The end-tidal carbon dioxide partial pressure (PCO2) response curves for the flow velocity in the middle cerebral artery were studied in 31 normal subjects with transcranial Doppler techniques. An exponential curve with an exponent of 0.034 mm Hg-1 was found to be a good fit to the recorded data. By means of this relationship, recordings of flow velocity in cerebral arteries can be normalized to a standard value of PCO2. Physiological aspects of cerebrovascular reactivity to PCO2 and the clinical implications of the PCO2 response curve are discussed. The normal material provides a reference for assessing pathological responses.

Noninvasive transcranial Doppler recordings were correlated to the angiographic findings in 77 patients with carotid artery disease. Stenoses reducing the luminal area of the internal carotid artery by 75% or more also reduced the pulsatility transmission index (PTI) of the ipsilateral middle cerebral artery (MCA). The PTI is the pulsatility index of the artery under study expressed as a percent of the pulsatility index of another intracranial artery with presumed unimpeded inflow in the same individual. For stenoses in the 75% to 89% category. PTI reduction was significantly greater in patients with bilateral carotid stenosis, indicating an impaired potential for collateral flow in these patients. The PTI reduction probably reflects both the pressure drop across the stenosis and the cerebral autoregulatory response. Two criteria proved useful in demonstrating collateral MCA supply through the circle of Willis. On the recipient side, retrograde flow in the proximal anterior cerebral artery was demonstrated in 29 of the 31 patients when this flow pattern was disclosed angiographically. In 26 of these patients, the anterior cerebral artery on the supplying side also had clearly increased flow velocity. Increased flow velocities in the proximal posterior cerebral artery were present in 26 of the 30 vessels that were acting as a collateral channel to the ipsilateral MCA.

No abstract

Results from 1,039 combined cervical and transcranial Doppler examinations are reported. Satisfactory transcranial signals were not found in 2.7% of the cases. Compared with angiography, the accuracy of transcranial criteria in assessing collateral flow over the circle of Willis was 94 and 88% for anterior and posterior circulation, respectively. The method also appeared very promising for detection of lesions of the intracranial arteries although the number of such cases with angiographic verification was limited in the present series. Arterial narrowing due to cerebral vasospasm was diagnosed with a sensitivity of 80%. In patients with ruptured intracranial aneurysms, an incidence of 93% arterial narrowing in basal cerebral arteries was found. Patients with subarachnold hemorrhage and no aneurysm on angiography also showed arterial narrowing with an incidence of 56%. It was possible to monitor the time course and severity of cerebral vasospasm. Arteriovenous malformations were characterized by Doppler findings of high velocities and low pulsatilities. These lesions were diagnosed with an accuracy of 95%. (Stroke 1987;18:1018-1024)

No abstract

Thirty-five patients were admitted to the hospital with Glasgow coma scale scores of 4 to 7 after severe, blunt head injury. Blood flow velocities of the middle cerebral artery (MCA) and the ipsilateral internal carotid artery (ICA), high in the neck, were recorded noninvasively by Doppler ultrasonography. Serial examinations were begun within 48 hours of trauma and continued until the patient either died or was discharged. Spasm of the MCA was assumed if the ratio of the velocity of blood flow in the MCA (VMCA) to the velocity of blood flow in the ICA (VICA) exceeded 3 (normal value, 1.7 +/- 0.4). In cases of severe MCA spasm, this ratio is higher because of increased flow velocity in the MCA and reduced flow in the ICA due to increased cerebrovascular resistance. Higher MCA velocities with VMCA/VICA above 3, consistent with MCA spasm, were found in 14 of the 35 patients (40%). The increase began as early as 48 hours after injury, reached a maximum between Days 5 and 7, and lasted until 2 weeks after injury. There was a significant correlation between the occurrence of vasospasm and the quantity of cisternal or intracerebral blood seen on a computed tomographic scan. No correlation was found with the age of the patients, the Glasgow coma scale score at admission, the intracranial pressure, or the functional outcome 6 months after injury. The occurrence of a secondary infarction in a patient with severe MCA spasm suggests that, at least in some cases, spasm may influence the prognosis.

In 39 patients with a proven subarachnoid hemorrhage (SAH), the clinical status, the amount of subarachnoid blood on a computerized tomography scan obtained within 5 days after SAH, and the flow velocities (FV's) in both middle cerebral arteries (MCA's) measured by transcranial Doppler sonography were recorded daily and correlated. All patients had pathological FV's over 80 cm/sec between Day 4 and Day 10 after SAH. The side of the ruptured aneurysm showed higher FV's than did the unaffected side in cases of laterally localized aneurysms. Increase in FV preceded clinical manifestation of ischemia. A step early increase of FV's portended severe ischemia and impending infarction. Maximum FV's in the range of 120 to 140 cm/sec were not critical and in no case led to brain infarction. Maximum FV's over 200 cm/sec were associated with a tendency for ischemia, but the patients may remain clinically asymptomatic. In cases of no or only a little blood in the basal cisterns, mean FV's in both MCA's increased only moderately whereas, with thick clots of subarachnoid blood, there was a steeper and higher increase of mean FV's.

Blood flow velocities in basal cerebral arteries were recorded noninvasively in 28 patients with cerebral arteriovenous malformations (AVM's) and were correlated with the angiographic findings. In normal arteries remote from the AVM, flow velocities ranged from 44 to 94 cm/sec (median 65 cm/sec) with pulsatility indexes from 0.65 to 1.10 (median 0.87). This is consistent with findings in normal individuals. Arteries feeding the AVM's were identified by the high flow velocities (ranging from 75 to 237 cm/sec, median 124 cm/sec). The pulsatility index ranged from 0.22 to 0.74 (median 0.48). The difference of these results from findings in normal remote arteries was highly significant (p less than 0.001). Hyperventilation tests illustrated the hemodynamic difference between an AVM and normal cerebrovascular beds. Flow velocity measurements permitted noninvasive diagnosis of AVM's in 26 of the 28 patients. Furthermore, the identification of individual feeding arteries permitted good definition of the anatomical localization of individual AVM's. Flow velocity measurements combined with computerized tomography scans are useful in the diagnosis of AVM's. With the feeding artery's configuration identified on angiography, flow velocity measurements permit a new insight into the "hemodynamic dimension" of an AVM and its possible effects on adjacent normal brain-tissue perfusion in the individual patient.