Read WESTERGREN text version

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

3rd C o n f e r e n c e Clin ical Trials o n Alzh eimer's Disease

LECTURES

November 3-5, 2010 Hôtel-Dieu Saint Jacques, Toulouse, France

pathology, and to draft recommendations for the process of defining the preclinical stages of AD. We chose the term "preclinical AD" to refer to the full spectrum from completely asymptomatic individuals with evidence of AD pathologic change on biomarkers to individuals manifesting subtle cognitive decline but who do not yet meet accepted clinical criteria for MCI. The workgroup reached consensus on several key concepts. First, both the pathophysiological process and clinical course of AD are best conceptualized as continuums. Second, emerging evidence suggests that these two processes typically evolve with parallel but temporally offset trajectories with pathologic changes occurring prior to clinical changes. Third, based on evidence from both genetic- and age-at-risk cohorts, amyloid-ß accumulation appears to represent one of the earliest inciting events and, through mechanisms that remain to be elucidated, leads to "downstream" synaptic dysfunction, neurodegeneration, and neurofibrillary tangle formation, eventually resulting in the clinical symptoms of AD. While amyloid dysregulation may be a necessary event, it may not be sufficient to result in clinical signs and symptoms of AD. Furthermore, it is clear from clinical-pathologic studies that some individuals with AD pathologic changes may not manifest clinical symptoms in life. We propose operational research criteria in order to facilitate a common framework for research to further characterize the preclinical stages of AD. Research cohorts could be selected on the basis of staging criteria, which would optimize the ability to ascertain the specific outcomes important for a given type (e.g. natural history or treatment trial) and duration of the study. Stage 1:Asymptomatic Cerebral Amyloidosis ­ evidence of amyloid positivity on CSF or PET amyloid imaging; Stage 2:Amyloid positivity + Neurodegeneration - Evidence of synaptic dysfunction and/or neurodegeneration on FDG-PET, CSF tau, or volumetric MRI; Stage 3:Amyloid positivity + Neurodegeneration + Subtle cognitive decline - Evidence of cognitive decline over time or poor performance on sensitive cognitive measures, but not yet meeting accepted criteria for MCI. The extent to which it will be feasible to distinguish among these stages is not yet clear. Recognizing that current biomarkers of amyloid-ß primarily reflect fibrillar amyloid deposition, and that oligomeric forms of amyloid may be critical in the pathological cascade, it may be possible in the future to detect a stage of disease that precedes Stage 1. Significant work is required to standardize biomarker and cognitive thresholds. These research criteria are not intended to serve as diagnostic criteria, and do not have any clinical implications at the present time, but are intended to provide a common language to help researchers advance the preclinical study of AD. Further study is required to determine the clinical course of individuals with biomarker evidence of AD pathologic changes, and whether treatment at this stage can delay the progression on biomarkers and emergence of clinical symptoms. We must not delay in beginning these studies, despite the challenges, as similar to the examples of cardiovascular disease and cancer, it is likely that treatment at the preclinical stage of AD holds the greatest promise for maximal efficacy.

ADNI is a $67 million observational longitudinal clinical trial with the overall goal of determining the value of MRI and PET imaging together with blood and CSF biomarkers for disease modifying Alzheimer's treatment trials. Specific aims are to: improve methods for clinical trials; determine the optimum methods for acquiring and processing images and biomarkers; "Validate" imaging and biomarker data by correlating with neuropsych and behavioral data. We have longitudinally studied: MCI (n= 400); AD (n= 200); Controls (n= 220). Clinical visits include neuropsychological assessments, MMSE, CDR, ADAS-cog, MRI (1.5 T), FDG PET blood and urine and CSF. The rate of hippocampus atrophy appears to have the high statistical power for measuring change over time, and further that changes in this same region predict conversion from MCI to AD. FDG PET measures also have high predictive value of MCI conversion to AD and cognitive decline, and some FDG PET measures have high power as outcomes. Furthermore, analysis of data from the normal controls suggests that normal healthy elders with APOE4 and/or low CSF A amyloid have worse memory scores and higher rates of hippocampal atrophy than the non carrier controls or subjects with high CSF A; this is consistent with the hypothesis that some controls have preclinical AD pathology. Similar ADNI-like projects, with similar methods, are underway in Japan, Australia, and Europe. We recently received a $24 million grant to enroll 200 early MCI, and to perform F18 amyloid imaging on all ADNI and GO subjects using AVID's AV-45 ligand. World-wide standards and a world-wide network have been created for clinical trials. Updated analysis of ADNI followup data will be presented. CRITERIA FOR ALZHEIMER'S DISEASE DEMENTIA. D. KNOPMAN (Mayo Clinic, USA )

DESIGNING PRIMARY PREVENTION TRIALS USING BIOMARKER AND CLINICAL DATA FROM THE ALZHEIMER'S DISEASE NEUROIMAGING INITIATIVE. M.W. WEINER1, P. AISEN2, R. PETERSEN3, C. JACK3, W. JAGUST4, J. TROJANOWSKI 5, L. SHAW 5, A. TOGA 6, L. BECKETT 7, A. GAMST 2 (1. San Francisco V A Medical Center/University of California San Francisco, San Francisco, CA , USA ; 2. University of California San Diego, San Diego, CA , USA ; 3. Mayo Clinic, R ochester, M N , US A ; 4. Univ ersity of Calif ornia B erk eley , B erk eley , CA , US A ; 5. University of Pennsylvania, Philadelphia, PA , USA ; 6. University of California Los A ngeles, Los A ngeles, CA , USA ; 7. University of California Davis, Davis, CA , USA )

Redefining Alzheimer's Disease: Revising the NINCDS-ADRDA Criteria. In the fall of 1983, a group was convened to establish and to describe the clinical diagnosis of Alzheimer's disease (AD). The group convened subgroups which addressed issues of medical history, clinical examination, neuropsychological testing, and laboratory assessments and then produced an integrated report, published in July 1984. The criteria in this report commonly referred to as the NINCDS-ADRDA criteria, have been quite successful, surviving for over 25 years. These criteria have been reliable for the diagnosis of probable AD, with a sensitivity of 80% and specificity of 70%. In 2009, the Alzheimer's Association convened a Research Roundtable on Alzheimer's Disease Classification. A review of recent research advances presented at this meeting indicated that it was time to revisit the NINCDS-ADRDA Criteria and perhaps revise the criteria to include modern methodology and address whether there is sufficient evidence for an earlier diagnosis. The National Institute on Aging and the Alzheimer's Association has convened three workgroups to examine evidence for the following three clinical categories: · Criteria for Alzheimer's Disease Dementia; · Criteria for Symptomatic Pre-Dementia Alzheimer's Disease; · Criteria for Preclinical Alzheimer's Disease Dementia.

TOWARDS DEFINING THE PRECLINICAL STAGES OF ALZHEIMER'S DISEASE. R. SPERLING, E. REIMAN, D. BENNETT, C. JACK, A. FAGAN, Y. STERN, E. SIEMERS, D. PARK, J. KAYE, S. CRAFT, L. BECKETT, K. YAFFE, T. MONTINE, P. AISEN, W. THIES, M. CARILLO, M. WAGSTER, C. PHELPS (The NIA /A lzheimer A ssociation Preclinical W orkgroup) Converging evidence suggests that the pathophysiological process of Alzheimer's disease (AD) begins years, if not decades, prior to the diagnosis of dementia. This long preclinical phase of AD provides a particularly advantageous opportunity for therapeutic intervention, but it is critical to link what is known about the pathological cascade of AD to the emergence of clinical symptoms. The National Institute on Aging and Alzheimer's Association convened a workgroup to review the biomarker, epidemiological, and neuropsychological factors that best predict the risk of progression from "normal" to mild cognitive impairment (MCI) and clinical dementia thought to be due to underlying AD

S2

Aim: To review the potential use of amyloid PET imaging to better understand the role of -amyloid accumulation in pathological aging, progressive late-life cognitive impairment, and neurodegenerative dementing syndromes with mixed pathology. Method: Review of the imaging properties of florbetapir F 18 and the correlation of a florbetapirPET signal and the presence of -amyloid aggregation found at autopsy. Review the current gaps in understanding the role -amyloid aggregation in conditions associated with age related physical and mental decline and how florbetapir-PET imaging can be incorporated into research protocols designed to address these gaps. Results: Florbetapir F 18 has favorable imaging characteristics including rapid brain uptake and clearance, strong selective binding affinity for -amyloid aggregation, early separation of cortical to cerebellar standard uptake value ratio with stability between 30 and 90 minutes in subjects with cortical -amyloid, excellent test-retest reliability and an excellent safety profile. An imaging to pathology study has established the ability of florbetapir F 18 to identify the presence and density of -amyloid aggregation in living subjects. Discussion: Amyloid PET imaging in research studies of elderly subjects can now be obtained in both academic and community based PET facilities, facilitating the ability to include measures of cerebral -amyloid aggregation to address the impact of this neurodegenerative protein on the development of cognitive impairment, behavioral changes, and frailty associated with pathological aging. Improved knowledge should translate into improvements in clinical care and provide strategies for the development of treatments to reduce the incidence and severity of impairments that are often thought to be unavoidable components of aging. Amyloid PET will be useful to address the following questions:1. What is temporal relationship between -amyloid deposition and cognitive impairment? Is amyloid deposition in cognitively normal elderly subjects the earliest objective predictor of later

INCORPORATING AMYLOID IMAGING IN CURRENT AND/OR FUTURE COLLABORATIVE RESEARCH STUDIES. C.M. CLARK, A.P. CARPENTER, F. HEFTI, M.A. MINTUN, M.J. PONTECORVO, D.M. SKOVRONSKY (A v id Radiopharmaceuticals, Philadelphia, PA , USA 19104)

cognitive decline? Does the rate of amyloid accumulation change in a linear manner as individuals progress through the preclinical, prodromal, and dementia stages of AD? Is there a rapid buildup of amyloid in the prodromal stage followed by relative homeostasis? 2. What is the relationship between reductions in -amyloid levels and cognitive improvements in patients given amyloid-lowering drugs? Are such drugs effective in subjects with preclinical AD? During anti-amyloid treatment, does longitudinal -amyloid imaging have value in predicting failure or success of treatment? Can -amyloid imaging assist in choices of dosage or therapy type? 3. How effective are putative nonpharmacologic `therapies' for AD (e.g. exercise, mental stimulation, diet) in altering amyloid plaque levels? 4. How can -amyloid imaging be used in conjunction with biomarkers of inflammation to better understand the role of inflammation in AD pathology and treatment? 5. What is the role of -amyloid in conditions related to AD, such as Down's syndrome?

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

RE-THINKING ALZHEIMER DISEASE THERAPY; LESSONS LEARNED FROM ONGOING CLINICAL TRIALS. B. WINBLAD, F. MANGIALASCHE, A. SOLOMON, P. MECOCCI, M. KIVIPELTO (KI-A lzheimer Disease Research Center (KI-A DRC), Sweden) Research into AD has been at least partly successful in terms of developing symptomatic treatments, but has also had several failures in terms of developing diseasemodifying therapies. These successes and failures have led to debate about the potential deficiencies in our understanding of the pathogenesis of AD and potential pitfalls in diagnosis, choice of therapeutic targets, development of drug candidates, and design of clinical trials. Many clinical and experimental studies are ongoing, but we need to acknowledge that a single cure for AD is unlikely to be found ant that the approach to drug development for this disorder needs to be reconsidered. The amyloid cascade hypothesis dominates current drug development strategies, but whether A is more pathognomonic than pathogenetic is not yet clear, and so is the therapeutic role of A removal. Preclinical research is constantly providing us with new information on pieces of the complex AD puzzle, and an analysis of this information might reveal patterns of pharmacological interactions instead of single potential drug targets. The first signs of a shift away from linear one protein - one drug thinking have already appeared: research is moving from proteins to focus on organelles (eg mitochondria) and also multi-target-directed ligands. A better understanding of the disease pathogenesis, but also solving methodological problems in clinical trials on AD - eg standardized diagnostic criteria to identify homogeneous group of patients, appropriate treatment duration and measures of disease-modifying effects - will help finding a cure for AD.

Background and Aim. An increased rate of brain atrophy is often observed in older subjects, in particular those who suffer from cognitive decline. Homocysteine is a risk factor for brain atrophy, cognitive impairment and dementia. Plasma concentrations of total homocysteine can be lowered by dietary administration of B vitamins. The aim was to determine whether supplementation with B vitamins that lower levels of plasma total homocysteine can slow the rate of brain atrophy in subjects with mild cognitive impairment in a randomised controlled trial (www.controlled-trials.com ISRCTN 94410159). Methods. Single-center, randomized, double-blind controlled trial of high-dose folic acid, vitamins B6 and B12 in 271 individuals (of 646 screened) over 70 y old with mild cognitive impairment. A subset (187) volunteered to have cranial MRI scans at the start and finish of the study. Participants were randomly assigned to two groups of equal size, one treated with folic acid (0.8 mg/d), vitamin B12 (0.5 mg/d) and vitamin B6 (20 mg/d), the other with placebo; treatment was for 24 months. The main outcome measure was the change in the rate of atrophy of the whole brain assessed by serial volumetric MRI scans (SIENA protocol). The trial had power of 90% to detect a 20% slowing of the rate of atrophy in 70 subjects in each group. It was not powered to detect changes in cognitive test scores, although several such tests were carried out. Results. A total of 168 participants (85 in active treatment group; 83 receiving placebo) completed the MRI section of the trial. The atrophy data were analysed on a per protocol basis to answer the following questions: 1. Does treatment slow down the rate of atrophy of the brain compared with placebo? 2. If so, is the effect of treatment related to baseline homocysteine levels? 3. If so, is the effect of treatment related to the degree of lowering of homocysteine levels? Finally, in an ITT analysis, can any significant effect of treatment on cognition be shown, in spite of apparent lack of power? Discussion. The possible significance of the results as a new approach to disease modification will be discussed.

RESULTS OF THE VITACOG TRIAL OF HOMOCYSTEINE-LOWERING B VITAMINS IN SUBJECTS WITH MCI. A.D. SMITH 1 , S.M. SMITH, C.A. DE JAGER, P. WHITBREAD, C. JOHNSTON, G. AGACINSKI, A. OULHAJ, K.M. BRADLEY, R. JACOBY, H. REFSUM (1. University of Oxford, U.K, H. Refsum, University of Oslo, Norway; Oxford Project to Investigate Memory and A geing (OPTIMA ), University of Oxford, UK)

The detection of Alzheimer's disease (AD) in its presymptomatic stages is becoming a reality. Our recent studies show this using imaging modalities including: FDG-PET and high resolution serial MRI. However, while FDG-PET and MRI studies of the hippocampal formation show sensitivity for early features of AD at both the normal and MCI stages of cognition and show good correlation with clinical progression, these measures alone are not specific for AD. Additional valid markers for the neurofibrillary and amyloid plaque pathology of AD are required for diagnostic confidence. Our recent studies show that CSF measures of hyperphosphorylated (P) P-tau contribute this added potential and correlate with brain changes even in the presymptomatic stages. CSF and PET biomarkers of amyloid beta (Ab) to a lesser extent show diagnostic and correlative potential in MCI and NL stages. Overall, the data suggest a combination of both sensitive and specific markers is useful for advance diagnosis. Moreover, as AD specific biomarkers do not show meaningful progression effects, imaging will be of continued value for following the course of AD. We will present longitudinal evidence supporting the combined use of imaging and AD specific biomarkers in the early diagnosis of AD. Perhaps the most important objective for imaging and biomarkers is the identification of new mechanisms of AD progression in NL subjects. In a series of longitudinal NL aging studies we observe two emergent mechanisms: 1) presymptomatic subjects with a maternal inheritance of AD, show reduced glucose metabolism cross-sectionally and longitudinally and evidence for brain amyloid deposits; and 2) using MRI-ASL, NL elderly with elevated plasma Ab40 levels show impaired CO2 induced vasodilation of hippocampal blood vessels that is independent of the vascular reactivity deficits associated with cerebrovascular disease. In summary, imaging has provided sensitive measures for the detection of AD in MCI and presymptomatic stages. Both FDG-PET and MRI imaging when combined with AD specific biomarkers improve diagnostic classifications at all stages of AD. Thus, the presymptomatic recognition of AD is a realistic goal. However, biomarkers and imaging have also provided new evidence for potential mechanisms related to AD progression, where those marked by maternal inheritance suggest a possible mitochondrial pathology and that marked by impaired vasoreactivity, a possible vascular mediated basis for progression. Key words: Presymptomatic diagnosis of AD, biomarkers, imaging, disease mechanism DESIGNING SECONDARY PREVENTION TRIALS IN AD. P.S. AISEN (Department of Neurosciences, University of California San Diego, USA )

PRESYMPTOMATIC TESTS FOR AD AND THE SEARCH FOR MECHANISM. M. J. DE LEON1,2, L. MOSCONI1, L. GLODZIK1, H. RUSINEK1, S. DESANTI1, Y. LI 1, W. TSUI 1,2 , E. PIRRAGLIA 1 , D. PRATICO 3 , S. VALLABHAJOSULA 4 , R. ZINKOWSKI5, P.D. MEHTA6 , H. ZETTERBERG7, K. BLENNOW7 (1. New Y ork University School of Medicine, 145 East 32nd Street, New Y ork, NY , 10016, USA ; 2. Nathan Kline Institute, Orangeburg; 3. Temple University, Philadelphia, PA ; 4. Cornell University School of Medicine; 5. A pplied NeuroSolutions, V ernon Hills IL; 6. Institute for Basic Research, Staten Island; 7. Salgrenska University Goteborg, Sweden)

S3

Aim: Alzheimer disease (AD) is multifactorial and involves several different etiopathogenic mechanisms which all end in the same histopathological disease hallmarks, the Aß plaques and the phosphotau neurofibrillary tangles. Thus, it is critical to identify the disease-relevant biomarker signatures for each subgroup of AD that can be employed accordingly in the therapeutic clinical trials. Methods: Self-assembly of abnormally hyperphosphorylated tau isolated from AD brains was carried out following its dephosphorylation with protein phosphatase-2A (PP2A) and rephosphorylation by different combinations of tau protein kinases. The CSF levels of Aß1­42 as a marker of Aß plaques, total tau and ubiquitin as markers of neurofibrillary tangles were determined by ELISAs in 353 clinically diagnosed AD and 115 non-AD control patients, and subjected

DIFFERENT ETIOPATHOGENIC MECHANISMS AND SUBGROUPS OF AD BASED ON CSF MOLECULAR MARKERS. K. IQBAL, I. GRUNDKE-IQBAL (New Y ork State Institute for Basic Research in Developmental Disabilities, Staten Island, New Y ork, USA )

Although several definitions exist, secondary prevention can be considered to encompass early disease detection allowing interventions to prevent progression of the disease and emergence of symptoms. Alzheimer's disease therapeutic research has now reached the stage at which secondary prevention can be addressed. A plausible strategy for identifying preclinical (ie, asymptomatic) AD has been proposed, involving the use of amyloid biomarkers such as low CSF abeta42 or an abnormal amyloid PET scan to select individuals with underlying AD pathology in brain but no demonstrable cognitive impairment. It remains to be proven that this represents an early stage of AD (meaning, clinical manifestations will emerge with time assuming sufficient longevity); much more data on the long-term course of such individuals is required. But some supporting evidence has been accruing. For example, among asymptomatic older individuals, an abnormal amyloid biomarker is associated with accelerated atrophy of hippocampus and other brain structures associated with AD pathology. In fact, if we consider that the difference in regional atrophy rates between amyloid positive and amyloid negative asymptomatic individuals may represent "amyloid-mediated atrophy," we can design trials to assess the impact of anti-amyloid and neuroprotective treatments on reduction of this pathological process. Data from ADNI suggests that feasible sample sizes (eg, a few hundred subjects per group) may provide adequate power to demonstrate a 50% reduction in amyloid-mediated atrophy in a two year treatment trial involving asymptomatic older individuals. While the regulatory pathway for preclinical AD remains uncertain, much may be learned by evaluating the impact of putative disease-modifiers in this population.

to cluster analysis by Latent Profile Analysis (LPA) using Latent Gold 3.0 software. Results: Studies on self-assembly of tau revealed that phosphorylation of this protein by different combinations of tau protein kinases could generate bundles of paired helical filaments, suggesting the involvement of different etiopathogenic mechanisms of this pathology. The cluster analysis of mixture of AD and control patients based on their CSF levels of Aß1­42, tau and ubiquitin revealed that there were at least five different subgroups of AD. Each of these subgroups showed a specific clinical profile. Out of these 468 cases, 280 were from Kuopio University Hospital, Kuopio, Finland. All of the Finnish patients had been scored for the presence or absence of the same seven clinical symptoms, namely depression, hallucination, hypokinesia, myoclonus, paranoia, rigidity, and tremor. Each of the five AD subgroups was found to have its own signature set of these clinical symptoms. These clinical features, along with the CSF disease marker profiles, can be employed to define each subgroup of AD. Conclusions: The findings (1) demonstrate the multifactorial nature and involvement of different etiopathogenic mechanisms of AD, (2) show the identification of different subgroups of AD based on CSF biomarkers and clinical symptoms, and (3) suggest the need for stratification of AD patients in different disease subgroups that can considerably improve the outcome of the clinical trials and development of specific and potent disease modifying drugs. (These findings were supported in part by the New York State Office of Mental Retardation and Developmental Disabilities, and grants AG028538 from NIH/NIA, and IIRG-06-025836 from Alzheimer's Association, Chicago, IL, U.S.A.) THE PAD2020 SURVEY OF COGNITIVE INSTRUMENTATION USED IN CLINICAL TRIALS FOR MCI & PRODROMAL AD. P.J. SNYDER (W arren A lpert Medical School of Brown University, USA )

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

A RANDOMIZED, DOUBLE-BLIND, PLACEBO-CONTROLLED DOSE-FINDING TRIAL OF INTRAVENOUS IMMUNOGLOBULIN (OCTAGAM® 10%, OCTAPHARMA AG) IN PATIENTS WITH ALZHEIMER'S DISEASE (GAM10-04). R. DODEL1, F. JESSEN2, M.R. FARLOW3, S. HAAG4, T. MELSEN5, S. WIETEK6 (1. Philips-Universität Marburg, Marburg, Germany; 2. Rheinische-Friedrich-W ilhelmsUniversität Bonn, Bonn, Germany; 3. Indiana University School of Medicine, Indianapolis, IN, USA ; 4. Octapharma A G, Lachen, Switzerland; 5. Octapharma Pharmazeutika Produk tionsges.m .b.H., V ienna, A ustria;6. Octapharm a Pharm az eutik a Produktionsges.m.b.H., V ienna, A ustria) Aim: To describe the design and status of a phase 2 randomized placebo-controlled trial to establish the safety and identify the most efficacious dose and infusion interval for octagam 10% in AD patients. Background: Three small randomized trials (37 patients in total; one controlled with 24 patients and two uncontrolled trials with 5 and 8 patients respectively) suggested that intravenous immunoglobulin (IVIg) may have significant beneficial effects on biomarkers and may improve clinical symptoms in AD patients. The minimal optimal dose and the right treatment interval to maximize IVIg treatment effects while not negatively influencing safety are not known, although there are preliminary data available supporting a 2-week infusion schedule over a 4-week schedule. Methods: The trial design had been reviewed by FDA and the German regulatory authority, PEI (PaulEhrlich-Institute). Trial GAM10-04 has 8 treatment arms. A total of 58 subjects with AD have been included and randomized to receive either one of three monthly doses of octagam® 10%, 0.2 g/kg, 0.5 g/kg or 0.8 g/kg, or placebo infusions every 4 weeks or half of each dose every 2 weeks. For each patient the trial will last 24 weeks. Patients in the 4 arms on 4-week intervals will receive 6 infusions; patients on the 4 arms on 2-week intervals will receive 12 infusions.. The principal outcome is the mean AUC value of the blood plasma change in total A after last infusion with IVIg or placebo from the trough level value prior to last infusion. Secondary outcome measures at week 24 compared to baseline will be A1-42, total A and anti-A autoantibody concentrations and changes in tau and ptau181in CSF; and A1-42, total A and anti-A autoantibodies concentrations in blood plasma; ADAScog, MMSE, CDR-SOB, ADCS-ADL questionnaires; MRI and 18FDG-PET scans. Results: All 56 patients have been randomized to their blinded treatment and received at minimum their first study infusion. The trial will be completed in September 2010. The final report will be available around turn of the year. The trial is registered in the US and Europe on clinicaltrials.gov and controlled-trials.com, respectively, and the results will be published in a peer-reviewed journal and on scientific congresses. Discussion: This trial will generate data on the safety and efficacy of octagam® 10% in AD and will help identifying the optimal IVIg dose and treatment interval for treatment of AD patients. This clinical trial will advance our understanding of A- and tau-mediated neurodegeneration in AD by aiming at using biochemical surrogate markers as primary endpoints in phase III trial(s). Further, GAM10-04 will also provide information on the potentially disease-modifying character of the clinical benefit achieved by treating patients with mild to moderate AD with IVIg. Study supported by: Octapharma AG

individual change. 3) Assessment delivery. The technology is available to collect the finegrained data needed to capture potential therapeutic effects. Techniques to conveniently and reliably administer testing outside of the clinic or laboratory must be perfected. These approaches will improve the feasibility of the proposed assessment approaches and might increase overall compliance. While a great deal of effort has been devoted to identifying potential therapeutic targets and to drug development, cognitive instrumentation in clinical trials has lagged behind. The poor sensitively and precision of available instruments has been a source of frustration, prompting some to search for alternate outcomes such as neuroimaging or biomarkers. In the end, however, a key aim of clinical trials in dementia is to improve or protect cognition. It is therefore that imperative that a new generation of cognitive instruments for clinical trials be developed. Fortunately, the building blocks for these instruments are already in hand. All we need are studies to implement and validate these approaches.

The new generation of clinical trials to prevent the onset of AD will require instruments that can reliably detect subtle cognitive changes over time. Optimally these performance changes should be associated with progression of the underlying AD pathology. Most extant instruments cannot meet these criteria, particularly in the very early or clinically asymptomatic stages of disease. Part of the problem has been the reliance on simple global performance measures, or on measures that have been designed for diagnosis but not for ascertaining change in performance over time. I will review three inter-related considerations for developing the instruments we need: task selection, measurement design, and assessment delivery. 1) Task selection. We need simple, short, reliable, repeatable tasks that are sensitive to AD pathology. Short tasks allow us to utilize appropriate psychometric approaches to detecting performance change. Cognitive studies have already identified several tasks that may be sensitive to underlying pathology. Clearly, well-designed memory tasks, for example the one card recognition task, should have that property. Two other examples of potential tasks that appear to be sensitive to AD pathology are drawn from a rich cognitive neuroscience literature: attentional tasks that demand integration of sensory modalities via cortico-cortico connections, and reaction times tasks that use discordant cues. A further step builds on this straightforward approach and could be particularly useful for very long-term studies: embed one or more tasks into engaging, more complex games. These complex games could allow the participant to maximize performance over a period of time; subsequent testing would then monitor maintenance of performance. 2) Measurement design. The design and implementation of our measures must maximize our ability to detect change over time. One desirable quality of short repeatable tasks is that they allow the quantification of within-subject variance in performance at each measurement session. For long-term studies, each individual's performance trajectory can be captured prior to initiation of treatment of over time, or group data can be used to develop indicators of reliable change. Also, implementing a measurement burst design can improve detection of long-term intra-

HOW, WHEN AND WHY TO DEVELOP A NEW GENERATION OF COGNITIVE INSTRUMENTS FOR CLINICAL TRIALS. Y. STERN (Columbia University, USA)

A plethora of cognitive instruments are commercially available for use in clinical trials with older adults who are at risk for MCI or early AD. In determining the suitability of individual instruments for a specific study, several critically important factors should be considered including (but not limited to) training issues, rater/staff requirements, data security and regulatory compliance, sensitivity to change over time within-subjects, convergent and divergent validity, practice effects and test-retest reliability, task delivery options and more. In order to identify a test or battery that best addresses these factors, a systematic and thorough review of public and proprietary cognitive batteries is required. Such information - in its entirety - is not frequently published in peer-reviewed journal articles, and therefore necessitates a more comprehensive investigation of these measures using information provided directly by the test authors or publishers. The Campaign to Prevent AD by 2020 (PAD2020), an independent not-for-profit organization, has sponsored an inclusive and systematic review of roughly 28 cognitive assessment tests and batteries for use with older adults at risk for dementia. We have been compiling a massive amount of information, provided directly from the vendors of these batteries, and the specific issues and questions, addressed in this review, will be described in this talk. All information received from test vendors/authors have passed through a quality control review process in order to ensure the veracity of all responses to this survey. These data are being compiled into a review of the current state of neuropsychological assessment batteries focused on aging and dementia (with a specific focus on the utility of these instruments for tracking cognitive change in the earliest ("pre-symptomatic") stages of the disease), with the final review to be published in Alzheimer's & Dementia: The Journal of the Alzheimer's Association. Concurrent with this publication, all supporting documents and materials will be uploaded and archived by PAD2020 and these materials will be available for free download from the www.pad2020.org website.

