Magnetic resonance imaging of the brain in patients with migraine
Hisaka Igarashi, Fumihiko Sakai, Shinichi Kan, Jun Okada, Yoshiaki Tazaki
Cephalalgia Igarashi H, Sakai F, Kan S, Okada J, Tazaki Y. Magnestic resonance imaging of the brain in patients with migraine. Cephalalgia 1991;11:69-74. Oslo. ISSN 0333-1024 Magnetic resonance imaging (MRI) was studied in 91 patients with migraine and in 98 controls. Risk factors known to cause MRI lesions were carefully examined. In 36 patients with migraine (39.6%), small foci of high intensity on T2-weighted and proton-density-weighted images were seen in the white matter. Of patients with migraine who were less than 40 years old and without any risk factor, 29.4% showed lesions on MRI; this was significantly higher than the 11.2% for the group of age-matched controls (n = 98). The lesions were distributed predominantly in the centrum semiovale and frontal white matter in young patients, but extended to the deeper white matter at the level of basal ganglia in the older age group. The side of the MRI lesions did not always correspond to the side of usual aura or headache. Migraine-related variables such as type of migraine, frequency, duration or intensity of headache or consumption of ergotamine showed no significant correlation with the incidence of MRI abnormalities. Our data indicated that migraine may be associated with early pathologic changes in the brain. • Anti-cardiolipin antibody, magnetic resonance imaging, migraine, white matter lesion Hisaka Igarashi, Fumihiko Sakai, Jun Okada, Yoshiaki Tazaki, Department of Medicine; Shinichi Kan, Department of Radiology, School of Medicine, Kitasato University; Correspondence to Hisaka lgarashi, M.D., Department of Medicine, Kitasato University 1-15-1 Kitasato, Sagamihara, Kanagawa, Japan (228); Accepted 11 December 1990 Reduction of regional cerebral blood flow below the level of ischaemic threshold has been reported to be responsible for the aura of migraine (1-4). It is clinically important to know whether or not the repetition of migraine pathophysiology may cause permanent ischaemic changes in the brain. Some previous studies by CT scan have shown brain atrophy in patients with a-long history of frequent migraine attacks (5-7). Whilst the significance of this result has been questioned in other studies (8, 9), some recent studies using MRI have suggested parenchymal abnormalities in the brain (10, 11). We studied magnetic resonance imaging (MRI) of the brain in a large number of patients with migraine in order to investigate if migraine causes early parenchymal changes. Other risk factors known to cause MRI lesions were carefully examined and the data were compared with age-matched controls Subjects and methods
A total of 91 patients with migraine were studied. Diagnosis of migraine was made according to the criteria of the Headache Classification Committee of the International Headache Society (12). Fifty-three had migraine with aura and their mean age was 33.2 ± 14.6 (mean ± SD) years. Thirty-three patients had migraine without aura (36.7 ± 11.9 years) and 5 had migraine with prolonged aura (24.8 ± 5.7 years). Their aura symptoms with visual disturbances followed by hemiparaesthesia (n = 5) and dysphasia (n = 2) lasted 1 to 12 h. The mean duration of the history of migraine was 13.4 ± 11.6 years and there was no significant difference between different groups. The control group consisted of a random sample of 98 subjects from the consecutive studies during the study period of the patients with migraine. This group was investigated because of dizziness, tension-type headache, idiopathic focal spasm, psychological disorders or history of epilepsy. They were less than 40 years old (32.4 ± 18.2 years) and did not have any focal neurological signs. Risk factors which may possibly cause lesions on MRI were carefully examined clinically and in laboratory examination in all the patients with migraine. They included multiple sclerosis, diabetes mellitus, hypertension, collagen disease, valvular heart disease, hyperlipidemia and polycythemia. Serological tests for syphilis, LE cells, antinuclear antibody and anti-ENA (extractable nuclear anti-gens) were also carried out. IgG anti-cardiolipin antibody was measured in 41 patients with migraine by ELISA and was evaluated using Harris's standard cut-off point (13). MRI studies were done in 33 of these 41 patients. MRI studies were carried out with the RESONA 0.5 Tesla super-conductive MRI unit (Yokogawa Medical Systems) using 2000/120(TR/TE) T2-weighted images, 2000/60(TR/TE) proton density weighted images and 500/25 T1-weighted images. Slice thickness was 10 mm. Assessment of MRI was
