INTRODUCTION
Since their introduction in 1999, robot-assisted surgeries have gradually expanded owing to several advantages, such as enhanced surgical precision and reduced tissue trauma [1,2]. Particularly in prostatectomy, robotic assistance has become the standard approach, demonstrating fewer postoperative complications than open surgery [3]. However, robot-assisted laparoscopic prostatectomy (RALP) presents unique respiratory challenges. The required pneumoperitoneum with CO2 insufflation, which is essential for surgical exposure, compromises ventilation. This effect is further exacerbated by the steep Trendelenburg position necessary during the procedure, leading to a ventilation-perfusion mismatch and an increased risk of postoperative pulmonary complications [4]. Recent studies have shown that appropriate alveolar recruitment maneuvers followed by adequate positive end-expiratory pressure (PEEP) can improve respiratory parameters and oxygenation during laparoscopic procedures [5,6].
The optimal PEEP level remains controversial, as fixed PEEP strategies do not account for individual variations in the chest wall and lung mechanics. Driving pressure (DP, ΔP), defined as the difference between plateau pressure (Pplat) and PEEP (ΔP=Pplat − PEEP), represents the dynamic strain applied to ventilated alveoli and serves as a surrogate for the functional size of the “baby lung,” which is the portion of lung available for ventilation in a patient [7]. Unlike fixed PEEP, which applies a uniform end-expiratory pressure regardless of individual lung compliance, DP-guided PEEP titration offers a physiologically individualized approach by dynamically adapting to the respiratory mechanics of the patient. This individualized strategy aims to minimize regional overdistension in non-dependent lung regions while preventing cyclic atelectasis in dependent regions, thereby promoting a more homogeneous ventilation distribution [8-11].
Minimizing DP has been strongly associated with improved survival in patients with acute respiratory distress syndrome (ARDS) [8,10], and recent surgical studies have demonstrated improved oxygenation and reduced pulmonary stress when intraoperative DP is optimized [9,11]. A recent meta-analysis of over 2,000 surgical patients demonstrated a significant dose-dependent association between increased intraoperative DP and postoperative pulmonary complications [11], supporting the clinical relevance of DP minimization during surgery.
We designed this prospective randomized controlled trial to evaluate whether DP-guided PEEP titration during RALP, compared with conventional fixed PEEP, reduced intraoperative atelectasis, as assessed by lung ultrasound (LUS) score, improved oxygenation, and reduced the incidence of postoperative pulmonary complications.
MATERIALS AND METHODS
Study design and ethical approval
This single-center, prospective, randomized controlled trial was approved by the Institutional Review Board of Seoul National University Hospital (No. 2002-163-1107, 28 April 2020). The study protocol was registered at ClinicalTrials.gov (NCT04327193, 26 March 2020). Written informed consent was obtained from all patients before enrollment.
Participants
1. Inclusion criteria
We included adult patients (age ≥ 18 years) with American Society of Anesthesiologists (ASA) physical status I or II scheduled for elective RALP under general anesthesia between May and October 2020.
2. Exclusion criteria
We excluded patients with ASA physical status ≥ III, cardiovascular impairment, severe chronic obstructive pulmonary disease (preoperative forced expiratory volume in 1 s [FEV1]/forced vital capacity [FVC] ≤ 60%), emphysema, bullae, pneumothorax, those who had undergone previous lung resection surgery, and those with increased intracranial pressure. Patients were also excluded due to protocol violations, massive bleeding, or unexpected open conversion.
Randomization and blinding
Permuted block randomization (block sizes of 4 and 6) without stratification was performed using R software (version 3.6.0, R Foundation for Statistical Computing) to assign patients to the DP or control groups in a 1:1 ratio. An independent statistician not involved in patient care generated the randomization sequence. Group assignments were concealed in sequentially numbered opaque sealed envelopes. The sealed assignment was disclosed only to the attending anesthesiologist immediately before anesthesia induction. The anesthesiologist induced general anesthesia and set the ventilator according to the study protocol and group assignment. The investigators responsible for lung ultrasound image acquisition and scoring were blinded to group allocation.
