Detection of Cortical Lesions is Dependent on Choice of Slice Thickness in Patients with Multiple Sclerosis

Abstract

Understanding the importance of cortical lesions in MS pathogenesis has changed. Histopathologic studies using new immunohistochemical methods show that cortical lesions can be detected more frequently than previously reported. Newer MRI sequences also detect cortical lesions more accurately. The aim of this study was to evaluate whether the effect of slice thickness (th) is an important factor for detection of cortical lesions in patients with multiple sclerosis (MS). We aimed also to investigate the relationship of cortical lesions with clinical status or other MRI variables. Forty-one patients with relapsing-remitting (RR) MS (11 males, 30 females with mean EDSS 2.3) underwent scans of Two-dimensional (2D)-fluid-attenuated inversion recovery (FLAIR) and 3D-T1-WI at 1.5-, 3-, and 5-mm slice thicknesses on 1.5-T MRI. Cortical and juxtacortical lesions were volumetrically assessed using a semiautomated method. 2D-FLAIR and 3D-T1-WI were coregistered and the matrix of the neocortical volume (NCV) segmentation mask (SIENAX-generated) was used to classify the location of the cortical-subcortical lesions. Cortical lesions fell into three classes: (1) class 1 were defined as lesions located in the NCV, (2) class 2 were juxtacortical lesions in contact with the NCV mask, and (3) class 3 were cortical-juxtacortical lesions situated in both regions. We measured NCV and normalized gray matter (GM) volume as well. We used partial correlation and multiple regressions to investigate the relationship between cortical lesions and other clinical and MRI variables. Of the total T2-lesion volume (T2-LV) measured on 1.5-mm th scans (mean 16108 mm(3)), cortical lesions represented 2.4% (276 mm(3)), juxtacortical lesions 6.1% (760 mm(3)), and cortical-juxtacortical 3.7% (491 mm(3)). Compared to 1.5-mm th scan, cortical LV was reduced by -28.3%, p < 0.001 on 3-mm th and by -40.78%, p < 0.001 on 5-mm th scans. Results for juxtacortical LV were for 3-mm th scans (-17.9%, p < 0.01) and for 5-mm th scans (-30.3%, p < 0.01). The figures for cortical-juxtacortical LV were also for 3-mm th scans (-16.2%, p < 0.01) and for 5-mm th scans (-26.7%, p < 0.01). We observed a significant correlation between T2-LV and GM atrophy in all slice thickness (r = -0.4 to -0.48, p = 0.001-0.003) and a modest relationship between cortical and cortical-juxtacortical LVs and disability, especially at 1.5-mm slice thickness (r = 0.35, p = 0.025). Use of thinner slices (1.5 mm) on 2D-FLAIR images can significantly increase the sensitivity and precision of detecting cortical and juxtracotical lesions in patients with MS. Cortical and juxtacortical lesions contribute more to disability development than total T2-LV alone.