Table Incidence of positive MRI studies in patients with migraine. Duration of Positive Type of migraine n (men:women) Age (years) history MRI Brain atrophy (years) studies Migraine with aura 53 (19:34) 33.2 ± 14.6 13.1 ± 12.7 23 (43.4%) 0 Migraine without aura 33 (7:26) 36.7 ± 11.9 14.3 ± 10.1 11 (33.3%) 1 Migraine with prolonged aura 5 (3:2) 24.8 ± 5.7 11.2 ± 6.4 2 (40.0%) 0 Total 91 (29:62) 34.0 ± 13.6 13.4 ± 11.6 36 (39.6%) 1 Positive MRI studies: focal areas of high intensity images in the white matter on both T2-weighted and proton-density-weighted. made by two observers, and discrepancies were resolved by a third judge. All patients had MRI studies done during the interictal period of migraine. None of the patients, including those with migraine with prolonged aura, had any neurological deficit at the time of the MRI study. The statistics used were the Student's t-test for parametric variables and the Mann-Whitney U test for non-parametric variables. Results
Table 1 summarizes the incidence of positive MRI studies in patients with migraine. The characteristics of the lesion were focal areas of high intensity on both T2-weighted and proton density weighted MR images distributed bilaterally in the white matter of the brain. Etat crible or Virchow-Robin's space was carefully excluded from the positive findings (14). Positive MRI studies were seen in 23 out of 53 patients (43.4%) with migraine with aura, in 11 out of 33 patients (33.3%) with migraine without aura, and in 2 out of 5 patients (40.0%) with migraine with prolonged aura. Altogether, 36 out of 91 patients (39.6%) with migraine showed small white matter lesions on MRI. Only one patient showed brain atrophy. The incidence of MRI lesions was rot significantly different among different types of migraine. Figs 1 and 2 demonstrate typical MRI lesions in our patients. Fig. 1 is a 42-year-old woman with migraine with aura. Since the age of 14 she has had severe migraine headaches preceded by visual aura. Her blood pressure was normal, and there was no risk factor for atherosclerotic or other diseases known to be responsible for MRI lesions. Multiple small high-intensity lesions were seen in bilateral coror a radiata and centrum semiovale. Fig. 2 is that of a 26-year-old woman with migraine with aura. She had a 7-year history of frequent attacks of migraine preceded by scintillating scotoma. CT scan, electroencephalogram and all the blood tests for risk factors were negative. MRI also showed small high intensity lesions in both centrum semiovale. Patients with positive MRI studies were compared with those with negative MRI studies for various physiological parameters (Table 2). The mean age of the MRI positive group was 39.4 ± 15.2 years and was significantly older than the MRI negative group (30.1 ± 11.0 years). There were no significant differences between the two groups as regards the duration of the history, the frequency of migraine attacks or the amount of consumption of ergotamine. Risk factors were found in 16 patients with migraine (for details see Table 2). The incidence of these risk factors was higher in the MRI positive group (22.2%) than in the MRI negative group (14.8%), but the difference was not statistically significant. The time elapsed between the last migraine attack and the MRI study was no different between the MRI positive group (38.1 ± 39.2 days) and the MRI negative group (33.7 ± 25.2 days). Anti-cardiolipin antibody was abnormally high in only 3 patients (Table 3); 2 of these were MRI positive, the remaining patient MRI negative. Those patients without any risk factor were subjected to further analysis. Fig. 3 illustrates the frequency of MRI lesions comparing the different age groups. The frequency of white matter lesions increased with advancing age. While MRI studies were positive in all the patients over 60 years of age, the incidence of a positive MRI study was high even in younger patients. Lesions on MRI were present in 29.4% of patients under 40 years. The distribution of MRI lesions in all patients was plotted in the different brain levels (Fig. 4). In patients younger than 40 years the distribution of the lesion was predominantly in the centrum semiovale and frontal white matter (Fig. 4A), but it extended to the deeper white matter in patients aged 40 or over (Fig. 4B). The side of lesions on MRI did not always correspond to the side of the usual aura or headache. The distribution of lesions was no different among different types of migraine. Table 4 shows the comparison between patients with migraine and controls. Both groups included only those subjects younger than 40 years who did
not have risk factors. The percentage of positive MRI studies in patients with migraine became 29.4% but was still significantly greater than the 11.2% in age-matched controls. Discussion
