Multimodal brain imaging demonstrates that both amyloid burden and glucose metabolism predict the severity of cognitive impairment in patients with Alzheimer’s disease

Abstract

AbstractBackgroundThe purpose of this study is to assess brain glucose metabolism and Amyloid‐β burden in patients with Alzheimer’s disease with 3 Tesla PET/MR imaging and to seek possible relation of imaging findings with the extent of cognitive impairment.Method45 consecutive subjects with memory complaints were evaluated with a detailed neuropsychological test battery and F18‐Fluorodeoxyglucose(FDG)‐PET/MRI and F18‐Flutemetamol‐PET/MRI scans. The imaging modalities were performed with a hybrid 3T PET / MR camera at a maximum of 7 days intervals. Images were visually evaluated by two nuclear medicine experts. The centiloid scores were calculated through amyloid standardized uptake value ratios (SUVR) for all participants and regional glucose metabolism characteristics were analyzed.ResultThe study population consisted of 30 males and 15 females with a mean age of 65.9±8.2 years. Mean MMSE was 21.7±7.2 for the overall group. Five patients had imaging features suggestive of non‐AD type of dementias and were excluded from further analyses. Visual examination of F18‐Flutemetamol PET/MRI was consistent with the Alzheimer’s disease continuum in 29 cases, 9 of them demonstrated normal glucose metabolism, whereas 20 remaining cases showed F18‐FDG‐PET/MRI patterns characteristic to Alzheimer’s disease (hypometabolism in parieto‐temporal cortices and/or posterior cingulate cortex). In bivariate analysis, MMSE was positively correlated with the glucose metabolism in temporal lobe and posterior cingulate gyrus (r= 0.382, p=0.020; r=0.330, p=0.046; respectively), additionally, a significant negative correlation was observed between MMSE and amyloid burden designated by the centiloid score (r=‐0.461, p=0.004). Memory performance tested by Enhanced Cued Recall (ECR) was also positively associated with the glucose metabolism in posterior cingulate gyrus (r= 0.446, p=0.006), and negatively associated with the centiloid value (r= ‐0.458, p=0.005). The relationship between the ECR test and the centiloid value persisted when adjusted for age, gender, and year of education (β=‐0.121, p=0.021) in multivariate models.ConclusionGeneral cognitive ability and memory were predicted by imaging markers of regional glucose metabolism and amyloid deposition in this study population with early‐stage Alzheimer’s disease. Further research should investigate the predictive value of those imaging markers for the rate of cognitive decline. Acknowledgment: We acknowledge the support of GE Health Care for the [F18]flutemetamol doses.