Lewy body co-pathology in Alzheimer's disease and primary age-related tauopathy contributes to differential neuropathological, cognitive, and brain atrophy patterns

Abstract

Introduction: Alzheimer's disease (AD) co-pathology with Lewy bodies (LB) is frequent and influences clinical manifestations and outcomes. Its significance in primary age-related tauopathy (PART) is unknown. We investigated the influence of LB on cognition and brain atrophy in AD and PART.

Methods: We performed a retrospective cohort study in a large sample of autopsied participants with AD neuropathological change (ADNC) with and without LB and PART with and without LB, with corresponding ante mortem magnetic resonance imaging (MRI) data from the National Alzheimer's Coordinating Center dataset.

Results: LB co-pathology worsened cognitive impairment in both PART and ADNC. On longitudinal follow-up, LB impacted cognitive decline in multiple domains. Additionally, LB influenced brain atrophy on MRI across groups and LB regional staging was different in PART and ADNC, accompanying tauopathy progression.

Discussion: These results suggest that LB co-pathology is associated with divergent patterns of cognitive impairment, brain atrophy, and regional pathological distribution in PART and AD.

Highlights: Lewy body (LB) co-pathology is frequent in Alzheimer's disease (AD) with important clinical implications. LB co-pathology is also present in primary age-related tauopathy (PART), but its significance is still understudied. In PART and AD, LB leads to higher cognitive impairment and brain regional atrophy. In PART and AD, LB tends to accompany neurofibrillary tangle progression, suggesting amyloid pathology might be a trigger for regional pathology progression.

Keywords: Alzheimer's disease; Lewy bodies; MRI; co‐pathology; primary age‐related tauopathy.

FIGURE 1

The presence of LB associates with cognitive impairment in PART. (A) The comparison of CDR‐SB residuals after linear regression with age across the four groups. ADNC+LB showed the highest dementia severity followed by ADNC w/o LB, PART+LB and PART w/o LB, with all pairwise comparisons statistically significant (*p < 0.05; **p < 0.01; ***p < 0.001, with Dunn's test after Holm correction). (B) The proportion of dementia severity based on CDR global shows the same gradient across groups. All pairwise comparisons were statistically significant (p < 0.05 with chi‐squared test after Holm correction). Sample size, n = 1955 (PART w/o LB, n = 263; PART+LB, n = 75; ADNC w/o LB, n = 953; ADNC+LB, n = 664). ADNC, Alzheimer's disease neuropathological change; ADNC+LB, ADNC with LB; ADNC w/o LB, ADNC without LB; CDR, Cognitive Dementia Rating Scale; Dem, dementia; LB, Lewy bodies; MCI, mild cognitive impairment; NC, normal cognition; PART w/o LB, PART without LB; PART, primary age‐related tauopathy; PART+LB, PART with LB.

FIGURE 2

The presence of LB is associated with higher regional atrophy ADNC and PART. This figure shows the comparison of cortical and subcortical volumes for ADNC and PART with and without LBs. z‐scores based on mean difference in brain volume residuals after linear regression with age at MRI are shown in the brain graph in (A) PART+LB versus PART w/o LB; (B) PART w/o LB versus ADNC w/o LB; (C) PART w/o LB versus ADNC+LB; (D) PART+LB versus ADNC w/o LB; (E) PART+LB versus ADNC+LB; (F) ADNC+LB versus ADNC w/o LB. The presence of LB pathology was associated with lower cortical and subcortical volume in both ADNC and PART, whereas PART w/o LB pathology showed higher volumes in widespread cortical and subcortical areas compared with ADNC w/o LB and ADNC+LB. Blue regions represent lower volumes for the first group in the comparison, whereas yellow represents higher volumes. Only significantly different comparisons are shown (p < 0.05 after FDR correction for Welch's ANOVA or Kruskall–Wallis Test and p < 0.05 after TukeyHSD or Dunn's test with Holm correction). When not significant, group comparisons are not shown. Sample size, n = 322 (PART w/o LB, n = 38; PART+LB, n = 13; ADNC w/o LB, n = 157; ADNC+LB, n = 114). ADNC, Alzheimer's disease neuropathological change; ADNC+LB, ADNC with LB; ADNC w/o LB, ADNC without LB; FDR, false discovery rate; LB, Lewy bodies; MRI, magnetic resonance imaging; PART w/o LB, PART without LB; PART, primary age‐related tauopathy; PART+LB, PART with LB.

FIGURE 3

ADNC+LB and PART+LB show different topographical patterns of LB deposition. This figure shows the proportion of participants according to LB location and Braak stage in (A) PART+LB; (B) Mild NP+LB; (C) Moderate NP+LB; and (D) Severe NP+LB. In PART+LB and mild NP+LB, most patients located in early tau Braak stages (I–IV) and balanced LB stages, whereas in moderate to severe NP+LB, most patients localized to neocortical tau Braak stages (V–VI) and limbic and neocortical LB stages. Sample size, n = 1955 (PART w/o LB, n = 263; PART+LB, n = 75; ADNC w/o LB, n = 953; ADNC+LB, n = 664). ADNC, Alzheimer's disease neuropathological change; ADNC+LB, ADNC with LB; ADNC w/o LB, ADNC without LB; LB, Lewy bodies; NP, neuritic plaques; PART w/o LB, PART without LB; PART, primary age‐related tauopathy; PART+LB, PART with LB.