Abstract
Cortical spreading depression (CSD) is associated with a dramatic failure of brain ion homeostasis as well as efflux of excitatory amino acids from nerve cells and increased energy metabolism. There is strong clinical and experimental evidence to suggest that CSD is involved in the mechanism of migraine. This paper will, based on the experience related to the detection of CSD in humans, discuss pitfalls and possible strategies for detection of CSD in man. Development of reliable methods for detection of CSD in humans will determine the extent to which the large body of experimental findings from animal studies of CSD can be applied to the investigation and treatment of human brain disease. The paper is based on the experience that has been gained from two decades of studies of CSD in relation to clinical neurological diseases.
The strongest evidence for cortical spreading depression in man is the migraine aura
The migraine aura may be defined as any neurological disturbance that appears shortly before or during the development of a migraine headache. Seemingly similar migraine auras may have different features, suggesting that different brain regions are involved (1). A typical migraine aura from the visual cortex is a scintillationscotoma with a characteristic distribution of fortification figures. Usually, the disturbance starts at the visual field centre and propagates to the peripheral parts within 10–15 min. Function returns to normal within another 10–15 min. The symptoms indicate that a wave of intense excitation in the primary visual cortex moves at speed of approximately 3 mm/min, followed by a longer period of inhibition. Milner noted that this was the same speed of propagation as Leao's spreading depression (2, 3). Similar calculations can be made with respect to the somatosensory symptoms developing along the sensory homunculus.
Comparison between migraine aura and CSD
Migraine auras are usually visual, apparently starting in area 17, i.e. the part of the brain with the highest neuronal density. This is consistent with the observation that CSD is much easier to elicit in brains with a high neuronal density. Functioning glial cells tend to decrease the probability of successful elicitation of a CSD. Both the migraine aura and CSD propagate along the cortical surface. Both tend to become extinct for reasons we do not yet understand. For the human aspects it is important that CSD tends to become extinct when propagating into a sulcus (4, 5). The rate of spread of the migraine aura is usually 2–6 mm/min, while the rate of spread of CSD is variable, especially in gyrencephalic animals. CSD, as well as the aura, tends to be restricted to one hemisphere, both can repeat themselves, and both are accompanied by a transient neurological deficit (1).
The second strongest evidence for a CSD in migraine is the ‘spreading oligemia’ observed in migraine patients
Cerebral blood flow (CBF) undergoes a sequence of changes during attacks of spontaneous migraine that has not been observed in any other patient categories with a neurological disorder (1, 6). At the very start of migraine attacks CBF decreases in the posterior part of the brain. Subsequently, the low flow region spreads into the parietal and temporal lobes at a rate of 2–3 mm/min, the maximal decrease of CBF being 30–40%, i.e. far above ischaemic levels (1, 7). Tests of CBF dynamics have revealed preserved autoregulation in the oligaemic region, while the CO2 reactivity and the functional coupling between neuronal activity and CBF is attenuated (8). The original findings have been confirmed repeatedly with PET and MR techniques (7, 9). Two recently published studies using high resolution functional magnetic resonance imaging are particularly interesting (10, 11). These two studies examined the development of the changes of the BOLD signal (BOLD = blood oxygen level dependent) during the course of migraine. In one study the onset of headache or visual symptoms, or both, was preceded by suppression of initial activation. The suppression slowly propagated into the contiguous visual cortex at a rate of 3–6 mm/min, i.e. at the same rate of propagation as CSD (10). The second study examined the vascular changes in more detail (11). A focal increase in the BOLD signal, possibly vasodilatation, developed within the extrastriate cortex at the beginning of the attack. This BOLD change propagated at a rate of approximately 3.5 mm/min over the occipital cortex at the same time as the visual aura developed. The BOLD signal then diminished, possibly reflecting vasoconstriction. The reactivity to functional activation tests decreased as previously observed with the 133Xenon technique (6, 8). During periods with no visual stimulation, but while the subject was still having visual symptoms (scintillations), the BOLD signal change followed the retinotopic progression of the visual percept.
Thus, the CBF changes in migraine are consistent with a process affecting blood vessels that propagates at a rate of 2–3 mm/min. The vascular changes consist of a vasodilatation that is capricious and precedes the oligaemia and the development of neurological symptoms. The changes of blood flow regulation consist of preserved autoregulation, attenuated CO2 reactivity, and attenuated reactivity to mental stimuli. This ‘fingerprint’ is unique for this type of neurological patient. This pattern of flow changes was used for comparison with the changes of CBF during CSD.
