Lung protective ventilation during pulmonary resection in children: a prospective, single-centre, randomised controlled trial

British Journal of Anaesthesia, xxx (xxx): xxx (xxxx) doi: 10.1016/j.bja.2019.02.013 Advance Access Publication Date: xxx Clinical Investigation

CLINICAL INVESTIGATION

Lung protective ventilation during pulmonary resection in children: a prospective, single-centre, randomised controlled trial Ji-Hyun Leey, Jung-il Baey, Young-Eun Jang, Eun-Hee Kim, Hee-Soo Kim and Jin-Tae Kim* Department of Anaesthesiology and Pain Medicine, Seoul National University Hospital, Seoul National University College of Medicine, Seoul, Republic of Korea *Corresponding author. E-mail: [email protected] y

These two authors contributed equally as first authors of this manuscript.

Abstract Background: Perioperative ventilatory strategies for lung protection in children are underexplored. This study evaluated the effects of lung protective ventilation (LPV) on postoperative clinical outcomes in children requiring one-lung ventilation (OLV) for pulmonary resection. Methods: Children age 5 yr scheduled for video-assisted thoracoscopic lung lobectomy or segmentectomy were randomly assigned to LPV or control ventilation. For LPV, tidal volume (VT) was 6 ml kg1 during two-lung ventilation (TLV(VT)), 4 ml kg1 during OLV, with 6 cm H2O PEEP maintained throughout. In the control group, TLV(VT) was 10 ml kg1, 8 ml kg1 during OLV, but without PEEP. The primary outcome was the incidence of pulmonary complications within 72 h after operation. Secondary outcomes included intraoperative desaturation, arterial oxygen partial pressure/inspiratory fraction of oxygen (P/F) ratio >40 kPa, and development of consolidation and B-lines (assessed by lung ultrasound at the end of surgery, by an investigator masked to group allocation). Odds ratio (OR) with 95% confidence intervals are reported. Results: Overall, 19/110 (17.3%) children sustained pulmonary complications after surgery. LPV reduced pulmonary complications (5/55; 9.1%), compared with 14/55 (25.5%) children sustaining complications in the control group (OR¼0.29 [0.10e0.88]; P¼0.02). Masked ultrasound assessment showed less consolidation, and fewer B-lines, after LPV (P<0.001). Intraoperative desaturation was more common in control mode (eight/55; 14.5%), compared with 1/55 (1.8%) after LPV (OR¼9.2 [1.1e76]; P¼0.015). LPV maintained (P/F) ratio >40 more frequently (53/55; 96.4%) than control-mode (45/55; 81.8%) ventilation (OR¼5.9 [1.2e28.3%]; P<0.01). Conclusions: Lung protective ventilation decreased postoperative pulmonary complications compared with conventional ventilation in children requiring one-lung ventilation for pulmonary resection. Clinical trial registration: NCT02680925. Keywords: anaesthesia, general; mechanical ventilation; paediatrics; postoperative complications; pulmonary atelectasis

Editorial decision: 05 Feb 2019; Accepted: 5 February 2019 © 2019 British Journal of Anaesthesia. Published by Elsevier Ltd. All rights reserved. For Permissions, please email: [email protected]

1

2

-

Lee et al.

Editor’s key points  Studies of perioperative lung protective ventilation strategies in children are scarce.  This landmark, single-centre study assessed the effect of a protective ventilatory strategy on perioperative outcomes in children < 5 yr old undergoing pulmonary resection and requiring one-lung ventilation.  Lung protective ventilation reduced postoperative pulmonary complications (number needed to treat six children (95% confidence interval, 3e39; P¼0.03).

each surgery, an anaesthesiologist met with the child’s parents, explained the study protocol, and obtained written informed consent from the parents.

Inclusion criteria Children younger than 5 yr with an ASA physical status of 1e2 undergoing video-assisted thoracoscopic lung lobectomy or segmentectomy because of congenital cystic adenomatoid malformation from February 2016 to December 2017 were enrolled.

Exclusion criteria Postoperative lung complications occur in around 5% of patients after general anaesthesia and are associated with increased mortality and morbidity.1,2 Mechanical ventilation strategies during general anaesthesia can limit pulmonary morbidity.3 Lung protective ventilation (LPV) strategies minimise ventilator-induced lung injuries and improve clinical outcomes in patients with acute respiratory distress syndrome or acute lung injury.4 LPV includes the use of low tidal volumes, PEEP, and recruitment manoeuvres. Perioperatively, LPV improves respiratory function and reduces hospital stay in adults undergoing laparotomy.5,6 In contrast to adults, the impact of intraoperative LPV on clinical outcomes after paediatric surgery is unclear. Studies in LPV are limited to neonates or preterm infants who require mechanical ventilation in the ICU because of respiratory distress syndrome.7 Because of smaller functional residual capacities, larger closing volumes, and greater chest wall compliance compared with adults, pulmonary atelectasis occurs more frequently in children after general anaesthesia. Therefore, young children are vulnerable to ventilatorinduced lung injury. Because of technological advances in ventilation, LPV for paediatric patients has become more frequently implemented in the operating room.8 However, it remains unclear who will benefit from LPV among paediatric surgical patients. It was proposed that an optimal mechanical ventilation strategy during thoracic surgery requiring one-lung ventilation (OLV) was important as greater lung damage, as a result of surgical manipulation, is expected.9 We hypothesised that the LPV during pulmonary resection could reduce postoperative pulmonary complications in children. This single-centre, randomised controlled study evaluated the effect of LPV on postoperative clinical outcomes and intraoperative oxygenation in young children (age <5 yr) undergoing pulmonary resection requiring OLV.

