P3-091 Regional distribution of white matter changes in Alzheimer's disease

Poster Session P3: Diagnosis and Disease Progression mechanism of contrast at 3 Tesla. T2 was measured with a whole-brain spin echo prepared 3D fast spin echo sequence using echo times of 16 and 76 ms. Thirteen subjects with AD (mean age 76, mean MMSE 19) and 11 age-matched control subjects (mean age 73) were studied. Images of the ratio of the short and long echo time were transformed to a standard space using SPM 99 software and compared for significant differences between patients and controls on a voxel by voxel basis. Results: Shortened T2 indicating elevated ferritin iron concentration was found in Alzheimer's patients in a bilateral distribution in the inferior putamen, globus pallidus and internal capsule. (Figure 1) Iron related changes were not seen in most of the cortex, perhaps due to masking from high T2 signal of CSF in areas of atrophy, though indications of shortened T2 were also seen in the posterior cingulate cortex. Conclusions: High field MRI can readily demonstrate differences in iron sensitive relaxation at high field strength in patients with Alzheimer's disease. High field MRI may be useful for longitudinal measurements to monitor the biochemical efficacy of metal chelation therapy. The spatial distribution of the iron related differences in AD encompasses the nucleus basalis of Meynert and may parallel the loss of cortical cholinergic enervation. A C O V A R I A N C E R E S T I N G P E T PATTERN THAT DISCRIMINATES BETWEEN HEALTHY ELDERLY AND EARLY AD PATIENTS CORRELATES WITH F U N C T I O N A L A N D C O G N I T I V E S E V E R I T Y IN SUBJECTS WITH C O G N I T I V E I M P A I R M E N T B U T NO DEMENTIA Nikolaos Scarmeas *1 , Christian G. Habeck I , Eric Zarahn 1, Karen E. Anderson 2, Aileen Park ~, John Hilton I, Gregory Pelton 1, Matthias H. Tabert 1 , Lawrence S. Honig 1, James R. Moeller 1, Davangere E Devanand 1 , Yaakov Stern I . 1Columbia University, New York, NY, USA; 2University of Baltimore, Baltimore, MD, USA. Contact e-mail: ns257@ columbia, edu Background: Multivariate techniques capture a different dimension of the imaging data (eovariance among brain regions) than voxel-based or ROI analyses, but have rarely been used to in detecting AD. Objective(s): We sought to identify a single covariance pattern of relative rest blood flow that would help in discriminating between AD patients and elderly controls, and would display meaningful associations with cognitive performance measures in subjects with minimal to mild cognitive impairment (CI) but not dementia. Methods: Non-quantitative H215 0 PET scans during rest were acquired in 16 probable AD subjects selected for mild severity (modified Mini Mental Status examination [mMMS] 46/57; sd 5.1) (CDR = 1), 16 cognitively intact elderly controls (mMMS 54/57; sd 2.5) (CDR = 0) and 23 CI subjects (mMMS 54/57; sd 2.6) (broadly defined, including subjects with CDR = 0 and CDR = 0.5). AD-Control discrimination was attempted via (i) voxel-wise comparisons, (ii) ROI analyses and (iii) multivariate voxel-wise regional covariance analysis. The covariance analysis derived pattern was then prospectively applied to the CI subjects. Results: There were no significant mean flow differences in either voxel-wise or ROI analyses. However, the mukivariate analyses identified a covariance pattern whose mean expression was significantly higher in the AD patients as compared to controls (p = 0.03) (sensitivity 76-94%; specificity 63-81%). Sites of increased concomitant flow included insuta, cuneus, pulvinar, lingual, fusiform, superior occipital and parahippocampal gyri, whereas decreased concomitant flow was found in cingulate, inferior parietal lobule, middle and inferior frontal, supramarginal and precentral gyri. When prospectively applied to the CI subjects, the covariance pattern discriminated well between subjects with CDR = 0 and CDR = 0.5 (p = 0.009). Expression of this pattern correlated inversely with Selective Reminding Test total recall (r = -0.401, p = 0.002), delayed recall (r = -0.351, p = 0.008) and mMMS scores (r = -0.401, p = 0.002) in all 3 groups combined, and in the CI group alone. Conclusions: Multivariate techniques may be of use in early AD diagnosis even when univariate methods fall. They also provide a sensitive tool for differentiating subjects with CI into those with higher and lower functional and cognitive abilities, and perhaps into those with greater probability of subsequent conversion to AD. •f• Neuroimaging MRI SIGNAL HYPERINTENSITIES PREDICT MORE RAPID C O G N I T I V E D E C L I N E IN ALZHEIMER'S DISEASE (AD) Lawrence S. Honig* 1, Nikolaos Scarmeas 1, Aliza Riba 1, Marilyn Albert 2, Jason Brandt 2, Yaakov Stern s . 