Body habitus, pneumoperitoneum, and Trendelenburg positioning may each independently impair lung mechanics during robotic laparoscopic surgery. This study hypothesized that increasing body mass index is associated with more mechanical strain and alveolar collapse, and these impairments are exacerbated by pneumoperitoneum and Trendelenburg positioning.
This cross-sectional study measured respiratory flow, airway pressures, and esophageal pressures in 91 subjects with body mass index ranging from 18.3 to 60.6 kg/m2. Pulmonary mechanics were quantified at four stages: (1) supine and level after intubation, (2) with pneumoperitoneum, (3) in Trendelenburg docked with the surgical robot, and (4) level without pneumoperitoneum. Subjects were stratified into five body mass index categories (less than 25, 25 to 29.9, 30 to 34.9, 35 to 39.9, and 40 or higher), and respiratory mechanics were compared over surgical stages using generalized estimating equations. The optimal positive end-expiratory pressure settings needed to achieve positive end-expiratory transpulmonary pressures were calculated.
At baseline, transpulmonary driving pressures increased in each body mass index category (1.9 ± 0.5 cm H2O; mean difference ± SD; P < 0.006), and subjects with a body mass index of 40 or higher had decreased mean end-expiratory transpulmonary pressures compared with those with body mass index of less than 25 (–7.5 ± 6.3 vs. –1.3 ± 3.4 cm H2O; P < 0.001). Pneumoperitoneum and Trendelenburg each further elevated transpulmonary driving pressures (2.8 ± 0.7 and 4.7 ± 1.0 cm H2O, respectively; P < 0.001) and depressed end-expiratory transpulmonary pressures (–3.4 ± 1.3 and –4.5 ± 1.5 cm H2O, respectively; P < 0.001) compared with baseline. Optimal positive end-expiratory pressure was greater than set positive end-expiratory pressure in 79% of subjects at baseline, 88% with pneumoperitoneum, 95% in Trendelenburg, and ranged from 0 to 36.6 cm H2O depending on body mass index and surgical stage.
Increasing body mass index induces significant alterations in lung mechanics during robotic laparoscopic surgery, but there is a wide range in the degree of impairment. Positive end-expiratory pressure settings may need individualization based on body mass index and surgical conditions.
Prior studies suggest that intraoperative use of standardized tidal volumes based on ideal body weight are beneficial. However, attempts to define optimal positive end-expiratory pressure levels remain elusive given varying effects of body habitus and dynamic surgical conditions (pneumoperitoneum and Trendelenberg positioning).
Using esophageal manometry, the authors partitioned respiratory mechanical properties into lung and chest wall components in patients undergoing robotic laparoscopic surgery to assess the effects of obesity, pneumoperitoneum and Trendelenberg positioning on transpulmonary driving pressures and estimate optimal positive end-expiratory pressure for a given degree of obesity or surgical condition.
Obese patients demonstrated increased driving pressures and decreased mean end-expiratory transpulmonary pressures. Pneumoperitoneum and Trendelenberg position further accentuated these differences. The optimal positive end-expiratory pressure was greater than set positive end-expiratory pressure in most subjects at all stages, ranging from 0 to 36.6 cm H2O.
Intraoperative positive end-expiratory pressure settings should account for dynamic changes in transpulmonary driving and end-expiratory pressures related to these factors. How best to accomplish this clinically remains uncertain.
A key challenge of mechanical ventilation during surgery, beyond simply achieving adequate gas exchange, is optimizing ventilation parameters to avoid ventilator-induced lung injury and postoperative pulmonary complications. Intraoperative lung protective ventilation strategies using standardized tidal volumes based on predicted body weight have proven beneficial, but attempts to standardize positive end-expiratory pressure (PEEP) settings have not robustly accounted for body habitus or dynamic surgical conditions.1–4 There is a growing recognition that PEEP settings may need to both be personalized and adjust to variability in surgical conditions and patient physique.2,5,6 Laparoscopic abdominal surgery in the Trendelenburg (head-down) position is an increasingly common surgical modality that presents a unique physiologic challenge to the pulmonary system.
