Epilepsia, 48(2):211–219, 2007 Blackwell Publishing, Inc. C 2007 International League Against Epilepsy
Original Research
Cerebral Cortical Gyrification: A Preliminary Investigation in Temporal Lobe Epilepsy ∗ Lisa Ronan, †Kevin Murphy, †Norman Delanty, ‡Colin Doherty, §Sinead Maguire, ∗ Cathy Scanlon, and ∗ Mary Fitzsimons Departments of ∗ Neurophysics; and †Neurology, Beaumont Hospital; ‡Department of Neurology, St James’ Hospital; and §Department of Radiology, Beaumont Hospital, Dublin, Ireland
Summary: Purpose: To introduce a measure of global cortical folding in epilepsy by using stereology. Subtle developmental abnormalities associated with temporal lobe epilepsy may encompass brain morphologic changes such as an aberrant degree of cortical folding. Methods: Stereologic methods of volume and surface-area estimation were applied to in vivo MR brain-image data of a cohort of 20 temporal lobe epilepsy (TLE) patients (10 men, 10 women), and 20 neurologically normal controls (10 men, 10 women). Indices of cerebral gyrification and cerebral atrophy were generated. The impact of side of seizure onset, age at onset, history of febrile seizures, presence or absence of lesions, and presence or absence of secondarily generalized seizures on cerebral gyrification was assessed. Results: Although no significant group mean difference was
found in the degree of cerebral gyrification between patients and controls, five of 10 of male patients had an abnormal gyrification when compared with male controls. One female patient had a significant change in gyrification compared with female controls. In general, patients with TLE demonstrated a significant degree of global cerebral atrophy compared with controls. Clinical factors were not demonstrated to affect significantly any of the quantitative parameters. Conclusions: The results of this study suggest that an aberrant degree of global cerebral gyrification may occur in certain clinical groups of TLE patients. These findings have implications for general theories of developmental susceptibility in TLE. Key Words: Epilepsy—Gyrification—Stereology— Quantitative magnetic resonance imaging.
The human cerebral cortex is a highly folded structure composed of a series of functional elements organized into columnar units. The number of functional units in the cerebral cortex is more closely linked to its surface area than to its volume (Rockel et al., 1974, 1980). A uniformity of connectivity is found across the cortex. This means that each unit of cortical area has the same total crosssectional area of long-distance connection fibers as has every other. These fibers are the interconnections between cortical regions (Zhang and Sejnowski, 2000). In terms of evolution, a trend exists to increase the surface area of the cortex and hence the number of functional units, relative to brain size. This is achieved through extrinsic folding of the cortex. Developmentally, this process is called gyrification and begins in the fetal stages, increasing to a maximum
in childhood and then decreasing to a stable adult value (Magnotta et al., 1999). Although the development of cortical gyrification is poorly understood, it is thought to be a genetically influenced process (Tramo et al., 1995). Because the number of cortical neurons and their connections is far in excess of the number of genes, it implies some underlying general principle for cortical development (Sisodiya et al., 1995). Cortical gyrification is thought to be a mechanical process, controlled by tension factors determining cortico–cortical connections (Griffin, 1994). As such, it is a useful marker of cerebral connectivity. Any disruption, whether by genetic or environmental factors of the normal mechanisms responsible for the formation of the cerebral structure, may result in cortical developmental malformations (CDMs) (Armstrong et al., 1995). This may be manifested as an abnormal gyrification. The phenomenon of aberrant cortical gyrification has previously been investigated in schizophrenia (Kulynych et al., 1997; Vogeley et al., 2000, 2001; Sallet et al., 2003), Williams syndrome (Schmitt et al., 2002), dyslexia
Accepted June 28, 2006. Address correspondence and reprint requests to Dr. L. Ronan at Department of Neurophysics, Beaumont Hospital, Dublin 9, Ireland. E-mail:
[email protected] doi: 10.1111/j.1528-1167.2006.00928.x
211
212
L. RONAN ET AL. In this study, we introduce a measure of global cortical folding and, by extension, cortico–cortical connectivity in TLE patients with and without localized lesions. We applied stereologic methods of volume and surface area estimation to in vivo MR-image data for a cohort of 20 TLE patients and 20 controls with no known neurologic deficits. A resulting index of cerebral gyrification is determined, the isoperimetric ratio (IPR), which is defined as surface area corrected for volume (Eq. 1, Fig. 1a). These techniques have previously been validated by the authors (Ronan et al., 2006). This study was motivated by an interest in improving the yield of diagnostic MR imaging and furthering the understanding of the underlying causes of epilepsy through quantitative analysis of subtle brain morphologic features. The application of these measures may prove useful in clinical settings, in particular in patients in whom routine brain images appear unremarkable. METHODS Subjects and materials
FIG. 1. (A) A 3D representation of the brain reconstructed from a set of magnetic resonance images. The structure is bisected through the frontal lobes. The IPR quantifies the degree of folding of the surface of the brain and is derived from a ratio of cortical surface area to cerebral volume (Eq. 1 in text). In this cross-sectional image, the red outline indicates the boundary used to determine the cortical surface area and to specify the cerebral volume. (B) A 2D MRI coronal view of the brain. Both the intracranial boundary (in green), and the cerebral surface boundary (in red) are represented. The IPR is the ratio between the cerebral surface area and the cerebral volume (defined as grey and white matter interior to the cerebral surface boundary). The CV/ICV ratio used as a measure of cerebral atrophy is a ratio of the cerebral volume to the intracranial volume.
