P2-208 BASELINE ASSESSMENT OF EEG BIS-AD IN LONGITUDINAL TRIAL PREDICTS SUBJECTS WHO LATER SOUGHT PHARMACOLOGICAL COGNITIVE THERAPY Scott D. Greenwald 1 , Vikas Saini 2 , Charles Smith 1 , Philip Devlin 1 , 1 Aspect Medical Systems, Inc., Newton, MA, USA; 2 The Cardiovascular Specialists, Inc., Hyannis, MA, USA. Contact e-mail: sgreenwald@aspectms.com Background: BIS-AD, an EEG-derived index, was developed as a sensi- tive, simple-to-use biomarker for the early detection of dementia. Previous reports have demonstrated that BIS-AD correlates with metrics of cogni- tive function (e.g., Mini-Mental State Exam (MMSE), Alzheimer’s Disease Assessment Scale (ADAScog)) in normal and demented subjects. Objec- tive(s): This analysis evaluated BIS-AD in subjects in an ongoing longi- tudinal trial who were assessed at enrollment (i.e., unmedicated baseline), and before and after initial treatment with pharmacological cognitive en- hancers. Methods: Following IRB approval, elderly subjects living inde- pendently in their community (age75; baseline MMSE 25) entered an ongoing longitudinal trial designed to explore early detection of dementia. EEG, MMSE and ADAScog were recorded every 3 months. Unpaired t-tests, paired t-tests, and Fisher’s Exact Test were used to measure differ- ences between groups or visits. P0.05 was considered statistically sig- nificant. Results: To date, 205 subjects have enrolled (format: mean +/- SD [range]; age: 80.5 +/ 3.7 [75-91]; MMSE: 28.9 +/- 1.1 [25-30] ; ADAScog: 7.4 +/- 3.3 [1-17]; BIS-AD: 94.8 +/- 3.2 [85-100]; Months in study: 9.4 +/- 8.0 [0-30].) Four subjects (2%) began pharmacological cognitive enhancing therapy (donepezil HCL (n=3) or mementine HCL (n=1)) during the study (14.3 +/- 6.2 [6-21] months after enrolling) . BIS-AD was lower (91.1 vs. 94.9, p=0.019) and MMSE (27.8 vs. 28.9, p=0.029) and ADAScog (13.6 vs 7.3, p=0.008) were worse at baseline in these 4 subjects who sought treatment compared to all others. In the 3 subjects monitored for more than 6 months prior to treatment, BIS-AD decreased from baseline prior to treatment (-2.3 +/- 0.90, p=0.049). At the visit following initial treatment, BIS-AD increased from pre-treatment in the 3 subjects with improved cognitive assessments (2.5 +/- 2.1 [0.5-4.6]), and decreased in the subject whose cognition worsened (-1.0 +/- DNE; p=0.25). Conclusions: BIS-AD was lower at baseline in subjects who later sought pharmacological cognitive therapy. Change in BIS-AD fol- lowing treatment onset reflected the change in cognitive ability. BIS-AD might be useful for early detection of dementia, tracking the course of change of Alzheimer’s disease, and evaluating the impact of potential therapies on the progression of the disease. P2-209 PROGRESSIVE DECLINE IN MUSICAL ABILITIES WITHOUT LANGUAGE DEFICITS Rodrigo Rizek Schultz, Marcel Abramo, Paulo Henrique Ferreira Bertolucci, Aline Martins Pinto, Cle ´o Monteiro Franc ¸a Correia, Federal University of Sa ˜o Paulo, Sa ˜o Paulo, Brazil. Contact e-mail: rodrigo@cian.med.br Background: Mild cognitive decline (MCI) is defined as an isolated decline of one area of cognition with general cognition spared. The interest in this condition is the much higher probability of AD in comparison with general population. Though amnesic MCI is reported more often, there may be isolated impairment of other areas, such as language and executive function. Objective(s): To present a case of progressive decline of musical abilities sparing other areas of cognition, in what could be considered a musical MCI. Methods: A 61 year old right handed professional musician presented for evaluation complaining of increasing difficulty in playing new songs starting 10 years before. As the symptoms worsened she had difficulty with well-learned songs as well, and had to retire from her former activity. Otherwise she was in good health and had a transient visual and speech disturbance 10 years before. Two grandparents had a diagnosis of AD. She started galantamine 18 months before and felt much better. Results: Her MMSE score was 29, she had an excellent performance in the WMS word