This study was approved by the Ethics Committee of the First Affiliated Hospital of China University of Science and Technology (Ref:2022-ky-027). This study was registered in the Chinese Clinical Trial Registry on 24/11/2021 (ChiCTR2100053598). We used the CONSORT checklist when writing our report. 180 patients from February 2022 to September 2022. All patients were assigned in a 1:1:1 ratio to the staged GDT group (group A), GDT group (group B), and conventional fluid therapy group (group C). The randomization was stratified by sequential blocking based on the computer random number generator. The anesthesiologists responsible for intraoperative management were informed of the patient’s’ grouping, whereas the rest of the research team and the patients were blinded to the grouping.

Inclusion criteria were as follows: (1) those who were classified ASA II–III, (2) those who were scheduled for McKeown esophagectomy. Exclusion criteria were as follows: (1) patients with severe liver or kidney dysfunction (Blood creatinine > 442µmol/L, Child C), (2) those with cardiac dysfunction (New York Heart Association (NYHA) III-V) or abnormal preoperative pulmonary function (forced expiratory volume in 1 s < 50% of predicted), (3) those with severe arrhythmia, (4) those with intraoperative blood loss > 500 mL or those who required blood transfusion.

The patients received a restricted diet 6–8 h prior to surgery. Before the induction of anesthesia, a standardized catheterization of the right internal jugular vein was performed, and a 20-G arterial catheter (B. Braun Medical Inc, Bethlehem, PA, USA) was inserted into the radial artery of the left forearm. Physiological parameters such as heart rate (HR), mean arterial pressure (MAP), central venous pressure (CVP), pulse oximetry (SpO2), partial pressure of end-tidal carbon dioxide (PetCO2), body temperature, and bispectral index (BIS) were continuously monitored and recorded. In addition, the LiDCO system (LiDCO, Ltd, London, UK) was utilized to continuously monitor SVV, cardiac index (CI), CO, and stroke volume (SV). Then etomidate 0.3–0.4 mg·kg-1, sufentanil 0.5–0.7 µg·kg-1, and rocuronium 1.0 mg·kg-1 were administered intravenously for induction of general anesthesia. Intravenous infusion of propofol (3–6 mg·kg-1·h-1) and remifentanil (0.1–0.3 µg·kg-1·min-1) were maintained throughout the procedure to achieve the BIS target value of 45–60. After the induction of anesthesia, a right bronchial occluder (Yichang Humanwell, China) was located by bronchoscopy. Ventilation mode was set to volume controlled mode, the ventilation parameters are set as follows. In OLV, tidal volume (VT): 7 mL·kg-1; FiO2: 100%; expiratory time ratio (I/E):1:1.5. In TLV, tidal volume (VT): 8 mL·kg-1; FiO2: 50%; expiratory time ratio (I/E):1:2. Positive end-expiratory pressure (PEEP) of 5 cmH2O was applied. The ventilation parameter were adjusted to maintain PetCO2 at 35–45 mmHg, and the nasopharynx temperature was maintained at > 36 °C by a fluid warmer and medical warming blanket. As needed, additional muscle relaxants and analgesics were intermittently administered, and all patients used patient-controlled intravenous analgesia (PCIA) postoperatively.

First, the patient is placed in the left lateral position with OLV, and thoracic procedures are performed under thoracoscopy, including the dissection of the azygos vein arch and the thoracic esophagus. Lymph nodes in the paraesophageal, subcarinal, bilateral recurrent laryngeal nerve, and inferior pulmonary ligament are removed based on the tumor location and extent of invasion. Subsequently, the patient is placed in the reverse Trendelenburg position with TLV, and laparoscopy is used to free the stomach and create a tubular stomach. Finally, an incision is made in front of the left sternocleidomastoid muscle to complete the dissection, severance of the cervical esophagus, and esophagogastric anastomosis.

Groups A, B, and C received a continuous infusion of lactated Ringer’s solution at a rate of 3 mL·kg-1·h-1 after induction of anesthesia. Group A followed the protocol illustrated in Fig. 1. During OLV, when SVV ≥ 10% persisted for more than 3 min, colloid 4 ml·kg-1 (6% hydroxyethyl starch 130/0.4, Fresenius Kabi AG, Bad Homburg, Germany) was administered over 15 min. The above process was repeated until the SVV < 10%. An infusion of dobutamine was given at 3–5 µg·kg-1·min-1 if SVV drop < 2% after infusion or SVV < 10% with concurrent CI < 2.5 L·min-1·m-2. Additionally, norepinephrine was pumped at 4–10 µg·min-1 if the MAP fell below 65 mmHg after fluid infusion and CI ≥ 2.5 L·min·m2. The infusion was slowed or stopped when SVV < 8%. Furthermore, hemodynamic parameters were evaluated every 5 min. During the TLV, above protocol was applied with the target SVV set at 8–12%. Similarly, group B followed the same protocol as group A, with the exception that the target SVV range was set at 8–12% during the whole surgery. Conversely, in group C, hemodynamic management was determined by the discretion of the attending anesthesiologist, if necessary, aiming at MAP > 65 mmHg, HR 60,100 bpm, CVP 6–12 mmH2O, and the urine output > 0.5 mL/kg/h.