(Casanova et al., 2004), and autism (Hardan et al., 2004). Although no similar studies have been reported in epilepsy, some evidence suggests abnormal formation of the temporal lobe neocortex in temporal lobe epilepsy (TLE) (Hardiman et al., 1988). Furthermore, work in our laboratory has shown that the degree of lobar atrophy in TLE is often underestimated by qualitative analysis (Doherty et al., 2003). Additionally, many studies have demonstrated extratemporal changes in TLE (i.e., gross morphologic and atrophic changes not simply attributable to regional or lobar abnormalities) (Keller et al., 2002; Liu et al., 2005). It may be that developmental changes in global morphology of the brain are associated with TLE, and that this may be reflected in an aberrant degree of cortical folding. Epilepsia, Vol. 48, No. 2, 2007
Subjects Twenty control subjects with no neurologic deficit (10 male, 10 female patients) were chosen at random from a bank of controls in MR-image data library at Beaumont Hospital. The age range was 22 to 43 years (mean, 31.1 years) for women and 21 to 40 years (mean, 29.3) for men. A sample of 20 TLE patients (10 men, 10 women) who attended the epilepsy service at Beaumont Hospital, Dublin, was included in the study. The age range was 25 to 35 years (mean, 29.3 years) for women and 17 to 40 years (mean, 32.8) for men. Three-dimensional (3D) MR images were acquired for each patient as part of their routine epilepsy care. In each case, MR image data were reviewed by a radiologist, and this qualitative report used to aid diagnosis. Full clinical, imaging, and EEG data were used to categorize patients as follows: right TLE (n = 8), left TLE (n = 6), and nonlateralizing TLE (n = 6) (Table 1). Fourteen patients were classified as lesional (defined to include mesial temporal sclerosis as determined by qualitative MRI assessment). The remaining six patients were determined to be radiologically nonlesional. Informed consent was obtained from subjects involved. The medical research ethics committee at Beaumont Hospital Dublin approved this study. Imaging data Patient and control subjects were imaged with a 1.5T GE MRI system (General Electric, Milwaukee, WI, U.S.A.) with a standard head coil. A 3D spoiled gradient (SPGR) sequence was acquired from a sagittal localizer in the coronal plane. The following imaging parameters were used: 10.1, 4.2, 450 ms (TR/TE/TI); one excitation;
CEREBRAL CORTICAL GYRIFICATION TABLE 1.
PM1 PM2 PM3 PM4 PM5 PM6 PM7 PM8 PM9 PM10 PF1 PF2 PF3 PF4 PF5 PF6 PF7 PF8 PF9 PF10
213
Patient demographics
Sex
Diagnosis
Lesional
Febrile seizures
M M M M M M M M M M F F F F F F F F F F
R TLE Nl TLE R TLE L TLE R TLE L TLE R TLE L TLE R TLE R TLE R TLE L TLE L TLE Nl TLE Nl TLE Nl TLE Nl TLE R TLE L TLE Nl TLE
NL NL Othera NL R MTS L MTS R MTS NL R MTS Otherb R MTS L MTS L MTS L MTS NL NL Otherc R MTS L MTS Otherd
No Yes No No No Yes Yes No No No Yes Yes No No No No No Yes No Yes
Secondarily generalized convulsions
Patient age at date of scan
Age at seizure onset
Duration of epilepsy
No Yes Yes Yes Yes Yes Yes No Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes
32 17 23 36 39 40 31 35 40 35 29 32 31 28 25 31 26 26 30 35
21 8 4 5 5 12 2 30 29 25 7 17 6 17 7 9 19 14 6 3
11 9 19 31 34 28 29 5 11 10 22 15 25 11 18 22 7 12 24 32
M, male; F, female; R/L/Nl TLE, right, left and nonlateralizing TLE; NL, nonlesional; R/L MTS, right/left mesial temporal sclerosis. a Abnormal right posterior hippocampus. b Right anterior venous malformation. c Lesional cavernoma. d Right temporal focal polymicrogyria.
