Clinical Implications of an Integrated Clinical and Biological Staging Scheme for Alzheimer's Disease

Abstract

Background and purpose: This study aimed to characterize the clinical heterogeneity of Alzheimer's disease (AD) by implementing an integrated biological and clinical staging scheme for AD.

Methods: Clinical staging was performed in 193 participants from the Alzheimer's Disease Neuroimaging Initiative based on cognitive scores, while biological staging was performed based on global tau deposition in tau positron-emission tomography (PET). The discrepancy between clinical and biological stages was quantified as standardized residuals (W-scores), and classified into the following groups: concordant clinical and biological stages (W₀), worse clinical stage (W_-), and better clinical stage (W₊). Longitudinal changes in cognition, clinical progression, copathology burden, and brain metabolism on [18F]-fluorodeoxyglucose PET scans were compared between these groups.

Results: Relative to the W₀ group, the W_- group showed a faster cognitive decline and higher progression risk (hazard ratio [HR]=2.40, 95% confidence interval [CI]=1.20-4.83), while the W₊ group had a lower progression risk (HR=0.43, 95% CI=0.19-0.96). The copathology burden at autopsy (n=7) was correlated with the W-score (partial r=-0.87, p=0.023); however, this finding should be interpreted with caution due to the small sample. The ratio of cerebrospinal fluid α-synuclein positivity differed significantly between the groups, reaching 56.3% in the W_- group. Brain metabolism in the occipital, orbitofrontal, dorsolateral frontal, inferior and medial temporal cortex, and precuneus was lower in the W_- group than in the W₀ group, whereas it was higher in the W₊ group in the prefrontal, parietal, and temporal cortex.

Conclusions: The integration of clinical and biological staging has significant potential in clinical practice by providing information about copathologies, underlying neurodegeneration, and the progression of AD.

Keywords: Alzheimer's disease; copathologies; heterogeneity; metabolism; tau.

Fig. 1. Associations between the global tau PET SUVR and z-scores of mPACC5. Data are from general linear models of the mPACC5 z-scores using the global tau PET SUVR as a predictor. Covariates include age and sex. Each dot represents an individual, and predicted regression lines (black lines) with 95% confidence intervals (gray shading) are displayed. mPACC5, modified Preclinical Alzheimer Cognitive Composite 5; PET, positron emission tomography; SUVR, standardized uptake value ratio.

Fig. 2. Longitudinal changes in cognition among groups according to clinical and biological staging. Data are from linear mixed-effects models for longitudinal changes in memory function score (A), executive function score (B), ADAS-Cog13 score (C), and mPACC5 score (D) using W-score group, time, and W-score group×time as predictors, with APOE-ε4 carrier status, time, APOE-ε4 carrier status×time, age at baseline, and sex as covariates. High scores on the ADAS-Cog13 indicate worsening of cognitive function. ADAS-Cog13, Alzheimer’s Disease Assessment Scale Cognitive subscale 13-item; APOE, apolipoprotein E; mPACC5, modified Preclinical Alzheimer Cognitive Composite 5; W₀, concordant clinical and biological stages; W₋, worse clinical stage; W₊, better clinical stage.

Fig. 3. Progression to next clinical stage among groups according to clinical and biological staging. Progression to the next clinical stage was defined as occurring when the CDR global score at baseline progressed from 0 (normal) to 0.5 (MCI), 0.5 (MCI) to 1 (mild dementia), or 1 (mild dementia) to 2 (moderate dementia). A: Kaplan–Meier plot showing the actual duration in months to the progression event. B: Forest plots showing the HRs and 95% CIs in each W-score group for Cox proportional-hazards models with age, sex, and APOE-ε4 carrier status as covariates. CDR, Clinical Dementia Rating; CI, confidence interval; HR, hazard ratio; MCI, mild cognitive impairment; W₀, concordant clinical and biological stages; W₋, worse clinical stage; W₊, better clinical stage.

Fig. 4. Copathologies and CSF α-synuclein SAA positivity rates in the groups according to clinical and biological staging. A: Partial correlations between standardized residuals of clinical and biological staging (W-score) and sum of copathologies were assessed in autopsy-confirmed cases (n=7), controlling for the time interval between tau PET and autopsy. The sum of copathologies was calculated as the total sum of dichotomized statuses for Lewy body disease, limbic-predominant age-related TDP-43 encephalopathy, and cerebrovascular disease. B: Rates of CSF α-synuclein SAA positivity for each W-score group (38 positive cases). The number of positive cases over the total number of tested individuals in that group is indicated for each bar, along with the percentage of positive cases. CSF, cerebrospinal fluid; PET, positron-emission tomography; SAA, seed amplification assay; W₀, concordant clinical and biological stages; W₋, worse clinical stage; W₊, better clinical stage.

Fig. 5. Comparison of brain metabolism between the groups according to clinical and biological staging. A: Data from general linear models of ROI-based FDG metabolism, with group as a predictor, after controlling for age, sex, and education duration as covariates in all participants. B and C: Results from the same analyses in subgroups stratified by CSF α-synuclein SAA status. CSF, cerebrospinal fluid; FDG, [18F]-fluorodeoxyglucose; ROI, region of interest; SAA, seed amplification assay; W₀, concordant clinical and biological stages; W₋, worse clinical stage; W₊, better clinical stage.