Identification of Posttraumatic Ischemia and Hyperperfusion by Determination of the Effect of Induced Arterial Hypertension on Carbon Dioxide Reactivity
Background and Purpose Both ischemia and hyperperfusion are known phenomena that follow traumatic brain injury. Cerebral carbon dioxide reactivity is diminished in both conditions. Differentiation is important because ischemia is thought to be a major factor of secondary neuronal loss and is potentially amenable to therapy by manipulation of cerebral perfusion pressure.
Methods The response of transcranial Doppler–based carbon dioxide reactivity to pharmacologically induced hypertension was studied sequentially in 29 patients with severe to moderate head injury to identify ischemia and luxury perfusion. After simultaneous baseline registration of the carbon dioxide reactivity of both middle cerebral arteries by two-channel transcranial Doppler, systolic arterial pressure was raised approximately 20 mm Hg by means of phenylephrine (Neosynephrine) infusion, and the carbon dioxide reactivity test was repeated.
Results A significant improvement of impaired (<2%/mm Hg) carbon dioxide reactivity after arterial pressure was raised by 20 mm Hg (signaling ischemia) was found in 32 of 124 evaluated middle cerebral arteries. Further deterioration of impaired reactivity occurred in only four tested hemispheres. While ischemic conditions were identified during the entire study period, hyperperfusion was encountered only after day 3.
Conclusions Ischemia after traumatic brain injury is a frequent phenomenon, whereas hyperperfusion is rare. Whether therapeutic optimization of carbon dioxide reactivity can improve the outcome of head-injury patients needs to be evaluated in further studies.
Recent hemodynamic studies have shown that both high and low cerebral perfusion are common events during the first few days after severe head injury.1 2 3 4 5 6 7 8 9 Low flow conditions can either be due to true ischemia or merely be secondary to neuronal functional depression as a consequence of the trauma or medications such as barbiturates.10 Posttraumatic true ischemia is thought to be a significant cause of secondary neuronal loss. Hyperperfusion, on the other hand, may be equally detrimental by causing edema and hemorrhage.11 Recording of the oxygen saturation in the jugular bulb has recently been proposed as a means to identify global ischemia and hyperperfusion.8 12 13 14 15 In the present study, we performed TCD-based CO2 reactivity tests at two different arterial pressures to identify ischemic and hyperemic conditions. Improvement of vasoreactivities after pharmacologically increasing arterial pressure is expected in ischemic situations, and vice versa, worsening is expected with hyperperfusion.
Subjects and Methods
The investigations were performed in 29 patients with severe and moderate head injury admitted to our institution between August 1992 and October 1993. Patients with a lucid posttraumatic interval and secondary loss of consciousness due to progressive intracranial hematoma were excluded, as well as patients showing a pure extradural hematoma on CT. The frequency distribution with respect to initial scoring on the GCS and GOS is given in Fig 1⇓. The study was approved by the institutional ethics committee and conducted according to the respective guidelines.
The general procedures applied to trauma patients at our institution have been outlined before.9 In short, cranial CT was carried out after cardiopulmonary resuscitation and stabilization, and possible space-occupying hematomas were evacuated via craniotomy. The bone flap in patients undergoing craniotomy was not replaced in the acute stage. An epidural pressure probe (Gaeltec Ltd) was implanted in all patients of this series. Treatment in the intensive care unit aimed at maintaining ICP below 20 to 25 mm Hg and CPP above 70 mm Hg. Patients were positioned with their heads 30° elevated. Moderate hyperventilation to a Pco2 of 30 to 35 mm Hg, 10 to 20 g mannitol every 4 hours, and 100 to 200 mg thiopentone per hour were instituted in this sequence according to the needs for control of ICP.
Sequential Doppler investigations were performed during the first week after the injury. On average, patients were examined two to three times beginning 12 to 36 hours after the injury. No testing was done within the first 12 hours because of concern of increasing evolving intracranial hematomas.
A dedicated two-channel TCD system (Hemodop, DWL Electronics) was used. The software of the system facilitates determination of CO2 reactivity, which was accomplished by decreasing end-tidal Pco2 by 8 mm Hg. After baseline registration of both cervical ICA and MCA flow velocities, as well as MCA CO2 reactivity, systolic arterial pressure was raised approximately 20 mm Hg by means of a phenylephrine (Neosynephrine, Winthrop) infusion. A standard rate of 80 μg/min was used initially. Adjustments of the infusion rate were made after 3 minutes if necessary. MCA flow velocity and vasoreactivity measurements were then repeated.
The effect of induced hypertension on CO2 reactivity and ICP was analyzed with regard to time after the injury and outcome. Significance levels of differences between subgroups were determined by t statistics.
Incidence of Ischemia and Hyperperfusion
Improvement of subnormal CO2 reactivity (<2%/mm Hg) after induced hypertension was found in 32 of 124 evaluated MCAs (26%) (Fig 2⇓). Further deterioration of subnormal CO2 reactivity after induced hypertension was seen in only 4 of 124 evaluated MCAs (3%). While improvement of subnormal CO2 reactivity was observed unilaterally or bilaterally, the four instances of further deterioration of subnormal CO2 reactivity were unilateral. Improvement of subnormal CO2 reactivity was seen at CPPs up to 80 mm Hg (see Fig 3⇓).