Obesity is a global epidemic.7 The pathophysiological effects of obesity on pulmonary function confer a high risk of developing postoperative pulmonary complications after mechanical ventilation. Patients with obesity have preexisting pulmonary dysfunction arising from chronic low-grade inflammation, decreased functional residual capacity, persistent atelectasis, restrictive physiology, and altered airway architecture, all of which lead to changes in normal and intraoperative pulmonary mechanics.6,8–11 Common PEEP settings may be inadequate to minimize intraoperative pulmonary impairments in patients with obesity.2,6,12 Due to lower lung volumes from loss of functional residual capacity, atelectrauma from repetitive alveolar closure and recruitment may be more likely with obesity. Knowing the ideal PEEP for these patients, however, is complicated by the significant interpatient variation in body mass and surgical positioning that now confronts the anesthesiologist.12–15
To delineate the impact of body habitus, pneumoperitoneum, and surgical positioning on intraoperative pulmonary mechanics, we conducted an observational study of patients undergoing robotic-assisted laparoscopic abdominal surgery in the Trendelenburg position. Using esophageal manometry, we partitioned the mechanical properties of the respiratory system into its lung and chest wall components and evaluated the effects of pneumoperitoneum, surgical position, and body mass index on transpulmonary pressures, airway and transpulmonary driving pressures, and lung elastance. We hypothesized that increasing body mass index would be associated with evidence of increasing atelectasis, increased driving pressures, and elevated lung elastance and that these changes would be exacerbated by pneumoperitoneum and Trendelenburg positioning.16
Materials and Methods
The study procedures were approved by the University of Vermont Institutional Review Board (Burlington, Vermont), and all subjects gave written informed consent to participate. The study’s protocols have been deposited at ClinicalTrials.gov under accession number NCT04329910.
We conducted a cross-sectional study of pulmonary mechanics in patients undergoing robotic-assisted laparoscopic abdominal surgery at the University of Vermont Medical Center between November 2017 and July 2019. Inclusion criteria were presentation for robotic-assisted laparoscopic abdominal surgery in the supine position, age ≥ 18 yr, and ability to provide informed consent. Exclusion criteria included intrinsic lung disease, smoking history of 20 pack/year or more, reactive airways disease, home oxygen requirement, inability to provide informed consent, emergent surgery, or esophageal pathology (i.e., strictures, varices, history of esophageal dilatation or surgery). All eligible subjects were approached for inclusion, and no decisions for recruitment were made based on body habitus.
Surgical maneuvers and anesthetic management were carried out according to standard institutional practice. Subjects underwent induction and maintenance of anesthesia at the discretion of clinical anesthetists, all of whom were blinded to the study data. All subjects received neuromuscular blockade with rocuronium or vecuronium for the entirety of the surgical procedure. Mechanical ventilation using volume control mode with set tidal volumes of 6 to 8 ml/kg of ideal body weight, and PEEP of 5 cm H2O or higher was recommended, but the discrete settings were chosen by the clinical anesthetist. Inspired oxygen fraction and respiratory rates were titrated to maintain oxygen saturation of more than 90% and end-tidal CO2 between 35 and 45 mmHg.
A pneumotachometer (Hamilton Medical, USA) was inserted at the Y-piece of the breathing circuit. After induction and endotracheal intubation, an orogastric tube was passed into the stomach, suctioned, and removed. An esophageal balloon catheter (Cooper Surgical, USA) was then placed into the midesophagus using established methods.17,18 Briefly, the balloon was first passed into the stomach, confirmed by positive pressure deflection with abdominal manipulation, then it was retracted into the esophagus to the point of maximal cardiac artifact, finally it was withdrawn 8 cm proximally to position it in the midesophagus. The balloon was deflated and reinflated every 30 min to prevent artifact from gas transfer or temperature effects. Airway pressure, esophageal pressure, and tracheal gas flow (V) were recorded continuously throughout the surgical procedure using a custom pressure transducer array (Silicon Microstructures, Inc., USA) and WinDaq engineering software (DATAQ Inc., USA).
Age, sex, height, and surgical indication were obtained from the medical record. The subjects were weighed before surgery by study personnel. Anesthesia times, intraoperative ventilator settings, gas analysis, pulse oximetry, fluid balance, hemodynamics, and train-of-four data were collected from the anesthesia record. Intraoperative surgical events including surgical duration, trocar placements, insufflation times and pressures, bed angles, blood loss, and specimen removal times were recorded by study personnel during the procedure.