Methods Study population This prospective randomised, controlled, parallel-designed trial was approved by the Institutional Review Board of Seoul National University Hospital, Seoul, Korea (H1511-124-725; date of approval: December 12, 2015) and registered at https:// clinicaltrials.gov (number: NCT02680925; principal investigator: KJT; date of registration: February 6, 2016). The study was performed according to the ethical standards set by the 1964 Declaration of Helsinki and its later amendments. Before

Patients with a history of asthma, active upper airway or systemic infection, interstitial lung disease, and cardiac problems and who were scheduled for open thoracotomy were excluded.

Group allocation Patients were randomly assigned to one of two groupsdthe LPV group or the control groupdafter a simple randomisation procedure (computerised random number; https://www. randomizer.org). The allocation ratio was 1:1, and an anaesthetic nurse, not associated with the study, performed a random allocation sequence by preparing coded and sealed opaque envelopes for allocation concealment. The patients, guardians, surgeon, nurse, and outcome assessors who performed the lung ultrasonography were blinded to the group allocation. However, the bedside anaesthesiologist in charge of setting the ventilator was not blinded.

Video-assisted thoracoscopic lung resection surgery A protocolised surgical pathway for the video-assisted thoracoscopic lung resection for congenital cystic adenomatoid malformation was followed in all patients. Patients were admitted 1 day before surgery, and preoperative routine tests, including chest radiography and venous blood gas analysis, were performed. After surgical lung resection, an air-leak test was performed, and one mini-tube was inserted into the thoracic cavity. Follow-up chest X-rays were performed daily until there was no evidence of pneumothorax, haemothorax, pleural effusion, or other complications. The chest tube was removed after normal X-ray findings were confirmed; patients were discharged 1 day after chest tube removal. When respiratory infection was suspected, additional antibiotics (e.g. piperacillin/tazobactam) were started.

Anaesthesia induction and maintenance The pre-use leak and compliance tests were performed before anaesthesia induction; the volume of the breathing circuit was not changed thereafter. Anaesthesia was induced with atropine 0.02 mg kg1, thiopental sodium 5 mg kg1, and rocuronium 0.6 mg kg1, then maintained using sevoflurane (1.5e2.5 vol%) and a continuous infusion of remifentanil 0.1e0.2 m kg1 min1. An additional 0.3 mg kg1 of rocuronium was added when surgery started. Electrocardiography, peripheral oxygen saturation, and invasive arterial blood pressure monitoring were performed during anaesthesia.

Lung protective ventilation in children

Ventilator equipment

Outcome measures

We used a single type of mechanical ventilator (Avance CS2; GE Healthcare, Milwaukee, WI, USA). According to the manufacturer’s handbook, this ventilator provides tidal volume compensation during volume-controlled ventilation, fresh gas flow compensation (from 150 ml min1 to 15 L min1), and temperature and atmospheric pressure compensation. Moreover, compliance is automatically compensated for compression losses within the absorber and bellows assembly. This compliance compensation enables accurate tidal volumes as low as 20 ml, thereby making neonatal mechanical ventilation possible. Flow sensors in the breathing system acquired volume and flow measurements; the flow sensor sensitivity for volume was as low as 5 ml.

Arterial blood gas measurement

Study protocol: two-lung ventilation After tracheal intubation, a recruitment manoeuvre was performed on patients of both groups before initiating mechanical ventilation. Mechanical ventilation started with a tidal volume of 10 ml kg1 without PEEP in the control group and 6 ml kg1 with PEEP at 6 cm H2O in the LPV group. Volumecontrolled mode with a constant (square waveform) inspiratory flow was used for mechanical ventilation. The tidal volume was set based on the actual body weight. The set inspired oxygen fraction was 0.5, and the initial ventilatory frequency rate was set to deliver 20e25 breaths min1 and adjusted to maintain an end-tidal carbon dioxide (ETCO2) concentration to 4.7e5.3 kPa, with an inspiratory/expiratory (I/E) ratio of 1:2.

Study protocol: one-lung ventilation All lung resections were performed using video-assisted thoracoscopy. Therefore, we started OLV using a 4 French (Fr) Fogarty catheter (Fogarty® Occlusion Catheter; Edwards Lifesciences, Irvine, CA, USA) or endobronchial blocker 5 Fr (Arndt blocker; Cook Critical Care, Bloomington, IL, USA) according to the main bronchus size of the operative side. The Fogarty catheter and bronchial blocker were introduced using fibreoptic bronchoscopy (Olympus LF-DP; Olympus Corporation, Tokyo, Japan) guidance. After lateral decubitus positioning and lung separation, OLV commenced with a tidal volume of 8 ml kg1 without PEEP in the control group and 4 ml kg1 with PEEP at 6 cm H2O in the LPV group. The set inspired oxygen fraction was maintained at 1.0 during OLV. After lung lobectomy or segmentectomy was completed, both lungs were ventilated with each ventilation technique that was applied after anaesthetic induction according to the groups. Mechanical ventilation of all patients was maintained in a volumecontrolled mode with a constant flow rate during inspiration, without pressure regulation. Regarding fluid management, both groups were uniformly managed to avoid excessive fluid administration, and only crystalloids were used.

Study protocol: recruitment manoeuvres Recruitment manoeuvres were performed three times in all patients: after tracheal intubation, after commencement of OLV, and at the end of OLV before restarting two-lung ventilation (TLV). In this study, lung recruitment was done according to previous reports: sustained airway pressure of 30 cm H2O for 15e20 s.10e12

-

3

Arterial blood gas was checked four times: 15 min after tracheal intubation (T1), 15 min after OLV commencement (T2), 1 h after OLV commencement but still during OLV (T3), and at the end of surgery (T4). Immediately after surgery, all patients were extubated after gentle tracheal suction and lung recruitment, then transferred to the PACU.

