Visually Evoked Cerebral Blood Flow Velocity Changes in Different States of Brain Dysfunction
Background and Purpose By assessment of metabolically induced cerebral blood flow velocity changes, transcranial Doppler sonography offers the opportunity to evaluate vasoneuronal coupling in different states of brain activation and in critically ill patients.
Methods With simultaneous transcranial Doppler monitoring of the posterior cerebral artery (PCA) and the middle cerebral artery (MCA), 27 control subjects, 11 patients under general anesthesia, 5 patients in the vegetative state, and 12 patients with aneurysmal subarachnoid hemorrhage were stimulated with a 10-Hz flashlight for 30 seconds. Ten cycles of stimulation were averaged, and a specific flow response (SFR) was computed as the normalized ratio of PCA/MCA mean flow velocity.
Results Maximal SFR was 14.2% in control subjects. Eye closure significantly reduced maximal SFR (11.6% versus 15.4%, P<.01). In subarachnoid hemorrhage, SFR was markedly decreased in the early phase (4.8%, P<.01) but became normal later on. Four of 5 patients with abolished SFR suffered delayed ischemia due to vasospasm. Of 7 patients with preserved SFR, 5 had vasospasm but none had delayed ischemia. No SFR was observed in patients under general anesthesia or in the vegetative state.
Conclusions Although reflecting fast and local neuronal activity patterns, metabolically induced blood flow response is highly dependent on stimulus-directed attention. In subarachnoid hemorrhage, decreased metabolic flow response suggests severe depression of vasoneuronal coupling, and abolished SFR might indicate increased vulnerability to vasospasm and a higher risk for delayed ischemia.
Increase of rCBF in the visual cortex due to metabolic demand after appropriate stimulation is well established, as demonstrated by various functional imaging studies.1 2 3 Aaslid4 first described corresponding findings using TCD examination, documenting a flow velocity increase in the PCA after visual stimulation. Compared with functional imaging methods such as PET or functional MRI, TCD allows much better temporal resolution, although spatial resolution is poor. This disadvantage is less important for the PCA feeding the occipital lobe because most of the supplied tissue involves the visual system. Therefore, TCD might be valuable for clinical investigation of cerebral hemodynamics. A noninvasive technique and easier instrumentation allow examination of critically ill patients for whom PET and MRI studies are not suitable. Visually evoked blood flow velocity has been examined under various conditions in healthy subjects,4 5 6 7 but experience in pathological or altered states of brain function is restricted to patients with occipital lobe infarction8 9 and migraine.10 We address the question of how states of reduced stimulus-directed attention affect flow velocity changes in the PCA after visual stimulation. In addition, patients with SAH, a disorder with global alteration of CBF and metabolism, were studied.
Subjects and Methods
TCD examination was performed with bilateral 2-MHz probes monitoring the PCA and the contralateral MCA (DWL). In most cases, the left P1 segment and right M1 segment were insonated. Identification of blood vessels followed standard criteria for transcranial ultrasound examination.11 After optimal adjustment, probes were tightly fixed by a headband. Subjects were stimulated with a binocular flashlight at a frequency of 10 Hz, with conventional visual evoked potential stimulation goggles (stimulus time, 10 milliseconds). Stimulation and rest periods were both 30 seconds. Ten cycles of stimulation were performed for each trial. Envelope curves of flow velocities were stored on hard disk. Off-line analysis averaged all cycles of a trial, using a sampling frequency of 5 Hz for estimation of averaged mean flow velocity. Furthermore, the ratio of PCA/MCA flow velocity was computed and normalized to 100%. As baseline, the last 2 seconds before each stimulation period were averaged. Changes of this ratio in percent were identified as SFR. On and off latency and the slopes of increasing and decreasing leg were obtained by visual determination as shown in Fig 1⇓. By using the contralateral MCA as a reference, specific velocity changes in the PCA are separated from blood pressure changes or a more global, attention-related reaction of cerebral perfusion.
