Comparing Researches Between MRI And Pathology And Postoperative Follow-up In Epileptogenic Focal Cortical Dysplasia

Objectives1. To explore the best MRI scanning protocols and MRI characteristics of different pathological subtypes in epileptogenic focal cortical dysplasia (FCD).2. To discuss the differences of surgical outcomes in MRI positive group and MRI negative group, temporal FCD group and extra-temporal FCD group, dual pathology group and pure FCD group.Subjects and Methods1. Study subjects and pathology classification The study subjects were 36 patients (40 lesions) between October 2005 and December 2010, who have been diagnosed as FCD by pathology. According to Palmini's classification system, these lesions were categorized as FCD typeⅠ(FCDⅠA and FCDⅠB) and FCD typeⅡ(FCDⅡA and FCDⅡB). The different surgical outcomes of MRI positive group and MRI negative group, temporal FCD group and extra-temporal FCD group, dual pathology group and pure FCD group were evaluated after follow-up.2. MRI scanning protocols Axial FSE T2WI and FLAIR, sagittal 3D T1WI, oblique coronal FSE T2WI and FLAIR, axial SE T1WI, coronal FLAIR. Both routine and optimization scanning protocols were performed in all patients.3. Image analysesThe following signs were mainly observed:(1) focal cortical thickening,(2) blurring of grey–white matter junction,(3)focal brain atrophy/hypoplasia,(4)abnormal sulci or gyri(deep sulci,broad gyri),(5)aberrent intensity of the subcortical white matter and its morphometry,(6)aberrent intensity of the gray matter and its morphometry. With at least one of the abnormal signs were considered as MRI positive group, those without considered as MRI negative group (including the normal MRI performance and hippocampal sclerosis). The correlation of the different subtypes FCD and the scanning position/sequence with the above signs were evaluated. Compare blind film-reading with binding clinic film-reading.4. Statistical analysesStatistical analyses were performed by SPSS 17.0 software. Fourfold tableχ2 test, correctedχ2 test, fourfold table Fisher's exact test and McNemar test were used. P<0.05 was considered to indicate a statistically significant difference, P>0.05 was indicated no statistically significant difference.Results1. Histopathological diagnosis and location of lesions According to Palmini's classification system, 27 cases (29 lesions,72.5%) in FCD typeⅠgroup and 10 cases (11 lesions,27.5%) in FCD typeⅡgroup. 5 cases (6 lesions,15%) in FCD typeⅠA group, 22 cases (23 lesions,57.5%) in FCD typeⅠB group, 8 cases (8 lesions,20%) in FCD typeⅡA group, 2 cases (3 lesions,7.5%) in FCD typeⅡB group.14 cases were dual pathology (FCD with hippocampal sclerosis).Among 36 cases (40 lesions), 29 were in temporal lobe (72.5%), 9 in frontal lobe (22.5%), 2 in parietal lobe (5%). 4 cases located in multi-lobes (frontal and temporal lobes).FCD was categorized as two groups (non-FCD typeⅡB group and FCD type ⅡB group) while FCD location was put into temporal lobe group and extra-temporal lobe group. By fourfold table Fisher's exact test, there were statistically significant differences between non-FCD typeⅡB group and FCD typeⅡB group (P=0.017). Non-FCD typeⅡB group was found in temporal lobe at most, while FCD typeⅡB group was found in extra-temporal lobe at most.In 29 lesions of temporal lobes, 25 were FCD typeⅠgroup, 4 were FCD typeⅡgroup, which respectively took up 86.2% (25/29) and 36.4% (4/11). By correctedχ2 test, there were statistically significant differences in two groups (P=0.006), there were more FCD typeⅠgroup in temporal lobes.2. MRI findingsIn 40 FCD lesions, 31 were MRI positive lesions (77.5%). By correctedχ2 test, there were statistically significant differences in FCD typeⅠgroup and FCD typeⅡgroup (P=1.000).The blurring of the gray-white matter junction has been found in 42.5% of all the cases,which covered 34.5% in FCD typeⅠgroup and 63.6% in