The impact of hyperoxia on outcome has been studied in a number of experimental models of head injury. In a brain contusion model, produced by the application of negative pressure to the cortex, normobaric and hyperbaric hyperoxia reduced evidence of apoptosis in peri-lesional brain (22
A number of experimental and clinical studies have investigated the impact of normobaric hyperoxia in TBI. Initial studies that utilized microdialysis to assess the metabolic response in TBI reported reduced lactate levels. (23
) These data were interpreted to indicate reduced production of lactate due to a shift from anaerobic to aerobic metabolism. However interpretation of lactate levels is far more complex as they may be influenced by a number of factors. CBF determines lactate clearance, independent of production. Thus if CBF were to rise and lactate production was unchanged levels would fall. Furthermore, white blood cells produce lactate. Infiltration by inflammatory cells occurs in TBI (26
) and the movement of these cells in and out of the brain could influence lactate levels. Finally lactate may be used as fuel by neurons (27
) complicating the interpretation of interstitial levels.
Interpreting levels of lactate in relation to pyruvate (lactate/pyruvate ratio) provides an assessment of the brain's redox state. (29
) Elevated ratios are considered to be indicative of ischemia (31
) Administration of normobaric hyperoxia to TBI patients has yielded inconsistent results, demonstrating either no change or a fall in the lactate/pyruvate ratio (23
). Still these measurements represent only two components of the brain's extraordinarily complex metabolic machinery.
The ability of normobaric hyperoxia to improve brain metabolism has been studied primarily using indirect physiologic measures, the interpretation of which remains disputed. Recently, direct measurement of brain metabolism was performed using PET in a small group of patients with severe TBI (34
). Cerebral blood flow, blood volume and CMRO2
were measured during baseline ventilation with 40% oxygen and again after one hour of ventilation with 100% oxygen. Cerebral metabolic rate for oxygen did not change with hyperoxia indicating that the improved values of PbrO2
, lactate and lactate/pyruvate ratio do not necessarily indicate improved utilization of oxygen.
A recent study compared the impact of 24 hours of 100% oxygen on 52 severe TBI patients to historical control. (33
) The groups were matched based on age, sex, post resuscitation GCS and ICP during the first 12 hours post injury. Both groups were monitored with microdialysis and brain tissue O2
sensors. The results mirror those of previous studies. With hyperoxia ICP was lower, dialysate glucose levels rose, lactate fell and the lactate/pyruvate ratio declined. The magnitude of the difference in ICP is not clinically significant (12 vs. 15 mm Hg) and the implications of the biochemical changes are unknown. None of these changes were seen the normoxia group. At 6 months after injury the mean Glasgow Outcome Score with hyperoxia was higher than in the control group (3.2 compared with 2.8), but these differences were not statistically significant.
Thus the data on the impact of normobaric hyperoxia in TBI almost all use indirect measures of the impact on brain oxygen metabolism. Interpretation of the measures used is controversial an depending on perspective of the interpreter the same data can be said to support or refute improved brain oxygen metabolism with hyperoxia (12
). It has been suggested that this controversy could be resolved by direct measurement of cerebral metabolic rate for oxygen with PET or brain high energy phosphates with MRI. Such data are now available in a recent PET study. (34
) Patients with acute severe TBI were studied within 24 hours of injury under conditions of stable cerebral perfusion pressure, hemoglobin and arterial pressure of carbon dioxide. Although the number of patients studied was small, the results were consistent: CMRO2
did not change when inspired oxygen concentration was raised from ~40 to 100% for one hour. These data suggest hyperoxia does not produce the desired effect, increased brain oxygen utilization. They also indicate that indirect measures of oxygen metabolism should be interpreted with caution and cannot reliably predict a change in CMRO2