hypothesized that VH is associated with less neurofibrillary tangle (NFT) pathology. Objectives: This study is to investigate the association between visual hallucinations in patients with Alzheimer’s (AD) and/or Lewy body pathology. Method: Clinical assessments were conducted as part of the University of Washington/Group Health Cooperative Alzheimer’s Disease Patient Registry on an annual basis. Each subject underwent detailed neuropsychiatric and cognitive assessments. A sub- set of subjects agreed to autopsy. All autopsied material underwent a standard neuropathological (np) examination. In addition, alpha- synuclein immunostaining was performed to fully characterize LBP in each case. One-hundred and ten AD and/ or LBP subjects with sufficient clinical and np data were eligible for this study. Results: Cases with np AD and concomitant LBP had the highest frequency of VH, compared to those with AD alone (34% versus 12%, Fisher’s exact test, p=.015). VH were present in 28% of subjects with LBP alone. Among LBP cases (n=68) those with limbic LBP have a higher frequency of VH, com- pared to those with LBP predominantly in the amygdala (42% versus 12%, p = 0.045). Although differences were not statistically significant, there is also an indication that, relative to cases with amygdala pre- dominant LBP, cases with brainstem LBP also have greater frequency of VH (43% versus 12%, p = 0.126). VH were present in 33% of subjects with neocortical predominant LBP. There was no significant association between Braak NFT staging and presence of VH. Conclu- sions: In our sample, cases with AD and concomitant LBP have the highest frequency of VH. Within LBP cases, frequencies of VH varied with LBP subtypes. Understanding the pathophysiological bases for the development of VH in LBP is critical in the diagnosis and treatment of patients with DLB. O1-04-02 BRAIN CORTICAL THICKNESS IN AN FTD3 PATIENT AND MUTATION CARRIERS Peter Johannsen 1 , Simon F. Eskildsen 2 , Jørgen Nielsen 3,1 , Elizabeth Fisher 4 , Dora Zeidler 5 , Anders Rodell 6 , Leif Østergaard 6 , 1 Memory Clinic, Department of Neurology, Rigshospitalet, Copenhagen University Hospital, Copenhagen, Denmark; 2 Department of Health Science and Technology, Aalborg University, Aalborg, Denmark; 3 Section of Neurogenetics, Institute of Medical Biochemistry & Genetics, The Panum Institute, University of Copenhagen, Copenhagen, Denmark; 4 Department of Neurodegenerative Disease, Institute of Neurology, London, United Kingdom; 5 Center for Functionally Integrative Neuroscience, Aarhus University Hospital, Aarhus, Denmark; 6 Center for Functionally Integrative Neuroscience, Aarhus University Hospital, Aarhus, Denmark. Contact e-mail: peterj@dadlnet.dk Background: Frontotemporal dementia linked to chromosome 3 (FTD3) (1) is caused by mutations in the CHMP2B gene in the endo- somal ESCRT-III complex (2). Objective: To study changes of cortical thickness in structural neuroimaging MR data from a large Danish FTD3 family. Methods: One FTD3 patient and 15 asymptomatic 50% risk family members were scanned in a 3 Tesla GE-Excite scanner (FSPGR 3D sequence). The 50% risk persons were genetically tested for the mutation and the investigators involved in subject assessment were blinded to the mutational status. The patient (male, 60 yr, MMSE = 22) had the FTD3 diagnosis confirmed by direct sequencing of the CHMP2B gene, showing the pathogenic G-to-C transition in the acceptor splice site of exon 6. Eight of the 50% risk persons were carrying the mutation and 7 were non-carriers. Cortical thickness was calculated with an updated previously published method (3) where 3D non-parametric surfaces were iteratively fitted to the inner and outer boundary of the cortex. The method calculates the cortical thickness with sub-voxel accuracy at more than 150,000 vertices equally distrib- uted over the cortical surface. The average cortical thickness in the frontal and