Abstracts

Rationale: Original MRI criteria of hippocampal sclerosis (HS) consisted of atrophy, signal abnormality (high T2, low T1), and loss of hippocampal internal architecture (HIA) clarity. HIA refers to the dark band on T2w images of tissue that consists of the strata radiatum, lacunosum, and moleculare (SRLM) that separates the subiculum, CA1, and CA2 regions from the hilus/CA4 and defines the characteristic spiral appearance of Ammon’s horn in coronal MRI images. While atrophy and signal abnormality in HS have been widely studied, the significance of loss of HIA in has been neglected. This is due to the difficulty of clearly imaging HIA in healthy hippocampi, therefore pathologic loss of HIA clarity is difficult to detect with confidence. We have previously reported that loss of HIA clarity is associated with the side of seizure onset in TLE, and we recently published an MRI method (HR-MICRA) that clearly and consistently produces images with good to excellent HIA clarity. Here we apply that method in a small sample of surgical TLE patients and compare the HIA clarity to patterns of astrogliosis (AG) on pathologic samples. Our hypothesis is that loss of HIA clarity in HS is due to increased AG in the SRLM, which causes increased T2 signal intensity of this normally dark band, thus decreasing its contrast relative to the surrounding gray matter.

Methods: Seven patients were identified from a larger study of the HR-MICRA method in TLE who had histopathological samples through the body of the hippocampus that allowed for identification of the granule cell layer of the dentate gyrus and the adjacent CA4 and SRLM. Slides were reviewed by a neuropathologist and rated (mild, moderate, or severe AG) in CA4 and SRLM separately. The HR-MICRA method produces images with resolution of 0.5 mm isotropic with high gray/white contrast and good SNR. Coronal images through the hippocampus were given a single rating for each hippocampus as “no loss of HIA clarity,” “obvious blurring of HIA,” or “partial loss of HIA clarity sparing the subiculum”; the latter category emerged as a distinct pattern through review of images.

Results: Of the 7 participants, 2 had no loss of HIA clarity, 3 had obvious blurring of HIA, and 2 had partial loss of HIA clarity sparing the subiculum (Fig. 1). Patients with no loss of HIA clarity had only mild AG in the SRLM but moderate or severe AG in CA4, thus showing a marked difference in the degree of AG between regions. By contrast, patients with obvious blurring of HIA had severe (2) or moderate (1) AG in SRLM and severe (3) AG in CA4, thus showing a minor degree of AG between regions. Patients with partial loss of HIA had an intermediate pattern with mild-moderate or moderate AG in the SRLM and moderate-severe or severe AG in CA4.

Conclusions: This preliminary sample of patients with surgical TLE provides evidence supportive of our hypothesis that loss of HIA clarity (blurring) reflects increased AG in the SRLM relative to CA4. AG is known to produce increased T2w signal on MRI. This is the first known report of a layer-specific pattern of AG in HS to be correlated with HIA on MRI.

Funding: Please list any funding that was received in support of this abstract.: R01NS094743.