Association Between Postmortem Pathologic Burden and the Rate of Clinical Progression in Patients With Frontotemporal Lobar Degeneration.
BACKGROUND AND OBJECTIVES Histopathologic staging of Alzheimer disease has led to validation of imaging techniques that guide diagnosis and treatment. We previously constructed preliminary phases of the sequential progression of TDP-43 and tau to guide similar efforts in behavioral-variant frontotemporal dementia (bvFTD). In this article, we expand this work using digital pathology and longitudinal clinical data to more comprehensively model the relationship between clinical progression and the distribution and severity of postmortem frontotemporal lobar degeneration (FTLD) pathology. METHODS In this retrospective cohort study, 101 patients (42% female, median age at symptom onset = 63 years) were selected from the Penn Integrated Neurodegenerative Disease Database and had both longitudinal assessments and primary neuropathologic diagnosis of FTLD-Tau or FTLD-TDP. We used validated methods to quantify the burden of primary pathology from up to 6 cortical regions across hemispheres. FTLD-TDP pathologic phase was constructed from diagnostic pathology data based on published criteria. We tested the association between pathologic metrics and (1) disease duration or (2) the rate of clinic progression measured by 2 independent global measures (Clinical Dementia Rating Scale-Sum of Boxes [CDR-SB] and Mini-Mental State Examination [MMSE]). Linear regression and linear mixed-effects models were adjusted for hemisphere sampled, sex, age at onset, pathogenic variant status, and pathologic subtype. RESULTS Disease duration did not associate with pathologic burden in multiple regression (FTLD-TDP β = 0.01 [-0.06, 0.09]; p = 0.7; FTLD-Tau β = 0.1 [-0.4, 0.7]; p = 0.7). By contrast, mean TDP-43 burden, but not FTLD-Tau burden, was associated with both worse relative CDR-SB (β = 0.1 [0.06, 0.2]; p = 0.0001) and MMSE (β = -0.1 [-0.2, -0.03]; p = 0.009) among all FTLD-TDP patients. TDP-43 phase also associated with worse CDR-SB (β = 0.07 [0.02, 0.1]; p = 0.005) and MMSE (β = -0.2 [-0.3, -0.1]; p = 0.000005). TDP-43 burden (CDR-SB (β = 0.1 [0.03, 0.2]; p = 0.005 and MMSE (β = -0.2 [-0.4, -0.05]; p = 0.009)), but not phase (CDR-SB (β = 0.02 [-0.03, 0.08]; p = 0.4 and MMSE (β = -0.04 [-1, 0.07]; p = 0.5)), associated with relative decline in sensitivity analyses limited to bvFTD. DISCUSSION Greater TDP-43 burden was most closely associated with antemortem clinical decline rather than cumulative aggregation through the disease course. These human data suggest that the temporal dynamics of protein aggregation may differ among FTLD proteinopathies, with implications for the interpretation of FTLD-Tau and FTLD-TDP‑specific biomarkers as these are developed.