5. CEREBRAL CIRCULATION

5. CEREBRAL CIRCULATION

5.1 BLOOD-VELOCITY/TIME WAVEFORM ANALYSIS

It is possible to record blood velocity information from many of the head and neck blood vessels. Most of the clinically useful information up to the present time has involved waveforms from the carotid arteries, the terminal branches of the ophthalmic, and the vertebral arteries.

5.1a Carotid Arteries

Typical blood-velocity/time waveforms for the common, internal and external carotids are shown in Fig. 5. 1. It can be seen that in the internal carotid artery there is a relatively higher

diastolic flow. Based on the hypothesis that the shape of the blood-velocity/time waveform over the cardiac cycle is due in part to the pathophysiology of the circulation, several attempts have been made to characterize these waveforms and relate them to the known pathophysiological properties of the vessels. The various techniques of characterization can be explained with reference to Fig. 5.2, which shows a typically normal common carotid waveform and also a waveform from the common carotid artery proximal to which is a major

stenosis.

Mol (I973) investigated the systolic and diastolic amplitudes of these waveforms and showed that in the normal situation there is no significant difference in the parameters from
both common carotid arteries. However, it was shown that systolic and diastolic amplitudes decrease almost proportionally with progressive age. Planiol and Pourcelot (1971) introduced an index which is related to the peripheral resistance of the cerebral circulation. This index is

defined as the difference between the systolic and end-diastolic amplitudes divided by the systolic amplitude, i.e.

The normal range regardless of age was found to be 0'55-0'75. Values higher than this relate to an increased peripheral resistance, and lower values relate to a decreased resistance. This index, unlike that described by Mol, is normalized and so the angle of inclination of the ultrasound beam to the flow direction does not need to be measured.

Baskett et al. (1977) investigated a number of features of the carotid artery waveforms and found that the most consistent parameter which demonstrates change with age and disease is the ratio of the two peaks in systole, A/B (see Fig. 5.2). The peak height A is equivalent to Pourcelot's S. From a total of 800 asymptomatic volunteers from the ages 5-90 years the ratio A/B decreases with age until the sixth decade after which it remains at the value of approximately 1^.2.

In the search for a routine diagnostic test for occlusive carotid artery disease, Baskett et al. measured the A/B ratios at both the common carotid artery and the supraorbital artery. The technique for obtaining the signals on both vessels is as follows. The common carotid artery is insonated at the base of the neck with the transducer pointing towards the carotid junction. Twenty consecutive heartbeats are obtained and the mean value of A/B calculated. The supraorbital artery is insonated under the eyebrow just below the supraorbital ridge. Twenty consecutive heartbeats are again recorded and A/B calculated. The two values of the A/B ratio (one for the common carotid, and one for the supraorbital) are then plotted on a graph, as illustrated in Fig. 5.3. Baskett et al. compared the A/B ratios from these two vessels with the angiographic appearance in 101 cases. Empirical diagnostic thresholds are drawn at 1^.05 on both axes. Their results show that if the A/B ratio in either vessel is less than 1^.05 then there is an 88%. probability of disease at the carotid bifurcation. When the A/B ratio in both vessels is greater than 1'05, then there is an 80%. probability of a normal bifurcation. Horrocks et al. (l979) have repeated this work on a smaller sample of cerebral vessels. Their results do not appear as convincing as those of Baskett et al.

Reinertson and Barnes (1976) showed that in order to detect accurately changes in blood velocity produced by minor stenoses, it is important to investigate the whole Doppler-shift spectrum. The normal spectrum of a carotid artery shows the greatest signal amplitude is associated with the maximum frequency envelope. Distal to stenoses which reduce the lumen by between 25-50%, the spectrum during systole exhibits increasing amplitudes in the lower frequency range. This is referred to as spectral broadening. In stenoses in the 50-75%. range spectral broadening is noted in both systole and diastole. Distal to stenoses of greater than 75%. there is a loss of the normal structure of the waveform with a dominance of the tower frequency signals. Actually in the stenosis very high Doppler shifts are recorded. These changes in spectral characteristics are shown in Fig. 5.4.

Spencer and Reid (1979) have quantified the degree of carotid stenosis using these techniques. It is interesting to note that these authors found Doppler shifts of between 15-16KHz in very tight stenoses, where the Doppler probe is inclined at an angle of 60° to the direction of flow. Spencer and Reid have produced an empirical relationship between the maximum Doppler shift and the minimal diameter shown on contrast arteriography in 95 internal carotid arteries. For 77 diameters greater than 1 '5 mm the diameter D = 8^.77¦ ^{-0 57} with a correlation coefficient of 0^.74, where ¦ is the maximum Doppler shift frequency in the stenosis, in kHz.

A second method has also been developed in which the ratio of the maximum frequencies recorded in the internal carotid artery at the angle of the jaw is divided by the maximum frequency recorded in the stenosis. Spencer and Reid quote an accuracy of ±20%. in 80%. of cases, in the assessment of the size of the stenosis. For stenoses greater than 70%, the technique provides a 63%. sensitivity, 85%. specificity and an overall accuracy of 95%..

Blackshear et al. (l979) analysed the full Doppler shift spectrum in 66 carotid arteries. Of eight normal carotids, diagnosed using contrast arteriography, one had minor abnormalities. Sixty-six per cent of vessels with mild irregularity of the wall had minimal spectral broadening. In the range 1 0-49%. diameter stenosis, 78%. exhibited spectral broadening, while all of the 22 high grade stenoses (50-99%. stenosis) had a positive examination. Thus 89%. of vessels with any disease demonstrable on contrast arteriography, were identified using spectral analysis.

