Relationship between Brain Tissue Oxygen Tension and Transcranial Doppler Ultrasonography

Abstract

Multimodal monitoring of intracranial pressure and brain tissue oxygen tension (PbtO2) have been increasingly used to detect delayed cerebral ischemia (DCI) after subarachnoid hemorrhage. At our center, patients who cannot be easily assessed clinically will undergo intracranial pressure and PbtO2 monitoring via a NEUROVENT-PTO bolt. We aimed to determine whether the Lindegaard ratios (LRs) computed from transcranial Doppler ultrasonography (TCDU) would correlate with, or can predict, the simultaneously recorded PbtO2 value. Patients with aneurysmal subarachnoid hemorrhage, PbtO2 recordings from the middle cerebral artery territory, and simultaneous TCDU scans available from the ipsilateral middle cerebral artery and internal carotid artery from August 2018 to 2019 were included in the present study. The index test result was vasospasm (LR of ≥3) found on TCDU. The reference standard was the presence of regional hypoxia (PbtO2 <20 mm Hg). The PbtO2 results were compared with those from computed tomography angiography as a radiological standard. The predictive values were calculated using a contingency table and receiver operating characteristic curve. A total of 28 patients (6 men and 22 women; age, 59.04 ± 13.75 years) were identified with simultaneous brain tissue oxygen and TCDU recordings available. Of the 28 patients, 7 had cerebral hypoxia (PbtO2 <20 mm Hg). We found no correlation between the PbtO2 measurements and simultaneously recorded LRs (r2= 0.048; P= 0.26). A LR of ≥3 had high specificity (95.24%) for hypoxia but relatively low sensitivity (42.86%; P= 0.037). We find TCDU to be specific for predicting cerebral hypoxia (measured via an intraparenchymal probe). Therefore, it could be a useful and noninvasive tool in the context of preventative DCI monitoring. However, given the low sensitivity, the lack of vasospasm on TCDU should not preclude the possibility of the presence of evolving DCI.