In the present study MRI revealed small areas of high intensity in the white matter of the brain in patients with migraine, suggesting early parenchymal lesions. The overall frequency of lesions detected by MRI studies was 39.6% and was similar to the previous report by Soges et al. (11) who studied 24 patients with migraine and found 46% of them showing abnormal MRI studies. Their slightly higher percentage for positive MRI studies may be explained by the slightly older mean age (36.8 years) than our patients (34.0 years). The importance of our study is that we studied a large number of patients with migraine and compared these with similarly studied age-matched controls. A group of patients with migraine under 40 years old without risk factors showed a significantly higher incidence of positive MRI changes (29.4%) than the controls (11.2%). In order to discover whether lesions on MRI are the result of the migraine pathophysiology or of other variables, we screened all other possible risk factors by clinical and laboratory examination. The comparison of the two groups with either positive MRI or negative MRI revealed that ageing was the only variable significantly different. Risk factors such as hypertension, hyperlipidemia or others were not significantly different between the groups. Anti-cardiolipin antibody, which has been reported to be one of the important risk factors for cerebrovascular diseases in young subjects (15, 16), also showed no correlation with the occurrence of MRI lesions. The characteristics of MRI lesions in migraine were well defined small foci on T2-weighted and proton density weighted images in the white matter. These white matter lesions resemble so-called UBOs, unidentified bright objects (17) which are often related to ageing, hypertension and other arteriosclerotic risk factors. However, the pathogenesis of the lesions on MRI in migraine seemed to be different from the lesions caused by arteriosclerotic or atheromatous mechanisms, because our patients consisted of a young population and did not
Table 2 Comparison between MRI positive group and negative group. Duration Frequency Total Risk n Age (years) of of ergotamine factors history (years) attacks (month)consumption (tablets) Positive group 36 39.4 ± 15.2* 15.7 ± 13.7 2.7 ± 5.2 54.7 ± 169.2 8 (22.2%) Negative group 54 30.1 ± 11.0 11.6 ± 9.5 3.2 ± 3.5 62.0 ± 159.9 8 (14.8%) MRI positive group: Patients with small high intensity lesions. A patient with brain atrophy was not included. Risk Factors: hypertension (6), TIA (2), diabetes mellitus (1), mitral valve insufficiency (1), hyperlipidemia (3), thrombocytopenia (1), positive LE (1), positive ANA (3), positive anti-cardiolipin antibody (3). * p < 0.01 (positive group vs. negative group: Student's t-test). have risk factors such as hypertension or diabetes mellitus. None of the patients showed any sign of dementia. In addition, lesions on MRI in younger patients with migraine were seen less frequently in the basal ganglia, where lacunar infarctions are usually seen. The most tempting speculation to explain the parenchymatous lesions in migraine is that repeated attacks of hypoperfusion during aura cause permanent ischaemic changes. This assumption was proved to be an unlikely mechanism because lesions were no more frequent in migraine with aura than in migraine without aura. Neither did the frequency of the lesion correlate with such variables as the
Table 3. Anti-cardiolipin antibody and MRI in patients with migraine. MRI lesion Anti-cardiolipin n Positive Negative antibody Positive 3 2 1 Negative 30 11 19 Total 33 13 20
Anti-cardiolipin antibody: ELISA (Workshop 1986). Cut-off point of 5GCL unit.
Table 4. Incidence of MRI lesions: migraine vs. controls; younger than 40 years and without risk factors. No. of Percent of n positive positive MRI studies MRI studies Migraineurs 51 15 29.4%* Controls 98 11 11.2%
* p < 0.05 (Student's t-test). frequency of attacks, the duration of the disease, or the type of migraine. Lesions in the brain were not localized to the occipital region but distributed diffusely in the centrum semiovale and corona radiata, If ischaemia has any relation to the present finding, it has to be of global hypoperfusion and not of regional hypoperfusion. One possible mechanism is the reduction of cerebral perfusion pressure due to hypotension during orthostatic posture or other causes. Not infrequently do patients with migraine experience orthostatic hypotension and may suffer from fainting sensation or even syncopal episodes. This hypothesis, however, remains to be investigated. Another possible mechanism which may cause patchy white matter lesions is the production of multiple microemboli. During attacks of migraine increased aggregability of platelets has been shown by several investigators (18-21). Abnormal platelet function has also been suggested. It is possible that increased platelet aggregation may trigger the formation of multiple microemboli. In addition, vasoactive substances such as 5-HT are released from the platelets and may cause secondary vasoconstriction (22) which may produce cerebral parenchymatous lesions. Currently, one of the major hypotheses is that migraine is a vascular headache (23), While it should be reasonable to relate our MRI findings to the vascular events, there is the remaining possibility that they are the result of primary neuronal damage related to the migraine pathophysiology. There are investigators who suggest the neural mechanism of migraine (24-26). Although the present study did not provide data to suggest certain aetiology for migraine attacks, it suggested that there was a significant risk in migraine for the early occurrence of parenchymatous lesions on MRI. Further studies are necessary to clarify the significance of changes in the MRI in the pathophysiology and treatment of migraine. References
1.
