Design
This prospective, patient- and sonographer-blinded, single-center, parallel, randomized, controlled trial was approved by the Institutional Review Board of Seoul National University Hospital (No. 1903-137-1020, 22 April 2019) and registered at ClinicalTrials.gov (NCT03943433, 7 May 2019). The study was conducted in accordance with CONSORT guidelines. We enrolled adult patients scheduled to undergo elective laparoscopic gynecologic surgery or colorectal surgery in the Trendelenburg position from May to November 2019 after obtaining written informed consent. The inclusion criterion was adult patients aged 20–70 years with an American Society of Anesthesiologists physical status I–II. The exclusion criteria were patients with body mass index ≥35 kg m-2, cardiovascular impairment, severe chronic obstructive pulmonary disease (preoperative forced expiratory volume in 1 second/forced vital capacity of 60% or lower) or emphysema, pneumothorax or bullae, previous lung resection surgery, and increased intracranial pressure. Some patients dropped out because of protocol violation, massive bleeding with hemodynamic compromise, or unexpected open conversion.
Patients were randomly assigned to two groups based on the applied Fio2 during RM, in a 1:1 ratio, by computer-generated randomization, using R software (version 3.5.1, R Foundation for Statistical Computing, Vienna, Austria). Allocation was concealed in an opaque envelope by an assistant not involved in the study and was delivered to the attending anesthesiologist before general anesthesia induction. The sonographer (BRK or HB) was completely blinded to the group assignment.
Anesthesia and ventilator strategy
General anesthesia was induced according to the predetermined protocol with standard monitoring of pulse oximetry (Spo2), non-invasive blood pressure, electrocardiography, bispectral index (A-2000 XP; Aspect Medical Systems, Newton, MA), and end-tidal carbon dioxide concentration. After preoxygenation with 100% oxygen, propofol 1.5–2.0 mg kg-1 was administered intravenously with a continuous target-controlled remifentanil infusion (Orchestra; Fresenius Kabi, Brézins, France). Rocuronium 0.6–0.8 mg kg-1 was administered for neuromuscular blockade, and tracheal intubation was performed. General anesthesia was maintained with sevoflurane and remifentanil to maintain the bispectral index within 40‒60. A radial arterial catheter was placed and connected to an arterial waveform analysis system (Flotrac; Edwards Lifesciences, Irvine, CA) to monitor continuous arterial blood pressure and the cardiac output.
Mechanical ventilation was maintained intraoperatively with the Fio2 at 0.4, tidal volume at 8 ml kg-1 of ideal body weight, PEEP at 5 cmH2O, inspiration to expiration ratio of 1:2, and end-inspiration pause 10% at volume-controlled ventilation mode. Respiratory rate was adjusted to maintain partial pressure of arterial carbon dioxide at 35‒45 mmHg. If the peak airway pressure exceeded 35 cmH2O, the tidal volume was decreased stepwise by 1 ml kg-1 until the peak pressure was <35 cmH2O.
At the end of the surgery, sugammadex 2–4 mg kg-1 was administered after train-of-four count monitoring for reversal of neuromuscular blockade. The Fio2 was changed to 1.0 when the first spontaneous breathing was observed. After extubation, patients were transferred to the post-anesthesia care unit (PACU). Intravenous patient-controlled analgesia was routinely used for postoperative pain control. Patients were discharged from the PACU when they met the Modified Aldrete Score criteria (20).
