Lewy body pathology exacerbates brain hypometabolism and cognitive decline in Alzheimer's disease

Abstract

Identifying concomitant Lewy body (LB) pathology through seed amplification assays (SAA) might enhance the diagnostic and prognostic work-up of Alzheimer's disease (AD) in clinical practice and trials. This study examined whether LB pathology exacerbates AD-related disease progression in 795 cognitively impaired individuals (Mild Cognitive Impairment and dementia) from the longitudinal multi-center observational ADNI cohort. Participants were on average 75 years of age (SD = 7.89), 40.8% were female, 184 (23.1%) had no biomarker evidence of AD/LB pathology, 39 (4.9%) had isolated LB pathology (AD-LB+), 395 (49.7%) had only AD pathology (AD+LB-), and 177 (22.3%) had both pathologies (AD+LB+). The AD+LB+ group showed worst baseline performance for most cognitive outcomes and compared to the AD+LB- group faster global cognitive decline and more cortical hypometabolism, particularly in posterior brain regions. Neuropathological examination (n = 61) showed high sensitivity (26/27, 96.3%) and specificity (27/28, 96.4%) of the SAA-test. We showed that co-existing LB-positivity exacerbates cognitive decline and cortical brain hypometabolism in AD. In vivo LB pathology detection could enhance prognostic evaluations in clinical practice and could have implications for clinical AD trial design.

Fig. 1. Prevalence of CSF AD/LB groups in clinical and post-mortem populations.

A Barplots illustrate the prevalence of AD/LB groups across the whole cohort and split by MCI and dementia individuals. AD+ is defined as CSF p-tau₁₈₁/Aβ₄₂ positive and LB+ as α-synuclein SAA positive. B CSF AD/LB groups distribution within the post-mortem subpopulation is shown across neuropathological scores of regional distribution of Lewy body pathology, amyloid plaque Thal stages, neuritic plaques CERAD grading, tau Braak stages, and presence of TDP-43 in the hippocampus.

Fig. 2. AD/LB groups and global cognitive functioning.

Figure illustrates AD/LB group differences for (A, B) cross-sectional cognition based on the two-sided generalized linear model (n = 795) and (C, D) longitudinal cognitive performance based on the two-sided linear mixed models. All models were corrected for baseline age, sex, cognitive state, and level of education. Boxplots show the median, lower, and upper quartiles with whiskers representing minimum and maximum values. The spaghetti plots illustrate raw data regarding cognitive performance over time, while lines represent model fits (shaded area reflect 95% confidence interval). Models including an additional quadratic term for time better described the data. Lines within figure represent significant differences in AD/LB group*time². Only p-values adjusted for multiple comparison are shown. *p_adjusted < 0.05, **p_adjusted < 0.01, ***p_adjusted < 0.001.

Fig. 3. Regional hypometabolism as measured with FDG-PET.

Figure illustrates group differences assessed with two-sided linear mixed models in regional metabolism as measured with FDG-PET between AD/LB groups in the CI population (n = 568), after correction for baseline age, sex, and cognitive state. Only results that survived FDR-correction are shown. A Effects of AD/LB pathological groups vs the AD-LB- reference group, while (B) illustrates regions for which the AD+LB+ group demonstrated more severe hypometabolism at baseline compared to the AD+LB− group. C The spaghetti plots illustrate raw data regarding FDG-PET over time, while lines represent model fits (shaded area reflect 95% confidence interval), demonstrating a non-significant trend towards more apparent hypometabolism over time in the meta-ROI, consisting of regions identified in (B). ^#p = 0.06.