Differentiating Mild Cognitive Impairment, Alzheimer's Disease, and Dementia With Lewy Bodies Using Cingulate Island Sign on Perfusion IMP-SPECT

• Main Authors: Hidekazu Kanetaka, Soichiro Shimizu, Yuta Inagawa, Daisuke Hirose, Naoto Takenoshita, Hirofumi Sakurai, Haruo Hanyu
• Format: Article
• Language: English
• Published: Frontiers Media S.A. 2020-11-01
• Series: Frontiers in Neurology
• Subjects: cingulate island sign; IMP-SPECT; CIS; Alzheimer's disease; dementia with Lewy bodies (DLB)
• Online Access: https://www.frontiersin.org/articles/10.3389/fneur.2020.568438/full

The cingulate island sign (CIS) on fludeoxyglucose (FDG)-positron emission tomography (PET) is a supporting biomarker of dementia with Lewy bodies (DLB). Its diagnostic accuracy has only been investigated in FDG-PET, however. The present prospective study compared the CIS on I-iodoamphetamine-single photon emission computed tomography (SPECT) among patients with mild cognitive impairment (MCI), AD, or DLB. Fifty-eight patients with MCI, 42 with probable AD, and 58 with probable DLB were enrolled. The “CIScore” used to evaluate the CIS was defined as the ratio of volume of interest (VOI)-1 (indicating posterior cingulate gyrus [PCG]) to VOI-2 (area of significantly reduced regional cerebral blood perfusion [rCBF] in DLB patients compared with in healthy controls). It was calculated using eZIS software. The CIScore for MCI, DLB, and AD was 0.22, 0.23, and 0.28, respectively. The CIScore in the AD group was significantly higher than that in the DLB or MCI groups (AD vs. DLB: p < 0.001, AD vs. MCI: p < 0.005). This suggests that the CIScore can discriminate DLB from AD, if the decrease in rCBF in the PCG is similar between them. We believe that it is difficult to identify MCI based on the CIScore, as the decrease in rCBF in the PCG is not severe. The diagnostic accuracy of the CIScore may be low as it often shows an increase in elderly DLB patients, in whom the pathologically common form is most prevalent (1). Further study should include assessment of multiple components such as symptom classification and age.