Near-Infrared Spectroscopy: Shedding Light on the Injured Brain

Cerebral hypoxia/ischemia (H-I) is a major contributor to poor neurologic outcome in a variety of clinical scenarios but its detection remains problematic. Measurement of cerebral oxygenation is widely used to assess the balance between cerebral metabolic supply and demand although standard bedside methods of measuring cerebral oxygenation have significant limitations. Jugular venous oxygen saturation is a global, flow-weighted measure that may miss regional ischemia,1 whereas measurement of intraparenchymal brain tissue oxygen tension is a focal technique2 and the identification of ischemia is crucially dependent on the correct location of the probe.3 As well as being invasive, these techniques are associated with a degree of technical difficulty and are not widely available outside specialist centers. Noninvasive techniques, such as continuous electroencephalography (EEG), are increasingly being used in the intensive care setting because, in addition to identifying seizure activity, EEG changes correlate closely with cerebral metabolism and ischemia.4 Although digital processing and automated analysis is making EEG more accessible, it remains difficult to interpret. In neonatal practice, identification of H-I is particularly difficult and is often diagnosed after the event by clinical and radiological findings.5 There is therefore a need for a noninvasive, bedside monitor of cerebral well-being that can provide reliable and real-time data from several regions of the brain simultaneously. Near-infrared spectroscopy (NIRS) is a noninvasive technique that offers the potential for cerebral monitoring over multiple regions of interest. The technology is based on the transmission and absorption of near-infrared light (700–1000 nm) at multiple wavelengths as it passes through tissue. NIRS allows interrogation of the cerebral cortex using reflectance spectroscopy via optodes, light transmitting and detecting devices, placed on the scalp.6 In simple terms, oxygenated and deoxygenated hemoglobin have different absorption spectra and cerebral oxygenation and hemodynamic status can be determined by their relative absorption of near-infrared light. Earlier NIRS methodology was predominantly limited to differential spectroscopy methods that provide trend monitoring of the changes in tissue chromophore concentration (e.g., oxy- and deoxyhemoglobin).7 These variables are generally unfamiliar to clinicians, even if the changes are quantified in micromolar units. Technical developments, for example, the use of spatially resolved spectroscopy, have allowed the introduction of clinical monitors that incorporate an absolute measure of cerebral tissue oxygen saturation (ScO2), an easily accessible and continuous measure of the balance between cerebral oxygen delivery and utilization. Because NIRS interrogates arterial, venous, and capillary blood within the field of view, the derived saturation represents a “tissue” oxygen saturation measured from these three compartments8 and can be used to identify tissue H-I.9 Although trends in NIRS variables have been used as markers of cerebral well-being during the intraoperative management of cardiopulmonary bypass10 and carotid endarterectomy,11 and during the intensive care management of adult2 and neonatal brain injury,12 ScO2 thresholds for the onset of H-I are unclear. In this issue of the journal, Kurth et al.13 investigate the application of NIRS to identify ScO2 viability-time thresholds predictive of neurological outcome. Viability-time thresholds have been studied most widely in adult stroke research in which the concept of “time is brain” is increasingly recognized,14 and interventions to restore perfusion before the onset of irreversible tissue damage are associated with improved functional outcomes.15 Although the mechanisms of injury and potential interventions are different in other types of brain injury, it is likely that similar windows for targeted intervention can be identified.16,17 A reliable monitor of H-I, in association with the ability to determine viability-time thresholds, would clearly have wide clinical application. The study by Kurth et al.13 used a survival piglet model to determine the temporal development of brain damage when ScO2 was maintained at 35%, a threshold previously demonstrated by the same group to be associated with neurophysiologic impairment.9 The authors concluded that H-I lasting for 2 hr or less was not associated with subsequent neurologic deficit, whereas for longer episodes of H-I, the incidence of neurologic injury increased by about 15% per hour and was heralded by abnormalities in NIRS variables during reperfusion. These findings indicate that not only is it possible to define a viability-time threshold using ScO2 but also that there is a several hour window of opportunity during severe H-I that might be used to deliver targeted neuroprotective strategies and potentially prevent or minimize subsequent neurologic injury. The possibility that NIRS might identify treatment windows before irreversible tissue changes have occurred, and that changes during reperfusion might also predict outcome, is an exciting prospect. Defining the degree and duration of reduction in ScO2 associated with the development of permanent neurologic injury is an important clinical goal and the study by Kurth et al.13 is a valuable contribution to this debate. Al-Rawi