Brain tissue oxygen reactivity: clinical implications and pathophysiology

It is generally accepted that PbO2 reflects the balance between O2 delivery and consumption (Diringer et al., 2007; Diringer, 2008). However, implementation in the perioperative period of various ventilatory modes using high FiO2 leads to a dramatic and non-physiologic increase in PbO2 with approximating levels of 147 ± 36 mmHg (McLeod et al., 2003). This phenomenon doesn't correlate with the extent of slight increase in arterial O2 content. At the same time, the jugular venous PO2 increases only slightly (37–40 mmHg) (Forkner et al., 2007). Moreover, hyperoxia does not affect significantly the regional CBF, and there is no improvement in cerebral metabolism with oxygen therapy (Magnoni et al., 2003; Diringer et al., 2007; Diringer, 2008; Xu et al., 2012). 
 
The PbO2 increase is more pronounced in edematous (but not necrotized) brain tissues compared to normal areas (Meixensberger et al., 1993). Although, this can be considered a positive phenomenon, it masks the real state of rCBF and local oxidative metabolism. Recording of high PbO2 absolute values may create a false impression of safety and negatively impact the clinical decision making. Apparently, better indicators of the status of energy exchange in the brain tissue are needed for practical use in the perioperative and critical care settings.


INTRODUCTION
It is generally accepted that PbO 2 reflects the balance between O 2 delivery and consumption (Diringer et al., 2007;Diringer, 2008). However, implementation in the perioperative period of various ventilatory modes using high FiO 2 leads to a dramatic and non-physiologic increase in PbO 2 with approximating levels of 147 ± 36 mmHg (McLeod et al., 2003). This phenomenon doesn't correlate with the extent of slight increase in arterial O 2 content. At the same time, the jugular venous PO 2 increases only slightly (37-40 mmHg) (Forkner et al., 2007). Moreover, hyperoxia does not affect significantly the regional CBF, and there is no improvement in cerebral metabolism with oxygen therapy (Magnoni et al., 2003;Diringer et al., 2007;Diringer, 2008;Xu et al., 2012).
The PbO 2 increase is more pronounced in edematous (but not necrotized) brain tissues compared to normal areas (Meixensberger et al., 1993). Although, this can be considered a positive phenomenon, it masks the real state of rCBF and local oxidative metabolism. Recording of high PbO 2 absolute values may create a false impression of safety and negatively impact the clinical decision making. Apparently, better indicators of the status of energy exchange in the brain tissue are needed for practical use in the perioperative and critical care settings.

BRAIN TISSUE OXYGEN REACTIVITY: CLINICAL IMPLICATIONS
Dynamic assessment of relative changes in brain oxygenation to monitor the brain functionality is a better approach compared to relying on a single parameter. With such monitoring, both the current status of brain tissue oxygenation and the functional reserve capabilities can be accomplished.
Brain tissue oxygen reactivity (BTOR) is the measure (in percents) of PbO 2 changes relative to changes in PaO 2 ( PbO 2 / PaO 2 ) with oxygen inhalation (Johnston et al., 2003). The latter parameter can be easily adjusted to reach BTOR optimal values. The technique of measurement includes increasing the FiO 2 up to 1.0 with simultaneous recording of the PaO 2 and PbO 2 values.
Literature reports indicate that high BTOR values within the first 24 h after TBI are considered an indicator of unfavorable outcome and negatively correlate with the Glasgow Outcome Score (van Santbrink et al., 1996;Menzel et al., 1999).
It is not mandatory to apply the maximal FiO 2 of 1.0 to calculate the BTOR. Any other high inspired O 2 levels can be applied that will produce significant PbO 2 changes within 20 min. Such a time period is considered the minimal required interval adequate for equilibration and meaningful assessment. During this short period, the respiration, regional metabolism and the rCBF are assumed to remain stable, and the calculated values of PbO 2 / PaO 2 will indirectly characterize the rCBF.
Low BTOR is considered a positive phenomenon even when the absolute PbO 2 values decrease, unless regional hypoperfusion (<20 ml/100 g/min) exists (Hlatky et al., 2008). Simultaneous elevations of PbO 2 and PbO 2 / PaO 2 values reflect the imbalance between the oxygen delivery and consumption.
Under normal cardio-respiratory conditions, when the right to left pulmonary shunting is negligible, the FiO 2 is proportional to PaO 2 . On the other hand, PbO 2 itself correlates with PaO 2 . Therefore, one can presume that FiO 2 is proportional to PbO 2 . Taking this into account, the formula used to calculate the BTOR can be modified to evaluate the correlation between the changes in PbO 2 and FiO 2 . This new parameter ( PbO 2 / FiO 2 ) is considered an equivalent of BTOR and can be easily calculated. This is a simple and practical approach to BTOR assessment that can be readily used at bedside. Such an approach will allow for dynamic assessment of tissue oxygen reactivity.