Assessment time points
Four assessment points were predefined in this study (Fig. 1). T1 (baseline) was defined as 10 min after the initiation of mechanical ventilation in the supine position before the surgical incision. T2 (30 min-Trendelenburg) was measured 30 min after Trendelenburg positioning with CO2 insufflation. T3 (end-Trendelenburg) was assessed at the completion of the Trendelenburg position before returning to the supine position. T4 (postoperative) was evaluated 30 min after admission to the post-anesthesia care unit (PACU).
General anesthesia management
Standard monitoring included pulse oximetry, blood pressure, electrocardiography, bispectral index (A-2000 XP, Aspect Medical Systems), and end-tidal CO2. Anesthesia was induced with propofol 1.5-2.0 mg/kg and remifentanil using target-controlled infusion (Orchestra, Fresenius Kabi), followed by rocuronium 0.6-0.8 mg/kg for tracheal intubation. Fraction of inspired oxygen (FiO2) was set to 1.0 during induction. Radial arterial catheterization enabled continuous hemodynamic monitoring using the FloTrac system (Edwards Lifesciences). Anesthesia was maintained with sevoflurane and remifentanil targeting Bispectral Index 40-60, with continuous rocuronium infusion (0.01 mg/kg/min) during surgery. FiO2 was maintained at 0.40 unless adjustment was needed for adequate oxygenation. Fluid management and vasoactive drugs including phenylephrine and ephedrine boluses were administered at the discretion of the anesthesiologist. Upon completion of the procedure, the neuromuscular blockade was reversed with sugammadex at 2-4 mg/kg. FiO2 was increased to 1.0 during emergence. After extubation, the patients were transferred to the PACU, where discharge timing was determined using the modified Aldrete score [12]. Supplemental oxygen was delivered via a facial mask at a rate of 5 L/min. All the patients received intravenous patient-controlled analgesia for postoperative pain management.
Ventilation strategy
Upon induction of general anesthesia, alveolar recruitment was performed identically in both groups using volume-controlled ventilation (Primus Infinity Empowered, Dräger). PEEP was increased every four breaths from 5 to 15 cmH2O in 5 cmH2O increments, with a tidal volume of 8 ml/kg ideal body weight, respiratory rate 10/min, inspiration:expiration ratio of 1:1, and end-inspiratory pause of 30%. After recruitment, a ventilation strategy was applied according to group allocation.
1. Intervention group (DP-guided PEEP titration)
PEEP was titrated to minimize DP (Pplat minus PEEP), which was calculated after maintaining five respiratory cycles at each decremental PEEP level from 10 to 0 cmH2O. During PEEP titration, the ventilation parameters included a tidal volume of 8 ml/kg of ideal body weight, respiratory rate 12/min, inspiration:expiration ratio of 1:2, and end-inspiratory pause of 30%. The lowest PEEP level was selected when the lowest DP was observed at multiple PEEP levels. The determined PEEP was maintained during mechanical ventilation with a tidal volume of 8 ml/kg ideal body weight, inspiration:expiration ratio of 1:2, and end-inspiratory pause of 30%. Furthermore, the respiratory rate was adjusted to maintain an arterial carbon dioxide tension (PaCO2) of 35-45 mmHg.
2. Control group (Fixed PEEP)
The control group received a fixed PEEP of 5 cmH2O [13] after alveolar recruitment.
Recruitment with or without PEEP titration was performed at three time points: 10 min after mechanical ventilation (after initial LUS examination and arterial sampling), after Trendelenburg positioning, and after returning to the supine position (Fig. 1). In both groups, if the peak inspiratory pressure (PIP) exceeded 35 cmH2O, ventilation was switched to the pressure-controlled mode to prevent barotrauma. Rescue interventions (increased FiO2, recruitment maneuvers, or adjustments in tidal volume and PEEP) were implemented and documented for SpO2 < 90%.
