Background Deposits of fibrillar amyloid β (Aβ) protein are important in the pathophysiology of Alzheimer's disease (AD), but not in frontotemporal dementia (FTD). In vivo imaging of Aβ protein deposition, therefore, is a promising tool in the differential diagnosis of dementia syndromes. Previous pathological and imaging studies suggest that a minority of FTD patients may have coexistent Aβ deposition, possibly due to possession of the apolipoprotein E (APOE) ε4 allele. We report a PET study using the novel Aβ ligand [18F]–florbetapir in patients with AD, FTD and healthy controls.

Methods Ten healthy controls (mean age 62.5±5.2, MMSE 30), 10 AD patients (mean age 62.6±4.5 years, MMSE 10–24) and eight FTD patients (mean age 62.5±9.6, MMSE 0–30) were recruited to the study. All patients underwent detailed clinical and neuropsychological assessment, and were genotyped for APOE status. All participants underwent imaging with [18F]–florbetapir using the high–resolution research tomograph, and FTD patients also underwent [18F]–fluorodeoxyglucose (FDG) PET. Motion–corrected static florbetapir PET images (50–60 minutes post–injection) were coregistered with 3T volumetric T1–weighted MR images. Grey matter uptake values, intensity–normalised to the mean grey matter cerebellar uptake, were sampled from cortical and subcortical areas using a probabilistic anatomical atlas. Group differences were analysed using non–parametric Kruskal–Wallis test.

Results Total cortical grey matter florbetapir uptake values were significantly higher in AD (mean uptake ratio 1.77±0.38) compared to FTD patients (1.25±0.36, p=0.002) and controls (1.29±0.11, p=0.04). Florbetapir uptake was also higher in AD in frontal, parietal, occipital and central subcortical regions. PET images of seven of the FTD patients were visually classed as amyloid negative. One FTD patient (homozygous for APOE ε4) displayed high cortical florbetapir retention, with prominent mesiofrontal hypometabolism on FDG PET consistent with the clinical diagnosis. Three AD patients were APOE ε4 homozygous, while three AD and three FTD patients were ε3/4 heterozygotes. APOE ε4 gene dosage did not appear to be directly related to amyloid load.

Discussion Imaging amyloid deposition with [18F]–florbetapir in dementia contributes to differential diagnosis between AD and non–amyloid diseases such as FTD. Our findings are in line with previous reports of coexistent amyloid deposition in a small number of FTD patients. Florbetapir imaging will be of particular interest in subjects with progressive cognitive deficits that are difficult to classify clinically.