PET differences between cognitively unimpaired stable and declining subtypes: Multiple approaches show higher amyloid and tau among those with sub‐clinical cognitive decline

Abstract

AbstractBackgroundBiomarker‐related cognitive decline begins well before mild cognitive impairment (MCI), yet detecting subclinical decline is challenging. The presence of differences in AD biomarkers between cognitively unimpaired stable and declining subtypes (CU‐S and CU‐D) might indicate validity evidence for CU‐D criteria. This study examines whether brain amyloid and tau biomarkers differed across concurrent determinations of CU‐S, CU‐D and MCI/Dementia statuses using consensus conference and algorithm‐based cognitive statuses.MethodWRAP participants with >=3 Preclinical Alzheimer’s Cognitive Composite (PACC3) scores, free of dementia at baseline PACC3, and >=1 positron emission tomography (PET) [C‐11]Pittsburgh Compound B (PiB) scan were included (n=269); n=215 also had [F‐18]MK‐6240 tau scans. The mean(sd) age at last cognitive assessment was 67.0(6); PET PiB and MK scans occurred within means of 1.6(2) and .8(1.1) years of cognitive assessment. Cognitive statuses of CU‐S, CU‐D, and MCI/Dementia were assigned via consensus conference review. Six algorithm‐based cognitive status indicators were also examined (3 PACC3 composite‐based and 3 based on combinations of individual tests; for all, CU‐D range = centiles [4, 16), or ∼1 to 1.75 SD below expected; see Table 1). We compared continuous Global PiB DVR and MK‐6240 regional SUVR outcomes across cognitive statuses using analysis of variance and proportions PiB+ (Global PiB DVR>1.2) across cognitive statuses using Fisher’s exact tests. Significant omnibus tests (p<.05) were followed with pairwise comparisons.ResultTable 1 shows cognitive statuses by determination method. All omnibus tests comparing biomarkers across CU‐S, CU‐D and MCI groups were significant. CU‐D had higher proportion PiB+ than CU‐S for 4 of 7 methods (Figure 1). CU‐D differed from CU‐S on average Global PiB DVR (5 methods; Figure 2), MK‐6240 entorhinal cortex SUVR (6 methods, Figure 3), and MK‐6240 hippocampus SUVR (6 methods, Figure 4).ConclusionReaching a clinical diagnosis of Alzheimer’s or other dementia is the culmination of years of brain changes and initially subtle cognitive decline. The observed increased PET amyloid and tau pathology among cognitively unimpaired‐declining participants, particularly in cognitive status determination methods using longitudinal patterns, suggest CU‐D criteria can be sensitive to preclinical changes. Detecting subclinical cognitive decline may improve opportunity for early intervention that could ultimately delay clinical impairment.