Quantification of Pulmonary Mechanics
During the surgical procedure, 2- to 3-min segments of stable pressure and flow signals were extracted under four specific conditions: (1) Baseline: while supine and level, immediately after intubation; (2) Pneumoperitoneum: after abdominal insufflation; (3) Trendelenburg: steep Trendelenburg positioning with pneumoperitoneum and docked with the surgical robot (Da Vinci, Intuitive Surgical, USA); and (4) Desufflation: bed returned to level (0°) with the pneumoperitoneum released but still with neuromuscular blockade. A condition was excluded from analysis if there were not at least 2 min of artifact-free data available or if the esophageal pressure signal became dampened, suggesting a migration of the esophageal balloon into the stomach (table 1 for number of data points at each stage). Mechanical parameters were derived from the recorded signals as follows. First, a tidal volume (VT) relative to functional residual capacity was calculated by numerical integration of V with respect to time (t). Transpulmonary pressure was then calculated as follows:
Next, the pressure and flow data for each individual breath from the extracted segments were ensemble-averaged to produce a single breath for each signal with reduced noise. The following equation of motion was then fit to the averaged signals using least squares,
where elastance and resistance of the respiratory system, chest wall, and lung are calculated from the airway, esophageal, and transpulmonary pressure signals, respectively. End-expiratory airway pressure corresponds to the level of PEEP set on the ventilator together with any intrinsic PEEP that may have been present. End-expiratory transpulmonary pressure corresponds to the difference in end-expiratory airway and esophageal pressures (Eq. 1). When end-expiratory transpulmonary pressure is negative, distal or dependent alveoli are likely closed.
Driving pressures were calculated from airway, esophageal, and transpulmonary pressures as follows:
Finally, we estimated optimal PEEP for each subject at each surgical stage as follows,
where Set PEEP is the ventilator setting. When end-expiratory transpulmonary pressure is negative, the PEEP set on the ventilator would need to increase to optimal PEEP to make the end-expiratory transpulmonary pressure at least 0 cm H2O.
Initial sample size calculations were made assuming a between-factors repeated measures ANOVA with four repeated measurements in five groups, 50% correlation between measurements, and α of 0.05 at 80% power using G*Power 18.104.22.168 (Universitat Kiel, Germany). An estimated total sample size of 90 subjects was needed to measure a change in pressure of 1.5 ± 5.0 cm H2O (a 30% effect size) between groups. This effect size was established a priori during the study design. The a priori analysis plan was to examine differences in pulmonary mechanical parameters between body mass index categories and surgical conditions. The primary parameter of interest was transpulmonary pressure, but all respiratory parameters were examined.
Subjects were categorized according to their body mass index: lean (body mass index of less than 25 kg/m2), overweight (body mass index of 25 to 29.9 kg/m2), class I obesity (body mass index of 30 to 34.9 kg/m2), class II obesity (body mass index of 35 to 39.9 kg/m2), or class III obesity (body mass index of 40 kg/m2 or higher). Descriptive statistics were used to compare demographic, anthropometric, and perioperative characteristics. Categorical data are presented as counts or proportions. All continuous data were tested for normality by Shapiro–Wilks test and presented as means ± SD or median [interquartile range], as appropriate. All tests were two-tailed, and a P value of < 0.05 was considered significant, unless otherwise noted.
We modeled pulmonary mechanics data with linear generalized estimating equations using robust variance estimation to account for repeated measurements. Missing data were treated as missing at random. Interactions between body mass index grouping and position for each parameter were modeled. We made 36 pairwise comparisons of the marginal estimates and calculated the appropriate Bonferroni correction (0.05/36 = 0.0014). Thus, a multiple testing threshold of P ≤ 0.001 for the unadjusted P values was considered significant for these interactions. Secondary analyses of these data were conducted adding train-of-four count of 1 or lower and 2 or higher as a dichotomous variable; a P value < 0.05 was considered a significant contribution to the model. All analyses were carried out in STATA 15.1 (STATA Corp, USA).
We recruited and enrolled 99 subjects. Eight subjects were ultimately excluded from analysis either because the surgeon requested an orogastric tube for the entire surgical case (n = 4), the esophageal balloon was misplaced into the trachea (n = 1), the pressure transducer array malfunctioned (n = 2), or intraoperative bronchospasm required albuterol treatment (n = 1). The demographic and perioperative data for the remaining 91 subjects categorized by body mass index are presented in table 2. The ratio of males to females was similar among body mass index groupings, but there were more males in the overweight group. Surgical type mirrored the sex distribution. American Society of Anesthesiologists Physical Status classification increased appropriately with body mass index. Nine subjects (10%) were current smokers, 29 (32%) were former smokers, and 52 (58%) had never smoked. No subjects with obesity were current smokers. Obstructive sleep apnea was present in 10 subjects, 7 of whom had a body mass index of 35 kg/m2 or more.
