Effects of volume-targeted pressure-controlled inverse ratio ventilation on functional residual capacity and dead space in obese patients undergoing robot-assisted laparoscopic radical prostatectomy

Background The effect of inverse inspiration:expiration (I:E) ratio on functional residual capacity (FRC) during pneumoperitoneum is unclear. We hypothesised that volume-targeted pressure-controlled inverse ratio ventilation (vtPC-IRV) would increase FRC by increasing the level of auto-PEEP in low respiratory compliance situations. Methods During robot-assisted laparoscopic radical prostatectomy, 20 obese patients were sequentially ventilated with four different settings for 30 min in each setting: (1) control, I:E ratio of 1:2 and baseline airway pressure (BAP) of 5 cm H2O; (2) IRV2, I:E ratio of 2:1 and BAP off; (3) IRV3, I:E ratio of 3:1 and BAP off; and (4) IRV4, I:E ratio of 4:1 and BAP off. The changes in FRC were identified and compared among these settings. Results The FRC significantly increased as the I:E ratio increased. The FRC values expressed as median (inter-quartile range) during control, IRV2, IRV3, and IRV4 were 1149 (898–1386), 1485 (1018–1717), 1602 (1209–1775), and 1757 (1337–1955) ml, respectively. Auto-PEEP increased significantly as the I:E ratio increased and correlated with FRC (rho=0.303; P=0.006). Shunt and physiological dead space were significantly lower in all IRV groups than in the control group; however, there were no significant differences among the IRV groups. Conclusions vtPC-IRV with shortened expiratory time and increased auto-PEEP effectively increases FRC during robot-assisted laparoscopic radical prostatectomy in obese patients. FRC increases progressively as the I:E ratio increases from 1:2 to 4:1; however, an I:E ratio higher than 2:1 does not further improve the dead space. Clinical trial registration UMIN000038989.

ratio with short expiratory time on FRC during pneumoperitoneum in the Trendelenburg position. Thus, we aimed to evaluate the FRC with various I:E ratios of vtPC-IRV in obese patients undergoing robot-assisted laparoscopic radical prostatectomy.
We hypothesised that vtPC-IRV with a short expiratory time would increase FRC by increasing the level of auto-PEEP. Our study contributes to the practice of open-lung ventilation in situations with low static compliance.

Study design and patients
This single-centre prospective interventional study was conducted at the Mizonokuchi Hospital, Teikyo University School of Medicine, Kanagawa, Japan, and was approved by the Ethics Committee of the Teikyo University School of Medicine, Tokyo, Japan (chairperson and dean: M. Kawamura) on 1. Novel theory of volumetric capnography. The expired tidal volume on the Y axis is plotted against the partial pressure of expired CO 2 on the X axis to express the curve of 'Volume¼VCO 2 /FCO 2 '. Phase I represents CO 2 -free and pure dead space, phase II represents the transition between the airway and alveolar gas, and phase III represents the alveolar gas. Phase II ends at the inner square area maximum point. VD aw is determined by applying Fowler's equal area method (area A equals to B). The volume from the start of expiration to the partial pressure of the mixed expired CO 2 point on the Y axis is defined as respiratory dead space (VD resp ). Thus, alveolar dead space (VD alv ) is calculated as VD resp eVD aw and shunt dead space (VD shunt ) is calculated as VD phys eVD resp . VD phys , physiological dead space; VD aw , airway dead space; VD alv , alveolar dead space; VD shunt , shunt dead space; V TE , expired tidal volume; VCO 2 , expired tidal volume of CO 2 ; FCO 2 , fractional concentration of CO 2 [FCO 2 ¼PCO 2 (P B eP H2O ) À1 ¼PCO 2 (760e47) À1 ; P B , barometric pressure; P H2O , water vapor pressure at 37 C]; F E CO 2 , mixed expired FCO 2 ; F ET CO 2 , end-tidal FCO 2 ; F A CO 2 , alveolar FCO 2 ; and FaCO 2 , arterial FCO 2 .
Hospital Medical Information Network Clinical Trials Registry (UMIN000038989). Written informed consent was obtained from all participants. The study included patients with BMI >25 kg m À2 aged 40e75 yr with ASA physical status score 1e2, scheduled to undergo robot-assisted laparoscopic radical prostatectomy. The exclusion criteria were ASA physical status of 3e5, a history of asthma, pulmonary emphysema, pneumothorax, or lung surgery. The patients were enrolled between 15 June 2019 and 30 July 2020.