Lung ultrasonography and scoring system Lung ultrasound examination was performed at the end of surgery in all patients. All ultrasound scans were performed by a single anaesthesiologist, who was blinded to the group allocation and had experience with more than 100 lung ultrasound scans in paediatric patients, using a Logiq e ultrasound machine (GE Healthcare, Wauwatosa, WI, USA) with a 4e10 MHz linear transducer. Patients were scanned according to a previous report.13 Each hemithorax was divided into six regions using three longitudinal lines (parasternal, anterior, and posterior axillary) and two axial lines (one above the diaphragm and the other 1 cm above the nipples). The 12 total lung regions were assessed for the following signs: lung sliding sign, A-line, B-line, or air bronchograms. As juxtapleural consolidation of various sizes and the presence of B-lines are the two most commonly diagnosed lung ultrasound signs of atelectasis during anaesthesia,13 the degree of consolidation and B-lines were scored for each region separately.14 The degree of juxtapleural consolidation was divided into four grades and scored between 0 and 3 with the following interpretations: 0, no consolidation; 1, minimal juxtapleural consolidation; 2, small-sized consolidation; and 3, large-sized consolidation. B-line degree was also divided into four grades and scored between 0 and 3 in the following manner: 0, fewer than three isolated B-lines; 1, multiple welldefined B-lines; 2, multiple coalescent B-lines; and 3, white lung. The graded degree for consolidation and B-lines were added for each region.

Data collection The intraoperative respiratory parameters, including inspiratory fraction of oxygen (FiO2), peak and mean airway pressure, expiratory tidal volume measured by the ventilator, ventilatory frequency, and arterial blood gas data, were collected at each epoch. The arterial oxygen partial pressure PaO2/FiO2 ratio and oxygenation index (FiO2 mean airway pressure PaO1 2 ) were calculated. Additionally, postoperative laboratory and imaging data and vital signs, including peripheral oxygen saturation, until patients’ discharge were collected.

Primary outcome The primary outcome of this study was the incidence of postoperative pulmonary complications within 72 h of surgery. The pulmonary complications were defined as respiratory infection, respiratory failure, pleural effusion, atelectasis, pneumothorax, bronchospasm, and aspiration pneumonitis, recommended by Jammer and colleagues15 and Gallart and Canet.16 Respiratory infection was defined by sputum, fever, leucocytosis, and increased opacity on chest radiography. Pleural effusion, atelectasis, and pneumothorax were defined based on chest radiography and CT findings. The images were

4

-

Lee et al.

Enrollment

Assessed for eligibility (n=114)

Excluded (n=0) ♦ Not meeting inclusion criteria (n=0) ♦ Declined to participate (n=0) ♦ Other reasons (n=0)

Randomized (n=114)

Allocation Allocated to intervention (n=57) ♦ Received allocated intervention (n=57)

Allocated to intervention (n=57) ♦ Received allocated intervention (n=57)

♦

♦

Did not receive allocated intervention (give reasons) (n=0)

Did not receive allocated intervention (give reasons) (n=0)

Follow-up Lost to follow-up (give reasons) (n=0)

Lost to follow-up (give reasons) (n=0 )

Discontinued intervention (give reasons) (n=0)

Discontinued intervention (give reasons) (n=0)

Analysis Analysed (n=55) ♦ Excluded from analysis (Fail to perform lung ultrasound) (n=2)

Analysed (n=55 ) ♦ Excluded from analysis (Fail to perform lung ultrasound) (n=2)

Fig 1. CONSORT diagram. CONSORT, Consolidated Standards of Reporting Trials.

examined in a blinded manner by an independent specialist in radiology who was not involved in the study. Complications were comprehensively assessed using chest radiography or CT if available, laboratory data, clinical symptoms, and vital signs.

Secondary outcomes The secondary outcomes included the lung ultrasound scores assessed at the end of surgery, the number of patients with intraoperative hypoxia (peripheral oxygen saturation (SpO2) <90%), intraoperative PaO2/FiO2 <40 kPa, postoperative extrapulmonary complications, and total hospital stay length. Post-hoc testing was also undertaken to assess perioperative determinants of postoperative pulmonary complications.

Sample size estimation The required sample size of the present study was determined using a previous study performed in adults undergoing lung resection in which the incidence of postoperative pulmonary dysfunction was decreased from 22% to 4% when LPV was applied during OLV.17 Therefore, it was assumed that the primary outcome of the present study would occur in 22% of the control group and 4% in the LPV group. To detect a difference

in the primary outcome using a c2 statistic (two-sided Z-test with pooled variance) with a power of 80% and a significance level of 5%, we calculated a sample size of 47 patients for each group using PASS 2008 software (version 8.0.16, NCSS statistical software; NCSS LLC, Kaysville, UT, USA). Considering a 20% attrition rate, a total sample size was increased to 114 patients (57 patients in each group).

Statistical analysis All data were analysed using SPSS for Windows (version 23.0; SPSS Inc., Chicago, IL, USA). Data normality was assessed using the KolmogoroveSmirnov test. Categorical variables were expressed as numbers and percentages, and continuous variables were expressed as means (standard deviation, SD) or medians and inter-quartile (IQR) ranges. c2 test was used to test categorical data significance, and a Fisher’s exact test was used when the expected count of >20% cells was less than 5. The number needed to treat with 95% confidence interval (CI) was calculated to evaluate the effect of LPV on primary outcome. Student’s t-test or ManneWhitney U test were used to determine the significance of continuous data. Univariable logistic regression analysis was initially used to identify the possible risk factors for postoperative pulmonary complications within 72 h of surgery. The presence of

Lung protective ventilation in children

-

5

Table 1 Baseline characteristics and intraoperative variables of all patients. Data are presented as mean (standard deviation), median (inter-quartile range), or n (%). CI, confidence interval; LPV, lung protective ventilation; OLV, one-lung ventilation; SpO2, peripheral oxygen saturation Parameters