Surgical Case Data
Surgical cases included 36 prostatectomies, 1 cystectomy, and 54 hysterectomies with or without salpingo-oophorectomy. Surgical duration averaged 199 ± 62 min with a mean anesthesia time of 235 ± 63 min; neither of these differed across body mass index categories (P = 0.422 and 0.306, respectively; table 2). An average of 1.9 ± 0.6 l of crystalloid were administered, with no differences among body mass index groups (P = 0.695). There were statistically different amounts of estimated blood loss and urine output among body mass index groupings (P = 0.041 and 0.048, respectively). However, these were clinically insignificant differences. Median pneumoperitoneum pressure was 15 mmHg [12 to 16], and slightly lower pressures were used in higher body mass index categories (P = 0.024). Bed angles in Trendelenburg ranged from 11 to 36°, with a median of 30° [23 to 31°] and did not differ between body mass index categories (P = 0.209; Supplemental Digital Content, table 1, http://links.lww.com/ALN/C424). Bed angle at all other surgical stages was 0° (level).
Baseline Pulmonary Mechanics
There were 89 extractable data points in the Baseline position (supine after intubation; table 1). Volume control was used for 88 subjects, pressure control was used for 1 subject, and tidal volumes averaged 6.8 ± 0.9 ml/kg ideal body weight (Supplemental Digital Content, tables 2 and 3, http://links.lww.com/ALN/C424). The fraction of inspired oxygen averaged 59 ± 14% (Supplemental Digital Content, table 4, http://links.lww.com/ALN/C424). At Baseline, airway plateau pressure significantly increased in each body mass index category compared with lean, and airway driving pressure was elevated in subjects with a body mass index of 30 or higher compared with lean (fig. 1, A and B, and table 3). Similarly, esophageal plateau and end-expiratory pressures were significantly increased in subjects with body mass index of 30 or higher compared with lean subjects (fig. 2, A and C; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). After intubation, 79% of subjects had negative end-expiratory and positive plateau transpulmonary pressures (fig. 3 and table 4; Supplemental Digital Content, table 6, http://links.lww.com/ALN/C424). Transpulmonary driving pressure increased an average of 1.9 ± 0.5 cm H2O with each increase in body mass index category compared with lean subjects (P < 0.006). At Baseline, more driving pressure was partitioned to the lungs of subjects with body mass index of 30 or higher compared with lean subjects (80 ± 13% vs. 62 ± 19%, respectively; Supplemental Digital Content, table 7, http://links.lww.com/ALN/C424). There were no differences in transpulmonary plateau pressures among body mass index groups (fig. 3A and table 4). In subjects with body mass index of 30 or higher, lung elastance was increased compared with lean subjects (fig. 4C and table 4), and in subjects with body mass index of 35 or higher, lung resistance and respiratory system elastance and resistance were increased compared with lean subjects (figs. 4A and 5, A and C, and tables 3 and 4).
Effect of Pneumoperitoneum on Pulmonary Mechanics
There were 82 extractable data points in the Pneumoperitoneum position (table 2), and all subjects received volume control ventilation with tidal volumes 6.8 ± 0.9 ml/kg ideal body weight (Supplemental Digital Content, tables 2 and 3, http://links.lww.com/ALN/C424). The fraction of inspired oxygen averaged 54 ± 13% (Supplemental Digital Content, table 4, http://links.lww.com/ALN/C424). With the pneumoperitoneum, airway plateau and driving pressures increased in all subjects compared with Baseline (fig. 1, A and B, and table 3). Although airway plateau pressure was increased in each body mass index category compared with lean, driving pressure was increased in subjects with a body mass index of 35 or higher (fig. 1, A and B, and table 3). With the pneumoperitoneum, esophageal plateau and driving pressure increased in all subjects compared with Baseline, whereas end-expiratory esophageal pressure was increased in lean subjects and those with a body mass index of 35 or higher (fig. 2; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). After establishment of pneumoperitoneum, 88% of subjects had negative end-expiratory and positive plateau transpulmonary pressures (Supplemental Digital Content, table 6, http://links.lww.com/ALN/C424). Although transpulmonary plateau pressure was not changed compared with Baseline, transpulmonary driving pressure was increased by an average of 2.8 ± 0.7 cm H2O in essentially all subjects compared with Baseline, although the increase in subjects with body mass index 35 to 39.9 was not significant using the multiple testing threshold (P = 0.004; fig. 3 and table 4). End-expiratory transpulmonary pressure was decreased by an average of –3.4 ± 1.3 cm H2O in all subjects but did not reach significance in subjects with body mass index of 25 to 29.9 (P = 0.002). With pneumoperitoneum, transpulmonary driving pressure was increased, and end-expiratory transpulmonary pressure was decreased in subjects with a body mass index of 35 or higher compared with lean (fig. 3 and table 4). Less driving pressure partitioned to the lung with pneumoperitoneum in all subjects (Supplemental Digital Content, table 7, http://links.lww.com/ALN/C424). With pneumoperitoneum, subjects with a body mass index of 40 or higher had more driving pressure partitioned to the lung compared with lean subjects (Supplemental Digital Content, table 7, http://links.lww.com/ALN/C424). All subjects had increased respiratory system elastance and resistance, chest wall elastance, and lung resistance with pneumoperitoneum compared with Baseline (figs. 4, A and B and 5, A and C, and tables 3 and 4; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). Subjects with a body mass index of 40 or higher had increased elastance and resistance in both respiratory system and lung compared with lean subjects (figs. 4, A and C, and 5, A and C; tables 3 and 4).