Anaesthesia protocol
Routine patient monitoring included ECG, pulse oximetry, noninvasive arterial blood pressure measurement, and inspired and expired CO 2 analysis. Continuous radial arterial pressure, cardiac index, and stroke volume variation (SVV) were monitored using a Vigileo system with a Flo-Trac sensor (Edwards Lifesciences, Irvine, CA, USA). Mainstream CO 2 and flow sensors were attached to the proximal end of the tracheal tube to enable volumetric capnography (Senko Medical Instrument Co. Ltd., Tokyo, Japan). Anaesthesia was induced by administering i.v. propofol 1e3 mg kg À1 and fentanyl 2e4 mg kg À1 . Tracheal intubation was performed after the administration of rocuronium 0.8e1.0 mg kg À1 . Anaesthesia was maintained by total intravenous administration of propofol 100e200 mg kg À1 min À1 and remifentanil 0.2e0.3 mg kg À1 min À1 . Intermittent intravenous injections of rocuronium 0.1e0.2 mg kg À1 and fentanyl 1e2 mg kg À1 were administered, as needed.

Interventions and ventilatory settings
After the induction of anaesthesia, we ventilated the patients' lungs using a ventilator (Engstr€ om Carestation™; Datex-Ohmeda, GE Healthcare, Helsinki, Finland) that measures FRC and auto-PEEP. After positioning in the 25e30 Trendelenburg position and CO 2 pneumoperitoneum at 12 mm Hg, the patients' lungs were sequentially ventilated using four different settings for 30 min each: control, IRV2, IRV3, and IRV4 settings. Each ventilator setting included the volume targeted pressure control (expressed as 'pressure control ventilationvolume guarantee' in Engstr€ om Carestation™) mode, in which the airway pressure was adjusted to achieve a target tidal volume. The target tidal volume was adjusted with the plateau pressures permitted to increase to an upper limit of 30 cm H 2 O. The initial fraction of inspiratory oxygen (F i O 2 ) was 0.5, and the ventilatory frequency was set at 12 bpm. The target tidal volume and ventilatory frequency remained constant throughout the study providing there were no respiratory complications. For the control setting, the I:E ratio was 1:2, and the baseline airway pressure (BAP; baseline at a set airway-pressure level, replacing a PEEP setting 13 ) was 5 cm H 2 O. In IRV2, the I:E ratio was 2:1 and BAP was 0 cm H 2 O (off). In IRV3, the I:E ratio was 3:1 and BAP was off. In IRV4, the I:E ratio was 4:1 and BAP was off.
Auto-PEEP was allowed to increase to 10 cm H 2 O, peripheral oxygen saturation was allowed to drop to 93%, whereas endtidal carbon dioxide (E T CO 2 ) was permitted to increase to 60 mm Hg. When any of these predetermined limits were exceeded, the study was halted, and the ventilator setting was changed to an I:E ratio of 1:1, increasing the ventilatory frequency and F i O 2 and increasing or decreasing the tidal volume.
The MAP was maintained >70 mm Hg using IV ephedrine (4e8 mg). An intravenous fluid challenge was provided with 10 ml kg À1 of Ringer's acetate solution or hydroxyethyl starch if the SVV exceeded 15%.