et al.18 previously attempted to determine the reduction in tissue oxygenation index, a measure of ScO2 using a NIRO 300 spectrometer (Hamamatsu Photonics, Japan), associated with cerebral ischemia. EEG was used to define the presence of cerebral ischemia after carotid artery clamping during carotid surgery and no patient with a reduction in tissue oxygenation index <13% developed EEG evidence of ischemia. Other studies that have used NIRS to evaluate cerebral oxygenation status and sufficiency have concluded that it is not easily possible to determine accurate reference ranges for normal subjects or ischemic thresholds.19 One reason for this might be the significant baseline variability of ScO2 demonstrated in most studies. The “normal” range varies between 60% and 75%, with a coefficient of variation for absolute baseline values of almost 10%.20 It is also important to note that ScO2 thresholds for the onset of H-I are likely to be individual and disease specific.21 The lack of standardization of currently available NIRS devices also contributes to the difficulty in defining thresholds. Although different devices use different nomenclature, most do measure ScO2 in some form and display a simple percentage value. However, the algorithms used to assess cerebral oxygenation, and even the variables actually measured, vary between devices and among studies.22 It is important to understand these limitations when making comparisons between studies and between devices. Kurth et al.13 also highlight the importance of cellular energy failure as a component of the pathophysiology of acute brain injury. The exact etiology of this energy failure is poorly understood, but both reduced substrate delivery below critical thresholds and impaired mitochondrial substrate use appear to be implicated.23 Cytochrome c oxidase (CCO) is the terminal complex of the electron transfer chain responsible for over 95% of oxygen metabolism. NIRS-derived CCO has been validated in animal studies as a measure of changes in cellular energy status24 and offers the potential to assess cerebral mitochondrial redox state and adequacy of oxygen delivery and use after brain injury in humans.2,8 Although there are enormous challenges when using NIRS as a clinical monitor of CCO (particularly in adults), recent NIRS systems that have been optimized for CCO measurements are providing promising results.8,25,26 NIRS-measured changes in the oxidation status of CCO have been shown to correlate with estimated changes in cerebral oxygen delivery during hypoxemia in healthy adult humans,25 and NIRS has been used to demonstrate oxidation in cerebral mitochondrial redox state during normobaric hyperoxia in adult head-injured patients.26 The further development of technology capable of reliably measuring changes in CCO offers the real potential for a single NIRS-based device to provide multisite, regional monitoring, not only of cerebral oxygenation and hemodynamics but also of cerebral cellular energy status. An essential component of the translation of NIRS systems and methods into the clinical arena is an understanding of the physiologic basis of the measured NIRS signals. In this context, it is important to note that, in the case of both ScO2 and CCO, there are no obvious “gold standard” measurements against which a direct experimental validation can be performed. However, mathematical models of cerebral circulation and energy metabolism have been developed, which can be used to interpret NIRS-derived data and maximize their clinical usefulness.27 There are several concerns over the clinical application of NIRS, not least the potential for “contamination” of the NIRS signal from extracerebral sources.22 The wider application of spatially resolved spectroscopy, which has been shown in adults to have high sensitivity and specificity to intracranial changes,28 goes some way to resolving this issue. Furthermore, because mitochondrial density is higher in brain than in tissues with lower metabolic rates, such as the skull and skin, NIRS techniques incorporating CCO measurement will also be less prone to extracerebal contamination. The efficacy of brain oxygenation monitoring using NIRS has also been questioned because of the absence of abnormalities in NIRS variables in patients undergoing cardiac surgery who are subsequently shown to have developed intraoperative ischemic brain injury.29 Current commercially available NIRS devices are usually designed to be placed on the forehead and, as with other regional monitoring techniques, it is impossible to detect changes in areas located distant from the monitored site although global cerebral oxygen sufficiency can be evaluated.30 However, multiprobe NIRS devices are now available that provide regional measurements from multiple areas simultaneously. Although modern NIRS systems already provide high temporal and spectral resolution, continued instrumentation and algorithm development is required to provide monitors suitable for widespread clinical application. Further investigations, including large randomized controlled trials, are also required to establish the clinical efficacy of NIRS, particularly to determine its role in providing robust markers of thresholds for H-I. Of equal importance is the incorporation of continuous, noninvasive, and multisite NIRS measurement of cerebral oxygenation, hemodynamics, and cellular energy status into multimodal monitoring strategies in brain injury.