Effect of Trendelenburg with Pneumoperitoneum and Surgical Robot Docking on Pulmonary Mechanics
There were 88 extractable data points in the Trendelenburg position with pneumoperitoneum and docked with the surgical robot. At this stage, 84 subjects received volume control ventilation, 4 received pressure control, and tidal volumes averaged 6.8 ± 0.8 ml/kg ideal body weight (Supplemental Digital Content, tables 2 and 3, http://links.lww.com/ALN/C424). The fraction of inspired oxygen averaged 55 ± 14% (Supplemental Digital Content, table 4, http://links.lww.com/ALN/C424). In this position, airway plateau and driving pressure were increased in all subjects compared with Baseline and to Pneumoperitoneum (fig. 1, A and B, and table 3). Airway plateau and driving pressure were increased in each body mass index category compared with lean. Esophageal pressures were increased in all subjects compared with Baseline (fig. 2; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). In Trendelenburg, esophageal plateau pressure was increased in all subjects compared with Pneumoperitoneum, but this was not significant in subjects with a body mass index of 25 to 29.9 or a body mass index of 40 or higher using the multiple testing threshold (P = 0.002 and 0.004, respectively). Similarly, esophageal driving pressure was increased in all subjects compared with Pneumoperitoneum, but this was not significant in subjects with body mass index 25 to 29.9 using the multiple testing threshold (P = 0.033). In Trendelenburg, end-expiratory esophageal pressure was significantly elevated in all subjects compared with Baseline, but only in subjects with a body mass index of 30 to 34.9 compared with Pneumoperitoneum (fig. 2; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). In this position, esophageal plateau and end-expiratory pressures were elevated in subjects with a body mass index of 30 or higher compared with lean, whereas esophageal driving pressure did not differ between body mass index categories. Negative end-expiratory and positive plateau transpulmonary pressures were observed in 95% of subjects in Trendelenburg (Supplemental Digital Content, table 6, http://links.lww.com/ALN/C424). Transpulmonary plateau pressure was not increased in any subjects compared with Baseline or Pneumoperitoneum (fig. 3A and table 4). Compared with Baseline, transpulmonary driving pressure was increased by an average of 4.7 ± 1.0 cm H2O in all subjects in Trendelenburg but was not further increased compared with Pneumoperitoneum (fig. 3B and table 4). Similarly, end-expiratory transpulmonary pressure was decreased by an average of –4.5 ± 1.5 cm H2O in all subjects compared with Baseline (except subjects with body mass index 25 to 29.9; P = 0.002) but did not differ compared with Pneumoperitoneum (fig. 3C and table 4). In Trendelenburg, more driving pressure partitioned to the chest wall in all subjects compared with Baseline but did not differ from Pneumoperitoneum (Supplemental Digital Content, table 7, http://links.lww.com/ALN/C424). Coordinately, less pressure partitioned to the lungs in all subjects compared with Baseline. There was no significant effect of body mass index on driving pressures partitioned to lungs, but subjects with a body mass index 40 or higher had less driving pressure partitioned to the chest wall compared with subjects with a body mass index of less than 25 (Supplemental Digital Content, table 7, http://links.lww.com/ALN/C424). In Trendelenburg, elastances were increased in all subjects compared with Baseline (fig. 4 and tables 3 and 4; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). Compared with Pneumoperitoneum, respiratory system elastance was significantly increased in all subjects (fig. 4A and table 3), and chest wall elastance was increased in all subjects except those with a body mass index of 25 to 29.9 (P = 0.030; fig. 4B; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424). In Trendelenburg, lung elastance was increased in all subjects compared with Baseline but was not further increased over Pneumoperitoneum (fig. 4C and table 4). In this position, subjects with a body mass index of 35 or higher had increased respiratory system and lung elastance compared with lean subjects. Respiratory system and lung resistance were increased in all subjects compared with Baseline and only differed from Pneumoperitoneum in subjects with a body mass index of 30 to 34.9 (fig. 5, A and C, and tables 3 and 4). Chest wall resistance was decreased in subjects with a body mass index of less than 35 compared with Baseline and did not differ from Pneumoperitoneum (fig. 5B; Supplemental Digital Content, table 5, http://links.lww.com/ALN/C424).
