Centiloid method evaluation for amyloid PET of subcortical vascular dementia

Hyuk Jin Yun,Jong-Min Lee,Duk L Na,Kyung Han Lee,Hee Jin Kim,Yearn Seong Choe,Sang Won Seo,Samuel N Lockhart,Seung Hwan Moon

doi:10.1038/s41598-017-16236-1

Abstract

Reference region selection is important for proper amyloid PET analysis, especially in subcortical vascular dementia (SVaD) patients. We investigated reference region differences between SVaD and Alzheimer’s disease (AD) using Centiloid scores. In 57 [C-11] Pittsburgh compound B (PiB) positive (+) AD and 23 PiB (+) SVaD patients, we assessed standardized PiB uptake and Centiloid scores in disease-specific cortical regions, with several reference regions: cerebellar gray (CG), whole cerebellum (WC), WC with brainstem (WC + B), pons, and white matter (WM). We calculated disease group differences from young controls (YC) and YC variance according to reference region. SVaD patients showed large effect sizes (Cohen’s d > 0.8) using all reference regions. WM and pons showed larger YC variances than other regions. Findings were similar for AD patients. CG, WC, and WC + B, but not WM or pons, are reliable reference regions for amyloid imaging analysis in SVaD.

Highlights

The use of amyloid PET scans for quantitative measurement of amyloid-beta deposition has grown in recent years
There is concern that cerebellum may not be suitable as a reference region for analyses of amyloid burden in conditions other than late-onset Alzheimer’s disease (AD), as cerebellar amyloid deposits may be present in cerebral amyloid angiopathy (CAA)[13], prion diseases[14], and genetic AD15,16
Our findings show that cerebellar gray matter (CG), whole cerebellum (WC), and WC + B could be used as reference regions for amyloid imaging analysis in patients with Pittsburgh compound B (PiB) (+) subcortical vascular dementia (SVaD), which are the same for AD patients

Summary

Introduction

The use of amyloid PET scans for quantitative measurement of amyloid-beta deposition has grown in recent years. The ratio of target to reference region is an effective method for calculating amyloid load, with sufficient discriminatory power between Alzheimer’s disease (AD) and matched controls[2,4] This method is easy, useful, has obvious merit for clinical use as it requires no arterial blood sampling[2], and has been adopted widely for semi-quantification of amyloid load. Small differences in quantification result according to the reference region used can present an issue especially in subjects with borderline uptake. Mixed pathology is present in approximately half of all clinically diagnosed AD cases[6,7,8,9], even in clinical trials with participants extensively screened for pure AD10 Conditions such as subcortical vascular dementia (SVaD), which exhibit both CVD and www.nature.com/scientificreports/. We investigated the use of different reference regions in analyzing amyloid uptake in SVaD patients. We used five reference regions cerebellar gray matter (CG), whole cerebellum (WC), WC with brainstem (WC + B), pons, and WM which have been frequently used in previous studies

Methods

Results

Conclusion