-1 Lhanges in Regional Cerebral Blood Flow During the Course of Classic Migraine Attacks Martin Lauritzen, MD,” Tom Skyhqjj Olsen, MD,? Niels A. Lassen, MD,? and Olaf B. Paulson, MD” Regional cerebral blood flow (rCBF) following carotid arteriography was studied in thirteen patients with classic migraine. Using the ‘jjxenon intraarterial injection method, rCBF was measured in 254 areas in one hemisphere. Nine patients developed a characteristic attack following arteriography and were examined by a series of rCBF studies, spaced by intervals of 5 to 10 minutes. A wave of reduced blood flow originating in the posterior part of the brain and progressing anteriorly was obderved in eight of the nine patients. The oligemia advanced at a speed of 2 mm per minute over the hemisphere, progressing anteriorly but not crossing the rolandic or sylvian sulcus. Typically, the spreading oligemia reached the primary sensorimotor area after symptoms from that area had begun and persisted there long after the focal symptoms had disappeared. The observed time course suggests that the focal symptoms are not secondary to the oligemia. We suggest that focal symptoms and blood flow changes may be secondary to spreading depression of Leao. Lauritzen M, Skyhgij Olsen T, Lassen NA, Paulson OB: Changes in regional cerebral blood flow during the course of classic migraine attacks. Ann Neurol 13:633-641, 1983 evaluation. Informed consent was obtained after full explanaStudies on regional cerebral blood flow (rCBF) during tion of the procedure, including the risk of developing a provoked migrainous headache were obtained by Olesen and colleagues using the intraarterial 1 3 3 ~ e n ~ n migraine attack during the study. One patient was found to have a venous malformation in the left frontal lobe and is not (133Xe) injection method E21). From a group of two included in the following discussion; the remaining twelve hundred fifty patients these researchers selected eight had normal angiograms. Computed tomographic scans and patients in whom K B F studies revealed occipielectroencephalograms were also normal, except in one patoparietal oligemia spreading anteriorly during the intient (No. 452) who had a slightly abnormal pattern with vestigation. They noted that of the eight patients, seven preponderance of 3 Ht activity in the left frontal region. developed migrainous attacks during the investigation; Nine patients developed a classic migrainous attack with focal all seven had a history of spontaneous migraine, six of neurological symptoms and headache after angiography and the classic and one of the common type. The finding of during the subsequent ongoing rCBF study. They constitute the patients described in the following discussion. “spreading oligemia” led to experiments of the rCBF All nine patients were studied in the morning while fasting. changes accompanying Leao’s spreading depression in After the percutaneous carotid arteriography, a catheter (1.2 rats [15). These studies demonstrated that the brief mm outside diameter) was introduced in the internal carotid wave of hyperperfusion accompanying the front was artery using the Seldinger technique. A bolus of 5 to 10 mCi followed by long-lasting moderate cortical oligemia ‘33Xedissolved in 3 ml of saline was injected through the lasting for more than 60 minutes [13]. catheter, and the washout of isotope was recorded by 254 O n this basis, the present clinical studies were undercollimated detectors covering the lateral aspect of one hemitaken to analyze the spreading oligemia as to intensity, sphere [30}. Regional blood flow was thus measured in all localization, speed and mode of spread, and relation to cortical regions supplied by the internal carotid artery. In the clinical symptoms. seven patients with filling of the posterior cerebral artery, the In a series of forty-five patients with hemiplegic migraine, arteriography was carried out in thirteen as part of the clinical area measured included the occipital lobe. In the remaining six patients the most posterior part of the hemisphere in which blood flow was measured corresponded to the posterior part of the parietal lobe and the anterior part of the From the *Department of Neurology, Rigshospitalet, and the tDepartment of Clinical Physiology, Bispebjerg Hospital, Copenhagen, Denmark. Address reprint requests to Dr Lauritzen, Department of Neurology, Rigshospitalet, 9, Blegdarnsvej, DK-2 100 Copenhagen 8, Denmark. Materials and Methods Received July 7, 1982, and in revised form Nov 2, 1982. Accepted for publication Nov 12, 1782. 