flip angle of 20◦ ; field of view, 24 × 24 cm; matrix, 256 × 256, resulting in 124 × 1.5- mm-thick image slices.
of febrile seizures, and history of secondarily generalized seizures.
Stereologic software Surface area stereologic estimates were carried out by a single rater by using specialized interactive software Measure developed at Johns Hopkins, Baltimore, Maryland, U.S.A. (Barta et al., 1995). Volume estimates were carried out by using EasyMeasure software (MARIARC Institute, University of Liverpool, Liverpool, England).
Methods
Data Measurements The IPR is calculated based on surface area (SA) corrected for cerebral volume (CV) (Eq. 1). The normal relation between cerebral volume and intracranial volume (ICV) is fixed (CV/ICV). This ratio can be used to determine whether cerebral atrophy exists. This ratio is also important in determining whether an abnormal index of gyrification (abnormal IPR) maybe a result of a CDM or a manifestation of cerebral atrophy. In this project, we measured cerebral surface area, cerebral volume, and intracranial volume, and generated a measure of cerebral gyrification (IPR), as well as a measure of cortical atrophy (CV/ICV) (Fig. 1b). Clinical factors From a review of patient charts, the following clinical data were collected: age, sex, age at onset of seizures, side of seizures onset, presence or absence of lesions, history
Image data MR image data were transferred to a work station and converted from DICOM into Analyse format by using the software MRIcro (http://www.psychology.nottingham. ac.uk/staff/crl/mricro.html). Stereologic estimates Stereology is a mathematically unbiased method for estimation of geometric properties of an object, based on two-dimensional slices of that object. It is ideally suited to applications of quantitative MRI. The stereologic methods used to estimate cerebral surface area and volume have previously been described and validated by the authors (Doherty et al., 2000, 2003; Ronan et al., 2006) (see Fig. 2). Experimental method To generate unbiased estimates, stereologically generated slices of the object under investigation are required to be isotropic and uniform-random. This is automatically implemented by using the Measure and EasyMeasure software. In this study, volume estimates were generated by using a sample-slice separation of 0.75 cm and an area per test point of 1.98 cm2 . Surface area estimates were generated by using a slice separation of 2.25 cm and an area per length for cycloid probes of 3.02 cm. Twenty-four Epilepsia, Vol. 48, No. 2, 2007
214
L. RONAN ET AL.
FIG. 2. (A) Volume estimates are generated by counting the number of points from a uniform test grid that fall within the volume of interest. The volume is then calculated as the sum of all points within the volume multiplied by the distance between each sample slice, and the area per point on the test grid. In this example, 179 points are marked on six consecutive sample slices through the brain, each separated by 3 cm. The area per test point is 1.9775 cm2 . The resulting volume estimate of this brain is 1,062 cc (179 × 3 cm × 1.9775 cm2 ). (See Ronan et al., 2006). (B) Surface area estimates are generated by counting the number of intersections (in yellow) generated when the test curves (in red) intersect with the surface of interest (pial surface of cortex). The surface area is then calculated as the sum of all intersections multiplied by the distance between each sample slice, the area per length of the test curves, and a factor of two. In this example, 119 intersections are marked on seven sample slices through the brain, each separated by 2.25 cm. The area per length of the test curves is 3.02 cm. The resulting surface area estimate of this brain is 1,617 cm2 (119 × 3.02 cm × 2.25 cm × 2). This process is repeated for several different orientations through the brain, and all estimates averaged to produce a final surface area estimate (see Ronan et al., 2006). Epilepsia, Vol. 48, No. 2, 2007
CEREBRAL CORTICAL GYRIFICATION orientations were used to generate an average surface area estimate for each subject. Rules were established for identification of stereologic intersections. For cerebral volume estimates, the total cerebral volume comprised all cerebral grey and white matter but excluded CSF of ventricles. Intracranial volume was defined as including all grey and white matter and CSF within the cranium. Surface-area estimates were based on the pial layer of the cortex. Isoperimetric ratio estimation The IPR for each subject was calculated by using Eq. 1, based on each subject’s volume and surface-area estimates. IPR =
S V 2/3