Time Course of Ischemia and Hyperperfusion
Conditions of improvement of subnormal CO2 reactivity on induced hypertension were identified during the entire study period (Table, Fig 4).⇓⇓ Further deterioration of subnormal reactivity was seen only after the third day after injury.
Incidence of Vasospasm
Unilateral or bilateral vasospasm with an MCA/ICA flow velocity ratio >3 occurred in a total of 7 patients. Related secondary ischemia (hypertensive improvement of subnormal CO2 reactivity) was seen unilaterally or bilaterally in 4 of these cases. Three of them died, possibly because of delayed ischemia, and 1 survived with moderate disability.
Effect of Induced Hypertension on ICP
Induced hypertension tended to result in an increase of ICP in cases of ischemia as determined by CO2 reactivity and otherwise in a decrease (Table). The correlation coefficient between response of CO2 reactivity and ICP response was calculated as r=.4. There was also an inverse correlation between baseline CO2 reactivity and ICP response (r=−.3).
Methodological Considerations and Safety of Pharmacologically Induced Arterial Hypertension
The present study used a novel technique to diagnose ischemia and hyperemia in trauma patients. In contrast to patients with vascular occlusive disease, determination of diminished vasoreactivity alone is not sufficient for the diagnosis of ischemia in trauma patients. Only the response of the vasoreactivities to induced blood pressure changes allows for differentiation. Established methods for identification of ischemia and hyperemia are based on the demonstration of uncoupled flow and metabolism by means of positron emission tomography or jugular vein oximetry.4 8 12 13 14 16 17 Cerebrospinal fluid and parenchymal lactate content have been used as indicators of ischemia.15 18 19 The global methods of jugular vein oximetry and cerebrospinal fluid lactate determination have the disadvantage of lacking any regional correlation. Furthermore, jugular venous oximetry misses ischemic areas coexisting with hyperemic areas because such combinations can result in a normal net venous O2 content. The present method allows at least separate evaluation of the two MCA territories.
Because of concern about the safety of pharmacologically raising systemic arterial pressure, no tests were performed during the first 12 hours after injury. Follow-up CT scans in this series did not show any new intracranial hematomas as a consequence of the arterial pressure manipulations. More discrete negative effects, such as increase of edema, cannot be excluded with certainty. The routine use of these tests is certainly only justified by the potential of therapeutic consequences.
Relations Between CPP and Ischemia and Hyperemia, Respectively
The actual analysis showed that the common physiological variables ICP and CPP are not reliable predictors of hypoperfusion and hyperperfusion. Although there was some negative correlation between CPP and the occurrence of ischemic patterns, considerable variability was seen. There were ischemic patterns with improvement of vasoreactivities by induced hypertension in patients with baseline CPPs above 70 mm Hg. This finding is in agreement with reports by Chan and colleagues4 5 8 who found that a much higher CPP is required in trauma patients than in normal subjects. This phenomenon is explained by injury to the peripheral vascular bed and by edema affecting diffusion of O2 and CO2. The evolving evidence of a major variability of the required CPP suggests that a corresponding individual assessment may be important in the management of head injury. The present bilateral evaluations commonly showed some differences between the responses of the two hemispheres to induced hypertension. While ischemia reaching the used arbitrary limits was identified unilaterally or bilaterally, the four instances of hyperperfusion were seen unilaterally. The small number of events does not allow a statement with regard to the possibility of bilateral hyperperfusion. Earlier studies using regional cerebral blood measurements demonstrated hyperperfusion predominantly around focal brain injuries.17 18
Time Course of Ischemia and Hyperemia
The actual investigations showed that ischemia is a common event during the first 72 hours. After this period, three groups can be separated: (1) normalization of perfusion, (2) persistent ischemia, and (3) hyperemia. The finding that hyperperfusion is rare during the early phase is corroborated by several other recent reports.1 2 3 8 The above-mentioned evolution suggests that early ischemic territories can secondarily become hyperemic exactly as ischemic cerebral infarctions. Persistent or delayed ischemia, on the other hand, may be a consequence of edema, reduced perfusion pressure, or vasospasm.11 20 21 The significance of vasospasm after head injury is debated.5 22 23 In a previous series of 86 patients, we saw no instance of an unfavorable outcome due to vasospasm.9 In the present series, there were 3 patients who died, possibly as a consequence of vasospasm.
Should Dysemia Be Treated?
Our evaluations demonstrated that in some patients induced hypertension led to an improvement of cerebrovascular reactivity, probably signaling improved ischemia. However, at the present time it appears to be premature to suggest therapeutic hypertension in these instances. Further research is necessary to document that the short-term beneficial effect could be maintained during prolonged periods. It is possible that induced hypertension could cause a more pronounced formation of edema, which would annihilate the benefit of hypertension after a few hours.
Selected Abbreviations and Acronyms
|CPP||=||cerebral perfusion pressure|
|GCS||=||Glasgow Coma Scale|
|GOS||=||Glasgow Outcome Scale|
|ICA||=||internal carotid artery|
|MCA||=||middle cerebral artery|
The TCD research program is supported by the Swiss National Fund, grant 31-32569.91. The authors are indebted to Dr L. Studer for his help with the statistical analysis.
- Received May 13, 1996.
- Revision received July 13, 1996.
- Accepted July 13, 1996.
- Copyright © 1996 by American Heart Association