Abstract SNACC-87

Return to Poster Listing

Continuous Non-Invasive Measurement of Cerebral Blood Flow, Cerebral Metabolic Rate for Oxygen and Oxygen Extraction Fraction in Critically Ill Brain Injured Patients.

Chandra M, Balu R, Yodh A, Frangos S, Park S, Kofke W
University of Pennsylvania, Philadelphia, PA, USA

INTRODUCTION: Prevention of secondary ischemic brain injury by maintaining perfusion and oxygenation is a core principle of neurocritical care and neuroanesthesia. Identifying episodes of reduced cerebral perfusion in brain-injured patients remains a challenge. Previous reports by our group indicate that acute changes in regional cerebral blood flow (rCBF) by diffuse correlation spectroscopy(DCS) correlate with changes on stable XeCTCBF(1). We used DCS, a novel non-invasive optical method that measures light scattering by moving particles in optically thick tissues, combined with static near infrared spectroscopy (NIRS) to provide real time measurements of relative changes in rCBF, cerebral metabolic rate for oxygen (rCMRO2), and oxygen extraction fraction (rOEF) in patients with severe brain injury.

METHODS: We recruited 16 patients with acute subarachnoid hemorrhage (SAH), severe traumatic brain injury (TBI) or primary intracerebral hemorrhage (ICH) from a single NeuroICU. Assumptions in calculating CMRO2 and OEF were that baseline values for total Hb=50uM and StO2=55%. Relative CBF and tissue oxygen saturation (StO2) in both hemispheres were measured for ~2 hours daily for up to 7 consecutive days using a custom built, NIRS-DCS apparatus. For each recording epoch, relative changes in OEF were calculated based on StO2 and peripheral arterial O2 saturation, and time-varying changes in CMRO2 were calculated by multiplying rOEF by rCBF.

RESULTS: Overall, rCMRO2 was highly correlated with rCBF, suggesting that for the majority of time points changes in rCBF matched the metabolic demands of brain tissue. There were multiple instances, however, where we identified abrupt decreases in the correlation between rCMRO2 and rCBF; these instances occurred during abrupt changes in OEF. The strength of the correlation between rCMRO2 and rCBF varied with initial assumed values for mean total hemoglobin (THB) and mean tissue oxygenation (StO2). Time points with abrupt increases in rOEF and time matched reductions in rCMRO2/CBF correlation may correspond to acute ischemic events, even though these events were not always accompanied by reductions in rCBF.

CONCLUSION: Our observation of changes in rCBF associated with corresponding changes in CMRO2 is not unexpected but does raise the possibility that current notions on the value of CBF-MAP correlations may be flawed, absent consideration of metabolic rate. We demonstrate that combined non-invasive real time measurements of rCBF, rOEF and rCMRO2 in brain-injured patients is feasible in the NCCU population and may be particularly useful in identifying episodes of compromised tissue perfusion or of dysautoregulation. Supported by NIH: 5R21NS61074-2
(1) Kim MN, etal: Neurocrit Care. 2010 Apr;12(2):173-80.

  • SNACC-87 Image 1

Back to Top