S4

Aim: Lithium, a first line drug for the treatment of bipolar depression, has recently been shown to regulate glycogen synthase kinase 3 (GSK-3), a kinase that is involved in the phosphorylation of the tau protein. Since hyperphosphorylation of tau is a core pathological feature in Alzheimer's disease (AD), lithium-induced inhibition of GSK-3 activity may have therapeutic effects in AD. In the current study, we tested the effect of short-term lithium treatment in patients with AD. Method: A total of 71 patients with mild AD (MMSE 21 and 26) were successfully randomized towards placebo (N = 38) or lithium treatment (N = 33) at six academic expert memory clinics. The 10-week treatment included a six-week titration phase to reach the target serum level of lithium (0.5 ­ 0.8 mmol/L). The primary outcome measures were cerebrospinal fluid (CSF) concentrations of total tau protein, tau protein phosphorylated at the epitope serine 181 (p-tau181) or at threonine 231 (p-tau231), as well as the activity of GSK-3 (including both alpha and beta forms) measured indirectly by the pCREB/total CREB ratio in lymphocytes. Results: No treatment effect on GSK-3 activity or CSF-based biomarker concentrations (p > 0.05) was observed. Lithium treatment did not lead to change in global cognitive performance as measured by the ADAS-cog subtest (p = 0.11) or in depressive symptoms (p = 0.82). Discussion: The current results do not support the notion that lithium may lead to reduced hyperphosphorylation of tau protein after a short 10-week treatment in the AD target

LITHIUM TRIAL IN ALZHEIMER'S DISEASE: A RANDOMIZED, SINGLEBLINDED, PLACEBO-CONTROLLED MULTICENTRE 10-WEEK STUDY. H. HAMPEL (Department of Psychiatry, Psychosomatic Medicine & Psychotherapy, Goethe University, Frankfurt, Germany)

population. Interpretation based on study data analysis and potential weaknesses of studydesign and follow-up studies will be discussed. 1) Hampel H. et al. Lithium trial in Alzheimer's disease: a randomized, single-blind, placebo-controlled, multicenter 10-week study. Journal of Clinical Psychiatry, 2009, 70(6):922-31. 2) Hampel H. et al. Total and phosphorylated tau protein as biological markers of Alzheimer's Disease, Experimental Neurology, 2010, Jan;45(1):30-40. Epub 2009 Oct 22. 3) Hampel H. et al. Biomarkers for diagnosis and as endpoints for clinical trials in Alzheimer's disease. Nature Reviews Drug Discovery, 2010, accepted. 4) Blennow K., Hampel H. et al. Cerebrospinal fluid and plasma biomarkers in Alzheimer's disease. Nature Reviews Neurology, 2010, Mar;6(3):131-44. Epub 2010 Feb 16.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Several authoritative groups have published consensus guidelines for the care of patients with Alzheimer disease and suggested regular follow-up with evaluation and management of behavioural disturbances, psychoses and depression, active monitoring and support of the caregiver's emotional and physical health and consideration of treatment with specific anti-dementia drugs. They are mostly based on the scientific literature. They also include some empirical guidelines, for example concerning the frequency of follow-up assessments. Our objective was to test the effectiveness of a comprehensive specific care plan in decreasing the rate of functional decline in patients with mild-to-moderate Alzheimer disease compared with usual care in memory clinics. Our study is a cluster randomised 24-month trial with memory clinics as the randomised unit (2 arms: intervention group and usual care-control group). Patients included in each centre had a Mini Mental State Examination score between 12 and 26. The intervention includes a comprehensive standardised twice-yearly consultation for patients and their caregivers, with standardised guidelines for the management of problems identified during the assessment. The primary outcome measure was change on the Alzheimer Disease Cooperative Study-Activities of Daily Living (ADCS-ADL) scale assessed at 12 and 24 months. 50 memory clinics were randomised giving a total of 1131 patients (intervention group: 26 clinics, 574 patients; usual care-control group: 24 clinics, 557 patients). Fiftyeight percent of patients completed the two-year assessment in the intervention group and 62% in the usual care-control group. The annual rate of ADCS-ADL change was -5.73 in the intervention group and -5.96 in usual care-control group (p=0.78). There was no difference in the risk of institutionalization (p=.79) or mortality (p=.32) between groups. Interpretation: The results of the primary intention to treat analysis of this trial showed no

EFFECTIVENESS OF A SPECIFIC CARE PLAN IN PATIENTS WITH ALZHEIMER DISEASE (PLASA STUDY). F. NOURHASHEMI 1,2,3 , S. ANDRIEU1,2,3,4, S. GILLETTE-GUYONNET1,2,3, B. GIRAUDEAU5, C. CANTET2,3, B. VELLAS1,2,3 AND THE PLASA GROUP (1. Gérontopôle, CMRR, F-31059 Toulouse, France; 2. Inserm U558, F-31073 Toulouse, France; 3. University of Toulouse III, Toulouse, F-31073, France; 4. Toulouse University Hospital, Department of Epidemiology and Public Health, Toulouse, F-31073, France; 5- Inserm CIC-202, France; François Rabelais University, Tours, France; CHRU de Tours, France)

Tarenflurbil is a gamma secretase modulator that was shown in animal models of AD to reduce Beta amyloid production and therefore to be a possible disease modifying treatment for the disease. A Phase 2 study was followed by two negative phase 3 studies. Why might this be? It is possible that the outcome in the Phase 2 study was mis- or over interpreted. A per protocol analysis of mild and moderately affected participants was undertaken, indicating that the mild group showed some limited signs of efficacy with the higher of two doses, whilst the moderate group deteriorated. This study was one of the first in AD trials to use analysis of slopes as the primary outcome and it is possible that symptomatic effects rather than disease modification may have affected the slopes, rather than a disease modifying process. The ITT analysis in both mild and moderate subjects demonstrated no significant benefit on the primary outcome measures and it was only the additional analysis by subgroup that identified benefit. One would have expected trends to have moved in the same direction for the mild and the moderate sub groups, but this did not happen and is counter intuitive. This may have been a hint that perhaps things were not as they seemed. Although animal studies demonstrated less amyloid deposition and associated pathology in the brain after short-term treatment with Tarenflurbil, studies in human control subjects demonstrated only a dose dependent blood-brain penetration but no change in CSF or plasma levels of A42, with a non-significant trend for a greater reduction in the CSF markers in the placebo group. This raises a question about the validity of transgenic mouse models as an indicator of what might happen in the human brain. And also whether the dose levels were appropriate. Was the Phase 2 study outcome spurious? Only drugs that have a positive outcome at Phase 2 level stimulate Phase 3 trials and some Phase 2 studies are positive by chance, in which case the larger Phase 3 study may produce a result that is closer to the truth, i.e. a negative outcome. In the larger of the two 18 month phase 3 studies the trial protocol was modified early on, based on the Phase 2 data, removing the low dose of tarenflurbil dose and restricting the participants to mild subjects only (MMSE 20-26). This reflects current thoughts that anti-amyloid strategies may have to be given early in the course of the disease. But perhaps it wasn't early enough? Did Tarenflurbil reach appropriate levels within the brain? Although plasma concentrations indicate that it was well absorbed only a small proportion (0.5% - 1%) appears in the cerebrospinal fluid. This may not be adequate for the larger human, compared to mouse, brain. A subset of subjects in the Phase 2 study should have undergone CSF collection to establish how much drug had crossed the blood brain barrier, and its impact on CSF biomarkers, allowing dose adjustment. Finally, is targeting amyloid deposition in the Alzheimer brain the way ahead?

WHAT HAVE WE LEARNT FROM THE MYRIAD TRIALS OF TARENFLURBIL? G. WILCOCK (University of Bristol, United Kingdom)

Aim: There are many clinical contenders for predictors of Alzheimer's disease (AD) costs, such as cognition, functional, and behavioural statuses. However, the relative importance of one clinical indicator over another is uncertain. Patient characteristics, such as age, gender, and co-morbidities also influence costs, and need to be controlled for in any evaluation of clinical contenders. Finally, few studies explore AD cost drivers in longitudinal settings. Objective: We focused on the impact of cognitive, functional, behavioural impairments, and caregivers' burden on AD costs using longitudinal data. Our objective was to explore their ability to explain variations in costs over a 2-year period. Methods: We used a large longitudinal dataset of 1,131 community-dwelling patients with mild to moderate AD in France. We used a validated instrument for the assessment of informal and formal costs, and we measured AD severity using measures of cognition, function, and behaviour. We used random-effects models to explore changes in cost over time. Results: Overtime, the mean total cost of AD per patient was 3,166.75 per month. Informal cost was the most important cost component as it represented more than 80% of total cost. Controlling for functional assessment, behavioural impairment, and informal caregivers' burden, a marginal decrease in MMSE lead to an increase in +5.20% of informal costs, +1.48% of non-medical costs, +0.39% of medical costs, and +4.26% of total costs. A marginal loss of functional status, measured by Katz's activities of daily living scale contributed to large increases in informal costs (+27.64 %), non-medical costs (+55.08%), medical costs (+32.92%), and total costs (+31.07%). Conclusion: The impacts of cognitive and functional declines on costs were evident in our results, even if their influence varied across cost components. The impact of cognitive decline on costs was reduced and became non-significant when we controlled for behavioural disturbances and function scales. This research was supported by a research grant from Janssen Alzheimer Immunotherapy and Wyeth (which was acquired by Pfizer in October 2009).

FACTORS INFLUENCING LONGITUDINAL COSTS IN MILD TO MODERATE ALZHEIMER DISEASE: RESULTS FROM THE PLASA STUDY. T. RAPP, S. ANDRIEU, B. VELLAS (Gérontopôle de Toulouse et INSERM U558. Institut du vieillissement, Faculté de Médecine, 37, allées Jules Guesde, 31073 Toulouse Cedex, France)

difference in the rate of functional decline measured by the ADCS-ADL scale in the intervention group in comparison with the usual-care control group in mild to moderate Alzheimer disease patients. To our knowledge, this is the first nationwide randomised clinical trial testing the effectiveness of guideline-based care interventions for Alzheimer disease delivered in memory clinics. Alzheimer's disease is a complex and heterogeneous disease. Consequently, in order to have a beneficial effect on disease progression, either it may be that interventions must be targeted towards patients at particular risk of decline, or we may need to develop a more effective intervention, and ensure that it is correctly implemented in all patients. Future research is needed to determine whether the effects obtained herein can be improved with more direct involvement of the GP. Trial registration: clinicaltrials.gov; Identifier: NCT00480220

MABT5102A IS AN EFFECTOR-REDUCED ANTI-A ANTIBODY WITH UNIQUE BINDING PROPERTIES THAT PROMOTES NEUROPROTECTION AND GLIAL ENGULFMENT OF A. O. ADOLFSSON 1 , M. PIHLGREN 1 , Y. VARISCO1, A.-L. BUCARRELLO1, N. TONI1, K. ANTONIELLO1, S. LOHMANN1, K. PIORKOWSKA 1 , V. GAFNER 1 , J. ATWAL 2 , R. WEIMER 2 , A.L. BOXER 3 , D.L. MORTENSEN2, J. LI2, K. PHAM2, M. FRIESENHAHN2, C. HO2, R. PAUL2, A. PFEIFER1, A. MUHS1, R.J. WATTS2 (1. A C Immune SA , Lausanne, Switzerland; 2. Genentech Inc., South San Francisco, USA ; 3. Memory and A ging Center, University of California, San Francisco, CA , USA ) Passive immunization against beta-amyloid (A) has become an increasingly desirable strategy as a therapeutic treatment for Alzheimer's disease (AD). The effectiveness of passive immunization has been demonstrated in transgenic animal models of AD, where anti-A therapies have been shown to reduce plaque burden and reverse behavioral deficits. In spite of overcoming hyper-activation of cytotoxic T-cells, a risk of active immunization with A, passive immunization still carries the risk of Fc receptor (FcR)mediated over-activation of microglia cells and complement activation, which may contribute to an inappropriate pro-inflammatory response and vasogenic edema. We describe the generation of a humanized anti-A monoclonal antibody (mAb) of an IgG4 isotype, known as MABT5102A (MABT). An IgG4 isotype was selected to reduce the risk of Fc receptor-mediated over-activation of microglia and to avoid complement activation. MABT bound with high affinity to multiple forms of A1-42 and A1-40, protected

The societal and individual costs of Alzheimer's disease are significant, worldwide. As the world ages, these costs are increasing rapidly, while health systems face finite budgets. As a result, many regulators and payers will require phase III cost-effectiveness data (in addition to safety and efficacy data) to consider drug approval and reimbursement, increasing the risks and costs of drug development. Incorporating pharmacoeconomic studies in phase III clinical trials for Alzheimer's disease presents a number of challenges. In this presentation, we propose and discuss several specific ways to improve the design of pharmacoeconomic studies in phase III clinical trials. In summary, since we are entering an era in which pharmacoeconomic studies will be essential in drug development, we must consider the design and incorporation of pharmacoeconomic studies in phase III clinical trials more seriously and more creatively.

NOVEL APPROACHES TO INCORPORATING PHARMACOECONOMIC STUDIES INTO PHASE III CLINICAL TRIALS FOR ALZHEIMER'S DISEASE. H. FILLIT (A DDF, USA )

S5

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

against A1-42 oligomer-induced cytotoxicity, mediated uptake of neurotoxic A by microglia both in vitro and in vivo. When compared to a human IgG1 wild-type subclass, containing the same antigen-binding variable domains and with equal binding avidity to A, MABT showed reduced activation of the stress-activated p38 mitogen-activated protein kinase (p38MAPK) in microglia, and induced less release of pro-inflammatory mediators. We propose that a humanized IgG4 anti-A antibody that takes advantage of both an effective neutralizing capability based on a select A binding profile, while also possessing reduced effector function, may provide a safer therapeutic alternative for passive immunotherapy for AD. Supported by preclinical data, a single and multiple dose, multicenter, randomized, placebo-controlled, double-blind Phase I study to assess safety, pharmacokinetics, and pharmacodynamics was conducted in patients with mild to moderate Alzheimer's disease. Data from this clinical study will also be presented. O1 - DOMINANTLY INHERITED ALZHEIMER'S NETWORK TREATMENT TRIALS INITIATIVE AT CTAD. R. BATEMAN

ORAL COMMUNICATIONS

Introduction: The main objective of this study is to study the effect of Donepezil on the rate of atrophy of the hippocampus, a MRI marker of the disease progression, in a sample of patients with amnestic disorders evoking the prodromal stage of AD. Methods and material: In this double-blind, randomized, placebo-controlled, two-parallel groups study, the objective was to enroll 240 patients that received 52 weeks of treatment with donepezil (10mg/d) or placebo. To be eligible, patients must meet all following criteria: (1) amnestic syndrome of the "hippocampal type" measured by the Free and Cued Selective Reminding Test (free recall < or =17 or total recall <40 ); (2) preserved activities of daily living; and (3) global clinical dementia rating stage of 0.5. Subjects were required to have structural brain MRI scan prior to baseline visit validated by a central reading structure. A second MRI was realized at the final visit (12 months) or in case of premature withdrawal after 6 months. The primary outcome measure was the change from baseline in hippocampal volumetry (left+right) measured by MRI, using a semi-automatic segmentation technique. Secondary outcome of interests included evolution of neuropsychological performances and change in other brain MRI markers. Results: 28 French clinical sites and 24 MRI centres were involved in the study; 337 patients have been selected and 216 randomised on August 31st 2009. A full description of the sample for baseline neuroimaging and clinical characteristics will be provided. Results on cross-sectional associations between neuroimaging and neuropsychological tests performances will also be presented. Conclusion: This study was one of the first to rely on the amnestic syndrome of hippocampal type for inclusion criteria and on hippocampal volume for primary efficacy criteria and it was aimed at evaluating the impact of Donepezil on MRI markers at early stage of AD disease.

O2 - HIPPOCAMPUS COLLABORATIVE STUDY: DESIGN AND FIRST RESULTS. B. DUBOIS, M. SARAZIN, S. LEHERICY, M. CHUPIN, I. TONELLI, L. GARNERO, D. DORMONT FOR THE "DONEPEZIL HIPPOCAMPUS STUDY GROUP "

The Dominantly Inherited Alzheimer's Network (DIAN) has been established as a cooperative agreement by the National Institute on Aging (NIA) to enable international longitudinal studies of persons with dominantly inherited mutations that cause AD. DIAN aims to determine the chronological changes in cognition and biomarkers in relation to clinical onset and progression of dementia, in a well characterized and uniformly studied group of autosomal dominant AD individuals. This is the first large scale, multicenter, systematic effort to identify and uniformly evaluate individuals with dominantly inherited AD with standardized instruments. Many proposed therapeutics for Alzheimer's disease currently target slowing or halting the underlying disease (disease modifying), but are not likely to reverse the extensive neuronal death present at advanced stages of disease. For individuals and families at risk for the autosomal dominantly inherited form of Alzheimer's disease, potential disease modifying therapies offer the opportunity to delay or even prevent dementia in asymptomatic individuals, in addition to treating those with symptoms. An additional scientific aim for DIAN is the evaluation of potential diseasemodifying compounds for the treatment of AD. To this end, DIAN recently formed a Clinical Trials Committee to direct the design and management of interventional therapeutic trials of DIAN participants. The DIAN Clinical Trials Committee goal is to implement the safest trials that have the highest likelihood of success while providing advancement of treatments, scientific understanding and clinical effects of proposed therapies. The DIAN Clinical Trials Committee scientific and medical aims include: 1) Determine the timing of treatment important for improved clinical outcomes. 2) Determine changes in physiologic or pathologic biomarkers that can be used to track therapeutic effectiveness of treatments. 3) Determine which therapeutic targets are most amenable to treatment at different stages of Alzheimer's disease. 4) Test the hypotheses of the causes of Alzheimer's disease (e.g. amyloid hypothesis) through therapeutic treatment trials. A review of the development of clinical trials in the DIAN will be presented.

O3 - DO UNDETECTED ERRORS CAUSE SOME AD DRUG DEVELOPMENTS TO FAIL? R.E. BECKER, N.H. GREIG (Drug Design & Development Section, NIA , Baltimore, MD 21224, USA )

Aim: Recent studies of Alzheimer's disease (AD) and other neuropsychiatric drug developments raise questions whether failures in clinical trials (CT) of some drugs occur due to flaws in methods. We seek to identify these sources of errors, their impact on drug developments and clinical trials, and routes to overcoming error interference with the

S6

Cerebrospinal fluid (CSF) biomarkers are now widely used for diagnosis of Alzheimer Disease (AD) in atypical clinical forms, for differential and early diagnosis, or for stratification of patients in clinical trials. Among these biomarkers, different forms of amyloid peptides (A) produced by the cleavage of a transmembrane precursor protein called APP (amyloid precursor protein) have a major role. A peptides exist in different length the most common ones having 40 (A40), 42 (A42) or 38 (A38) amino acids in length. APP processing by -secretase releases also an aminoterminal secreted fragment called sAPP while an alternative non-amyloidogenic cleavage of APP, through a secretase, liberates an other fragment called sAPP. To decipher the molecular and pathological mechanisms leading to the production and the detection of these entities is essential for the comprehension and the prevention of AD. In this report, we present the results of the multiplex measurement of CSF A38, A40, A42, sAPP and sAPP in 60 patients with dementia segregated between neurochemical dementia diagnostic (NDD)

O4 - CORRELATIONS BETWEEN SOLUBLE / FORMS OF AMYLOID PRECURSOR PROTEIN AND A38, 40 AND 42 IN HUMAN CEREBROSPINAL FLUID. A. GABELLE1,2,3, S. ROCHE4, A.C. GENY1, K. BENNYS1, P. LABAUGE5, Y. THOLANCE 6 , I. QUADRIO 6 , L. TIERS 4 , B. GOR 4 , C. CHAULET 1 , A. VIGHETTO7,10, B. CROISILE8,10, P. KROLAK-SALMON9,10, J. TOUCHON1,2,11, A. PERRET-LIAUDET6, S. LEHMANN2,3,4 (1. Centre Mémoire Recherche Ressources Montpellier, Hôpital Gui de Chauliac, CHU de Montpellier, 34025 Montpellier cedex 5; 2. Université Montpellier1, Faculté de Médecine, rue de l'Ecole de Médecine, 34000 Montpellier, France; 3. UPR 1142, CNRS, Institut de Génétique Humaine, 141 rue de la Cardonille, 34396 Montpellier Cedex 5, France; 4. Laboratoire de Biochimie Protéomique Clinique, Institut de Recherche en Biothérapie, Centre Hospitalier Universitaire Saint Eloi, 80 A venue A ugustin Fliche, 34295 Montpellier Cedex 5, France; 5. Service de Neurologie, Centre Hospitalier Universitaire de Nîmes, Hôpital Caremeau, Place du Professeur Robert Debré, 30029 Nîmes Cedex 9, France; 6. Service de Neurobiologie; Biochimie et Biologie moléculaire; 7. Service de Neurologie; Unité 402; 8. Service de Neuropsychologie; Groupement Hospitalier Est, Hospices Civils de Lyon / Université Claude Bernard Lyon 1, 69677 Bron cedex, France; 9. Médecine Gériatrique, Hôpital des Charpennes, 27, Hospices Civils de Lyon, 69100 V illeurbanne; 10. Centre Mémoire Recherche Ressources Lyon, Hospices Civils de Lyon, 3 quai des Célestins, 69002 Lyon; 11. INSERM U888, Hôpital de la Colombière, 34000 Montpellier, France)

effectiveness of AD CTs. Background: Two recent reviews of AD drug developments found over 100 (1) and 172 (2) failed trials giving a current estimated current success rate in AD drug developments of 3%. In a survey of 40 of these AD and MCI drug developments we found investigators giving little attention to 56 potential error sources able to potentially account for these or other failures (1). In three case studies of recent AD drug development failures we identified methodological lapses able to account for the failures of these investigations (3). Further study of the 56 potential errors and of the errors associated with failures in the three AD case studies showed the clear majority of them to be undetectable at the times of occurrence and of interference with the success of the CT or development (4). In this feature error risks to AD drug developments, as aspects of working with complex systems, resemble error risks already well characterized in the nuclear and aviation industries: almost inescapable due to the inevitability of human mistakes and most frequently hidden at the time of occurrence and thereafter until a mishap suggests an error source (5). In response we proposed the need for preemptive error management as a preventive strategy to exclude or control error intrusions into AD drug developments and the development of checklists to facilitate this preemption (3). To further characterize the potential effectiveness of checklists, to understand the relations of checklists to milestones in AD drug developments, to evaluate the potential effectiveness and limitations to be expected from these checklists, and to pilot routes to further studies of errors and interventions against errors we conducted studies of six drug developments or CTs already characterized as questionable because of evidence that errors, not drug efficacy, had accounted for the observed outcomes. Methods: We studied six problematic drug development or clinical trial reports to determine the nature of errors, their effects, and the presence or absence of core scientific practices. Results: Error intrusions into the examples studied were found undetectable at the time of occurrence and indicating the need for preemptive interventions to exclude or control most error intrusions into AD drug developments. Core scientific practice standards used in other areas of science and medicine to address error problems were not found present in the neuropsychiatric drug developments we studied. Discussion: A preemptive error strategy, by introducing methodological interventions able to assure the quality of practices and data, adds to regulatory technologies that already structure AD drug developments core scientific practices better able to insure the validity of data. We discuss how wider introductions of scientific practices make more practical and effective checklist reviews of planning and executions at milestones of a drug development, add credibility and confidence to outcomes realized by AD drug developments and CTs, and more effectively implement the aims of translational medicine to realize clinical benefits from new discoveries. 1. Becker RE, Greig NH, Giacobini E. W hy Do So Many Drugs for A lzheimer's Disease Fail in Development ? Time for New Methods and New Practices? J A lzheimers Dis. 15: 303-325 (2008). 2. L indner M D, M cA rthur R A , Deadw y ler S A , Ham pson R E, T ariot PN . Development, Optimization and Use of Preclinical Behavioral Models to Maximize the Productivity of Drug Discovery for A lzheimer's Disease. In: (Ed.) McA rthur RA , Borsini F. A nimal and Translational Models for CNS Drug Discovery: Neurologic Disorders. San Diego: A cademic Press (2008) p. 93-157. 3. Becker RE, Greig NH. W hy So Few Drugs for A lzheimer's Disease? A re Methods Failing Drugs? Curr A lz Research (In press 2010). 4. Becker RE, Greig NH. Neuropsychiatric Clinical Trials: Should They A ccommodate to `Real-W orld' Practices or Set Standards for Clinical Practices? J Clin Psychopharm. 29: 56-64 (2009). 5. Reason J. Human Error. Cambridge: Cambridge University Press (1990).

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

positive and negative groups. The NDD classification was based on our routine Tau, Ptau181 and A42 cut off values. We confirmed previous findings regarding the correlation between sAPP and sAPP, as well as the potential interest of these new biomarkers. We also studied the correlation between sAPPs and A peptides, as well as between A peptides themselves. We observed a strong correlation between A38 and sAPP which suggested that the production of this peptide was in direct relation with secretase activities. We also reported a strong correlation between A38 and A40, while A42 was correlated to these fragments only in non pathological situations. These results enlighten the complex relationships between these molecular markers in both physiological and pathological situations. Our results are important for the further use of these analytes for AD diagnosis, as well as, to validate cell biological hypotheses of APP processing and A fragment production.

Aim: It is widely accepted that a neuronal energy crisis, cerebral hypometabolism and vascular hypoperfusion are major and potentially treatable contributors to the loss of function in patients with Alzheimer disease (AD). In addition, recently we have found an unexpected capacity for melanin to dissociate water and aid in the production of energy in the process. We hypothesize that melanin can be used as a new and more effective therapeutic approach in the treatment of AD patients. The E4 isoform of apolipoprotein E (ApoE) is involved in cardiovascular and cerebrovascular disorders and is the most prevalent risk factor for late onset of sporadic AD. Moreover, ApoE4 transgenic (Tg+) mice are appropriate models for studying the pathogenesis and preclinical treatment of ApoE-related cognitive deficits associated with late onset and sporadic AD. The goal of our work is to potentially target melanin's dissociation properties for therapy of the disease by using a combination of a selective mitochondrial antioxidant plus a melanin analogue developed by Dr. Solis (QUIAPI-1) both in combination with our recently developed brain exercise program. Clinical data were compared to our animal study on ApoE4 mice who received selective mitochondrial antioxidants acetyl-L-Carnitine and R- -Lipoic acid (ALCAR+LA). Methods: This study applies the brain hypometabolism paradigm to AD patients in order to analyze cognitive function in patients who receive ALCAR+LA, Omega-3-6-9 Fish, Flax, Borage oil as well as Coenzyme Q-10 and QUIAPI-1 , along with diet changes in combination with our recently developed brain activation program (a home-based protocol involving mild physical exercise and cognitive training). The average

O7 - OXIDATIVE STRESS-INDUCED BRAIN HYPOMETABOLISM UNDERLIES THE PATHOPHYSIOLOGY OF ALZHEIMER DISEASE AND OFFERS NEW AND SUCCESSFUL TARGETS FOR TREATMENT: ASTONISHING THERAPEUTICS EFFECTS FROM A COMBINATION OF A MELANIN PRECURSOR (QIAPI-1) AND SELECTIVE MITOCHONDRIAL ANTIOXIDANTS (ALCAR+LA): NEW SCENT ON THE TRAIL? G. ALIEV 1,2,3, H.H. PALACIOS 4, J. LESZEK4, M.E. OBRENOVICH5, V. BRAGIN2, A. SOLÍS HERRERA6 (1. University of A tlanta, A tlanta Georgia, US A ; 2. S tress R elief Center, B rook ly n, N Y , US A ; 3. Department of Biology, University of Texas at San A ntonio, San A ntonio, Texas, USA ; 4. Departm ent of Psy chiatry , W roclaw M edical Univ ersity , W roclaw , Poland; 5. Department of Pathology, School of Medicine, Case W estern Reserve University, Clev eland, Ohio, US A ; 6. Centro de Estudios de la Fotosíntesis Hum ana. S .C., A guascalientes, 20000, México)

Aim: In the absence of validated surrogate biomarker endpoints, clinical measures continue to be used as primary endpoints for disease-modifying trials for Alzheimer's disease (AD). Currently, two co-primary endpoints must be specified, which measure cognitive and functional impairment. Generally, the ADAS-Cog is one of the co-primary endpoints, but high variability in this measure leads to large sample sizes. We evaluated the psychometric properties of the Clinical Dementia Rating-Sum of Boxes (CDR-SB) in order to assess its suitability as a single primary endpoint as an alternative to the traditional co-primary approach. Method: Subjects were 667 very mild to moderate (CDR 0.5-2) AD patients from the REAL.FR study. The CDR-SB was assessed in terms of internal consistency, structural and convergent validity, and 2-year internal and external responsiveness. Data quality and sample sizes were also evaluated. Results: The CDR-SB showed good internal consistency (Cronbach's alpha=0.88), and acceptable structural (separate "cognitive" and "functional" factors) and convergent validity. Variability in mean changes over time was low, leading to excellent internal responsiveness (effect size=1.2; standardized response mean=1.17 at 2 years) and smaller sample sizes compared to the ADAS-Cog. External responsiveness was acceptable when compared to "clinically meaningful" changes on the ADL scale but only borderline acceptable when compared to the ADAS-Cog and IADL. Data quality was high for the CDR-SB and there were few floor or ceiling effects. Discussion: The CDR-SB measures cognitive and functional impairment simultaneously, and has excellent 2-year internal responsiveness. The rate of change on the CDR-SB was similar in both very mild and mild-moderate stages of AD. This makes it a promising candidate as a sole primary endpoint for AD trials, although more work is required to determine the clinical relevance of CDR-SB changes, and its usefulness as an endpoint at other stages of the disease.

O5 - SUITABILITY OF THE CLINICAL DEMENTIA RATING-SUM OF BOXES AS A SINGLE PRIMARY ENDPOINT FOR ALZHEIMER'S DISEASE TRIALS. N. COLEY 1 , S. ANDRIEU 1,2 , M. JAROS 3 , M. WEINER 4 , J. CEDARBAUM 3 , B. VELLAS1,5 (1. INSERM U558 ­ University of Toulouse III, F-31073, France; 2. CHU Toulouse, Department of Epidemiology and Public Health, Toulouse, France; 3. Elan Pharmaceuticals, South San Francisco, CA , USA ; 4. University of California, San Francisco; Department of V eterans A ffairs Medical Center, San Francisco, CA , USA ; 5. CHU Toulouse, Department of Geriatric Medicine, A lzheimer Centre, Gerontopole, Toulouse, France)

S7

Aim: To test the ability of long-term supplementation with vitamin D (2000 IU/day) or EPA/DHA (500 mg of each/day) to reduce rates of cognitive decline in older Caucasian and African-American men and women. In particular, we will test the interaction between race and vitamin D supplementation, with the hypothesis that supplementation will be

O9 - A LARGE-SCALE RANDOMIZED CLINICAL TRIAL OF VITAMIN D AND EPA/DHA FOR PREVENTION OF COGNITIVE DECLINE IN A MULTI-ETHNIC POPULATION: THE VITAL-COG TRIAL F. GRODSTEIN, J.E. MANSON, J.H. KANG (Department of Medicine, Brigham and W omen's Hospital, Harvard Medical School, Boston, MA , USA )

Aim: To estimate the enrichment of MCI patients with true AD cases by means of hippocampal atrophy on MRI, temporoparietal hypometabolism, CSF biomarkers, and cortical amyloid deposition. Method: Four hundred five MCI patients (143 converters and 262 nonconverters to AD within 2 years) of the Alzheimer's disease Neuroimaging Initative (ADNI) were used. Markers for enrichment were hippocampal atrophy on magnetic resonance (M0R), temporoparietal hypometabolism on FDG PET, cerebrospinal fluid (CSF) biomarkers (Abeta42, tau, and phospho-tau), and cortical amyloid deposition (11C­PIB positron emission tomography (PET)). Two separate enrichment strategies were tested to A) maximize the proportion of MCI converters screened in, and B) minimize the proportion of MCI converters screened out. Results: Based on strategy A, when compared with no enrichment and ADAS-Cog as an outcome measure (sample size of 834), enrichment with 18F-FDG PET and hippocampal volume lowered samples size to 260 and 277 cases per arm, but at the cost of screening out 1,597 and 434 cases per arm. When compared with no enrichment and clinical dementia rating (CDR-SOB) as an outcome measure (sample size of 674), enrichment with hippocampal volume and Abeta42 lowered sample sizes to 191 and 291 cases per arm, with 639 and 157 screened out cases. Strategy B reduced the number of screened out cases (740 for [11C]-PIB PET, 101 hippocampal volume, 82 ADAS-COG and 330 for [18F]-FDG PET) but at the expense of decreased power and a relative increase size (740 for [11C]-PIB PET, 676 for hippocampal volume, 744 for ADAS-Cog and 517 for [18F]-FDG PET). Discussion: Groups of MCI patients enrolled in clinical trials of AD drugs can be enriched with true AD patients through screening with imaging markers. Enrichment comes at the price of an often relevant proportion of screened out cases, and in clinical trial settings, the balance between enrichment of screened in and loss of screened out patients should be critically discussed.

