PATTERN-INDUCED VISUAL STRESS AND CORTICAL ACTIVATION IN MIGRAINE PATIENTS

Jie Huang¹, Thomas G. Cooper¹, Banu Satana²,
David I. Kaufman², Yue Cao¹

Departments of Radiology¹ and Neurology²
Michigan State University,
East Lansing, Michigan, USA

INTRODUCTION

Migraine affects nearly 12% of the population in the United States. People suffering from migraine (migraineurs) are highly susceptible to visual discomfort and distortion [1-3]. During migraine attacks, photophobia is one of the most common complaints. Between migraine attacks, migraineurs are also sensitive to bright light and certain visual patterns. For example, migraineurs frequently report illusions of color, shape, and motion while viewing a black-and-white stripe pattern [1,2]. In this preliminary study, we investigate visual neuronal activity and its association with self-described visual distortion and discomfort in migraineurs.

METHODS AND MATERIALS

Subjects:  Four migraineurs with visual aura (MwA) and two without aura (4 female, 2 male, age 39±8 years) participated in the study (Table 1). Diagnosis was made according to the criteria of the International Headache Society [4]. Six non-headache, age-matched controls (5 female, 1 male, age 39±9 years) were also enrolled in the study. 

Visual distortion test:  Before the fMRI scan, all subjects had a 15 minute test of visual discomfort and distortion by using a black-and-white stripe pattern with spatial frequency (SF) varying from 0.25 to 19.2 cycles per degree (cpd). The subjects were instructed to describe freely their visual perceptions after viewing the pattern for 5 s. Their comments were recorded by an investigator.

fMRI:  During fMRI, the same pattern with five selected SFs (0.3, 1.2, 3.0, 6.0, and 9.0 cpd) were used to acquire functional images. Ten 15 s long fixation blocks interleaved with ten 15 s long stimulation blocks with each SF repeated twice. The stimulus was displayed with a mean luminance of 5.4 cd/m², a contrast of 73%, and a visual angle of 13°.  T2*-weighted images of 10 axial-oblique sections parallel to the Calcarine Fissure (CF) were acquired on a GE 1.5 T clinical scanner using a Gradient Echo Echo-Planar-Imaging pulse sequence (field-of-view 24 cm, TE/TR=50/1000 ms, flip angle 90°, matrix size 64x64, slice thickness 6 mm). This resulted in 300 images per anatomic section acquired. 

Data pre-processing:  Images were assessed and corrected for possible in-plane translation and rotation of the head [5]. The signal intensity time course was corrected for possible slow baseline drifts by fitting 0, 1^st, and 2^nd order polynomials to the time series of each voxel. 

Statistical analysis of activation:  Time courses of images were cross-correlated [6] with sine and cosine reference functions to obtain a pair of complex cross-correlation coefficients (ccc) voxel by voxel [7]. Both magnitude and phase of ccc were then calculated [7], and activation maps were thresholded with an estimated p < 0.001. The blood oxygenation level dependent (BOLD) responses to stimulation were averaged in six voxels with the largest ccc values at the center of activation clusters adjacent to CF in the left-hemisphere or in the right-hemisphere (Figure 1). The BOLD responses to each SF were also averaged over the two repeated blocks. In order to quantitatively compare the BOLD responses between migraineurs and non-headache controls, the mean MR signal intensity change for each SF was calculated by averaging five consecutive data points around the response peak.  

RESULTS AND DISCUSSION

Visual distortion:  Any self-reported illusions involving blurring lines, zigzag lines, waving or moving lines and patterns, flickering spots and lines, three-dimensional effects, and colors were accounted as visual distortion. Migraineurs and non-headache individuals are most sensitive to SFs ranging from 1.2 to 3.0 cpd, and less sensitive to lower or higher SFs. Thus, a 0-2 scale was adopted to score subject’s self-reported visual distortion. A score of zero represented no visual distortion reported, one denoted distortion reported only at the most sensitive SF range, and two denoted distortion reported both at the most sensitive SF range and the lower or higher SFs (<1.2 cpd or >3.0 cpd). Migraineurs with and without aura were significantly more susceptible to visual distortion than non-headache controls (Mann-Whitney rank U test, p<0.016), see Table 1.

Visual discomfort:  Visual discomfort included eye-ache, queasy feeling, headache, nausea, and dizziness. A 0-5 scale was used to score visual discomfort, according to the number of items of visual discomfort reported. No control subjects reported any visual discomfort, whereas five patients experienced discomfort (Table 1). The difference between the two groups was significant (Mann-Whitney rank U test, p<0.016). 

Activation in the putative V1:  Stimulus-correlated BOLD signal changes were observed in the V1 of all subjects. In the non-headache control (NHC) group, averaged BOLD signal intensity changes did not show any significant difference between the left and right putative V1 at all five SFs (Figures 2 and 4). However, asymmetric visual cortical activity was observed in the migraineurs (Table 1). The BOLD signal intensity changes in the migraine-affected side (MAS) of visual cortex of the migraineurs increased at SFs of 1.2, 3.0, and 6.0 cpd, compared with the opposite side (OS) (Figures 3 and 5, Table 1). The increased neuronal activity was significant at 1.2 cpd SF (Wilcoxon signed-rank test, p<0.028, n=6), indicating hyperactivity in the migraine-affected side of the occipital lobe of the migraineurs. Compared with controls, cortical activity in the MAS was significantly greater at SF of 1.2 cpd (Mann-Whitney U test, p<0.05), while activation in the OS did not statistically differ from the controls’ (Mann-Whitney U test, p>0.4) (Figures 3 and 5, Table 1).

Correlation between visual distortion and neuronal activity:  A significant correlation was found between visual distortion scores and BOLD signal intensity changes at SF of 1.2 cpd in all subjects (Spearman rank correlation, one-tail, rs=0.544, n=12, p=0.036). In this calculation, BOLD signal intensity changes in the non-headache controls were averaged over the two hemispheres and in the migraineurs were in the migraine-affected side of the occipital cortex.

Our preliminary study demonstrates a significant relationship between visual distortion and visual neuronal activity in migraine. We found that: (1) the migraineurs were highly susceptible to visual distortion and stress, compared with the non-headache controls (Table 1); and, (2) severe visual distortions induced by the linear stimuli in the migraineurs were associated with hyperactivity in the visual cortex. Furthermore, the visual hyperactivity in the migraineurs exhibits: (a) asymmetry between the primary migraine-affected side and the opposite side of the visual cortex; (b) spatial frequency dependency of the linear stimulus; and (c) significant correlation with sensitivity to visual distortion. These data support the notion that visual distortion has a neurological origin.

REFERENCES

1. A. Wilkins, et al., Brain, 107: 989 (1984).
2. D. A. Marcus, et al., Arch Neurol., 46, 1129 (1989). 
3. Y. Cao, et al., Arch Neurol., 56, 548 (1999).
4. Headache Classification Committee of the International Headache Society, Cephalalgia 8, 1 (1988).
5. Y. Cao, et al., J. Magn. Reson. Med., 3, 869 (1993).
6. P. A. Bandettini, et al., Magn. Reson. Med. 30, 161 (1993).
7. A. T. Lee, et al., Magn. Reson. Med. 33, 745 (1995).