633 Table I . Medical Histories of Individual PatientJ ~~ ~ Patient No., Age, and Sex Family History Duration of Migraine History (yr) 384, 4 2 , M 417, 37, F No Yes 1 23 440, 26, F 447, 29, F 449, 25, M 450, 25, M 452, 26, F 451, 44, F No Yes No Yes Yes Yes 457, 27, M Yes Neurological Examination Moderate dementia Slight weakness of right arm and leg; hypoesthesia of right hand Normal Normal Normal Normal Normal Slight weakness and hyporeflexia of right arm Normal 8 10 s 8 4 10 6 Table 2. Characteristics of Habitual Attacks in Individual Patients Patient No. Frequency of Attacks Duration of Attacks (hr) 384 limo 24 417 limo 48 440 Variable 24-48 44 7 liwk 48 449 Variable 3-4 450 4iyr 24 452 limo 24 453 2imo 24-48 457 limo 1-2 occipital lobe. Initial slope analysis was used to calculate rCBF 1221, employing the semilogarithmically displayed clearance curves 15 to 60 seconds after injection of the isotope. In each patient, rCBF was measured 5 to 10 times at 5- to 10-minute intervals except between the first and second measurement, when an interval of 20 to 30 minutes was used. The patients were studied during rest (eyes closed, ears plugged), during hypertension induced by intravenous infusion of angiotensin 11, during hyperventilation, and during physiological stimulation (e.g., talking, listening). Intraarterial 634 Annals of Neurology Vol 13 No 6 June 1983 Symptoms fight-sided scintillations and occasional right-sided paresthesias, paresis, and dysphasia, followed by diffuse pressing headache, phonophobia, and nausea Left-sided scintillations; left-sided facial paresthesias and hypoesthesias; left-sided central facial palsy; rightsided arm paresis; hypocsthesias and paresthesias; d ysphasia; left-sided throbbing headache; nausea; vomiting Vertigo; bilateral scintillations; paresthesias of right forehead, chin, and hand; dysphasia; right-sided headache Right-sided scotoma; paresthesias of left hand and arm; right-sided throbbing headache; photophobia; nausea; vomiting Bilateral scintillations; paresthesias of right arm and chin; bifrontal, throbbing headache; ten attacks of amnesia and automatisms not followed by headache Left-sided scintillations; paresthesias and hypoesthesias of left arm, hand, and chin; pressing left-sided headache; nausea; diarrhea; photophobia Paresthesias and heaviness of right arm; dysphasia; left-sided headache; nausea; vomiting Bilateral scintillations; right-sided paresthesias and paresis; dysphasia; throbbing left-sided headache; nausea; vomiting Paresthesias and hypoesthesias of right hand, arm, lips, teeth, and tongue; slight paresis of right arm and hand; bilateral, pressing headache; nausea; photophobia blood pressure and pulse were continuously recorded with a transducer connected to an electromanometer. Arterial carbon dioxide tension (PaCOI) was measured after every rCBF investigation. The rCBF values from the 254 different brain regions were stored as a matrix. A flow distribution map was obtained by dividing each flow value by the mean blood flow for the hemisphere. On the basis of 9557 confidence limits for variations in data from eleven normal brains (retrospectively characterized as neurologically normal), we defined a significant change in the rCBF pattern as a deviation Table 3 . Regional Cerebral Blood Flow in Classic Migraine" Patient No. 384 417 440 447 449 450 452 453 457 No. of lnvestigations Oligemic areah (mi. 100 gm-' .min-') Range' (ml ,100 gm-' min I ) Decrease" (57) .inin- 38 i 3 43 t 2 67 t 2 64 34 t 1 (20-46) (23-58) (56-74) (41-79) (21-42) (23-47) (2 3-43 ) ... (23-45) 24 f 4 29 5 3 19 i 2 26 22 f 2 25 f 2 31 r 3 ... 24 f 3 53 i 6 67 i 5 75 i 2 93 48 i 2 51 5 4 51 i 6 64 t 5 54 2 4 35 i 3 35 2 3 ... 37 *5 . ~ Normal Brain' ( m l . 100 gm- PaC02' linm Hg) MAP' (mm Hg) 43 f 2 124 t s 100 2 3 33 t 1 40 t 1 95 i 7 39 41 ? 1 42 t 1 18 2 I 38 c 1 41 t 1 100 110 t 7 123 i 10 91 i 8 113 2 7 108 2 10 "All values except range are expressed as means 2 standard deviations. "Absolute rCBF values. 'KBF values in pixel with lowest and highest flow values. 'Expressed as percentage of mean hemispheric flows. 'For the corresponding rCBF measurements. PaCOz = arterial cabon dioxide tension; MAP = mean arterial pressure of 15% or more from the expected value. Thus, the term oligemia is used to characterize a region with low flow as compared with other areas in the same hemisphere, and flow in the oligemic areas may therefore be within the normal range for mean hemispheric flow values. Exact registration of the time interval between the individual measurements and the increasing number of channels with reduced rCBF (each covering 1.5 cm2 of brain tissue) made calculation of the rate of spreading of oligemia possible. The direction of spreading was determined by direct inspection of the flow maps. The spreading of the oligemia was related to the central sulcus. Movement of the hand and the mouth (speech) increases the blood flow in this