ABSTRACT Frontotemporal dementia (FTD) is a common neurodegenerative cause of early age-of-onset dementia. The behavioral variant of FTD (bvFTD) results in profound changes in personality, as well as social and emotional functioning. Diagnostic uncertainty remains a common concern in bvFTD in spite of improved diagnostic criteria that focus on particular behavioral and cognitive features accompanied by characteristic neuroimaging patterns. Accurate diagnoses at the first visit are especially challenging, when cognitive impairment can be mild and functional abilities are more preserved. Other dementia syndromes can mimic features of bvFTD and there is partial overlap with the symptoms of psychiatric illnesses. With the clarity provided by longitudinal follow-up clinicians sometimes change bvFTD diagnoses. There are over 15 potential neuropathological diagnoses that underlie bvFTD. Even when a bvFTD diagnosis is clear it is challenging to predict the specific molecular subtype causing an individual patient’s symptoms. There is substantial heterogeneity in the clinical course in bvFTD as well, with some declining rapidly within 2-3 years and others surviving over more than a decade. Motor features may be present or absent, and their severity in the context of bvFTD can impact the level of care a patient requires. There are few reliable indicators to prognosticate a patient’s disease course. The proposed research will study 60 patients with bvFTD longitudinally with neurological examinations, cognitive testing, and structural and functional neuroimaging, and will retrospectively review the clinical features of 284 autopsied patients with bvFTD. The central hypothesis of this proposal is that in spite of the similarities between patients with bvFTD there will be clinical, neuropsychological, neuroimaging, serum marker, genetic, and gene expression differences that permit improved predictive certainty at the first visit. In Aim 1 we will use longitudinal assessment of diagnostic stability in order to determine factors that predict certainty in the bvFTD diagnosis. In Aim 2 we will identify clinical, gene expression, and imaging profiles in a cohort of autopsied patients with bvFTD that allow accurate prediction of a patient’s pathological diagnosis. In Aim 3 we will use longitudinal data on patients with bvFTD to determine factors that allow accurate prognostication of the clinical course. The results of the proposed research will provide guidance to clinicians who are considering a diagnosis of bvFTD and will improve the diagnostic and prognostic information available to patients, families, and clinicians, leading to improved clinical care from the time of the first visit. It will also facilitate enrollment of patients into observational research and molecularly targeted clinical trials.

PROJECT NARRATIVE Diagnostic pitfalls and clinicopathological heterogeneity limit the certainty of prognostication for patients with the behavioral variant of frontotemporal dementia (bvFTD). The focus of this project is to identify patterns of clinical features in patients at their first visit that increase certainty in the clinical diagnosis, prediction of the underlying pathological diagnosis, and precision of prognosis. This research will provide guidance to clinicians, increase the accuracy of information provided to patients and families, lead to refinement of diagnostic criteria, and facilitate enrollment in clinical trials.