O8 - ENRICHMENT THROUGH BIOMARKERS IN CLINICAL TRIALS OF ALZHEIMER'S DRUGS IN PATIENTS WITH MILD COGNITIVE IMPAIRMENT M. LORENZI 1 , M. DONOHUE 2 , D. PATERNICO 1 , C. SCARPAZZA 1 , S. OSTROWITZKI3, O. BLIN4, E. IRVING5, G.B. FRISONI1,6, THE ALZHEIMER'S DISEASE NEUROIMAGING INITIATIVE (1. LENITEM Laboratory of Epidemiology, Neuroimaging and Telemedicine, IRCCS San Giovanni di Dio-FBF, Brescia, Italy; 2. Division of Biostatistics and Bioinformatics, University of California, San Diego, USA ; 3. Translational Medicine, Neuroscience, F. Hoffmann-La Roche Ltd, Basel, Switzerland; 4. Clinical Investigation Centre (CIC-UPCET) and Department of Clinical Pharmacology, UMR-CNRS, 6193 Institute of Cognitive Neurosciences, CHU, Timone, Marseille, France; 5. Neurosciences CEDD, GlaxoSmithKline, Harlow, Essex, UK; 6. A FaR A ssociazione Fatebenefratelli per la Ricerca, Rome, Italy)

age of the patients was 73.5. The patients were evaluated at baseline and in one and two years post treatment. In addition, our animal study applies the vascular dementia paradigm to ApoE4 Tg+ mice in order to analyze the effects of the selective mitochondrial antioxidants ALCAR+LA on cerebral blood flow (CBF), neuropathology, brain and vessel ultrastructural abnormalities and behavior. Results: Our clinical results showed that patients who received a combination of mitochondrial antioxidants and QIAPI-1 exhibited the maximum cognitive improvement at the end of 24 months of treatment. The maximum cognitive improvement was seen with the combined treatment in MMSE, attention, memory, naming, construction, clock drawing, verbal fluency, and Ruff Frontal Fluency tests. By the end of 24 months of treatment, significant improvement was observed, especially in attention, construction and clock drawing, when patients received the combined treatment. In addition, this group also showed that in all categories there were no signs of a decrease and/or decline below the base line for the entire period of treatment. ApoE4 associated factors reduced the CBF gradually and created brain hypoperfusion when compared to the wild-type (WT). The differences in CBF were greatest as animals aged from 6 weeks to 12 months. Transmission electron microscopy (TEM) with colloidal gold immunocytochemistry and in situ hybridization using human and mouse DNA probes showed structural damage and mitochondrial DNA overproliferation and/or deletion in the young and aged microvessels endothelium of ApoE4 animals, extending to the cytoplasm of perivascular cells, perivascular nerve terminals, hippocampal neurons, and glial cells. Spatial and temporal memory tests showed a trend in improving cognitive function in ApoE4 Tg+ mice who were fed the selective mitochondrial antioxidants (ALCAR+LA). Conclusions: Our conclusion is that the potential pharmacologic modulation of brain hypometabolism by using human photosynthesis compounds such as selective mitochondrial antioxidants and a melanin analogue, and/or precursor represents a completely new and more effective strategy to treat Alzheimer and/or other dementia types. Moreover, further expanding the examination of the ultrastructural degeneration caused by aging, especially under cardiovascular disease complications, is likely to contribute to our understanding of neurodegenerative etiology and will indicate a new avenue for the development of novel prophylactic and treatment strategies by offering selective mitochondrial antioxidants like ALCAR+LA and QIAPI 1 to the demented patients.

particularly beneficial in African-Americans, who have a high prevalence of vitamin D deficiency. To our knowledge, this will be the first large-scale prevention trial to test an agent with special relevance to African-Americans, who are at higher risk of cognitive impairments than Caucasians. We focus on cognitive decline rather than dementia, to better capture the continuum of cognitive function compared to a binary dementia diagnosis. Method: VITAL (VITamin D and OmegA-3 TriaL) will be initiated in the fall of 2010 to test whether these supplements prevent chronic diseases in men and women over a planned 5 years of treatment/placebo. We are in the process of recruiting 2000 Caucasian and 1000 African-American men and women aged 65 years and older. A 2x2 factorial design will be utilized, and all data will be collected from questionnaires, medical records, and telephone interviews; these simple data collection methods were chosen to maximize efficiency and continued participation, which are critical to valid results in randomized trials. The telephone cognitive assessment will include tests of global function, episodic memory, executive function, verbal fluency, and working memory. Results: Preliminary studies support our proposed telephone methodology. First, we have conducted extensive validation studies of our telephone cognitive battery, establishing its utility in population research. In one validation study, our brief telephone battery correlated highly with a 21-test cognitive assessment administered in-person (r=0.81). In addition, in a sample of men and women who had completed this battery and agreed to autopsy after death, cognitive performance was strongly related to Alzheimer Disease pathology. In secondary analyses, we will also assess dementia diagnosis, using a telephone informant interview (the Dementia Questionnaire). We have found that ratings on the Dementia Questionnaire corresponded highly to dementia diagnosis as determined by experienced neurologists who evaluated older subjects at the Massachusetts Alzheimer Disease Research Center. Finally, in previous clinical trials, we have found that telephone instruments for measuring cognitive health produce over 90% follow-up rates. Discussion: Large-scale, long-term trials are required for assessing agents to preserve cognitive health with aging. We are able to conduct the VITAL-Cog trial of 3000 older men and women by utilizing telephone cognitive testing and telephone informant interviewing, which reduces costs and increases the validity of study findings by maximizing participation and followup. These methods should be carefully considered for additional large-scale prevention trials.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Background: The ability to visually identify and quantify brain -amyloid in individuals with progressive late-life cognitive impairment should increase the accuracy of a clinical diagnosis of Alzheimer's disease. Florbetapir F 18 is an amyloid PET ligand with the potential to achieve that goal because of its high affinity and selectivity for -amyloid. Methods: Florbetapir-PET brain scans were acquired in 152 subjects (mean age 78.1, mean MMSE 21.2) with a variety of end-of-life medical conditions, including 37% with a clinical diagnosis of Alzheimer's disease, and results compared to postmortem measures of -amyloid using immunohistochemistry and Bielchowsky silver staining in the first 35 subjects coming to autopsy. Florbetapir-PET scans were also obtained in 74 cognitively normal non-autopsied young subjects (mean MMSE 29.7) presumed to be free of amyloid pathology. Results: The mean interval from PET scan to death was 89.4 days. Nineteen of 35 autopsy subjects (54.3%) met pathological criteria for AD. Both the visual interpretation and quantitative assessment of the florbetapir-PET images were highly correlated with the presence and quantity of -amyloid pathology found at autopsy (correlation coefficients = 0.70 to 0.78). There was agreement between the florbetapirPET results and the postmortem results in 97% of the autopsied subjects. In addition 100% of the young control subjects were appropriately rated as amyloid negative. Conclusions: Florbetapir-PET accurately measures the presence and density of -amyloid during life. This study provides direct evidence that a molecular imaging procedure available in most communities can identify and quantify Alzheimer's pathology in the brains of individuals during life.

O10 - VALIDATION OF FLORBETAPIR-PET FOR IMAGING ALZHEIMER'S DISEASE AMYLOID PATHOLOGY: A PROSPECTIVE MULTICENTER CLINICO-PATHOLOGIC CORRELATIVE STUDY. D.M. SKOVRONSKY 1,2 , C.M. CLARK1,2, J.A. SCHNEIDER3, B.J. BEDELL4,5, T.G. BEACH6, M.A. MINTUN1,7, M.J. PONTECORVO 1 , F. HEFTI 1 , A.P. CARPENTER 1 , M.L. FLITTER 1 , M.J. KRAUTKRAMER 1 , H.F. KUNG 2 , R.E. COLEMAN, M. DORAISWAMY 8 , A.S. FLEISCHER9,10, M.N. SABBAGH6, C.H. SADOWSKY11, P.E.M. REIMAN9,12 AND THE AV45-A07 STUDY GROUP (1. A vid Radiopharmaceuticals, Philadelphia; 2. University of Pennsylvania School of Medicine, Philadelphia; 3; Rush University Medical Center, Chicago; 4. Biospective, Montreal Quebec; 5. Montreal Neurological Institute, McGill University, Montreal Quebec; 6. Banner Sun Health Research Institute, Phoenix; 7. W ashington University School of Medicine, St. Louis; 8. Duke University Medical Center, Durham; 9. Banner A lzheimer's Institute, Phoenix; 10. University of California, San Diego; 11. Nova SE University, Ft. Lauderdale; 12. A rizona A lzheimer's Consortium, Phoenix and Department of Psychiatry, University of A rizona College of Medicine Phoenix)

O12 - HOURLY VARIABILITY OF CEREBROSPINAL FLUID BIOMARKERS IN ALZHEIMER PATIENTS AND HEALTHY ELDERLY CONTROLS. D. SLATS 1,6,7 , J. CLAASSEN 1,6,7 , P. SPIES 1,6,7 , G. BORM 2 , T. BESSE 3 , W. VAN AALST 1,6 , J. TSENG 8 , M. SJOGREN 1,6, M. OLDE RIKKERT 1,6,7 , M. VERBEEK4,5,6 (Radboud University Nijmegen Medical Centre, 1. Department of Geriatric Medicine; 2. Department of Epidemiology, Biostatistics and HTA ; 3. Department of A nesthesiology; 4. Department of Neurology, 5Department of Laboratory Medicine, Nijmegen, The Netherlands; 6. A lzheimer Centre Nijmegen, Nijmegen, The Netherlands; 7. Donders Institute for Brain, Cognition and Behaviour, Nijmegen, The Netherlands; 8. Merck Research Labs, Kenilworth, NJ, USA ) Aim: Large hour-to-hour variability in the cerebrospinal fluid (CSF) concentrations of Alzheimer's disease (AD) biomarkers amyloid 42 (A42) and A40 has been demonstrated in healthy subjects, with a suggested diurnal pattern. We investigated the within-subject variability of CSF A and tau proteins in patients with AD and older controls, since hourly variability may affect implementation of CSF analysis in clinical practice. Method: Six patients with mild stage AD (59-85 yrs, MMSE 16-26) and six healthy older volunteers (64-77 yrs) received an intrathecal catheter from which each hour during 36 hours 6 ml of CSF was drawn. Concentrations of A42, A40, total tau and phosphorylated tau were determined and variability was analyzed using linear mixed models. Results: The coefficients of variation for the four biomarkers were comparable in AD (3.1 to 5.8%) and controls (4.2 to 4.6%). We did not observe a diurnal pattern for any of the biomarkers. Discussion: Within-subject variability of CSF biomarkers is low and does not differ between AD and older controls. Moreover, CSF levels do not exhibit a diurnal pattern. These results imply that concentrations of CSF biomarkers are not affected by the time of day the lumbar puncture is performed. This, combined with the low variability, validates the diagnostic use of a single, random time-point CSF measurement in clinical practice. Finally, these data bear relevance for the validity of repeated CSF measurements in monitoring disease modifying effects in intervention studies.

Background and aim: Plasma total homocysteine (tHcy) concentrations are inversely associated with cognitive function in healthy elderly and are elevated in Alzheimer's disease. Our objective was to determine whether treatment with vitamins B12, B6 and folic acid to lower homocysteine has an effect on cognitive decline in older people with mild cognitive impairment (MCI). Methods: We conducted a 2 year, placebo-controlled, randomized clinical trial in 266 participants aged 70 years and older with MCI assessed with telephone screening. Treatment was a tablet containing 0.8 mg folic acid, 0.5 mg B12 and 20 mg B6, or placebo. Cognitive performance in the MMSE, Hopkins Verbal Learning Test, category fluency and CLOX was assessed at baseline and at follow-up; final assessment completed by 223 subjects. Generalized linear mixed effects models were fitted to the longitudinal cognitive scores using binomial and Poisson distributions for discrete outcomes and the Gaussian distribution for continuous outcomes. Registration: www.controlled-trials.com ISRCTN 94410159. Results: tHcy levels were reduced from 11.8 (3.4) mol/L at baseline to 8.9 (2.3) mol/L after 2 years in the treatment group. We asked the following questions: Is there a direct effect of treatment on cognitive test scores at outcome? Does treatment interact with the baseline concentration of tHcy? Does treatment influence the proportion of subjects with a CDR rating of 0.5? Discussion: Possible implications of the trial for the prevention of dementia will be considered.

O11 - DOES B VITAMIN TREATMENT SLOW COGNITIVE DECLINE IN MILD COGNITIVE IMPAIRMENT FOR THOSE WITH RAISED PLASMA HOMOCYSTEINE? A RANDOMIZED CONTROLLED TRIAL (VITACOG). C.A. DE JAGER 1 , A. OULHAJ 1 , R. JACOBY 2 , H. REFSUM 3,4 , A.D. SMITH 3 (1. OPTIMA , Nuffield Department of Medicine, University of Oxford, Oxford, UK; 2. University Department of Psychiatry, W arneford Hospital, Oxford, UK; 3. OPTIMA , B vitamin Laboratory, Department of Physiology, A natomy and Genetics, University of Oxford, Oxford, UK; 4. Department of Nutrition, Institute of Basic Medical Sciences, University of Oslo, Norway)

S8

Aim: Glycogen Synthase Kinase-3 (GSK-3) is an intracellular enzyme which seems to be overactive in the nerve cells of Alzheimer's disease (AD) patients and involved in its physiopathology. The inhibition of GSK-3 is expected to slow down the neurodegenerative process of AD. The clinical effects of Nypta® (NP-12, Tideglusib), a non-ATP competitive GSK-3 inhibitor of the thiadiazolidindione family, have been assessed in AD patients in a pilot, 20 week, double-blind, placebo-controlled, randomized, escalating dose, clinical trial. Method: Thirty male and female mild to moderate AD patients (Mini Mental State Examination (MMSE) 16 to 26), under stable and well tolerated anticholinesterasic treatment, were orally administered escalated doses of 400, 600, 800 and 1000 mg of Nypta® or placebo (ratio 2:1) for successive periods of 4, 4, 6 and 6 weeks respectively. The Primary Objective of the trial was to evaluate the safety and tolerability of NP12 and strict criteria for dose escalation and drug withdrawal were stated. As Secondary Objectives changes in cognition (MMSE, ADAS-cog, Word fluency) and depressive mood (Geriatric Depression Scale (GDS)) were assessed to estimate the clinical effects of the compound. A final Global Clinical Assessment (GCA) was also performed. Results: Treatment was well tolerated. The incidence of adverse events was similar in Active and Placebo groups except for some mild-moderate, asymptomatic and fully reversible increases of transaminases in.the Active group. Positive effects were observed in patients treated with Nypta® in four (MMSE, ADAS-cog, GDS and GCA) out of five efficacy

O13 - A PILOT CLINICAL TRIAL ON ALZHEIMER'S DISEASE WITH THE GSK-3 INHIBITOR NYPTA® (TIDEGLUSIB). T. DEL SER (Noscira SA , Madrid, Spain)

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

variables, although the small sample size precluded to reach statistical significance. Changes from baseline were in favour of the Active group by more than 1.5 points in the MMSE and more than 4 points in the ADAS-cog. The frequency of responders in MMSE was significantly higher in the Active group. Conclusions: This pilot study has provided valuable safety information about treatment of AD patients with NP12. Although only efficacy estimates have been obtained due to the small sample size and the dose escalating design, the current results are promising. More extended clinical trials are planned to confirm these findings.

Aim: The objective is to show how a simple, non-invasive and objective biomarker (AclarusDxTM) can assist investigators and the pharmaceutical industry in designing AD trials and in optimizing patient selection and recruitment. Methods: The utility of a biomarker in AD clinical trials can be considered in multiple ways including patient selection, stratification and recruitment, all of which remain challenging and may be detrimental to study completion timelines and successful study outcomes. The different concepts are discussed and arguments for use are provided. Results: Recent failures in well-run AD trials support the fact that it is time to change our approach to AD, its assessment in clinical studies and in treatment evaluation. AD is now recognized as being a heterogeneous disease with multiple clinical manifestations contributing to significant random noise in trial results, thus making the potential efficacy of a drug difficult to demonstrate. The scientific community needs to revise how patients are selected for clinical trials, moving beyond the current paradigm based on using clinical criteria supported by brain imaging. There are well-known CSF biomarkers (Abeta, tau proteins) which can contribute to this objective, however as they are invasive and not easily accepted by patients, this limits their widespread use in clinical studies. Adding genomic homogeneity to the selection criteria may contribute to improving study power to detect a real effect or equivalently reducing the number of patients needed to detect such an effect. The genomic approach is all the more interesting given that it provides objective "physical" information related to the individual patient, and is not related to human judgment which may be subject to variability. This is particularly important when running proof-of-concept studies as well as in subsequent studies in which AD subpopulations of interest may be identified. In addition, using a simple and non-invasive biomarker (AclarusDxTM) as an objective pre-selection tool may reduce a study's inter-patient variability while enlarging the patient selection pool by increasing the number of centers involved. A drug's clinical efficacy and safety are based on multiple factors (i.e. pharmacokinetic and pharmacodynamic profiles, genetic and environmental interactions, concomitant medication and diseases, etc.), the complexity of which can be captured at the level of genomic/transcriptomic regulation. Monitoring such regulation with genome-wide expression profiling tools can provide valuable information on the study drug's clinical response and offer new avenues for patient stratification. Discussion: New biomarkers can be identified to refine AD diagnosis and to provide medical and biological information in AD clinical trials. Simple and non-traumatic access to a biomarker is critical to its widespread use in the fragile AD population. Transcriptomic expression in blood can provide multiple disease and/or treatment specific biomarkers which fulfill these requirements. It is critical to expand the identification and use of clinically relevant biomarkers which can contribute to the renewal of AD drug development. The time has come to benchmark what has been done in cancer for many years and apply it to the AD field.

O14 - HOW BIOMARKERS CAN HELP INVESTIGATORS AND THE PHARMACEUTICAL INDUSTRY IN AD CLINICAL TRIALS. FROM CONCEPT TO APPLICATION. S. GAUTHIER 1 , O. SOL 2 , J.C. LEMARIE 3 , M. PANDO 2 (1. M cGill Centre f or S tudies in A ging, V erdun, Québec, Canada; 2. Ex onHit Therapeutics, Paris, France; 3. Effi-Stat, Paris, France)

(AD) and whether they are involved in goal directed behaviors. Scenario: The proposed scenario includes three parts: 1) guided actions (e.g., walk, sit, lay) 2) semi guided actions (e.g., make coffee) to be done by the patient without obligation concerning the order and 3) free activities (several potential activities are available, in the room; eg TV, reading, playing card...). The patient had to realize overall the scenario in one of the room of the Memory center. This smart room is fully was equipped with captors (video camera and microphone) and can be used to acquire training data. Captors: The proposed event recognition (the activity described on the previous paragraph) approach is based on a video interpretation Framework. The video event framework takes as input video streams and a priori knowledge. The a priori knowledge is all the information used by the event recognition process to infer high semantically representation of the scene. This knowledge is composed of 3D geometric information (i.e. empty scene model, camera calibration) and pre-defined event - behavior models. The framework contains a vision component (e.g. detection, classification and tracking tasks) (figure 2) and an event recognition component. The vision component allows to detect the person in the scene and to track his different movements over time. To compare the behavior of the volunteers the Normalized Difference of mean duration of Activity (NDA) was computed using the following equation; NDA = m1 ­ m2 / m1 + m2 here m1 and m2 are the mean duration of a certain activity. Results and discussion: Three demented patients (2 AD and 1 mixed dementia), 4 MCI and 2 control subjects were included at the time of the writing of the abstract. Each of the activities done by the patient has been considered as an event. Analysis focused on the following parameters: - Event recognition; - Event non-event realization; - Speed of realization of an event; - Order of realization of the events. At the end the scenario will be considered as efficient if the following conditions are respected: - Goodness of fit: do the criteria for assigning a condition to a particular cognitive/ behavioral symptom of dementia fit the case that are seen in practice ? - Ease of use: is the application of the scenario easy ? Can be used in differents AAL, different countries ? Is it easy to learn how to use it ? Reliability: is the test ­ retest and sensitivity to change acceptable ? Key words:,Alzheimer's disease, neuropsychiatry, behavior, Information and communication technologies O16 - AMYLOID-BETA ASSOCIATED CORTICAL THINNING IN CLINICALLY NORMAL ELDERLY. K.A. JOHNSON, T. HEDDEN, R.L. BUCKNER, R.A. SPERLING, J.A. BECKER (Massachusetts General Hospital, Brigham and W omen's Hospital, Harvard Medical School)

O1 5 - CLIN ICA L A S S ES S MEN T U S IN G IN FORMA TION A N D COMMUNICATION TECHNOLOGIES IN ALZHEIMER DISEASE: INTEREST FOR CLINICAL TRIALS. P.H. ROBERT 1 , R. ROMDHANE 2 , F. BREMONT 2 , E. MULIN 1 (1. Memory Center (CMRR) & Nice; 2. France INRIA Pulsar, Sophia A ntipolis, France) Care strategies for Alzheimer Disease (AD) are not only limited to pharmacological treatment and there is a need to promote clinical trials in other fields. Ambient Assisted Living (known as AAL) includes methods, concepts using Information and Communication Technologies (ICT) as well as services that are providing unobtrusive support for daily life based on context and the situation of the assisted person. More specifically the concept of AAL is understood as: · to extend the time people can live in their preferred environment by increasing their autonomy, self-confidence and mobility, · to support maintaining health and functional capability of the elderly individuals, · to promote a better and healthier lifestyle for individuals at risk, · to enhance the security, to prevent social isolation and to support maintaining the multifunctional network around the individual, · to support careers, families and care organizations, · to increase the efficiency and productivity of used resources in the ageing societies. The number of AAL projects using ICT for supporting elderly is vast and spread rapidly to the Alzheimer disease's field. However several questions are still in debate, including the quality and the standardization of clinical data provided and the way to assess the efficacy of such technologies. Objective: To present the research scenario used within the SWEET-Home project. Methods: The SWEET-Home project (ANR TecSan ­ 2009 program including the Nice CMRR, INRIA Pulsar, CNRS MICA center and Actis ingenierie) aims to build an innovative framework for modeling behavior and activities of daily living and most particularly focus on the assessment the cognitive and initiative ability of patient presenting Alzheimer's disease

Aim: Both amyloid- (A) deposition and brain atrophy are invariably associated with Alzheimer's disease (AD) and the disease process likely begins many years before symptoms appear. We sought to determine whether clinically normal (CN) older individuals with A deposition revealed by PET imaging using Pittsburgh Compound B (PiB) also have evidence of both cortical thickness and hippocampal volume reductions in a pattern similar to that seen in AD. Methods: One hundred and nineteen older individuals (87 CN subjects, CDR0; and 32 patients with mild AD, CDR1) underwent PiB PET amyloid imaging and high-resolution structural MR. Cortical thickness and volumes of interest were estimated using FreeSurfer. PiB retention over 40-60 minutes post-injection was calculated in a precuneus/posterior cingulate region. Regression models were used to relate PiB retention to cortical thickness and hippocampal volume. Results: We found that higher PiB retention in CN subjects was associated with cortical thinning, most prominently in precuneus, posterior cingulate and lateral parietal regions (p<0.0002), in a pattern similar to that observed in mild AD. Amyloid-associated medial temporal cortical thinning and hippocampal atrophy were also observed in CN, but were not statistically significant when age was included in the model. PiB retention and parietal/precuneus thickness were more strongly correlated in AD than in CN. Age-adjusted PiB and thickness data were evaluated using a model in which A deposition and cortical thickness follow sigmoid shaped curves in time. The predicted temporal lag between the dynamic phase of amyloid accumulation and the dynamic phase of cortical thinning was approximately 35% of the total time required for amyloid to increase from minimum to maximum (r-sq=0.49). For example, for an estimated PiB total increase time of 10 years, the lag time between the rapid phase of PiB increase and rapid phase cortical thinning would be 3.5 years. Conclusions: We conclude that A deposition is associated with a pattern of cortical thinning consistent with AD, prior to the development of cognitive impairment. Moreover, these data support a recently proposed dynamic biomarker model, in which A deposition occurs early and begins to plateau prior to maximal atrophy. These findings may have important implications for therapeutic trials in preclinical AD, particularly in tracking rate of cortical thinning prior to the onset of cognitive symptoms.

S9

Aim: Several recent studies in mild AD and pre-dementia AD have failed to show treatment effects. This failure may be a result of lack of efficacy of the treatments under consideration, but lack of sensitivity of the clinical scales that were used could also be to blame. Our goal is to provide diagnostic information on individual cognitive items from the ADAS-cog and from neuropsychological test batteries that could help inform the construction of cognitive scales that have improved sensitivity in mild and pre-dementia AD patients. Methods: Natural History data from patients in ADNI and clinical trial data from placebo patients in ADCS trials were used to assess scale properties. Separate ADAScog items, as well as the neuropsychological test items were assessed to determine changes at different stages of Alzheimer's disease. Reverse "S-curves" were calculated to show the

O17 - TIME COURSE OF COGNITIVE DECLINE IN SUBJECTS WITH PREDEMENTIA ALZHEIMER'S DISEASE AND MILD ALZHEIMER'S DISEASE BASED ON ADAS-COG SUBSCALES AND NEUROPSYCHOLOGICAL TESTS. S.B. HENDRIX, B.M. WELLS (Pentara Corporation, Salt Lake City, UT, USA )

average progression on each item relative to time. Several different composite scores were constructed by including items measuring different cognitive domains. These composite scores were assessed for sensitivity to progression over time. Results: The mean to standard deviation ratio of the ADAS-cog 11 item score was 0.53 over 12 months. Three ADAS-cog items were identified as contributing to the lack of responsiveness of the ADAS-cog in mild AD patients: language comprehension, naming objects and following test instructions. A composite score excluding these three items shows improved sensitivity to change over time in mild AD patients with a mean to standard deviation ratio of 0.57. The addition of three neuropsychological tests, category fluency and two scores from verbal paired associates, improves the sensitivity to 0.76 over 12 months. This "composite" score is substantially more sensitive to change in Mild AD, resulting in better power for smaller sample sizes. Results for several other composite scores are shown in Mild AD and in pre-dementia AD. Discussion: This analysis illustrates the improved sensitivity to change that is possible with a composite score constructed from commonly used cognitive items. This type of composite score could be used in a proof of concept study. These types of analyses could be used in conjunction with a theoretical understanding of the disease to construct a valid cognitive instrument for measuring cognitive change in pivotal Alzheimer's disease studies. O18 - IDENTIFICATION OF BLOOD TRANSCRIPTOMIC SIGNATURES IN AD PATIENTS RELATED TO EHT 0202 TREATMENT RESPONSE AND EFFICACY. P. BEURDELEY 1 , J.-C. LEMARIE 3 , E. BLONDIAUX 3 , O. SOL 1 , J. CARRIERE1, R. HADDAD1, L. DESIRE1, M. PANDO1, R. EINSTEIN2 (1. ExonHit Therapeutics SA , Paris, France; 2. ExonHit Therapeutics, Inc., Gaithersburg, MD 20877, USA ; 3. Effi-Stat, Paris, France)

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

O19 - A PHARMACOGENETIC-ASSISTED ALZHEIMER'S DISEASE DELAY OF ONSET TRIAL DESIGN WITH DIAGNOSTIC PREDICTOR VALIDATION. A.D. ROSES, K. WELSH-BOHMER1, O. MAKEEVA, S. BREWSTER, J. ALTMAN, J. ARBUCKLE, Y. MARUYAMA, A. SAUNDERS1, D. CRENSHAW1, M. LUTZ1, D.K. BURNS1 (Zinfandel Pharmaceuticals, Inc.; 1. Duke University, Durham, NC, USA )

Background: To date, treatment response in Alzheimer's Disease (AD) is mainly assessed using psychometric scales. Due to the recent development of new biotechnology tools, drug response can be measured also by interrogating genomic expression profiles. The current study is based on the identification of blood-based transcriptomic signatures using Exonhit's proprietary SpliceArrayTM technology in AD patients who were enrolled in a clinical trial to assess EHT0202, a new compound sharing both potential diseasemodifying and symptomatic properties. Methods: EHT0202/002 study was a randomized, double-blind, placebo-controlled, parallel group, phase IIA study aimed at determining the clinical safety, tolerability and exploratory efficacy of EHT0202 (40 and 80mg bid) as adjunctive therapy to one cholinesterase inhibitor (AchEI) over a 3-month period in mild to moderate AD patients. A research endpoint was to identify prospective blood expression biomarkers using the SpliceArrayTM technology for predicting patient response to EHT0202 as measured by a change in the ADAS-cog total score. During the study, blood samples were collected from each randomized patient at two time-points; before study treatment initiation and at study completion. Results: Sixty AD patients (20 in each of the 3 study groups, placebo, 40 and 80 bid) were selected that either improved or declined during the study period with regards to ADAS-cog total score by more than 3 points. Another 20 cognitive impairment free subjects served as controls. Analyses were performed to identify predictive and efficacy biomarkers from blood transcriptome using the SpliceArrayTM technology in relation to treatment response of EHT0202. Biomarkers were identified to associate with response and the selected biomarkers were correlated with BMI, a known risk factor for dementia. In addition, the involved genes play significant roles in metabolic activity, supporting the biological implication of the biomarkers. Efficacy biomarkers were also identified and will be presented in addition. Conclusions: Pharmacogenomic analysis can be applied in clinical trials to discover predictive or monitoring biomarkers that can be used to identify patients which will benefit from a treatment or follow their response to that treatment. In this study, we begin to address this approach with the identification of biomarkers for prediction response and efficacy; the next step will be to validate those markers with a larger cohort.

presented and discussed via the FDA Voluntary Exploratory Data Submission (VxDS) process with the FDA from January ­ October 2009. The design was independent of any particular drug and dealt with issues of companion diagnostic testing according to current FDA guidance. , Trial design considerations included a low risk group to accurately determine clinical performance characteristics for the test. The FDA also expressed strong opinions concerning the safety of a drug that would be tested in normal, non-demented individuals. Results: A design was agreed by the FDA VxDS as meeting the criteria of a "single adequate and well-controlled clinical investigation" for registration of a candidate drug for an Alzheimer's Disease delay of onset clinical trial. Based on incidence rates of cognitive impairment/dementia in prior studies, as well as outcome assumptions (screen normal, age 60-87, allele frequency, drop-out rate, etc.) the sample size projected for each arm of the study was 387, considerably less than epidemiological studies and conventional primary prevention studies. As a conservative approach, we have planned for at least 1,000 subjects per arm. Pre-recruitment ascertainments for such a trial performed at several geographical locations have been ongoing for almost two years (www.opalstudy.org). Discussion: It was clear from the discussion that the FDA Offices responsible for drugs trials and for diagnostic test reviews had, at the beginning of the consultation, heterogeneity of views regarding the application of a new genetic test as a companion diagnostic. Validation of the test had two distinct meanings: the technical robustness of the test and its clinical validity, or its predictive accuracy. The design of the trial is a prospective study of geographically-ascertained normal Caucasian subjects of 60-87 years, with a high or low risk of cognitive impairment over the subsequent 5-7 year period risk on entry. The high risk group would be randomized to receive either drug or placebo, while the low risk group will receive placebo. If the drug does not delay age of onset of disease over the course of the trial, the trial would still provide measures of the diagnostic performance characteristics. O20 - TOWARDS A MOLECULAR ANALYSIS OF -SYNUCLEIN PROTEIN IN CEREBROSPINAL FLUID. H. DECRAEMER HILDE, S. DE MEY, C. CINDY, P. GROGNET, D. JACOBS, E. VANMECHELEN (Innogenetics NV , Technologiepark Zwijnaarde 6, 9052 Gent-Zwijnaarde, Belgium)

Detection and quantification of -synuclein in cerebrospinal fluid (CSF) has been difficult with concentrations of -synuclein ranging from ng/mL to pg/mL. Furthermore synuclein in brains of patients with Lewy body diseases is quite heterogeneous ranging from post-translational modification such as ubiquitination, nitration, abnormal phosphorylation, truncation, alternative splicing and oligomeric forms. In order to establish -synuclein as a possible marker for Lewy body pathology in the brain highly robust quantifications are necessary using well-characterized tools for detection of these species. We now report the initial characterization of highly selective -synuclein monoclonal antibodies (mAbs) using recombinant -synuclein constructs and synthetic peptides. Standardization of an -synuclein assay is strongly influenced by the use of different mAb combinations, buffers and detergents. One combination of two monoclonal antibodies was further used in spiking-experiments with recoveries ranging from 80 and 120 % in CSF. Using this combination we were able to quantify -synuclein in CSF of normal individuals in the low pg/ml range. As previously noted, samples contaminated with red blood cells have higher CSF--synuclein levels, thus suggesting that hemorrhagic CSF is not suitable for -synuclein determinations. Developing this test into a multi-analyte assay, the combination of -synuclein quantification with total tau, phosphorylated tau, and A1-42 levels, will allow us to determine the clinical utility of this new -synuclein assay in a range of dementias, including Lewy body diseases.