Skinhøj E. Hemodynamic studies within the brain during migraine. Arch Neurol 1973;29:95-8
2.
Norris JW, Hachinski VC, Cooper PW. Changes in cerebral blood flow during a migraine attack. Br Med J 1975;3:676-7
3.
Sakai F, Meyer JS. Regional cerebral hemodynamics during migraine and cluster headaches measured by the 133Xe inhalation method. Headache 1978;18:122-32
4.
Clesen J, Larsen B, Lauritzen M. Focal hyperemia followed b,, spreading oligemia and impaired activation of rCBF in classic migraine. Ann Neurol 1981;9:344-52
5
Hangerford GD, du Boulay GH, Zilkha KJ. Computerised axial tomography in patients with severe migraine: a preliminary report. J Neurol Neurosurg Psychiatry 1976;39:990- 4
6.
Mathew NT, Meyer JS, Welch KMA, Neblett CR. Abnormal CT-scans in migraine. Headache 1977;16:272-9
7.
du Boulay GH, Ruiz JS, Rose FC, Stevens JM, Zilkha KJ. CT changes associated with migraine. AJNR 1983;4:472-3
8.
Masland WS, Friedman AP, Buchsbaum HW. Computerized axial tomography of migraine. Res Clin Stud Headache 1973;6:136-40
9.
Cuetter AC, Aita JF. CT scanning in classic migraine. Headache 1983;23:195
10.
Kaplan RD, Solomon GD, Diamond S, Freitag FG. The role of MRI in the evaluation of a migraine population: preliminary data. Headache 1987;27:315-18
11.
Soges LJ, Cacayorin ED, Petro GR, Ramachandran TS. Migraine: Evaluation by MR. AJNR 1988;9:425-9
12.
Headache Classification Committee of the international Headache Society: Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988;(suppl 7):1-96
I3.
Harris EN, Gharavi AE, Patel SP, Hughes GRV. Evaluation of the anti-cardiolipin antibody test: report of an international workshop held 4 April 1986. Clin Exp Immunol 1987;68:215- 22
14.
Jungreis CA, Kanal E, Hirsch WL, Martinez AJ, Moossy J. Normal perivascular spaces mimicking lacunar infarction: MR imaging. Radiology 1988;169:101-4
15.
Syrjänen J, Vaarala O, Iivanainen M, Palosuo T, Valtonen W, Aho K. Anticardiolipin response and its association with infections in young and middle-aged patients with cerebral infarction. Acta Neurol Scand 1988;78:381-6
16.
Levine SR, Deegan MJ, Futrell N, Welch KMA. Cerebrovascular and neurologic disease associated with antiphospholipid antibodies: 48 cases. Neurology 1990;40:1181-9
17.
Bradley Jr WG. Investigators solve mystery of unidentified bright spots. Diagn Imaging 1988;57:322-6, 446
18.
Hanington E, Jones RJ, Amess JAL, Wachowicz B. Migraine: a platelet disorder. Lancet 1981;2:720-3
19.
Kalendovsky Z, Austin JH. "Complicated migraine". Its association with increased platelet aggregability and abnormal plasma coagulation factors. Headache 1978;15:18-35
20.
Couch JR, Hassanein RS. Platelet aggregability in migraine. Neurology 1977;27:843-8
21.
Deshmukh SV, Meyer JS. Cyclic changes in platelet dynamics and the pathogenesis and prophylaxis of migraine. Headache 1977;17:101-8
22.
Lance JW, Anthony M, Gonski A. Serotonin, the carotid body and cranial vessels in migraine. Arch Neurol 1967;16:553-8
23.
Wolff HG. Headache and other head pain. New York: Oxford University Press, 1963
24.
Lance JW, Lambert GA, Goadsby PJ, Duckworth JW. Brainstem influences on the cephalic circulation: Experimental data from cat and monkey of relevance to the mechanism of migraine. Headache 1983;23:258-65
25.
Blau JN. Migraine pathogenesis: the neural hypothesis reexamined. J Neurol Neurosurg Psychiatry 1984;47:437-42
26.
Lauritzen M, Olesen J. Regional cerebral blood flow during migraine attacks by Xenon-133 inhalation and emission tomography, Brain 1984;107:447-61
Hisaka Igarashi, Fumihiko Sakai, Shinichi Kan, Jun Okada, Yoshiaki Tazaki
Cephalalgia Igarashi H, Sakai F, Kan S, Okada J, Tazaki Y. Magnestic resonance imaging of the brain in patients with migraine. Cephalalgia 1991;11:69-74. Oslo. ISSN 0333-1024 Magnetic resonance imaging (MRI) was studied in 91 patients with migraine and in 98 controls. Risk factors known to cause MRI lesions were carefully examined. In 36 patients with migraine (39.6%), small foci of high intensity on T2-weighted and proton-density-weighted images were seen in the white matter. Of patients with migraine who were less than 40 years old and without any risk factor, 29.4% showed lesions on MRI; this was significantly higher than the 11.2% for the group of age-matched controls (n = 98). The lesions were distributed predominantly in the centrum semiovale and frontal white matter in young patients, but extended to the deeper white matter at the level of basal ganglia in the older age group. The side of the MRI lesions did not always correspond to the side of usual aura or headache. Migraine-related variables such as type of migraine, frequency, duration or intensity of headache or consumption of ergotamine showed no significant correlation with the incidence of MRI abnormalities. Our data indicated that migraine may be associated with early pathologic changes in the brain. • Anti-cardiolipin antibody, magnetic resonance imaging, migraine, white matter lesion Hisaka Igarashi, Fumihiko Sakai, Jun Okada, Yoshiaki Tazaki, Department of Medicine; Shinichi Kan, Department of Radiology, School of Medicine, Kitasato University; Correspondence to Hisaka lgarashi, M.D., Department of Medicine, Kitasato University 1-15-1 Kitasato, Sagamihara, Kanagawa, Japan (228); Accepted 11 December 1990 Reduction of regional cerebral blood flow below the level of ischaemic threshold has been reported to be responsible for the aura of migraine (1-4). It is clinically important to know whether or not the repetition of migraine pathophysiology may cause permanent ischaemic changes in the brain. Some previous studies by CT scan have shown brain atrophy in patients with a-long history of frequent migraine attacks (5-7). Whilst the significance of this result has been questioned in other studies (8, 9), some recent studies using MRI have suggested parenchymal abnormalities in the brain (10, 11). We studied magnetic resonance imaging (MRI) of the brain in a large number of patients with migraine in order to investigate if migraine causes early parenchymal changes. Other risk factors known to cause MRI lesions were carefully examined and the data were compared with age-matched controls Subjects and methods
A total of 91 patients with migraine were studied. Diagnosis of migraine was made according to the criteria of the Headache Classification Committee of the International Headache Society (12). Fifty-three had migraine with aura and their mean age was 33.2 ± 14.6 (mean ± SD) years. Thirty-three patients had migraine without aura (36.7 ± 11.9 years) and 5 had migraine with prolonged aura (24.8 ± 5.7 years). Their aura symptoms with visual disturbances followed by hemiparaesthesia (n = 5) and dysphasia (n = 2) lasted 1 to 12 h. The mean duration of the history of migraine was 13.4 ± 11.6 years and there was no significant difference between different groups. The control group consisted of a random sample of 98 subjects from the consecutive studies during the study period of the patients with migraine. This group was investigated because of dizziness, tension-type headache, idiopathic focal spasm, psychological disorders or history of epilepsy. They were less than 40 years old (32.4 ± 18.2 years) and did not have any focal neurological signs. Risk factors which may possibly cause lesions on MRI were carefully examined clinically and in laboratory examination in all the patients with migraine. They included multiple sclerosis, diabetes mellitus, hypertension, collagen disease, valvular heart disease, hyperlipidemia and polycythemia. Serological tests for syphilis, LE cells, antinuclear antibody and anti-ENA (extractable nuclear anti-gens) were also carried out. IgG anti-cardiolipin antibody was measured in 41 patients with migraine by ELISA and was evaluated using Harris's standard cut-off point (13). MRI studies were done in 33 of these 41 patients. MRI studies were carried out with the RESONA 0.5 Tesla super-conductive MRI unit (Yokogawa Medical Systems) using 2000/120(TR/TE) T2-weighted images, 2000/60(TR/TE) proton density weighted images and 500/25 T1-weighted images. Slice thickness was 10 mm. Assessment of MRI was