[1]
Data analysis Confidence intervals (CIs) at 95% were constructed for the difference between group means of surface area, cerebral volume, intracranial volume, IPR, and CV/ICV. Student’s t test p values were also reported for these. Oneway analysis of variance (ANOVA) was used to examine the effects of sex, side of seizure onset, history of febrile seizures, and history of secondarily generalized convulsions on quantitative parameters measured. To examine the effects of subject age, age at onset of seizures, and duration of epilepsy on measured parameters, analysis of covariance (ANCOVA) was used, with sex as a covariate. RESULTS For all controls (n = 20), we estimated an average cerebral surface area of 1,816 cm2 (range, 1,542 cm2 to 2,014 cm2 ; SD, 133), cerebral volume of 1,120 cc (range, 1,009 cc to 1,269 cc; SD, 70), and intracranial volume of 1,967 cc (range, 1,716 cc to 2,231 cc; SD, 124). The average value for IPR was 16.85 (range, 15.01–18.79; SD, 1.05), and for CV/ICV was 0.57 (range, 0.53–0.61; SD, 0.02). Table 2 shows cortical surface area, cerebral volume, intracranial volume, isoperimetric ratio, and cerebral atrophy measures for the individual male and female control subjects and TLE patients. Measurements for individual patients falling outside 2 SDs, as defined by the control population, are marked in red. Confidence intervals at 95% indicate that the patient group demonstrated a significantly reduced cerebral volume (CI, 24–129; p = 0.006) and significant cerebral atrophy (CI, 0.007–0.04; p = 0.008), as indicated by the ratio CV/ICV. No statistically significant differences were found between the patient and control groups in terms of surface area, intracranial volume, and IPR (Table 3).
215
With normal limits (±2 SDs) as defined by control results, five of 10 male patients and one of 10 female patients had an abnormal IPR measure. Three of 10 male patients had an abnormal degree of cerebral atrophy. In comparison, no female patients were found to have a significant degree of atrophy. Other individual results for absolute measures are highlighted in Table 2. When data were separated according to gender, no statistically significant male–female differences were found in the control group for any of the quantitative measures (Table 3). However, a small increase in gyrification was noted in male controls compared with female controls (CI, −0.52–1.48; p = 0.33; see Fig. 3). ANCOVA indicated no significant effects of subject age and sex on any of the parameters measured. Unlike the controls, the patient group demonstrated statistically significant male–female differences in measures of surface area (CI, 112–363; p = 0.001) and cerebral volume (CI, 5–164; p = 0.04), which is most likely attributable to the significantly smaller ICV for the women (CI, 80–346; p = 0.003). No difference was found in the magnitude of cerebral atrophy between male and female patients; however, male patients had a significantly greater IPR than their female counterparts (CI, 0.27–2.53; p = 0.02) (see Fig. 3). ANCOVA indicated no significant effects of patient age and sex on cerebral surface area, volume, or IPR, but a significant effect on cerebral atrophy (p = 0.049). In a comparison between patients and controls, male brains were comparable in terms of intracranial volume, surface area, and IPR, although a trend was observed toward an increase in IPR in the male patients (CI, −1.68– 0.25; p = 0.14). A significant level of cerebral atrophy was noted for male patients compared with male controls (CI, 0.01–0.05; p = 0.02). In similar comparisons between female patients and controls, female patients demonstrated a near-significant reduction in overall intracranial volume (CI, −7–252; p = 0.06), which is further echoed in a nearsignificant reduction in surface area (CI, −0.3–272; p = 0.051) and a significant reduction in cerebral volume (CI, 37–169; p = 0.004). However, unlike male patients, female patients did not reveal a significant level of atrophy when compared with controls. Effects of side of seizure onset One-way ANOVA revealed a significant influence of side of seizure onset on patient surface area and cerebral volume (p = 0.05, 0.04, respectively), where the category “side of seizure onset” had the categoric variables left TLE (LTLE) (n = 6), right TLE (RTLE) (n = 8), and nonlateralizing TLE (NTLE) (n = 6). No significant effects were seen on IPR or cerebral atrophy (p = 0.41, 0.65, respectively). When sex effects were factored in, the significance of side of seizure onset was removed (p = 0.9 for surface area, 0.42 for cerebral volume). Epilepsia, Vol. 48, No. 2, 2007