Outcome measures
The primary outcome measure was FRC measured using the oxygen wash-in and wash-out methods provided within the ventilator (Engstr€ om Carestation™). The International Organization for Standardization (ISO) redefined the baseline and PEEP terminology. 13 PEEP is the actual and measured value of respiratory pressure at the end of an expiratory phase. BAP is the quantity by which the BAP positively offsets from the ambient pressure, replacing a PEEP setting. Auto-PEEP is the portion of the stabilised airway pressure above that set for the endexpiratory pressure, at the end of an expiratory-hold procedure that temporarily occludes the airway in the absence of any respiratory activity. Total PEEP is the stabilised airway pressure Table 1 Interpretation of dead space components. V_ A /Q_, ventilation to perfusion ratio. at the end of an expiratory-hold procedure that temporarily occludes the airway in the absence of any respiratory activity. The ventilator measured auto-PEEP (expressed as 'PEEPi' in Engstr€ om Carestation™) and total PEEP were calculated as follows: total PEEP¼auto-PEEPþPEEP. Static compliance (C stat ) was calculated as C stat ¼V TI (P plat etotal PEEP) À1 , where V TI is the inspired tidal volume and P plat is the plateau pressure.
Physiological dead space (VD phys ) is a functional evaluation of the difference in the partial pressure of CO 2 between the arterial blood and mixed expired gas [VD phys ¼V TE ･(PaCO 2 eP E CO 2 )･PaCO 2 À1 ], where V TE is the expired tidal volume and PeCO 2 is the mixed expired gas partial pressure of CO 2 , representing the overall CO 2 elimination efficiency of the respiratory and circulatory dynamics. Using the volumetric capnography theory (see Supplementary data), VD phys was divided into airway dead space (VD aw ), alveolar dead space (VD alv ), and shunt dead space (VD shunt ) (VD phys ¼VD aw þVD alv þVD shunt ) (Fig 1). VD aw is a functional evaluation of the airway space volume, including the gas diffusion effect, representing extra-alveolar V_ A /Q_¼∞ mismatch. Respiratory dead space (VD resp ) is a functional evaluation of the difference in CO 2 partial pressure between the alveolar and mixed expired gas, representing V_ A /Q_>1 mismatch. VD alv , calculated as VD resp eVD aw , is a functional evaluation of the relative hyperinflation in the alveolar units, representing intraalveolar V_ A /Q_>1 mismatch. VD shunt , calculated as VD phys eV-D resp or VD phys eVD aw eVD alv , represented functional evaluation of relative hyperperfusion and the difference in CO 2 partial pressure between the pulmonary artery and mixed expired gas, representing V_ A /Q_<1 mismatch.
Each component had two different influential factors: the plateau-dependent dead space change caused by different ventilator settings and the plateau-independent dead space change caused by the time elapsed under the same ventilator settings. Plateau-independent VD aw change represented extra-alveolar V_ A /Q_¼∞ mismatch derived from airway space volume change. Plateau-dependent VD aw change represented extra-alveolar V_ A /Q_¼∞ mismatch caused by gas diffusion in the airway. Plateau-dependent VD alv change represented intra-alveolar V_ A /Q_>1 mismatch caused by hyperinflation with circulatory suppression. The plateau-independent VD alv change represented intra-alveolar V_ A /Q_¼∞ mismatch such as pulmonary infarction. Plateau-dependent VD shunt change represented intra-alveolar V_ A /Q_<1 mismatch caused by heterogeneous expansion disorders. Plateau-independent VD shunt change represented intra-alveolar V_ A /Q_¼0 mismatch caused by atelectasis and extra-alveolar V_ A /Q_¼0 mismatch caused by an extra-alveolar shunt (Table 1).
In this study, we did not explore the mechanism by which the plateau-dependent or independent factors affected each component of the dead space because the dead space components were analysed only at the end of each ventilator setting for a 30-min period but not evaluated over time.

Statistical analysis
The sample size was calculated based on previous data 14,15 using paired samples t-test as 20 subjects per group to detect the differences in the FRC with a power of 0.8 and a type I error rate of 0.05, based on an estimated difference of 0.66 of the parameter's estimated standard deviation (SD). Nonparametric data distribution was assumed owing to the small number of patients included in this study, and the descriptive parameters were generally expressed as median (inter-quartile range [IQR]).
The Friedman test followed by the Wilcoxon signed-rank test with Bonferroni correction were used to identify the changes and differences in respiratory and haemodynamic variables among the ventilator settings. Multiple regression analysis was used to identify the influence of various factors on the primary outcome measure. The dependent variable was FRC, and the independent variables were selected from patient characteristics and ventilator settings based on clinical prediction. We eliminated the calculated or estimated variables from the potential independent factors. The selected independent variables were BMI, FVC (% predicted), forced expiratory volume in 1 s (FEV 1 , % predicted), target tidal volume, BAP, and I:E ratio. Spearman's rank correlation was used to assess the relationship between the variables. All statistical analyses were performed using R© version 3.5.2 (R Foundation for Statistical Computing, Vienna, Austria). Statistical significance was set at P<0.05.

Patient characteristics
Out of the 99 patients screened, 73 patients did not meet the inclusion criteria and six patients refused participation. Thus, 20 patients were included in the analysis. No patient was lost to follow-up (Fig 2). The patient characteristics and respiratory measurements before intervention in the supine position without CO 2 pneumoperitoneum are summarised in Table 2. The patients' lungs were ventilated with no respiratory difficulties in the supine position without CO 2 pneumoperitoneum.