Pulmonary Mechanics after Undocking, Leveling the Operating Room Table, and Abdominal Desufflation
There were 58 extractable data points after returning the subjects to a level position and releasing the pneumoperitoneum (Desufflation). At this stage, 44 subjects received volume control ventilation, 1 received pressure control, 13 received some form of spontaneous or triggered/synchronized ventilation, and tidal volumes averaged 6.9 ± 1.0 ml/kg ideal body weight (Supplemental Digital Content, tables 2 and 3, http://links.lww.com/ALN/C424). The fraction of inspired oxygen averaged 57 ± 18% (Supplemental Digital Content, table 4, http://links.lww.com/ALN/C424). At Desufflation, pulmonary mechanical parameters and driving pressure partitioning generally returned to values similar to those measured at Baseline. Subjects with body mass index of less than 25 had increased airway plateau and transpulmonary driving pressure compared with Baseline (figs. 1A and 3B and tables 3 and 4). At this stage, 80% of subjects had negative end-expiratory transpulmonary pressure (Supplemental Digital Content, table 6, http://links.lww.com/ALN/C424). The effects of increasing body mass index were similar to the patterns observed at Baseline.
Depth of Neuromuscular Blockade and Pulmonary Mechanics
Depth of neuromuscular blockade was recorded for the majority of subjects at each position (Supplemental Digital Content, table 8, http://links.lww.com/ALN/C424). At Baseline, 94% had 1 twitch or less on train-of-four testing. During Pneumoperitoneum 92% had 1 twitch or less, and in Trendelenburg 88% had 1 twitch or less. After Desufflation, 62% of subjects had 1 twitch or less. The addition of depth of neuromuscular blockade to the body mass index and position analyses did not significantly alter the modeling of any mechanical parameter (Supplemental Digital Content, table 9, http://links.lww.com/ALN/C424).
Optimal PEEP Estimation
Using the intraoperative pulmonary mechanics data, we estimated optimal PEEP settings for all subjects at each surgical stage. At each surgical stage, optimal PEEP settings were significantly increased for subjects with body mass index of 30 or higher compared with body mass index of less than 25 (fig. 6 and table 5). For all subjects with a body mass index of less than 30, optimal PEEP increased during Pneumoperitoneum compared with Baseline. After Trendelenburg positioning and robot docking, optimal PEEP was increased for all subjects compared with Baseline and increased in subjects with a body mass index of 30 to 39.9 compared with Pneumoperitoneum. At Desufflation, optimal PEEP estimates were not different than Baseline estimates.