Aim: The specific aim is to design a combination genetic diagnostic validation and disease prevention/delay of onset clinical trial, based on a new genetic diagnostic/predictor for the onset age of Alzheimer's disease, as reported in 2009. The genetic test measures the length of a variable polyT repeat inherited on both strands of DNA. The locus, rs10524523 (523) is within intron 6 of the mitochondrial membrane channel gene, TOMM40, through which virtually all cytoplasmic peptides and proteins pass into the mitochondrion in order to support mitochondrial fission. TOMM40 is adjacent to and in linkage disequilibrium with APOE and the three APOE alleles have characteristic TOMM40 polyT lengths linked to them through evolution. APOE4 strands have long polyT lengthss (L=19-29 Ts) linked to them approximately 98% of the time in Caucasians. APOE3 strands have either short polyTs (S<18), or very long polyTs (VL>30). Previous reports have demonstrated in APOE3/3 subjects that the age of onset distributions for 523 length variants, the ages of incident cases in prospective cohorts, changes in gray matter density using MRI, and neuropsychology changes occur earlier when these individuals carry two longer length 523 polyTs (VL-VL), and relatively later with S-VL, or S-S genotypes. Thus the inheritance of APOE3 is not neutral in riski f this allele is linked to VL 523 polyTs. APOE3, then, can be effectively differentiated for age of onset distributions when 523 genotype is considered ­ a very useful tool for stratifying a prevention trial. Methods: A proposed design for prevention/delay of onset of AD was

O21 - PLASMA EXCHANGE WITH 5% HUMAN ALBUMIN FOR AMYLOID- MOBILIZATION IN PATIENTS WITH MILD-MODERATE ALZHEIMER'S DISEASE. M. BOADA1,2, P. ORTIZ3, F. ANAYA4, I. HERNANDEZ1, J. MUNOZ3, L. NUNEZ 5 , M. TORRES 5 , J. OLAZARAN 6 , I. ROCA 7 , G. CUBERAS 7 , LL. TARRAGA1, M. BUENDIA1, R. PLA3, I. FERRER8, A. PAEZ5 (1. Fundació A CE. Institut Català de Neurociències A plicades. Barcelona. Spain; 2. Hospital Universitari V all d'Hebron - Institut de Recerca, Universitat A utònoma de Barcelona (V HIR-UA B). Barcelona. Spain; 3. BST (Banc de Sang i Teixits). Barcelona. Spain; 4. Nephrology Departament, Hospital Gregorio Marañón. Madrid. Spain; 5. Clinical Trials Department, Instituto Grifols S.A . Barcelona. Spain; 6. Neurology Department. Hospital Gregorio Marañón. Madrid. Spain; 7. Nuclear Medicine Department. Hospital V all d'Hebron. Barcelona. Spain; 8. Neuropathology Institute, Hospital Bellvitge. Barcelona. Spain) Based on the fact that most circulating amyloid- (A) is bound to plasma albumin, a clinical investigation program was carried out to replace endogenous albumin in mildmoderate Alzheimer's disease (AD) patients with 5% Human Albumin Grifols® through a plasma exchange (PE) schedule. This therapeutic approach was aimed to alter the dynamic equilibrium existing between albumin-bound A in plasma and A in cerebrospinal fluid. The first study of the clinical investigation program was a pilot, proof-of-concept study was performed based on 7 patients underwent 6 PE in 3 weeks followed by 1 year of follow-up. Plasma A determinations demonstrated a consistent variation pattern in plasma A levels in relation with each of the PEs. In addition, cognitive status scores (MMSE and ADAS-Cog) were more stable than expected in this patient population. Following the pilot study, A phase II randomized, controlled, clinical trial, was carried out. Forty-two patients were randomized into PE-treated and control groups with 1 year follow-up. Three different PE schedules were planned: 2 PEs/week for 3 weeks, 1 PE/week for 6 weeks and 1 PE/2 weeks for 12 weeks. Although all patients have already been recruited, two patients are still finalizing the follow-up period. However, the interim results with a sample of 29 patients pointed toward the occurrence of A40 mobilization in the PE-treated patients with the same characteristics of the variations seen in the pilot study. No such A40

S10

There is widespread recognition in the urgency to understand the causes and mechanisms of senile dementia. Attempts to find cures for Alzheimer's disease (AD) have, however, failed so far, in spite of enormous investments, intellectual and financial. We therefore have to reconsider the problem from new angles. AD is regarded as a disease because of its clinical manifestations and underlying pathology. However, this combination does not define a disease but rather a syndrome, just like hepatic cirrhosis in which liver pathology causes metabolic changes, but which can result from many different etiologies. It is unlikely that attacking a downstream phenomenon, like apoptosis or amyloid accumulation, can cure AD, or prevent the progression of the disease. Epidemiological studies have identified many risk factors for "senile dementia of the Alzheimer type", some genetic but most environmental and therefore modifiable. It is probable that senile dementia is the result of a combination of several processes, working differently in each person. Therefore a concerted action to fight the dementia epidemic must be made by aggressive action against its risk factors, and this battle must begin in midlife, not in old age.

O23 - WHY HAVE WE FAILED TO CURE ALZHEIMER'S DISEASE? A.D. KORCZYN

mobilization was observed in the control group. Additionally, PE-treated patients scored better in cognitive tests (differences at 9 months: 2.5 in MMSE and 5.5 in ADAS-cog). These results suggest that a PE program with 5% Human Albumin Grifols may have a promising role in the treatment of mild-moderate AD patients.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Aim: Typical Alzheimer's disease (AD) clinical or observational trials include a vast array of cognitive, functional, and health outcome test batteries that can range across many possible total scores for each disease domain. Due to the differing score ranges it can be difficult to graphically assess between-test changes (relative to each tests range), and to present all results in a common graph. The aim of this work is to develop a plotting technique that facilitates test score comparisons with respect to each single measure on a common scale. Further, a modification of the standard boxplot is developed to better express percentile results relative to the range of each test. Methods: For a given test, individual scores or summary measures are transformed relative to the total score of the test as follows. Let sij denote the score on the jth test for the ith subject. Let Tj denote the total score of the jth test. The proportion worsening relative to the total score of test j is given by sij divided by Tj for tests that increase with cognitive worsening and (Tj - sij) divided by Tj for tests that decrease with cognitive worsening. Denote this quantity by ij. Each subjects' data can be plotted on a unit circle by connecting the origin (0,0) with the following points. x ij = sin (2*ij) y ij = cos (2*ij). Summary measures, e.g. mean or median, can be plotted as well (using the summary measure in place of the sij's). When plotting individual subject data, percentile ranges can be color coded to identify the inter quartile range, outliers, median, and other relevant percentile identifiers. Important summary statistics such as the mean can be plotted as well. Data from a non-interventional observational study which enrolled 196 AD, 70 MCI, and 75 cognitively healthy volunteers was used for all analyses. Results from several cognitive (ADAS-Cog, MMSE, and NTB) and functional (CDR-SB, DAD, and FAQ) assessments will be used in the graphical analyses. Results/Discussion: The graphical method presented provides a simple visual tool to assess differences among cognitive and functional tests, and between the study subpopulations. In particular, substantial movement (such as CDR-SB) and relatively slight movement or learning effects (the tests that comprise the NTB) relative to the total test score is easily identifiable. This type of visual inspection can be used to compare many tests from a multitude of data bases when considering optimal outcome measures for novel clinical trials. Additionally, the new method of presenting percentile data that define boxplots might be more effective visually as these statistics can be gauged relative to the ranges of each test. Two limitations are worth considering. First, this technique cannot be used for tests without a total score upper bound. Second, the graphs do not describe the variability associated with each outcome. In conclusion, the graphical technique described in this presentation facilitates easy visual identification of test score movement within its range of possible scores, allows comparisons among tests that assess a common disease domain, and provides a means of effectively comparing different study subpopulations.

P1 - GRA PHICAL METHOD FOR THE PRESENTATION OF MULTIVARIATE TEST SCORE DATA. G.G. CRANS (Elan Pharmaceuticals, South San Francisco, USA )

POSTERS

modifications, and biochemical properties. Method: Samples: Serum and plasma were obtained from AD patients with typical clinical course and brain imaging results, age matched normal subjects, and amnestic MCI subjects with subsequent evidences which support progression to AD afterwards. Pre-treatment: Samples were either subjected to immunoaffinity column chromato- graphy to remove abundant proteins such as albumin and immunoglobulins, or directly immunoprecipitated to enrich proteins of interest, such as tau and beta-amyloid. Ettan DIGE (differential gel electrophoresis) analysis: In order to search candidate epitope(s) expressed on proteins (>10 kD), it was employed to differentially label samples derived from diseased and normal subject with different fluorescent dyes, after which 2D separation in the same gel was performed, followed by fluorography by a scanner equipped with a sophisticated analytical software for separated protein spots. Mass spectrometric analyses: As a direct workflow from the above step, isolated protein spot was subjected to peptide mass fingerprinting analysis using a MALDI-TOF mass spectrometer (MS) to identify the protein of interest. For analyses of peptides (< 10 kD) and low molecular mass compounds, a liquid chromatography (LC) MS system equipped with a multivariate analysis software was used. Targets were identified by peptide mass fingerprinting by MALDI-TOF-MS or directly by LC-QIT-MS/MS. Also modification to peptides, such as phosphorylation, glycosylation, and nitrosylation were analyzed and identified by using LC-MS and LC-QIT-MS/MS. Elaboration of monoclonal antibody: For identified target epitopes through the analyses, such as stereo-specific structure generated by modified amino acids and other post- translational modifications, a rat lymphocyte system was used to efficiently screen a large number of hybridoma clones generated. Results: At present, candidate peptides were found and identified through analysis of plasma derived from AD patients and amnestic MCI subjects by immunoprecipitation using an antibody which recognizes an epitope of candidate biomarker protein, such as human brain carboxypeptidase B, a beta-amyloid modifying enzyme expressed in brain. Other candidate biomarker protein(s) are also recognized by the Ettan DIGE analysis, and are now under investigation for further study. Discussion: The Ettan DIGE separation and analytical system in combination with suitable mass spectrometric analyses for identification is recognized as the most precise analytical measures for detection of protein targets of interest. For peptides and low molecular mass targets, LC-QIT-MS/MS analysis in combination with multivariate analysis seems to be the most practical approach.

P2 - A PROTEOMIC APPROACH FOR SEARCH AND IDENTIFICATION OF SURROGATE BLOOD BIOMARKERS FOR ALZHEIMER'S DISEASE AND MILD COGNITIVE IMPAIRMENT. A. MATSUMOTO 1 , H. TSUMOTO 1 , N. SHIKAKU1, M. TOYOMOTO1, R. MATSUMOTO2, H. SUGIMOTO1 (1. Graduate School of Pharmaceutical Sciences, Kyoto University 1, Kyoto; 2. Medical Corporation Midorikawa Clinic of Neurology 2, Kyoto, Japan) Aim: It is widely approved that there is an urgent need for establishing surrogate biomarkers for early-stage AD (Alzheimer's disease) and amnestic MCI (Mild Cognitive Impairment), especially in developed countries and developing countries where proportion of the elderly is increasing. Through a significant progress in mass spectrometry and its applications, this study was intended to search and identify blood biomarker candidates, with a wide range of characteristics such as, molecular mass, post-translational

Background: Patients with mild cognitive impairment (MCI) often describe changes in behavior and personality. However, the use of self reports has been questioned since loss of insight and unawareness of cognitive deficits may bias the accuracy of their information. To increase reliability, reports from informants and objective measures have been used. In order to increase knowledge about changes in early, preclinical stages of dementia, the aim was to examine degree of agreement between self and informant reports on personality in relation to cognitive function in patients with subjective cognitive impairment (SCI), mild cognitive impairment (MCI) and healthy controls. Methods: Patterns of personality was assessed in 24 patients with SCI, 35 with MCI and 26 healthy controls with the Swedish universities Scales of Personality (SSP) questionnaire. The participants' self reports were compared to parallel versions of informants' reports on 13 SSP scales, measuring aspects of neuroticism/anxiety proneness, extraversion, and aggression-hostility. An incongruence index, based on the differences between patient and informant ratings, was employed to estimate degree of agreement between patients/controls and their informants. The MiniMental State Examination (MMSE) was used to assess global cognitive function. Results: The correlations between all participants' self ratings of their personality and ratings by their informants' were moderate to high in most SSP scales. Ratings differed significantly only in the Detachment scale, but not in the other scales. Moreover, patients with MCI and their informants scored higher than patients vs. informants with SCI followed by controls vs. informants in most SSP scales. Patients with MCI showed significantly larger incongruence than patients with SCI and controls across SSP scales, meaning lower agreement between patient and informant. There was a significant negative correlation between the incongruence index and the MMSE score for all subjects. Discussion: The results indicate that a high degree of disagreement between patients' self reports and that of their informants' could be an early indicator of cognitive impairment. Informant reports ought to be a mandatory tool in dementia assessment. P4 - ANTEMORTEM MODELING OF BRAAK & BRAAK NEUROPATHOLOGY. R.E. TRACTENBERG (Departments of Neurology, Biostatistics, Bioinformatics & Biomathematics, and Psychiatry; Georgetown University Medical Center, W ashington DC, USA )

P3 - PATIENT VERSUS INFORMANT REPORTS ON PERSONALITY IN SUBJECTIVE COGNITIVE IMPAIRMENT AND MILD COGNITIVE IMPAIRMENT. N. BOGDANOVIC 1, B. AUSEN 1, G. EDMAN 2, O. ALMKVIST 1,3 (1. Department of Neurobiology, Care Sciences and Society, Karolinska Institutet, Stockholm, Sweden; 2. Department of Psychiatry, Danderyd Hospital, Stockholm, Sweden; 3. Department of Psychology, Stockholm University, Stockholm, Sweden)

S11

Aim: Investigators have recently published statistically significant relationships between many different biomarkers and various clinical constructs including "brain damage" (Querbes et al. 2009), "atrophy" (Jack Jr, et al. 2009), "AD pathology" (Schuff et al. 2009), "neuropathology" (Nestor et al. 2009) and "disease severity" (Desikan et al. 2009). Three specific components in AD neuropathology are: "tangle formation", "synapse loss or failure", and "atrophy" (Braak & Braak 1991; Jack Jr, et al. 2009, p. 1356). This poster describes causal measurement models representing key constructs for which the biomarkers are theorized to be indicative (synapse loss, neuropathology, atrophy (Braak &

The essential fatty acid DHA (docosahexanoic acid) comprises 3% of the dry mass of the human brain. Dietary deficiency of DHA is associated with increased incidence of neurodegenerative disorders including Alzheimer's disease and subsequent brain cell death. Nutritional DHA deficiency also affects brain zinc homeostasis. Our previous work showed that DHA deficiency increased hippocampal zinc levels and raised the expression of the ZnT3 putative zinc transporter in the rat brain*. We hypothesize that DHA deficiency induces neuronal cell death through zinc-induced apoptosis. To elucidate the link between DHA, zinc and brain cell death, we grew cultured human neuronal M17 cells in DHA-deficient and DHA-enriched culture medium. Exposure of M17 cells to DHAdeficient medium reduced the levels of active caspase-3 and increased the levels of Bcl2, relative to levels in DHA-replete cells, confirming the adverse effects of DHA deficiency in promoting neuronal cell death. To investigate the role of zinc in DHA-induced apoptotic cell death we grew cells in DHA-deficient and DHA-replete culture medium and measured zinc uptake using radiolabelled zinc. In DHA­deficient M17 cells, zinc uptake was 30% less in compared with that of DHA-replete cells. Furthermore, in DHA-treated cells, ZnT3 mRNA and protein levels were reduced compared to the levels in DHA-replete cells. Based on previous reports showing free zinc mediates brain cell death through apoptosis, we propose that neuroprotective function of DHA is exerted through a reduction in cellular zinc levels that in turn inhibits apoptosis. * Jayasooriya AP et al. 2005. Perinatal {omega}-3 polyunsaturated fatty acid supply modifies brain zinc homeostasis during adulthood. Proc Natl Acad Sci U S A. 102(20):7133-7138.

P5 - THE OMEGA-3 FATTY ACID, DHA, ACTS THROUGH A ZINCDEPENDENT MECHANISM TO PREVENT NEURONAL CELL DEATH. D. DE MEL 1 , C. SUPHIOGLU 1 , L. KUMAR 1 , N. SADLI 1 , D. FREESTONE 1 , A. MICHALCZYK 1 , A. SINCLAIR 2 , M.L. ACKLAND 1 (1. S chool of L if e and Environmental Sciences; 2; School of Exercise and Nutrition Sciences, Deakin University, 221 Burwood Highway, Burwood, V ictoria 3125, A ustralia)

Braak, 1991). Variables in the models capitalize on the multitude of indicators for each of these processes being collected in ADNI (and ADNI-type protocols). Method: Modeling of MRI-derived volumes and PET activation values was accomplished with confirmatory factor analyses. Akaike's information criterion and four other fit indices were used to choose the best fitting models at two visits 12 months apart. Temporal replication of model fit was the sixth criterion for identifying the best fitting model(s). Results: A one-factor model is an insufficient fit to the ADNI data-failing to fit both baseline and 12 month visit data. A second, general neurological status factor, possibly representing the robustness of the brain to the atrophy or neuropathology that is theorized to cause the volume and activation variables to vary/covary, yielded excellent fit to the baseline data and to the 12 month data when the two factors were modeled to be correlated. Discussion: PET values were associated with cognitive performance variability; volume values were unassociated with PET and with variability, supporting modeling of "synapse failure" separate from volume-based "atrophy". The constructs "synapse failure", represented by PET values and intra-individual variability, and "atrophy" represented by volumetrics should be modeled separately, as implied by Braak & Braak (1991). Biomarkers should not be analyzed as if they are independent but as systems, capturing their covariation (and common causes thereof) and representing neuropathology to further scientific knowledge and to be useful in clinical trials- as endpoints as well as in the identification of participants at greatest risk for developing MCI and/or AD.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

P7 - OPTIMIZATION OF THE TREATMENT REGIMEN WITH ACTIVE Aß IMMUNOTHERAPY CAD106 IN ALZHEIMER PATIENTS. G. IMBERT 1 , N. ANDREASEN2, M.E. RIVIERE1, J. ROS1, J. MOREAU1, J. SEVIGNY1, P. QUARG1, A. CAPUTO1, L.A. FINELLI1, B. WINBLAD2, A. GRAF1 (1. Novartis Pharma A G, Basel, Switzerland; 2. Karolinska Universitetssjukhuset, Huddinge, Sweden) Background: CAD106 is an active immunotherapy being developed for Alzheimers Disease (AD), which comprises the amyloid (Aß) 1-6 peptide coupled to the Qß virus-like particle. In animals, CAD106 induced Aß-antibody titers without activating Aß-reactive Tcells. Administration of CAD106 to APP transgenic mice showed a reduction of amyloid accumulation in the brain. Objective: Two regimens were tested in two Phase IIa studies, with the objective to assess safety, tolerability and immunogenicity. These studies were 52week, multi-center, randomized, double-blind, placebo-controlled, parallel group studies in patients with mild AD. CAD106 dose of 150 g was administered subcutaneously (s.c.) at weeks 0/6/12 in Study 1, and s.c. or intramuscularly (i.m.) at weeks 0/2/6 in Study 2. Results: At baseline, out of 58 patients, none had Aß-specific IgG antibody titers above the Lower Limit of Quantitation (LLOQ, 8.9 units) of the assay. Aß IgG antibody response was defined as increase above 16 units (LLOQ + 3 SD) at least once during the study. Following treatment with CAD106 in Study 1, antibody response was observed in 20/22 patients. Similar mean peak titers were achieved following both the second and third injection (week 8: 44 units, week 14: 40 units). In addition, antibodies were detected for longer than 6 months. Preliminary data from Study 2 are available. These data show a single antibody titer peak at week 8 of similar magnitude at the same time-point in Study 1. Responder rate and duration of response seem also to be comparable at this early stage. Antibody titer profiles with s.c. and i.m. routes of administration from Study 2 will be presented. Conclusions: Preliminary data from these studies confirm previous information from Phase I that injection at week 2 does not enhance antibody response. A week 0/6/12 regimen is therefore selected for the next study. Further injections at 12-week intervals are also being evaluated, along with addition of adjuvants.

with a greater decrease of p-tau-181 and a smaller increase of total tau in the follow-up. Discussion: Hyperphosphorylated tau decreases in the late stages of the AD process. The decrease of p-tau-181 may correlate to clinical cognitive functioning, probably as a reflector to neuronal loss. More longitudinal study of CSF biomarker dynamics is needed, especially in the preclinical stage of the disease.

Aim: Longitudinal changes of Alzheimer's disease (AD) biomarkers have been studied mainly in small groups with inconclusive results. Despite the well-shown evidence of the clinical predictive value of the cerebrospinal fluid biomarkers, little is known of their longitudinal variation in different stages of the disease course. We hypothesized that there would be a connection between the progression of the cognitive decline and the CSF biomarker changes. Method: 131 memory clinic patients (56 AD, 57 MCI, 10 other neurological disorders, 8 unimpaired) underwent a clinical follow-up with repeated MMSE tests and two lumbar punctures with a median interval of 2.98 years (range 0.48-8.15). The clinical routine included dementia screening with medical history, medical and neurological status and neuroimaging. The clinical diagnostics were made blinded to CSF results. CSF A42, tau and p-tau-181 concentrations were measured using commercially available ELISA (Innotest, Belgium). Results: During the clinical follow-up of 36 months, 21 subjects progressed to AD. 56 subjects had AD already at the baseline. 26 subjects remained stable with MCI. Subjects with the fastest MMSE decrease rate (lowest quintile) had the lowest baseline A42, highest tau and highest p-tau-181 CSF concentrations at the baseline. The baseline biomarker levels and A42/tau ratios differed linearly by MMSE decrease rate quintiles. The CSF A42 concentration decreased significantly (p<0.05) faster in the AD-AD group (decrease of -11.9 pg/ml/year) compared to stable MCI-MCI group (increase of 3.42 pg/ml/year), but overall differences in A decrease rate between other groups remained statistically insignificant. An annual decrease of -2.20 pg/ml/year in the CSF p-tau-181 concentration of AD-AD patients was noticed. The difference was significant compared to stable MCI-MCI (increase of 1.24 pg/ml/year) and cognitively healthy (increase of 0.84 pg/ml/year) subjects (p for group differences 0.004). The decrease rate of p-tau-181 correlated to MMSE decrease rate (r=0.402, p<0.001) in the entire cohort as well as in the AD-AD group (r=0.470, p<0.001). The APOE 4 allele was associated

P6 - THE LONGITUDINAL CHANGES OF CSF BIOMARKERS IN ALZHEIMER'S DISEASE ­ CSF HYPERPHOSPHORYLATED TAU DECREASES IN CONNECTION WITH COGNITIVE DECLINE DURING THE AD PROCESS. T. SEPPALA, A. KOIVISTO, P. HARTIKAINEN, S. HELISALMI, H. SOININEN, S.-K. HERUKKA (University of Eastern Finland, Kuopio University Hospital, Kuopio, Finland)

P8 - SURVEY OF SEGMENTATION PROTOCOLS FOR MANUAL HIPPOCAMPAL VOLUMETRY: PREPARATORY PHASE FOR AN EADC-ADNI HARMONIZATION EFFORT. M. BOCCARDI1, R. GANZOLA1, S. DUCHESNE2, N. ROBITAILLE 2 , M. BOCCHETTA, A. REDOLFI 1 , G. BARTZOKIS 3 , J.G. CSERNANSKY4, M.J. DE LEON5, L. DETOLEDO-MORRELL6, R.J. KILLIANY7, S. LEHERICY8, R. CAMICIOLI9, J. PANTEL10, J.C. PRUESSNER11, H. SOININEN12, C. WATSON 13 , C. JACK 14 , G.B. FRISONI 1 (1. IR CCS ­ S . Giov anni di Dio ­ Fatebenefratelli Brescia, Italy; 2. Université Laval and Centre de Recherche Université Laval ­ Robert Giffard, Quebec City, Canada; 3. David Geffen School of Medicine at UCLA , Los A ngeles, CA ; 4. Northwestern University Feinberg School of Medicine, Chicago, IL, USA ; 5. Center for Brain Health, New Y ork, NY ; 6. Rush University, Chicago, Illinois; 7. Boston University School of Medicine; 8. CENIR and Dept of Neuroradiology, Université Pierre et Marie Curie-Paris 6, Groupe Hospitalier PitiéSalpêtrière, Paris, France; 9. University of A lberta, Edmonton, A lberta, Canada; 10. Univ ersity of Frank furt/Main, Germany ; 11. McGill Centre for S tudies in A ging, Department of Psychiatry, McGill University, Montreal, Quebec, Canada; 12. Kuopio University Hospital, Kuopio, Finland; 13. W ayne State University School of Medicine, 8D-University Health Center, St. A ntoine, Detroit, MI; 14. Mayo Clinic and Foundation, Rochester, MN) Aim: Hippocampal atrophy is a key diagnostic marker for early and incipient Alzheimer's disease (AD) but different procedures give rise to estimates of normal hippocampal volume that vary up to 2.5 times across laboratories. Internationally harmonized procedures are required to use hippocampal volume as a diagnostic marker in clinical settings, surrogate measure in clinical trials for AD drugs, and gold standard for automated segmentation algorithms. Methods: The most frequently used segmentation protocols for hippocampal volumetry in the AD literature (see following) were surveyed to extract anatomical landmarks. The hippocampi of one healthy control and one AD patient taken form the ADNI database were segmented by a single rater following each protocol. Individual interactive web conferences were arranged with the primary authors to check correctness of interpretation. Semantically harmonized landmark differences among protocols were extracted and operationalized into segmentation units (SU). SU were quantitatively evaluated in AD and controls. The protocols were [B] Bartzokis et al., 1998, [C] Convit et al., 1997, [dTM] deToledo-Morrell et al., 2004; [H] Haller et al., 1997, [J] Jack et al., 1994, [K] Killiany 1993 et al., [L] Lehericy et al., 1994, [M] Malykhin et al., 2007 [P] Pantel et al., 2000, [Pr] Pruessner et al., 2000, and [S] Soininen et al., 1994; [W] Watson et al., 1992. Results: The survey of anatomical landmarks of the protocols, and the hippocampi segmented following each of the protocols and checked with their authors can be found at http://www.hippocampal-protocol.net. Landmark differences among protocols (http://www.hippocampal-protocol.net/site/sops-project-outlines.html) regard: (a) inclusion [C,dTM,J,L,M,P,PR,W] and exclusion [B,H,K,S] of hippocampal white matter (alveus and fimbria); (b) specification of the most posterior slice: most restrictive [B], most inclusive [H,M,P,Pr]; (c) separation between the subiculum and the parahippocampal gyrus, consisting of variable combinations of four criteria. Heterogeneities in the definition of the boundary with the amygdala, while a source of variance in the earlier protocols, are no longer an issue due to currently available visualization tools allowing 3D navigation.

S12

Aim: In addition to memory impairment, other cognitive dysfunctions, decline of activities of daily living and neuropsychiatric symptoms are common manifestations of Alzheimer's disease (AD) already in the early phases of disease process. The present study was designed to examine whether cognitive performance is associated with demographic variables, dementia severity, ADL and neuropsychiatric symptoms of patients with very mild or mild AD. Method: We analyzed the baseline data of 236 patients with recently diagnosed very mild or mild AD from a prospective follow-up study (ALSOVA) with a family caregiver. The primary aim of the ongoing ALSOVA study is to evaluate the effects and cost-efficiency of psychosocial rehabilitation. At baseline the cognitive evaluation was performed with The Consortium to Establish a Registry for Alzheimer's disease neuropsychological test battery (CERAD-NB) total score and the Mini-Mental State Examination (MMSE). Severity of dementia was evaluated with the Clinical Dementia Rating (CDR), ADL functions with the ADCS-ADL and neuropsychiatric symptoms with the Neuropsychiatric Inventory (NPI). Results: Cognitive performance was associated with education but not with age. Dementia severity and ADL functions were associated with education and age, but such association was not found with demographic variables and neuropsychiatric symptoms. Women got better scores on the ADCS-ADL than men but there were no gender differences in other measures. Cognitive performance was associated with dementia severity and ADL functions but not with neuropsychiatric symptoms. ADL functions were associated with both cognitive performance and neuropsychiatric symptoms. Discussion: Results indicate that cognitive and neuropsychiatric symptoms are not associated with each other but are relatively independent dimensions of AD, possibly reflecting different aspects of the disease process. Both cognitive decline and behavioral symptoms may cumulatively affect the patients' ability to perform activities of daily living. Results emphasize the importance of a multidimensional approach for the diagnostic and prognostic evaluation of patients with AD in clinical and research settings. P10 - MARKERS OF DISEASE PROGRESSION: FROM MCI TO AD. V. DRAGO, M. PIEVANI, C. BABILONI, A. CAROLI, L. VENTURI, P. SPITZER, J. WILTFANG, M. DIDIC, H. KLAFKI, F. VECCHIO, A. REDOLFI, G.B. FRISONI (Laboratorio LENITEM, I.R.C.C.S. San Giovanni di Dio Fatebenefratelli, Brescia, Italy)

P9 - COGNITIVE PERFORMANCE, FUNCTIONAL ABILITY AND NEUROPSYCHIATRIC SYMPTOMS OF PATIENTS WITH ALZHEIMER'S DISEASE. I. HALLIKAINEN 1 , A. HILTUNEN 2 , T. HÄNNINEN 3 , P. KARPPI 4 , T. PAAJANEN 1 , H. SOININEN 1,3 , M. VANHANEN 5 , T. VÄLIMÄKI 6 , T. PIRTTILÄ1,3 (1. Institute of Clinical Medicine, Neurology, University of Eastern Finland, Kuopio, Finland; 2. Department of Neurology, North Karelia Central Hospital, Joensuu, Finland; 3. Department of Neurology, Kuopio University Hospital, Kuopio, Finland; 4. Mikkeli Central Hospital, Mikkeli, Finland; 5. Niuvanniemi Hospital, Kuopio, Finland; 6. Department of Nursing Sciences, University of Eastern Finland, Kuopio, Finland)

Discussion: Operationalization of such differences into segmentation units, quantification of their contribution to total hippocampal volume and to the difference between AD and controls, and reliability measures for the tracing of each are also provided, and will constitute the quantitative evidence helping consensual decisions about a harmonized protocol for hippocampal segmentation. Segmentation units as operationalized from the survey of the 12 protocols for hippocampal segmentation.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

added to the cultured astrocytes at the concentrations of 25 nM, 50 nM and 100 nM . After 48 hours, the expression of TGF1mRNA and TGF2mRNA were examined by real-time RT-PCR in the treated cells. Results: Our findings have shown that treated cells express higher amounts of TGF1 and TGF2 in comparison with the controls and treated cells with 17--estradiol have a different morphology in comparison with untreated cells. Discussion: To seem that the inhibition of GSK3- by 17--estradiol increases of the expression of neuroprotective factors which points at the key role of some signaling pathway in regulating survival and function of astrocytes. Key words: neurodegenerative disease, Neuroprotective factors, GSK3-, 17-estradiol, astrocytes

P13 - GLUTAMINYL CYCLASE AS NOVEL PHARMACOLOGICAL TARGET FOR AD THERAPY. S. ROßNER1, M. HARTLAGE-RUBSAMEN1, S. SCHILLING2, H.-U. DEMUTH 2 (1. Univ ersity of L eipz ig, M edical Faculty , L eipz ig, Germ any ; 2. Probiodrug A G, Halle/S., Germany) Background: Brains of AD patients are characterized by the formation of high molecular weight Abeta aggregates. In particular, pyroglutamate (pE)-modified Abeta peptides are disease-specific, display fast aggregation, co-aggregate with non-modified Abeta peptides and may, therefore, act as seed for Abeta deposition. Recent evidence indicates that the generation of pE-Abeta is catalyzed by glutaminyl cyclase (QC). Thus, inhibition of QC could prevent the generation of pE-Abeta, accompanying gliosis and ameliorate behavioral deficits. Aim: The aim of this study was to reveal the cell type- and brain region-specific QC expression and its regulation under disease-related conditions. Additionally, we tested whether inhibition of QC reduces pE-Abeta generation and total Abeta aggregation in brain of transgenic mice. Methods: QC expression during aging of mice under physiological conditions and in animal models of neurodegeneration was studied by immunohistochemical double labeling. APP transgenic mice were treated with the QC inhibitor PQ50 for 6 to 10 months in prophylactic and therapeutic paradigms. At the end of the treatment period, the performance of mice was assessed in contextual fear and Morris water maze paradigms. Brains of mice were subjected to histological and biochemical analyses to reveal any effects of QC inhibition on pathological processes. Results: QC was found to be expressed by a relatively low proportion of interneurons in neocortex and in defined hippocampal layers. Additionally, robust QC expression was observed in brain nuclei known to be affected in AD, such as locus coeruleus, EdingerWestphal nucleus and nucleus basalis Meynert. Intracellularly, QC was present along the secretory pathway, including ER, Golgi and secretory vesicles. In animal models of neurodegeneration QC was also expressed by reactive astrocytes. Chronic inhibition of QC in APP transgenic mouse models resulted in reduced pE-Abeta and total Abeta levels, ameliorated Abeta plaque pathology and gliosis and in improved performance of mice in Morris water maze and contextual fear paradigms. Discussion: Our data demonstrate a QCmediated pE-Abeta formation in brain in vivo. The brain region-specific and subcellular localization of QC implies a role in Abeta pathology and in neurodegeneration of defined vulnerable cell populations. Thus, inhibition of QC may be beneficial with regard to Abeta pathology and neuronal cell loss. We conclude that inhibition of QC is a novel therapeutic approach for AD treatment. P14 - THE EFFECT OF DCNP ON AMYLOID BETA PEPTIDE TOXICITY AND PREGNENOLONE SULFATE SYNTHESIS. O.G. CALAN, P. AKAN, M. ORMEN, M. FADILOGLU (Department of Medical Biochemistry, Faculty of Medicine, Dokuz Eylul University, Izmir, Turkey)

Aim: There is growing epidemiological evidence that physical activity (PA) reduces the risk of dementia. We asked if this protective effect was present in the Interdisciplinary Longitudinal Study on Adult Development and Aging (ILSE) and thus modulated the risk of developing mild cognitive impairment (MCI) or Alzheimer's disease (AD). Methods: 500 participants of the ILSE were examined in 1993-1996 (t1), 1997-2000 (t2) and 20052008 (t3). In all examination waves, participants were carefully screened for physical and mental health through extensive medical interviews, physical and neuropsychological examinations. In addition at t1, all the participants completed a questionnaire about current and past PA and were asked to balance on one foot for 15 seconds as an objective index for PA. The participants were classified as physically active if they reported a regular sport activity since at least one year and at least once a week for at least 2 hours per week. Results: 381 participants of the cohort were re-examined at t3. Prevalence of MCI increased from 13% to 23% and 29% over time; in addition, 7% of the participants developed AD at t3. Subjects who passed the one-foot-balance-test at t1 showed a reduced risk for developing MCI or AD in the 12 year follow-up (odds ratio [OR] =0.38, 95% confidence interval [CI] =0.24-0.70, p<0.01) compared to healthy controls (scores adjusted for gender, education, socioeconomic status and depressive symptoms). Self-rated physical activity did not predict MCI/AD development in our sample. Discussion: Our findings confirm the hypotheses that PA constitutes a protective factor for the development of MCI and AD.