Lung ultrasound protocol
For atelectasis assessment, we employed lung ultrasound, which offers comparable sensitivity compared to computed tomography in detecting atelectasis while avoiding radiation exposure and invasive procedures [14]. Examinations were performed using a Vivid-I ultrasound device (GE HealthCare) with a 2.5-7.5 MHz convex probe. Each hemithorax was divided into six regions using the parasternal, anterior, and posterior axillary lines vertically and two horizontal lines at the level of the diaphragm and 1 cm above the nipple line. All 12 regions were examined at T1, T3, and T4. Images were stored as video clips for subsequent analysis.
Two blinded investigators scored the images from 0 to 3 using a validated LUS scoring system [15]. A score of 0 indicated 0-2 B-lines, 1 indicated ≥ 3 B-lines or small consolidations with a normal pleural line, 2 indicated coalescent B-lines or subpleural consolidations with an irregular pleural line, and 3 indicated a consolidation larger than 1 × 2 cm. When scoring discrepancies occurred, a consensus was reached through discussion. The B-line and consolidation scores were calculated separately and summed across all regions. Significant atelectasis was defined as consolidation score ≥ 2 in any region [16].
Data collection and measurements
1. Preoperative assessment
Baseline demographics, smoking status (categorized as never, former if cessation occurred 4 weeks or more before surgery, or current if smoking occurred within 4 weeks), comorbidities, ASA status, laboratory values including hemoglobin and SpO2, and pulmonary function tests were collected. The Assess Respiratory Risk in Surgical Patients in Catalonia score, which ranges from 0 to 123, was calculated to predict postoperative pulmonary complications [17].
2. Intraoperative monitoring
Hemodynamic variables, including heart rate (HR), mean arterial pressure (MAP), cardiac output, and cardiac index, were recorded at T1-T3 using the FloTrac system. Monitoring of respiratory mechanics included dynamic compliance recorded from the ventilator display (Primus Infinity Empowered) during PEEP titrations, PIP, Pplat, DP, and applied PEEP levels throughout the procedure. The documented surgical data included CO2 insufflation pressure, Trendelenburg angle and duration, operation time, and total anesthesia duration. Fluid balance was assessed through documentation of crystalloid and colloid volumes administered, estimated blood loss, urine output, and all vasoactive medications used. Arterial blood gas analysis was performed at all four predetermined time points.
3. Postoperative surveillance
Body temperature was monitored every 8 h for 72 h postoperatively, with fever defined as a temperature of ≥ 37.5°C. Patients underwent daily clinical assessments to identify pulmonary complications within 72 h. Details regarding the definitions of postoperative pulmonary complications are available in the Supplementary File 1. The length of hospital stay from surgery to discharge was recorded.
Outcome measures
1. Primary outcome
The primary outcome was the modified LUS score at T3 (at the end of the Trendelenburg position).
2. Secondary outcomes
Secondary respiratory outcomes included LUS score at T4, incidence of significant atelectasis at T3 and T4, arterial oxygen tension (PaO2) values at T2, T3, and T4, and occurrence of intraoperative oxygen desaturation defined as SpO2 < 90%. Clinical outcomes included the incidence of postoperative fever within 3 days, duration of hospital stay, and postoperative pulmonary complications within 72 h. Hemodynamic stability was assessed by comparing the HR, MAP, cardiac output, and cardiac index between the groups at T1-T3. The ventilatory parameters compared between groups included dynamic compliance, PIP, Pplat, and DP. Safety outcomes included the incidence of PIP exceeding 35 cmH2O, frequency of oxygen desaturation events requiring intervention, and all documented rescue interventions.