This study quantified the impact of body habitus on intraoperative pulmonary mechanics during discrete stages of robotic laparoscopic abdominal surgery. Increasing body mass index was associated with higher airway plateau and driving pressures, elevated esophageal pressures, augmented lung elastance, and decreased end-expiratory transpulmonary pressures. These differences were exacerbated by pneumoperitoneum alone and by Trendelenburg with pneumoperitoneum while docked with the surgical robot. Increased airway pressures with pneumoperitoneum and during Trendelenburg largely partitioned to the chest wall causing the lung to experience lower driving pressures. Essentially all subjects had mechanical evidence of atelectasis with negative end-expiratory transpulmonary pressures, indicating that they might benefit from PEEP optimization. Higher body mass index was associated with more negative end-expiratory transpulmonary pressures and larger transpulmonary driving pressures, suggesting obese subjects were at higher risk for impaired gas exchange, atelectrauma, and intensified mechanical strain despite use of common lung protective ventilation strategies. Estimated optimal PEEP was higher than set PEEP in the majority of subjects in every position and body mass index category. To achieve open lung ventilation in robotic laparoscopic abdominal surgery, PEEP should increase with body mass index and adapt to discrete surgical stages. Individualized PEEP settings derived from previous studies of intraoperative ventilation in patients with obesity ranged from 13 to 25 cm H2O.12,13 A PEEP of 20 cm H2O or more may be required to maintain open lung ventilation during pneumoperitoneum or Trendelenburg positioning in patients with morbid obesity. In balance with known reductions in venous return, right ventricular afterload, and cerebral perfusion pressures and the potential for barotrauma and increased intracranial pressures, PEEP of more than 20 cm H2O should be used judiciously in the broader context of a patient’s anesthetic care.19 Notably, not all subjects with higher body mass index appear to need a higher PEEP to maintain positive end-expiratory transpulmonary pressures and optimize driving pressures, highlighting the need for individualization of ventilation parameters.
Our data are consistent with recent findings by Brandao et al.5 that increased airway pressures during pneumoperitoneum and Trendelenburg for robotic laparoscopic abdominal surgery largely partition to the chest wall. Similarly, we observed an increased contribution of the chest wall to total respiratory system elastance when subjects were docked in Trendelenburg. It is possible that the robotic frame adds rigidity to the chest wall, although we did not test this explicitly in this study. Our data support the concept that high airway pressures can be tolerated safely in robotic laparoscopic abdominal surgery because the increased forces are transmitted mostly to the chest wall. Obesity causes a progressively restrictive pulmonary physiology with increased esophageal plateau and end-expiratory pressures. Imagine increasing chest wall pressure like a brick on a balloon. Although the brick can increase in size, resulting in higher chest wall pressures, it does not change during the ventilation cycle, hence the lack of difference in chest wall elastance or esophageal driving pressures. More pressure is required to inflate the balloon, resulting in higher transpulmonary driving pressures with increasing body mass index. Interestingly, the pneumoperitoneum and Trendelenburg, although clearly adding to intrathoracic pressures, do not behave like a larger brick because less driving pressure is partitioned to the lungs at these stages. Abdominal distention and head-down positioning may alter the anatomic relationships of the diaphragm, rib cage, and lungs, but ultimately these conditions only partially and temporarily offset the effects of obesity on lung mechanics. In contrast to Brandao et al.,5 we only found persistently increased airway plateau and transpulmonary driving pressures at Desufflation in subjects with a body mass index of less than 25. Other subjects had no residual impairments in pulmonary mechanics at Desufflation.
Brandao et al.5 and a recent study by Shono et al.20 used electrical impedance tomography to evaluate regional aeration of the lungs during robotic laparoscopic surgery. Both found the center of aeration shifted ventrally in Trendelenburg, likely from a cephalad shift in the diaphragm and abdominal contents. They also found a predominant collapse of the ventral regions corresponding to apical subsegments while in Trendelenburg with partial collapse of dorsal (basilar) regions. The latter study showed application of 15 cm H2O PEEP improved aeration of dorsal regions compared with 5 cm H2O PEEP in lean subjects. Subjects still had negative end-expiratory transpulmonary pressures. These findings show increased PEEP improves alveolar recruitment during robotic laparoscopic abdominal surgery, supporting the conceptual foundation of open lung ventilation.
The differences in transpulmonary pressures in our study compared with Shono et al. are possibly related to esophageal balloon catheter position. Esophageal pressures vary along the esophagus, with the largest variations related to the lower esophagus in conjunction with cardiac artifact.21,22 A balloon catheter provides a regional average of pressures from surrounding structures. Changes in patient position and lung inflation or collapse can alter the intrathoracic pleural pressure gradient, measured esophageal pressures, and the subsequent calculation of pulmonary mechanics.23 Supine positioning results in a change in the measured esophageal pressures that can vary depending on catheter position.22,24 However, even with changes in the dependent and nondependent lung aeration, esophageal pressures are a reasonable estimate of pleural pressure in the midlung.25 In the midesophagus a change from sitting to supine results in a change in pressure of less than 3 cm H2O.24 If placed in the lower esophagus, the cephalad shift of the diaphragm and abdominal contents may lead to overestimating the pleural pressure because the balloon is closer to the diaphragm where the pressures are higher. We placed our catheters in the midesophagus in anticipation of this shift. If used to guide clinical therapy, a second measurement, such as assessment of regional aeration by electrical impedance tomography or change in total respiratory compliance, may be useful for confirming pulmonary recruitment. This is especially germane for patients with obesity because they are more likely to have large-airway closure in addition to atelectasis.
