One year changes in PiB SUVR values in Alzheimer's disease, amnestic cognitive impairment, and cognitively normal subjects.

Objectives With standard uptake value ratio (SUVR) of amyloid PET we can estimate density of amyloid plaque representing progression of Alzheimer's disease (AD) and therapeutic effect of disease modifying drug, target of which is amyloid plaque. The objectives of this study were to investigate one-year changes of SUVR value of PiB PET in cognitively normal (CN), amnestic mild cognitive impairment (aMCI), and AD and to estimate the necessary sample size to detect significant change of the SUVR value.

Methods Subjects selected from an in-house study database were nine CN subjects (CNp) (age: 73.4±3.9 y.o.), nine patients with MCI (age:75.6±4.9), and five patients with AD (age:75.8±4.2) whose PiB PET images were visually rated as positive scans. Four PiB-negative CN (CNn) (age: 68.8±1.5 y.o) were also selected from the database. They underwent PiB PET and MRI with an interval of one year. PET images of 50-70 min after the injection of 555 (370-740) MBq PiB were masked with individual gray matter segmented MRI images and spatially normalized using 3D-T1 weighted MRI and DARTEL. The mean cortical SUVR (mcSUVR) values were calculated with the Automated Anatomical Labeling Atlas. Percent change of mcSUVR in one year was calculated in each subject. The sample sizes necessary to detect 5% changes of mcSUVR in a one-year interval at 80% power and a significance level of two-tailed 5% were calculated using GPower 3.1 software.

Results At the baseline the mcSUVR values in CNn, CNp, aMCI, and AD were 1.10±0.03, 1.42±0.15, 1.77±0.24, and 1.95±0.23, respectively. The mean and S.D. of percent changes of mcSUVR at the visit of one year in CNn, CNp, aMCI, and AD were 0.005±0.025, 5.76±4.06, 5.40±6.59, and 6.19±3.95, respectively. Twelve, 24, 58, and 22 cases were necessary to detect significant changes of mcSUVR in CNn, CNp, aMCI, and AD, respectively.

Conclusions Mean cortical PiB PET SUVR values on atlas-based ROI have sufficient power to monitor initial amyloid deposition and effects of anti-amyloid therapies