Twenty-seven control subjects were asked to keep their eyes open during rest and stimulation periods. Thirteen of these subjects were also examined with their eyes closed during the same session, and 8 subjects were examined twice with eyes open. The intervals between examinations ranged from 7 days to 2 months. Twelve patients with spontaneous SAH were examined; in 6 patients, serial examinations (3 to 5) were performed. All patients had suffered aneurysmal bleeding with a Hunt and Hess score of 1 to 4 at admission; 11 of them underwent clipping or coil embolization of the aneurysm. Aneurysms were localized in the MCA (6), internal carotid artery (2), anterior communicating artery (3), and basilar artery (1). Serial routine TCD examinations or serial angiograms for diagnosis of cerebral vasospasm were available in all patients. Nimodipine was administered intravenously to all patients. Patients were asked to keep their eyes open during the session. Compliance was checked by direct control of eye opening. No examination had to be aborted for incomplete compliance. Eleven patients under general anesthesia were intubated, mechanically ventilated, and received fentanyl and midazolam by continuous infusion. Indication for anesthesia was abdominal surgery (8 patients) and trauma without head injury (3 patients). All patients had stable cardiac and respiratory function. Carbon dioxide pressure was measured by arterial blood gas sampling immediately before examination and was in the normal range (34 to 41 mm Hg). In 4 of these patients, eyes were kept open by eyelid clamps during stimulation. Five patients were in the vegetative state after head injury with diffuse axonal trauma. None of those patients had visible lesions in the occipital region on CT scan. Eye opening was inconstant in these patients. For statistical analysis, the Wilcoxon test for paired and unpaired groups and Spearman’s correlation coefficient were used.
SFR characteristics for control subjects are shown in the Table⇓. SFR was evoked in all subjects (range, 6.8% to 26.2%). Onset latency was shorter than offset latency. In 15 of 27 subjects (56%), a transient SFR increase after stimulus cessation was present (off-phenomenon). Decrease of SFR after stimulus cessation below the level of the prestimulus period (undershooting) was present in 7 patients (26%). The slope of the increasing leg varied between 0.84% and 7.46% (mean, 3.17% per second) and correlated with the maximal SFR (r=.59, P<.01). The slope of the descending leg varied between 1.24% and 6.16% (mean, 3.15% per second) and also correlated with maximal SFR (r=.61, P<.01). In the subgroup examined with both modalities, maximal SFR with open eyes (mean±SD, 15.4±4.9%) was significantly higher than with closed eyes (11.6±5.3%; P<.01). Although the range of maximal SFR was wide, intraindividual variability was low. Maximal SFR after repeated stimulation showed no significant difference (Fig 2⇓).
Maximal SFR in SAH during the early stage of disease was invariably less than in control subjects (Table⇑). Off-phenomena were present in 3 patients and undershooting in only 1 patient. Maximal SFR correlated with the slope of the increasing leg (r=.54, P<.05) but not with the slope of the decreasing leg. On and off latencies were similar to those in control subjects. In 5 patients, no SFR was recorded. Four of these patients suffered delayed ischemia. Symptoms were dysphasia (2 patients) and mild hemiparesis (2 patients) but not visual field deficit related to the PCA territory. All patients had vasospasm in the MCA, 1 patient also in the PCA. This patient suffered dysphasia, suggesting ischemia in the MCA territory. Of 7 patients with preserved SFR, 5 had vasospasm in the MCA, but none had delayed ischemia. Serial examinations showed distinct but not significant increase of maximal SFR in all but 1 patient (Fig 2⇑). There was no difference in age or clinical findings on admission between the subgroups of patients with serial and single SFR measurements. All patients were alert, followed instructions, and kept eyes open. Hunt and Hess scores were 1 or 2 during examination.
No reliable SFR was recorded in patients in the vegetative state or under general anesthesia.
In this study, we found an immediate increase of blood flow velocity after stimulation of neurons involved in visual processing in healthy control subjects. This increase was less pronounced in those with closed eyes. In patients with SAH, it was markedly reduced in the acute phase but showed some improvement during the clinical course. An abolished response was detected in patients under general anesthesia and in vegetative patients. Although TCD does not measure absolute values of rCBF or absolute rCBF changes, changes of flow velocity reflect relative rCBF changes.12 In this way, TCD provides information about rCBF regulation in relation to metabolic activity.
