Abstract
Background
Resuming two lung ventilation (2LV) from one lung ventilation (OLV) has been proven to induce significant oxidative stress, mainly by superoxide release. Although total intravenous anesthesia and inhalational anesthesia are both used in thoracic surgery, different anesthetics may alter the oxidant/antioxidant balance.
Methods
Thirty patients undergoing thoracic surgery were randomly allocated to the propofol infusion (intravenous) group or isoflurane (inhalational) group after induction and placement of a double lumen endobronchial tube. Reactive oxygen species (ROS) production and total antioxidant status (TAS) were measured during OLV and 2LV manipulations. Blood samples were taken in the lateral position before OLV (T1), immediately before resuming 2LV (T2), and 5 minutes (T3) and 20 minutes (T4) after resuming 2LV, for measurement of ROS.
Results
ROS production increased significantly at T3 and T4 in both groups but to a lesser extent in the propofol infusion group. TAS levels increased with time in the propofol group but not in the isoflurane group.
Conclusion
Propofol infusion, compared with isoflurane inhalation, attenuates ROS production and limits it for 20 minutes after resuming 2LV from OLV. Propofol infusion may be beneficial to patients with inadequate antioxidant capacity.
Keywords
anesthetics, intravenous; propofol; reactive oxygen species; respiration, artificial;
1. Introduction
A sudden and significant increase in reactive oxygen species (ROS) production immediately following the resumption of two lung ventilation (2LV) from one lung ventilation (OLV) was found in our previous study.1 The magnitude of oxidative stress resulting from OLV-2LV management has been reported to be greater than that resulting from lung tumor resection.2 However, while there is little discussion on the effects of oxidative stress relating to the anesthetics used for thoracic surgery, there have been reports of cases where complications might be associated with oxidative stress.3,4 Although most patients tolerate OLV-2LV well, and have adequate antioxidant capacity to counteract sudden ROS production, it is not unusual in clinical anesthesia to apply OLV to patients with compromised antioxidant capacity, such as in critical illness5 and sepsis.6 Severe injury with oxidative stress may be induced by overwhelming ROS production when the patient lacks adequate antioxidant capacity.
Total intravenous anesthesia with propofol is frequently applied as are inhalational agents7,8 in thoracic surgery needing OLV since it causes no air pollution and less hypoxic pulmonary vasoconstriction. Propofol, with a chemical structure similar to alpha-tocopherol, is also considered to have antioxidant effects9,10 and has been shown in clinical investigations to be able to prevent the increase in ROS after tourniquet release even in a small dose.11
In this study, we evaluated the effects of propofol infusion on oxidant/antioxidant balance during OLV-2LV manipulation and compared them with those of isoflurane inhalation. ROS production and total antioxidant status (TAS) were measured during the scheduled procedure with minimal pulmonary trauma and a long period of intraoperative OLV. The goals of this study were to investigate if propofol infusion could prevent sudden ROS production and to verify if propofol infusion could change the pattern of ROS production after resuming 2LV.
2. Methods
With the approval of the Ethics Committee of National Taiwan University Hospital and patients’ written informed consent, 30 patients (ASA I−II, 20−60 years of age) scheduled to undergo thoracoscopic surgery or esophageal surgery, with limited lung trauma but lengthy OLV were enrolled in the study. After induction of anesthesia and double-lumen endobronchial tube placement, anesthesia was maintained either with isoflurane inhalation or propofol infusion in accordance with randomization. One anesthesiologist and three surgeons took part in this study. The investigators responsible for analyses of ROS production and TAS level were blinded to the anesthetic technique.
Anesthesia was induced with thiopental (3−5 mg/ kg), fentanyl (3 μg/kg), and pancuronium (0.1 mg/ kg). After induction, a left-sided double lumen endobronchial tube was placed in all patients and the position of the tube was adjusted by fiberoptic bronchoscope before and after patients turned to the lateral decubitus position. In the propofol group, propofol infusion was started at a rate of 10 mg/kg per hour for 10 minutes and then switched to 8 mg/ kg per hour throughout the study. Incremental doses of pancuronium were given for sufficient muscle relaxation. In the isoflurane group, 1−2% isoflurane was used to maintain anesthesia. Direct systolic arterial blood pressure (SABP) and diastolic arterial blood pressure (DABP), heart rate (HR), central venous pressure (CVP), ECG, pulse oximetry, temperature, urine output, and peak airway pressure were monitored. Intravenous crystalloids and fentanyl were titrated as needed to maintain systolic blood pressure within 15% of pre-induction values. Ventilation was mechanically controlled. OLV was performed with the following settings: fraction of inspired O2 (FiO2) = 1, tidal volume (VT) = 8−10 mL/kg, respiratory rate = 12−16 per minute, finely adjusted to maintain the arterial CO2 between 35−45 mmHg and I:E = 1:2.