Keller et al. (1976) have produced a set of eight criteria for the diagnosis of haemodynamic disturbances in the carotid artery: (1) reverse flow in the frontal artery; (Z) flow pulse amplitude differences between frontal arteries of more than 40%; (3) reverse flow in the supraorbital artery; (4) flow pulse amplitude difference between supraorbital arteries of more than 80%.; (5) pathological flow changes in frontal arteries on common carotid artery compression; (6) low diastolic flow in the common carotid; (7) time delay of more than 30 ms in the rising phase of systolic flow wave in either the frontal artery and/or the supraorbital with/without compression of the external carotid branches; (8) shape of the frontal artery flow waveform. particularly the absence of the two systolic components and/or the incisura.

Keller et al. in studying criteria 2, the amplitude difference between both normal frontal arteries, state that the averaged flow waveform amplitudes should not differ by more than 20%:. If a difference of more than 40%,, is observed, this is probably due to a stenosed carotid artery. A 20-40',, difference has been associated with disease distal to the carotid syphon, i.e. intracranial.

In the study described by Keller et al. 186 patients underwent carotid arteriography. The Doppler investigations detected 138 pathological flow situations in 123 patients. Angiography showed 110 obstructions, 11 cases of selective stenosis distal from the ophthalmic artery, and 11 normal findings. The diagnostic value of the individual criteria listed above was calculated from the ratio of the number of cases with false positive occurrence to the total number of cases with positive occurrence. In this way a lower quotient corresponds to a lower ranking. The criteria were ranked as follows: criterion 1, rank 1: criterion 2. rank 6; criterion 3. rank 5: criterion 4, rank 7: criterion 5, rank 3; criterion 6, rank 2: criterion 7, rank 4.

Keller et al.'s overall conclusions are that with these diagnostic criteria haemodynamically significant obstructions from the intrathoracic origin to the branching of the ophthalmic artery can be made with 90%, reliability. When criterion 1 or 6 is present the Doppler diagnosis was always correct. One or other of these occurred in 20%. of cases. Criteria 3, 5 and 7 are of approximately equal value. having a high probability of correlation with morphological findings (43;;, of cases including 6% false-positive cases). Criteria 2 and 4 gave an indication of abnormal haemodynamics (37% of cases including 14% false positives). When three or more positive criteria were present on a given patient, Doppler diagnosis was always correct. There seems to be good correlation between the diagnostic rank of a criterion and the degree of obstruction, but little correlation between the degree of a stenosis and the occurrence of a particular criterion. This is probably due to the multi-factor control and compensation processes which occur as a result of a stenosis.

5.1b Vertebral Artery

Similar characterization methods have been applied to the study of the vertebral artery. Miyazaki et al. (1966) showed it is possible to study flow patterns in the vertebral artery but Kaneda et al. (1977) have used a quantitative method to assess the severity of vertebral artery disease. Their initial results are very promising but liable to inaccuracies because their index is based on a measure of the absolute systolic and diastolic heights on the sonagram. As explained in a previous chapter the absolute height depends on both the velocity of the blood and on the angle of inclination of the ultrasound beam to the flow direction, and in order that these parameters may be compared either within a patient or from patient to patient, this angle of inclination must be known. However, in the Kaneda et al. study, 64 vertebral arteries (53 patients) were studied. The results show that if the systolic height is less than or equal to 12 mm or the diastolic height less than or equal to 4 mm then the vessels are abnormal, when compared with the results of angiography. All normal vertebral arteries, except one, satisfied the criteria of S > 12mm, d > 4mm. In the 64 vertebral arteries in 53 patients, a Flow signal was detected in 55 cases. In nine patients no vertebral artery signal was detected; four were found to be occluded, on angiography, in two cases the artery was missing altogether but three patients had normal looking vessels.

Although the index used is not normalized, the results are very interesting. The diagnostic reliability of the percutaneous vertebral Doppler technique was 67%: (6/9) where no flow was detected, 94%. (16/17) for the "poor flow" type, 97%. (37/38) for normal vessels and 92%, (59/64) for all vessels examined.

The technique for obtaining the Doppler signals is obviously important when used with a non-normalized characterisation procedure. The patient is placed in the supine position with the head slightly extended and rotated about 45" to the opposite side. The probe is then placed on the skin just below the mastoid process and pointed towards the orbit on the opposite side. The probe is then slowly moved to find the optimum angle for a maximum Doppler shift. If there is any doubt as to whether it is the vertebral being insonated, ipsilateral common carotid compression may be tried,

Keller et al. (1976) have investigated the vertebral artery via the peroral route. The technique here is to place the probe on the anaesthetised oropharynx between the transverse processes of two adjacent cervical vertebrae, and angled in a caudad direction. Transient carotid compression aids in identifying the vertebral artery. Normally no change in the vertebral artery is observed, although on occasions flow is increased, with either ipsilateral or contralateral compression. If no increase occurs the vessel under the probe is independent of the particular carotid artery perfusion, whereas if flow increases, the posterior communicating artery ipsilateral to the compression is functioning. If a particularly large increase is noted then a shunt from the intracranial portion of the carotid, such as vertebrobasilar anastomosis or a dominant posterior communicating artery, could be the cause, or the vertebral artery itself has a reduced blood pressure. If, on compression, the vertebral Flow decreases markedly, this could be due to subclavian steal.

In a study of 42 patients with contrast angiograms Doppler investigations of the vertebral artery were carried out. The results are shown in Table 5.1.