Mechanical Ventilation During Cardiac Surgery

BACKGROUND Respiratory failure (RF) is a common complication in cardiac surgery, with a global incidence of 20-25% . Its clinical manifestation ranges from a mild form of respiratory failure up to an acute respiratory distress syndrome (ARDS) requiring prolonged mechanical ventilation (MV) and intensive care unit (ICU) stay.

The pathophysiologic mechanism of RF is quite complex, but it is known that cardiopulmonary bypass (CPB) plays a main role in determining lung injury. A number of factors contribute to this injury: atelectasis, hyper-oxygenation causing the release of free radicals and CPB-related systemic inflammatory response.

It's a common practice to suspend ventilation during CPB, since lung function is carried out by an extracorporeal gas-exchanger. Moreover the absence of lung movements clearly facilitates surgery. However, the interruption of MV during CPB is associated with the development of micro-atelectasis, hydrostatic pulmonary edema, reduced lung compliance and surfactant diffusion.

A recent observational study has identified the duration of CPB as an important risk factor for the development of microbiologically-documented pneumonia.

During last years' several preventive lung protective strategies have been investigated and proposed: ultrafiltration to remove neutrophils, controlled hemodilution (with hematocrit higher than 23%), steroids and MV settings during CPB, such as the application of a Positive End-Expiratory Pressure (PEEP) or a Continuous Positive Airway Pressure (CPAP) 5-15 of cmH2O, low tidal-high frequency ventilation (100 acts-per-minute), application of 100% oxygen inspired fraction (FiO2), and bilateral CPB, which involves lungs for blood-oxygenation.

A recent meta-analysis based on sixteen clinical trials found an increase in oxygenation and a reduction in shunt fraction immediately after the weaning from CPB if CPAP was applied during CPB. Similar results were obtained with a lung recruitment maneuver at the end of CPB. Furthermore, maintaining MV during the whole duration of extracorporeal circulation would reduce the CPB-related inflammatory response and tissue damage. Unfortunately, although adequately planned, studies are not powered enough to recommend maintaining MV during CPB as an evidence-based strategy to prevent respiratory complications, because major indicators of clinical outcome (i.e., duration of postoperative MV, length of ICU and hospital stay, and long-term follow-up) have not been investigated. Therefore, according to the available data in literature, an unquestionable standardized strategy of lung protection during CPB cannot be recommended.

Objectives We designed a randomized controlled trial to investigate the effects of three different ventilator strategies in the short, medium and long term. We are testing the hypothesis that MV during CPB would reduce lung damage, defined as the incidence of RF (PaO2/FiO2 < 200) and post-operative pulmonary complications (PPCs). See Appendix 1 for complete definition of PPCs.

METHODS Trial design The CPBVENT study is a non-pharmacological, multi-center, single-blind, randomized controlled trial.

The study has been registered on ClinicalTrials.gov with the registration number NCT02090205 and was endorsed by the Study Group on Cardiothoracic and Vascular Anesthesia of the SIAARTI (the Italian Society of Anesthesia and Intensive Care Medicine).

Participants After Ethics Committee approval, we are going to enroll patients aged 18 or over undergoing elective cardiac surgery with planned use of CPB, aortic cross-clamping, median sternotomy and two lung ventilation. All patients will provide written informed consent before their inclusion in the trial.

End-points The primary end-point will be the reduction of incidence of PaO2/FiO2 ratio <200 until the discharge from the Intensive Care Unit (ICU).

The secondary end-points will be the evaluation of the following:

• readmission to the ICU for RF,
• need for re-intubation,
• need for noninvasive ventilation,
• duration of mechanical ventilation,
• length of the ICU and hospital stay,
• cardiovascular complications,
• short-term and long-term mortality,
• post-operative infections,
• Post-Operative Residual Curarization (PORC): measured with a Train Of Four (TOF) and defined as need for pharmacological reversal.