The concentrations of the inspiratory and expiratory gases (FiO2, end tidal [ET] CO2, fraction of inspired isoflurane, ET isoflurane) were monitored continuously (Capnomac Datex, Finland). Direct observation of the collapsed non-ventilated lung was made through the open hemithorax. Arterial blood gases were analyzed 10 and 30 minutes after OLV to avoid hypoxemia. Blood samples (6 mL each) were drawn at designated time points, i.e. after induction of anesthesia with hemodynamic stabilization in the decubitus position and before OLV (T1) as a baseline reference, after surgical incision and at the end of OLV immediately before resuming 2LV (T2), and 5 minutes (T3) and 20 minutes (T4) after resuming 2LV for the measurement of ROS production and TAS. Of each blood sample, 1 mL was transferred to a tube, which was immediately wrapped in aluminum foil and kept in ice until chemiluminescence (CL) was used to determine ROS within 2 hours.12 The remaining 5 mL of the blood sample was placed in a tube containing EDTA. The sera were separated (2000 rpm, 10 minutes) immediately after sampling and stored at −80ºC until the determination of TAS was performed in due course (within 4 weeks).
2.1. CL testing of ROS production
Immediately before CL testing, 0.1 mL of phosphate buffered saline (PBS) (pH 7.4) was added to 0.2 mL of each blood sample. CL testing was performed in a completely dark chamber with the CL analysis system. After a 100-second background level determination, 1.0 mL of 0.1 mM lucigenin in PBS (pH 7.4) was injected into each sample. CL was continuously monitored for an additional 600 seconds. Integrating the area under the curve and subtracting the background level yielded the total amount of CL. This assay was performed twice for each sample and was expressed as CL counts/10 seconds for whole blood CL. The mean ± standard deviation of the CL level of each sample was calculated.
2.2. TAS assay
Commercially available laboratory kits were used for measuring TAS. The chain-breaking antioxidant capacity of the sera was determined using Randox Total Antioxidant Status kits (Cat. #NX2332; Randox, San Francisco, CA, USA) in accordance with the manufacturer’s instructions.13
2.3. Statistical analysis
The observed continuous variables were expressed as mean ± standard deviation. Hemodynamic and blood gas data were compared between the two groups and between the four time points by twoway analysis of variance (ANOVA). ROS production and TAS levels were further analyzed using multiple marginal regression models. The generalized estimating equation (GEE) method was used with repeatedly measured response variables to investigate differences between the isoflurane and propofol groups over time and with regard to patient characteristics (sex, age, operative site, anesthetic). Basic model-fitting techniques for regression analysis, including variable selection, goodness-of-fit assessment, and regression diagnostics, were used to assure the quality of the analytical results. If the first-order autocorrelation, i.e. AR(1), structure fitted the repeated measures data well, the model-based standard error estimates were used in GEE analysis; otherwise, empirical standard error estimates were reported instead. Regression analysis was performed using the Reg and GenMod procedures in SAS version 9.1 (SAS Institute Inc., Cary, NC, USA). Two-sided p< 0.05 was considered statistically significant.
3. Results
Thirty patients were enrolled randomly into the isoflurane inhalation and propofol infusion groups. The time from induction to T1 was 45 ± 9 minutes in the propofol group and 42 ± 8 minutes in the isoflurane group. The basic characteristics of the patients and the duration of OLV in each group are shown in Table 1. Hemodynamic data in each group during the OLV and 2LV periods are listed in Table 2. SABP, DABP, HR and CVP did not change significantly during the study. PO2 values decreased significantly at 30 minutes during OLV in both groups but returned to a level comparable with baseline values after 2LV. The decrease in PO2/FiO2 during OLV was comparable in both the isoflurane and propofol groups. Positive end-expiratory pressure was not applied in any patients because of satisfactory oxygenation during OLV. Postoperative recovery was smooth and without complications in all patients.