Control group (n¼55)

LPV group (n¼55)

P-value

Age (yr) Height (cm) Weight (kg) Sex (male) Type of operation Segmentectomy or wedge resection Single lobectomy Bilobectomy OLV time (min) Total operation time (min) Anaesthesia time (min) Total infused fluid (ml) Minimum SpO2 during OLV (%)

1.8 (range 0.5e4.7) 85.8 (6.6) 11.4 (10.3, 12.3) 30 (54.5)

1.6 (range 0.8e4.3) 86.7 (7.5) 11.3 (10.6, 13.0) 24 (43.6)

0.850 0.492 0.432 0.340 0.228

12 (21.8) 42 (76.4) 1 (1.8) 60 (40, 70) 63 (50, 75) 109.3 (27.1) 72 (53, 89) 98 (95, 99)

12 (21.8) 43 (78.2) 0 55 (50, 75) 60 (55, 80) 111.8 (30.5) 75 (49, 92) 99 (97, 100)

multicollinearity between related parameters with a P-value 0.1 for the univariable analysis was determined using the variance inflation factor >2.5 (or high correlation coefficient >0.7).18 If there were variables with multicollinearity, a variable was selected according to clinical significance. Subsequently, multivariable logistic regression was performed to identify risk factors associated with pulmonary complications using remaining covariates with a P-value 0.1 for the univariable analysis. Stepwise selection by backward elimination method was used, and no additional steps were required to change the covariates to develop further risk prediction models. The HosmereLemeshow goodness-of-fit test was used to compare the estimated-to-observed likelihood of outcomes. All P-values were two-sided, and P<0.05 was considered statistically significant.

Results Patient characteristics A total of 114 paediatric patients were enrolled and randomised into two groups from February 2016 to December 2017. Among them, four patients were excluded as a result of failure to obtain lung ultrasonography data because of unavailability of the ultrasound machine. Therefore, data for 110

0.554 0.600 0.644 0.694 0.017

children (55 in the control group and 55 in the LPV group) were analysed according to intention-to-treat analysis (Fig. 1). Table 1 summarises the baseline characteristics and intraoperative variables of patients. No patients experienced respiratory dysfunction or needed oxygen supply before surgery. All patients were transferred to the PACU after tracheal extubation.

Primary outcome There were 19 patients (17.3%) who showed postoperative pulmonary complications within 72 h of surgery including respiratory infection, pleural effusion, atelectasis, and pneumothorax. Postoperative pulmonary complications occurred more frequently in the control group, compared with LPV (25.5% vs 9.1; odds ratio [OR]¼0.29; 95% CI, 0.10e0.88; P¼0.02; Table 2). The number needed to treat with LPV was six children (95% CI, 3e39; P¼0.03). All pneumothorax and pleural effusion occurred on the same side as lung resection. Atelectasis, pneumothorax, and pleural effusion were detected on the day of surgery and continuously observed until the chest tube was removed between postoperative days 1 and 3, and after normal chest radiography was confirmed.

Table 2 Dependent lung ultrasound scores and incidence of postoperative pulmonary complications of patients included in the study. Data are present as median (inter-quartile range) and n (%). *Complications included pneumothorax, chylothorax and pleural effusion requiring chest tube insertion. LPV, lung protective ventilation

Lung ultrasound Consolidation score Anterior Lateral Posterior B-lines score Anterior Lateral Posterior Postoperative pulmonary complications* Respiratory infection Pleural effusion Atelectasis Pneumothorax

Control group (n¼55)

LPV group (n¼55)

9 (6, 11.25) 3 (1, 4) 4 (3, 5) 3 (1, 4) 11 (9, 13) 4 (2, 5) 5 (3, 5) 3 (1, 4) 14 (25.5) 3 (5.5) 3 (5.5) 3 (5.5) 5 (9.1)

2 (1, 4) 0 (0, 1) 1 (0, 2) 0 (0, 1) 3.5 (1, 6) 1 (0, 2) 2 (0, 3) 0 (0, 2) 5 (9.1) 1 (1.8) 2 (3.6) 0 2 (3.6)

Odds ratio (95% CI)

P-value

3.41 (1.13e10.27)

<0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 <0.001 0.021

6

-

Lee et al.

PaO2/FiO2 ratio In the control group, 10 patients (18.2%) had a PaO2/FiO2 <40 kPa at the end of surgery (T4) compared with two (3.6%) in the LPV group (OR¼5.9; 95% CI, 1.2e28.3; P<0.01; Table 3). At all times, the patients in the LPV group had higher ETCO2 and PaCO2 values than those in the control group (P<0.001). From T2 to T4, peak airway pressure was lower in the LPV group (P<0.001), but mean airway pressure was higher in the LPV group than in the control group at all epochs (all P<0.001). The oxygenation index was higher in the LPV group at T1 and T2 (all P<0.001). There was no difference in the amount of fluid infused during anaesthesia between the two groups.

Hospital stay The mean (SD) hospital stay length of patients with and without pulmonary complications was 3.6 (1.9) and 4.2 (2.0) days, respectively (mean difference, e0.61; 95% CI, e1.6 to 0.4; P¼0.23). The median (IQR) hospital stay length was similar between groups (both 3 [3e4]; P¼0.838).

Perioperative factors contributing to postoperative pulmonary complications

Fig 2. Lung ultrasound findings of dependent lung in the control and LPV groups. The images were obtained from the lateral region, between the anterior and posterior axillary lines, and 1 cm above the nipples. In the control group (a), small to largesized juxtapleural consolidation and multiple coalescent Blines were observed. In the LPV group (b), two isolated B-lines without juxtapleural consolidation were observed. Asterisks indicate juxtapleural consolidation and arrows indicate B-lines. LPV, lung protective ventilation.