216
L. RONAN ET AL.
TABLE 2. Results of cerebral surface area, cerebral volume, intracranial volume, IPR, and CV/ICV for TLE patients (n = 20) and controls (n = 20)
Control
Mean SD Patient
Mean SD
Male SA
CV
ICV
IPR
CV/ICV
1,800 1,967 1,914 1,791 1,976 1,832 1,893 1,779 1,827 1,793 1,857 74 2,113 1,759 1,834 1,740 1,950 2,021 1,666 1,908 2,016 1,758 1,876 147
1,009 1,162 1,086 1,136 1,263 1,068 1,172 1,198 1,138 1,124 1,135 71 1,212 929 970 1,105 1,055 1,071 1,120 1,129 1,231 1,034 1,085 96
1,848 2,093 1,951 2,044 2,088 1,810 2,112 2,120 2,020 1,957 2,004 110 2,167 1,747 2,044 2,023 1,939 1,937 2,061 2,244 2,094 1,955 2,021 138
17.9 17.8 18.12 16.44 16.91 17.53 17.03 15.77 16.76 16.58 17.09 0.74 18.59 18.48 18.72 16.28 18.83 19.3 15.45 17.6 17.55 17.2 17.8 1.23
0.55 0.56 0.56 0.56 0.6 0.59 0.56 0.57 0.56 0.57 0.57 0.02 0.56 0.53 0.47 0.55 0.54 0.55 0.54 0.5 0.59 0.53 0.54 0.03
Female SA 1,560 1,954 1,630 1,822 1,859 1,921 1,680 2,014 1,768 1,542 1,775 167 1,593 1,776 1,587 1,628 1,677 1,405 1,661 1,715 1,802 1,545 1,639 116
CV
ICV
IPR
CV/ICV
1,053 1,061 1,067 1,142 1,058 1,126 1,077 1,269 1,151 1,041 1,104 70 1,001 1,053 1,043 1,036 1,052 902 951 1,114 910 952 1,001 70
1,716 1,954 1,822 1,899 1,894 1,946 1,952 2,231 1,916 1,968 1,930 131 1,725 1,923 1,914 1,985 1,801 1,694 1,571 2,013 1,718 1,734 1,807 144
15.07 18.79 15.62 16.68 17.92 17.75 15.98 17.18 16.1 15.01 16.61 1.28 15.93 17.16 15.43 15.9 16.22 15.05 17.18 15.96 19.2 15.97 16.4 1.18
0.61 0.54 0.59 0.6 0.56 0.58 0.55 0.57 0.6 0.53 0.57 0.03 0.58 0.55 0.55 0.52 0.58 0.53 0.61 0.55 0.53 0.55 0.55 0.03
SA, surface area; CV, cerebral volume; ICV, intracranial volume; IPR, isoperimetric ratio.
Effects of febrile seizures Seven of 13 patients had a history of febrile seizures (three men, four women). One-way ANOVA revealed no significant influence of a history of febrile seizures on patient surface area (p = 0.56), cerebral volume (p = 0.76), IPR (p = 0.64) and cerebral atrophy (p = 0.58). Additionally, no significant sex effects were demonstrated for any of the measures. Effects of history of secondarily generalized convulsions One-way ANOVA revealed no significant differences between patients with (n = 17) and those without (n = 3)
TABLE 3. Surface area Controls versus patients All (−42–159) p = 0.25 Men (−132–93) p = 0.72 Women (−0.3–272) p = 0.051 Men versus women Controls (−43–208) p = 0.18 Patients (122–363) p < 0.01
a history of secondarily generalized convulsions for any measure. Effects of lesions One-way ANOVA revealed no significant differences between patients with (n = 14) and those without (n = 6) lesions for any measure. Additionally, no significant sex effects were demonstrated for any of the measures. Effects of age, age at onset of seizures, and duration of epilepsy ANCOVA indicated no significant interactions between sex and age at seizure onset and sex and duration of epilepsy on any of the measures in the patient population.
Confidence intervals and associated p values for the difference between means Cerebral volume
Intracranial volume
Cerebral gyrification
Cerebral atrophy (CV/ICV)
(24–129) p < 0.01 (−30– 130) p = 0.2 (38–169) p < 0.01
(−45–150) p = 0.28 (−135–101) p = 0.77 (−7–252) p = 0.06
(−1.04– 0.5) p = 0.52 (−1.68–0.25) p = 0.14 (−0.95–1.37) p = 0.71
(0.007–0.04) p = 0.008 (0.005–0.05) p = 0.02 (−0.01–0.04) p = 0.15
(−35–97) p = 0.34 (5–164) p = 0.04
(−39–188) p = 0.19 (80–346) p < 0.01
(−0.52–1.48) p = 0.33 (0.27–2.53) p = 0.02
(−0.03–0.02) p = 0.55 (−0.05–0.01) p = 0.19
Results in bold indicate a statistically significant difference for α = 0.05. CV, cerebral volume; ICV, intracranial volume.
