Transcranial Doppler in Spontaneous Attacks of Migraine

Transcription

1 680 Transcranial Doppler in Spontaneous s of Migraine Elietta M. Zanette, MD; Alessandro Agnoli, MD; Cinzia Robert, MD; Flavia Chiarotti, MD; Rosanna Cerbo, MD; and Cesare Fieschi, MD Background and Purpose: Our aim in this study was to compare headache-free and spontaneous migraine measurements of blood flow velocity and the pulsatility index in the anterior cerebral artery, middle cerebral artery, and posterior cerebral artery. Methods: Thirty-one patients (nine having experienced migraine with aura and 22 migraine without aura) were studied in headache-free periods and during spontaneous migraine attacks with transcranial Doppler ultrasonography. Results: During attacks of migraine with aura, blood flow velocities (particularly the diastolic velocity [p=0.0]) were reduced while the pulsatility index increased (p=0.0), whereas a generalized increase in diastolic velocity (p<0.02) and a decrease in the pulsatility index (/?=0.0) were observed during attacks of migraine without aura. Significant variations of blood pressure and heart rate were never found. Conclusions: These findings are consistent with constriction of resistance vessels in migraine with aura and dilatation of the vessels in migraine without aura. This disparity could be due to a difference between the two migraine types or could be related to the fact that in this study the time interval between headache onset and transcranial Doppler was shorter in the migraine-with-aura group. The latter explanation would apply if, in fact, both types of migraine evolve from hypoperfusion to hyperperfusion during their time course, although perhaps with a difference in intensity. (Stroke 1992^:680-68) KEY WORDS cerebral blood flow migraine ultrasonics Regional cerebral blood flow (rcbf) studies in patients during migraine with aura have shown hypoperfusion outlasting the aura, usually up to 6 hours, 1-2 sometimes followed by hyperperfusion. 2-4 Cerebral blood flow changes have not been described in cases of migraine without aura, 1 - except for some instances of hyperperfusion. 6-8 Recently, transcranial Doppler ultrasonography (TCD), a noninvasive technique used to assess the dynamics of the intracranial circulation in various physiological and pathological conditions, 9 has been used to study migraine. The CO 2 reactivity of the cerebral arteries 10 and the blood flow velocity during the migraine-free phase 11 have been measured and the flow velocity values estimated during both attack and headache-free periods. 12 The present study compares blood flow velocities using TCD in the headache-free period and within the first 6 hours of the attack in cases of migraine with and without aura. Materials and Methods We observed thirty-one patients, nine males and 22 females with a mean±sd age of.2±7.4 years, recruited from the Outpatients Service for Headache of the Department of Neurological Sciences of the Uni- From the Department of Neurological Sciences, University "La Sapienza" (E.M.Z., /VA., C.R., F.C., R.C., C.F.), and Istituto Superiore di Sanita, Rome. Supported in part by CNR grant Address for correspondence: Prof. Elietta M. Zanette, Dipartimento di Scienze Neurologiche, V.le delltjniversita', 0, 0018 Rome, Italy. Received July 2, 1990; accepted January 29, versity of Rome "La Sapienza", between 1987 and Inclusion criteria were: 1) a diagnosis of migraine with aura or migraine without aura according to the Headache Classification of the International Headache Society 1 ; 2) a normal neurological examination, duplex scan of the extracranial arteries, and brain computed tomography; and ) the absence of other pathologies (in particular, cardiovascular diseases and hypertension). The clinical profile of the patients is shown in Table 1. All migraine prophylactic therapies were withdrawn for at least 1 month before the TCD studies. Furthermore, patients were not treated for the acute attack studied by TCD. A control group of 20 healthy volunteers (10 males and 10 females, mean±sd age.7±.4 years) with no personal or familial history of headache were recruited from among the hospital employees in These control subjects underwent TCD twice, with an interval of 21 days between the two exams. Transcranial Doppler examination was performed in the headache-free period (at least % hours after the last migraine attack) and within 6 hours of the onset of the migraine attack with a TC2-64 B EME (Uberlingen, FRG) with a 2-MHz probe and a spectral analyzer. Patients were studied supine following at least 10 minutes' rest with eyes closed. The anterior cerebral artery (ACA), middle cerebral artery (MCA), and posterior cerebral artery (PCA) were insonated through the temporal window, and systolic velocity, diastolic velocity, and mean velocity (time-mean velocity) were measured. The best signal was sought for each artery, and