In a small case-control study, Grieco et al.6 demonstrated that airway closure in obese patients undergoing robotic-assisted laparoscopic abdominal surgery is associated with low end-expiratory transpulmonary pressure and a subsequent overestimation of pulmonary elastic forces. Citing an airway closure incidence of ~20% in obesity, they suggest that previous studies of pulmonary mechanics in obesity may not have considered airway closure as a contributor to high elastic forces in the lungs.6,26 This phenomenon is undoubtedly found in some of our study subjects and may contribute to the larger variability in pulmonary mechanical parameters in subjects with obesity. However, we have not explicitly measured airway opening pressures. Although there are multiple small differences in the ventilator setting used in their study compared with ours, we agree with their interpretations: airway closure physiology is prevalent in obesity and may result in damage to the more proximal bronchioles in addition to alveolar damage. This must be taken into account in developing future studies aimed at lung protective ventilation in patients with obesity.
There are some limitations to note with this study beyond those discussed above. Only adult, nonpregnant patients without lung disease undergoing robotic laparoscopic abdominal surgery were included, although some subjects were current or former smokers. We did not make any measurements of regional aeration by imaging or electrical impedance tomography, so our mechanical data can only infer changes relative to the position of the esophageal balloon. Because of variation in surgeon and anesthetist practice, the data from the Desufflation stage were not robustly captured in all subjects making definitive interpretation more difficult. This is a result of a deviation from planned study design (i.e., surgeon desufflating the abdomen while still in Trendelenburg) or subjects beginning to breathe as their neuromuscular blockade was reversed or wearing off, causing the mechanical modeling to fail. We were able to model pulmonary mechanics in 58 subjects (64%), although 21 of them had train-of-four 2 or higher despite adequate data extraction. As such, our data from the Desufflation stage should be interpreted carefully. However, inclusion of train-of-four data into the statistical modeling suggests that even with these limitations the depth of neuromuscular blockade did not adversely affect our analyses. This type of missing data can cause biased interpretation of the statistical results. Our missing data are best classified as missing at random, which satisfies the underlying assumptions of generalized estimating equation modeling. We attempted to minimize the potential bias by using postestimation pairwise comparisons with a stringent multiple testing correction. There are a number of interesting associations in the data that are apparent if this stringency is loosened. However, to retain the most parsimonious interpretation, we have not included unadjusted analyses in this study.
Altered intraoperative pulmonary mechanics may contribute to ventilator-induced lung injury and increased risk for postoperative pulmonary complications by exposing alveoli to overdistention, repetitive recruitment and collapse, or atelectasis. Identifying patients in whom impaired pulmonary mechanics leads to injury needs to be prioritized to utilize lung protective techniques in an evidence–based manner. Individualized protective ventilation for patients with obesity may require increased PEEP settings above what have been traditionally used. Development of lung protective ventilation strategies will require larger clinical trials to address the complex intraoperative pulmonary physiology of patients with obesity.
The authors acknowledge Dr. Stephen H. Loring, M.D., Ph.D., Beth Israel Deaconess Medical Center, Harvard Medical Center, Boston, Massachusetts, who taught us how to build and use our esophageal manometry sensor array; Dr. Charles G. Irvin, Ph.D., Larner College of Medicine, Burlington, Vermont, who taught us how to design and troubleshoot an esophageal manometry clinical protocol; and Bradley Holcomb, B.Sc., Respiratory Care Services and Pulmonary Function Lab, University of Vermont Medical Center, Burlington, Vermont, who gave us access to the respiratory care teaching center to test and calibrate our sensors.
Support was provided primarily from institutional and/or departmental sources. Ms. Booms received support from a Foundation for Anesthesia Education and Research Medical Student Anesthesia Research Fellowship.
Dr. Bates notes financial relationships with the American Physiological Society (Rockville, Maryland), Healthy Design LLC (Westminster, Colorado), and Oscillavent LLC (Iowa City, Iowa). The other authors declare no competing interests.