P12 - PHYSICAL ACTIVITY AS A PROTECTIVE FACTOR IN A PROSPECTIVE AGEING STUDY IN GERMANY. C. SATTLER, H. JILG, P. TORO, J. SCHRÖDER (Section of Geriatric Psychiatry, University of Heidelberg, Heidelberg, Germany)

Introduction: Recent studies have shown 17-estradiol exerts profound effects on the physiology and pathology of the central nervous system (CNS). Also there is a positive correlation between astrocytes apoptosis and rapid disease progression in persons with neurodegenerative disease. These cells are one of the most important cells of the "central nervous system" and They have important functions such as protection of neurons against injuries, ischemia, etc. Findings from other researchers suggest that GSK3- "glycogen synthase kinase-3", a key molecule of the Wnt signaling pathway. Aims: The aim of this study was to study the possible role of 17--estradiol on the expression of neuroprotective factors such as TGF1 and TGF2 in cultured astrocytes. Methods: 17 -esteradiol was

P11 - THE EFFECT OF 17--ESTRADIOL ON THE EXPRESSION OF NEUROPROTECTIVE FACTORS IN CULTURED ASTROCYTES. A. PIROUZI, M. MOHSENZADEH, M. JAFFARI, A. ABDOLLAHI, Z. YARAHMADI (Health Department, Gerash University of Medical Sciences, Gerash, Iran)

Currently available evidence suggests that in persons affected by Mild Cognitive Impairment (MCI), some biochemical, neuroimaging, electrophysiological, and neuropsychological markers of Alzheimer's disease might be sensitive to cognitive deterioration. The amount of evidence of validity is variable, some markers being backed by rigorous longitudinal studies, and others by merely suggestive cross sectional observations. Aim: We revised the literature on markers for AD progression based on their evidence of validity. Method: The strength of the evidence has been categorized into three classes: class A, markers studied by multiple longitudinal clinical studies; class B, data derived from a single serial study or multiple cross sectional studies of patients at different stage of Alzheimer's disease including MCI; and class C, data derived from multiple cross sectional studies of patients at different stages of Alzheimer's dementia, not including MCI. Results: The markers that have the greatest evidence of being good markers of disease progression based on multiple longitudinal studies (Class A) are markers of structural neuroimaging, Electrophisiology, CSF markers as well as neuropsychological markers. Discussion: The majority of the markers have been studied using cross sectional designs. For some markers the literature reports mixed results for some other the results are more consistent.

S13

Amyloid beta (A) is a 4 kDa amino acid peptide which can accumulate as insoluble extracellular deposits in senile plaque and has a causal role in pathogenesis of Alzheimer's disease. Neuroactive steroid hormones which are biosynthesized in brain or nervous system are called `neurosteroids' due to their extraordinary origins and different functions. Pregnenolone (P), the main precursor of the steroids, and its sulfate ester, pregnenolone sulfate (PS), are the major neurosteroids produced in the neural tissue. After synthesis of

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

P16 - SPOKEN LANGUAGE AS AN INDICATOR OF RISK FOR ALZHEIMER'S DISEASE: ANALYSES OF BIOGRAPHICAL INTERVIEWS OF A LONGITUDINAL AGEING STUDY IN GERMANY. B. WENDELSTEIN 1,2,3 , E. FELDER 2,3, J. SCHRÖDER 1,3 (1. Section of Geriatric Psychiatry, University of Heidelberg, Heidelberg, Germany; 2. Germanistisches Seminar, University of Heidelberg, Germany; 3. Marsilius-Kolleg, University of Heidelberg, Germany)

Aim: Analysis of the relationship between nutritional and functional, cognitive and psychological indicators in elderly patients with Alzheimer's disease. Method: The sample consisted of 83 free-living patients (29 men, 54 women) with mild-moderate Alzheimer's disease, aged 66 to 96 years, and 468 age-matched controls (202 men, 266 women). Nutritional status was evaluated by anthropometry (weight, skinfolds, circumferences, and body mass index, BMI), bioimpedance analysis (phase angle, vector length) and Mini Nutritional Assessment (MNA). Functional, cognitive and psychological status was assessed at the moment of the nutritional survey by the following tests: Mini Mental State Examination (MMSE), Geriatric Depression Scale (GDS), Activities of Daily Living (ADL), Instrumental Activities of Daily Living (IADL) and Alzheimer's Disease Assessment Scale (ADAS). Historical data were collected from the hospital database. Statistical analyses (Student's t-test and correlation) were performed using the R program. Results: On the basis of MNA, the patients with mild-moderate AD showed a low prevalence of malnutrition (no men, 6.9% of women) and a widespread risk of malnutrition (men: 27.0%, women: 51.4%). According to BMI cut-offs in the elderly, 3.4% of men and 7.4% of women were classified as malnourished (BMI<20 kg/m2), and 31.0% of men and 25.9% of women at risk of malnutrition (BMI<24 kg/m2). Compared to the control groups, patients with Alzheimer's disease showed similar MNA and BMI, and smaller skinfolds and phase angle (p<0.01). Women also had lower weight and waist circumference values (p<0.05). With progression of the disease, the anthropometric and MNA values diminished significantly. Cognitive and psycho-functional status was partly compromised (men: MMSE 19.4±5.6; GDS 3.3±2.9; ADL 4.3±1.3; IADL 1.8±2.0; ADAS 22.2±10.2; women: MMSE 19.1±4.9; GDS 4.2±3.5; ADL 3.7±1.6; IADL 2.0±1.7; ADAS 21.7± 11.0). With the exception of psychological functions (GDS), these capabilities showed a deterioration with time. A better functional, cognitive and psychological status (especially IADL) was associated with reduced anthropometric values. Discussion: As shown in the literature and particularly in a previous study of this sample (Buffa et al., in press), patients with Alzheimer's disease have a worse nutritional status than healthy controls (Gillette Guyonnet et al., 2007; Belmin et al., 2007) and their condition tends to worsen with progression of the disease (White et al., 1998; Ousset et al., 2008). However, a better functional, cognitive and psychological status was associated with smaller body dimensions. Previous studies showed that overweight and abdominal obesity are risk factors for Alzheimer's disease (Gustafson et al., 2003; Luchsinger et al., 2007; Whitmer et al., 2008). These insights suggest that patients with mild-moderate AD, although characterized by a tendency to malnutrition which increases with progression of the disease, are disadvantaged by excessive body mass.

P15 - NUTRITIONAL AND PSYCHO-FUNCTIONAL STATUS IN ELDERLY PATIENTS WITH ALZHEIMER'S DISEASE. B. SARAGAT 1 , R. BUFFA 1 , R.M. MEREU2, D. VIALE2, P.F. PUTZU2, E. MARINI1 (1. Department of Experimental Biology, A nthropological Science Section, University of Cagliari, Italy; 2. Geriatric Division, SS Trinità Hospital, A SL 8, Cagliari, Italy)

pregnenolone in mitochondria, it can be converted to its sulfate ester by hydroxysteroid sulfotransferase. The synthesis of P and PS was suggested as an early event of neuronal apoptosis and the inhibition of their synthesis has been shown to attenuate excitotoxicity in neuronal tissue. This work aims to investigate the effects of 2,4-dichloro-6-nitrophenol (DCNP), an inhibitor of hydroxysteroid sulfotransferase enzyme, on A peptide toxicity and steroid synthesis in neuronal cells. Rat pheochromocytoma (PC-12) and human neuroblastoma (SHSY-5Y) cell lines were used as the neuronal model. A 25-35 fragment which is the C-terminal peptide of A 1­42 located in its hydrophobic functional domain was used for the toxicity. The cell viability was evaluated with MTT reduction, LDH release and light microscopy. The changes of P and PS levels were also determined by HPLC method. All data were obtained from at least three separate examinations. After 72 hours(h) of exposure to A at 20µM, the viability of PC-12 cells significantly decreased compared with the control group(p<0.001). On the other hand, a significant decrease in viability of SHSY-5Y cells could be obtained with 72 h of exposure to 40µM A which was previously aggregated at 37°C for 3 days. The viability of A treated cells was significantly increased after 72 h exposure to 1µM DCNP (p<0.05). DCNP also decreased PS levels while had no significant effect on P levels in both cell lines compared with A treated and control groups. This study shows that DCNP exerts protective effect on A toxicity in PC-12 and SHSY-5Y cells. This effect may be due to the limiting role on the concentration of PS levels in neuronal cells. The possible role of sulfating processes in A toxicity needs further investigation.

P18 - LOOKING FOR ALZHEIMER'S DISEASE VESTIGE IN THE MEMORY OF PERIPHERAL BLOOD. F. CICCOCIOPPO 1,2, A. THOMAS 1,2, P. LANUTI 3,4, L. BONANNI 1,2 , D. MONACO 1,2 , S. BIFOLCHETTI 1,2 , M.C. D'AMICO 1,2 , I. BORRELLI1,2, M. MARCHISIO3,4, L. PIERDOMENICO3,4, D. GAMBI1,2, F. KERN5, S. MISCIA3,4, M. ONOFRJ1,2 (1. Department of Neuroscience and Imaging, University "G. d'A nnunzio" of Chieti-Pescara, Chieti, Italy; 2. Behavioural Neurology and Movement Disorders Unit, A ging Research Centre (Ce.S.I.), "Università G. d'A nnunzio" Foundation, Chieti, Italy; 3. Cell Signalling Unit, Department of Medicine and A ging Sciences, University "G. d'A nnunzio" of Chieti-Pescara, Chieti, Italy; 4. Citomorphology Unit, A ging Research Centre (Ce.S.I.), "Università G. d'A nnunzio" Foundation, Chieti, Italy; 5. Division of Medicine, Brighton and Sussex Medical School, Brighton, UK.

Aim: While decreased hemoglobin concentration is common in the elderly, the relationship of the level of hemoglobin concentrations with cognitive function is not well understood. Method: Cross-sectional analyses were conducted from community-dwelling, the elderly Korean. Participants were assessed with level of hemoglobin and they were administered to measure global cognitive function along with mini-mental status examination. Results: For 1,079 participants without clinical dementia, the mean age was 67.81 years (SD=4.93); 52.7% were women, and the mean level of education was 9.15 years (SD=4.76). The mean hemoglobin concentration was 13.33 g/dl (SD=1.40). Using linear regression models adjusted for age, gender, education and body mass index, low hemoglobin level was associated with lower global cognitive function (P=0.014). Discussion: Low hemoglobin concentration in the elderly are associated with a lower level of cognitive function in this study. Anemia is one factor to bear in mind when evaluating risk of cognitive dysfunction. If this relationship is causal, it is possible that treatment of anemia could prevent or postpone cognitive dysfunction.

P17 - THE RELATIONSHIP OF HEMOGLOBIN AND COGNITIVE FUNCTION IN THE ELDERLY KOREAN. M.H. PARK (Departm ent of N eurology , K orea University A nsan Hospital, A nsan, South Korea)

included subjects who developed AD as well as matched healthy controls. All of the subjects were cognitively healthy at t1. At t3 the subjects differed in cognitive diagnosis with AD or remained healthy. The linguistic analyses were based on biographical interviews collected at t1. The mean followup time between t1 and diagnosis at t3 was 12 years. The interviews of the subsample were transcribed and analyzed by "morphemes per t-unit". Results: Already at t1, when all patients have been cognitively healthy, differences show up in language production. In the complexity measure "morphemes per T-unit" a considerable difference with values 14.69 (SD = 2.5) for the AD patients and 18.75 (SD = 3.8) for the healthy subjects can be noted. Discussion: Spoken language shows to be a sensitive measurement to differentiate cognitive function. Differences in linguistic measures between healthy subjects and subjects which develop an AD can be noted years before the onset of first symptoms. This could be an important indication for early detection and diagnosis.

Accurate diagnosis of Alzheimer's Disease may have therapeutic implications. Current attempts to identify Alzheimer's disease biomarkers are principally focused on cerebrospinal fluid studies yet, B- and T-lymphocytes of Alzheimer's disease patients are involved in Amyloid-beta1-42 removal and in driving inflammation. As an alternative attempt to identify blood biomarkers of Alzheimer's disease we focused our attention on Amyloid-beta1-42-specific T-cells. We analyzed cytokine production and Protein-Kinase C phosphorilation of in vitro Amyloid-beta1-42-stimulated T-cells in patients with Alzheimer's Disease, Lewy Body Dementia, Inclusion Body Myositis, Cerebral Amyloid Angiopathy and in healthy controls. We show that a subset of memory T-cells characterised by bright expression of Phosphorylated-Protein-Kinase C delta and zeta is specific to Abeta1-42 and distinguishes Alzheimer's Disease Inclusion Body Myositis (also characterised by Amyloid-beta1-42 deposition) from controls and other neurodegenerative pathologies where amyloid deposition consistent of different peptides or unshareprominent inclusion do not consist of amyloid. This marker could be tested to facilitate the diagnosis of Alzheimer's disease and to discriminate from Lewy Body dementia and other neurodegenerative conditions. P19 - CHOLESTEROL PROFILE IN ALZHEIMER'S DISEASE. J. JAFARBAY 1, H. ANGLARET2, M. SABER1, J.J. ARVIEU1 (1. Centre hospitalier de Gériatrie « Les A bondances » 56, rue des A bondances, 92100 Boulogne Billancourt. France; 2. Hôpital Saint Cloud, France)

In the present study language in preclinical stages of Alzheimer's disease (AD) was analyzed. The aim was to identify indicators for healthy ageing and pathological ageing on the surface of language. It was observed whether there are differences in preclinical stages between subjects which develop AD and subjects which remain healthy. Methods: Spoken language was examined on the basis of semi standardised biographic interviews of the Interdisciplinary Longitudinal Study on Adult Development and Aging (ILSE). The ILSEstudy includes a birth cohort born between 1930 and 1932 which was observed at three examination waves so far (t1: 1993/94, t2: 1997/98, t3: 2005/07). The subsample (n = 6)

S14

Introduction: Links between lipid profile and dementia (Alzheimer, vascular,...) are still unclear. A correlation between cholesterol and depression is suggested in some studies. Objective: The aim of this study was to investigate the serum level of cholesterol (HDL-C, LDL-C) in patients with cognitive impairment (Alzheimer or other dementia) and patients with depression. Methods: We measured the serum level of cholesterol (highdensity lipoprotein HDL-C, lower-density lipoprotein LDL-C) in fasting samples of 43 patients with dementia (MMSE <19 ). 13 of them had possible Alzhiemer's disease (AD). 23 patients had been treated for depression. Patients with thyroid and renal deficiency were excluded of this study. Results: The serum levels of LDL-C, and HDL-C were slightly higher in the patients with AD (mean of serum level LDL-C=2.88 mmol/l, HDL-C =1.51mmol/l) in comparison of other patients (mean of level: LDL-C =2.69 mmol/l, HDLC= 1.06 mmol/l). We didn't observe any significant difference of cholesterol level in patients with depression compared to patients with behavioural and psychological

Background and Purpose: New research criteria for subcortical vascular dementia (SVaD) have been suggested to define a more homogeneous subgroup of vascular dementia. Besides the defining features of cerebrovascular disease, hippocampal (hc) atrophy has been found in vascular dementias. So far, it is unknown to which extent hc atrophy is present in SVaD. Methods: Eleven patients fulfilling the research criteria for SVaD were carefully matched with comparison groups of healthy controls and patients with Alzheimer's disease (AD). To estimate the extent of hc atrophy in SVaD, both hc volumetry and visual rating of medial temporal lobe atrophy (MTA) were applied. Results: In SVaD, significant hc atrophy occurred. The extent was intermediate between controls and patients with AD, both on hippocampal volumetry and visual MTA ratings. At the same level of global cognition, hc volumes were reduced by 11.6% in SVaD and 16.6% in AD, relative to controls. Conclusions: Hippocampal atrophy is an important feature in cognition disorders of both vascular and neurodegenerative origin. Further studies are needed to elucidate the mechanisms of hc atrophy in SVaD. In future trials on SVaD, the degree of hippocampal atrophy should be taken into account in addition to the defining features.

P20 - HIPPOCAMPAL ATROPHY IN SUBCORTICAL VASCULAR DEMENTIA. L. VAN DE POL1, H.-J. GERTZ2, P. SCHELTENS1, H. WOLF2,3 (1. Department of Neurology and A lzheimer Center, V U University Medical, Center A msterdam, PO Box 7057, 1007 MB A msterdam, The Netherlands; 2. Department of Psychiatry, University of Leipzig, Memory Clinic, Johannisallee 34, 04103 Leipzig, Germany; 3. Department of Psychiatry Research and Geriatric Psychiatry at the Psychiatric University Hospitals (PUK), University of Zurich, Switzerland)

disorders. Conclusions: Results of this study showed a slightly higher level of cholesterol in patients with AD. We didn't observe any correlation between cholesterol and depression in this population. Further studies are required to confirm and understand this relation.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Aim and background: Apathy is the most common neuropsychiatric disturbance in Alzheimer's Dementia (AD). It is defined by loss of motivation, diminished social interaction and reduced emotional responsiveness. Today, there is no high-level evidence for any effective treatment of apathy in AD. Reduced dopamine transmission contributes to apathy in AD. The antidepressant Bupropion is a dopamine reuptake inhibitor. This study is designed to test the safety and efficacy of Bupropion in the treatment of apathy in Alzheimer's dementia. Method: During a time period of 21 months 270 AD patients will be assessed for eligibility in 11 centers in Germany specialized in diagnosis and treatment of Alzheimer's Disease. The key inclusion criteria are for the male and female participants with mild to moderate Alzheimer's dementia (MMSE 10-23) are the presence of clinically relevant apathy defined by the Neuropsychiatric Inventory (NPI) apathy item (score of >/= 4 points) and the revised Marin/Starkstein criteria for apathy. The participants should be 55-90 years old and still living at home with a reliable caregiver. The treatment with antidementia drugs should be stable for at least three months prior to entry. The key exclusion criteria are the presence of a clinically relevant depressed mood defined by either the NPI depression item (score >/= 4 points) or DSM-IV criteria for major depressive episode (with depressed mood). Further exclusion criteria are for example the treatment with antipsychotics, antidepressants, dopaminergic agents or MAO inhibitor. The participants which fulfill the eligibility criteria will be allocated to either the verum (n=108) or placebo (n=108) arm by central randomization. The randomization will be stratified with regard to the use of Donepezil/Galantamine or no use of either drug, because Bupropion may affect plasma levels of both. The treatment period for the individual will take 12-weeks with a once-daily oral intake of the dopamine and norepinephrine reuptake inhibitor Bupropion extended Release or placebo in a doubleblind fashion. The overall design of the trial follows a pragmatic approach with flexible dosing of the drug under investigation to reflect the "real-world" treatment situation of AD patients with apathy. According to this approach the initial dose of 150 mg Bupropion can be increased after four weeks to 300mg, which will be the final dose throughout the trial. If the participant does not tolerate 300mg, he/she may be down titrated to 150mg (1 tablet), which will then be the fixed dose throughout the study. Downtitration must be done at week six at the latest. The aim is to analyze a total of 172 patients for the change in the Apathy evaluation scale (AES). The Apathy Evaluation Scale (AES) has been constructed to assess apathy in clinical populations and consists of 18 items. The AES-C is based on an interview with the caregiver by the clinician. Since the AES assesses apathy in great detail and is constructed

P21 - STUDY DESIGN OF THE "12-WEEK, MULTICENTER, RANDOMIZED, DOUBLE-BLIND, PLACEBO-CONTROLLED TRIAL OF BUPROPION FOR THE TREATMENT OF APATHY IN ALZHEIMER'S DEMENTIA (APA-AD)". A. SPOTTKE1,2, M. MAHNE5, H. STÜTZER6, K. BÜRGER7, R. DODEL8, A. HEINZ9, I. HEUSER 10 , H. JAHN 11 , J. KORNHUBER 12 , O. PETERS 10 , J. SCHRODER 13 , J. WILTFANG14, M. HÜLL4, L. FRÖLICH3, F. JESSEN2 (1. Department of Neurology, University of Bonn, Germany; 2. Department of Psychiatry, University of Bonn, Germany; 3. Central Institute for Mental Health, Mannheim, Germany; 4. Center for Geriatric Psychiatry, University of Freiburg, Germany; 5. Center for Clinical Trials Cologne, Germany; 6. Institute of Medical Statistics, Informatics and Epidemiology, University of Cologne, Germany; 7. Institute for Stroke and Dementia Research, University of Munich, Germany; 8. Department of Neurology, University Marburg, Germany; 9. Department of Psychiatry, Charité Campus Mitte, Germany; 10. Department of Psychiatry, Charité Campus Benjamin Franklin, Germany; 11. Department of Psychiatry, University of Hamburg, Germany; 12. Department of Psychiatry, University of Erlangen, Germany; 13. Department of Psychiatry, University of Heidelberg, Germany; 14. Department of Psychiatry, University of Essen, Germany)

Aim : Patients with amnestic mild cognitive impairment (aMCI) are characterized by memory complaints and impaired performance on memory tasks while general cognition and functional ability in basic everyday tasks are preserved (Petersen 2004). Although they do not meet the criteria for dementia, 20­50% of these persons will develop dementia over a period of 2­3 years. Hence, aMCI appears to capture a group of individuals in a transitional state between normal aging and Alzheimer Disease (AD). While several clinical trials for symptomatic drug treatment have been proposed for these patients, few studies have explored the effect of cognitive intervention in aMCI. Belleville et al. (2006) showed an improvement of aMCI's performance on episodic memory when the intervention focused on teaching episodic memory strategies. Greenaway et al. (2008) suggest that with appropriate training, aMCI can use a memory notebook system to help compensate for memory loss and functional difficulties. These studies encourage towards further investigation, especially as the direct benefit on daily functioning and the long-term effects of these interventions have not been evaluated. In this preliminary study, we hypothesis that combination of using a memory notebook with applying cognitive interventions, will reduce functional impacts of cognitive deficits in aMCI patients'everyday life. Method : A cognitive intervention adapted from Belleville (2006) was administrated in seven weekly sessions to a group of 5 aMCI patients (age : 72,7 ±8,42; MMSE : 27,2 ±1,48). Each session lasted 2 hours and consisted on teaching episodic memory strategies. In addition, psychoeducation basic knowledge was taught. To assess the effect of our program, cognitive and functional outcome measures were collected at three time-delays : one week pre-intervention (baseline) and the first and sixth weeks post-intervention (assessments 1 and 2). To enhance the ecological aspect, a memory notebook was used during the sessions. Furthermore, homework were enriched, to allow the patients making a more direct link between the learning of the memory strategies and their use in the everyday life. Results : The intervention effect (differences between pre- and post-intervention cognitive and functional measures) did not reach statistical significance. Nevertheless, results suggest that 3 patients improved their performances on episodic memory tasks 6 weeks after the intervention and 4 patients utilized memory notebook. Furthermore, we observed an increase of their memory complaints and modification of their metacognition. Discussion : This preliminary study allowed to examine the feasibility of a multifactorial cognitive training in a group of patients with aMCI. Results tend to suggest a long-term improvement both at cognitive and functional levels. Studies conducted with larger groups of patients and during longer periods will be necessary to evaluate the long term benefits of such intervention. P23 - MANAGING RATER DRIFT AND RATINGS ERRORS WITH IN-STUDY RATINGS SURVEILLANCE. D.S. MILLER, P. SAMUELSON, J. CARPENTER, K. FOULDS (United BioSource Corporation, W ayne, PA , USA )

P22 - MULTIFACTORIAL COGNITIVE TRAINING INTERVENTION IN aMCI. C. LATGER-FLORENCE 1 , E. TRAMONI 1,2 , C. ARNAUD 1 , E. DA FONSECA 1 , M. CECCALDI1,2 (1. Service de Neurologie et de Neuropsychologie, CHU Timone, Marseille, France; 2. Laboratoire Epilepsies et Cognition, INSERM U 751, Faculté de Médecine, Marseille)

as a full instrument for the assessment of apathy, it is chosen as the primary endpoint rather than the NPI apathy item. Secondary endpoints will be quality of life of patients, caregivers' distress, ability of patients to perform activities of daily living, utilization of heathcare resources by patients and by caregivers, and cognitive functions. Discussion: The Apa-AD trial is the first high-quality trial for the treatment of apathy in AD. If Bupropion is safe and effective it will be the first evidence -based treatment option for this most frequent neuropsychiatric symptom in AD.

S15

Aim: The Alzheimer's Disease Assessment Scale ­ Cognitive subscale (ADAS-Cog) is the most commonly used primary efficacy measure in Alzheimer's disease clinical trials. In general, the (11 item) ADAS-Cog has demonstrated sensitivity to change across a range of dementia severities. Some clinical research studies now include some additional items (eg ­ delayed word recall, maze test and digit cancellation) hoping to better assess the early stages of dementia as well as executive functioning. Earlier research has demonstrated 1) considerable variability in how raters have been trained to administer and score the (11 item) ADAS-Cog in clinical trials, and 2) that the introduction of these additional items appears to provide additional challenges to raters who attempt to qualify to rate the scale in studies where they are used. We evaluated the impact of an In-Study rater surveillance program, which identifies rater errors in scale administration and scoring, and remediates raters when those errors occur, on the raters' subsequent performance in the course of a clinical trial. Method: Two separate, multi-national Alzheimer's disease clinical trials where such a customized surveillance program was employed were evaluated to assess the program's impact. All raters received extensive didactic training on the proper administration and scoring of the ADAS-Cog and also had to demonstrate their competency to rate by successfully watching and scoring a video of the scale being administered to a patient. Rater performance (proper ADAS-Cog administration and scoring) was assessed by worksheet review at 2 time points ­ baseline and 1 year. Results: When the 2 studies are combined, to date a total of 2,410 patients have been assessed at baseline, and 781 patients have progressed to the second assessment time point. 772/2410 (32%) of the baseline worksheets required a contact, and of these, the majority (60%) required remediation on administration and/or scoring technique, rather than a clerical or math error. At the second assessment, the number of assessments requiring a contact was markedly reduced to 68/781 (9%), and the proportion requiring remediation on administration and/or scoring technique, was also reduced to 49%. When the studies are separated, a similar pattern exists ­ in one the number of assessments decreased from 36%

Background: A number of compounds with potential to attenuate or stop the progression of Alzheimer disease (AD) from its pre-symptomatic stage to dementia, socalled disease-course altering drugs, are being or will soon be tested in man. The conventional study design is the randomized, placebo-controlled clinical trial (RCCT), which implies that a high proportion of patients will receive placebo treatment. Considering that studies in this indication last two years or longer, it is ethically problematic to expose patients with pre-symptomatic AD, who by definition run a very high risk of developing dementia, to prolonged placebo treatment. Aim: To develop an ethically acceptable and scientifically sound alternative to long-term RCCT designs with placebo, for use in studies of novel drugs with possible disease-altering properties in patients with pre-symptomatic AD. Method: Making use of existing clinical, psychometric and biological data, we are developing mathematical algorithms that reliably predict disease trajectories of pre-symptomatic AD patient groups. These quantified trajectories constitute the background ("virtual placebo group") against which potential drug effects can be tested. The example presented uses the data base of the Alzheimer's Disease Neuroimaging Initiative (ADNI), with a focus on 397 MCI subjects followed for up to 36 months. One (univariate) outcome tested was the modified ADAScog score after 24 mo, the other (multivariate) one the trajectory over 36 mo of a compound neuropsychological test score (NP-Batt) based on the averaged z-scores of 9 subtests. We tested baseline values of these measures, demographics (age, gender, years of education, BMI, number of APO E4 alleles), Hachinski modified score, a functional assessment score (FAQ) and the MMSE

P25 - PREDICTING COGNITIVE DECLINE IN PATIENTS WITH PRESYMPTOMATIC AD: THE PLACEBO GROUP SIMULATION APPROACH. R. SPIEGEL, M. BERRES, A.U. MONSCH, A.R. MISEREZ (Memory Clinic, Geriatric University Hospital, Basel, Switzerland)

Aim: Intra individual variability (IIV) is sometimes regarded as measurement error or `noise', obscuring detection of "true change" in Alzheimer's disease (AD) and mild cognitive impairment (MCI), and for their early detection and monitoring. IIV is most often based on response time tasks, but these types of tasks are not often used in clinical studies of AD or MCI. There are also multiple definitions of IIV, complicating the selection and use of IIV as a performance summary. IIV may reflect neuropathology independent of the sex or education effects that are characteristic of common tests. Before the advantages of IIV can be realized, a single definition of IIV must be selected. The "best" IIV formula would be the one that is most sensitive to neuropathology, possibly contributing independent information beyond the total test score(s). This study explored different definitions of intra-individual variability (IIV) to summarize performance on common cognitive tests and examined their utility in predicting clinical change in individuals from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and neuropathology values from the Religious Orders Study and Memory and Aging Project, two longitudinal cohort studies in the US. Method: Sample statistics (mean, variance) were obtained from the individuals participating in the studies with no cognitive disorder diagnosis ("normal"), with a diagnosis of mild cognitive impairment ("MCI") and those with a diagnosis of possible or probable Alzheimer's disease ("AD"). These groupings, with their different levels of (observable) disease burden, serve as proxies for levels of structural brain integrity; the longitudinal cohorts also had over 300 neuropathology results each (including Braak and CERAD staging and a global neuropathology score). Using these sample statistics as `population' parameters, 500 replicate `samples' were simulated from each population, all of size n=200. Using the 500 replicates, standardized mean difference effect sizes (Cohen's d) were computed creating sampling distributions of 500 effect size (ES) estimates. These ES values were computed based on comparison of samples generated as `normal' and `MCI', and comparison of samples generated as `MCI' and `AD'. Effect size estimates were computed using the items of the Mini-Mental State Exam (MMSE), a Clock-Drawing test, and the ADAS-Cog, separately (`within test' IIV) and using the MMSE & Clock-Drawing test total scores together (`across test' IIV). ES values were compared across method of ES estimation qualitatively or by t-test. Results: ES estimates differed substantially depending on the definition of IIV and the test(s) on which IIV was based. Four-year change in IIV was associated with neuropathology as strongly as total (MMSE) or global (composite) scores, but were not associated with education (total scores were). Discussion: This study estimated the degree to which intraindividual variability (IIV), defined as individual standard deviation, coefficient of variation, or level-independent variation- is sensitive to group differences as compared to the standard, total score-based summary of test performance. IIV can be computed within (items) or across (totals) items on commonly-utilized cognitive tests, and may be a useful summary of neuropsychological test performance. However, the definition of IIV will affect the utility (ES).