area [ l l , 261; because these tests were performed in seven of the nine patients, we used the corresponding blood flow increases as indications of the location of the central sulcus. Table 4.Oligemic Spread in Classic Migraine in Selected Patients Results The medical histories of the patients are given in Tables 1 and 2. During the migrainous attacks eight of the nine patients developed significant changes in the rCBF. In eight patients spreading oligemia was recorded. The mean decrease in blood flow relative to the mean hemispheric blood flow was 24 t 396 (standard deviation) (Table 3). Variations in PaC02 and mean arterial pressure were negligible and allowed us to average rCBF in the individual patients. The values in Table 3 are for the resting state only. Oligemia began in the posterior part of the brain and progressed anteriorly through the parietal and temporal lobes. In seven patients oligemia also appeared in the frontal lobe, apparently independently of the posterior oligemia; that is, the oligemia did not cross the central or lateral sulcus. The progression rate of the spreading oligemia was calculated to be 2.2 k 0.3 mm per minute (Table 4 ) , assuming a uniform spread of the oligemia within the temporal, parietal, and frontal lobes. Among individ- uals, rCBF in the oligemic areas varied from 34 to 67 mI * 100 gm-' * min-'. In one patient (No. 440) frontal oligemia and parietal hyperemia were recorded in the regions bordering the central sulcus without any spreading. In another patient (No. 453) the decrease in rCBF during a full-blown hemiplegic migraine attack was not statistically significant. Nevertheless, the flow map showed the characteristic pattern of spreading oligemia. Focal symptoms (scintillations, monoparesis or hemiparesis or paresthesia, and sometimes aphasia) did not coincide with regional oligemia in a corresponding primary sensorimotor area. Typically, spreading oligemia began before the patient experienced any focal symptoms and reached the primary sensorimotor area after symptoms from that area had begun. Having reached this area, the oligemia persisted after the focal symptoms had ceased. In addition, the oligemia spread to other regions without giving rise to new focal symptoms. Patient No. 384 417 447 449 450 452 457 No. of Observations Rate of Spread" (mrdmin) 3 4 2.1 ? 0.1 2.6 & 1.0 2.1 2.6 2 0.6 1.8 ? 0.3 1.9 t 0.3 2.5 2 0.9 2.2 ? 0.3 1 4 5 3 3 Mean "On cortex surface. Expressed as means standard deviatlons Lauritzen et al: Spreading Oligemia in Migraine 635 Six patients developed headache only on the side of arteriography, two patients on both sides, and one patient on the side opposite to the arteriography. Five patients developed headache at the time of rCBF investigation, four at a time of global blood flow reduction and one when experiencing parietal hyperemia and frontal oligemia. The induced migraine attacks sometimes differed from the patients’ habitual attacks in the sequence, intensity, or quality of the neurological symptoms. The discrepancies are commented on in the case reports where a detailed account of the symptoms during the rCBF studies is given. Figures 1-6 show rCBF in six patients. Patient 384 During the rCBF measurements the patient experienced scintillations in both visual fields, heaviness of the right arm and leg, and difficulty in finding words (the time course of the symptoms was not recorded). After the investigation headache ensued. (Fig 1) During the first flow measurement (rest), the flow distribution of the frontal, temporal, and anterior parts of the parietal lobe was normal, whereas the flow was low in the posterior part of the parietal lobe. In the Flow 4 measurement (repetition of words), the oligemia progressed anteriorly through the parietal lobe and down into the temporal lobe. A small region of low blood flow in the frontal lobe appeared to be independent of the posterior oligemia. In the regions of low blood flow, notably the auditory cortex, no increase occurred. In the regions of normal blood flow, notably the mouth and supplementary motor areas, blood flow increased as expected during the investigation { 111. In the Flow 5 measurement (listening), the oligemia spread further. Blood flow remained low in the auditory cortex during the test 1271. In the Flow 6 measurement (talking), the oligemia continued to spread in the frontal lobe. During this test the blood flow remained low in the auditory cortex and no longer increased in the mouth area, but it increased in the supplementary motor area in the normally perfused part of the brain 1111. During the Flow 7 measurement (hand movement and counting aloud), blood flow increased in the supplementary motor, hand, and mouth areas but remained low in the superior temporal area of the oligemic regions [ll, LEFT HEMISPHERIC rCBF. 261. This patient exhibited a habitual migraine attack. Oligemia began in the posterior part of the brain and progressed anteriorly without crossing the primary sulci. Activation of rCBF was impaired in the oligemic regions. COMMENT. 