list, Boston naming test, reading, and writing. Her performance in the recognition of pitch, duration, intensity and timbre of sounds was within normal range. However, she presented a deficit to recognize the environment sounds. These results suggest the presence of auditory agnosia for non verbal sounds and it represents an evolutive way to a global agnosia or can be associated to a deficit of perception of musical sounds. Conclu- sions: While for the non musicians music abilities are lateralized to the right hemisphere, for musicians these functions are more often co-localized with speech. Though a professional musician, this patient showed a disso- ciation between these functions. This case is compatible with progressive auditory agnosia, and her difficulty was discovered only because of her occupation. It is possible that cases of MCI affecting this kind of function, not necessary for “formal” functioning may be often missed. P2-210 THE ARCHITECTURE OF MILD COGNITIVE IMPAIRMENT: A PRELIMINARY STUDY William Britt, III, Sofia Bhaskerrao, Floyd Petersen, Waheed Baqai, James Larsen, Cindy Dickson, Wolff M. Kirsch, Loma Linda University, Loma Linda, CA, USA. Contact e-mail: wbritt@llu.edu Background: Correlations of cognitive test-retest score changes of elderly patients in dementia studies remains controversial. Objectives: To deter- mine associations between novel test-retest neuropsychological indicators of elderly controls and Mild Cognitive Impairment (MCI) participants (Petersen Mayo criteria). Methods: 1) Serial cognitive performance at 6-month to 1-year-intervals using standard neuropsychological and infor- mant-based measures. 2) Practice effects on performance between the two groups. 3) Characterization of participants failing to benefit from practice effects. 4) Determining the effect of education on brain resiliency. 5) Correlations with blood and neuroimaging studies. Results: Screening 1300 volunteers selected 183 individuals for testing with (comprehensive neuropsychological batteries) and videotaped Clinical Dementia Ratings (CDR). Twenty-eight controls and 76 MCI participants have been studied the past 36 months. Sixteen MCI participants have withdrawn; all controls remain. Serial MR (SWI, MP-RAGE, MRS, FLAIR, T 1 ,T 2 ) and blood studies (iron regulatory proteins, cytokines) are correlated with test-retest cognitive status. Control and MCI cohorts test significantly different for MMSE, Logical Memory raw score, CDR raw score, total sum of CDR boxes, Trail Making Test B, Word Fluency: Phonemic and Semantic, Boston Naming Test (short version), Wisconsin Card Sorting Perseverative Errors and Categories Complete, and Draw-A-Clock at baseline. Of the 76 MCI cases at baseline, 68 are multiple domain and 8 are single domain MCI amnestic. Subsequent tests performed at 6-months to 1-year after initial examination for 36 months revealed significant differences between control and MCI performance for the same set of tests. MR and blood studies form baseline values for further comparison. Conclusions: 1) Control participants performed significantly better than MCI participants on both sequential neuropsychological and informant-based measures. 2) Control participants have improved cognitive performance over time due to practice effects in semantic memory. 3) MCI participants did not benefit from practice effects. 4) Progressive MCI participants have low semantic memory scores, those who have not progressed show more variance. The cognitive decline of the progressive MCI participants became rapidly precipitous during the time frame of this study and can be correlated with MR and blood studies. This research was funded by NIH Grant #AG20948. P2-211 PERFORMANCE OF THE MCDR: A MODIFICATION OF THE CLINICAL DEMENTIA RATING SCALE FOR DIAGNOSING AND STAGING MCI Ranjan Duara 1 , Maria T. Greig-Custo 1 , Warren W. Barker 1 , Silvia Strauman 1 , Maria Aviles 2 , Karina Lifschitz 2 , David Loewenstein 1 , 1 Mount Sinai Medical Center, Miami Beach, FL, USA; 2 Johnny B. Byrd Alzheimer’s Institute, Tampa, FL, USA. Contact e-mail: billy@msmc.com S297 Poster P2: Monday Posters