BTOR: PATHOPHYSIOLOGY
In order to illustrate the importance of BTOR as an ultimate indicator of balance between the rCBF, oxygen delivery and consumption and justify the need for its monitoring, the hypothesis of hyperreactive, non-physiologic, luxurious PbO 2 elevation is proposed.
We hypothesize that the significant increase of PbO 2 with hyperoxia in the injured brain is explained by an excessive right shift of the oxyhemoglobin dissociation curve with resultant significant reduction in hemoglobin's affinity to oxygen molecules at the microcirculatory level. This is a result of a mismatch between the rCBF and existing cerebral metabolic rate of oxygen (CMRO 2 ), which leads to accumulation of CO 2 , converted by erythrocyte carboanhydrase into HCO − 3 and H + ions (at a 1000 times faster rate compared to plasma and extracellular space). It is known that CO 2 and H + , which are produced during the tissue metabolism, are heterotropic effectors of hemoglobin that enhance oxygen release (Berg et al., 2002). The latter ions bind to hemoglobin with release of oxygen. With decrease in rCBF and/or relative increase of CMRO 2 , hemoglobin gets saturated with protons and practically loses its affinity to oxygen in the microcirculatory bed.
The role of CO 2 -induced local increase of PO 2 is particularly important in the brain tissue where, under normal conditions (glucose-dependant metabolism without chronic fasting), the respiratory quotient equals 1 and the CO 2 production almost 1.25 times exceeds that of the other tissues.
According to the above mentioned considerations, the rCBF determines PbO 2 values via two principal mechanisms: (a) as an oxygen delivery mechanism within the arterial compartment and (b) via a "non-physiologic" right shift of the oxyhemoglobin dissociation curve as a result of decreased removal rate of the flowdependent metabolites in the microcirculatory bed.
Many drugs and techniques used commonly during therapy of severe TBI, including manitol, sodium thiopenthal, ketorolac, nimodipine, intra-arterial papaverine, hypothermia, deep sedation, etc., can reduce the PbO 2 in the damaged tissue (Steiner et al., 2001;Gupta et al., 2002;Stiefel et al., 2004Stiefel et al., , 2006Sakowitz et al., 2007). On the other hand, the effects of medically induced augmentation of cerebral perfusion pressure on cerebral oxygenation are difficult to predict (Sahuquillo et al., 2000;Imberti et al., 2002;Le Roux and Oddo, 2013). In addition, Zygun et al. (2009) showed that even though transfusion of packed red blood cells in TBI patients may improve the brain tissue oxygenation, it won't have an appreciable effect on cerebral metabolism (Zygun et al., 2009). Thus, there is a complex interaction of multiple factors influencing the functional and metabolic activity of the injured brain including injury-related pathological mechanisms, drugs and methods used to manage these patients. Their overall effects are not straightforward and cannot be anticipated easily in an individual case. Apparently, Monitoring of PbO 2 in these patients will not provide reliable feedback and may be misleading in some cases. It is not justified to treat the severe TBI patients relying only on the PbO 2 as an indicator of adequacy of cerebral metabolism. Instead, dynamic oxygen reactivity should be routinely monitored as an indicator of overall brain tissue oxygenation and metabolism.

CALCULATIONS
Assuming CBF and CMRO 2 stability during oxygen therapy and equivalence of PbO 2 with capillary PO 2 , (Kett-White et al., 2002) we can modify the standard formula for calculation of arteriovenous difference in oxygen (Kett-White et al., 2002) to determine the changes in hemoglobin saturation in the capillary blood: where Sv. a O 2 and Sv. b O 2 are oxygen saturation at distal microcirculatory level after and before inhalation of oxygen; Ct a O 2 (a) and Ct b O 2 (a) are arterial oxygen content values after and before initiating oxygen therapy; Pb a O 2 and Pb b O 2 are PbO 2 values after and before starting inhalation of oxygen; and Hb is hemoglobin concentration in g/dL.
For example, if we increase PbO 2 from P 50 = 35 mmHg (if hemoglobin saturation is 0.5 or 50%) to 100 mmHg and assume a change in CtO 2 (a) equal to 1 vol. %, the hemoglobin saturation at distal microcirculatory level will change in the following way (assuming a hemoglobin concentration 12 g/dL): /1.34x12 = 0.05 or 5 % This means that the distal microcirculatory oxygen saturation under these arterial conditions (PbO 2 = 100 mmHg) will only increase 50% + 5% = 55%.
Calculations show the weak affinity of hemoglobin to oxygen under these conditions which results in allocation of additional oxygen amounts out of hemoglobin with creation of abnormally high PbO 2 in injured brain tissue areas.

Monitoring of BTOR or its equivalent
PbO 2 / FiO 2 is indicated during the intensive therapy of TBI patients. Both indices reflect the actual status of cerebral oxidative metabolism and help to reduce the risk of management errors which are otherwise masked by high FiO 2 -induced "adequate" PbO 2 absolute values.
Blood transfusions, controlled hyperventilation and restoration of the regional acid-base balance should be performed under the guidance of above mentioned indices.
Further studies will help to establish the role of BTOR and PbO 2 / FiO 2 monitoring in assessment of metabolic changes and adaptations taking place in the injured brain during the acute phase of TBI.

DISCLOSURE
The authors did not receive any financial or other support related to this work.