Table Incidence of positive MRI studies in patients with migraine. Duration of Positive Type of migraine n (men:women) Age (years) history MRI Brain atrophy (years) studies Migraine with aura 53 (19:34) 33.2 ± 14.6 13.1 ± 12.7 23 (43.4%) 0 Migraine without aura 33 (7:26) 36.7 ± 11.9 14.3 ± 10.1 11 (33.3%) 1 Migraine with prolonged aura 5 (3:2) 24.8 ± 5.7 11.2 ± 6.4 2 (40.0%) 0 Total 91 (29:62) 34.0 ± 13.6 13.4 ± 11.6 36 (39.6%) 1 Positive MRI studies: focal areas of high intensity images in the white matter on both T2-weighted and proton-density-weighted. made by two observers, and discrepancies were resolved by a third judge. All patients had MRI studies done during the interictal period of migraine. None of the patients, including those with migraine with prolonged aura, had any neurological deficit at the time of the MRI study. The statistics used were the Student's t-test for parametric variables and the Mann-Whitney U test for non-parametric variables. Results
Table 1 summarizes the incidence of positive MRI studies in patients with migraine. The characteristics of the lesion were focal areas of high intensity on both T2-weighted and proton density weighted MR images distributed bilaterally in the white matter of the brain. Etat crible or Virchow-Robin's space was carefully excluded from the positive findings (14). Positive MRI studies were seen in 23 out of 53 patients (43.4%) with migraine with aura, in 11 out of 33 patients (33.3%) with migraine without aura, and in 2 out of 5 patients (40.0%) with migraine with prolonged aura. Altogether, 36 out of 91 patients (39.6%) with migraine showed small white matter lesions on MRI. Only one patient showed brain atrophy. The incidence of MRI lesions was rot significantly different among different types of migraine. Figs 1 and 2 demonstrate typical MRI lesions in our patients. Fig. 1 is a 42-year-old woman with migraine with aura. Since the age of 14 she has had severe migraine headaches preceded by visual aura. Her blood pressure was normal, and there was no risk factor for atherosclerotic or other diseases known to be responsible for MRI lesions. Multiple small high-intensity lesions were seen in bilateral coror a radiata and centrum semiovale. Fig. 2 is that of a 26-year-old woman with migraine with aura. She had a 7-year history of frequent attacks of migraine preceded by scintillating scotoma. CT scan, electroencephalogram and all the blood tests for risk factors were negative. MRI also showed small high intensity lesions in both centrum semiovale. Patients with positive MRI studies were compared with those with negative MRI studies for various physiological parameters (Table 2). The mean age of the MRI positive group was 39.4 ± 15.2 years and was significantly older than the MRI negative group (30.1 ± 11.0 years). There were no significant differences between the two groups as regards the duration of the history, the frequency of migraine attacks or the amount of consumption of ergotamine. Risk factors were found in 16 patients with migraine (for details see Table 2). The incidence of these risk factors was higher in the MRI positive group (22.2%) than in the MRI negative group (14.8%), but the difference was not statistically significant. The time elapsed between the last migraine attack and the MRI study was no different between the MRI positive group (38.1 ± 39.2 days) and the MRI negative group (33.7 ± 25.2 days). Anti-cardiolipin antibody was abnormally high in only 3 patients (Table 3); 2 of these were MRI positive, the remaining patient MRI negative. Those patients without any risk factor were subjected to further analysis. Fig. 3 illustrates the frequency of MRI lesions comparing the different age groups. The frequency of white matter lesions increased with advancing age. While MRI studies were positive in all the patients over 60 years of age, the incidence of a positive MRI study was high even in younger patients. Lesions on MRI were present in 29.4% of patients under 40 years. The distribution of MRI lesions in all patients was plotted in the different brain levels (Fig. 4). In patients younger than 40 years the distribution of the lesion was predominantly in the centrum semiovale and frontal white matter (Fig. 4A), but it extended to the deeper white matter in patients aged 40 or over (Fig. 4B). The side of lesions on MRI did not always correspond to the side of the usual aura or headache. The distribution of lesions was no different among different types of migraine. Table 4 shows the comparison between patients with migraine and controls. Both groups included only those subjects younger than 40 years who did
not have risk factors. The percentage of positive MRI studies in patients with migraine became 29.4% but was still significantly greater than the 11.2% in age-matched controls. Discussion
