monoclonal antibody, in patients with prodromal or mild Alzheimer’s dis- ease (AD). Retrospective analyses of both bapineuzumab and solanezumab data have shown substantially higher percentages of negative amyloid PET findings in APOE ε4 non-carriers than in APOE ε4 carriers in the mild and moderate AD sub-study populations. We report our amyloid PET screening results by APOE ε4 status in the prodromal and mild AD study populations. Methods: During screening, patients fulfilling clinical criteria for either prodromal or mild AD underwent florbetapir PET scanning and APOE gen- otyping. Florbetapir PET scans were visually evaluated for amyloid plaque burden. Results: Data from the first 250 patients were included in this anal- ysis. Similar to the bapineuzumab and solanezumab results, we have observed a substantially higher percentage of negative amyloid PET find- ings in APOE ε4 non-carriers than in APOE ε4 carriers (highlighted in bold in Table 1). However, the overall incidence of negative amyloid scans observed in 221AD103 is substantially higher than that reported in the two Phase III studies, likely attributable to the earlier stage of AD patients being recruited in this study (prodromal/mild, mean MMSE w 25 vs. mild/mod- erate AD, mean MMSE w21 in the Phase III studies). This finding is also consistent with the solanezumab results in that the percentage of negative amyloid PET findings was higher in mild (27%) than in moderate (13%) AD groups. Conclusions: These results suggest that: (1) selecting subjects having AD pathology based on clinical criteria remains a challenge; (2) enrichment by using amyloid PET imaging is effective and feasible; (3) enrichment by assessing amyloid plaque burden is critically important to clinical studies in early stages of AD because of a higher incidence of nega- tive amyloid findings; and (4) APOE genotyping may potentially improve the economy of enrichment by amyloid PET, via lowering the likelihood of negative amyloid findings. Table 1 Amyloid PET screening results by APOE ε4 status. Results from bapineuzumab and solanezumab Phase III studies (reported at HAI 2013) are also shown for comparison. APOE ε4 non-carrier APOE ε4 carrier All BIIB037 study 221AD103 Amyloid PET + 51 107 158 Amyloid PET - 68 24 92 % Amyloid PET - 57 18 37 Bapineuzumab study 1 % Amyloid PET - 36 7 16 Solanezumab study 1 % Amyloid PET - 38 9.5 22 1 Mintun M, et al. HAI Conference, 2013 O4-07-03 PATHOLOGIC VALIDATION OF THE EADC-ADNI HARMONIZED HIPPOCAMPAL PROTOCOL Liana Apostolova 1 , Chris Zarow 2 , Kristina Biado 3 , Sona Hurtz 4 , Marina Boccardi 5 , Johanne Somme 6 , Hedieh Honarpisheh 7 , Anna Blanken 1 , Jenny Brook 3 , Spencer Tung 3 , Denise Ng 3 , Jeffrey Alger 3 , Harry Vinters 8 , Martina Bocchetta 9 , Henri Duvernoy 10 , Clifford Jack, Jr., 11 , Giovanni Frisoni 12 , 1 UCLA, Los Angeles, California, United States; 2 University of Southern California, Downey, California, United States; 3 UCLA, Los Angeles, California, United States; 4 San Francisco State University, Los Angeles, California, United States; 5 LENTINEM, Brescia, Italy; 6 Alava Univesrsity Hospital, Victoria-Gasteiz, Spain; 7 Yale University, New Haven, Connecticut, United States; 8 University of California, Los Angeles, Los Angeles, California, United States; 9 IRCCS Centro S. Giovanni di Dio, Fatebenefratelli, Brescia, Italy; 10 Affiliation not available; 11 Mayo Clinic, Rochester, Minnesota, United States; 12 LENITEM, Brescia, Italy. Contact e-mail: lapostolova@mednet.ucla.edu Background: The gold standard for Alzheimer’s Disease (AD) diagnosis is post-mortem examination of brain tissue. Pathologic validation is the only definitive way to ascertain the validity of disease biomarkers. Hippocampal atrophy is the most established structural imaging biomarker for AD to date. European AD Consortium and AD Neuroimaging Initiative investigators recently developed a Harmonized Protocol for Hippocampal Segmentation (EADC-ADNI HarP). EADC-ADNI HarP has not yet been pathologically validated. Methods: The temporal lobes of 9 AD and 7 cognitively normal subjects (NC) were scanned post-mortem at 7 Tesla. Pathologic diagnosis of AD was based on Braak and Braak and CERAD criteria. The temporal lobes were scanned for 60 hours on a 7T Bruker Biospec MRI scanner. Hippocam- pal volumes were obtained with the EADC-ADNI HarP. 6 mm-thick hippo- campal slices were stained for amyloid beta (Ab1-40), tau and cresyl violet. The demarcations of each hippocampal subfield were manually drawn with Aperio ImageScope Ò CS on the digitally scanned stained tissue. Subfield margins were identified based on cytoarchitectonic features. Neuronal counts, Ab and tau burden for each hippocampal subfield were obtained. Results: Kruskal-Wallis comparison of medians showed significant differ- ences between the two groups for total hippocampal tau and Ab burden (p¼0.01 for both) but not neuronal count (p¼0.12). Significant differences in the medians were also seen in all subfields for tau and in the subiculum, CA1 and CA3 for Ab. We found significant correlations between hippocam- pal volume and fresh brain weight (r¼0.69, p¼0.003), Braak and Braak staging (r¼-0.71, p¼0.002), tau (r¼-0.53, p¼0.034) and Ab burden (r¼-0.61, p¼0.012). Subfield-wise significant association were found for Ab in CA1 (r¼-0.58, p¼0.019) and subiculum (r¼-0.75, p¼0.001), as well as tau in CA2 (r¼-0.59, p¼0.016) and CA3 (r¼-0.5, p¼0.047). Con- clusions: The observed associations provide pathologic validation for the EADC-ADNI HarP and pathologic confirmation of hippocampal morphom- etry as a valid AD biomarker. O4-07-04 AUTOMATED REPORTING OFAMYLOID PET QUANTIFICATION ON BRAIN SURFACE THROUGH AWEB INTERFACE Pierrick Bourgeat 1 , Vincent Dor e 2 , Luping Zhou 3 , Jurgen Fripp 4 , Ralph Martins 5 , Lance Macaulay 6 , Colin Louis Masters 7 , David Ames 8 , Belinda Brown 9 , Christopher Cleon Rowe 10 , Olivier Salvado 3 , Victor L. Villemagne 11 , 1 CSIRO, Herston, Queensland, Australia; 2 CSIRO, Melbourne, Australia; 3 CSIRO, Herston, Australia; 4 AeHRC, Herston, Queensland, Australia; 5 Edith Cowan University, Joondalup, Western Australia, Australia; 6 CSIRO, Parkville, Australia; 7 Florey Institute, Parkville, Australia; 8 National Ageing Research Institute Inc. (NARI), Parkville, Australia; 9 Edith Cowan University, Perth, WA, Australia; 10 Austin Hospital, Melbourne, Australia; 11 Austin Health, Melbourne, Australia. Contact e-mail: vincent.dore@csiro.au Background: Molecular brain imaging using Positron Emission Tomogra- phy (PET) is a robust diagnostic tool for which several tracers labelled with either 11C or 18F are available. For visual inspection of the images, cortical surface based visualisation presents the advantage of providing a compact and more convenient display than volumetric scans. We have developed an automated reporting tool to display the semiquantitative PET signal for several common tracers without the need of a MRI and using several Oral Sessions: O4-07: Neuroimaging: Clinical Trials and Clinical Tools P264