Lung ultrasound examination and RM strategy
Lung ultrasound examination was performed at three time-points: 1 minute after starting mechanical ventilation, at the end of surgery (before emergence), and after breathing room air for 20 minutes at PACU (Figure 1). Lung ultrasound was performed by two investigators (BRK and HB) blinded to the group assignment. Both investigators had performed more than 100 cases of lung ultrasound. The ultrasound was performed in the supine position using a Vivid-I ultrasound device (GE Healthcare, Chalfont St. Giles, Bucks, UK) and a convex probe, with a frequency of 2.5 MHz–7.5 MHz. All intercostal spaces were examined as previously described: each hemithorax was divided into six regions with three longitudinal lines (parasternal, anterior, and posterior axillary) and two axial lines (one above the diaphragm and another at 1 cm above the nipples) (15). Each region was scored according to the modified LUSS system suggested by Monastesse et al., which showed sufficient sensitivity to detect loss of aeration during laparoscopic surgery (21). The degree of deaeration was rated from 0 to 3 as follows (Figure 2): 0, 0–2 B lines; 1, ≥3 B lines or 1 or multiple subpleural consolidations separated by a normal pleural line; 2, multiple coalescent B lines or multiple subpleural consolidations separated by a thickened or irregular pleural line; and 3, consolidation or small subpleural consolidation exceeding 1×2 cm in diameter (21). The points for the 12 regions were summed for analysis. Furthermore, we defined significant atelectasis as a score of 2 or 3 assigned to any region.
RMs were performed after lung ultrasound examinations (twice) under real-time ultrasound guidance, with the probe placed at the region with the highest score. After setting the Fio2 (1.0 or 0.4) according to the assignment, continuous positive airway pressure was applied from 15 cmH2O in 5-cmH2O stepwise increments, up to the pressure at which no collapsed area was observed. The maximum continuous airway pressure applied during RM was 40 cmH2O. The applied pressure (opening pressure) and the duration of the RM were recorded. Additional intraoperative RMs were performed at several time-points: at the time of Trendelenburg positioning and at every 30 minutes thereafter, and after a return to supine position after procedure completion (Figure 1). Intraoperative RMs were performed using the initial pressure and duration after adjustment of Fio2 according to the group assignment. The pre-designated Fio2 was applied only during the RM, after which it was maintained at 0.4 throughout mechanical ventilation in both groups.
Outcomes
The primary outcome was the modified LUSS at surgery completion (before emergence), reflecting an aeration loss during general anesthesia. The secondary outcomes were the modified LUSS at PACU, significant atelectasis observed on lung ultrasound, intraoperative and postoperative Pao2 to Fio2 ratios, and incidences of intraoperative desaturation (Spo2 <95%), postoperative fever (body temperature ≥38°C during hospital stay), and postoperative pulmonary complications during hospital stay. Arterial blood samples were obtained 20 minutes after a change in position from supine to Trendelenburg and after breathing room air for 20 minutes at the PACU. Postoperative atelectasis, pneumonia, acute respiratory distress syndrome, and pulmonary aspiration data were collected by reviewing medical records. Their severity was evaluated based on previous consensus definitions for standardized perioperative pulmonary complications (22). In our study, in-hospital pulmonary complications were atelectasis, pneumonia, acute respiratory distress syndrome and mild-to-severe pulmonary aspiration. Data on postoperative pulmonary complications were collected during the hospital stay. Additionally, data on age, height, weight, sex, type of operation, duration of anesthesia and surgery, pressure and duration of RM, and ventilator parameters were collected. Significant hemodynamic deterioration during RM (>20% of baseline) was documented and treated with vasoactive drugs or crystalloid agents.
Statistical analysis
In our pilot study on patients undergoing laparoscopic surgery in the Trendelenburg position (n=10), the modified LUSS [mean (SD)] before and at the end of surgery were 3.88 (1.26) and 8.66 (2.82), respectively. Considering a 20% decrease in the modified LUSS in the low Fio2 group, we calculated that 44 patients would be needed in each group, with a type-I error risk of 0.05 and a power of 0.8 for two-tailed analysis.
Continuous variables were summarized as mean (SD) or median (interquartile range). The variables were analyzed using unpaired or paired t-tests and the Mann‒Whitney U or Wilcoxon signed-rank tests, after assessing the normality of data distribution with the Shapiro‒Wilk test. Number of patients (%) was compared using the chi-squared test or Fisher’s exact test. Statistical analyses were performed with R software (version 3.5.1, R Foundation for Statistical Computing, Vienna, Austria). For all analyses, P<0.05 was statistically significant.