In the present study MRI revealed small areas of high intensity in the white matter of the brain in patients with migraine, suggesting early parenchymal lesions. The overall frequency of lesions detected by MRI studies was 39.6% and was similar to the previous report by Soges et al. (11) who studied 24 patients with migraine and found 46% of them showing abnormal MRI studies. Their slightly higher percentage for positive MRI studies may be explained by the slightly older mean age (36.8 years) than our patients (34.0 years). The importance of our study is that we studied a large number of patients with migraine and compared these with similarly studied age-matched controls. A group of patients with migraine under 40 years old without risk factors showed a significantly higher incidence of positive MRI changes (29.4%) than the controls (11.2%). In order to discover whether lesions on MRI are the result of the migraine pathophysiology or of other variables, we screened all other possible risk factors by clinical and laboratory examination. The comparison of the two groups with either positive MRI or negative MRI revealed that ageing was the only variable significantly different. Risk factors such as hypertension, hyperlipidemia or others were not significantly different between the groups. Anti-cardiolipin antibody, which has been reported to be one of the important risk factors for cerebrovascular diseases in young subjects (15, 16), also showed no correlation with the occurrence of MRI lesions. The characteristics of MRI lesions in migraine were well defined small foci on T2-weighted and proton density weighted images in the white matter. These white matter lesions resemble so-called UBOs, unidentified bright objects (17) which are often related to ageing, hypertension and other arteriosclerotic risk factors. However, the pathogenesis of the lesions on MRI in migraine seemed to be different from the lesions caused by arteriosclerotic or atheromatous mechanisms, because our patients consisted of a young population and did not
Table 2 Comparison between MRI positive group and negative group. Duration Frequency Total Risk n Age (years) of of ergotamine factors history (years) attacks (month)consumption (tablets) Positive group 36 39.4 ± 15.2* 15.7 ± 13.7 2.7 ± 5.2 54.7 ± 169.2 8 (22.2%) Negative group 54 30.1 ± 11.0 11.6 ± 9.5 3.2 ± 3.5 62.0 ± 159.9 8 (14.8%) MRI positive group: Patients with small high intensity lesions. A patient with brain atrophy was not included. Risk Factors: hypertension (6), TIA (2), diabetes mellitus (1), mitral valve insufficiency (1), hyperlipidemia (3), thrombocytopenia (1), positive LE (1), positive ANA (3), positive anti-cardiolipin antibody (3). * p < 0.01 (positive group vs. negative group: Student's t-test). have risk factors such as hypertension or diabetes mellitus. None of the patients showed any sign of dementia. In addition, lesions on MRI in younger patients with migraine were seen less frequently in the basal ganglia, where lacunar infarctions are usually seen. The most tempting speculation to explain the parenchymatous lesions in migraine is that repeated attacks of hypoperfusion during aura cause permanent ischaemic changes. This assumption was proved to be an unlikely mechanism because lesions were no more frequent in migraine with aura than in migraine without aura. Neither did the frequency of the lesion correlate with such variables as the
Table 3. Anti-cardiolipin antibody and MRI in patients with migraine. MRI lesion Anti-cardiolipin n Positive Negative antibody Positive 3 2 1 Negative 30 11 19 Total 33 13 20
Anti-cardiolipin antibody: ELISA (Workshop 1986). Cut-off point of 5GCL unit.
Table 4. Incidence of MRI lesions: migraine vs. controls; younger than 40 years and without risk factors. No. of Percent of n positive positive MRI studies MRI studies Migraineurs 51 15 29.4%* Controls 98 11 11.2%
* p < 0.05 (Student's t-test). frequency of attacks, the duration of the disease, or the type of migraine. Lesions in the brain were not localized to the occipital region but distributed diffusely in the centrum semiovale and corona radiata, If ischaemia has any relation to the present finding, it has to be of global hypoperfusion and not of regional hypoperfusion. One possible mechanism is the reduction of cerebral perfusion pressure due to hypotension during orthostatic posture or other causes. Not infrequently do patients with migraine experience orthostatic hypotension and may suffer from fainting sensation or even syncopal episodes. This hypothesis, however, remains to be investigated. Another possible mechanism which may cause patchy white matter lesions is the production of multiple microemboli. During attacks of migraine increased aggregability of platelets has been shown by several investigators (18-21). Abnormal platelet function has also been suggested. It is possible that increased platelet aggregation may trigger the formation of multiple microemboli. In addition, vasoactive substances such as 5-HT are released from the platelets and may cause secondary vasoconstriction (22) which may produce cerebral parenchymatous lesions. Currently, one of the major hypotheses is that migraine is a vascular headache (23), While it should be reasonable to relate our MRI findings to the vascular events, there is the remaining possibility that they are the result of primary neuronal damage related to the migraine pathophysiology. There are investigators who suggest the neural mechanism of migraine (24-26). Although the present study did not provide data to suggest certain aetiology for migraine attacks, it suggested that there was a significant risk in migraine for the early occurrence of parenchymatous lesions on MRI. Further studies are necessary to clarify the significance of changes in the MRI in the pathophysiology and treatment of migraine. References
1.