636 Annals of Neurology Vol 13 No 6 June 1983 Patient 41 7 LEFT HEMISPHERIC rCBF. (Fig 2) During the first flow measurement (rest), the patient displayed no symptoms. The blood flow was normal in the frontal, temporal, and anterior parts of the parietal lobe but low in the posterior part of the parietal lobe. During the Flow 2 measurement (conversation), the patient complained of paresthesias and hypoesthesias of the right arm and the face, slight paresis of the right arm, and a slight right-sided facial palsy. The oligemia spread anteriorly through the parietal lobe and into the temporal lobe. Blood flow increased in the mouth and supplementary motor areas and in the frontal lobe but remained low in the auditory cortex in the oligemic region Ell]. During the Flow 3 measurement (rest), the symptoms were unchanged. The oligemia had progressed further through the parietal and temporal lobes. Paresis did not disappear until the Flow 5 measurement (eyes open), but the sensory disturbances persisted. Blood flow was not activated in the oligemic parietal lobe [lS}. In the Flow 6 and 7 measurements (right hand gripping, not shown), symptoms and flow distribution were unchanged. Then, during the Flow 8 measurement (rest), the patient exhibited no symptoms. The oligemia persisted in the parietal and temporal lobes, and low blood flow also appeared in the anterior part of the frontal lobe. Ten minutes after the investigation, the patient developed severe throbbing left-sided headache similar to her usual migraine headaches. The patient developed a migraine attack slightly different from her habitual attacks in focal symptoms (she did not experience scintillations or symptoms of the left side of the face). The headache on the side of arteriography was identical to her migraine headaches. Oligemia began in the posterior part of the brain and progressed anteriorly without crossing the primary sulci. Activation of rCBF was impaired in the oligemic areas. The region of oligemia and the focal symptoms did not correlate. COMMENT. Patient 440 During the rCBF studies, this patient developed her habitual focal symptoms and a severe bilateral headache. The rCBF abnormalities differed from those of the other patients in that no spreading oligemia was recorded. The observed blood flow abnormalities were Figs 1-6. Computer-generated,color-scaled images of regional cerebral bloodjow in each of six patients. To the right, a scale shws the relation between colors and d B F in ml. 100 gm-’ min-’. The images are turned with the nose to the left. PaCO2, blood pressure, and other data are given in Table 3. Description of the relationship between symptoms and K B F changes is presented in the case reports in the Results section. - a parietal hyperemia and a small frontal oligemia. The focal symptoms and the rCBF changes did not correlate. No image is shown of rCBF in this patient. Patient 44 7 During the rCBF study this patient developed a migraine attack with bilateral hand paresthesias, followed by headache after the investigation. The headache, on the side of arteriography, was identical to her usual migraine headaches, although the focal symptoms were dissimilar in that she experienced no visual disturbances. Oligemia starting in the posterior part of the brain and spreading anteriorly was noted in two studies. The focal symptoms and the region of oligemia did not correlate. No image is shown of rCBF in this patient. Patient 449 (Fig 3) During the Flow 1 measurement (rest), the patient displayed no symptoms and the flow distribution was normal. In the Flow 2 measurement (rest; not shown), a low-flow area close to the central sulcus was observed which spread posteriorly during the Flow 3 measurement (rest). During these two studies the patient complained of right-sided paresthesias. In the Flow 5 measurement (rest), the focal symptoms were unchanged, but the patient at this time complained also of a right-sided headache. The oligemia had spread through the temporal, parietal, and frontal lobes and continued to do so in the Flow 7 and 8 measurements, during which the paresthesias disappeared and the right-sided headache was the only symptom. LEFT HEMISPHERIC rCBF. COMMENT. The invoked migraine attack differed from the patient’s habitual attack in that he had not previously experienced symptoms in his right leg. The headache was opposite to the side of arteriographyright sided-whereas his habitual migraine headaches were