18F-Florbetapir PET in Patients with Frontotemporal Dementia and Alzheimer Disease

Abstract

Pathologic deposition of amyloid β (Aβ) protein is a key component in the pathogenesis of Alzheimer disease (AD) but not a feature of frontotemporal dementia (FTD). PET ligands for Aβ protein are increasingly used in diagnosis and research of dementia syndromes. Here, we report a PET study using (18)F-florbetapir in healthy controls and patients with AD and FTD. Ten healthy controls (mean age ± SD, 62.5 ± 5.2 y), 10 AD patients (mean age ± SD, 62.6 ± 4.5), and 8 FTD patients (mean age ± SD, 62.5 ± 9.6) were recruited to the study. All patients underwent detailed clinical and neuropsychologic assessment and T1-weighted MR imaging and were genotyped for apolipoprotein E status. All participants underwent dynamic (18)F-florbetapir PET on a high-resolution research tomograph, and FTD patients also underwent (18)F-FDG PET scans. Standardized uptake value ratios (SUVRs) were extracted for predefined gray and white matter regions of interest using cerebellar gray matter as a reference region. Static PET images were evaluated by trained raters masked to clinical status and regional analysis. Total cortical gray matter (18)F-florbetapir uptake values were significantly higher in AD patients (median SUVR, 1.73) than FTD patients (SUVR, 1.13, P = 0.002) and controls (SUVR, 1.26, P = 0.04). (18)F-Florbetapir uptake was also higher in AD patients than FTD patients and controls in the frontal, parietal, occipital, and cingulate cortices and in the central subcortical regions. Only 1 FTD patient (homozygous for apolipoprotein E ε4) displayed high cortical (18)F-florbetapir retention, whereas (18)F-FDG PET demonstrated mesiofrontal hypometabolism consistent with the clinical diagnosis of FTD. Most visual raters classified 1 control (10%) and 8 AD (80%) and 2 FTD (25%) patients as amyloid-positive, whereas ratings were tied in another 2 FTD patients and 1 healthy control. Cortical (18)F-florbetapir uptake is low in most FTD patients, providing good discrimination from AD. However, visual rating of FTD scans was challenging, with a higher rate of discordance between interpreters than in AD and control subjects.