Epilepsia, Vol. 48, No. 2, 2007
CEREBRAL CORTICAL GYRIFICATION
FIG. 3. Box plots comparing male (M) and female (F) data demonstrating the nonsignificant trend toward increased gyrification in male controls compared with female controls (A). This trend becomes significant in the patient group (B).
DISCUSSION The aim of this study was to introduce a method of quantitation of cerebral gyrification into an epilepsy population and to inform the direction of future clinical studies. Studies of this type may help elucidate subtle developmental malformations in the disease. In this study, our results for normal controls in measures of total cerebral surface area, volume, and ICV agree well with previously reported studies (Tramo et al., 1995; Nopoulos et al., 2000; Moran et al., 2001; Barta and Dazzan, 2003; Liu et al., 2005). Additionally, no significant group mean differences were found between the genders for any of the quantitative measures in this group. In the patient group, our findings of global cerebral atrophy, as well as the finding of increased susceptibility for atrophy in men correspond with findings of other TLE investigations (Briellmann et al., 2000; Hermann et al., 2002; Theodore et al., 2003; Liu et al., 2005). Reports in the literature (Lee et al., 1998; Moran et al., 2001; Hermann et al., 2002; Keller et al., 2002; Theodore et al., 2003; Liu et al., 2005) are diverse and contradictory relating to the the significance of the impact of clinical features on brain morphologic changes in TLE. Our results reveal no significant effects of side of seizure onset, history of febrile seizures, history of secondarily generalized convulsions, presence or absence of lesions, age at onset of seizures, and duration of epilepsy on brain atrophy once age and sex are controlled for. We further found that these clinical factors had no impact on cerebral surface area, intracranial volume, or IPR. The demonstrated significant difference between male and female patients in terms of IPR may reflect the normal trend for an increased gyrification in male over female patients, confounded by an additional level of cerebral atrophy in male TLE patients (see Fig. 3). Although the difference in IPR between male and female controls was not found to be significant, the nonsignificant increase
217
for men raises the possibility of a sexual dimorphism in normal cerebral gyrification and underscores the need to control for sex in this type of study. In this study, we found no significant group difference in cerebral surface area or IPR between patients and controls, despite significant atrophy (CV/ICV) in the patient cohort. The surface-area contribution to IPR is based on cortical grey matter. Meanwhile, the cerebral volume estimates incorporate grey and white matter. A significant loss in white matter volume might not be expected to affect surface area estimates to the same extent as do cerebral volume estimates. Conversely, we would expect significant cortical grey matter atrophy to affect both surface area and volume estimates. Because of this, we hypothesize that a greater magnitude of atrophy occurs in cerebral white matter over cortical grey matter in TLE. Differences in atrophic changes of grey and white matter in TLE have previously been reported (Hermann et al., 2002). These authors postulated that a nonuniform cerebral atrophy affecting cerebral white matter more than cortical grey matter necessarily affects the degree of cortical connectivity by eroding the normal connective relation between cortical grey matter and cerebral white matter. The experimental design chosen for this study incorporated an examination of global cortical gyrification. The rationale for this was based on previous reported findings of global changes to brain morphology in TLE (Moran et al., 2001; Keller et al., 2002; Theodore et al., 2003; Liu et al., 2005). However, reported studies of cortical gyrification have limited the area of research to localized regions most likely to have aberrant cortical development (e.g., the frontal lobes in schizophrenia) (Vogeley et al., 2000; Jou et al., 2005). Localized measures of cortical folding (e.g., the segmented temporal lobe in TLE) will increase the power of future gyrification studies in TLE. As discussed in a previous study (Ronan et al., 2006), the variance of stereologic measures of global cerebral gyrification are less than those produced by other methods reported in the literature. Given the relatively small variance of our stereologic estimates, coupled with the lack of bias in the measurements produced, we conclude that the method of stereologic estimation is optimal in studies of cerebral gyrification. Although this study did not reveal significant group mean differences in whole-brain gyrification in TLE, five of 10 male patients and one of 10 female patients were demonstrated to have an abnormal index of cerebral gyrification. The study group used in this project demonstrated a wide range of clinical features within the umbrella diagnosis of TLE. Through correlation of clinical information with quantitative morphologic results of the present study, future studies of cerebral gyrification in TLE may reveal definitive results for more homogeneous clinical cohorts. Epilepsia, Vol. 48, No. 2, 2007