Interventions (Randomization and treatment protocol) The randomization list was created by the coordinating center with a dedicated software and was stratified per center, in a 1:1:1 ratio, in blocks of 30. Once the patient releases informed consent, the investigator will log in on a dedicated on line portal and he will obtain the allocation arm. From that moment it will be impossible to remove the patient's record card from the online platform and, in any case, the patient will be analyzed according to the intention-to-treat principle. Any deviation from the ventilation protocol, together with reason for deviation, will be recorded on the CRF. All the patients will be kept blind to the allocation.

Patients will be randomly assigned to receive one of the following ventilator strategies:

• First arm. No mechanical ventilation during CPB: patient will be disconnected from the respiratory circuit.
• Second arm. Patients will receive CPAP with PEEP of 5 mmH2O and FiO2 < 80%. To perform CPAP the ventilator will be set in manual/spontaneous mode, with a flow of 1-2 L/min and the adjustable pressure valve (APL) set at 5 cmH2O. The actual pressure will be checked with a pressure gauge integrated in the ventilator and a pressure gauge connected to the proximal end of the endotracheal tube.
• Third arm. Patient will be ventilated with a respiratory rate of 5 per minute, with tidal volume (TV) of 2-3 mL/Kg of Ideal Body Weight (IBW) and PEEP of 3-5 cm H2O.

Before and after CPB patients will receive a lung-protective ventilator strategy, with an Intermittent Positive Pressure Ventilation (IPPV) mode along with the following parameters:

• Tidal Volume (TV) = 6-8 ml/kg of IBW:
• PEEP = 5 cmH2O
• FiO2 <80%
• I:E = 1:2 (inspiration:expiration ratio). During CPB our goal will be to maintain a PaO2 between 200 and 250 mmHg, in order to avoid hyperoxia-induced lung injury; moreover the hematocrit will be maintained above 24%. During weaning from CPB we will perform a single alveolar recruitment maneuver, with an airway pressure of 40 cmH2O maintained at least for 7 seconds.

Post-operative ventilation The anesthesiologist will report in the Case Report Form (CRF) the mechanical ventilation setting used during the transfer of the patient from the operating theater to ICU. In ICU we will apply an IPPV with the same parameters used in the operating room. Blood oxygen saturation will be constantly monitored with a pulseoxymeter. We will report the extubation-time, the duration of mechanical ventilation and the need of re-intubation. Blood gas analyses will be performed by the clinician according with clinical needs.

Data collection Investigators will collect all the data on the dedicated CRF and will insert all the information required in the online platform. The coordinator center will directly receive all the information in a very simple data flow, with safe mechanisms for the protection of personal clinical information. The website uses an https format and all patients' data will be collected anonymously. We have also implemented regular backups, in order to minimize the risk of data corruption.

After discharge from the hospital, patients will be phone-called for the follow-up. We will record any re-admission in hospital or exitus. Follow-up will be performed 30 days, 60 days and one year after randomization.

Statistical considerations Sample size Sample-size calculation was based on a two-sided alpha error of 0.05 and a 80% power. On the basis of respiratory insufficiency incidence after cardiac surgery we anticipate a 25% of patients with a PaO2/FiO2 < 200 ratio. We expect the incidence of this parameter to be reduced of the 35%. We calculate that we will need a sample size of 263 patients per group. Including a drop-out fraction of 10%, we calculate that we will be needing 870patients to complete the trial.

Data analysis We will analyze patients in the treatment group to which they are allocated. Data will be analyzed with a professional statistical software. Data will be analyzed according to the intention-to-treat principle and following a pre-established analysis plan. Dichotomous variables will be compared with the two-tailed Χ2 test, using the Yates correction when appropriate. Continuous variables will be compared by analysis of variance or the non-parametric Kruskal-Wallis test, when appropriate. Relative risks with 95% confidence intervals and differences between medians with 95% confidence intervals (using the Hodges-Lehmann estimation) will be calculated when appropriate. Two-sided significance tests will be used throughout.

Subgroup analyses We will infer a subgroup effect if the interaction term of treatment and subgroup is statistically significant at P <0.05.