FCD typeⅡgroup. Focal brain atrophy/hypoplasia was found in 32.5% of all the cases, which covered 44.8% in FCD typeⅠgroup. Focal cortical thickening has been found in 25% of all the cases, which covered 17.2% in FCD typeⅠgroup and 45.5% in FCD typeⅡgroup. Abnormal sulci or gyri have been found in 25% of all the cases, which covered 24.1% in FCD typeⅠgroup and 27.3% in FCD typeⅡgroup. Increased signal intensity of subcortical white matter on FLAIR/T2WI has been found in 20% of all the cases, which covered 6.9% in FCD typeⅠgroup and 54.5% in FCD typeⅡgroup. Increased signal intensity of the gray matter on FLAIR/T2WI has been found in 15% of all the cases, which covered 6.9% in FCD typeⅠgroup and 36.4% in FCD typeⅡgroup.By statistical analyses, 2 kinds of MRI findings indicated statistically significant difference in FCD typeⅠgroup and FCD typeⅡgroup. These were focal brain atrophy/hypoplasia, increased signal intensity of subcortical white matter on FLAIR/T2WI. The former sign was found more in FCD typeⅠ, while the latter was more in FCD typeⅡ.3. Correlation of different scanning sequence and FCD signs Blurring of the gray-white matter junction was showed well in T1WI, T2WI and FLAIR (T2WI was the best). Focal brain atrophy/hypoplasia showed better in T2WI than FLAIR and T1WI. Aberrant intensity of the subcortical white matter and gray matter showed better in FLAIR and T2WI than T1WI.There were no advantages for T1WI in showing FCD signs. Temporal lobe FCD and hippocampal sclerosis can be shown more distinctly by optimization scanning rather than routine scanning.4. Comparison between two film-reading methods 40 FCD lesions were read respectively by blind and carefully binding clinic film-reading methods. The number of FCD positive lesions was 21(52.5%) and 31 (77.5%) by the above two methods separately, while the number of FCD negative lesions was 19 (47.5%) and 9 (22.5%). By McNemar test, there were statistically significant differences in two methods (P=0.019). The method of carefully binding clinic film-reading can improve the accuracy of FCD diagnosis.5. Surgical outcomes There were 35 cases that had been followed up successfully (25 cases were effective). MRI positive group covered 90% (18/20) of all the cases, while MRI negative group covered 46.7% (7/15). There were statistically significant differences in two groups (P=0.008). Surgical effect of MRI positive group was better than MRI negative group. Temporal FCD covered 76% of all the cases, while extra-temporal FCD covered 60%. There was no statistically significant difference in two groups (P=0.421). Pure FCD group covered 57.1% (12/21) of all the cases, while dual pathology group covered 92.9%. There were statistically significant differences in two groups (P=0.028). Surgical effect of dual pathology group was better than pure FCD group.Conclusions1. There are more FCD typeⅠB, which locate in temporal lobe at most. FCD typeⅡB group is found in extra-temporal lobe at most. 2. FCD has the following MRI characteristics:①Blurring of the gray-white matter junction is the most common marker of FCD; the following are focal brain atrophy/hypoplasia, focal cortical thickening and normal sulci or gyri.②MRI is useful to discriminate the subtypes of FCD. Focal brain atrophy/hypoplasia is typical in FCD typeⅠ; the feature of FCD typeⅡis increased signal intensity of subcortical white matter on FLAIR/T2WI.3. Optimization scanning is easier to show temporal lobe FCD and hippocampal sclerosis than routine scanning.4. The method of binding clinic film-reading is better, which can improve diagnostic accuracy of FCD.5. A higher MRI diagnostic accuracy contributes to a better surgical outcome. MRI performances could be viewed as predictors of postoperative curative effect. There is no significant difference in temporal FCD and extra-temporal FCD. Surgical effect of dual pathology FCD is better than the effect of pure FCD.