bilateral. Oligemia began in the central part of the brain and spread in all directions without crossing the primary sulci. The headache occurred simultaneously with extensive oligemia. The regions of oligemia did not correlate to the focal symptoms. Patient 450 RIGHT HEMISPHERIC rCBF. (Fig 4 ) During the first two flow measurements, the patient displayed no symptoms. In the Flow 1 measurement (rest), the flow distribution was normal, whereas in the Flow 2 measurement (not shown) the flow in the posterior part of the parietal lobe was low. In the Flow 3 measurement (rest), the oligemia spread through the parietal and temporal lobes, and ascending paresthesias from the left hand reached the elbow. In the Flow 4 measurement (rest), the oligemia did not progress, and the patient now complained of paresthesias of his left arm and 638 Annals of Neurology Vol 13 No 6 June 1983 chin. During the Flow 6 measurement (rest), the paresthesias were very weak, and a right-sided headache was beginning. The oligemia had progressed further through the temporal and parietal lobes. The focal symptoms disappeared during the Flow 7 measurement (sensory hand test), but the headache worsened. A low-flow area in the inferior part of the frontal lobe was apparent at this time. During this test the blood flow increased normally in the area close to the central sulcus in the normally perfused parts of the brain 1251. During the Flow 8 and 9 measurements (rest), the headache intensified and the patient complained of paresthesias of the left arm. The low blood flow now occupied most of the hemisphere. COMMENT. This patient developed a migraine attack, although the headache was right sided, on the side of arteriography, whereas his habitual migraine headaches were left sided. Oligemia began in the posterior part of the brain and progressed anteriorly without crossing the primary sulci. The headache occurred when the oligemia occupied almost the entire hemisphere. The region of oligemia did not correlate with the focal symptoms. Patient 452 (Fig 5) During the first flow measurement (rest), the patient displayed no symptoms. The flow distribution was normal in the frontal and temporal lobes but low in the posterior part of the parietal lobe. During the Flow 3 measurement (rest), the patient complained of paresthesias and heaviness of the right arm. The oligemia had now progressed through the parietal and temporal lobes. A small region of low flow in the middle part of the frontal lobe appeared to be independent of the posterior oligemia. In the Flow 5 and 6 measurements, the patient complained of a left-sided, severe, throbbing headache; the focal symptoms had completely disappeared. The oligemia continued to spread in these two measurements to include most of the hemisphere and remained low in the oligemic regions during the hand test of the Flow 5 measurement [26]. LEFT HEMISPHERIC rCBF. COMMENT. The invoked migraine attack was similar to the patient’s habitual migraine attacks. The low-flow area began in the posterior part of the brain and progressed anteriorly without crossing the primary sulci. There was no correlation between the focal symptoms and the region of oligemia. The headache occurred simultaneously with extensive oligemia. The activation of rCBF was impaired in the oligemic regions. Patient 4.53 During the rCBF studies this patient developed a migraine attack with paresthesia and hypoesthesia and paresis of her right arm and, later, a moderate left-sided headache. The headache was on the side of arteriography and identical to her migraine headaches. The reduction in blood flow was not significant according to our definitions; that is, it was below 15%. The patient did, however, develop oligemia in the posterior part of the brain which progressed anteriorly without crossing the primary sulci. No correlation was found between the focal symptoms and the rCBF changes. The headache occurred at a time when the blood flow was decreased. One day before the investigation the patient suffered a spontaneous migraine attack, which may account for the relatively small blood flow decrease. No image is shown of rCBF in this patient. Patient 45 7 LEFT HEMISPHERIC rCBF. (Fig 6) This patient displayed no migraine symptoms during the rCBF studies. In the first flow measurement (rest), the blood flow in the frontal and anterior parts of the parietal -lobe was normal, but flow was low in the posterior part of the parietal and temporal lobes. In the Flow 3 measurement (rest), the oligemia had progressed through the temporal and parietal lobes, and a region of low flow appeared in the middle part of the frontal lobe. The oligemia was unchanged in the Flow 5 measurement (rest). In the Flow 6 measurement (hand gripping with eyes