MAP, HR, CI, and CO were recorded before surgery (T0), 30 min after OLV (T1), 60 min after OLV (T2), 30 min after TLV (T3), and at the end of surgery (T4). In addition to these parameters, the total infusions of crystalloids and colloids, blood loss, urine output, and vasoactive drug requirements were recorded and analyzed.

Goal-directed fluid therapy protocol. SVV, stroke volume variation; CI, cardiac index; MAP, mean arterial pressure

The primary outcome was the incidence rate of pulmonary complications within 7 d after surgery. The occurrence of pulmonary complications was assessed using the Melbourne Group Scale Version 2 (MGS-2), which comprises eight parameters. The diagnosis of pulmonary complications was confirmed if four or more of the following criteria were met: (1) body temperature > 38 °C, (2) white blood cell count > 11.0 × 109 L-1 or administration of antibiotics for respiratory infections, (3) pneumonia or pulmonary infection diagnosed by physicians, (4) chest radiograph showing atelectasis/consolidation, (5) yellow-green sputum of a different nature from the preoperative one, (6) positive sputum culture results, (7) blood oxygen saturation (SpO2) < 90% on room air. or (8) readmission to the intensive care unit (ICU) for respiratory problems or staying in the ICU for > 36 h [13].

Radial artery blood was collected at T0, T1, T2, T3, and T4. Additionally, the oxygenation index (OI = PaO2 / FiO2), alveolar (A) - arterial (a) oxygen partial pressure difference (PA-aDO2 = FiO2 (PB - PH2O) - PaCO2 / R - PaO2, PB, Atmospheric Pressure; R, respiratory quotient), and respiratory index (RI = PA - aDO2 / PaO2) were calculated and recorded.

At T0, 2 h after surgery (T5), and 48 h after surgery (T6), LUS was performed according to international evidence-based recommendations for point-of-care LUS based on a simplified four-sector scan [14] by a skilled operator using a portable ultrasound device. The left thorax was examined to avoid potential confounding factors from subcutaneous emphysema and surgical effects. The thorax was partitioned into four areas on one side: upper anterior, upper lateral, lower anterior, and lower lateral, delimited by the anterior axillary line and the fourth intercostal space. B-lines artifacts (BLAs) are defined as vertical, laser-like, hyperechoic images that originate from the pleural line and move with respiration. The number of B-lines detected in each area was tallied and subsequently scored. A score of 0 points was assigned for no BLA, 1 point for 1–3 BLAs, 2 points for 3–5 BLAs, and 3 points for > 5 BLAs. The total B-lines score was calculated as the sum of the scores of the four areas.

This study is a randomized controlled trial, with the incidence of postoperative pulmonary complications as the primary endpoint. Based on previous research reports on similar cases, the incidence of pulmonary complications in esophagectomy cases treated with conventional fluid therapy is approximately 35%. Similarly, the incidence of pulmonary complications after GDT is approximately 10%.The estimated incidence of pulmonary complications in the staged GDT group is 5%.the sample size was calculated using PASS V.15 software, setting α = 0.05 (two-tailed) and (1-β) = 0.99. The total sample size required was found to be 165 cases. Taking into account potential dropouts, the sample size was increased to 180 cases, with 60 cases in each group.

Statistical analysis was conducted using SPSS 15.0. Continuous data are presented as mean ± standard deviation, and their normality was assessed using the Kolmogorov-Smirnov test. For normally distributed continuous data, t-tests (paired or unpaired) were used for analysis, while for non-normally distributed continuous data, the Mann-Whitney U test and Wilcoxon signed-rank test were employed for unpaired and paired analyses, respectively. Repeated measures one-way analysis of variance (ANOVA) and Bonferroni-correction was used to assess significant differences among three groups. The analysis of categorical data employed the chi-square test or Fisher’s exact test. P < 0.05 was considered statistically significant.