P24 - INTRAINDIVIDUAL VARIABILITY & NEUROPATHOLOGY FOR STUDY OF AD, MCI AND BRAIN AGING. R.E. TRACTENBERG (Departments of Neurology, Biostatistics, Bioinformatics & Biomathematics, and Psychiatry; Georgetown University Medical Center, W ashington DC, USA )

to 9%, and in the other, from 25% to 5%. Discussion: There is a trend to look at the disease modifying ability of potential Alzheimer's disease therapies. As a result, trials are of longer duration, and the likelihood of rater drift increases. The use of a customized In-Study rater surveillance program can both identify instances where raters deviate from proper ADASCog administration and/or scoring technique, and by remediating raters in question, can effectively decrease the likelihood of errors recurring over the course of an Alzheimer's dementia clinical trial. Such a program can also be effectively utilized in multi-national trials. Given the apparent effectiveness of this type of program, its use is bears consideration.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

P26 - SAPP IS SIGNIFICANTLY CORRELATED WITH TAU PROTEIN IN THE CEREBROSPINAL FLUID: A LINK BETWEEN -AMYLOID AND TAU PATHOLOGY IN ALZHEIMERS'S DISEASE? P. ALEXOPOULOS 1,2 , A. TSOLAKIDOU1, F. ROSELLI3, T. EISELE1, C. WESTERTEICHER1, A. ARNOLD1, H. FORSTL2, A. KURZ2, R. PERNECZKY1,2 (1. Neurobiological Laboratory, Department of Psychiatry and Psychotherapy, Technische Universität München, Munich, Germany; 2. Department of Psychiatry and Psychotherapy, Technische Universität München, Munich, Germany; 3. Department of Neurological and Psychiatric Sciences, University of Bari, Bari, Italy) Introduction: Two of the pathological hallmarks of Alzheimer's Disease (AD) are the extracellular deposits of -amyloid (A), including products of the amyloid cascade, and the intracellular neurofibrillary tagles, which are formed by abnormally phosphorylated tau protein. The proteolytic breakdown of the amyloid precursor protein (APP) by -secretase generates the - secretase cleaved soluble APP (sAPP) and the peptide C99. A is produced by the proteolysis of C99 by -secretase. Recently evidence on the toxicity of soluble A oligomers has accumulated. Since in contrast to several types of A oligomers, sAPP is not prone to aggregation and due to the fact that it can be detected in the cerebrospinal fluid, its levels in CSF reflect the extent of APP processing by - secretase, as well as the production of soluble A oligomers. It has been hypothesized that the production of -amyloid leads to the hyperphosphorylation of tau, resulting subsequently to axonal degeneration. However the link between the amyloid cascade and the tau pathology remains elusive. Aims: The aim of the present study was to investigate the relationship between sAPP and total tau concentrations in the CSF of patients with mild dementia in AD and Mild Cognitive Impairment (MCI), as well as in cognitively healthy individuals. Methods: CSF was obtained from 76 patients with MCI, 64 with mild dementia in AD and 13 healthy controls. sAPP and total tau concentrations were determined with commercially available ELISA kits. Normal distribution of data was checked using the Kolmogorov-Smirnov test.A parametric correlation analysis was employed. P values of less than 0.05 were considered statistically significant. Results: There was a significant positive correlation between sAPP and total tau CSF levels in all groups of participants (in all cases P<0.0001). Conclusions: Though not establishing any straightforward facilitatory causal effect of sAPP on tau hyperphosphorylation, the observed positive correlation between sAPP and total tau in the CSF offers a link between the amyloid and the tau pathology of the AD. P27 - PERIVENTRICULAR WMHS ARE RELATED TO VASCULAR RISK-- EVEN IN ADNI. C. DECARLI, D. HARVEY, L. BECKETT, C. SCHWARZ, D. DRUCKER, E. FLETCHER, O. CARMICHAEL (Imaging of Dementia and A ging laboratory , Departm ent of Neurology and Center for Neuroscience, Univ ersity of California at Davis, Davis, California, USA )

as potential predictors. In the univariate analysis we predict the square-root transformed ADASCog after 24 mo from predictor variables in a linear regression model. Variable selection is based on the Akaike Information Criterion (AIC) and, after the main effects model is found, all pairwise interactions are tested for inclusion in the model. The longitudinal model is a mixed model with random intercepts and random slopes versus time. It is computed for the NP-Batt at all follow-up visits. Results: ADASCog after 24 mo can be predicted from Gender, Obesity, FAQ, MMSE and the baseline scores of ADASCog and NP-Batt with an R-square of 0.63 and a residual SD of 0.67. This allows reasonably precise estimates of sample means. The longitudinal model of the NP-Batt has random intercepts and slopes und fixed effects for BMI, time, APOE4, age, FAQ, baseline scores of ADASCog and NP-Batt and four interaction terms. Estimates of the residual SD range from 0.3 to 0.5 on a standard normal scale. Treatment effects are expected to diminish the negative slope of scores with time. A change of 0.04 per year can be detected in samples of 400 (real vs. virtual) with a power of about 80%. Discussion: Our first predictive cross-sectional and longitudinal models based on ADNI data show high correspondence of predicted and real observed values. Corroboration of these models with data from other studies is ongoing. It is hoped that the "Placebo Group Simulation Approach" will replace or complement the current placebo-based study designs in conditions such as pre-symptomatic AD. Although existing regulatory guidelines appear to support RCCTs with placebo control even for patients with slowly evolving, irreversible and ultimately fatal disorders such as AD, other - ethically more acceptable - study designs need to be considered in the future.

S16

Aim: To examine whether regional WMH is a viable biological marker for vascular brain injury in ADNI, a clinical trial cohort with an extremely mild profile of vascular risk. Methods: A total of 409 subjects were selected from the ADNI cohort based on the presence of Hachinski Scale Score (HSS), CSF measures of Tau and amyloid beta and MRI-based WMH measurement at baseline. Subject mean age was 75.6 + 7 years, 40% were female, 25% were AD, 48% MCI and 28% normal controls. WMH were measured from MRI as previously reported (1) and placed in a common template space for regional parcellation of WMH burden as previously reported (2). Multiple linear regressions were used initially to predict log-transformed WMH volumes (lWMH) in anterior, middle and posterior periventricular ROIs with independent variables of age, gender, HSS (as ordinal score), diagnosis, ApoE4 genotype, and CSF Tau and CSF abeta concentrations. Confirmation of the results used analogous zero-inflated Poisson (ZIP) modeling of WMH volume transformed into a count variable (cWMH). Results: HSS scores varied from 0-3 with 50% having a score of 0 and an additional 40% having a score of 1. Using the multiple regression approach, HSS (p=0.0058) and age (p=0.03) were significant predictors of anterior WMH. HSS (p=0.03), age (p=0.0056) and CSF Tau (p=0.027) were significant predictors of middle WMH. Only age (p <0.0001) was significant predictor of

Aim: It has been shown in several studies that the agreement between clinical diagnosis and true Alzheimer pathology can vary significantly between different studies and between different clinical sites and may mask a good agreement between a diagnostic test and true pathology. The implications of using an imperfect reference standard when assessing biomarker performance have to be taken into consideration. The aim of this presentation is to provide support for a view that agreement between a disease prediction of an Alzheimer's disease (AD) test and true pathology might be significantly higher than what an agreement with clinical diagnosis indicates. Methods: We have developed a blood test (ADtect®) to aid in the early diagnosis of AD. Our unique approach detects the presence of disease by measuring the pattern of specific gene expression changes in peripheral blood cells. This is a novel diagnostic approach as it measures the systemic effects of the disease. The performance of each of 96 the gene assays in a real time RTPCR analysis is evaluated by an algorithm that results in a test score indicating the presence or absence of AD. Results: In a multicenter study of 412 subjects the test is able to discriminate AD subjects from cognitively healthy controls with a 72% overall agreement with clinical diagnosis. Since clinical diagnostic routine differs between clinics within countries and between countries the agreement of ADtect® between clinical sites and countries have been compared. We find clear differences in test agreement with clinical diagnosis between sites (70-96%) and the agreement of the diagnosis overall at the Swedish clinics are significantly higher (85.5%) than what is found at the Norwegian clinics (69.5%). Also, the tests agreement with biomarkers in CSF was compared. Of 18 samples were data from CSF biomarkers was available the prediction was the same for 15 samples. Together these data suggest that the accuracy of the clinical diagnosis to which the tests performance is compared can be a significant confounding factor when estimating a novel test's accuracy. Assuming a clinical accuracy of 80% in the multicenter study, suggests an overall test accuracy of 85-90%. The test performance is confirmed in two independent validation studies separated in time with a year (72% accuracy in both), and shows similar performance in both mild and moderate AD cases. Discussion: Our current findings indicate that the AD-specific gene signature is a reliable and diagnostically relevant biomarker for the early detection of AD. The high agreement of ADtect® with CSF biomarkers and its higher performance on samples from Swedish clinics where the clinical diagnosis likely is more accurate supports the calculations that the agreement between ADtect® and true AD pathology likely is well above 80%. ADtect® is particularly valuable as an aid in the diagnosis of mild AD cases with minor cognitive decline, the most difficult cases clinically to diagnose.

P28 - IMPACT OF CLINICAL ACCURACY IN DEVELOPMENT AND VALIDATION OF A BLOOD BASED TEST FOR EARLY DETECTION OF ALZHEIMER'S DISEASE. A. LÖNNEBORG, G. GRAVE, B.B. BOOIJ, P.D. RYE, L. KRISTIANSEN, T. LINDAHL, P. SHARMA (Diagenic A sa, Oslo, Norway)

posterior WMH. Using the ZIP model, an HSS of 1 was associated with reductions in odds of 39% and 50% of having exactly zero WMH volume in the anterior and middle regions, compared to an HSS score of 0 (OR=0.50, 95% CI: 0.28, 0.89, p=0.02). Among individuals with non-zero regional WMH volumes, increased HSS was associated with increased WMH volume in each of the three regions. A HSS of 1 was associated with increased WMH volumes of 46%, 43%, and 41% compared to an HSS of 0 in anterior, middle, and posterior regions (ratios = 1.46, 1.43, 1.41; 95% CIs: [1.21, 1.77] [1.11, 1.84] [1.25, 1.59]; p<0.001, p=.02, p<0.001). Conclusion: We have previously shown that WMH volumes are associated with cross-sectional and 1-year longitudinal changes in ADAS-Cog, MMSE, and CDR scores amongst the full ADNI cohort (3). This study finds that the extent of WMH measures is associated with HSS independent of ApoE genotype and CSF Tau and amyloid beta measures in ADNI. We interpret these results in two ways. First, vascular disease, even when crudely measured by the HSS, is associated with WMH in ADNI, a research trial designed to represent clinical trials of AD therapeutics. This is true even though ADNI participants were intentionally selected to have a relatively mild profile of vascular risks as indicated by an HSS of 3 or less. Second, periventricular WMH are associated with HSS independent of typical AD markers such as ApoE4 genotype, CSF Tau and amyloid beta, and diagnosis. We believe these results suggest that periventricular WMH may be a biological marker of vascular brain injury sufficient for further study in clinical trials of dementia prevention therapies. References: 1. Schwarz C, Fletcher E, DeCarli C, Carmichael O. Fully-automated white matter hyperintensity detection with anatomical prior knowledge and without flair. Inf Process Med Imaging. 2009;21:239-251; 2. Yoshita M, Fletcher E, Harvey D, Ortega M, Martinez O, Mungas DM, Reed BR, DeCarli CS. Extent and distribution of white matter hyperintensities in normal aging, mci, and ad. Neurology. 2006;67:2192-2198; 3. Carmichael O, Schwarz C, Drucker D, Fletcher E, Harvey D, Beckett L, Jack CRJ, Weiner M, DeCarli C, Initiative. tADN. Longitudinal changes in white matter disease and cognition in the first year of the alzheimers disease neuroimaging initiative. . Arch Neruol. 2010; In Press

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

30.000 inhabitants town, in the province of Milan. Method: Starting at the end of November the population of the selected years was invited to participate at the comprehensive assessment: a first appointment for blood sample, social questionnaire, anthropometric and walking speed evaluation (taking about 1 hour and half); a second one for clinical examination and neuropsychological assessment of mood and cognitive function (2 hours). The instruments are partly derived from the CERAD and from other longitudinal study. Trained psychologists administer a neuropsychological battery exploring mood (GDS 15 items), verbal and visual memory (Rey 15-item Memory Test, and the Rey-Osterrieth Complex Figure Test (ROCF) recall; Babcock Story Recall Test ), executive function ( TMTA and TMTB; copy of ROCF), attention (Numerical Attention), abstract reasoning (Raven's Coloured Progressive Matrices), screening cognitive test (Clock test and MMSE). Clinical interview and visit were executed by expert geriatricians, members of the same geriatric staff. All the instruments were tested for inter rater reliability in a similar population attending a geriatric day hospital. The participation was solicited for any single age class, an identity mark for people in that area, using: 1 ­ one meeting for each class with an explanation of the aims and methods of the research , followed by a party with music, theater or some game. 2 - one letter with the date of the appointment already fixed 3- a phone call for everyone whose phone number was available 4- a second letter and a second call 5 ­ Finally involvement of the family doctor. Results: At the end of June 2010 , with the exclusion of 1939 born subjects not yet included in the evaluation, almost a thousand persons were fully evaluated with this percentage of recruitment : Fully assessed refuse absent To recall Wrong address ; no phone number total 229 44 11 15 0 1935 n 76,6 14,7 3,7 5 100 % 212 45 11 25 0 1936 n 72,4 15,3 3,8 8,5 100 % 243 41 33 35 3 355 1937 n % 68,45 11,55 9,30 9,85 0,85 100 1938 n % 271 24 65 33 3 396 68,44 6,06 16,41 8,33 0,76 955 154 120 108 6 total

P29 - INVECE.AB : A MULTIDIMENSIONAL POPULATION STUDY ON BRAIN AGING FOR THE 1935 ­ 1939 BORN PEOPLE, IN A SMALL TOWN NEAR MILAN. A GUAITA 1 , G.L. FORLONI 2 , V. FERRETTI 3 , S. VILLANI 3 , S. FOSSI 1 , M. COLOMBO 4 , G. SALVINI PORRO 5 (1. Golgi Cenci R esearch Foundation, A bbiategrasso, Milano (I); 2. Mario Negri Institute of Pharmacological Research, Milan (I); 3. Department of Health Sciences, Section of Epidemiology and Medical Statistic Pavia University, Pavia (I); 4. Geriatric Institute "Camillo Golgi", A bbiategrasso (I); 5. Federazione A lzheimer Italia, Milan (I)) Aim : To study cognitive function, physical health and biological parameters in the whole population of 70 ­ 75 years of age, 1770 subjects living in Abbiategrasso, a

S17

Aim: Given the multifactorial nature and the occurrence of several subgroups of Alzheimer disease (AD), without a reasonable understanding of the etiopathogenic mechanism involved in each disease subgroup the outcome of therapeutic clinical trials of drugs that inhibit a specific step in neurodegeneration is most uncertain. Thus, one therapeutic approach to this heterogeneous disease is to shift the balance between neurodegeneration and neurogenesis along with enhanced neuronal plasticity by employing the regenerative capacity of the brain. The aim of this study was to examine the feasibility of this approach in a transgenic mouse model of AD. Methods: The triple transgenic mouse model of AD, the 3xTg-AD which harbors PS1M146V, APPSWE, and tauP301L mutations, is believed to be one of the most biologically relevant animal models of AD described so far. We administered peripherally in these 6­7 month female 3xTg-AD mice Peptide 6, an 11-mer corresponding to an active region of Ciliary Neurotrophic Factor (CNTF). Peptide 6 is blood-brain barrier permeable and enhances neurogenesis through CNTF signaling by inhibiting Leukemia Inhibiting Factor (LIF) activity. We investigated the effect of the treatment with Peptide 6 for six weeks on neurogenesis, neuronal plasticity and neurobehavior. Results: Treatment with Peptide 6 had no effect on body weight and did not show any detectable effects on motor coordination or motoricity in 3xTg-AD or non-transgenic control animals matched for strain and genetic background, the 129/SVX C57BL/6 mice. By the time of ~8­9 months when the animals were sacrificed for immunohistochemical and biochemical studies, they showed no A plaques or neurofibrillary tangles in their brains. Intraneuronal accumulation of A1­42 and phosphotau were observed in 3xTg-AD mice. Treatment with Peptide 6 had no effect either on these early stages of A and tau pathologies. Peptide 6 enhanced dentate gyrus neurogenesis and inhibited ectopic birth in the granule cell layer. It increased neuronal plasticity in the hippocampus and cerebral cortex as measured by increase in MAP2 and synaptophysin. The treatment with Peptide 6 inhibited impairments in object discrimination task as well as in spatial reference memory task in 3xTg-AD mice. Conclusions: These findings showed (1) that neurogenesis, neuronal plasticity and cognition are impaired in 3xTg-AD mice several months before the formation of any A plaques or phosphotau neurofibrillary tangles, and (2) that it is feasible to pharmacologically shift the balance from neurodegeneration to regeneration of the brain and restore cognition in 3xTg-AD mice. AD therapeutic trials with this approach ought to be considered. (Supported in part by the New York State Office of Mental Retardation and Developmental Disabilities and a research grant from EVER NeuroPharma GmbH,

P30 - PHARMACOLOGICAL ENHANCEMENT OF NEUROGENESIS AND NEURONAL PLASTICITY: A NEW THERAPEUTIC APPROACH AWAITING CLINICAL TRIALS. I. GRUNDKE-IQBAL, J. BLANCHARD, K. IQBAL (New Y ork State Institute for Basic Research in Development Disabilities, Staten Island, New Y ork, USA )

Discussion : the main interest of this study can be summarized as follow : 1) a population not a champion study 2) the single small area involved 3) the age, a crucial moment of transition from late adulthood to the old age 4) the multidimensional approach able to asses, social, cognitive, medical and biological parameters. This first step is meant as the beginning of a longitudinal evaluation with at least 6 years of follow up, programming a new evaluation every third year.

299

293

100

1343

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010 P31 - INHIBITION OF GSK-3BETA FACILITATES THE FORMATION OF DENDRITIC SPINE AND LONG-TERM MEMORY. D. LIU 1 , L.-Q.ZHU 1,2 , J.-Z.WANG1,2 (1. Department of Pathophysiology, Institue of Neuroscience, Tongji Medical College, Huazhong University of Science and Technology, W uhan 430030, People's Republic of China; 2. Key Laboratory of Neurological Diseases of Education Ministry, Tongji Medical College, Huazhong University of Science and Technology, W uhan 430030, People's Republic of China) Unterach, Austria.)

Aim: Alzheimer disease(AD), an age-related neurodegenerative disorder with progressive loss of memory and deterioration of comprehensive cognition. Glycogen synthase kinase-3 (GSK-3), a Ser/Thr protein kinase abundantly expressed in neurons, plays a crucial role in several neurological disorders, such as Alzheimer's disease (AD), bipolar disorder and schizophrenia, all of which show synaptic abnormalities, and it is a prime drug target for a variety of neurological disorders. Our previous study shows that upregulation of GSK-3 suppress long term potentiation and presynaptic release of glutamate, however, the underlying mechanism is not clear. Methods: Colloidal gold immunoelectron microscopy, PSD fraction preparation and immunofluorescence costaining of GSK-3beta with specific postsynaptic markers in the primary hippocampal neurons were applied to explore the distribution of GSK-3beta in synapse. wtGSK-3beta or dnGSK-3beta or control plasmid was transfected to hippocampal neurons to exam the shape and density of dendritic spine. vena caudles with SB216763, in utero electroporation with wtGSK-3beta or dnGSK-3beta or control plasmid, lenti-virus injection of wtGSK3beta or dnGSK-3beta or control virus were used to investigate the learning and memory in rodents. The shape and density of dendritic spine was displayed by autofluorescence or Golgi staining. Results: We found that GSK-3beta is abundantly distributed in postsynaptic compartment. Inhibition of GSK-3beta can improve the learning and memory of rodents with higher density of dendritic spines and more percentage of mushroom-type spines, while activation of GSK-3beta has contrary effect. Inhibition of GSK-3beta can activate transcription factors and increase the expression of many postsynaptic proteins. Discussions: These results indicate that inhibition of GSK-3beta may improve learning and memory by synaptogenesis enhancement.

which aim to halt or slow the progression of the disease may require longer duration clinical trials to observe clinical improvement downstream of therapeutic affects on the underlying pathophysiology of AD. Measurements of biomarkers recognized as relevant to the disease process, in addition to the requisite cognitive and functional endpoints, would be required. Summary of Phase 3 Clinical Program: The bapineuzumab Phase 3 program consists of four 18-month clinical trials in subjects with mild to moderate AD. Each study utilizes a randomized, double-blind, placebo-controlled, parallel group design, with analysis of clinical variables supported by imaging and fluid biomarkers relevant to AD pathophysiology. A total of approximately 4000 subjects world-wide will participate. The program evaluates ApoE4 carriers and non-carriers separately, given evidence from Phase 2 trials of differential response to bapineuzumab therapy. Two trials are conducted primarily in North America and two primarily in Rest of World (ROW), with each region enrolling one trial of ApoE4 non-carriers and another trial of ApoE4 carriers. Bapineuzumab is administered by intravenous infusion 6 times in 13-week intervals. Measures of cognition, function, behavior, and global impression are assessed every 3 to 6 months. Biomarkers including cerebral spinal fluid proteins, brain volume, and brain amyloid burden are assessed at multiple time points. Discussion: The bapineuzumab Phase 3 program is designed to demonstrate disease modification based on a combination of clinical and biomarker evidence in trials of 18-month duration. Given that there is limited experience with the use these biomarkers in immunotherapy trials, whether they will prove to be sensitive or reliable measure of treatment related alterations in AD pathophysiology is not fully known. Thus whether bapineuzumab is effective in treating disease, altering disease pathology or slowing clinical progression will be determined by the overall strength of the data from these multiple clinical and biomarker analyses.

P34 - THE ECOG CORRELATES WITH COGNITIVE AND FUNCTIONAL PERFORMANCE. S. FARIAS1, G. CRANS2, R. MOTALLI3, T. E. LIU3, R. BLACK4, M. GRUNDMAN 3 FOR THE ELN-AIP-901 STUDY INVESTIGATOR GROUP (1. University of California, Davis; 2. Elan Pharmaceuticals, Inc., South San Francisco, CA , USA ; 3. Janssen A lzheimer Immunotherapy, South San Francisco, CA ; 4. Pfizer, Collegeville, PA , USA ) Background: Valid and reliable measures of function that are more sensitive to decline are needed for clinical trials of patients in the early stages of Alzheimer's disease (AD). The ECog (Everyday Cognition) was developed to assess functional impairment of a very mild nature, as can be seen in MCI. The ECog is an informant-rated questionnaire that comprises multiple subscales and takes approximately 10 minutes to administer. Prior results suggest that the ECOG may be a useful tool for the measurement of general and domain specific everyday functions. Objectives: To determine: 1) the correlation of the ECOG with other cognitive and functional instruments in an independent sample of healthy elderly controls, MCI and AD patients; 2) the correlation of the ECog with specific cognitive subtests included within the Neuropsychological Test Battery (NTB). Method: The ECog was added to an ongoing longitudinal cohort study of healthy elderly, MCI and AD patients initially enrolled at 40 study sites across the United States and Europe during 2006 and 2007. Recruitment criteria were similar to those used in the Alzheimer's Disease Neuroimaging Initiative. At month 24, 152 participants (76 AD, 39 MCI and 37 healthy elderly) received the ECog along with other standardized cognitive and functional instruments including the MMSE, ADAS-COG, DAD, NTB and CDR-SB. Results: The ECog showed significant cross-sectional correlations with all global cognitive and functional measures in MCI and AD subjects. In the healthy elderly, the ECOG correlated significantly with the FAQ and DAD but did not correlate significantly with the CDR-SB, MMSE or ADAS-COG. In AD, all NTB subtests correlated significantly with ECOG functional domains, while in MCI, there were significant correlations between tests of memory (e.g. Rey AVLT, WMS verbal paired associates) and ECog functional domains. In the healthy elderly, significant correlations were observed between the ECog everyday visual spatial domain and NTB subtests assessing similar cognitive functions (e.g. Wechsler visual paired associates). Conclusions: The ECog appears to be a sensitive measure of functional impairment across a broad spectrum of subjects ranging from healthy elderly to MCI and AD. The ECog was capable of eliciting functional impairments that were further corroborated on neuropsychological testing in these patient groups. The greater correlations observed in the MCI and AD patients compared to healthy elderly may reflect the greater range of impairment observed on cognitive and functional measures in these subjects.

Aim: Diagnosis of Alzheimer´s Disease (AD) in early stages may be substantiated by the measurement of biomarkers like Aß 1-42, Aß 1-40 and t-Tau in the cerebrospinal fluid (CSF). The German Dementia Competence Network (DCN) has accomplished a large biomaterial bank, comprising CSF from AD patients. Recently a number of different quantitative test systems have been developed and are now commercially available. All immunosorbent assays yield reliable results, although the reported absolute values differ. Since little is known about long-term stability of biomarker proteins in frozen CSF, we investigated the reliability of quantitative biomarker analysis in CSF samples that had been frozen for up to 5 years using different methods. Method: Here, we used CSF samples from singles patients and analyzed the concentrations of the biomarkers Aß 1-42, Aß 1-40 and tTau using different test systems. All samples were immediately frozen to -20°C after lumbar puncture and long-term storage was at -80°C. When investigating aged samples, aliquots of the same samples were re-investigated up to 5 years after collection. Assays obtained from Mesoscale Discovery Systems (MSD) were used for the quantification of all three biomarkers. Enzyme-linked immunosorbent assays (ELISAs) obtained from Innogenetics were used for the quantification of Aß 1-42 and t-Tau and from IBL for Aß 142 and 1-40. Results: We found that for all biomarkers, the absolute values obtained differ for individual CSF samples. However the data sets highly correlate for all comparisons, with the MSD test systems proving to be slightly more sensitive than the other systems. In CSF samples that have been for stored at -80°C for up to 5 years, all biomarkers can be determined without prior adaptation of the standard methods even when the protein in those CSF samples is denatured. Values varied within a range of ±20% compared to the initial measurements. Therefore, the determination of these levels in the CSF from cohorts of demented and non demented patients in large biomaterial banks assembled over many years, such as the DCN biomaterial bank, is feasible. Discussion: The quantitative analysis of the Abeta-species Aß 1-40 and Aß 1-42 as well as total-Tau in CSF in samples that have been long-term stored is reliable. The determination of proteins in the CSF stored in large biomaterial banks assembled over many years appears feasible.

P32 - STABILITY OF DIFFERENT BIOMARKER PROTEINS IN LONG-TERM STORED HUMAN CEREBROSPINAL FLUID FROM AD PATIENTS. O. PETERS 1, I. HEUSER 1, J. KORNHUBER 2, L. FRÖLICH 3, C.G. SCHIPKE 1 (1. Department of Psychiatry, Charité-CBF, Berlin, Germany; 2. Department of Psychiatry, University Erlangen, Germany; 3. CIMH, Mannheim, Germany)

Aim: To summarize the study design of pivotal clinical trials for a potential disease modifying therapy for mild to moderate Alzheimer's disease. Background: Alzheimer's disease (AD) onset is insidious with a pre-symptomatic disease course that could be decades long. Current marketed therapies for AD improve the function of surviving neuronal circuitry and their marketing approvals were based on short (6 month) trials, in expectation of relatively fast onset of symptomatic improvement. However, therapies

P33 - BAPINEUZUMAB PHASE 3 TRIALS IN MILD-TO-MODERATE ALZHEIMER'S DISEASE: TRIAL DESIGN FOR A POTENTIAL DISEASE MODIFYING THERAPY. E. LIU1, R. BLACK2, K. GREGG1, M. GRUNDMAN1 FOR THE ELN115727-301/302 INVESTIGATOR GROUP (1. JA N S S EN A lz heim er Immunotherapy Research & Development, LLC, South San Francisco, CA , USA ; 2. Pfizer Inc, Collegeville, PA , USA )

S18

Introduction. The development of efficient therapeutics for Alzheimer's disease (AD) represents one of the biggest unmet medical needs today. SynAging is a new and innovative spin-off in neurosciences providing pertinent novel models and time-efficient technologies for the development of neuroprotecive strategies targeting AD, even at early stages. Contrary to the classical models that focus on amyloid plaques (transgenic mice, or animals exposed to fibrillar A), we have used our expertise and savoir-faire to develop validated and standardized innovative models, both in vitro and in vivo using the oligomeric form of soluble A prepared according to a reproducible and calibrated protocol. This approach allowed us to identify specific therapeutic targets and validate candidate neuroprotective molecules able to inhibit deleterious intra-neuronal mechanisms triggered by soluble oligomers of A. Aim. In the present study, we demonstrated how these models allowed us to evaluate the neuroprotective effect mediated by two novel

P35 - NEW MODELS OF A OLIGOMER-INDUCED COGNITIVE IMPAIRMENT AND SYNAPTIC DEGENERATION FOR MOLECULE SCREENING AND THERAPEUTIC TARGETS IDENTIFICATION. C. MALAPLATE-ARMAND, V. KOZIEL, F.T. YEN, M.C. ESCANYE, T. OSTER, T. PILLOT (SynA ging SA S. Nancy, France)

GPCRx ligands (Syn11x and Syn21x) from our pipeline of original molecules. Methods. The neuroprotective potential of Syn11x and Syn21x were investigated in mice primary cortical neurons exposed to oligomeric A peptide. Neurodegeneration was monitored by MTT assay, DAPI and tubulin staining, as well as measurement of caspase proteolytic activities and synaptic protein levels. The cognitive performances of soluble A-injected mice were monitored by Y-maze and Morris water-maze reflecting immediate spatial working memory performance and learning and long-term memory capabilities. Oligonucleotide antisenses were used to determine whether GPCRx was involved in the neuroprotective pathway of both molecules. Results. Syn11x and Syn21x protected primary cortical neurons in culture from synaptic impairment and subsequent cell death induced by oligomeric A peptide by preventing cytoskeleton perturbations, caspase activation and by promoting neuronal-related survival pathways. In A oligomer-infused mice, intracerebral administration of Syn11x or Syn21x led to full recovery of cognitive functions. Furthermore GPRCx antisense oligonucleotides completely abolished neuroprotective effects of both molecules, indicating that the protective effect of Syn11x and Syn21x was mediated by a novel GPRCx. Conclusion. These data highlight the GPCRx-mediated signal transduction pathway as a novel therapeutic target for early stage AD intervention. Our models therefore represent powerful tools for screening or repositioning of candidate molecules as well as for identification of potential therapeutic targets. P36 - INHIBITION OF GLUTAMINYL CYCLASE(S) TARGET BOTH ­ NEURODEGENERATION AND NEUROINFLAMMATION ­ IN ALZHEIMER'S DISORDER. H.-U. DEMUTH 1 , H. CYNIS 1 , A. ALEXANDRU 2 , A. BECKER 2 , W. JAGLA 2 , S. GRAUBNER 2 , S. SCHILLING 1 (1. Probiodrug A G, Halle/S aale, Germany; 2. Ingenium Pharmaceuticals GmbH, Munich-Martinsried, Germany)

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Background: The metal-dependent enzyme Glutaminyl Cyclase (QC) facilitates pyroglutamate- (pGlu) formation from N-terminal glutamate or glutamine. Such modifications promote aggregation and toxicity of amyloid peptides, e.g. amyloid- (pGluA) in Alzheimer's disease. Moreover, pGlu-residues can be also found at the N-terminus of chemokines, which are upregulated in neurodegenerative processes. There, the pGluresidue mediates stability and biological activity, promoting immune cell invasion in overreacting peripheral immune response and recruitment of glial cells during neuroinflammation. Modeling Alzheimer's disease (AD)-like phenotypes in mice, however, does only rarely mimic features of AD such as early neuronal loss combined with neuroinflammation, indicating a lack of some pathogenic factors in commonly used models. Discrepancies in A constitution between transgenic mice and human AD, i.e. high content of N-truncated and pGlu-modified forms of A, might be a rationale for different AD pathogenesis in mice and human individuals. Aim: To specifically study the downstream effects of neuronal pGlu-A overexpression, we have developed two appropriate novel transgenic mouse models, in which we specifically aim to study the toxicity of pGlu3A-42 either generated in the secretory pathway in close proximity to QC or in the endosomal/lysosomal pathway after amyloidogenic processing of APP. Moreover, treatment of very different tg mice was conducted to further validate the concept. Methods: In both mouse models the appropriate constructs are expressed using the THY-1 promoter, directing the constructs to neurons. The TBA2.1 mice generate pGlu3A-42 within the secretory pathway after release by prohormone convertases. The construct utilizes the preprosequence of TRH which directs A into the regulated secretory pathway, e.g. into close proximity to QC. In contrast, tg NLQ animals express modified hAPP (containing the Swedish and London mutation, a deletion of Asp-597 and Ala-598 and a point mutation of Glu-599 to Gln-599). Results: In addition to discovering significant pGlu-A levels in the new tg mouse models, we found that animals display age-dependent cell loss, gliosis and A deposits, all of these being hallmarks of AD. Behavioral screening has also revealed that the neurodegenerative behavioral phenotype corresponds to the onset of pGlu-A expression. However, cell loss and inflammation were significantly less in the hAPP NLQ mice as compared to the TBA 2.1 animals. Significant behavioral changes were observed for both models, applying different tests of cognition and motor function. Treatment of all mouse strains investigated led to significant reduction of pGlu3A-42 accompanied by a rescue of the phenotype applying highly specific and save QC-inhibitors. Discussion: Our data demonstrate that QC-mediated formation of pGlu-A in vivo mediates neuronal dysfunction and cell death already in minor quantities. pGlu-A provokes formation of oligomers with severe neurotoxic potential when mixed with the full-length A peptides. Based on treatment results obtained with several mouse models, the application of QCinhibitors ­ attenuating A-meditaed toxity and the exuberant inflammatory response emerges as a new and innovative therapeutic approach for treatment of neurodegeneration. Accordingly, an intense drug discovery program has led to preclinical candidates, which will approach clinics early in 2011.