open), blood flow increased normally in the hand area but remained low in the posterior part of the parietal lobe El8, 261. Forty-five minutes after the investigation, the patient experienced a migraine attack, with paresthesias and hypoesthesias of his left arm and chin accompanied by bilateral headache. Three hours after the investigation he complained of paresthesias and hypoesthesias of his right arm and face, similar to his habitual prodromes. COMMENT. The focal symptoms of the invoked attack were primarily left sided, as opposed to the rightsided symptoms occurring during his habitual attacks. The headache was identical to his usual migraine headache. Although the considerable delay between onset of progressive oligemia and of symptoms is difficult to explain, it may have been caused by transient inhibition of the oligemia at the postcentral sulcus. Oligemia began in the posterior part of the brain and progressed anteriorly without crossing the primary sulci. Activation of rCBF was impaired in the low-flow areas. Discussion All patients studied fulfilled the US Public Health Service criteria for having classic migraine [ 11; intracranial processes were ruled out by the extensive investigation program. Nine of the thirteen patients developed migraine during or immediately following the rCBF investigation. No differences in habitual migraine symptoms existed between the patients who did and those who did not develop migraine during the study. All nine patients developed symptoms after arteriography, one before and eight after the introduction of the catheter in the internal carotid artery. Several mechanisms may have provoked the attacks. The short-lasting injection of hyperosmolar contrast medium does not impair the blood-brain barrier 1231but may give rise to a transient perturbation of cerebral blood flow and autoregulation [S] and to dehydration of the perivascular space 1241. The carotid puncture and the presence of the catheter do not change rCBF per se but may initiate vascular reflexes in susceptible subjects and thereby initiate an attack. The moderate discrepancy between the focal neurological symptoms of the induced and those of the habitual attacks raised the important question of whether a series of events was brought about which was accompanied by migrainelike symptoms or whether real migraine attacks were induced. The question cannot be answered fully until migraine patients have been studied atraumatically during the whole course of their spontaneous attacks. Because the disparities in symptoms were minor, we believe that the attacks were of migraine origin and that what we actually studied were pathophysiological components of migraine. K B F Changes and Migruine Symptoms The current view of migraine pathophysiology suggests that vasospasm causes cerebral ischemia and the focal symptoms during migraine prodromes, whereas reactive hyperemia accompanies the headache 141. Studies of cerebral blood flow in migraine are largely supportive 117, 20, 28, and other studies], although the recent report by Olesen and colleagues {2 11 reached a different conclusion. The present study demonstrates that the regional hypoperfusion in migraine appears before and outlasts the focal symptoms. The dissociation in time and place of the focal symptoms and the oligemia strongly suggests that the prodromes and accompaniments were not caused by the oligemia. Small areas of ischemia in the apparently homogeneously perfused oligemic region were observed in six patients; in addition, they cannot be excluded in the remaining three patients, because Compton scatter from the surrounding tissue and radiation from the underlying, normally perfused tissue tend to hide small hypoperfused areas [ S , lo]. Nevertheless, the mean regional perfusion in the oligemic areas was close to or above 40 ml * 100 gm-' * min-l, well above threshold for producing neuronal damage by ischemia (18 to 20 ml. 100 gm-' * min- ') 1311. This quantitative consideration supports our conclusion that ischemia is not of primary importance in the development of focal migraine symptoms. Four of the five patients developed headache during the rCBF study at the time of global oligemia. The Lauritzen et al: Spreading Oligemia in Migraine 639 headache may relate to either the size of the oligemic area or the time course of the unknown underlying process. Spreading Oligemia in Migraine Olesen and colleagues [2 11 described rCBF changes in induced migraine attacks using the same equipment employed in the present study. The changes included oligemia beginning in the posterior part of the brain and spreading anteriorly, irrespective of the supply territories of large arteries (spreading oligemia). Focal hyperemia sometimes preceded the