Pathological combinations in neurodegenerative disease are heterogeneous and disease-associated

Abstract

Pathologies that are causative for neurodegenerative disease (ND) are also frequently present in unimpaired, older individuals. In this retrospective study of 1647 autopsied individuals, we report the incidence of 10 pathologies across ND and normal ageing in attempt to clarify which pathological combinations are disease-associated and which are ageing-related. Eight clinically defined groups were examined including unimpaired individuals and those with clinical Alzheimer's disease, mixed dementia, amyotrophic lateral sclerosis, frontotemporal degeneration, multiple system atrophy, probable Lewy body disease or probable tauopathies. Up to seven pathologies were observed concurrently resulting in a heterogeneous mix of 161 pathological combinations. The presence of multiple additive pathologies associated with older age, increasing disease duration, APOE e4 allele and presence of dementia across the clinical groups. Fifteen to 67 combinations occurred in each group, with the unimpaired group defined by 35 combinations. Most combinations occurred at a <5% prevalence including 86 that were present in only one or two individuals. To better understand this heterogeneity, we organized the pathological combinations into five broad categories based on their age-related frequency: (i) 'Ageing only' for the unimpaired group combinations; (ii) 'ND only' if only the expected pathology for that individual's clinical phenotype was present; (iii) 'Other ND' if the expected pathology was not present; (iv) 'ND + ageing' if the expected pathology was present together with ageing-related pathologies at a similar prevalence as the unimpaired group; and (v) 'ND + associated' if the expected pathology was present together with other pathologies either not observed in the unimpaired group or observed at a greater frequency. ND only cases comprised a minority of cases (19-45%) except in the amyotrophic lateral sclerosis (56%) and multiple system atrophy (65%) groups. The ND + ageing category represented 9-28% of each group, but was rare in Alzheimer's disease (1%). ND + associated combinations were common in Alzheimer's disease (58%) and Lewy body disease (37%) and were observed in all groups. The Ageing only and Other ND categories accounted for a minority of individuals in each group. This observed heterogeneity indicates that the total pathological burden in ND is frequently more than a primary expected clinicopathological correlation with a high frequency of additional disease- or age-associated pathologies.

Keywords: Alzheimer’s disease; Lewy body disease; frontotemporal lobar degeneration; neurodegenerative disease; tauopathy.

Figure 1

ND pathologies, combinations and categories of combinations. Eight clinically defined cohorts were examined including unimpaired individuals and those with clinical Alzheimer’s disease, mixed dementia, ALS, FTD, MSA, probable LBD, or probable tauopathies. Ten assessed pathologies (see the ‘Materials and methods’ section) resulted in a heterogeneous mix of 161 pathological combinations, including 35 combinations in the unimpaired cohort. To better understand this heterogeneity, we organized the pathological combinations into five broad categories based on their age-related frequency: ‘Ageing only’ for the unimpaired group combinations; ‘ND only’ if only the expected pathology for that individual’s clinical phenotype was present; ‘ND + ageing’ if the expected pathology was present together with ageing-related pathologies at a similar prevalence as the unimpaired group; ‘ND + associated’ if the expected pathology was present together with other pathologies not observed in the unimpaired group or observed at a greater frequency and ‘Other ND’ if the expected pathology was not present. As an example, a subset of observed pathological combinations is shown for clinical Alzheimer’s disease. In this cohort, the combinations A-B-C and A-B-C-CAA are the ND only combinations. A-B-C-CVD was an ND + ageing combination, while A-B-C-TDP and A-B-C-LB-CAA were ND + associated combinations. Only rarely did clinical Alzheimer’s disease involve Other ND combinations such as TDP-B or Tau or Ageing only combinations such as CVD, A or A-B.

Figure 2

Prevalence of total number of pathologies and specific pathological combinations by clinical group. The total number of pathologies and the specific pathological combinations were variably distributed in each clinical group. (A) The presence of one or more pathologies was not uncommon in the unimpaired group and most cases in the clinical Alzheimer’s disease (AD) group had four or more pathologies. Between these two extremes, the prevalence of multiple pathologies was common in each ND group. (B) One hundred and sixty-one pathological combinations were observed across the cohort with 10 combinations accounting for approximately half of each clinical group. Consequently, the incidence of the remaining 151 combinations ranged from uncommon to rare, with 61 combinations only observed in individuals. The clinical LBD group was the most heterogeneous, the MSA group was the most homogeneous and 35 combinations were observed in the unimpaired group. Shown is a heat map showing the prevalence of the top 100 non-unique combinations after cubed-root transformation, generated using the heatmap function in R Version 4.2.1.

Figure 3

Pathological combination categories. The heterogeneous pathological combinations can be grouped into categories by expected ND including: ‘ND only’ if only the expected pathology or pathologies are present; ‘Other ND’ if the expected pathology is not present; ‘Ageing only’ for combinations also present in age-matched unimpaired cases; ‘ND + ageing’ if the expected pathology was present along with pathologies in age-matched unimpaired cases; and ‘ND + associated’ if the expected pathology was present together with other pathological combinations. We observed that ND only cases are uncommon except in the ALS and MSA groups and that ND + ageing only correlates with a minority of cases, with the implication that additional ND pathologies are frequently observed across disease.