2.Atypical depression with apathy, alteration of habitual activities of the daily life without autocare of the health as in the gown, personal neatness, foods and relation's life, with emotional unexpression. 3. Progressive inattention, resistant to the habitual stimuli of the daily life with absorption and isolation, which produces in consequence supposed loss of memory. 4. Expresses their desire to die, in direct or figurative form. 5. Begins to be disconnected of the reality, which already does not feel that it contains it, for which resorts to our animal ancient behavior to block of the motor or cognitive answers of the not priority sensory stimuli for the defense, that use it unconsciously. This drives to progressive faults in the sensory recognition in the ten channels (vision, hearing, tact, smell, balance, taste, spatial orientation, corporal perception and both developed ones in the relation's life: the sense of strange things and the ubiquity's sense).The installation of this process is often oscillating: they enter and go out of the reality, up to the permanent disconnection. The progressive disintegration of the sensory system, together with the atypical depression and the marked inattention explain the observed behaviors, and can be considered as to be the most robust biological indicators for the diagnosis and the followup of cases. We consider incipient when not more than three sensory channels are affected, habitually the vision, the smell and other one. Results: The first version of the protocol was applied in 2006 in Argentina, and from there 4 new versions happened, being applied in cases of cities of Argentina, Uruguay, Chile, Peru, Brazil, Bolivia and Paraguay until comes to 40 incipient cases with confirmed medical diagnosis and informed assent. The protocol also has applied in others countries cases by Internet. The recovery begins with a work with the patients for the production of the duel allows to recover his vital motivation.The daily administration of modafinil 25 mg and caffeine 50 mg favors the communication with the patient by way of alert attention raised,and is continued with a pre stimulation centered on the attention and autosteem's elevation, and then a massive autobiographical plurisensorial stimulations, using as source theirs family pics and the favorite songs, meanwhile offer of food, textures and perfumes that they preferred, that is realized in their home or in the welfare center. The last phase consists of the reassignment of a new daily life, labor that is carried out with the participation of the relatives in their habitual residence. It is reached by a monthly follow-up for six months with reinforcement in the vital remotivation and the new activities of the daily life. Twenty three cases could be recovered (57.5 %), and other seventeen aborted indebted to difficulties of participation of the relatives. Discussion: Nine of the recoveries was produced already in the pre stimulation, though then it was continued by the protocol. The commitment and participation of the relatives has turned out to be a bounding decisive factor. 77 % of the treated cases were women.The treated cases had remained between 1 to 5 years inside the process, and they showed neither cognitive nor motor sequels on having gone out. The disengaged explanation to the patient of the reasons of his problem enables clearly his recovery. P38 - COGNITIVE TRAINING OF ATTENTION AND EXECUTIVE FUNCTION IN PATIENTS WITH AMNESTIC MILD COGNITIVE IMPAIRMENT. M. TSOLAKI 1,2 , F. KOUNTI 1,4 , C. AGOGIATOU 1 , A. SOUBOUROU 1 , M. ZAFEIROPOULOU1,3, E. POPTSI1, E. BAKOGLIDOU1, S. ZAFEIROPOULOS1, G. BATSILA1, E. NIKOLAIDOU1, G. KIOSSEOGLOU4, A. EFKLIDES4 (1. Greek A ssociation f or A lz heim er's disease; 2. M edical S chool A ristotle Univ ersity of Thessaloniki; 3. Department of Psychology, New Y ork College; 4. Department of Psychology A ristotle University of Thessaloniki)

Aim:Presents the protocol developed and used for the recovery of incipient cases of alzheimer. Method: developed by deduction from the logic of production of the phenomenon (theory) which temporary sequence is as follow: 1. Very significant personal loss, without confrontation capacity and that drags besides the identity of the person.

P37 - EXPERIMENTAL PROTOCOL OF THEORETICAL BASE FOR THE RECOVERY OF INCIPIENT CASES OF THE INDIVIDUAL SO CALLED ALZHEIMER PROCESS AT SOUTH AMERICA. L.M. SANCHEZ, J.D. GARCIA, P. DEL C. TUDOR, J.A. VILLARREAL, D. PRIETO, M.A. LEITES, G.A. MARTINEZ, E.D. FERRARI, M.A. JALIFE, N.E. SANTA CRUZ, J. GOMEZ SALGADO, C. NIEMZ, C.R. CANTERO (Neuro-psycho-ethology Research Center, University of Entre Ríos, A rgentina)

Aim: The main aim of the present study was to investigate the effect of a cognitive training intervention on the overall intellectual ability and on the ADL in elderly patients with Mild Cognitive Impairment of amnesic type (aMCI). A secondary aim was to examine the duration of any cognitive or functional benefit after the end of the training. Method: Participants attended a 5 month intervention program targeting the enhancement attention and executive function. There were 181 participants randomly classified in one experimental group (n= 100) with S= 27.47 (1.95) and one control group (n= 81) with S= 27.49 (1.66), who were at the waiting list. The two groups were matched in gender (p= .934), education (p= .176) and age (p= .950). The experimental group attended an average of 20 weekly sessions each lasting for two hours. Cognitive and functional performance was assessed at baseline and 5 months later by a extensive neuropsychological battery. Sixteen experimental participants were reassessed 6 months after the end of the intervention. Results: Compared to the control group, the experimental has shown better performance in selective attention (p .007), executive function (p 0.008), language (p 0.015), verbal memory (p=0.001), visual memory (p0.014), ADL (p=0.002), and general cognitive status (p=0.000). The experimental group showed improvement at the end of the intervention, in selective attention (p 0 .007), visual memory (p=0.000), visual perception (p=0.005), verbal memory (p0.008), language (p=0.003), and ADL (p=0 .000). Controls appeared a deterioration in ADL (p=0.025) and verbal memory (p= 0.004), while there was improvement in digit span (p=0.000). Sixteen experimental participants were further examined 6 months later in absence of any kind of cognitive exercise with the same battery. There was shown a further benefit in executive function (p=0.001), attention (p=0.000), verbal episodic memory (p=0.002), language (p=0.000), and visual memory (p=0.003). Discussion: aMCI experimental patients were improved by the training, an effect that was maintained and further developed 6 months after the end of the intervention. They have also shown improvement in cognitive abilities that were not directly trained via the intervention. The cognitive improvement was transferred in activities of daily life. At the end of the intervention, their cognitive and functional performance did not meet any more the criteria of MCI but those of healthy elderly. Key words: cognitive training, Mild Cognitive Impairment, attention, executive function

S19

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

As clinical trials become more globalized, pharma and biotech companies are redrawing their maps and rethinking their strategies. By doing so, they are forced to explore otherwise unknown regions to them. In this presentation, we will educate and inform the audience about the benefits and challenges of conducting AD clinical trials in Latin America. This region is becoming an important part of global clinical development with trial growth increasing steadily each year. The reasons for such growth are explained by the vast resources available in Latin America: experienced investigators with large patient populations, naïve patient populations, well equipped study institutions, relatively streamlined regulatory processes, strong clinical research infrastructure (laboratories, drug depot services, and logistics), internationally recognized treatment standards, and quality data corroborated by various audits and inspections. Attendees will receive practical advice on critical points to take into account when conducting trials in Latin America and an indepth knowledge on how Latin America can be a successful region for their next clinical trial.

P39 - LATIN AMERICA A SUCCESSFUL EMERGING MARKET FOR ALZHEIMER´S DISEASE CLINICAL RESEARCH. F. MANES 1 , F. LERNER 2 (1. Director, Institute of Neurosciences - Favaloro University and Director, Institute of Cognitive Neurology (INECO); Buenos A ires, A rgentina; 2. Sr Director of Operations, Latin A merica; PRA International; Buenos A ires, A rgentina)

Aim: In recent years use of antipsychotics was widely debated for concerns about their safety in elderly patients affected with dementia (1). The aim of the present study was to make an update on the use of antipsychotic drugs in elderly persons affected with dementia. Materials and methods: A Medline search was made using as key words elderly, conventional and atypical antipsychotics, adverse events, dementia, behavioural and psychotic symptoms in dementia (BPSD). Some personal studies were considered too. Results: Conventional antipsychotics have been widely used for BPSD, even if available studies showed an efficacy comparable to placebo (2). A superior efficacy was only shown at high doses, but they were associated to several and severe side effects. Atypical antipsychotics showed an efficacy superior to placebo in randomized studies in BPSD treatment, with a better tolerability profile vs. conventional drugs (3). However, in 2002 trials with risperidone and olanzapine in elderly patients affected with dementia-related psychoses suggested the first warnings about the possible increase in cerebrovascular adverse events. Drug regulatory agencies have issued specific recommendations for undelining that treatment of BPSD with atypical antipsychotics is "off-label" (2). In 2005 a paper by Wang et al. showed that conventional antipsychotic medications are at least as likely as atypical agents to increase the risk of death among elderly persons and that conventional drugs should not be used to replace atypical agents discontinued in response to the FDA warning (2). In Italy atypical antipsychotic prescription is subordinated to a program of pharmacovigilance by Alzheimer's centers and to patient's informed consent on the potential risks of these drugs for adverse reactions. The presence of cardiovascular diseases, QTc interval on electrocardiogram, electrolytic imbalances, familiar history for torsades des pointes, concomitant treatments and use of drugs potentially able to lengthen QTc have to be closely taken into account (4). Discussion: Adverse events associated with typical antipsychotics make their use inadvisable in people with Alzheimer disease. Atypical antipsychotics are probably still the best option for short-term treatment of aggression (6­12 weeks) that is severe, persistent and treatment resistant, but serious adverse events are a major contraindication to long-term therapy (4). Elderly demented patients are often affected with concomitant diseases and they are frequently polytreated; therefore, use of atypical antipsychotics in dementia require a careful case-by-case assessment, together with the possible drug-drug interactions. References: 1. Jeste DV, Blazer D, Casey D, Meeks T, Salzman C, Schneider L, Tariot P, and Yaffe K: ACNP White Paper: update on use of antipsychotic drugs in elderly persons with dementia. Neuropsychopharmacology 2008, 33: 957-970. 2. Wang PS, Schneeweiss S, Avorn J, Fischer MA, Mogun H, Solomon DH, Brookhart MA: Risk of Death in Elderly Users of Conventional vs. Atypical Antipsychotic Medications. NEJM 2005, 353: 2335-2341. 3. Gareri P, De Fazio P, Stilo M, Ferreri G and De Sarro G: Conventional and atypical antipsychotics in the elderly. Clin Drug Invest 2003, 23: 287-322. 4. Gareri P, De Fazio P, De Fazio S, Marigliano NM, Ferreri Ibbadu G, De Sarro G: Adverse effects of atypical antipsychotics in the elderly: a review. Drugs Aging 2006, 23: 937-956.

P40 - USE OF ANTIPSYCHOTICS IN BEHAVIOURAL DISORDERS IN DEMENTIA. P. GARERI1,2, P. DE FAZIO3, N.M. MARIGLIANO1, R. LACAVA2, A. CASTAGNA2, D.S. COSTANTINO2, G. DE SARRO1 (1. Chair of Pharmacology and Psychiatry; 3. Department of Experimental and Clinical Medicine, Faculty of Medicine and S urgery , Univ ersity M agna Graecia of Catanz aro, Clinical Pharm acology and Pharmacovigilance Unit, Mater Domini University Hospital, 88100 Catanzaro, Italy; 2Geriatrist, Operative Unit Elderly Health Care, 88100 Catanzaro, Italy)

P43 - BEHAVIORAL-BASED SCREENING OF DRUG CANDIDATES FOR ALZHEIMER'S DISEASE TREATMENT. Z.-L. XIE 1 , W.-W. MA 1 , Y. ZHONG 2 (1. JoeKai Biotech. LLC. Beijing, China; 2. Cold Spring Harbor Lab., Cold Spring Harbor, NY 11724 USA ) We generate an Alzheimer's disease model through targeted pan-neuronal overexpression of a mini-transgene encoding a secretive form of Aß42 peptides in the Drosophila brain. Such expression leads to age-dependent memory loss and neurodegeneration in association with accumulation of Aß42 peptides in the forms of monomers, oligomers, and fibril deposits in the brain. We also show that such expressed Aß42 peptides affect glutamatergic synaptic transmission and synaptic plasticity in a manner similar to those observed in the vertebrate hippocampus. The Aß42 peptidedependent synaptic effects appear to involve the PI3 Kinase pathway. Since capable of recapitulating features of AD symptoms as well as vertebrate AD models, we thought of that this fly model might be helpful in elucidating mechanisms underlying Aß42 toxicity and in facilitating screening of drug candidates for developing novel AD treatments. We approached our goals through direct behavioral screening of chemicals that were capable of preventing or rescuing age-dependent memory loss in Aß42-expressing transgenic fruit flies (referred as AD flies). The new-born AD flies had a normal memory score assayed through the extensively used odor-electric shock paired fear conditioning paradigm. However though, 10-days old AD flies had significantly lower memory scores in comparison to control flies with similar genetic background at the same age. For drug screening, we fed AD flies with chemicals each day for two hours in consecutive 7 days starting at the day three after eclosion. Then drug-fed AD flies will be tested for memory assay at the day 10. The same AD flies fed with the sugar solution served as the negative control while positive controls were fed with the memantine, the FDA approved AD drug. Feeding of memantine in such a skim was capable of preventing memory loss in AD flies in a dosage dependent manner. The compounds tested were chosen on basis of two distinct resources. The first one is targeted. Since we have shown that the PI3 Kinase pathway is relevant, a group of selected chemicals affecting different components within this pathway was tested. The second was random. We tested a group of synthetic single compounds that were supposedly blood-brain-barrier permeable kinase inhibitors and natural products isolated from Chinese medicine herbal extracts. We screened through more than 5 thousands compounds and 67 leads were identified in AD flies. After database searching, similar structures for 26 of the 67 compounds were reported in previous literature with AD preventing activity. The memory rescue effect for 10 of the 67 leads was assayed on an AD mice model in the Morris Water Maze task. 3 of them could completely prevent AD mice memory decay and 4 partially while 3 of 10 failed to be effective. We are now also lunched a screening on basis of fly brain sectioning for chemicals that are capable of preventing neurodegeneration. Such data and its mechanistic implication will be discussed.

Twenty years passed by since we began our study about the three main causes of blindness in Mexico, this work allowed our team to detect at first in 1990 a biochemical process in human retina and twelve years we could understood the explanation. Our body has the astonishing capacity to use water as source of energy; furthermore 99 % of our energetic daily requirements comes from water. Human body has a photo-system analog to vegetables; therefore the eukaryotic cell could transform light (photonic energy) into chemical energy. The discovery is controversial indeed; however Russian Federation granted the patent in 2 April 2010. We must rethinking the role of the ventricles and subarachnoid space due to are the main source of energy of the central nervous system and not the blood vessels or ATP. Our therapeutic results are consisting with our discovery. Our Alzheimer's patients with medical enhancement of human photo synthesis treated improve markedly. In this work we will present an explanation of the biochemical basis and the clinical results that are extraordinary.

P42 - THE ENHANCEMENT OF THE CAPACITY OF HUMAN BODY TO TAKE ENERGY FROM WATER IMPROVES SIGNIFICANTLY THE PATIENTS WITH ALZHEIMER'S DISEASE. A. SOLIS HERRERA, M. DEL CARMEN ARIAS ESPARZA, R.I. SOLIS ARIAS, P. SOLIS ARIAS, M.P. SOLIS ARIAS (Hum an Photosynthesis Study Center, A guascalientes, Mexico)

following ROIs: hippocampi, amygdalae, posterior cingulate, temporoparietal and frontal cortices. The gray matter was modeled as a function of time, diagnosis, and the time x diagnosis interaction. Age and education were included as covariates and random coefficients were defined for intercept and time factors. Results: In all except the frontal ROIs, both time and time x diagnosis factors were highly significant (p<.001) with negative slopes, indicating an increasing of longitudinal gray matter loss in AD group compared to healthy controls. In the same regions, the mean ratios between gray matter loss in Alzheimer's compared to normal aging were 3.32 (amygdalae), 3.85 (hippocampus), 3.48 (medial temporal lobes) and 1.84 (posterior cingulate). On the contrary, in the frontal areas the interaction coefficients were either close to 0 or positive (max .0045 with p<.05). Discussion: Medial temporal and posterior cingulate structures show the greatest rates of change in incipient AD, a pattern significantly different from normal aging.

P41 - VOXEL-BASED LONGITUDINAL CHANGES OF GRAY MATTER DENSITY IN INCIPIENT ALZHEIMER'S DISEASE. M. LORENZI, G.B. FRISONI, THE ALZHEIMER'S DISEASE NEUROIMAGING INITIATIVE (LENITEM Laboratory of Epidemiology, Neuroimaging and Telemedicine, IRCCS San Giovanni di Dio-FBF, Brescia, Italy) Aim: To model the changes of gray matter density in the early stages of AD on a voxel-by-voxel basis. Method: MRI scans of 21 MCI subjects converted to dementia and 20 healthy elders from the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset were studied. Gray matter segmented images were normalized to a common reference using the DARTEL protocol and a linear mixed effect model was fitted voxel-wise in the

S20

Aim: Animal research and epidemiological studies identified cholesterol as a risk factor for mild cognitive impairment (MCI) and Alzheimer's disease (AD). However little

P44 - HDL, LDL AND TOTAL CHOLESTEROL IN MILD COGNITIVE IMPAIRMENT AND ALZHEIMER'S DISEASE. RESULTS FROM THE ILSESTUDY. J. SCHRODER, C. SATTLER, P. SCHÖNKNECHT, P. TORO (Section of Geriatric Psychiatry, University of Heidelberg, Heidelberg, Germany)

is known about the differential effects of HDL, LDL and plasma total cholesterol (TC) levels. We therefore investigated the potential impact of HDL, LDL and TC on the development of MCI and AD in a longitudinal study of a representative birth cohort born between 1930 and 1932. Methods: 500 participants of the Interdisciplinary Longitudinal Study on Adult Development and Aging (ILSE) were examined in 1993-1996 (t1), 19972000 (t2) and 2005-2008 (t3). Results: Prevalence of MCI increased from 13% to 23% and 29% over time; in addition, 7% of the participants developed AD at t3. At t1 and t2 patients showed significantly higher LDL and TC levels than controls. The diagnostic groups did not differ concerning their HDL levels. When compared to control subjects, patients showed a significant decrease of TC levels between t1 and t3. APOE4 carriers showed higher LDL and TC levels at all examination waves compared to the non-APOE4 carriers. With respect to LDL levels repeated analyses of variance yielded significant (p<0.05) main effects for diagnosis, presence of an APOE4 allele and the interaction between diagnosis*APOE4. Concerning TC levels we found significant main effects for APOE4, time as well as a time*diagnosis interaction. Discussion: Our findings confirm the hypothesis that high midlife serum TC constitutes a risk factor for the development of MCI and AD and demonstrate that this effect particularly strikes at the APOE4 carriers. Moreover we showed that patients and controls differed concerning their LDL but not their HDL levels. The results indicate that LDL levels in midlife need to be considered in the prevention of MCI and AD.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Aim: Diabetes mellitus type 2 (T2DM) is considered to be an important risk factor for mild cognitive impairment (MCI) and subsequent Alzheimer's disease (AD). The majority of studies relating T2DM to MCI and AD were performed in North America. We investigated the potential impact of T2DM on the development of MCI and AD in the Interdisciplinary Longitudinal Study on Adult Development and Aging (ILSE) which involves a representative birth cohort of subjects born between 1930 and 1932 in Germany. Methods: Subjects received a thorough psycho-geriatric examination and neuropsychological testing; particular care was taken to exclude subjects with severe medical or neurological conditions sufficient to explain the cognitive deficits, or other major psychiatric disorders. Results: When compared to healthy subjects (n=159), patients with MCI (n=108) or AD (n=26) showed a tendency towards increased prevalence rates for T2DM (16% vs. 23%; 2=1.7, p =0.18). In both patients with MCI and controls, T2DM was associated with psychomotor slowing but not deficits in other cognitive domains typically involved in MCI. Discussion: Our findings indicate that T2DM is involved in MCI and may aggravate the clinical picture as a concomitant factor.

P45 - DIABETES AS A RISK FACTOR FOR MILD COGNITIVE IMPAIRMENT IN A POPULATION-BASED SAMPLE IN GERMANY. J. SCHRODER, C. SATTLER, P. SCHONKNECHT, P. TORO (Section of Geriatric Psychiatry, University of Heidelberg, Heidelberg, Germany)

Aim: Diagnosis and treatment of Alzheimer's disease (AD) depend on clinical evaluation, and there is a strong need for an objective tool as a biomarker. We aimed to investigate event related oscillations in two groups of subjects with Alzheimer's disease (those treated with cholinergic drugs and drug-naïve subjects) and compare their results with healthy elderly control subjects. Method: Our group has investigated brain oscillatory responses in a group of AD subjects (n=22). Result: We found that the de novo (untreated) AD group (n=11) differs from both the cholinergically-treated AD (n=11) group and agedmatched healthy controls (n=20) in theta and delta responses over left frontal-central areas after cognitive stimulation. On the contrary, the difference observed in AD groups upon a sensory visual stimulation includes response increase over primary or secondary visual sensorial areas compared to controls. Multiple coherences are also very valuable in observing the group differences, especially when a cognitive stimulus is applied. In healthy controls, higher coherence values are elicited after a cognitive stimulus than after a sensory task. Discussion: These findings imply at least two different neural networks, depending on type of stimulation (i.e. cognitive or sensory). Our findings support the notion of disconnectivity of cortico-cortical connections in AD. The default mode defined as activity in resting state in AD seems to be affected as displayed by electrophysiological methods and cognitive stimulation elicits more evident deviations from controls.

P46 - DEFAULT MODE NETWORK IS AFFECTED IN ALZHEIMER'S DISEASE AS SHOWN BY BRAIN OSCILLATORY RESPONSES. G.G. YENER 1,2,3,4 , B. GÜNTEKIN4, E. BASAR4 (1.2.3. Dokuz Eylül University, Departments of Neurology and 2. Neurosciences; 3. Brain Dynamics Multidisciplinary Research Center, Izmir 35340, Turkey; 4. Istanbul Kultur University, Brain Dynamics, Cognition and Complex Systems Research Center, Istanbul 34156, Turkey)

Aim: While several prospective epidemiological studies have suggested a protective role of long-chain omega-3 fatty acids (n-3 PUFA) against dementia or cognitive decline, randomized clinical trials conducted so far failed to show any beneficial effect of n-3 PUFA in prevention of Alzheimer's Disease (AD). However, a protective effect has been found in AD subjects with mild cognitive deficits in post-hoc analyses, suggesting that n-3 PUFA may be efficient early in the course of neurodegeneration. However, the few primary prevention studies have been mostly disappointing to date. This failure may be due

P47 - PLASMA LONG-CHAIN OMEGA-3 FATTY ACIDS AND TRAJECTORIES OF COGNITIVE DECLINE AMONG OLDER SUBJECTS WITH A SUBJECTIVE MEMORY COMPLAINT AND MMSE SCORE ABOVE 24. C. SAMIERI, C. PROUST-LIMA, C. FEART, P. BARBERGER-GATEAU (Research Center Inserm, U897, ISPED, University Bordeaux 2, Bordeaux, France)

Introduction: Amyloid beta (A) appears to be a crucial molecule in Alzheimer disease (AD). Many researchers consider it as the root cause of the disease. The urgency for new and effective treatments for AD is matched by the need for new tests to detect and diagnose the condition. Uncertainties in the diagnosis of AD mean that the disease is often undiagnosed and under treated. Moreover, it is clear that clinical confirmation of AD, using cognitive tests, can only be made after substantial neuronal cell loss has occurred; a process that may have taken place over many years. Poor response to current therapies may therefore, in part, reflect the fact that such treatments are generally commenced only after extensive neuronal damage has occurred. The absence of tests to detect or diagnose presymptomatic AD also means there is no standard that can be applied to validate experimental findings (e.g. in drug discovery) without performing lengthy studies, and eventual confirmation by autopsy. Aim: The reliable recruitment of individuals for clinical trials in Alzheimer disease. Method: Direct determination of the -amyloid pool in blood by the quantification of both A1-40 and A1-42 free in plasma, bound to plasma proteins and bound to the blood cellular fraction with our ELISA sandwich colorimetric tests ABtest40® and ABtest42®. Results: We have carried out a pilot study including healthy controls, mild cognitive impairment (MCI) and Alzheimer disease patients and found that the pool in blood of A1-40 and A1-42 reached a sensitivity and specificity over 80% to discriminate between amnesic mild cognitive impairment patients and healthy controls. Also, we have found that the vaccination of dogs, which represent a naturally occurring animal model for the early phases of AD, produces a clean out of the A1-40 and A1-42 in blood. Discussion: The measure of A pool in blood (a cheap and non-invasive technique) in potential trial participants for a preventative or curative AD medicine, whether they are healthy individuals, MCI or early AD patients, may be an affective recruitment and screening tool for clinical trials to test medicines affecting the betaamyloid cascade. Ideally only those patients with relatively high levels of A in blood should be selected for the clinical trials in order to ascertain the expected decrease in the A after treatment with the assayed medicine.

P48 - AN EASY AND CONVENIENT WAY TO RECRUIT INDIVIDUALS FOR CLINICAL TRIALS IN ALZHEIMER DISEASE. P. PESINI 1 , I. MONLEON 1 , V. PEREZ-GRIJALBA 1 , A.Mª. LACOSTA 1 , I. SAN JOSE 1 , M. BOADA 2 , P. MARTINEZ-LAGE 2 , L. TARRAGA 2 , M.L. SUAREZ 3 , G. SANTAMARINA 3 , M. SARASA1 (1. A raclon Biotech Ltd, Zaragoza, Spain; 2. Fundación A CE, Barcelona, Spain; 3. Hospital universitario veterinario Rof Codina, Lugo, Spain)

to a wrong targeting of the population which may benefit from n-3 PUFA supplementation. Subjective memory complaint is a potential inclusion criterion in primary prevention clinical trials in AD. However, it may define a heterogeneous population, with only a subgroup having further cognitive decline. The aim of this exploratory study was to determine if older subjects with a subjective memory complaint from a population-based, prospective cohort study had different trajectories of cognitive decline over 8 years, and whether n-3 PUFA measured in plasma were associated with a particular trajectory of decline. Method: Among the Three-City (3C) study, 228 subjects were non-demented, had a subjective memory complaint, MMSE>24 and long-chain n-3 PUFA (eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA)) measurements at baseline, and were reexamined at least once over 7 years (2 yr, 4 yr, 7 yr follow-up visits). Cognitive performance was prospectively assessed on 4 neuropsychological tests: Mini-Mental State Examination (MMSE), Isaacs Set Test (IST), Benton Visual Retention Test (BVRT) and Trail Making Test (TMT). The Free and Cued Selective Reminding Test (FCSRT) was assessed at the 2 yr visit. Incident cases of dementia were validated by an independent expert committee of neurologists. Trajectories of cognitive decline in MMSE, IST, BVRT and TMT performances were explored using latent class linear mixed models, which assume that the population consists in a finite number of subpopulations, each one corresponding to a specific trajectory of the cognitive outcome. Results: Among the 224 subjects with no missing data for BVRT and covariates (age, sex, education, ApoE4 allele), 2 trajectories of decline in BVRT performances were observed (mean posterior probability in each class >85 %). A significant decline was found in a first class (class1, n=42, mean baseline BVRT=9.2, for slope=-3.22, P=0.02), but not in a second class (class2, n=182, mean baseline BVRT=12.4, for slope =-0.86, P=0.14). In class1, 40.5 % of subjects became demented during follow-up, and mean FCSRT score at 2 yr (n=24/42) was 17.1. In class2, 12.5 % of subjects became demented, and mean FCSRT score at 2 yr (n=147/182) was 23.6. Higher plasma EPA was significantly related to slower cognitive decline in class1 ( for slope in second and third tertiles grouped vs first tertile=3.17, P=0.04) but not in class 2. Two trajectories of decline in MMSE performances were observed, with a significant decline in a very small class only (class1, n=13/225), in which 46.2% of subjects became demented (mean FCSRT score (n=9/13)=21.2). Higher plasma EPA, but not DHA, was significantly related to slower cognitive decline in class1. No heterogeneity of decline in TMT and IST performances was found in this population, and plasma EPA and DHA were not associated with decline in these cognitive tests. Discussion: Nondemented older subjects with a subjective memory complaint, a BVRT score of 9 or less and a FCST score of 17 or less may exhibit a fast decline in working memory, and may be an interesting target population to assess the efficacy of long-chain n-3 PUFA supplementation in primary prevention of cognitive decline in AD. - preference: oral presentation; - presenting author:

S21

Introduction: Despite modest evidence of efficacy in randomized controlled trials and mounting post-marketing surveillance evidence of serious adverse events, including increased mortality, antipsychotic drugs continue to be used for behavioural and psychological disturbances associated with Alzheimer's disease (AD). Given the absence of any commercial interest of pharmaceutical companies on old drugs with possible new indications, studies on antipsychotics in dementia need public funding. Methods: The AdCare study, a no profit, randomised, placebo-controlled, double blind, multicentre, pragmatic trial coordinated by the Italian National Institute of Health aimed to evaluate the long-term safety and efficacy profile of three atypical (risperidone, olanzapine and quetiapine) and one conventional (haloperidol) antipsychotic drugs in treating psychosis, aggression and agitation in outpatients with AD. Results: The study was planned to be conducted in 19 clinical centres and to enrol 1000 patients. The legal requirement that, in patients lacking capacity to consent, consent had to be given by a legally authorised representative dramatically slowed down the recruitment process and for this reason the study was prematurely interrupted. From February 2009 to April 2010, 83 patients were enrolled in the study. 56 patients (68%) were included with consent given by the legal representative, while 27 patients (32%) were deemed able to give informed consent, according to a previously defined procedure. 23 patients experienced behavioural symptoms of clinically relevant severity and were randomized to the study drug; all the patients randomized except one were included in the study with consent given by the legal representative. After the trial interruption, all the patients that were taking an active drug went on assuming in clinical practice the same molecule their were assuming within the trial. Conclusions: Randomized controlled trials are recognized as the gold standard source of evidence on drug safety and efficacy. Patients with dementia have a great benefit to gain from getting the best evidence possible on the risks and benefits of the drugs they are assuming. Legal requirements currently requested in Italy to enrol patients lacking capacity to consent in clinical trials have the aim to protect patients' rights. Unfortunately the AdCare study demonstrated that they can become a major obstacle in conducting clinical research. Patients, not only in Italy, are facing in clinical practice the same risks as in the AdCare study without having the possibility to give a contribute in the assessment of the benefit-risk profile of these drugs. P50 - RELATIONAL FACTORS OF SUBJECTIVE WELL-BEING IN OLD AGE. R. MORARU, L. SPIRU, M. BLACIOTI (A na A slan International A cademy of A ging, Bucharest, Romania)

P49 - THE PARADOX OF ANTIPSYCHOTIC DRUGS IN ALZHEIMER'S DISEASE: THE PREMATURE INTERRUPTION OF A RANDOMIZED CONTROLLED TRIAL. F. GALEOTTI, N. VANACORE, F. IZZICUPO, S. GAINOTTI, F. MENNITI-IPPOLITO, R. RASCHETTI AND THE ADCARE STUDY GROUP (National Centre for Epidemiology, Surveillance and Health Promotion, National Institute of Health, Rome, Italy)

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

Aim: Nowadays Alzheimer's disease (AD) is a severe health problem. Despite huge financial efforts involved in research, there is neither a cure, nor a preventive strategy available yet. Moreover, underlying AD mechanism is still unknown. A new approach may provide new opportunities for identifying specific, early-stage biomarkers resulting in effective therapies and in developing prevention strategies. Besides its well known particular morphological and physiological traits Homo sapiens exhibits certain particularities of aging pathology. Humans suffer from AD, autoimmune diseases, and have a greater risk of developing type II diabetes and cancer than lab chimpanzees. The huge-sized human brain also has a specific pathology (besides AD) including psychiatric disorders, autism, etc. Therefore, a hypothesis capable of explaining the mechanisms of human evolution should also explain the onset and mechanisms of specific human brain diseases. Method: Three general classes of genetic differences have been proposed as factors separating humans from the great apes: (1) chromosomal differences, (2) small sequence differences that change gene expression, and (3) biochemical changes arising from gene inactivation. Molecular biology data supports the idea that epigenetic changes could result in genetic differences; important biochemical pathways could suffer alterations in terms of regulation and activity. The insulin/IGF1 signaling pathway involved in the metamorphosis of some species (as well as in cell proliferation and differentiation) is also altered in selected human pathological conditions. This suggests huge opportunities for designing (pre)clinical trials investigating possible new biomarkers linked to the insulin/IGF1 pathway. Results: The incidence of AD is 60% higher in type II diabetes patients. Factors involved in insulin resistance (i.e. streptosotocine) alter the activity of a key enzyme involved in the insulin/IGF1 pathway in a way that is similar to alterations noted in AD; AD could therefore be deemed as "brain's diabetes". Some data suggest that AD possible cell and tissue biomarkers (i.e. tau protein alterations, amyloid aggregations) could be the result of alterations in brain glucose and lipid metabolisms. The preamyloid protein enhances the pathway signaling. Moreover, apoE variant 4 is reported to alter glucose metabolism in brain. This variant, absent in the most related primates (but present in Microcebus murinus and Macacca mulata) may explain why macaques develop diabetes when adopting a rich western diet. The latter are very adaptable species, living in scarceresourced environments. Considering some biochemical aspects of the cholesterol metabolism, the presence of the 4 variant of ApoE may represent an adaptation to resource scarcity in these species. Discussion: Changes in the activity of insulin/IGF1 pathway resulting in altered cell proliferation/ differentiation patterns, possibly linked to the onset of certain human ­specific traits (i.e. huge sized brain) and disorders may provide new opportunities for preclinical study of AD mechanisms and therapeutic targets. This hypothesis could provide new theoretical grounds for new therapeutic strategies for some human diseases.