218
L. RONAN ET AL.
Quantitative analysis of MRI can improve its diagnostic yield and has the potential to elucidate the etiology of disease further. Measures of parameters such as IPR may be helpful in detecting subtle abnormalities of cortical structure in patients such as those with nonlesional focal epilepsy and idiopathic generalized epilepsy (e.g., juvenile myoclonic epilepsy), where MR imaging is unremarkable. Furthermore, these measures may be helpful in the understanding of refractory epilepsy. The quantitative MRI parameters described here may also be useful as endophenotypic markers in genetic association studies in TLE (Gottesman and Gould, 2003; Tan et al., 2004). To our knowledge, this is the first study of global cerebral gyrification in TLE. Our findings inform the direction for future research in this area. This study highlights the need to control absolute measurements of brain morphometry for intracranial volume and brain atrophy. Without these correction factors, results of cerebral surface area, volume, and IPR may be misinterpreted. Our future studies of cortical gyrification will focus on specific brain regions such as the temporal lobe in TLE and the frontal lobe in juvenile mycolonic epilepsy. These studies will include larger and more clinically homogeneous cohorts. Acknowledgment: We thank the Epilepsy Programme at Beaumont Hospital, Dublin, Ireland. This work was funded by a grant from the Irish Brain Research Foundation/Irish Institute of Clinical Neuroscience.
REFERENCES Armstrong E, Schleicher A, Omran H, Curtis M, Zilles K. (1995) The ontogeny of human gyrification. Cerebral Cortex 5:56– 63. Barta P, Petty R, McGilchrist I, Lewis RW, Jerram M, Casanova MF, Powers RE, Brill LB, Pearlson GD. (1995) Asymmetry of the planum temporale: methodological considerations and clinical associations. Psychiatry Research 61:137–150. Barta P, Dazzan P. (2003). Hemispheric surface area: sex, laterality and age effects. Cerebral Cortex 13:364–370. Briellmann RS, Berkovic SF, Jackson GD. (2000) Men may be more vulnerable to seizure-associated brain damage. Neurology 55:1479– 1485. Casanova MF, Araque J, Giedd J, Rumsey JM. (2004) Reduced brain size and gyrification in the brains of dyslexic patients. Journal of Child Neurology 19:275–281. Doherty CP, Fitzsimons M, Holohan T, Mohamed HB, Farrell M, Meredith GE, Staunton H. (2000) Accuracy and validity of stereology as a quantitative method for assessment of human temporal lobe volumes acquired by magnetic resonance imaging. Magnetic Resonance Imaging 18:1017–1025. Doherty CP, Fitzsimons M, Meredith G, Thornton J, McMackin D, Farrell M, Phillips J, Staunton H. (2003). Rapid stereological quantitation of temporal neocortex in TLE. Magnetic Resonance Imaging 21:511–518. Griffin LD. (1994) The intrinsic geometry of the cerebral cortex. Journal of Theoretical Biology 166:261–273. Gottesman II, Gould TD. (2003) The endophenotype concept in psychiatry: etymology and strategic intentions. American Journal of Psychiatry 160:636–645. Hardan AY, Jou RJ, Keshavan MS, Varma R, Minshew NJ. (2004) Increased frontal cortical folding in autism: a preliminary MRI study. Psychiatry Research 131:263–268.
