Correlated, simultaneous, multiple-wavelength optical monitoring in vivo of localized cerebrocortical NADH and brain microvessel hemoglobin oxygen saturation

Abstract

Current forms of brain monitoring, such as electroencephalography (EEG), have had limited clinical utility. The EEG records spontaneous cerebrocortical activity and thus is an indirect indicator of metabolic demand and, to a lesser extent, an indicator of mismatch of supply versus demand. Ischemia modulates EEG activity in ways that can usually be detected, but EEG patterns can be similarly modulated by many other factors, including temperature and pharmacologic manipulation. This in vivo study in physiologically monitored animals evaluated the use of correlated optical spectroscopy, performed with an instrument having a fiberoptic light-guide bundle in contact with the cerebral cortex, for the simultaneous monitoring of cerebrovascular oxygen availability and intracellular oxygen delivery. A highly specific monitor of cerebral intracellular oxygen supply, the cerebrocortical intramitochondrial NADH redox state, was monitored in vivo with a fluorescence technique. Absorption spectroscopy was used concurrently to monitor hemoglobin content (blood volume) and oxygen saturation in the microcirculation. Correlated changes in optical signals from cerebrocortical NADH and hemoglobin were studied in a swine model (n=7) of nitrogen hypoxia. Measurements were made at four wavelengths with a time-division, multiplexed fluorometer/reflectometer. Because the NADH fluorescence signal at 450 nm is affected by local changes in blood volume, a “corrected” fluorescence signal is usually calculated. In previous studies, where only two wave lengths have been measured, attempts at correction were based on reflectance at the excitation wavelength (366 nm). We compared estimators of changes in microcirculatory blood volume using reflection at two wavelengths: 366 nm and 585 nm, the wavelengths for maximum and isobestic absorption. The results of the studies were as follows: (1) during transient hypoxia, NADH and local hemoglobin saturation signals changed in concert with arterial pulse oximetry, with changes in NADH lagging behind changes in saturation by an average of 5.3 seconds; (2) after hypocapnic ventilation to a mean Paco ₂ of 20.2 ± 0.8 mm Hg, NADH increased by 11.5 ± 8.7% (as compared with maximal change during anoxia), local hemoglobin saturation decreased by 7.7 ± 6.4%, and local blood volume decreased by 12.5 ± 13%, while arterial SpO₂ was unchanged; (3) our two measures of local blood volume were closely correlated during carbon dioxide perturbations, but poorly correlated during hypoxic perturbation; and (4) NADH fluorescence provided a more rapid, sensitive indicator of oxygen deprivation than did the EEG. During transient hypoxia, EEG changes occurred 57.4 ± 10.4 seconds after the onset of decline in local hemoglobin saturation, after NADH had completed 50% of its maximal increase.