Secondary outcomes Lung ultrasound assessment Median lung ultrasound scores at the end of surgery (Table 2 and Fig. 2) were lower in the LPV group than in the control group for consolidation (2 [IQR, 1e4] vs 9 [6e11]; P<0.001) and B-lines (4 [1e6] vs 11 [9e13]; P<0.001).

OLV desaturation During OLV, desaturation (SpO2 <90%) occurred in eight/55 patients (14.5%) in the control group and one/55 patients (1.8%) in the LPV group (OR¼9.2; 95% CI, 1.1e76; P¼0.015). The minimum SpO2 during OLV was lower in the control group than in the LPV group (P¼0.017; Table 1).

Post-hoc testing was undertaken to assess perioperative factors associated with the development of postoperative pulmonary complications. Univariable analysis detected possible risk factors for postoperative pulmonary complications. Multicollinearity was found among groups, PaCO2 at T1 and T2, and mean airway pressure at T1. Therefore, group, OLV time, peak inspiratory pressure at T1 and T2, and the PaO2/FiO2 ratio at T4 were included into the multivariable logistic regression analysis. Among these possible risk factors, LPV (OR¼0.25; 95% CI, 0.071e0.942; P¼0.04) and OLV duration (OR¼1.024; 95% CI, 1.001e1.048; P¼0.037) were found to be independently associated with postoperative pulmonary complications within 72 h of surgery (Table 4). The HosmereLemeshow test showed a good fit (c2 test¼2.030, P¼0.154). There was only one case, in the control group, of an extrapulmonary complication, which required re-operation because of bleeding.

Discussion This randomised controlled study revealed that intraoperative LPV comprising lower tidal volumes with PEEP are associated with fewer postoperative pulmonary complications and better intraoperative oxygenation compared with conventional ventilation in children undergoing lung resection requiring OLV. Additionally, lung atelectasis could be reduced by LPV when assessed using lung ultrasound. However, total hospital stay length did not differ between children receiving different ventilatory strategies. Although characteristics of previous studies evaluating the association of applied intraoperative LPV and postoperative outcomes were heterogenous in terms of surgery type, patients’ risk for pulmonary complication, amount of tidal volume, and PEEP level, LPV reduced postoperative pulmonary complications, improved the pulmonary gas exchanges, and shortened hospital stay length in adults.5,6,17 Clinical data on lung protective ventilatory strategies in children are scarce. Avoiding high tidal volume and atelectrauma is considered to be rational,19,20 but the standard strategy of LPV has not been defined yet, as the respiratory physiology of children differs from that of adults. Because of

Lung protective ventilation in children

-

7

Table 3 Intraoperative respiratory parameters and results of arterial blood gas analysis. Data are presented as mean (standard deviation) or median (inter-quartile range). CI, confidence interval; ETCO2, end tidal carbon dioxide pressure; FiO2, inspiratory fraction of oxygen; LPV, lung protective ventilation; OLV, one lung ventilation; PaCO2, arterial partial pressure of carbon dioxide; PaO2, arterial partial pressure of oxygen; T1, 15 min after tracheal intubation; T2, 15 min after commencement of one lung ventilation; T3, 1 h after commencement of one lung ventilation but during one lung ventilation; T4, at the end of surgery

T1

T2

T3

T4

Parameters

Control group (n¼55)

LPV group (n¼55)

Tidal volume (ml) Respiratory rate (min1) PaO2 (kPa) PaCO2 (kPa) Peak airway pressure (cm H2O) Mean airway pressure (cm H2O) PaO2/FiO2 (kPa) Oxygenation index Tidal volume (ml) Respiratory rate (min1) PaO2 (kPa) PaCO2 (kPa) Peak airway pressure (cm H2O) Mean airway pressure (cm H2O) PaO2/FiO2 (kPa) Oxygenation index Tidal volume (ml) Respiratory rate (min1) PaO2 (kPa) PaCO2 (kPa) Peak airway pressure (cm H2O) Mean airway pressure (cm H2O) PaO2/FiO2 (kPa) Oxygenation index Tidal volume (ml) Respiratory rate (min1) PaO2 (kPa) PaCO2 (kPa) Peak airway pressure (cm H2O) Mean airway pressure (cm H2O) PaO2/FiO2 (kPa) Oxygenation index

110 (100, 120) 20 (20, 22) 35.8 (35.2, 37.2) 4.7 (0.5) 14 (13, 14) 5 (5, 5) 72.8 (7.3) 0.9 (0.9, 1.0) 89.3 (20.0) 23.2 (3.2) 40.4 (10.4) 5.6 (0.5) 21.1 (3.1) 7 (6, 7) 40.4 (10.4) 2.2 (1.8, 2.8) 89.5 (14.2) 23.7 (2.9) 26.3 (11.3) 5.9 (0.8) 22 (19, 26.5) 7 (6, 8) 29.2 (12.4) 4.2 (2.6) 110 (100, 120) 20 (18, 23) 30.0 (23.5, 32.5) 5.0 (4.7, 5.4) 19 (17, 22) 6 (6, 7) 56.8 (14.7) 1.6 (0.5)