Hypothesis of development of an MA attack based on aspects of CSD and migraine summarized in the text. The figures represent lateral views of the human brain at different time intervals after the start of the attack, spaced by approximately 30 min. The dotted area represents the region of reduced rCBF, the striped area represents the region of neuronal depolarization during the first minutes of CSD, and arrows represent the direction of progression of CSD. 1. Initially during an MA attack a CSD is elicited at the occipital pole, spreading anteriorly at the lateral, mesial and ventral sides of the brain. At the CSD wave front, transient ionic and metabolic disequilibria trigger perturbed neuronal function, rCBF changes and neurological symptoms. 2. Following CSD, cortical rCBF decreases by 20–30% for 2–6 h. 3. rCBF in regions not invaded by CSD remains normal until encountered by CSD. 4. The region of reduced rCBF expands as the CSD moves anteriorly. 5. Somatosensory symptoms from the extremities appear when the CSD invades the primary sensory cortex at the postcentral gyrus. 6. CSD usually stops on reaching the central sulcus, but in many patients it does not even propagate this far. The ventral spread of CSD causes activation of pain-sensitive fibres and headache. 7. Full-scale attack. The CSD has stopped and is now detectable as a persistent reduction of cortical rCBF. At this time the patient suffers from headache but has no focal deficits.
Comparison of CBF during CSD and migraine
Increases of blood flow preceding the spreading oligaemia are observed only rarely in patients, but commonly in CSD in animal preparations. However, it is noteworthy that in awake animals this increase of blood flow is absent (12). This might explain the absence of a hyperaemia preceding the oligaemia in the majority of patients. In both CSD and migraine patients there is a lasting oligaemia, and the degree of reduction of blood flow is moderate and time dependent. In both migraine and CSD, blood flow autoregulation is intact while CO2 reactivity is impaired, and a reactive hyperaemia is commonly observed in both conditions (13, 14). In brief the pattern of CBF changes in migraine and CSD are similar, suggesting that CSD is the mechanism of the migraine aura and the related blood flow changes. On the basis of this background, the following model has been suggested to explain migraine (1): At onset of the migraine attack a CSD is initiated in the occipital pole and the CSD propagates anteriorly while depolarizing nerve cells. The depolarization is responsible for the scintillations during the migraine aura, while decreased activity of the nerve cells after CSD is responsible for the scotoma (Fig. 1). Eventually the phenomenon becomes extinct and what is observed using the imaging techniques is reduced CBF.
Electrophysiological evidences that human cortical tissues do support CSD
Bures et al. obtained the first direct evidence that human grey matter in vivo supports CSD (15). They showed that micro-injections of potassium chloride into the caudate nucleus or the hippocampus elicited CSD. A large negative change of the DC potential was observed in the hippocampus, spreading at a rate of 3.2 mm/min away from its site of elicitation, consistent with the propagation rate of CSD in rodents. Similar recordings were obtained in the caudate nucleus. One interesting feature of these recordings is that there was no suppression of the spontaneous electrical activity in the caudate nucleus recorded locally (in the following termed EEG) while the CSD propagated. This is due to technical limitations of the method: the electrodes were not positioned along the path of propagation and the recordings were not bipolar but monopolar (15). Another issue is the pick-up area of the electrodes. The EEG depression is confined to a sphere with a maximum diameter of about 30 mm, which is surrounded by much larger volumes of tissue with preserved EEG activity. This may explain the preserved spontaneous activity due to volume conduction of activity from remote generators. Nevertheless, the DC potential changes were unequivocal and should be taken as definite evidence for the occurrence of spreading depression in human brain. DC changes very similar to CSD have been observed in excized temporal cortical tissue from humans. NMDA receptor antagonists inhibited the reaction. After washing out the blocker CSD could again be elicited (16). Still another line of evidence comes from the work of Mayevsky, who observed repetitive episodes of CSDs in a head-injured patient (in only one of 14 studied) starting approximately 30 h after beginning of the monitoring (17). The recordings showed characteristic transients of potassium, cerebral blood flow, cerebral blood volume, oxidation of the cortex and EEG suppression consistent with CSD. In summary, human cortical tissue supports the development of a CSD, but the report from Mayevsky et al. remains the only description of electrical evidence of CSD from the human neocortex in vivo.
Conclusion
Instruments are now available to detect CSD or CSD-like events non-invasively in animals and in man, but it is still difficult to obtain electrophysiological evidence of CSD or CSD-like reactions in humans. Rigorous protocols are needed. Longer observation times will probably be required in humans. Whatever method that emerges as the preferred one is likely to be less specific than direct imaging of the exposed cortex, and for this reason any method will best be supported by simultaneous use of an alternative method. Eventually this will determine to what extent the large body of knowledge about CSD in animals will be applicable to migraine.