3.1. ROS production during OLV-2LV
In the isoflurane group, the mean level of ROS production was significantly higher at T3, and even higher at T4 than at T1 (Figure 1). However, in the propofol group, the mean level of ROS production at T3 was significantly higher than at T1 and T2, but decreased significantly at T4. Covariates that significantly affected the mean level of ROS production are listed in Table 3. When the values of other covariates were fixed, the mean level of ROS production at T4 was significantly lower in male patients and patients under 35 years of age. Moreover, the mean level of ROS production was significantly lower in the propofol group at T3 and T4, compared with the isoflurane group, after adjusting for the effects of the other covariates.
Download full-size image
3.2. TAS capacity during OLV-2LV
TAS was significantly higher in the propofol group (1.19 ± 0.04) than in the isoflurane group (0.99 ± 0.05) from T1 onward (Figure 2). However, taking into account the TAS level difference at T1, changes in ROS production were not correlated with changes in TAS levels.
ROS production and TAS levels were further analyzed using multiple marginal regression models. The GEE method was used with repeatedly measured response variables to investigate differences between the isoflurane and propofol groups over time and with regard to patient characteristics (sex, age, operative site, anesthetic). Significantly less ROS production was also found in younger (age < 35 years) and male patients (Table 3). In the comparison of anesthetics used, propofol infusion induced significantly less ROS production at T3 and T4.
Download full-size image
4. Discussion
Our results confirmed that ROS production was significantly lower in the propofol infusion group at T3 and T4 after adjusting for the effects of the other covariates. Propofol infusion not only reduced ROS production after resuming 2LV from OLV but also shortened the duration of ROS production. The beneficial effects of propofol infusion might thus prevent pulmonary and cardiovascular complications arising from massive release of oxidative radicals from re-expansion of the collapsed lung. These results are in accord with those of previous animal studies on ischemia-reperfusion injury,14 which showed that free radicals decrease more rapidly under propofol infusion.15 However, it was also found in this study that a longer OLV time did not result in more ROS production after resuming 2LV. Younger, male patients also had lower ROS production during OLV2LV manipulations.
The comparable PO2/FiO2 during OLV in both groups did not support the possibility that lower oxidative stress resulted from reduced effects on hypoxic pulmonary vasoconstriction. In a previous study on OLV-2LV, it was revealed that sevoflurane inhalation did not increase shunt flow significantly more than propofol.12
A higher TAS level was also maintained under propofol infusion, compared with isoflurane, during OLV-2LV from T1 to T4. Since the data for T1 was obtained about 45 minutes after propofol infusion following induction, propofol infusion might induce a higher antioxidant status, though we did not measure the TAS level before anesthesia. However, our results showed that changes in ROS production were not correlated with changes in TAS levels and thus do not support the proposal that propofol infusion may increase TAS level (p= 0.0591).
It was shown that under adequate anesthesia, either by propofol infusion or isoflurane inhalation, surgical stress did not induce significant ROS production. The results of this study showed that there was massive ROS production during OLV-2LV manipulation which could not be fully captured by normal antioxidative capacity under adequate anesthesia. Compared with oxidative stress in normal thoracotomy, severe ROS production remains an important consideration in lung transplantation surgery where reperfusion injuries may alter the outcome.
It has been shown that TAS levels in critically ill16 and aged17 patients are lower than those in this study. However, in critically ill and aged patients, the use of total intravenous anesthesia with propofol infusion may be limited because of their unstable hemodynamics. Although we have not proven this to be true in this study, in view of the beneficial effects of propofol infusion on ROS production, a smaller dose of propofol,11 which does not affect blood pressure, should be considered.
While Lu et al’s results showed that, in animals under ischemia-reperfusion, a linear relationship existed between changes in pulmonary arterial pressure and extravascular albumin accumulation, it was also revealed that early lung injury was not neutrophildependent.18 However, if massive ROS production immediately after 2LV is prevented or reduced by anesthetics or other means, subsequent neutrophildependent lung complications will be milder.
We conclude that propofol infusion attenuates ROS production after resumption of 2LV from OLV, which may induce pulmonary and cardiovascular complications without affecting hypoxic pulmonary vasoconstriction. With higher maintenance of TAS levels, propofol infusion may be considered when the use of OLV is necessary in critically ill patients or in patients with reduced antioxidant capacity.
Acknowledgments
The authors would like to thank Dr. Fu-Chang Hu, National Center of Excellence for General Clinical Trial and Research, National Taiwan University Hospital and College of Public Health, National Taiwan University for the marginal regression analysis of the repeated measures data using the GEE method and Ms. Ling-Chu Wu for her assistance in the statistical computing. The study was supported by the National Science Council, NSC92-2314-B-002-259.