2 Zanette et al TCD in Migraine 681 TABLE 1. Characteristic Patients and Headache Characteristics No. patients Male Female Age (yr) (mean±sd) Male Female frequency (No. patients) >l/week 2-4/month <2/month Disease duration (yr) (mean±sd) Pain location during TCD Bilateral* Unilateral Type of aura Visual Visual+sensitive Visual+motor Sensitive Time interval between attack onset and TCD (hr) Mean Range Migraine with aura ±7.8.6±4. 4.± ± Values are mean±sd. TCD, transcranial Doppler ultrasonography. 'With side preference. Migraine without aura ± ±9 6.4± ± Total ±7.4 4±7.6.8± ± the probe position was recorded at the first exam. Gosling's pulsatility index was calculated according to the following formula: (SV-DV)/MV, where SV, DV, and MV stand for systolic velocity, diastolic velocity and mean velocity, respectively. 14 The above parameters were studied on the headache (symptomatic) and the contralateral (asymptomatic) sides. Blood pressure and heart rate were also recorded at the beginning of each TCD exam. Values of systolic velocity, diastolic velocity, mean velocity, and pulsatility index were analyzed separately, using parametric factorial analysis of variance (ANOVA), with migraine (without aura, with aura) as between-subject factor at two levels and status (basal condition, migtaine attack), side (symptomatic, asymptomatic), and artery (ACA, MCA, PCA) as withinsubject factors at two, two, and three levels, respectively. 1 Moreover, separate ANOVAs were applied to the two migraine groups, including only the within-subject factors in the linear model. Normality of parameter distributions and homoscedasticity were previously verified, and Greenhouse-Geisser correction was applied to control for the sphericity assumption violation, when present. When a significance (p^0.0) was found with ANOVA the data were analyzed by means of the Student's t test for paired samples, with Bonferroni's correction, comparing levels of parameters during basal condition and migraine attack, within each side x artery combination 16 ; values of /?<0.0 were considered as significant. Data from control subjects were analyzed using ANOVA, with status (first exam, second exam), side (right, left), and artery (ACA, MCA, PCA) as withinsubject factors at two, two, and three levels, respectively. These data were also used to assess reproducibility with the intraclass correlation coefficient. 1 The control group was compared with the migraine groups. No direct comparison could be performed because of the differences in the meaning of status and side factors in the different groups. In healthy subjects, the four values (right/left side x first/second exam) were not statistically different. Therefore, one of the four values within each subject was randomly selected, and this data set was separately compared with asymptomatic and symptomatic basal values in migraine patients by applying the ANOVA with one between-subject factor at three levels (control, migraine without aura, migraine with aura). Results No significant effect of status (first, second exam), or interaction between status, side, and artery was found in the 20 control subjects. Table 2 shows the means and standard deviations of TCD parameters in controls. The migraine patients showed systolic velocity, diastolic velocity, and mean velocity values within the normal range during the headache-free and the migraine periods. There were no significant differences in heart rate (74±9-74±8 beats per minute) and blood pressure (118±9/80±7-119±9/78± mm Hg) between the headache-free and migraine periods. When comparing the control group with the migraine groups during the migraine-free period, no significant differences in basal values were found among the three groups. In the migraine-with-aura patients, TCD was performed at a mean of 2.2 (range 1-) hours from onset of aura. The mean time between headache onset and TCD was 1.7 (range 0.-4.) hours in migraine-with-aura patients and.4 (range 2-6) hours in migraine-withoutaura patients, with a time-interval significantly shorter in the migraine-with-aura group (p<0.01, Student's t test). The effect of time on the differences between

3 682 Stroke Vol 2, No May 1992 TABLE 2. Mean Values and Standard Deviations of Transcranial Doppler Ultrasonograpby Parameters in Control Subjects Artery/side/examination MCA Right Left ACA Right Left PCA Right Left Systolic 104±14 106± ±9 92±1 90±ll ±1 72±17 70±ll 76±14 78±1 Velocities Diastolic 0±7 1±7 1± 1±6 42±7 42±7 41±8 42±7 4±9 4±6 7±7 _. 7±6 (cm/sec) Mean 68±10 69±9 69± ±10 8±9 +10 ±10 4±11 4±8 49±10 49±8 Pulsatility index 0.81 ± ± ± ± ± ± ± ± ± ± ± ±0.17 MCA, middle cerebral artery, ACA, anterior cerebral artery, PCA, posterior cerebral artery; exam 1, first Doppler study, exam 2, second study. headache-free and migraine period values was never significant. In fact, none of the correlation coefficients, computed between time from attack to TCD and the difference between the TCD measurements at the attack and the intercritical phase, was significantly different from zero. These coefficients ranged from to 0.29 (median 0.067) in the migraine-without-aura group and from to (median 0.279) in the migraine-with-aura group. When comparing the two migraine groups, ANOVA yielded a significant interaction status x migraine with regard to diastolic velocity (p=0.002), mean velocity (p=0.0018), and pulsatility index (p=0.006), and an almost significant interaction status x migraine with regard to systolic velocity (p=0.068). This was due to the difference in parameter variations between the migraine and migraine-free periods in the two migraine groups (generalized increase in the velocities and decrease in pulsatility index during the attack in the migraine-without-aura group and generalized decrease in the velocities and increase in pulsatility index during the attack in the migraine-withaura group). Results of the ANOVAs separately performed on the four measures in the two migraine groups are reported below. Migraine With Aura None of the patients underwent examination during an aura. Significant differences were observed in migraine with aura between the headache-free and the headache phase in all the arteries on both sides (diastolic velocity p=0.0 and pulsatility index /?