Neurological and respiratory effects of lung protective ventilation in acute brain injury patients without lung injury: brain vent, a single centre randomized interventional study

Abstract

Introduction: Lung protective ventilation (LPV) comprising low tidal volume (VT) and high positive end-expiratory pressure (PEEP) may compromise cerebral perfusion in acute brain injury (ABI). In patients with ABI, we investigated whether LPV is associated with increased intracranial pressure (ICP) and/or deranged cerebral autoregulation (CA), brain compensatory reserve and oxygenation.

Methods: In a prospective, crossover study, 30 intubated ABI patients with normal ICP and no lung injury were randomly assigned to receive low VT [6 ml/kg/predicted (pbw)]/at either low (5 cmH₂O) or high PEEP (12 cmH₂O). Between each intervention, baseline ventilation (VT 9 ml/kg/pbw and PEEP 5 cmH₂O) were resumed. The safety limit for interruption of the intervention was ICP above 22 mmHg for more than 5 min. Airway and transpulmonary pressures were continuously monitored to assess respiratory mechanics. We recorded ICP by using external ventricular drainage or a parenchymal probe. CA and brain compensatory reserve were derived from ICP waveform analysis.

Results: We included 27 patients (intracerebral haemorrhage, traumatic brain injury, subarachnoid haemorrhage), of whom 6 reached the safety limit, which required interruption of at least one intervention. For those without intervention interruption, the ICP change from baseline to "low VT/low PEEP" and "low VT/high PEEP" were 2.2 mmHg and 2.3 mmHg, respectively, and considered clinically non-relevant. None of the interventions affected CA or oxygenation significantly. Interrupted events were associated with high baseline ICP (p < 0.001), low brain compensatory reserve (p < 0.01) and mechanical power (p < 0.05). The transpulmonary driving pressure was 5 ± 2 cmH₂O in both interventions. Partial arterial pressure of carbon dioxide was kept in the range 34-36 mmHg by adjusting the respiratory rate, hence, changes in carbon dioxide were not associated with the increase in ICP.

Conclusions: The present study found that most patients did not experience any adverse effects of LPV, neither on ICP nor CA. However, in almost a quarter of patients, the ICP rose above the safety limit for interrupting the interventions. Baseline ICP, brain compensatory reserve, and mechanical power can predict a potentially deleterious effect of LPV and can be used to personalize ventilator settings. Trial registration NCT03278769 . Registered September 12, 2017.

Keywords: Acute brain injury; Cerebral autoregulation; Intracranial pressure; Lung protective ventilation; Positive end-expiratory pressure; Subarachnoid haemorrhage; Transpulmonary pressure; Traumatic brain injury.

Fig. 1

Study design: crossover design and intervention settings. Panel A shows the ventilator settings in the different study periods. Panel B shows the randomisation scheme and the study periods. During the baselines, patients were ventilated with VT 9 ml/kg/pbw and PEEP 5 cmH₂0. Group A was first exposed to “low VT/low PEEP” followed by “low VT/high PEEP”. Group B was subjected to the same interventions in reverse order. The length of the periods was chosen to allow PRx to stabilize and thus display less variance. A washout period with baseline ventilator settings was used to reduce the risk for carry-over effect of the intervention. VT, tidal volume; PEEP, positive end expiratory pressure; h, hours; ml/pbw, milliliter per kg predicted body weight

Fig. 2

CONSORT flow diagram for crossover study. VT, tidal volume; PEEP, positive end expiratory pressure; ICP, intracranial pressure

Fig. 3

Mean ICP and PRx induced changes. The interrupted interventions were presented in red. The value displayed is the mean value before interruption of the intervention, hence not entirely representative or comparable with not interrupted interventions. Panel (A). Mean ICP did not have a clinically important rise as compared with baseline, neither following the “Low VT/low PEEP” intervention nor during “Low VT/high PEEP” (one tailed paired t-test for noninferiority, both interventions p = 0.99). Similar results were achieved when including patients in whom the interventions were interrupted. In six of 27 (22%) patients, “Low VT/high PEEP” alone or both interventions were interrupted because the safety limit of ICP > 22 mmHg was reached. In one patient only the “low VT/low PEEP” intervention was interrupted. This patient had an increase in ICP over time, which might explain why the first attempted intervention “low VT/high PEEP” could be tolerated. Panel (B). None of the interventions produced significant changes from baseline in mean PRx, neither during intervention “Low VT/low PEEP” (paired t-test, p = 0.56) nor during intervention “Low VT/high PEEP” (p = 0.50). Similar results were achieved when including patients in whom the interventions were interrupted. VT, tidal volume; PEEP, positive end expiratory pressure; ICP, intracranial pressure; PRx, pressure reactivity index

Fig. 4

Differences in ICP (A), RAP (B), EtCO₂ (C), MP (D), TPP_ei (E) and TPP_ee (F) at baseline preceding interrupted and completed interventions. The data are presented with violin plots and boxplots. A. ICP (n = 8, median ICP = 17.1 (15.9 -19.7) mmHg) at the baseline in the interrupted interventions vs. completed interventions (n = 43, median ICP = 10.0 (4.6–13.3) mmHg). B. RAP at the baseline in the interrupted (n = 8, median RAP = 0.97 (0.86–1.18) vs. the completed (n = 41, median RAP = 0.42 (0.24–0.80) interventions. C. EtCO₂ at baseline in interrupted (n = 8, median EtCO₂ = 29.2 (25.5–30.9) mmHg) versus completed (n = 43, median EtCO₂ = 31.1 (29.5–32.8) mmHg) interventions. D. MP at the baseline in interrupted (n = 7, median MP = 12.1 (10.1–13.0) J/min) vs. completed (n = 33, median MP = 8 (6.4—9.3) J/min) interventions. E and F. Neither TPP_ei and TPP_ee were significantly different between completed and interrupted events. ICP, intracranial pressure; RAP, compensatory reserve index; EtCO₂, end-tidal partial pressure of CO₂; MP, mechanical power; TPP_ei, transpulmonary pressure end inspiration, absolute; TPP_ee, transpulmonary pressure end expiration, absolute