Changes in neuronal activity may influence rCBF locally via vasoactive substances. As shown by studies of induced brain ischemia, adenosine is a good candidate to act as a vasodilative agent in response to metabolic demand,13 14 and it is also possibly involved in fast autoregulation of the cerebral vasculature to maintain cerebral perfusion.14 15 Presumably, there are both feedback and “feed-forward” mechanisms of regulation. Therefore, vasodilation will be induced by imbalance between metabolic demand and oxygen supply but also in anticipation of local neuronal activity, probably by synaptic transmission and membrane depolarization.14 16 There is further evidence of modification of this messenger by other factors such as carbon dioxide.13 14
In this study, flow response after visual stimulation was less than in the study of Gomez et al,7 who applied a similar 10 Hz-flashlight stimulus. This difference might have been caused by our use of the MCA as a reference, which also reacts to visual stimuli in a minor degree, and by subjects keeping their eyes open in the off-stimulus period. Closing the eyes significantly changed flow response, which is in accordance with functional imaging studies.17 Reduced flow response with closed eyes suggests a decreased metabolic activity due to decreased stimulus intensity and diminished attention but preserved blood flow regulation.
The visually evoked flow response depends on stimulus characteristics (eg, flashlight frequency,7 complexity of stimuli,6 8 and diopter-corrected angle of stimulus presentation5 ) and on intrinsic factors (eg, stimulus evaluation6 ). The underlying neuronal mechanisms for these phenomena are excitation of associative visual areas by complex or color stimuli and activation of more and more complex cell assemblies by dynamic stimuli or illumination of the central regions of the retina.2 8
In contrast to functional imaging methods, TCD provides data about temporal properties of metabolically induced flow response: even very short stimuli of 50 milliseconds result in a flow response, suggesting a fast-acting mediation.5 Like others,6 we observed specific temporal patterns as off-phenomena and undershooting after stimulus cessation. Off-phenomena correspond to neuronal off-effects in the visual cortex, while delay in flow decrease corresponds to visual afterimages.6 The “undershooting” decrease of blood flow velocity in the poststimulus phase might be related to a counterregulation of excessive blood supply in an oscillatory mode.6 18 It is remarkable, however, that metabolic blood flow response, although reflecting fast and local neural activity patterns, is also dependent on global factors such as stimulus-directed attention. This might also be the explanation for abolished blood flow response in our vegetative patients.
A surprising finding in our study was the lack of flow response in patients under general anesthesia. While midazolam reduces both rCBF and cerebral metabolic rate for oxygen by as much as 30% in a parallel manner, fentanyl does not substantially alter both variables.19 Autoregulation of cerebral perfusion pressure and reactivity to carbon dioxide is not affected.19 Visual evoked potentials are obtainable under anesthesia with both drugs, suggesting that neural activity induced by flashlight stimulus is preserved.19 Since we used an averaging technique, it is rather unlikely that we missed blood flow response by quantity. Because both response to metabolic activity and autoregulation are probably mediated by the same vasodilating agents, emission of these agents and susceptibility of arterioles are presumably preserved. Instead, selective adjustment of the metabolic CBF control might be impaired. Further systematic studies of induced rCBF response in patients under anesthetic agents would help to clarify these questions.
Patients with aneurysmal SAH also exhibited diminished flow response that gradually improved over time, as shown in our serial examinations. Since intraindividual variability was low in our control subjects and compliance was good in our patients, this improvement was probably related to the course of the underlying disease. In 5 patients, no flow response was observed at all; 4 of the 5 suffered delayed ischemia due to vasospasm. There is ample evidence for decreased rCBF related to vasospasm of the feeding vessel, reflecting poststenotic compromise of perfusion.20 However, our patients had signs of MCA territory ischemia, and 1 of these patients underwent angiography 1 day before visual stimulation to ensure that no spasm affected the PCA territory. Therefore, we conclude that abolished flow response reflected global and severe disturbance of cerebral metabolism and rCBF, independent of vasospasm. This is in accordance with a recent PET study revealing decreased rCBF in the absence of spasm, elevated intracranial pressure, or cerebral hematoma; in the prespasm period, Carpenter and colleagues21 found decreased rCBF and cerebral metabolic rate for oxygen. Oxygen extraction was unchanged and the ratio of cerebral blood volume to CBF did not rise, as would be expected in a vasodilative response of autoregulation to a decline in CBF. The authors concluded that reduced CBF was secondary to reduced metabolism, which is probably a response to toxic effects of subarachnoid blood mediated through a brain stem mechanism.22 We hypothesize that similar toxic effects might also decrease metabolically induced rCBF response that could be related to augmented cerebral vulnerability to ischemia, as suggested by the results in our patients. Since flow velocity kinetics such as on and off latencies and correlation of increasing slope and maximal SFR were not substantially altered, downregulation of metabolically induced secretion of vasodilative agents with preserved susceptibility of arterioles is possible.