Epilepsia, Vol. 48, No. 2, 2007
Hardiman O, Burke T, Phillips J, Murphy S, O’Moore B, Staunton H, Farrell M. (1988) Microdysgenesis in resected temporal neocortex. Neurology 38:1041–1047. Hermann B, Seidenberg M, Bell B, Rutecki P, Sheth R, Wendt G, O’Leary D, Magnotta V. (2002) The neurodevelopmental impact of childhood-onset temporal lobe epilepsy on brain structure and function. Epilepsia 43:1062–1071. Jou RJ, Hardan AY, Keshavan MS. (2005) Reduced cortical folding in individuals at high risk for schizophrenia: a pilot study. Schizophrenia Research 75:309–313. Keller SS, Wieshmann UC, Mackay CE, Denby CE, Webb J, Roberts N. (2002) Voxel based morphometry of grey matter abnormalities in patients with medically intractable temporal lobe epilepsy; effects of side of seizure onset and epilepsy duration. Journal of Neurology, Neurosurgery and Psychiatry 73:648–655. Kulynych JJ, Luevano LF, Jones DW, Weinberger DR. (1997) Cortical abnormality in schizophrenia: an in vivo application of the gyrification index. Biological Psychiatry 41:995–999. Lee JW, Andermann F, Dubeau F, Bernasconi A, MacDonald D, Evans A, Reutens DC. (1998) Morphometric analysis of the temporal lobe in temporal lobe epilepsy. Epilepsia 39:727–736. Liu RS, Lemieux L, Bell GS, Sisodiya SM, Bartlett PA, Shorvon SD, Sander JW, Duncan JD. (2005) Cerebral damage in epilepsy: a population-based longitudinal quantitative MRI study. Epilepsia 46:1482–1494. Magnotta VA, Andresen NC, Schultz SK, Harris G, Cizaldo T, Heckel D, Nopoulos P, Flaum M. (1999) Quantitative in vivo measurement of gyrification in the human brain: changes associated with aging. Cerebral Cortex 9:151–160. Moran NF, Lemieux L, Kitchen ND, Fish DR, Shorvon SD. (2001) Extrahippocampal temporal lobe atrophy in temporal lobe epilepsy and mesial temporal sclerosis. Brain 124:167–175. Nopoulos P, Flaum M, O’Leary D, Andreasen NC. (2000) Sexual dimorphism in the human brain: evaluation of tissue volume, tissue composition and surface anatomy using magnetic resonance imaging. Psychiatry Research 98:1–13. Rockel AJ, Hiorns RW, Powell TPS. (1974) Numbers of neurons through the full depth of the neocortex. Journal of Anatomy 118:371. Rockel AJ, Hiorns RW, Powell TPS. (1980) The basic uniformity in structure of the neocortex. Brain 103:221–244. Ronan L, Doherty CP, Delanty N, Thornton J, Fitzsimons M. (2006) Quantitative MRI: a reliable protocol for measurement of cerebral gyrification using stereology. Magnetic Resonance Imaging 24:265– 272. Sallet PC, Elkis H, Alves TM, Oliveira JR, Sassi E, de Castro CC, Busatto GF, Gattaz WF. (2003) Reduced cortical folding in schizophrenia: an MRI morphometric study. American Journal of Psychiatry 160:1606–1613. Schmitt JE, Watts K, Eliez S, Bellugi U, Galaburda AM, Reiss AL. (2002) Increased gyrification in Williams syndrome: evidence using 3D MRI methods. Developmental Medicine and Child Neurology 44:292–295. Sisodiya SM, Free SL, Stevens JM, Fish DR, Shorvon SD. (1995) Widespread cerebral structural changes in patients with cortical dysgenesis and epilepsy. Brain 118:1039–1050. Tan NCK, Mulley JC, Berkovic SF. (2004) Genetic Association Studies in Epilepsy: “the truth is out there.” Epilepsia 45:1429– 1442. Theodore WH, DeCarli C, Gaillard WD. (2003) Total cerebral volume is reduced in patients with localization-related epilepsy and a history of complex febrile seizures. Archives of Neurology 60:250– 252. Tramo MJ, Loftus WC, Thomas CE, Green RL, Mott LA, Gazzaniga MS. (1995) Surface area of human cerebral cortex and its gross morphological subdivisions: in vivo measurements in monozygotic twins suggest differential hemisphere effects of genetic factors. Journal of Cognitive Neuroscience 7:292–301. Vogeley K, Schneider-Axmann T, Pfeiffer U, Tepest R, Bayer TA, Bogerts B, Honer WG, Falkai P. (2000) Disturbed gyrification of the prefrontal region in male schizophrenic patients: a morphometric post-mortem study. American Journal of Psychiatry 157:34– 39.
CEREBRAL CORTICAL GYRIFICATION Vogeley K, Tepest R, Pfeiffer U, Schneider-Axmann T, Maier W, Honer WG, Falkai P. (2001) Right frontal hypergyria differentiation in affected and unaffected siblings from families multiply affected with schizophrenia: a morphometric MRI study. American Journal of Psychiatry 158:494–496.
219
Zhang K and Sejnowski TJ. (2000) A universal scaling law between grey matter and white matter of cerebral cortex. Proceedings of the National Academy of Science USA 97:5621–5626. http://www.psychology.nottingham.ac.uk/staff/crl/mricro.html [Accessed 04/01/06].
Epilepsia, Vol. 48, No. 2, 2007