=0.0 by ANOVA), with a more significant reduction in the MCA diastolic velocity on the symptomatic side (p<0.01) by Student's t test (Figure 1). An almostsignificant interaction statusxside (p=0.07) could be attributed to the more marked tendency of diastolic velocity to decrease in all the arteries of the symptomatic side. The ANOVA showed a significant interaction status x side x artery in systolic velocity and mean velocity (p<0.02 and /?<0.04, respectively) also, and Student's t test showed the most significant reduction in systolic velocity (p=0.0) and mean velocity (p=0.0) of the MCA on the symptomatic side. Table reports mean values and standard deviations of TCD parameters in the migraine-with-aura group. Migraine Without Aura The data recorded during attacks of migraine without aura compared with the headache-free period showed a generalized increase in diastolic velocity (p<0.02) and a generalized decrease in pulsatility index (p=0.0) for each artery on both sides, as well as a tendency to increase in mean velocity (p=0.06 by ANOVA). Student's t test for paired samples revealed that the most significant increases in diastolic velocity (p<0.0) and decreases in pulsatility index (p<0.01) occurred in the MCA of the symptomatic side (Figures 1 and 2). Table 4 reports mean values and standard deviations of TCD parameters in the migraine without aura group. Discussion Control subjects failed to show any significant modification for status, side, and artery between their two TCD exams (ANOVA). Differences in TCD parameters were found, however, in both the migraine-with-aura and migraine-without-aura groups, in the absence of variations in blood pressure and heart rate. The above changes were observed in diastolic velocity and pulsatility index of all arteries on both sides but in opposite directions. In fact, in the migraine-with-aura group there was a decrease in diastolic velocity (/?=0.0) and an increase in pulsatility index (p=0.0), whereas in the

4 Zanette et al TCD in Migraine 68 ULmm. CONTROLS MGRAME WITHOUT AURA I0GRAME WITH AJRA Right L»ft AS. SS. CD boaol p < 0.0 CD basol p < 0.01 attack M attack FIGURE 1. Bar graph showing middle cerebral artery diastolic velocity values in patients with migraine without aura and migraine with aura from the asymptomatic (A.S.) and symptomatic (S.S.) sides and in control subjects from the right and the left sides. Lines on top of the bars represent 1 SD. Transcranial Doppler ultrasonography (TCD) was performed in migraine patients during attack-free periods (basal) and during migraine attacks (attack). In control subjects, two TCD exams (1 and 2) were performed with an interval of 21 days between them. migraine-without-aura group there was an increase in diastolic velocity (/?<0.02) and a decrease in pulsatility index (p=0.0). This divergent trend is seen in the interaction statusxmigraine: a statistically significant interaction was found for diastolic velocity (p=0.qo2), mean velocity (p = 0.006), and pulsatility index (/?=0.018). The fact that both diastolic velocity and pulsatility index are sensitive to changes in the resistance vessels deserves consideration. A decrease in diastolic velocity and an increase in pulsatility index are observed with resistance-artery constriction, whereas an increase in diastolic velocity and a decrease in pulsatility index are found with dilatation of the same arteries Our results concerning the differences between headache-free and migraine periods (diastolic velocity decrease/pulsatility index increase in migraine with aura and diastolic velocity increase/pulsatility index decrease in migraine without aura) led us to consider the possibility that in the first 6 hours of a migraine attack, the resistance vessels constrict in migraine with aura and dilate in migraine without aura. The phenomenon is most pronounced in the MCA of the symptomatic side in both migraine types. This observation should not be interpreted in the same way for artery TABLE. Mean Values and Standard Deviations of Transcranial Doppler Ultrasonography Parameters in Migraine-Wlth-Aura Patients Velocities (cm/sec) Artery/side/period Systolic Diastolic Mean Pulsatility index MCA ACA PCA 111±11 9±1 10±17 106±16 89±14 8±1 91 ±9 86±1 74±11 69± 68±28 72±12 ± 42±8' 2±8 48± 4 ±8 9±6 4±4 40±10 6±7 2±7 7±10 ±7 72± ± ± ± ± ± ± ± ± ± ± ± ± ± ±0.20 MCA, middle cerebral artery; ACA, anterior cerebral artery, PCA, posterior cerebral artery;, attack-free period;, migraine attack. V<0.01 by Student's t test.