72 (65, 85) 25 (23, 27) 36.1 (34.4, 37.6) 5.8 (0.7) 13 (12, 14) 8 (8, 8) 71.2 (7.8) 1.5 (1.4, 1.6) 53.9 (10.9) 31.8 (6.3) 40.4 (11.6) 8.4 (0.9) 17.4 (2.3) 9 (9, 10) 40.4 (11.6) 2.8 (2.4, 3.5) 59.6 (13.7) 30.5 (5.1) 32.3 (11.8) 6.7 (1.4) 18 (16, 20) 9 (9, 10) 35.5 (13.2) 4.0 (2.2) 70 (65, 80) 28 (26, 31) 31.0 (26.5, 35.0) 5.6 (5.3, 6.1) 18 (16, 19.75) 9 (8.25, 10) 61.4 (10.8) 2.0 (0.6)

smaller functional residual capacities, larger closing volumes, and greater chest wall compliance, younger children and infants are prone to have atelectasis during general anaesthesia and more postoperative pulmonary complications.21 Additionally, patients undergoing thoracic surgery may have a higher risk for postoperative pulmonary complications9 because of greater damage to the dependent lung during OLV, loss of lung parenchyma, and surgical manipulation of the non-dependent lung, which can cause tissue injury and trigger biomarker release.21,22 Therefore, potential greater benefit may be gained from protective ventilation during pulmonary resection in children. To the best of our knowledge, this is the first randomised trial to evaluate the lower tidal volume with PEEP during surgery in a paediatric population. The results of the present study were consistent with previous adult studies in terms of reduced postoperative pulmonary complications by LPV. For protective ventilation, the tidal volume of 6 ml kg1 during TLV was selected because a supraphysiological tidal volume of more than 5e7 ml kg1 is injurious.23,24 For protective OLV, a tidal volume of 4 ml kg1 was used, which was slightly lower than that used in studies involving adults.25 For using lower tidal volume, sufficient or moderate levels of PEEP should follow as low tidal volume is associated with

Mean differences (95% CI)

e1.1 (e1.4 to e0.9)

1.6 (e1.3 to 4.6) 35.4 (29.4e41.5) e8.6 (e10.4 to e6.7) 0.1 (e4.1 to 4.2) e2.8 (e3.9 to e1.4) 3.7 (2.7e4.7) 0.0 (e4.1 to 4.2) 30.1 (20.6e39.2) e6.8 (e9.6 to e4.1) e6.1 (e13.8 to 1.6) e0.8 (e1.5 to e0.1)

e6.3 (e14.8 to 2.3) 0.2 (e1.5 to 1.7)

e4.6 (e9.7 to 0.5) e0.4 (e0.6 to 0.1)

P-value <0.001 <0.001 0.550 <0.001 0.447 <0.001 0.267 <0.001 <0.001 <0.001 0.998 <0.001 <0.001 <0.001 0.998 <0.001 <0.001 <0.001 0.120 0.045 0.001 <0.001 0.148 0.156 <0.001 <0.001 0.385 <0.001 0.009 <0.001 0.072 0.156

atelectasis and postoperative adverse outcomes26; therefore, determining optimal PEEP was critical. In a paediatric animal model, lung compliance was improved at PEEP levels between 4 and 8 cm H2O.27 In children with acute respiratory failure, arterial oxygen tension improved when PEEP exceeded 6 cm H2O, whereas oxygen transport and the cardiac index decreased when PEEP was greater than 3 cm H2O.28 Considering our clinical experience in children with normal lung function, we applied a PEEP of 6 cm H2O in this study. The patients who received LPV showed lower peak inspiratory pressure and higher mean airway pressure. This might be associated with improved oxygenation. Although mean PaO2 and the PaO2/FiO2 ratio during anaesthesia did not differ significantly between the two groups, LPV reduced the incidence of low PaO2/FiO2 ratios at the end of surgery and was associated with less desaturation, implying that a low tidal volume with adequate PEEP was helpful in avoiding deoxygenation. These findings indicate that adequate gas exchange does not necessarily indicate that the ventilation strategy is optimal. In other words, simply monitoring gas exchange is not a sufficient guide to the optimal ventilation strategy but should be used in combination with an LPV strategy consisting of limited tidal volume with PEEP to reduce the incidence of postoperative pulmonary complications. Additionally, higher

8

-

Lee et al.

Table 4 Univariable and multivariable analysis for pulmonary complications after lung resection in children. *Type of surgery: reference is segmentectomy or wedge resection. LPV, lung protective ventilation; PaCO2, arterial partial pressure of carbon dioxide; PaO2/FiO2, arterial partial pressure of oxygen/inspiratory fraction of oxygen; PIP, peak inspiratory pressure Parameters

Age (yr) Height (cm) Weight (kg) Sex (female) Group (LPV group) Operation side (left side) Type of surgery* (unilobectomy or bilobectomy) One lung ventilation time (min) Anaesthesia time (min) Operation time (min) T1 PaCO2 (kPa) T1 PIP (cm H2O) T1 MAP (cm H2O) T1 PaO2/FiO2 ratio (kPa) T1 oxygenation index T2 PaCO2 (kPa) T2 PIP (cm H2O) T2 MAP (cm H2O) T2 PaO2/FiO2 ratio (kPa) T2 oxygenation index T4 PaCO2 (kPa) T4 PIP (cm H2O) T4 MAP (cm H2O) T4 PaO2/FiO2 ratio (kPa) T4 oxygenation index

Univariable analysis

Multivariable analysis

Odds ratio (95% CI)

P-value

1.146 1.003 1.111 1.406 0.293 1.087 0.739 1.018 1.011 1.013 0.877 1.230 0.723 0.999 0.382 0.921 1.161 0.776 1.000 0.816 0.956 1.100 0.768 0.995 1.109

0.676 0.930 0.296 0.504 0.029 0.869 0.603 0.054 0.207 0.150 0.011 0.096 0.076 0.906 0.200 0.026 0.05 0.140 0.908 0.353 0.405 0.299 0.107 0.063 0.810

(0.605e2.175) (0.935e1.076) (0.912e1.355) (0.518e3.817) (0.097e0.881) (0.404e2.925) (0.236e2.309) (1.000e1.037) (0.994e1.027) (0.996e1.030) (0.793e0.970) (0.964e1.570) (0.505e1.035) (0.990e1.009) (0.088e1.662) (0.857e0.990) (1.000e1.348) (0.554e1.087) (0.994e1.006) (0.531e1.254) (0.860e1.063) (0.919e1.317) (0.557e1.059) (0.990e1.000) (0.478e2.572)

