators in EBC and serum samples

Spearman correlation analysis found that the levels of IL-6, IL-8, TNF-α, and 8-isopgF2α in EBC were correlated significantly with those in the serum in the ARDS group (r = 0.7314, P < .0001; r = 0.898, P < .0001; r = 0.8386, P < .0001; r = 0.792, P < .0001) and control group (r = 0.6093, P < .0001; r = 0.8524, P < .0001; r = 0.7828, P < .0001; r = 0.6575, P < .0001) (Fig. 4A-H).

Figure 4

Correlations of EBC levels of IL-6, IL-8, TNF-α and 8-isopgF2α with levels in serum in the ARDS group and control group. A, Correlation of EBC levels of IL-6 with levels in serum in the ARDS group. B, Correlation of EBC levels of IL-6 with levels in serum in the control group. C, Correlation of EBC levels of IL-8 with levels in serum in the ARDS group. D, Correlation of EBC levels of IL-8 with levels in serum in the control group. E, Correlation of EBC levels of TNF-α with levels in serum in the ARDS group. F, Correlation of EBC levels of TNF-α with levels in serum in the control group. G, Correlation of EBC levels of 8-isopgF2α with levels in serum in the ARDS group. H, Correlation of EBC levels of 8-isopgF2α with levels in serum in the control group. 8-isoPGF2α = 8-iso-prostaglandin F2α, ARDS = acute respiratory distress syndrome, EBC = exhaled breath condensate, IL-6 = interleukin-6, IL-8 = interleukin-8, TNF-α = tumor necrosis factor-α.

The association between levels of mediators and postoperative acute respiratory distress syndrome

The ROC curves of mediators including IL-6, IL-8, TNF-α, and 8-isopgF2α at different time point (Fig. 5A-F, Table 3) showed that, the AUC of the IL-8 in EBC was significantly lower at T2 than in serum (P = .026), the AUC of the IL-6 in EBC was significantly higher at T4 than in serum (P = .001). There was no significant difference in AUC between mediators in EBC and serum samples at the other time points (P > .05). However, all of the mediators in EBC had a certain accuracy in diagnose of postoperative ARDS (P < .05).

Figure 5

The ROC curves of IL-6, IL-8, TNF-α and 8-isopgF2α in EBC and serum at T2–4. A, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in serum at T2. B, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in EBC at T2. C, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in serum at T3. D, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in EBC at T3. E, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in serum at T4. F, The ROC curve of IL-6, IL-8, TNF-α and 8-isopgF2α in EBC at T4. 8-isoPGF2α = 8-iso-prostaglandin F2α, EBC = exhaled breath condensate, IL-6 = interleukin-6, IL-8 = interleukin-8, ROC = receiver operating characteristic curve, TNF-α = tumor necrosis factor-α.

Table 3

The ability to predict postoperative acute respiratory distress syndrome of different variables.

Discussion

ARDS is a potentially life-threatening complication in patients undergoing cardiac surgery. Efforts should be made in order to identify those at higher risk aimed at decreasing the likelihood of developing this ominous condition and improve survival rates.[ EBC collection is a new, noninvasive method of studying the composition of fluid in the airway lining. Analysis of pulmonary biomarkers in EBC has been proven to be useful in characterizing the pathophysiology of pulmonary diseases and monitoring of airway inflammation.[ In our study, we compared the concentrations of IL-6, IL-8, TNF-α, and 8-isopgF2α in the patients undergoing cardiac valve replacement, calculated the correlation of them between EBC and serum. Then the ROC curves were used to evaluated the association between levels of mediators and postoperative ARDS. We choose these mediators because of they can predict the morbidity and mortality of ARDS.[ IL-6 function as a promoter of inflammatory reaction and was associated with development of lung injury. IL-8 can promote neutrophil degranulation, release elastase, endothelial injury, microcirculation of blood stasis, tissue necrosis, and organ function injury.[ TNF-α is an important cytokine and its concentration can reflect the degree of systemic inflammatory response after CPB.[ 8-isopgF2α is a sensitive indicator of oxidative stress.[ We found that the levels of IL-6, IL-8, TNF-α, and 8-isopgF2α increased after CBP and decreased at 6 hours after CPB in both EBC and serum, and the patients who were diagnosed as ARDS related to a higher level of these mediators. We also found the trends of these mediators in EBC were closely related to their trends in serum and most of the mediators had medium diagnostic value in diagnose of ARDS. The IL-6 in EBC at 6 hours after CPB had high diagnostic value and it was significantly higher in EBC than in the serum at that time point. But the IL-6 in serum at 6 hours after CPB and the 8-isopgF2α in serum at the end of CPB and 6 hours after CPB only had low diagnostic value. Therefore, we considered that EBC collection and evaluation are valuable for the prediction and diagnosis of ARDS, and sometimes may be more sensitive than serum.

Besides that, we compared the outcomes between ARDS and non-ARDS patients and found patients who suffered from ARDS experienced prolonged postoperative ventilator support, time to extubation, length of stay in the ICU, Post-ICU ventilator use, and postoperative length of stay. These results are consistent with previous studies.[ We also compared the hematologic and biochemistry indicators between them and found the ARDS patients related to higher WBC, BUN, Cr, Na and lower PLT, and ALB at 24 hours after operation. The WBC was significantly raised after operation and was significantly higher in the ARDS patients may be related to infection or systemic inflammatory response. The higher BUN and Cr indicated potential renal injury. The exact reason was unclear, but the disorder caused by CPB was probability involved. The higher level of natrium may be due to improper liquid management, but the patients with hypernatremia at the conclusion of CPB had longer hospital stays and higher postoperative complication rates.[ The lower platelet count may be caused by excessive consumption or death of platelet, and a research shows that thrombocytopenia is associated with an increased risk of ARDS.[ The lower albumin was likely caused by hypermetabolic like postoperative fever or excessive leakage due to pulmonary capillary injury, and albumin may be of great value in predicting and monitoring the severity and course of ARDS in critically patients with or at risk for the syndrome after new onset fever.[ Pulmonary oxygenation, an important indicator for evaluating lung function when lung injury occurs, may be directly reflected by P(A-a)O2 and RI.[ In this study, P(A-a)O2 and RI in the ARDS patients were higher than non-ARDS patients after CPB, suggesting that there are more serious damage of oxygenation and lung function in the ARDS patients. The reduction in pulmonary oxygenation and lung function may be caused by the imbalance of the ventilation/blood flow ratio, damage on the alveolar capillary barrier, or increase in inflammatory reactions and oxidative stress responses. The increase in postoperative Ppeak and Pplat in the ARDS patients may be related to the increase of lung water or decrease of pulmonary compliance, and may take part in the occurrence and development of ARDS. The comparison of PaO2/FiO2 further suggests a drastic decline in pulmonary ventilation and oxygenation function in the ARDS group. According to the above results, we considered that the ARDS patients were suffered from more damage on oxygenation, lung function, renal function and more disturbance on electrolyte, coagulation function, and nutrition metabolism. And these factors are likely to be the causes of the prolonged postoperative ventilator support, time to extubation, length of stay in the ICU, Post-ICU ventilator use and postoperative length of stay in ARDS patients.

However, there are some limitations in our research. The first one is the incidence of ARDS in our research was higher than article reports (ranges from 0.4%–20%).[ But it was lower than the report of Liu, et al.[ The different incidences may be caused by differences in design methodology of the studies, study populations, different ARDS definitions, and uses of preventive measures such as the use of intraoperative cell saver, use of leukocyte filtration, use of ultrafiltration, use of conservative fluid management, use of protective drugs, use of early extubation, use of lung protective ventilation and so on. Our diagnoses were made by 1 certain cardiac surgeon according to the patients condition and examination results concluded transthoracic echocardiography, coronary angiography, X-rays, blood gas analysis, N-terminal pro-brain natriuretic peptide (NT-Pro-BNP), et al. We considered that the possible causes of ARDS were systemic inflammatory response syndrome caused by surgical trauma, CPB, multiple blood transfusion, and infections.[ Patients undergoing cardiac surgery are particularly sensitive to lung damage due to these causes because cardiac operation alone with CPB can induce pulmonary ischemia/perfusion (I/R),[ the exposure of blood to abnormal surfaces and conditions would lead to inflammatory reactions and oxidative stress responses.[ Multiple blood transfusion can lead to transfusion related acute lung injury and infection is one of the most common reasons of ARDS. These factors can also explain the previous results in our study. Therefore, we must implement measures to prevent postoperative ARDS on the basis of the early detection of the EBC levels of biomarkers and it is our future research direction. The second limitation is the small sample size, which may cause the differences with the reported research. As we know, age, weight, smoking, hypertension, diabetes mellitus, duration of CPB, amount of bleeding, and transfusion volume are risk factors of ARDS. Although, the data of them were larger in ARDS patients, there were no significantly difference between the ARDS and non-ARDS patients. The third one is the patients were followed up only until discharge, and no statistical comparison was performed for long-term complications and mortality. The last one, EBC detection is not instantaneous, and this delay may have affected the accuracy of the results. And the components of EBC tend to become diluted.[

Conclusion

Our results suggest that EBC analysis of IL-6, IL-8, TNF-α, and 8-isopgF2α can be used to predict the occurrence of ARDS after cardiac valve replacement under CPB, and sometimes EBC may be more sensitive than serum. However, EBC detection has not been used as a routine method for the diagnosis of ARDS, and its practical value requires further study.

Author contributions

Conceptualization: Jie Song, Lei Yao, Jinliang Chen.

Data curation: Jie Song, Lei Yao, Lulin Zhao, Boxiang Du, Lin Liu, Jinliang Chen.

Formal analysis: Jie Song, Lei Yao, Lulin Zhao.

Investigation: Jie Song, Lei Yao, Lulin Zhao, Boxiang Du, Lin Liu, Jinliang Chen.

Methodology: Jie Song, Lei Yao, Jinliang Chen.

Supervision: Boxiang Du, Lin Liu.

Validation: Jie Song, Lei Yao, Lulin Zhao, Boxiang Du, Lin Liu, Jinliang Chen.

Writing – original draft: Jie Song, Lei Yao.

Writing – review & editing: Jinliang Chen.