5 684 Stroke Vol 2, No May 1992 CONTROLS UQRAME WITHOUT AURA IKRAME WITH AURA r r T asoo I~M ih ojoo A-\\_ p-i i J-M r~h H H H H 0.600' H H H H 0.400' H H H H H H H H H H H 1 H H I H M 1 H Right Laft AS. &S. AS. &S. CD 1 CD boaot»»p<0.01 C boaat M2 Mattock M attack FIGURE 2. Middle cerebral artery pulsatility index values in patients with migraine without aura and migraine with aura from the asymptomatic (A.S.) and symptomatic (S.S.) sides and in control subjects from the right and left sides. Lines on top of the bars represent 1 SD. Transcranial Doppler ultrasonography was performed as described in Figure 1 legend. (MCA) and side. The predominance of modifications in the MCA can be attributed to the fact that the detection of this artery is easier and reproducibility (r=0.79 for mean velocity) is better than in ACA (r=0.67) and PCA (r=0.4), as we found in the control population. On the contrary, the predominance of changes in the symptomatic side seems to be more significant because it derives from the comparison of the two homologous territories; moreover, in migraine with aura the interaction status x side showed a more marked, though not significant, modification of TCD parameters on the symptomatic side. There have been very few previous reports of TCD in migraine patients, and only some of these describe the comparison between the headache-free and migraine phases. Thie et al 19 describe a prolonged, diffuse vasospasm of the basal cerebral arteries in three migraine patients (one without and two with neurological focal symptoms). However, the differences in type of patient and timing of examinations are too great between our study and his for useful comparison. In a subsequent study, Thie et al examined five migraine-with-aura and 1 migraine-without-aura patients in headache-free periods and during migraine attacks. 12 They report a divergent trend in flow velocities between the two headache groups during the attack when compared with the headache-free period (mean velocity increase and pulsatility index decrease in migraine with aura and mean velocity decrease and pulsatility index increase in migraine without aura). He attributes these results to caliber modifications TABLE 4. Mean Values and Standard Deviations of Transcranial Doppler Ultrasonography Parameters in Migraine-Without-Aura Patients Artery/side/period MCA ACA PCA Systolic 97±16 97±17 97±17 101±17 81±12 84±1 69±10 77±1 68±1 68±10 66±1 69±1 Diastolic 4±6 0±10* 47±8 1±11 9±4 41±7 ±4 8±8 ± 6±7 +7 ±8 Velocities (cm/sec) Mean 64±10 68±1 6±12 71±1 4±7 7±10 46±7 2±11 4±8 46±8 44±8 47±11 Pulsatility index 0.79± ±0.1Ot 0.78± ± ± ± ± ± ± ± ± ±0.11 MCA, middle cerebral artery; ACA, anterior cerebral artery; PCA, posterior cerebral artery;, attack-free period;, migraine attack. p<0.0, t/><0.01 by Student's t test.