Sample size calculation
Given the absence of LUS data specific to robot-assisted laparoscopic surgery, our calculations were based on the most comparable available evidence. Using aggregated data from all patients in a previous Trendelenburg laparoscopic surgery study [18], we estimated a baseline modified LUS score of 11.90 (SD 4.48). Although the measurement timing in that study (post-extubation) differed from our primary endpoint, this represented the best approximation available during study design. We anticipated a 30% reduction with DP-guided PEEP based on related literature [19]. Sample size calculation (G*Power 3.1.9.4) indicated 26 patients per group would provide 80% power with α=0.05 (two-tailed). We enrolled 33 patients per group to account for 20% attrition.
Statistical analysis
All statistical analyses were performed using R software (version 3.6.0, R Foundation for Statistical Computing). Continuous variables were analyzed with unpaired or paired t-tests and the Mann-Whitney U or Wilcoxon signed-rank tests after assessment for normality using the Shapiro-Wilk test. The results of the analysis are presented as the mean ± standard deviation (SD) or median (1Q, 3Q). The number of patients (%) was compared using the chi-square test or Fisher’s exact test. Categorical variables were analyzed using the chi-square test or Fisher’s exact test and are presented as numbers (%). Statistical significance was set at P<0.05.
RESULTS
Of the 101 patients assessed for eligibility, 66 were enrolled and randomized (33 per group). Three patients were excluded because of delayed surgical schedules. The final analysis included 31 and 32 patients in the DP and control groups, respectively (Fig. 2). Baseline characteristics were comparable between the groups (Table 1).
In the DP group, individualized PEEP was applied with median (1Q, 3Q) values of 7.0 (5.0, 8.0), 8.0 (8.0, 9.0), and 7.0 (6.0, 8.0) cmH2O during induction (supine position), main procedure (Trendelenburg position), and end of surgery (supine position), respectively.
The primary endpoint (total modified LUS at T3) was lower in the DP group (9.0 [8.0, 11.5]) than in the control group (11.0 [9.0, 13.0]), with a between-group median difference of -2.0 (95% confidence interval [CI] -3.0 to 0.0, P=0.032). This difference was primarily driven by B-line scores (8.0 [7.0, 11.0] vs. 11.0 [8.0, 13.0], P=0.036), whereas consolidation scores and significant atelectasis incidence remained similar (8.0 [6.0, 10.0] vs. 9.5 [7.8, 11.0]; P=0.094) between groups. In the recovery room (T4), the intergroup difference in modified LUS scores became more pronounced (9.0 [7.5, 12.0] vs. 13.5 [11.0, 15.0], P<0.001). Although the B-line scores differed between the groups at T3 and T4, the incidence of significant atelectasis remained high at both time points, with no between-group differences observed (T3: 28/31 (90.6%) vs. 30/32 (93.8%), P>0.999; T4: 29/31 (93.5%) vs. 32/32 (100.0%), P=0.458) (Table 2).
During Trendelenburg positioning, mean DP was significantly lower in the DP group than in the control group (19.0 ± 3.4 vs. 21.3 ± 4.7 cmH2O, P=0.035). No significant intergroup differences in DP were observed during supine positioning (Table 3).
Arterial oxygenation during Trendelenburg positioning showed significantly higher PaO2 in the DP group at T2 (154.6 ± 33.1 vs. 133.3 ± 34.7 mmHg, P=0.015) and T3 (160.0 ± 31.8 vs. 136.5 ± 32.8 mmHg, P=0.006). At T3, both the PaO2 and arterial oxygen saturation demonstrated significant intergroup differences. However, despite persistent differences in LUS scores, the PaO2 differences were resolved by T4 (Table 2).
Across T1-T3, the HR, MAP, cardiac output, and cardiac index were comparable between the groups, with no significant group × time interaction for any variable (all P>0.05). The proportion of patients receiving inotropes was similar between groups (Supplementary Table 1). PIP exceeded 35 cmH2O in five patients (16.1%) in the DP group versus eight patients (25%) in the control group at T2 (P=0.535, Fisher’s exact test). As per the protocol, these patients were switched from volume-controlled to pressure-controlled ventilation while maintaining the allocated PEEP strategy. This intervention successfully reduced peak pressures below 35 cmH2O within 5 min in all cases. No episodes of oxygen desaturation (SpO2 < 90%) occurred during the mode transition, and no patients violated the protocol (Supplementary Table 2).
Postoperative outcomes showed no significant differences between groups, including the incidence of pulmonary complications, postoperative fever, length of hospital stay, and 30-day readmission rates. No mortality was observed in any group during the 30-day follow-up period (Table 4).
DISCUSSION
In this randomized controlled trial, we found that DP-guided PEEP titration during RALP resulted in a reduced continuous burden of lung deaeration, as measured by modified LUS scores, and improved oxygenation compared with conventional fixed PEEP. The DP group showed significantly lower LUS scores at the end of the Trendelenburg position and significantly higher PaO2 values than the control group during Trendelenburg positioning. However, these intraoperative benefits did not translate into reduced postoperative pulmonary complications.
Individualized PEEP selection based on minimizing DP achieved a 2.3 cmH2O reduction in mean DP during Trendelenburg positioning. While this reduction appears modest, a previous meta-analysis has demonstrated that each 1 cmH2O increase in intraoperative DP is associated with 16% higher odds of postoperative pulmonary complications (odds ratio [OR] 1.16, 95% CI 1.13-1.19) [11]. Therefore, our observed 2.3 cmH2O reduction could potentially translate to approximately 37% lower odds of complications. The clinical relevance is further emphasized by comparing our baseline DPs (21.3 cmH2O) to those in non-RALP procedures. Pereira et al. [13] reported mean DPs of only 8.0 cmH2O with individualized PEEP and 11.6 cmH2O with fixed PEEP in conventional abdominal surgery [20]. Given the inherently elevated baseline pressures in RALP, even small reductions become clinically meaningful, which is consistent with the consensus recommendations for minimizing intraoperative DP.
PEEP optimization is particularly challenging in RALP, owing to competing physiological demands. Although PEEP helps maintain lung recruitment, excessive PEEP may compromise venous return in the context of elevated intra-abdominal pressure and inferior vena cava compression [21,22]. Our approach successfully balanced these competing needs and improved the respiratory parameters without compromising hemodynamics. In non-RALP abdominal surgery, low-tidal-volume ventilation reduces postoperative pulmonary complications and respiratory failure [6]. In contrast, increasing PEEP alone from 2 to 12 cmH2O with the same low tidal volume did not reduce these complications and was associated with more intraoperative hypotension and vasopressor use [21].
Previous investigations into the optimal PEEP during RALP have yielded valuable but limited insights. Earlier studies concluded that 7 cmH2O represented an optimal balance between improved oxygenation and acceptable PIP [23]. PIP exceeded 45 cmH2O in six patients in the PEEP 10 cmH2O group, whereas PEEP 7 cmH2O achieved similar oxygenation with PIP around 32-35 mmHg, a value comparable to our median PEEP during Trendelenburg. However, such a higher fixed PEEP without individualization is rarely used clinically because of PIP constraints. Our DP approach showed substantial variability in optimal PEEP (5-12 cmH2O), with 28% requiring ≤ 5 cmH2O, supporting individualization over fixed strategies. Lestar et al. [24] demonstrated that during RALP, the central venous pressure nearly tripled, mean pulmonary artery and wedge pressures roughly doubled, and MAP increased by approximately 35%, while lung compliance was halved, highlighting the need for individualized strategies that preserve both respiratory mechanics and cardiovascular stability.
Recent research supports our individualized approach. Large-scale randomized trials have established a foundation for protective ventilation strategies in surgical patients [25], while specific studies on laparoscopic procedures have demonstrated the benefits of recruitment maneuvers and individualized PEEP settings [26]. Using spiral CT, Andersson et al. [27] demonstrated that CO2 pneumoperitoneum increases dependent-lung atelectasis by approximately 66% and decreases overall lung gas volume with an approximately 2 cm cranial diaphragmatic shift, thereby reducing aerated lung capacity and supporting individualized PEEP titration, which may be particularly beneficial during RALP.
