HIPPOCAMPAL HISTOPAHTOLOGY RELATES TO 1H SPECTOSCOPIC IMAGING IN MTLE
Abstract number :
1.117
Submission category :
Year :
2003
Submission ID :
4091
Source :
www.aesnet.org
Presentation date :
12/6/2003 12:00:00 AM
Published date :
Dec 1, 2003, 06:00 AM
Authors :
Hoby P. Hetherington, Jung H. Kim, Aaron A. Cohen-Gadol, Dennis D. Spencer, Jullie W. Pan Radiology, Albert Einstein College of Medicine, Bronx, NY; Pathology, Yale University, New Haven, CT; Neurosurgery, Yale University, New Haven, CT; Neurology, Albert
1H spectroscopic imaging of NAA/Cr in unilateral mTLE has been used extensively for lateralization and as an evaluation of neuronal mitochondrial function. The utility of NAA/Cr as a parameter of injury has arisen from the observation that decreases in NAA occur with neuronal mitochondrial dysfunction, while Cr is found in high concentrations in glia. Thus, given the potential role of bioenergetic sufficiency in seizures and seizure control, it is of importance to determine whether there is relation between neuropathology of mTLE with the non-invasive evaluations of NAA/Cr. In this study we evaluated high-resolution spectroscopic imaging of NAA/Cr in relation to post-operative histopathology in n=12 unilateral mTLE patients.
High-resolution 1H spectroscopic imaging of NAA/Cr (0.64cc nominal resolution) was taken in the hippocampal plane in 12 pre-operative patients and 12 controls. Five voxels spanned the anterior-posterior length of the hippocampi bilaterally, from the tail to the pes hippocampus. These regions and corresponding spectra are shown in Fig 1(left). The NAA/Cr ratio was determined and correlated to neuropathologic evaluation of hippocampal glial fibrillary acidic protein (GFAP) staining, gliosis and neuronal loss.
Linear regression analysis of the ipsilateral NAA/Cr (any voxel) did not reveal a statistically significant relation to neuronal loss. However, the ipsilateral NAA/Cr did show significant regressions with GFAP staining, from all sectors CA1-CA4 (R values ranged from -0.60, p[lt]0.05 with CA3 to -0.75, p[lt]0.01 with CA2). While the NAA/Cr did not relate linearly to the dentate GFAP staining, there was a clear logarithmic dependence with dentate GFAP staining, with R= -0.71, p[lt]0.01 (Fig 1 right).
The significant regression of NAA/Cr with all the CA sectors suggests that in the hippocampus, neuronal injury is linearly related to GFAP staining. GFAP expression is highly induced with repetitive seizures, is elevated in chronic epilepsy, and may be a complementary measure to neuronal loss. However given the dentate relationship, this suggests that the reactive GFAP based injury in the CA sectors (possibly in response to mossy fiber sprouting) occurs non-linearly and reaches a maximum in comparison to the dentate. Thus, the damage extending from the dentate into the CA sector can be assessed by 1H spectroscopic imaging, and may reflect a gatekeeper role for the dentate in seizure propagation.[figure1]
[Supported by: NIH NINDS P01 NS39092, R01 NS40550 and the Charles A. Dana Foundation]