oligemia. The present study corroborates the observation that spreading oligemia is a characteristic change in rCBF in classic migraine attacks. We calculated its rate of progression within the individual lobes to be close to 2 mm per minute. Spreading oligemia did not cross the central or lateral sulcus and was thus stopped by the primary sulci. The central sulcus represents a major cytoarchitectonic change: the eulaminate cortex of the postcentral region is abruptly replaced by the agranular cortex of the precentral region 12). The lateral sulcus represents a major change in the macroscopic structure, in that it separates the temporal lobe from the parietal and frontal lobes. The cytoarchitectonic structure of the temporal and parietal lobes is, however, not different from the structure of the insula, and the lateral sulcus is therefore not a major microstructural border 121. It was evident from inspection of the flow maps that the areas of frontal oligemia were not directly connected with the posterior oligemic areas. Because the frontal oligemia always appeared to correspond to the frontal or orbital operculum, it could have been caused by a spread of parietal or temporal oligemia through the insula. If so, the development of rCBF changes in migraine follows rules determined by the cytoarchitectonic structure of the cortex. We speculated that the oligemia was secondary to a perturbation of neuronal function that spread slowly across the brain. These characteristics led to further considerations of the possible relation to Leao’s spreading depression TlS]. Spreading Depression and Spreading Oligemia The concept that spreading depression (SD) is involved in migraine pathophysiology is not new. From an empirical viewpoint Gowers [7] and later Lashley 112) described the nature of potential neuronal disturbance underlying migraine prodromes, later noted to correspond in their time of spread to SD [lb, 191. Cortical SD in rat brain is followed by a persistent cerebral hypoperfusion 1131, a fact that justified an analysis of the spatial and temporal course of spreading oligemia in terms of its being a reflection of SD. In these patients spreading oligemia was noted to begin in the 640 Annals of Neurology Vol 13 No 6 J u n e 1983 posterior part of the brain, the sensory cortex, where the density of neurons is highest [2], corresponding to the observation that SD is more easily elicited in brains of high neuronal density than in brains of low neuronal density 131. Spreading oligemia progressed at a rate of 2 mm per minute over the cortex, corresponding to the known progression rate of SD (close to 3 mm per minute) [3]. The calculation of the progression rate of oligemia has the inherent error of neglecting the “foldedness” of the convoluted brain; thus, 2 mm per minute may be an underestimation. Spreading oligemia did not cross the central or lateral sulcus but probably circumvented the major macrostructural and microstructurd changes by passing through the insula to the frontal lobe, a mechanism similar to that of SD when stopped by similar structural changes 131. Having reached the differently composed cortex, the oligemia spread in a regular fashion that corresponds to the regular spread of SD within brain regions of different microstructure [b}.The uniform 25% reduction in rCBF in oligemic areas paralleled the 20 to 25% reduction in blood flow in rat brain [13]. The focal hyperemia previously found to precede oligemia in some patients with migraine 121) was not observed in this study. This hyperemia may correspond to the intense hyperemia accompanying the SD front and lasting for 1 to 3 minutes 191. The SD wavelength is 2 to 3 mm in the animal brain t31; in human nucleus caudatus and hippocampus it is also short but may vary [291, and exact values for the neocortex are unknown. Because the scintillation channels each cover 1.5 cm2 of the brain, such a small and short-lasting hyperemia could have been overlooked. Thus, the time course and intensity of the rCBF reduction in the induced migraine attacks suggest that spreading oligemia is not the cause of focal symptoms in classic migraine. The headache occurred at a time of cerebral hypoperfusion. How this hypoperfusion relates to the sensation of pain is unknown. The progression rate of spreading oligemia and its dependence on major macrostructural and microstructural changes may indicate a relation of Leao’s spreading depression. If so, spreading oligemia may be an epiphenomenon accompanying spreading depression, and focal migraine symptoms may result from the prolonged inhibition of cortical neurons after SD E3f. What determines the intensity and quality of the focal symptoms is unknown. The metabolic state of the cerebral cortex prior to the attack may be important. Presented in part at the 24th Scandinavian Congress of Neurology, Copenhagen, Denmark, June 9-12, 1982, and in ref 14. The authors thank secretaries Birgit Cederberg and Bodil K j m for the preparation of the manuscript. References 1. Ad Hoc Committee on Classification of Headache: Classification of headache. JAMA 179:717-718, 1962 2. Bailay P, von Bonin G: The Isocortex of Man. Urbana, University of Illinois Press, 1951 3. Bures J, Buresova 0, Krivanek J: The mechanism and applications of Leao’s spreading depression of electroencephalographic activity. New York, Academic, 1974 4. Dalessio DJ: Wolffs Headache and Other Head Pain. New York, Oxford, Oxford University Press, 1980 5. Donley RF, Sundt TM, Anderson RE, Sharbrough Fw:Blood flow measurements and the “look through artefact in cerebral ischemia. Stroke 6:121-131, 1975 6. Fifkova E: Spreading EEG depression in the neo-, paleo- and archicortical structures of the brain of the rat. Physiol BohemosIOV 13:l-15, 1964 7. Gowers WR: The Borderland of Epilepsy. London, Churchill, 1907 8. Grubb RL,Hernandez-Peres MJ, Raichle ME, Phelps ME: The effects of iodinated contrast agents on autoregulation of cerebral blood flow. Stroke 5:155-160, 1974 9. Hansen AJ, Quistorff B, Gjedde A: Relationship between local changes in cortical blood flow and extracellular K’ during spreading depression. Acta Physiol Scand 109: 1-6, 1980 10. Hanson EJ, Anderson RE, Sundt TM: Comparison of 85Krypton and 133Xenoncerebral blood flow measurements before, during and following incomplete ischemia in the squirrel monkey. Circ Res 36:18-26, 1975 11. Larsen B, Skinhej E, Lassen NA: Variation in regional cortical blood flow in right and left hemispheres during automatic speech. Brain 101:19~-209, 1978 12. Lashley Ks: Patterns of cerebral integration indicated by the scotomas of migraine. Arch Neurol Psychiatry 46:33 1-339, 1941 13. Lauritzen M, Jergensen MB, Diemer N H , Gjedde A, Hansen AJ: Persistent oligemia of rat cerebral cortex in the wake of spreading depression. Ann Neurol 12:469-474, 1982 14. Lauritzen M, Skyhej Olsen T, Paulson OB: Characteristics of the gradually spreading reduction of regional cerebral blood flow in classic migraine. Acta Neurol Scand (Suppl 90) 65:70-71, 1982 15. Lea0 AAP: Spreading depression of activity in cerebral cortex. J Neurophysiol 7:359-390, 1944 16. Leao AAP, Morrison RS: Propagation of spreading cortical depression. J Neurophysiol 8:33-45, 1945 17. Mathew NT, Krastnik F, Meyer JS: Regional cerebral blood flow in the diagnosis of vascular headache. Headache 15:252-260, 1976 18. Melamed E, Larsen B: Cortical activation pattern during saccadic eye movements in humans: localization by focal cerebral blood flow increases. Ann Neurol 5:79-88, 1979 19. Milner PM: Note on a possible correspondence between the scotomas of migraine and spreading depression of Leao. Electroencephalogr Clin Neurophysiol 10:705, 1958 20. OBrien M D Cerebral blood changes in migraine. Headache 10:139-143, 1971 21. Olesen J, Larsen B, Lauritzen M: Focal hyperemia followed by spreading oligemia and impaired activation of rCBF in classic migraine. Ann Neurol9:344-352, 1981 22. Olesen J, Paulson OB, Lassen NA: Regional cerebral blood flow in man determined by the initial slope of the clearance of intraarterially injected 13%Xe.Stroke 2:519-540, 1971 23. Paulson OB, Hertz MM: Blood-brain barrier permeability during shortlasting intravascular hyperosmolarity. Eur J Clin Invest 8~391-396, 1978 24. Paulson OB, Hertz MM, Bolwig TG, Lassen NA: Filtration and diffusion of water across the blood-brain barrier in man. Microvasc Res 13:113-124, 1977 25. Roland PE, Larsen B: Focal increase of cerebral blood flow during stereognostic testing in man. Arch Neurol 33:551-558, 1976 26. Roland PE, h s e n B, Skinhaj E, Lassen NA: The role of the supplementary motor area and other cortical areas in the organization of voluntary movement in man. J Neurophysiol43:118136, 1980 27. Roland PE, Skinhej E, Lassen NA: Focal activations of human cerebral cortex during auditory discrimination. J Neurophysiol 45:1139-1151, I981 28. Skinhoj E: Hemodynamic studies within the brain during migraine. Arch Neurol 29:95-98, 1973 29. Sramka M, Brozek G, Bures J, Nadvormik P Functional ablation by spreading depression: possible use in human stereotactic neurosurgery. Appl Neurophysiol40:48-61, 1977-1978 30. Sveinsdottir E, Larsen B, Rommer P, Lassen NA: A multidetector scintillation camera with 254 channels. J Nucl Med 18:168174, 1977 31. Trojaborg W, Boysen G: Relation between EEG, regional cerebral blood flow and internal carotid artery pressure during carotid endarterectomy. Electroencephalogr Clin Neurophysiol 34:61-69, 1973 Lauritzen et al: Spreading Oligemia in Migraine 641