1Columbia University College of Physicians & Surgeons, New York, NY, USA; 2Johns Hopkins University, Baltimore, MD, USA. Contact e-mail: lh456@columbia.edu Background: Cerebrovascular disease is common in the elderly, with or without AD. MRI frequently shows signal hyperintensities, of presumed microvascular origin, even in persons without clinical strokes. However, it is not established whether there is a relationship between onset, or course, of AD and these hyperintensities. Objective: Determine whether burden of MR[ T2-weighted signal hyperintensities relates to rapidity of cognitive de- cline in AD. Methods: We used data from a longitudinal, three-site study of patients with mild AD who were followed with neuropsychological testing including modified Mini-Mental State (mMMS; scale 0-57) every 6 months, for an average followup period of 2.3 years. Entry criteria excluded patients with cortical infarcts or lacunes > lcm. MRI films at baseline were available for 54 patients with mean age 75 yrs, education 14.9 yrs, and mMMS 43; this sample included 55% females, 56% apohpoproteinE4+, and 31% with hypertension, 12% diabetes, 18% hyperlipidemia, and 14% coronary heart disease. Each baseline FLAIR or T2-weighted MRI was rated, blind to clinical data, using the Scheltens scale (0-84 maximum) which includes assessment of periventricular, supratentorial, basal ganglionic, and infraten- torial signal hyperintensities. Results: Mean Scheltens total score was 8.3. Average decline in mMMS was 3.26 points/year. Using GEE analysis, we found a significant relationship (p = 0.04) between Scheltens MRI score and the rate of cognitive decline: 0.13 mMMS points/Scheltens point/year, or about 5% increased speed of decline per Scheltens point. This relationship remained significant, even with adjustments for baseline mMMS, age, sex, education, and apoE4. Examination of Scheltens component scores showed that supratentorial white matter lesions were most responsible for the ob- served association (p = 0.001). The effect of white matter abnormalities was not simply related to vascular risk factors (hypertension, diabetes, hyperlipidemia, or heart disease), since even after further adjustment for these variables, the association of cognitive decline and MR[ findings was preserved. Conclusions: Brain white matter signal change is associated with increased rapidity of cognitive decline in AD. This association does not appear to simply relate to underlying vascular risk factors, but may be due to white matter hyperintensifies providing a superimposed injurious factor on AD pathology, contributing to cognitive decline. REGIONAL DISTRIBUTION OF WHITE MATTER CHANGES IN ALZHEIMER'S DISEASE Christine Fennema-Notesfine .1,2 , Sarah L. Archibald 1,2, Anthony C. Gamst 3, Leon J. Thai 3, Terry L. Jernigan 1,2. 1UCSD Dept. of Psychiatry, La Jolla, CA, USA; 2VASDHS, San Diego, CA, USA; 3 UCSD Dept. of Neuroseience, La Jolla, CA, USA. Contact e-mail: fennema @ ucsd. edu Background: While previous work in Alzheimer's Disease (AD) has care- fully described gray matter changes, significant white matter (WM) volume loss and increases in abnormal signal in the WM have also been reported (e.g., Englund et al., 1987; Jemigan et al., 1991). Recent work has height- ened interest in better understanding WM changes in AD. Bartzokis (2004) proposed a model of AD-related neurodegeneration based on the normal pattern of myelination; later-myelinating regions (e.g., temporal and frontal cortices) may be affected first in AD, relative to early-myelinating regions (e.g., primary motor cortex). There is a clear need for more information on regional W M changes across the lifespan and in AD (Jernigan et aI., 2004). Objective: To systematically investigate the regional distribution of W M changes in AD with morphometric analyses of structural magnetic reso- nance imaging (MRI). Methods: Forty-eight AD and 44 elderly control (NC) participants were studied. Trained anatomists delineated frontal, parietal, occipital, temporal, and deep cerebral WM boundaries on tissue-segmented images. Regional WM volumes and amount of abnormal W M were mea