In summary, we conclude that visually evoked flow response might not only depend on local regulatory mechanisms but also might reflect global properties of brain function. This method is feasible in critically ill patients. Further studies are necessary to investigate whether functional TCD examinations could contribute to the clinical management of these patients, eg, monitoring of patients with SAH.
Selected Abbreviations and Acronyms
|CBF||=||cerebral blood flow|
|MCA||=||middle cerebral artery|
|PCA||=||posterior cerebral artery|
|PET||=||positron emission tomography|
|rCBF||=||regional cerebral blood flow|
|SFR||=||specific flow response|
|TCD||=||transcranial Doppler sonography|
- Received August 31, 1995.
- Revision received November 6, 1995.
- Accepted November 27, 1995.
- Copyright © 1996 by American Heart Association
Phelps ME, Kuhl DE, Mazziott JC. Metabolic mapping of the brain’s response to visual stimulation: studies in humans. Science. 1981;211:1445-1448.
Fox PT, Raichle ME. Stimulus rate dependence of regional cerebral blood flow in human striate cortex, demonstrated by positron emission tomography. J Neurophysiol. 1984;5:1109-1120.
Aaslid R. Visually evoked dynamic blood flow response of the human cerebral circulation. Stroke. 1987;18:771-775.
Wittich I, Klingelhöfer J, Matzander G, Sander D, Conrad B. Dynamics of visually evoked perfusion changes in the posterior cerebral artery territory. Cerebrovasc Dis. 1994;4(suppl 3):3. Abstract.
Gomez SM, Gomez CR, Hall IS. Transcranial Doppler ultrasound assessment of intermittent light stimulation at different frequencies. Stroke. 1990;21:1746-1748.
Sitzer M, Diehl RR, Hennerici M. Visually evoked cerebral blood flow responses: normal and pathological conditions. J Neuroimaging. 1992;2:65-70.
Niehaus LB. Visuell evozierte Perfusionsänderungen bei Posteriorinfarkten. In: Jerusalem F, Ries F, eds. Verhandlungen der Deutschen Gesellschaft für Neurologie. Berlin, Germany: Springer; 1994:935-936.
Fujioka KA, Douville CM. Anatomy and freehand techniques. In: Newell DW, Aaslid R, eds. Transcranial Doppler. New York, NY: Raven Press Publishers; 1992:9-31.
Sorteberg W. Cerebral artery blood velocity and cerebral blood flow. In: Newell DW, Aaslid R, eds. Transcranial Doppler. New York, NY: Raven Press Publishers; 1992:57-66.
Berne RM, Winn HR, Rubio R. The local regulation of cerebral blood flow. Prog Cardiovasc Dis. 1981;3:243-260.
Strandgaard S, Paulson OB. Cerebral autoregulation. Stroke. 1984;3:413-416.
Kuschinsky W. Is the EEG correlated with the brain metabolism and cerebral blood flow? In: Zschocke S, Speckmann EJ, eds. Basic Mechanism of the EEG. Boston, Mass: Birkhäuser; 1993:109-121.
Meyer JS, Hayman LA, Amano T, Nakajima S, Shaw T, Lauzon P, Derman S, Karacan I, Harati Y. Mapping of local blood flow of human brain by CT scanning during stable xenon inhalation. Stroke. 1981;12:426-436.
Aaslid R, Newell DW, Stoos R. Assessment of cerebral autoregulation dynamics from simultaneous arterial and venous transcranial Doppler recording in humans. Stroke. 1991;22:1148-1154.
Sakabe T, Nakakimura K. Effects of anesthetic agents and other drugs on cerebral blood flow, metabolism, and intracranial pressure. In: Cottrell JE, Smith DS, eds. Anesthesia and Neurosurgery. St Louis, Mo: CV Mosby; 1994:149-174.
Jakobsen M, Overgaard J, Marcussen E, Enevoldsen EM. Relation between angiographic cerebral vasospasm and regional CBF in patients with SAH. Acta Neurol Scand. 1990;80:109-115.
Fein JM. Cerebral energy metabolism after subarachnoid hemorrhage. Stroke. 1975;1:1-8.