Skinhøj E. Hemodynamic studies within the brain during migraine. Arch Neurol 1973;29:95-8
2.
Norris JW, Hachinski VC, Cooper PW. Changes in cerebral blood flow during a migraine attack. Br Med J 1975;3:676-7
3.
Sakai F, Meyer JS. Regional cerebral hemodynamics during migraine and cluster headaches measured by the 133Xe inhalation method. Headache 1978;18:122-32
4.
Clesen J, Larsen B, Lauritzen M. Focal hyperemia followed b,, spreading oligemia and impaired activation of rCBF in classic migraine. Ann Neurol 1981;9:344-52
5
Hangerford GD, du Boulay GH, Zilkha KJ. Computerised axial tomography in patients with severe migraine: a preliminary report. J Neurol Neurosurg Psychiatry 1976;39:990- 4
6.
Mathew NT, Meyer JS, Welch KMA, Neblett CR. Abnormal CT-scans in migraine. Headache 1977;16:272-9
7.
du Boulay GH, Ruiz JS, Rose FC, Stevens JM, Zilkha KJ. CT changes associated with migraine. AJNR 1983;4:472-3
8.
Masland WS, Friedman AP, Buchsbaum HW. Computerized axial tomography of migraine. Res Clin Stud Headache 1973;6:136-40
9.
Cuetter AC, Aita JF. CT scanning in classic migraine. Headache 1983;23:195
10.
Kaplan RD, Solomon GD, Diamond S, Freitag FG. The role of MRI in the evaluation of a migraine population: preliminary data. Headache 1987;27:315-18
11.
Soges LJ, Cacayorin ED, Petro GR, Ramachandran TS. Migraine: Evaluation by MR. AJNR 1988;9:425-9
12.
Headache Classification Committee of the international Headache Society: Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988;(suppl 7):1-96
I3.
Harris EN, Gharavi AE, Patel SP, Hughes GRV. Evaluation of the anti-cardiolipin antibody test: report of an international workshop held 4 April 1986. Clin Exp Immunol 1987;68:215- 22
14.
Jungreis CA, Kanal E, Hirsch WL, Martinez AJ, Moossy J. Normal perivascular spaces mimicking lacunar infarction: MR imaging. Radiology 1988;169:101-4
15.
Syrjänen J, Vaarala O, Iivanainen M, Palosuo T, Valtonen W, Aho K. Anticardiolipin response and its association with infections in young and middle-aged patients with cerebral infarction. Acta Neurol Scand 1988;78:381-6
16.
Levine SR, Deegan MJ, Futrell N, Welch KMA. Cerebrovascular and neurologic disease associated with antiphospholipid antibodies: 48 cases. Neurology 1990;40:1181-9
17.
Bradley Jr WG. Investigators solve mystery of unidentified bright spots. Diagn Imaging 1988;57:322-6, 446
18.
Hanington E, Jones RJ, Amess JAL, Wachowicz B. Migraine: a platelet disorder. Lancet 1981;2:720-3
19.
Kalendovsky Z, Austin JH. "Complicated migraine". Its association with increased platelet aggregability and abnormal plasma coagulation factors. Headache 1978;15:18-35
20.
Couch JR, Hassanein RS. Platelet aggregability in migraine. Neurology 1977;27:843-8
21.
Deshmukh SV, Meyer JS. Cyclic changes in platelet dynamics and the pathogenesis and prophylaxis of migraine. Headache 1977;17:101-8
22.
Lance JW, Anthony M, Gonski A. Serotonin, the carotid body and cranial vessels in migraine. Arch Neurol 1967;16:553-8
23.
Wolff HG. Headache and other head pain. New York: Oxford University Press, 1963
24.
Lance JW, Lambert GA, Goadsby PJ, Duckworth JW. Brainstem influences on the cephalic circulation: Experimental data from cat and monkey of relevance to the mechanism of migraine. Headache 1983;23:258-65
25.
Blau JN. Migraine pathogenesis: the neural hypothesis reexamined. J Neurol Neurosurg Psychiatry 1984;47:437-42
26.
Lauritzen M, Olesen J. Regional cerebral blood flow during migraine attacks by Xenon-133 inhalation and emission tomography, Brain 1984;107:447-61