temporal cortices for the patient, the pre-symptomatic mu- tation carriers and the first-degree no-mutation relatives are listed in the Table. Differences between those with and without the mutation had a trend but did not reach statistical significance (p=0.2, double-sided t-test). There was a highly statistical significant difference in cortical thickness between frontal and temporal cortices in both groups (p0.009). Age (years) Frontal cortex Temporal cortex mean SD mean SD Patient (III-53) 60 1.90 1.47* 2.22 1.90* Mutation carriers 54.8SD5.1 2.98 0.35¶ 3.57 0.42¶ No mutation 54.1SD5.8 3.24 0.40¶ 3.81 0.24¶ * = SD over cortex. ¶ = SD over subjects. Conclusion: Cortical thickness is reduced in FTD3. Pre-symptomatic sub- jects with the FTD3 CHMP2B exon 6 G-to-C mutation tend to have lower cortical thickness in frontal and temporal cortices compared to FTD3 family members without the mutation. References: (1) Gydesen. Neurol- ogy 2002; 59: 1585-94. (2) Skibinski. Nature Genetics 2005; 37: 806-8. (3) Eskildsen. Medical Imaging 2005; 5747: 1400-10. O1-04-03 CREUTZFELDT-JAKOB DISEASE (CJD) SURVEILLANCE IN JAPAN: TWO DISTINCT SUBTYPES OF CJD ASSOCIATED WITH CADAVERIC DURA MATER GRAFT AND THE FIRST JAPANESE CASE OF VARIANT CJD Masahito Yamada 1 , Moeko Shinohara-Noguchi 1 , Tsuyoshi Hamaguchi 1 , Yosikazu Nakamura 2 , Tetsuyuki Kitamoto 3 , Takeshi Sato 4 , Hidehiro Mizusawa 5 , vCJD Working Group, CJD Surveillance Committee, Japan, 1 Kanazawa University Graduate School of Medical Science, Kanazawa, Japan; 2 Jichi Medical School, Tochigi, Japan; 3 Tohoku University, Sendai, Japan; 4 Kohnodai Hospital, Ichikawa, Japan; 5 Tokyo Medical and Dental University, Tokyo, Japan. Contact e-mail: m-yamada@med.kanazawa-u.ac.jp Background: A large number of CJD cases associated with cadaveric dura mater graft (dCJD) have been reported from Japan. Recently, variant CJD (vCJD) has been also identified in Japan. Objective(s): To clarify the epidemiology and features of prion diseases in Japan. Meth- ods: Investigation of patients with prion diseases by the CJD Surveil- lance Committee, Japan, from April 1999 to September 2005. Results: We identified 712 patients with prion diseases including 552 patients (77.5%) with sporadic CJD, 95 (13.3%) with inherited prion diseases, 62 (8.7%) with infectious prion diseases, and 3 (0.4%) with unclassified CJD. For infectious prion diseases, all suffered from dCJD except for a case of vCJD. Taken together with dCJD cases identified by the previ- ous surveillance systems in Japan, we had 117 dCJD cases in total. Neuropathological analyses of the dCJD cases revealed two distinct subtypes, i.e., plaque type (p-dCJD) with formation of plaques (often, florid plaques) in the brain (57%), and non-plaque type (np-dCJD) without plaque formation (43%); np-dCJD presented with clinical man- ifestations typical of classical CJD, while p-dCJD presented with atyp- ical features such as slowly progressive ataxia and absence or late occurrence of periodic synchronous discharges (PSD) on EEG. The first Japanese case of vCJD had a history of short stay in the United Kingdom and other European countries in 1990. He developed mental changes in 2001, followed by painful dysesthesia, dementia, ataxia, and myoclonus, and died 42 months after onset. Autopsy established the diagnosis of vCJD. During the clinical course, PSD appeared on EEG, which had led to clinical misdiagnosis of probable sporadic CJD. Conclusions: The CJD surveillance in Japan revealed two distinct subtypes of dCJD: np-dCJD with classical CJD features, and p-dCJD with atypical clinicopathological features; the proportion of p-dCJD in dCJD would be higher than previously recognized. The first Japanese case of vCJD indicates that the appearance of the periodic EEG does not exclude the possibility of vCJD; we suggest revision of the WHO vCJD Case Definition (2001) to prevent missing cases of vCJD. S15 Oral O1-04: Other Dementias