PaCO2 levels in LPV might enhance hypoxic pulmonary vasoconstriction and reduce ventilation/perfusion mismatch during OLV in children.29 We also found that postoperative atelectasis was evident in the control group compared with the LPV group, as evaluated using lung ultrasonography. Considering that recruitment manoeuvres were applied in both groups, the importance of sufficient PEEP in the LPV group contributed to the prevention of anaesthesia-induced atelectasis; PEEP prevents atelectasis and improves oxygenation in healthy adult lungs during general anaesthesia.30 Additionally, our results suggested that recruitment manoeuvres alone could not prevent atelectasis, mirroring similar previous studies.31 Lung ultrasound scores for consolidation and B-lines were highest in the lateral regions at the end of surgery for both groups. Generally, anaesthesia-induced atelectasis is predominant in areas of the dependent lung in children when they were evaluated using MRI or ultrasonography.14,32 For unilateral lung resection in this study, children had to be placed in a lateral decubitus position, with the non-operative lung placed in the dependent area. The lateral part of lung was the most dependent area. Although LPV was associated with reduced pulmonary complications, no differences were found in the total hospital stay length between the two groups. The median hospital stay length was 3 days in this study, whereas adults undergoing lung resection typically require 5e8 days.9,17 Additionally, our mean operation time was about 60 min and no one required treatment in the ICU, whereas mean operation times were more than 120 min in adults requiring OLV, and postoperative mechanical ventilatory support was required in some patients.9,17,33 Furthermore, postoperative mortality or

Odds ratio (95% CI)

P-value

0.26 (0.07e0.94)

0.040

1.024 (1.001e1.05)

0.037

significant morbidities that were found in adult studies, including acute respiratory distress syndrome or pneumonia,9,17,33 were not observed in our patients. Finally, a common protocolised pathway was used for all surgeries, which reduces the variability in clinical practice. Whether LPV may impact on hospital stay requires further study across multiple centres. The effect of choice of anaesthetic technique on outcomes in the paediatric population undergoign lung resection also requires further investigation.34 Several limitations should be considered. First, evaluation of lung ultrasonography only occurred in the operating room, and chest radiographic findings were followed up to assess pulmonary complications. However, lung ultrasonography is more accurate than chest X-ray for the diagnosis of pneumonia35 and comparable with chest CT.36 Second, inflammatory biomarkers were not measured for direct assessment of lung damage. Third, the incidence of each clinical complication after surgery was low. Fourth, we used actual body weight, not predicted body weight, to set tidal volume. The predicted or lean body weight is recommended to avoid volutrauma in obese patients for LPV37; however, in the present study, most patients had a normal BMI. Finally, the volumecontrolled mode was used, and the peak airway pressure would have been influenced by various factors including inspiratory flow rate, resistance, and I/E ratio. Measurement of the plateau pressure would provide the best data to demonstrate the alveolar stress. However, the plateau pressure was not measured in this study. In summary, LPV reduced the incidence of postoperative pulmonary complications such as respiratory infection, pleural effusion, atelectasis, and pneumothorax in children

Lung protective ventilation in children

undergoing lung resection requiring OLV. However, the longterm benefit of protective lung ventilation should be further investigated.

13.

Authors’ contributions Patient recruitment: JHL Data collection: JHL, YEJ, EHK Data analysis: JHL, YEJ, EHK, HSK Writing of the first draft of the paper: JIB Approval of the manuscript: HSK Revision of the manuscript: JTK, JHL Approval of the final version: JTK

14.

15.

Funding Institutional and department sources.

Declaration of interest

16.

The authors declare that they have no conflicts of interest. 17.

References 1. Ghaferi AA, Birkmeyer JD, Dimick JB. Variation in hospital mortality associated with inpatient surgery. N Engl J Med 2009; 361: 1368e75 2. Canet J, Gallart L, Gomar C, et al. Prediction of postoperative pulmonary complications in a population-based surgical cohort. Anesthesiology 2010; 113: 1338e50 3. Tusman G, Bohm SH, Warner DO, Sprung J. Atelectasis and perioperative pulmonary complications in high-risk patients. Curr Opin Anaesthesiol 2012; 25: 1e10 4. Brower RG, Matthay MA, Morris A, Schoenfeld D, Thompson BT, Wheeler A. Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. N Engl J Med 2000; 342: 1301e8 5. Severgnini P, Selmo G, Lanza C, et al. Protective mechanical ventilation during general anesthesia for open abdominal surgery improves postoperative pulmonary function. Anesthesiology 2013; 118: 1307e21 6. Futier E, Constantin JM, Paugam-Burtz C, et al. A trial of intraoperative low-tidal-volume ventilation in abdominal surgery. N Engl J Med 2013; 369: 428e37 7. Clark RH, Slutsky AS, Gerstmann DR. Lung protective strategies of ventilation in the neonate: what are they? Pediatrics 2000; 105: 112e4 8. Feldman JM. Optimal ventilation of the anesthetized pediatric patient. Anesth Analg 2015; 120: 165e75 9. Amar D, Zhang H, Pedoto A, Desiderio DP, Shi W, Tan KS. Protective lung ventilation and morbidity after pulmonary resection: a propensity score-matched analysis. Anesth Analg 2017; 125: 190e9 10. Marcus RJ, Van Der Walt JH, Pettifer RJA. Pulmonary volume recruitment restores pulmonary compliance and resistance in anaesthetized young children. Pediatr Anesth 2002; 12: 579e84 11. Duff JP, Rosychuk RJ, Joffe AR. The safety and efficacy of sustained inflations as a lung recruitment maneuver in pediatric intensive care unit patients. Intensive Care Med 2007; 33: 1778e86 12. Morrow B, Futter M, Argent A. A recruitment manoeuvre performed after endotracheal suction does not increase

