The protocol was approved by the medical ethics committee of the affiliated hospital of Luzhou Medical college, and was registered in the Chinese clinical test registration center, with the registration number ChiCTR-TRC-11001402. Informed consent was obtained from all patients. Patients scheduled for esophagectomy were eligible for this study provided they agreed to postoperative analgesia. Exclusion criteria included New York Heart Association class (NYHA) III or IV, preexisting chronic obstructive pulmonary disease with preoperative forced expiratory volume (FEV1) of less than 80% in 1 second. Patients with coronary artery disease, morbid obesity (body mass index >35 kg/m2), cerebrovascular disease, or severe liver or renal malfunction were also excluded. Exit criteria included patient pulse oxygen saturation (SpO2) less than 90%, or surgery time less than 2 hours.
All patients received routine anesthesia, which included intravenous propofol (initially 2-3 mg/kg and subsequently 50-100 μg/kg/min), fentanyl (2-3 μg/kg) and remifentanil (0.1-0.2 μg/kg/min), and cisatracurium (0.2 mg/kg). A fiberoptic bronchoscope was employed for the insertion of the double-lumen tube. Mechanical ventilation was carried out using FiO2 (fraction of inspiration O2) of 0.8, a tidal volume (VT) of 9 ml/kg for TLV, and 7 ml/kg for OLV. The respiratory rate was adjusted to maintain arterial blood carbon dioxide partial pressure (PaCO2) at a level of 35 and 45 mm Hg throughout the surgery. OLV was initiated at the start of surgery, and was terminated when the definitive part of the surgical procedure ended. Ohmeda CAM anesthesia gas monitor (Datex-Ohmeda Inc., Tewkesbury, MA) was used to monitor the anesthesia gas and oxygen concentration.
All operations were carried out by the same experienced surgical team, which were also blinded to the trial protocol in this study. Surgical procedures included esophagectomy and esophageal reconstruction. The patients and the technician who performed the biomarker assays were also blinded to the randomization grouping.
Extubation was performed when the following criteria were fulfilled: (1) the patients were conscious and sufficiently alert; (2) SpO2>95% when breathing room air; and (3) PaCO2 was less than 50 mm Hg. All patients received intravenous patient controlled analgesia with fentanyl and tramadol, and visual analogue scale (VAS score) was recorded.
Patients were assigned randomly to four groups. Randomization was performed using computer generated random numbers, enclosed in sealed envelopes. In the control group (Control), OLV was initiated at the start of surgery, and the collapsed lung was reinflated with 80% oxygen. In the air reinflation group (AIR), room air was used to recover TLV, and to maintain ventilation for 5 minutes, followed by TLV with 80% oxygen. In the CPAP group, 5 cm H2O continuous CPAP with 80% oxygen was administered to the contralateral lung, and 80% oxygen was used for recovery of TLV. In the CPAP and AIR group (CPAP+AIR), the contralateral lung received CPAP with 80% oxygen during OLV, and room air was used to reinflate the lung and maintain ventilation during TLV.
A Simple Lightweight CPAP-Delivery Device
The CPAP-delivery device used in this trial was according to a previous study (10). It was composed with a three-way stopcock (Yi Xinda, Shenzhen, China) and a funnel-shaped piece of tube tightly connected to the proximal end of the bronchial lumen of a double-lumen tube. A pressure gauge (Fuyang Huayi meter factory, Hangzhou, China) was connected via a second stopcock for monitoring CPAP levels. The oxygen-flow rate was measured with an electrical flow meter (Novametrix Medical Systems, Wallingford, CT, USA) (11). All the materials used in this study are easy to acquire.
Blood Gas Analysis
Blood gas analysis and hemodynamic variables were measured and recorded at four time points: (1) 1 minute before initiation of OLV; (2) 30 minutes after initiation of OLV; (3) 60 minutes after initiation of OLV; (4) at the termination of OLV.
Bronchoalveolar Lavage Fluid (BALF)
Bronchoalveolar lavage fluid (BALF) was obtained 30 minutes after lung re-expansion using a flexible fiber-bronchoscope as described in a previous study. Seven successive 20-ml aliquots of 37 ℃ saline were instilled and aspirated immediately with 50 mm Hg suction (recovery 69±23 ml). BALF was centrifuged at 1,500 g for 10 minutes at 4 ℃. Cell-free BALF supernatant was stored at -80 ℃ for measurements of surfactant apoprotein A (SP-A) and surfactant apoprotein C (SP-C).
Arterial blood samples were collected from an indwelling arterial catheter for measurements of serum malondialdehyde (MDA), superoxide dismutase (SOD), tumor necrosis factor α (TNF-α), interleukin 1β (IL-1β), interleukin-8 (IL-8), and interleukin-6 (IL-6) levels at baseline (10 minutes after TLV) and 3 hours after reinflation. This plasma was centrifuged at 1,000 g for 15 minutes at 4 ℃. Supernatant was collected and stored at -80 ℃ until measurements were performed. Enzyme-linked immunoassays were used for measurements of these markers, and biomarker assays were performed using assay kits (Jianchen Bioengineering Institute, Nanjing, China) according to the manufacturer's instructions.
On the fourth postoperative day, patients were transported to the Department of Radiology for computed tomography (CT) scans. To avoid inter-observer variation, CT analysis was performed by the same investigator who was blinded to the patient's study group. Patients were told to lie in the supine position and raise their arms above their heads during CT scans. The examination was carried out at resting expiratory lung volume (11). Exposure time was 9 seconds for a 12 cm volume scan at 280 mA and 120 kV. Images were reconstructed with a slice thickness of 1.0 cm and a 512×512 matrix. The patients received a total estimated effective radiation dose of 1.5 millisieverts. Analysis of atelectasis and pulmonary aeration were done at 1 cm and 5 cm above the top of the left diaphragm. Lung area was delineated manually from the inner margins of the thoracic cage. Standard definitions of lung aeration according to attenuation values in Hounsfield Units (HU) were used. Aerated lung area was classified by volume elements with attenuation values between -100 HU and -1,000 HU, and atelectasis was defined by values between+100 and -100 HU (12). The amount of atelectasis was expressed as atelectatic area and the percentage of the total lung area.
Sample Size and Statistical Analysis
Power calculation was based on previous studies on ventilator-associated lung injury in esophagectomy (13). We calculated that 24 patients had to be included in each group to detect a difference in mean IL-6 concentration of 50%, an estimated SD of 60%, with a two-sided significance level of 0.05 and power of 80%. All quantitative data are expressed as means±SD. Male, ASA physical status, Smoke, and Stage pTNM (UICC) were expressed as a percentage, and differences between groups were compared using chi-square or Fisher exact tests. The intra group comparisons of respiratory and hemodynamic data at different time points were done using repeated measures analysis of variance (ANOVA). If significant differences between and within groups were found, multiple comparison with Bonferroni correction was applied. Differences with the rest of the data were examined by one-way ANOVA followed by the least-significant difference (LSD) test. P<0.05 was considered statistically significant. All analyses were carried out using SPSS version 13.0 (SPSS Inc, Chicago, IL, USA).