Effect of IRV
The ventilator settings and respiratory and haemodynamic variables during intervention under the Trendelenburg position and CO 2 pneumoperitoneum are shown in Table 3. The ventilatory frequency and target tidal volume did not change throughout the study in any patient. Static compliance was low, and there were no significant differences between the groups. Because of the low static compliance, the target tidal volume per kg body weight was 6.37 (5.97e6.64) ml kg À1 , with the plateau pressures' upper limit being 30 cm H 2 O. FRC and auto-PEEP increased significantly as the I:E ratio increased. For dead space, VD phys , VD shunt , and VD aw were significantly lower in all three IRVs than in the control; however, there were no significant differences among IRV2, IRV3, and IRV4. IRV resulted in lower cardiac index and higher SVV than the control.

Complications
There were no respiratory complications during the perioperative period with auto-PEEP, peripheral oxygen saturation, and E T CO 2 maintained within the predetermined limits in all patients. However, in one patient (a smoker with FEV 1 of 79%), auto-PEEP reached 9 cm H 2 O during the IRV4 settings with FRC of 1752 ml compared with 1157 ml at the control settings.

Discussion
The effect of an inverse I:E ratio on FRC during pneumoperitoneum remains unclear. We found that the larger the I:E ratio with vtPC-IRV, the higher was the FRC. Furthermore, the I:E ratio e but not BAP e was an independent predictor of FRC. FRC was correlated with auto-PEEP but not with PEEP or total PEEP. Tan and colleagues 16 reported that FRC was correlated with auto-PEEP during mechanical ventilation with a normal I:E ratio in patients with an exacerbation of chronic obstructive pulmonary disease. In contrast, auto-PEEP was generated by a short expiratory time with inverse I:E ratio in our study. Futier and colleagues 17 reported that a high BAP of 10 cm H 2 O with an I:E ratio of 1:2 effectively improved FRC and respiratory mechanics in non-obese patients. However, the effects of high BAP were diminished in patients with obesity during general anaesthesia with neuromuscular blocking agents. 17 In a pig model of intra-abdominal hypertension (low static compliance model), BAP of up to 15 cm H 2 O did not prevent FRC decreases. 18 Nestler and colleagues 19 reported that in individualised BAP titrated using electrical impedance tomography, a mean PEEP value of 18 cm H 2 O increased endexpiratory lung volume and improved PaO 2 /F I O 2 in obese patients during elective laparoscopic surgery.
In our study, patients with obesity must have had low static compliance during CO 2 pneumoperitoneum in the Trendelenburg position. Therefore, it was difficult to increase FRC because of the low levels of BAP with a long expiratory time (I:E ratio of 1:2), in the same way as previously reported. Thus, our results suggest that short expiratory time with IRV accompanied by increased auto-PEEP would effectively increase FRC under such conditions with a low static compliance. However, although increased FRC may prevent atelectasis, it has a potential risk of hyperinflation of the normal alveoli and circulatory compromise.
Atelectasis may obstruct alveolar expansion, inducing relative hyperperfusion (V_ A /Q_<1) in the impaired alveolar unit and may redistribute the inflated gas toward the healthy alveolar unit excessively, causing relative hyperinflation (V_ A / Q_>1). Similar to lung recruitment manoeuvres, PC-IRV can decrease pulmonary blood flow to the ventilated alveoli, inducing relative hyperinflation (V_ A /Q_>1). Moreover, it is likely to redistribute the pulmonary blood flow toward the nonventilated alveoli, promoting relative hyperperfusion (V_ A / Q_<1).
Our analysis of dead space helped detect the heterogeneous distribution of relative hyperinflation and hyperperfusion presenting as increased VD alv and increased VD shunt , respectively. Thus, VD alv and VD shunt improve significantly when the open-lung approach ventilation strategy improves the heterogeneous distribution of alveolar units with relative hyperinflation and hyperperfusion, resulting in minimised VD phys . 10,12,19e21  The reduction of VD phys during vtPC-IRV results from decreases in VD shunt and VD aw accompanied by unchanged VD alv . The prolonged plateau time likely afforded sufficient inflation of alveoli with compromised expansion. In addition, the short expiratory time with increased auto-PEEP led to an increased FRC, and both resulted in decreased VD shunt . vtPC-IRV slightly reduced VD aw , suggesting that prolonged plateau time enhanced diffusion of gas within the airway space. A prolonged plateau time may enhance gas diffusion from the pulmonary artery to the alveoli. Excessive plateau pressure or circulatory suppression might increase intra-alveolar V_ A /Q_>1 mismatch. Thus, PC-IRV can potentially increase VD alv under high-plateau pressure conditions with circulatory suppression. In this study, vtPC-IRV induced significant decreases in the mean arterial blood pressure and cardiac index, with increased SVV. However, the SVV and cardiac index were controlled within the normal range. No significant change in VD alv by vtPC-IRV indicated successful management of vtPC-IRV with moderate plateau pressure and circulatory dynamics.
VD phys in all IRV settings were significantly lower than those in the control setting, suggesting that vtPC-IRV would be an open-lung approach ventilation strategy suitable for a low static compliance situation. However, there were no significant differences in VD phys or VD shunts among IRV2, IRV3, and IRV4, which indicates that with respect to dead space, an I:E ratio of 3:1e4:1 has a smaller benefit than an I:E ratio of 2:1, although larger I:E ratios induced larger increases in FRC.
Earlier studies showed that the higher the BAP, the higher was the FRC 21,22 ; however, a high BAP of 14e20 cm H 2 O increased dead space under normal static compliance. 14,23, 24 Fengmei and colleagues 25 concluded that the optimal BAP was 12 cm H 2 O because it induced the highest compliance in conjunction with the lowest dead space, indicating a maximum amount of effectively expanded alveoli. These studies were usually conducted at a high BAP, with an I:E ratio of 1:2. A long expiratory time (I:E ratio of 1:2) may contribute to reduced FRC and induce atelectasis. Furthermore, increasing FRC by BAP under long expiration time (I:E ratio of 1:2) with low respiratory compliance is difficult. IRV with a short expiratory time and high auto-PEEP even with low respiratory compliance effectively increases FRC. However, IRV is not commonly used or safe. It needs an anaesthesia ventilator equipped with an expiratory flow-time wave monitor to avoid lung hyperinflation, adequate haemodynamic management, moderate muscle relaxation, and the expertise of an anaesthesiologist. Therefore, IRV with an I:E ratio of 3:1e4:1 is not recommended. However, a short expiratory time with IRV would enhance the BAP effect of increasing FRC in situations with low static compliance. We believe that vtPC-IRV with an I:E ratio of 1:1e2:1 and a BAP of 5e7 cm H 2 O is a safe and practical open-lung approach ventilation strategy in situations with low static compliance, which may provide sufficient total PEEP and appropriate inflation of the slow opening alveolar components.