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

-

9

dynamic compliance in ventilated paediatric patients: a randomised controlled trial. Aust J Physiother 2007; 53: 163e9 Acosta CM, Maidana GA, Jacovitti D, et al. Accuracy of transthoracic lung ultrasound for diagnosing anesthesiainduced atelectasis in children. Anesthesiology 2014; 120: 1370e9 Song IK, Kim EH, Lee JH, Ro S, Kim HS, Kim JT. Effects of an alveolar recruitment manoeuvre guided by lung ultrasound on anaesthesia-induced atelectasis in infants: a randomised, controlled trial. Anaesthesia 2017; 72: 214e22 Jammer I, Wickboldt N, Sander M, et al. Standards for definitions and use of outcome measures for clinical effectiveness research in perioperative medicine: European Perioperative Clinical Outcome (EPCO) definitions: a statement from the ESAeESICM joint taskforce on perioperative outcome measures. Eur J Anaesthesiol 2015; 32: 88e105 Gallart L, Canet J. Post-operative pulmonary complications: understanding definitions and risk assessment. Best Pract Res Clin Anaesthesiol 2015; 29: 315e30 Yang M, Ahn HJ, Kim K, et al. Does a protective ventilation strategy reduce the risk of pulmonary complications after lung cancer surgery?: a randomized controlled trial. Chest 2011; 139: 530e7 Indrayan A, Holt MP. Concise encyclopedia of biostatistics for medical professionals. Boca Raton, FL: CRC Press; 2016. p. 385 Santschi M, Randolph AG, Rimensberger PC, Jouvet P. Mechanical ventilation strategies in children with acute lung injury: a survey on stated practice pattern. Pediatr Crit Care Med 2013; 14: e332e7 Albuali WH, Singh RN, Fraser DD, et al. Have changes in ventilation practice improved outcome in children with acute lung injury? Pediatr Crit Care Med 2007; 8: 324e30 Bordet F, Allaouchiche B, Lansiaux S, et al. Risk factors for airway complications during general anaesthesia in paediatric patients. Paediatr Anaesth 2002; 12: 762e9 Gattinoni L, Caironi P. Refining ventilatory treatment for acute lung injury and acute respiratory distress syndrome. JAMA 2008; 299: 691e3 Kneyber MCJ. Intraoperative mechanical ventilation for the pediatric patient. Best Pract Res Clin Anaesthesiol 2015; 29: 371e9 Kneyber MCJ, Zhang H, Slutsky AS. Ventilator-induced lung injury: similarity and differences between children and adults. Am J Respir Crit Care Med 2014; 190: 258e65 Vegh T, Juhasz M, Szatmari S, et al. Effects of different tidal volumes for one-lung ventilation on oxygenation with open chest condition and surgical manipulation: a randomised cross-over trial. Minerva Anestesiol 2013; 79: 24e32 Levin MA, McCormick PJ, Lin HM, Hosseinian L, Fischer GW. Low intraoperative tidal volume ventilation with minimal PEEP is associated with increased mortality. Br J Anaesth 2014; 113: 97e108 Probyn ME, Hooper SB, Dargaville PA, et al. Positive end expiratory pressure during resuscitation of premature lambs rapidly improves blood gases without adversely affecting arterial pressure. Pediatr Res 2004; 56: 198e204 White MK, Galli SA, Chatburn RL, Blumer JL. Optimal positive end-expiratory pressure therapy in infants and children with acute respiratory failure. Pediatr Res 1988; 24: 217e21

10

-

Lee et al.

29. Corbelli R, Habre W. Ventilating the lungs safely: what’s new for infants and children? Curr Anesthesiol Rep 2013; 3: 42e8 € € m H, 30. Ostberg E, Thorisson A, Enlund M, Zetterstro Hedenstierna G, Edmark L. Positive end-expiratory pressure alone minimizes atelectasis formation in nonabdominal surgery: a randomized controlled trial. Anesthesiology 2018; 128: 1117e24 31. Reinius MDH, Jonsson MDL, Gustafsson MDPDS, et al. Prevention of atelectasis in morbidly obese patients during general anesthesia and paralysis: a computerized tomography study. Anesthesiology 2009; 111: 979e87 32. Acosta CM, Tusman G, Jacovitti D, et al. Anesthesiainduced atelectasis assessed by lung sonography. Crit Ultrasound J 2014; 6: A13 33. Shen Y, Zhong M, Wu W, et al. The impact of tidal volume on pulmonary complications following minimally invasive esophagectomy: a randomized and controlled study.

34.

35.

36.

37.

J Thorac Cardiovasc Surg 2013; 146: 1267e73. discussion 73e4 ~ eiro P, Reyes A, Vara E, Olmedilla L, Cruz P, de la Gala F, Pin Garutti I. Postoperative pulmonary complications, pulmonary and systemic inflammatory responses after lung resection surgery with prolonged one-lung ventilation. Randomized controlled trial comparing intravenous and inhalational anaesthesia. Br J Anaesth 2017; 119: 655e63 Ticinesi A, Lauretani F, Nouvenne A, et al. Lung ultrasound and chest x-ray for detecting pneumonia in an acute geriatric ward. Medicine (Baltimore) 2016; 95: e4153 Ye X, Xiao H, Chen B, Zhang S. Accuracy of lung ultrasonography versus chest radiography for the diagnosis of adult community-acquired pneumonia: review of the literature and meta-analysis. PLoS One 2015; 10: e0130066 Martin DC, Richards GN. Predicted body weight relationships for protective ventilation e unisex proposals from pre-term through to adult. BMC Pulm Med 2017; 17: 85 Handling editor: G.L. Ackland