Do Cerebral Cortex Perfusion, Oxygen Delivery, and Oxygen Saturation Responses Measured by Near-Infrared Spectroscopy Differ Between Patients Who Fail or Succeed in a Spontaneous Breathing Trial? A Prospective Observational Study

Background: Alterations in perfusion to the brain during the transition from mechanical ventilation (MV) to a spontaneous breathing trial (SBT) remain poorly understood. The aim of the study was to determine whether changes in cerebral cortex perfusion, oxygen delivery (DO₂), and oxygen saturation (%StiO₂) during the transition from MV to an SBT differ between patients who succeed or fail an SBT.

Methods: This was a single-center prospective observational study conducted in a 16-bed medical intensive care unit of the University Hospital Leuven, Belgium. Measurements were performed in 24 patients receiving MV immediately before and at the end of a 30-min SBT. Blood flow index (BFI), DO₂, and %StiO₂ in the prefrontal cortex, scalene, rectus abdominis, and thenar muscle were simultaneously assessed by near-infrared spectroscopy using the tracer indocyanine green dye. Cardiac output, arterial blood gases, and systemic oxygenation were also recorded.

Results: During the SBT, prefrontal cortex BFI and DO₂ responses did not differ between SBT-failure and SBT-success groups (p > 0.05). However, prefrontal cortex %StiO₂ decreased in six of eight patients (75%) in the SBT-failure group (median [interquartile range 25-75%]: MV = 57.2% [49.1-61.7] vs. SBT = 51.0% [41.5-62.5]) compared to 3 of 16 patients (19%) in the SBT-success group (median [interquartile range 25-75%]: MV = 65.0% [58.6-68.5] vs. SBT = 65.1% [59.5-71.1]), resulting in a significant differential %StiO₂ response between groups (p = 0.031). Similarly, a significant differential response in thenar muscle %StiO₂ (p = 0.018) was observed between groups. A receiver operating characteristic analysis identified a decrease in prefrontal cortex %StiO₂ > 1.6% during the SBT as an optimal cutoff, with a sensitivity of 94% and a specificity of 75% to predict SBT failure and an area under the curve of 0.79 (95% CI: 0.55-1.00). Cardiac output, systemic oxygenation, scalene, and rectus abdominis BFI, DO₂, and %StiO₂ responses did not differ between groups (p > 0.05); however, during the SBT, a significant positive association in prefrontal cortex BFI and partial pressure of arterial carbon dioxide was observed only in the SBT-success group (SBT success: Spearman's ρ = 0.728, p = 0.002 vs. SBT failure: ρ = 0.048, p = 0.934).

Conclusions: This study demonstrated a reduced differential response in prefrontal cortex %StiO₂ in the SBT-failure group compared with the SBT-success group possibly due to the insufficient increase in prefrontal cortex perfusion in SBT-failure patients. A > 1.6% drop in prefrontal cortex %StiO₂ during SBT was sensitive in predicting SBT failure. Further research is needed to validate these findings in a larger population and to evaluate whether cerebral cortex %StiO₂ measurements by near-infrared spectroscopy can assist in the decision-making process on liberation from MV.