Notably, in our study, over 90% of patients in both groups developed significant atelectasis at T3 and T4, with no between-group differences. This apparent paradox—lower modified LUS scores in the DP group despite comparable atelectasis incidence—reflects our scoring methodology: the binary atelectasis definition captures any regional consolidation, whereas continuous LUS scores quantify the overall deaeration burden across all lung regions. The lower B-line scores of the DP group suggest that while focal consolidation was inevitable under extreme physiological conditions, DP-guided PEEP limited the extent of interstitial edema and alveolar deaeration in the remaining lung parenchyma.
Despite improvements in intraoperative parameters, we observed no significant reduction in postoperative pulmonary complications. Several factors could explain this outcome. First, although LUS scores provide sensitive detection of atelectatic changes, our sample size may have had insufficient power to detect differences in clinical outcomes. Second, the median Trendelenburg duration of 70 min may have been inadequate to induce significant pulmonary changes that persisted postoperatively, suggesting that greater benefits might be observed with longer procedures. Third, the high incidence of atelectasis in both groups at PACU suggests that intraoperative improvements were likely mitigated by postoperative exposure to high inspired oxygen fractions, for example brief FiO2 1.0 at emergence and simple-mask oxygen at 5 L/min in the PACU, which can promote absorption atelectasis, as described by Hedenstierna and Edmark [28].
These findings are consistent with those of recent large-scale studies that examined perioperative ventilation strategies. The IMPROVE trial [29] demonstrated that while intraoperative lung-protective ventilation can improve certain parameters, its translation to clinical outcomes depends on multiple factors, including patient characteristics and postoperative care protocols. Similarly, registry-based studies [30] have shown that the benefits of protective ventilation strategies may be most apparent in high-risk populations, while a meta-analysis [31] has confirmed that protective ventilation approaches can reduce complications when applied consistently across surgical populations.
Our study had several limitations. First, we excluded patients with pulmonary disease who might benefit the most from DP optimization. As established in a landmark ARDS study [8], patients with compromised lung function may derive greater benefits from this approach. Second, our titration protocol did not include post-titration recruitment maneuvers, typically brief sustained inflations (40 cmH2O for 10-15 s, repeated once) or stepwise increases targeting end-inspiratory pressures of 40-50 cmH2O with immediate decremental PEEP stabilization. Consequently, its effectiveness might have been limited. Third, although we implemented measures to minimize LUS score assessment variability, the ultrasound interpretation remained somewhat operator-dependent. Fourth, the absence of significant differences in postoperative pulmonary complications may reflect limited statistical power, as the study was powered to detect differences in LUS scores rather than clinical outcomes. Fifth, FiO2 1.0 during emergence, used per institutional protocol, likely induced absorption atelectasis in both groups. Although this did not affect the primary outcome assessed intraoperatively, it may have contributed to similar postoperative complication rates between the groups.
Future studies should examine this approach in patients with compromised lung function and procedures requiring longer Trendelenburg positioning. Comprehensive postoperative assessments, including serial imaging, would provide greater insights into benefit and durability. Additionally, these findings should be evaluated in high-risk populations and procedures involving prolonged Trendelenburg positioning. The primary outcomes should include clinically relevant postoperative complications and the persistence of atelectasis on imaging.
In conclusion, this randomized controlled trial demonstrated that DP-guided PEEP titration during RALP reduces the continuous burden of lung deaeration, as measured using modified lung ultrasound scores, without hemodynamic compromise. The intervention achieved a 2.3 cmH2O reduction in DP and approximately 25% lower LUS scores, primarily through reduced B-line formation. While these intraoperative benefits did not translate to reduced postoperative pulmonary complications in our relatively healthy cohort (ASA I-II), this likely reflects the multifactorial nature of postoperative respiratory outcomes. Future studies should focus on high-risk populations where the physiological advantages of individualized ventilation may have a greater clinical impact.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Supplement1
