The authors measured regional cerebral 133xenon (133Xe) blood flow (rCBF), intraventricular pressure (IVP), cerebrospinal fluid (CSF) pH and lactate, systemic arterial blood pressure (SAP), and arterial blood gases during the acute phase in 23 comatose patients with severe head injuries. The IVP was kept below 45 mm Hg. The rCBF was measured repeatedly, and the response to induced hypertension and hyperventilation was tested. Most patients had reduced rCBF. No correlation was found between average CBF and clinical condition, and neither global nor regional ischemia contributed significantly to the reduced brain function. No correlation was found between CBF and IVP or CBF and cerebral perfusion pressure (CPP). The CSF lactate was elevated significantly in patients with brain-stem lesions, but not in patients with "pure" cortical lesiosn. The 133Xe clearance curves from areas of severe cortical lesions had very fast initial components called tissue peaks. The tissue peak areas correlated with areas of early veins in the angiograms, indicating a state of relative hyperemia, referred to as tissue-peak hyperemia. Tissue-peak hyperemia was found in all patients with cortical laceration or severe contusion but not in patients with brain-stem lesions without such cortical lesions. The peaks increased in number during clinical deterioration and disappeared during improvement. They could be provoked by induced hypertension and disappeared during hyperventilation. The changes in the tissue-peak areas appeared to be related to the clinical course of the cortical lesion.

Using the intra-arterial 133xenon (133Xe) method, the cerebrovascular response to acute Paco2 reduction was studied in 26 unconscious, brain-injured patients subjected to controlled ventilation. The CO2 reactivity was calculated as delta in CBF/delta Paco2. The perfusion pressure was defined as the difference between mean arterial pressure and mean intraventricular pressure. Although the CO2 reactivities did not differ significantly from that in awake, normocapnic subjects, it was low in the acute phase of injury, especially in those patients with severe outcome in whom the brain-stem reflexes were often affected. An increase of the CO2 reactivity with time was observed, indicating normal response after 1-2 weeks. Chronic hypocapnia in six unconscious patients resulted in sustained CSF pH adaptation. The question whether a delay in CSF pH adapation exerts an influence on the CO2 reactivity, and the influence of cerebral lactacidosis on the CO2 response are discussed.

In 26 unconscious patients with brain injuries, regional cerebral blood flow (rCBF) was measured with a 16-channel Cerebrograph before and after acute reduction of PaCO2. The intra-arterial 133xenon washout technique was used, and CBF was calculated regionally as initial slope index or stochastic flow. The CO2 reactivity was calculated as deltaln CBV/deltaPaCO2. In supratentorial cortical lesions, an acute fall in PaCO2 increased the homogeneity of the regional flow pattern (decrease in the standard deviation of the regional flow values), and reduced the number of focal hyperaemic regions (tissue peaks). The CO2 reactivity in tissue peak regions was generally higher than in regions without tissue peaks. In severely injured patients with a poor outcome (dementia, vegetative survival or death), inverse steal reaction was accounted for in 11% of all regions, but only in 3% of the regions in patients who survived without dementia. Inverse steal reaction was most frequently seen during the first 3 days after the trauma. In repeated CBF studies, an increase in the CO2 reactivity with time was observed after the acute trauma. In comparison with the CO2 reactivity found in normocapnic awake subjects, this increase was higher than expected in several cases.

No abstract

No abstract