Physiological Effects of High-Flow Tracheal Oxygen vs. Pressure Support Ventilation in Neurosurgical Patients with Brain Tumors Facing Delayed Extubation: A Pilot Study.

Background

Postoperative neurosurgical patients who successfully pass a spontaneous breathing trial (SBT) may still require delayed extubation due to impaired airway protection. Continued mechanical ventilation may contribute to diaphragm underuse. High-flow tracheal oxygen (HFTO) offers an alternative-discontinuing positive pressure while maintaining airway access-but its physiological effects in intubated neurosurgical patients remain poorly characterized.

Methods

In this prospective, randomized physiological pilot study, postoperative neurosurgical patients who required delayed extubation were assigned (1:1) to HFTO or low-level pressure support ventilation (PSV). Esophageal pressure monitoring and diaphragmatic ultrasound were performed at baseline and at 2, 12, 24, and 48 h. The primary endpoint was inspiratory effort assessed by esophageal pressure-time product (PTPes) per min. Secondary outcomes included diaphragmatic function, respiratory parameters, safety, and clinical outcomes.

Results

A total of 24 patients were randomized; 1 patient was withdrawn due to intracranial rebleeding, leaving 23 patients for analysis (HFTO, n = 12; PSV, n = 11). HFTO exhibited a significantly different temporal trajectory in inspiratory effort (group-by-time interaction P = 0.019), characterized by an acute physiological step-up immediately after withdrawing positive pressure support, followed by a stable plateau from 2 to 48 h. Diaphragmatic indices remained unchanged in both groups over 48 h. Airway pressures were significantly lower in the HFTO group (all P < 0.001), accompanied by modest declines in partial pressure of arterial oxygen (PaO2)/fraction of inspired oxygen (FiO2) and slight partial pressure of arterial carbon dioxide (PaCO2) increases. Respiratory deterioration rate was similar between groups (25.0% vs. 18.2%, P = 1.000). Ventilator-free days at day 28 were univariately higher with HFTO (27.0 [23.0-27.0] vs. 21.0 [17.0-24.0] days, P = 0.022); artificial airway-free days and ICU length of stay were comparable between groups.

Conclusions

HFTO was univariately associated with increased inspiratory effort compared with low-level PSV, without short-term diaphragmatic harm or increased clinical instability in postoperative neurosurgical patients requiring delayed extubation. These findings suggest HFTO as a feasible strategy for liberation from mechanical ventilation while maintaining airway safety. Larger multicenter studies are warranted.