6 of the large arteries. These data, which support our findings of different flow velocity modifications in migraine with and without aura, disagree with our results for the type of changes. In our opinion, this discordance may be due to the fact that patients were studied up to 24 hours after the onset of the attack as opposed to our study interval of 1-6 hours, and that Thie et al 12 considered only the mean velocity (which is less sensitive to dilatation and constriction of the resistance vessels than is diastolic velocity) and did not consider symptomatic and asymptomatic side separately. Our data seem more concordant with CBF studies, which are also more numerous in the literature. In fact, CBF studies describe a hypoperfusion in the early phase of migraine with aura episodes outlasting the aura period, 12 which is consistent with our observations. Interestingly, hypoperfusion in the symptomatic hemisphere, as we described, has been observed. 20 Hyperperfusion has also been observed in this migraine type when the study was performed in the late phase of the attack (after 6 hours). 2 However, this observation may not be applicable to our patients, who were studied earlier in their migraine. Most of the CBF studies have revealed no changes between the headache-free and headache phases in migraine-without-aura patients. 1>s However, hyperperfusion has been reported within 48 hours of the migraine attack 6 and, subsequently, also within 10 hours of the onset of the attack. 78 These observations seem to agree with our data, although the study interval is longer in these CBF studies. In conclusion, diastolic velocity and pulsatility index changes, reflecting variations in the resistance arteries, suggest that conditions during migraine-with-aura attacks are different from those during migraine-withoutaura attacks. This could be attributed to an intrinsic difference between the two headaches or may be part of the same phenomenon (i.e., vasoconstriction followed by vasodilatation), which may differ in timing (we emphasize that the time interval between headache onset and TCD exam is shorter in the migraine-withaura group) and intensity. Acknowledgments We gratefully acknowledge Prof. G.M. Von Reutern for his scientific help in the present study, and Prof. David G. Sherman for his contribution in the revision of the manuscript. Zanette et al TCD in Migraine 68 References 1. Lauritzen M, Olesen J: Regional cerebral blood flow during migraine attacks by Xenon-1 inhalation and emission tomography. Brain 1984;107: Andersen AR, Friberg L, Olsen TS, Olesen J: Delayed hyperemia following hypoperfusion in classic migraine: Single photon emission computed tomographic demonstration. Arch Neurol 1988;4: Norris JW, Hachinski VC, Cooper PW: Changes in cerebral blood flow during a migraine attack. Br Med J 197;: Olesen J, Larsen B, Lauritzen M: Focal hyperemia followed by spreading oligemia and impaired activation of rcbf in classic migraine. Ann Neurol 1981a;9:44-2. Olesen J: The ischemic hypotheses of migraine. Arch Neurol 1987;44: Sakai F, Meyer JS: Regional cerebral hemodynamics during migraine and cluster headache measured by the Xenon-1 inhalation method. Headache 1978;18: Friberg L, Olesen J, Iversen HK: Regional cerebral blood flow during attacks and when free of symptoms in a large group of migraine patients. Cephalalgia 1989;9(suppl 10): Kobari M, Meyer JS, Ichijo M, Kawamura J: Cortical and subcortical hyperperfusion during migraine and cluster headache measured by Xe CT-CBF. Neuroradiology 1990;2:4-ll 9. Aaslid R, Markwalder T, Nornes H: Noninvasive transcranial Doppler ultrasound recording of flow velocity in basal cerebral arteries. J Neurosurg 1982;7: Thomas TD, Harpold GJ, Troost BT: Cerebrovascular reactivity in migraineurs as measured by transcranial Doppler. Cephalalgia 1990;10: Thie A, Fuhlendorf A, Spitzer K, Kunze K: Transcranial Doppler evaluation of common and classic migraine: Part I. Ultrasonic features during the headache-free period. Headache 1990;0: Thie A, Fuhlendorf A, Spitzer K, Kunze K: Transcranial Doppler evaluation of common and classic migraine: Part II. Ultrasonic features during attacks. Headache 1990;0: Headache Classification Committee of the International Headache Society: Classification and diagnostic criteria for headache disorders, cranial neuralgias and facial pain. Cephalalgia 1988;8(suppl 7): Gosling RG, King DH: Arterial assessment by Doppler-shift ultrasound. Proc R Soc Med 1974;67: Winer BJ: Statistical Principles in Experimental Design, ed 2. New York, McGraw-Hill Book Co, Wilcox RR: New Statistical Procedures for the Social Sciences: Modern Solutions to Basic Problems. Hillsdale, NJ, Lawrence Erlbaum Associates, Harders A: Neurosurgical Applications of Transcranial Doppler Sonography. Wien, Austria, Springer-Verlag, Hassler W, Chioffi F: CO 2 reactivity of cerebral vasospasm after aneurysmal subarachnoid haemorrage. Ada Neurochir 1989;98: Thie A, Spitzer K, Lachenmayer L, Kunze K: Prolonged vasospasm in migraine detected by noninvasive transcranial Doppler ultrasound. Headache 1988;28: Olesen J, Friberg L, Olsen TS, Iversen HK, Lassen NA, Andersen AR, Karle A: Timing and topography of cerebral blood flow, aura, and headache during migraine attacks. Ann Neurol 1990;28: