Purpose: Cortical glucose metabolism, brain amyloid β accumulation and hippocampal atrophy imaging have all been suggested as potential biomarkers in predicting which patients with mild cognitive impairment (MCI) will convert to Alzheimer's disease (AD). The aim of this study was to compare the prognostic ability of [(11)C]PIB PET, [(18)F]FDG PET and quantitative hippocampal volumes measured with MR imaging in predicting conversion to AD in patients with MCI.
Methods: The study group comprised 29 patients with MCI who underwent [(11)C]PIB PET and MR imaging. Of these, 22 also underwent [(18)F]FDG PET. All subjects were invited back for clinical evaluation after 2 years.
Results: During the follow-up time 17 patients had converted to AD while 12 continued to meet the criteria for MCI. The two groups did not differ in age, gender or education level, but the converter group tended to have lower MMSE and Word List learning than the nonconverter group. High [(11)C]PIB retention in the frontotemporal regions and anterior and posterior cingulate (p < 0.05) predicted conversion to AD. Also reduced [(18)F]FDG uptake in the left lateral temporal cortex (LTC) predicted conversion (p < 0.05), but quantitative hippocampal volumes did not (p > 0.1). In receiver operating characteristic (ROC) analysis the measurements that best predicted the conversion were [(11)C]PIB retention in the lateral frontal cortex and [(18)F]FDG uptake in the left LTC. Both PET methods resulted in good sensitivity and specificity and neither was significantly superior to the other.
Conclusion: The findings indicate that [(11)C]PIB and [(18)F]FDG are superior to hippocampal volumes in predicting conversion to AD in patients with MCI.