Limitations
In this study, the ventilator settings were sequentially changed at intervals of 30 min, and the measurements were obtained at the end of each setting. As the sequence was not randomised, the results might be influenced by timedependent changes or trends in respiratory status during the 2-h intervention period. We cannot rule out this possibility. However, 12 VD phys and VD shunt increased within 2 h intraoperatively, and atelectasis probably developed in ventilator conditions with an I:E ratio of 1:2 or 2:1, suggesting that FRC may decrease, and the dead space components may increase with time. Therefore, it is conceivable that our findings of IRV acutely increasing FRC and decreasing VD phys and VD shunt were attributable to ventilator settings, rather than timedependent changes. Atelectasis likely develops heterogeneously within hours; however, we only measured acute changes in the respiratory and circulatory status in each 30-min period of ventilator setting. Changes over a longer period or the actual degree of atelectasis were not evaluated. Further studies evaluating the development of atelectasis using electrical impedance tomography, a useful tool for evaluating the heterogeneous development of atelectasis over time, 26,27 are warranted to establish a safe and practical open-lung approach ventilation strategy.
Conclusions vtPC-IRV with short expiratory time and high auto-PEEP effectively increases FRC during robot-assisted laparoscopic radical prostatectomy in patients with obesity. The higher FRC is observed with larger I:E ratio in vtPC-IRV. Furthermore, vtPC-IRV decreases the shunt and physiological dead space, indicating that vtPC-IRV can be used as an open-lung approach ventilation strategy in cases of low respiratory compliance. However, an I:E ratio of more than 2:1 does not improve dead space further, suggesting that using an I:E ratio of more than 2:1 is not required under these conditions.