P51 - ALZHEIMER'S DISEASE: A BRAIN'S DIABETES? A.G. DIACONEASA, L. SPIRU, I. TURCU (A na A slan International A cademy of A ging, Bucharest, Romania)

Aim: Our paper is based on the fact that old age is that time of life when individuals are confronting many crises and hardships, characterized either by acceptance and integration of these crises or by a reaction of despair in front of such issues (especially related to coping with death). Our leading question refers to what would precisely assist elders in coping with these life changes, what is keeping him alive when loosing their meaning and motivation for being. Most researchers agree that there are three operational and measurable specific components of subjective well-being. Two of them are within the affective domain, while the third is in the cognitive domain. The affective components include the presence of positive affects (i.e. happiness or good feelings) and the lack of negative affects. Conversely, negative affects refer to unpleasant feelings such as anxiety and anger. The cognitive component has been conceptualized as the intellectual evaluation of one's life satisfaction. Method: Our paper attempts to distinguish the following aspects: (1) the level of subjective well-being among our studied group of aged people; (2) the relationship between socio-demographic variables (age, marital status, levels of education, health assessed by a physician and perceived health) and subjective well-being; (3) the relationship between personality variables (i.e. self-esteem and extraversion) and subjective well-being; (4) the correlation between socio-psychological variables (i.e. frequency of social interactions with friends, social support) and the level of subjective well-being. 126 participants, aged 60 to 91 years, were recruited for our study (most of them recruited from the "Ana Aslan" Institute ambulatory. Initially we interviewed the participants, later on administering them the instruments in one-on-one groups. We used Satisfaction With Life Scale, Positive and Negative Affect Schedule (PANAS), the Rand Social Health Battery, the Duke Social Support Scale, Rosenberg Self-Esteem Scale and the E (extraversion) scale of Eysenck Personality Questionnaire. Results: After performing a statistical analysis of all results, we found that (out of all socio-demographic variables) it is solely the "marital status" that is influencing the level of subjective well-being, causing significant differences between the two groups of elders (widowers and married) on mean values of subjective well-being. The mean levels were highest for married participants. Subjective evaluation of health status caused significantly differences between the three groups; participants assessing their own health as "very good" had recorded the highest level of subjective well-being. "Health assessed by a physician" did not correlate with subjective well-being. As far as other socio-psychological and personality variables were concerned, we found that these were significantly correlated with subjective well-being. Discussions: Elevated levels of subjective well-being are usually associated with increased family social support, with a personality characterized by extraversion, with more abundant social interactions, with having a higher self-esteem and with being satisfied by one's own health status.

Aim: The development of supportive environments and Assistive Technology (AT) applications is a priority recommended by the International Plan of Action on Ageing (Madrid 2002). These technological applications are expected to support patients with physical and cognitive disabilities. Their main goal is to allow patients to remain in their homes for as long as possible, and to reduce health care financial burden. The first crucial aspect about the creation and implementation of a given Ambient Intelligence (AmI) system is to observe a specific user profile and to meet specific user requirements, thus boosting the functional capabilities of the intelligent and semi-autonomous assistive device. The design, testing and validation of smart assistive devices require multidisciplinary cooperation in which clinical trials on voluntary patients are of capital importance.Method: We performed an overall analysis of the results arising from 5 trials investigating physically and cognitively impaired voluntary patients involved in testing and validating 3 smart assistive platforms designed as part of a EU-IST-FP6 project in which we were medical partner. Results: Patient profiling is actually a very difficult task, given the huge variability of all possible combinations of disabilities. The emerging solution resided in defining clusters of disabilities to be used as templates specifically designed for creating AmI devices. Another significant pitfall was the human/machine interaction (albeit computer - patient or computer - caregiver interaction); specific attention needs to be paid to neglect patients. The solution resides in designing highly refined communication interfaces, easily usable and highly adaptable to the ever changing user needs. Additionally, the refinement of user's guide must be closely pursued. Such guides should be user-friendly and able to minimize any possible reluctance of the patient (possibly of the caregiver as well) when facing computers and device components. The ethical aspects concerning the involvement of voluntary subjects in testing and validating AmI applications should however observe both best practice policies and standards and specific ethical requirements and should also aim at improving the existing ethical guidelines. Discussions: Voluntary involvement in clinical trials for testing and validating smart assistive devices should always consider the specific features and requirements of cognitive patients as well as the requirements of their caregivers. The refinement of ethical aspects should be an ongoing process, vital for succeeding in the design of cost-effective and high-performance assistive tools. Keywords: cognitive variability, assistive technology, user profile

P52 - CLINICAL TRIALS FOR PATIENT PROFILING AND VALIDATION OF AMBIENT INTELLIGENCE ASSISTIVE DEVICES. L. SPIRU, I. TURCU, C. GHITA (A na A slan International A cademy of A ging, Bucharest, Romania)

S22

P54 - FIRST RESULTS OF ASCOMAVA TRIAL ON THE ASSOCIATION BETWEEN THE CHOLINESTERASE INHIBITOR DONEPEZIL AND THE CHOLINERGIC PRECURSOR CHOLINE ALPHOSCERATE IN ALZHEIMER'S DISEASE. F. AMENTA 1 , A. CAROTENUTO 1 , R. REA 1 , E. TRAINI 1 , A.M. FASANARO 2 (1. Centro R icerche Cliniche, T elem edicina e T elef arm acia, Università di Camerino, 62032 Camerino; 2. Unità V alutativa A lzheimer e Malattie Involutive Cerebrali, A zienda Ospedaliera Rilievo Nazionale A . Cardarelli, 80131 Napoli, Italy) Acetylcholinesterase/cholinesterase inhibitors (ChEIs) are the most prescribed drugs for Alzheimer's disease (AD) in spite of the controversy about their efficacy, the shortlasting effect and the high number of patients not responding to treatment. Underlying pathologies such as cerebrovascular diseases, as well as secondary ischemic brain injury contribute to aggravate cognitive loss. Cholinergic precursor loading strategy has represented the first approach to treat cholinergic dysfunction and cognitive decline in AD. This therapeutic option was leaved due to the lack of efficacy of several precursors investigated in clinical trials. This is not the case of choline alphoscerate (L-alphaglyceryl-phosphoryl-choline), a cholinergic precursor studied with encouraging effects both in preclinical paradigms and in clinical trials. In this paper we summarize the first results obtained in the ongoing trial on the "Effect of association between a cholinesterase inhibitor and choline alphoscerate on cognitive deficits in AD associated with cerebrovascular impairment" (ASCOMAVA). The trial was designed to assess if association between the ChEI donepezil at the daily dose of 10 mg and choline alphoscerate at the daily dose of 1,200 mg/day was accompanied by changes in Mini Mental State Evaluation (MMSE), Basic Activities of Daily Living (BADL), Instrumental Activities of Daily Living (IADL) and Neuropsychiatric Inventory (NPI). This latter

ABDOLLAHI A. ACKLAND M.L. ADOLFSSON O. AGACINSKI G. AGOGIATOU C. AISEN P. AKAN P. ALEXANDRU A. ALEXOPOULOS P. ALIEV G. ALMKVIST O. ALTMAN J. AMENTA F. ANAYA F. ANDREASEN N. ANDRIEU S. ANGLARET H. ANTONIELLO K. ARBUCKLE J. ARNAUD C. ARNOLD A. ARVIEU J.J. ATWAL J. AUSÉN B. BABILONI C. BAKOGLIDOU E. BARBERGER-GATEAU P. BARTZOKIS G. BASAR E. BATEMAN R. BATSILA G. BEACH T.G. BECKER J.A. BECKER R.E. BECKETT L.

Aim: To evaluate the ability to consent to biomedical studies for elderly people, cognitively normal, suffering for mild cognitive impairment (MCI) or Alzheimer Disease (AD) using the University of California Brief Assesment Of Capacity To Consent (UBACC), questionnaire Validated among schizophrenic subjects versus the subjective evaluation of a physician concerning the understanding of the patient. Method: We included 61 subjects who had consented to participate to a biomedical research and had signed the consent form during a consultation for memory troubles. These patients then benefited from a comprehensive neuro-psychological, clinical and biological assessment during day-care hospital. Three groups were individualized: cognitively normal, MCI, AD. For the biomedical study, patients had agreed to give a supplementary blood sample compared to the usual blood sample to test for a hormone level. The consent form was once again explicated to the patient by a physician who subjectively evaluated if the patient had understood the study. Then, the patients were asked for the 10 questions of the UBACC scale (max score 20, during time: 10 minutes). This scale evaluates the understanding of the study's aim, risks and benefits. A comparison was made between subjective assessment and the UBACC rate. Results: The physician considered that 18/61 patients had not understood. These one had a lower UBACC score (7.55±3 vs 17.7±3; p<0.001), a lower Mini Mental State Examination score (MMSE/30) (21±6 vs 27±3; p<0.001) and were older (84.8±8 vs 77±5) compared to those who had understood. Moreover, the patients who had not understood were all suffering for AD and had an UBACC score12. In the « AD » group, those who had understood had a higher mean MMSE (23.71±4) compared to those who had not understood. Nevertheless, taking into account standard deviations, MMSE appeared to be a poor discriminative test to assess capacity to consent. Discussion: There is a good concordance between the UBACC score and the subjective evaluation by a physician in order to assess the capacity to consent. The inferior UBACC score to have understood was 12. These results merit to be confirmed in a larger cohort.

P53 - ABILITY TO CONSENT TO BIOMEDICAL RESEARCH `S EVALUATION AMONG ELDERLY COGNITIVELY IMPAIRED PATIENTS. E. DURON 1 , M. BOULAY, A.-S. RIGAUD, L. HUGONOT-DIENER (1. Broca university hospital, Paris, France)

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

RESULTS OF GUIDAGE ­ A 5-YEAR PLACEBO-CONTROLLED STUDY ON THE EFFICACY OF EGB761® 120MG TO PREVENT OR DELAY ALZHEIMER'S DEMENTIA ONSET IN ELDERLY SUBJECTS WITH MEMORY COMPLAINT. B. VELLAS 1,2,3 , N. COLEY 1,2,3 , P.-J. OUSSET 1,2,3 , G. BERRUT, J.-F. DARTIGUES, B. DUBOIS, H. GRANDJEAN, F. PASQUIER, F. PIETTE, P. ROBERT, J. TOUCHON, P. GARNIER4, H. MATHIEX-FORTUNET4, S. ANDRIEU1,2,3 FOR THE GUIDAGE STUDY GROUP (1. INSERM U558, Toulouse, F-31073, France; 2. Univ ersity of T oulouse III, T oulouse, F-31073, France; 3. Gerontopole, Toulouse University Hospital, Toulouse, France; 4. Ipsen , Paris, France; 5. Toulouse University Hospital, Department of Epidemiology and Public Health, Toulouse, France) Background: Researchers predict that the global prevalence of Alzheimer's dementia (AD) will quadruple by 2050 to more than 100 million. In consequence, prevention strategies with a good clinical risk/benefit ratio are crucial to decrease the incidence of dementia and its associated medical, social, familial and economic impact. The objective of the GuidAge study was to assess the potential efficacy of EGb761® to prevent or delay AD onset in elderly with memory complaints. Methods: The GuidAge study was a doubleblind randomized placebo-controlled parallel group trial in which subjects were treated with EGb761® 120mg or placebo bid for a 5 year-period. Subjects aged 70 years or older having spontaneously complained of memory problems to a general practitioner (GP) and who were free of dementia, major depression, generalised anxiety or objective memory impairment were included. 4066 patients were selected by 686 GPs. These patients were tested in 25 French memory centres in order to confirm the lack of dementia at inclusion. Of these, 2854 patients were finally enrolled and randomised. These patients undertook a complete cognitive battery annually in the same memory centre (or before the yearly evaluation in case of cognitive deterioration detected by their GP). The tests permitted the detection of any occurrence of dementias (primary end point of the study), using the DSMIV classification, and of Alzheimer's type according to NINCDS-ADRDA. Results: During the study, 134 individuals developed AD, including 61 patients (4.3%) in the EGb 761® group and 73 patients (5.2%) in the placebo group. The difference was not statistically significant (p=0.31). However, a statistically significant difference between EGb 761® and placebo was observed in patients treated for at least 4 years. In this prespecified planned analysis, 15 out of 947 patients (1.6%) in the EGb 761® group converted to AD versus 29 out of 966 (3.0%) in the placebo group (p=0.03). Complementary analyses are on-going to further investigate these differences. The favourable long-term safety profile of EGb 761® was confirmed. Conclusions: Other studies are necessary to confirm a potential effect of EGB 761 on prevention of dementia after a long-term exposition

p. S8 p. S2 p. S6 p. S16 p. S23 p. S8 p. S10 p. S14 p. S22 p. S18 p. S17 p. S3 p. S7 p. S10 p. S10, 21 p. S12 p. S12 p. S11 p. S14 p. S17 p. S8 p. S14 p. S23 p. S5 p. S3 p. S7 p. S9 p. S10 p. S5 p. S9 p. S10 p. S14 p. S15 p. S10 p. S13 CANTERO C.R. CANTET C. CAPUTO A. CARILLO M. CARMICHAEL O. CAMICIOLI R. CAROLI A. CAROTENUTO A. CARPENTER A.P. CARPENTER J. CARRIERE J. CASTAGNA A. CECCALDI M. CEDARBAUM J. CHAULET C. CHUPIN M. CICCOCIOPPO F. CINDY C. CLAASSEN J. CLARK C.M. COLEMAN R.E. COLEY N. COLOMBO M. COSTANTINO D.S. CRAFT S. CRANS G.G. CRENSHAW D. CROISILE B. CSERNANSKY J.G. CUBERAS G. CYNIS H. D'AMICO M.C. DA FONSECA E. DARTIGUES J.F. DE FAZIO P. p. S19 p. S5 p. S12 p. S2 p. S16 p. S12 p. S13 p. S23 p. S2, 8 p. S15 p. S10 p. S20 p. S15 p. S7 p. S6 p. S6 p. S14 p. S10 p. S8 p. S2, 8 p. S8 p. S7, 23 p. S17 p. S20 p. S2 p. S11, 18 p. S10 p. S6 p. S12 p. S10 p. S19 p. S14 p. S15 p. S23 p. S20

included evaluation of severity and of caregiver stress measures (NPIF and NPIS). At the moment this double-blind trial has completed the observation of 70 patients of the 210 planned. Patients were aged between 56 and 86 years (mean 75 + 10 years) and were included in the protocol with a MMSE score between 15 and 23. According to the protocol, patients should suffer from ischemic brain damage documented by neuroimaging (MRI and CT scan), with a score 2 in at least one subfield (white matter or basal ganglia) according to the ARWMC (new rating scale for age-related white matter changes, Wahlund et al., Stroke 2001;32:1318­22). Recruited patients were then randomly allotted to an active treatment group (donepezil + choline alphoscerate) or to a reference treatment group (donepezil + placebo). The 70 patients so far analyzed were treated for 6 months and were examined at recruitement and after 3 and 6 months of treatment. Consistent with literature data, in patients allotted to the reference treatment group, a slight time- dependent worsening of MMSE, IADL and NPIS scores and no changes in the BADL and NPIF scores were noticeable. Treatment with donepezil plus choline alphoscerate improved compared to donepezil alone the different items analyzed except the BADL that was slightly worsened. The above results suggest that association of the cholinergic precursor choline alphoscerate to the standard treatment with a ChEI may represent a therapeutic option to prolong beneficial effects of cholinergic therapies in AD patients with concomitant ischemic cerebrovascular disease.

INDEX AUTHOR

p. S13 p. S12 p. S5 p. S3 p. S19 p. S2, 3 p. S13 p. S19 p. S16 p. S7 p. S11 p. S10 p. S23 p. S10 p. S12 p. S5, 7, 23 p. S14 p. S5 p. S10 p. S15 p. S16 p. S14 p. S5 p. S11 p. S13 p. S19 p. S21 p. S12 p. S21 p. S6 p. S19 p. S8 p. S9, 19 p. S6 p. S2, 16

BEDELL B.J. BENNETT D. BENNYS K. BERRES M. BERRUT G. BESSE T. BEURDELEY P. BIFOLCHETTI S. BLACIOTI M. BLACK R. BLANCHARD J. BLENNOW K. BLIN O. BLONDIAUX E. BOADA M. BOCCARDI M. BOCCHETTA M. BOGDANOVIC N. BONANNI L. BOOIJ B.B. BORM G. BORRELLI I. BOULAY M. BOXER A.L. BRADLEY K.M. BRAGIN V. BREMONT F. BREWSTER S. BUCARRELLO A.-L. BUCKNER R.L. BUENDIA M. BUFFA R. BÜRGER K. BURNS D.K. CALAN O.G.

S23

DE JAGER C.A. DE LEON M.J. DE MEY S. DE MEL D. DE SARRO G. DECARLI C. DECRAEMER HILDE H. DEL CARMEN ARIAS ESPARZA M. DEL SER T. DEMUTH H.-U. DESANTI S. DESIRE L. DETOLEDO-MORRELL L. DIACONEASA A.G. DIDIC M. DODEL R. DONOHUE M. DORAISWAMY M. DORMONT D. DRAGO V. DRUCKER D. DUBOIS B. DUCHESNE S. DURON E. EDMAN G. EFKLIDES A. EINSTEIN R. EISELE T. ESCANYÉ M.C. FADILOGLU M. FAGAN A. FARIAS S. FARLOW M.R. FASANARO A.M. FÉART C. FELDER E. FERRARI E.D. FERRER I. FERRETTI V. FILLIT H. FINELLI L.A. FLEISCHER A.S. FLETCHER E. FLITTER M.L. FORLONI G.L. FÖRSTL H. FOSSI S. FOULDS K. FREESTONE D. FRIESENHAHN M. FRISONI G.B. FRÖLICH L. GABELLE A. GAFNER V. GAINOTTI S. GALEOTTI F. GAMBI D. GAMST A. GANZOLA R. GARCIA J.D. GARERI P. GARNERO L. GARNIER P. GAUTHIER S. GÉNY A.C. GERTZ H.-J. GHITA C. GILLETTE-GUYONNET S. GIRAUDEAU B. GLODZIK L. GOMEZ SALGADO J. GOR B. GRAF A. GRANDJEAN H. GRAUBNER S. GRAVE G. GREIG N.H. GREGG K. GRODSTEIN F. GROGNET P. GRUNDKE-IQBAL I. GRUNDMAN M. GUAITA A. GÜNTEKIN B. HAAG S. HADDAD R. HALLIKAINEN I. HARTIKAINEN P. HAMPEL H. HÄNNINEN T. HARTLAGE-RÜBSAMEN M. HARVEY D. HEDDEN T. HEFTI F. HEINZ A. HELISALMI S. HENDRIX S.B. HERNANDEZ I. HERUKKA S.-K. HEUSER I. HILTUNEN A. HO C. HUGONOT-DIENER L. HÜLL M. IMBERT G. IQBAL K. IRVING E. IZZICUPO F. JACK C. JACOBS D. JACOBY R. JAFARBAY J. JAFFARI M. JAGLA W. JAGUST W. JAHN H. JALIFE M.A. JAROS M. JESSEN F. JILG H. JOHNSON K.A. JOHNSTON C. KANG J.H. KARPPI P. KAYE J. KERN F. KILLIANY R.J. KIOSSEOGLOU G. KIVIPELTO M. KLAFKI H. KNOPMAN D.

The Journal of Nutrition, Health & A ging© V olume 14, Supplement 2, 2010

p. S3, 8 p. S3, 12 p. S10 p. S12 p. S20 p. S16 p. S10 p. S20 p. S8 p. S13, 19 p. S3 p. S10 p. S12 p. S22 p. S13 p. S4, 15 p. S7 p. S8 p. S6 p. S13 p. S16 p. S6, 23 p. S12 p. S23 p. S11 p. S19 p. S10 p. S16 p. S18 p. S13 p. S2 p. S18 p. S4 p. S23 p. S21 p. S14 p. S19 p. S10 p. S17 p. S5 p. S12 p. S8 p. S16 p. S8 p. S17 p. S16 p. S17 p. S15 p. S12 p. S5 p. S7, 12, 13, 20 p. S15, 18 p. S6 p. S5 p. S22 p. S22 p. S14 p. S2 p. S12 p. S19 p. S20 p. S6 p. S23 p. S9 p. S6 p. S15 p. S22 p. S5 p. S5 p. S3 p. S19 p. S6 p. S12 p. S23 p. S19 p. S17 p. S6 p. S18 p. S7 p. S10 p. S3, 17 p. S18 p. S17 p. S21 p. S4 p. S10 p. S13 P. S12 p. S4 p. S13 p. S13 p. S16 p. S9 p. S2, 8 p. S15 p. S12 p. S9 p. S10 p. S12 p. S15, 18 p. S13 p. S5 p. S23 p. S15 p. S12 p. S3, 17 p. S7 p. S22 p. S2, 12 p. S10 p. S3, 8 p. S14 p. S13 p. S19 p. S2 p. S15 p. S19 p. S7 p. S4, 15 p. S13 p. S9 p. S3 p. S7 p. S13 p. S2 p. S14 p. S12 p. S19 p. S3 p. S13 p. S2

KOIVISTO A. KORCZYN A.D. KORNHUBER J. KOUNTI F. KOZIEL V. KRAUTKRAMER M.J. KRISTIANSEN L. KROLAK-SALMON P. KUMAR L. KUNG H.F. KURZ A. LABAUGE P. LACAVA R. LACOSTA A.M. LANUTI P. LATGER-FLORENCE C. LEHÉRICY S. LEHMANN S. LEITES M.A. LEMARIÉ J.C. LERNER F. LESZEK J. LI J. LI Y. LINDAHL T. LIU D. LIU E. LIU T.E. LOHMANN S. LÖNNEBORG A. LORENZI M. LUTZ M. MA W.W. MAHNE M. MAKEEVA O. MALAPLATE-ARMAND C. MANES F. MANGIALASCHE F. MANSON J.E. MARCHISIO M. MARIGLIANO N.M. MARINI E. MARTINEZ G.A. MARTÍNEZ-LAGE P. MARUYAMA Y. MATHIEX-FORTUNET H. MATSUMOTO A. MATSUMOTO R. MECOCCI P. MEHTA P.D. MELSEN T. MENNITI-IPPOLITO F. MEREU R.M. MICHALCZYK A. MILLER D.S. MINTUN M.A. MISCIA S. MISEREZ A.R. MOHSENZADEH M. MONACO D. MONLEÓN I. MONSCH A.U. MONTINE T. MORARU R. MOREAU J. MORTENSEN D.L. MOSCONI L. MOTALLI R. MUHS A. MULIN E. MUNOZ J. NIEMZ C. NIKOLAIDOU E. NOURHASHEMI F. NUNEZ L. OBRENOVICH M.E. OLAZARAN J. OLDE RIKKERT M. ONOFRJ M. ORMEN M. ORTIZ P. OSTER T. OSTROWITZKI S. OULHAJ A. OUSSET P.J. PAAJANEN T. PAEZ A. PALACIOS H.H. PANDO M. PANTEL J. PARK D. PARK M.H. PASQUIER F. PATERNICÒ D. PAUL R. PÉREZ-GRIJALBA V. PERNECZKY R. PESINI P. PETERS O. PETERSEN R. PFEIFER A. PHAM K. PHELPS C. PIERDOMENICO L. PIETTE F. PIEVANI M. PIHLGREN M. PILLOT T. PIORKOWSKA K. PIROUZI A. PIRRAGLIA E. PIRTTILÄ T. PLA R. PONTECORVO M.J. POPTSI E. PRATICO D. PRIETO D. PROUST-LIMA C. PRUESSNER J.C. PUTZU P.F. QUADRIO I. QUARG P. RAPP T. RASCHETTI R. REA R. REDOLFI A. REFSUM H. REIMAN E. REIMAN P.E.M. RIGAUD A.-S. RIVIERE M.E.

S24

p. S12 p. S11 p. S15, 18 p. S19 p. S18 p. S8 p. S17 p. S6 p. S12 p. S8 p. S16 p. S6 p. S20 p. S21 p. S14 p. S15 p. S6, 12 P. S6 p. S19 p. S9, 10 p. S20 p. S7 p. S3 p. S3 p. S17 p. S18 p. S18 p. S18 p. S5 p. S17 p. S7, 20 p. S10 p. S20 p. S15 p. S10 p. S18 p. S20 p. S3 p. S7 p. S14 p. S20 p. S14 p. S19 p. S21 p. S10 p. S23 p. S11 p. S11 p. S3 p. S3 p. S4 p. S22 p. S14 p. S12 p. S15 p. S2, 8 p. S14 p. S16 p. S13 p. S14 p. S21 p. S16 p. S2 p. S22 p. S12 p. S5 p. S3 p. S18 p. S5 p. S9 p. S10 p. S19 p. S19 p. S5 p. S10 p. S7 p. S10 p. S8 P. S14 p. S13 p. S10 p. S18 p. S7 p. S3, 8 p. S23 p. S13 p. S10 p. S7 p. S9, 10 p. S12 p. S2 p. S14 p. S23 p. S7 p. S5 p. S21 p. S16 p. S21 p. S15, 18 p. S2 p. S5 p. S5 p. S2 p. S14 p. S23 p. S13 p. S5 p. S18 p. S5 p. S13 p. S3 p. S13 p. S10 p. S2, 8 p. S19 p. S3 p. S19 p. S21 p. S12 p. S14 p. S6 p. S12 p. S5 p. S22 p. S23 p. S12, 13 p. S3, 8 p. S2 p. S8 p. S23 p. S12

ROBERT P. ROBITAILLE N. ROCA I. ROCHE S. ROMDHANE R. ROS J. ROSELLI F. ROSES A.D. ROßNER S. RUSINEK H. RYE P.D. SABBAGH M.N. SABER M. SADLI N. SADOWSKY C.H. SALVINI PORRO G. SAMIERI C. SAMUELSON P. SAN JOSÉ I. SANCHEZ L.M. SANTA CRUZ N.E. SANTAMARINA G. SARAGAT B. SARAZIN M. SATTLER C. SAUNDERS A. SCARPAZZA C. SCHELTENS P. SCHILLING S. SCHIPKE C.G. SCHNEIDER J.A. SCHÖNKNECHT P. SCHRÖDER J. SCHWARZ C. SEPPALA T. SEVIGNY J. SHARMA P. SHAW L. SHIKAKU N. SIEMERS E. SINCLAIR A. SJÖGREN M. SKOVRONSKY D.M. SLATS D. SMITH A.D. SMITH S.M. SNYDER P.J. SOININEN H. SOL O. SOLIS ARIAS M.P. SOLIS ARIAS P.E. SOLIS ARIAS R.I. SOLIS HERRERA A. SOLOMON A. SOUBOUROU A. SPERLING R. SPIEGEL R. SPIES P. SPIRU L. SPITZER P. SPOTTKE A. STERN Y. STÜTZER H. SUÁREZ M.L. SUGIMOTO H. SUPHIOGLU C. TÁRRAGA L. THIES W. THOLANCE Y. THOMAS A. TIERS L. TOGA A. TONELLI I. TONI N. TORRES M. TORO P. TOUCHON J. TOYOMOTO M. TRACTENBERG R.E. TRAINI E. TRAMONI E. TROJANOWSKI J. TSENG J. TSOLAKI M. TSOLAKIDOU A. TSUI W. TSUMOTO H. TUDOR P DEL C. TURCU I. VÄLIMÄKI T. VALLABHAJOSULA S. VAN AALST W. VAN DE POL L. VANACORE N. VANHANEN M. VANMECHELEN E. VARISCO Y. VECCHIO F. VELLAS B. VENTURI L. VIALE D. VIGHETTO A. VILLANI S. VILLARREAL J.A. WAGSTER M. WANG J.-Z. WATSON C. WATTS R.J. WEIMER R. WEINER M. W. WELLS B.M. WELSH-BOHMER K. WENDELSTEIN B. WESTERTEICHER C. WHITBREAD P. WIETEK S. WILCOCK G. WILTFANG J. WINBLAD B. WOLF H. XIE Z.-L. YAFFE K. YARAHMADI Z. YEN F.T. YENER G.G. ZAFEIROPOULOS S. ZAFEIROPOULOU M. ZETTERBERG H. ZHONG Y. ZHU L.-Q. ZINKOWSKI R.

p. S9, 23 p. S12 p. S10 p. S6 p. S9 p. S12 p. S16 p. S10 p. S13 p. S3 p. S17 p. S8 p. S14 p. S12 p. S8 p. S17 p. S21 p. S15 p. S21 p. S19 p. S19 p. S21 p. S14 p. S6 p. S13, 20, 21 p. S10 p. S7 p. S15 p. S13, 19 p. S18 p. S8 p. S20, 21 p. S13, 14, 15, 20, 21 p. S16 p. S12 p. S12 p. S17 p. S2 p. S11 p. S2 p. S12 p. S8 p. S2, 8 p. S8 p. S3, 8 p. S3 p. S4 p. S12, 13 p. S9, 10 p. S20 p. S20 p. S20 p. S7, 20 p. S3 p. S19 p. S2, 9 p. S16 p. S8 p. S22 p. S13 p. S15 p. S2, 4 p. S15 p. S21 p. S11 p. S12 p. S10, 21 p. S2 p. S6 p. S14 p. S6 p. S2 p. S6 p. S5 p. S10 p. S13, 20, 21 p. S6, 23 p. S11 p. S11, 16 p. S23 p. S15 p. S2 p. S8 p. S19 p. S16 p. S3 p. S11 p. S19 p. S22 p. S13 p. S3 p. S8 p. S15 p. S22 p. S13 p. S10 p. S5 p. S13 p. S5, 7, 23 p. S13 p. S14 p. S6 p. S17 p. S19 p. S2 p. S18 p. S12 p. S5 p. S5 p. S2, 7 p. S9 p. S10 p. S14 p. S16 p. S3 p. S4 p. S5 p. S13, 15 p. S3 ,12 p. S15 p. S20 p. S2 p. S13 p. S18 p. S21 p. S19 p. S19 p. S3 p. S20 p. S18 p. S3

Information

WESTERGREN

23 pages

Find more like this

Report File (DMCA)

Our content is added by our users. We aim to remove reported files within 1 working day. Please use this link to notify us:

Report this file as copyright or inappropriate

12603


You might also be interested in

BETA
2006 Eurocode - Interim Report.doc
Thomson
Microsoft Word - C05 SP Dehidroepiandrosterona.doc
WESTERGREN