DOI: 10.1002/alz.079883 ISSN: 1552-5260

Amygdalar pathology in early neuropsychiatric phenotypes due to 3R‐Pick disease

Rachel M Keszycki, Allegra Kawles, Grace Minogue, Antonia Zouridakis, Alyssa Macomber, Vivienne Lubbat, Christina Coventry, Emily J Rogalski, Sandra Weintraub, Margaret E Flanagan, Rudolph J Castellani, Marsel Mesulam, Changiz Geula, Tamar Gefen
  • Psychiatry and Mental health
  • Cellular and Molecular Neuroscience
  • Geriatrics and Gerontology
  • Neurology (clinical)
  • Developmental Neuroscience
  • Health Policy
  • Epidemiology



Persons with dementia syndromes such as primary progressive aphasia (PPA) and behavioral variant frontotemporal dementia (bvFTD) often develop neuropsychiatric symptoms (NPS). Various neurodegenerative diseases may cause PPA or bvFTD, including frontotemporal lobar degeneration (FTLD) with 3‐repeat tau pathology (Pick disease, “PiD”). Limbic regions implicated in mood and behavior, such as the amygdala, are highly impacted in FTLD with PiD. This study quantified Pick bodies and activated microglia in the amygdala of individuals with antemortem diagnosis of PPA or bvFTD and postmortem PiD neuropathology and examined associations with initial NPS.


We identified 16 right‐handed cases from the Northwestern University Alzheimer’s Disease Research Center brain bank with bvFTD (N = 8) or PPA (N = 8) due to PiD with unilateral amygdala tissue (basolateral region) available. Fifteen cases had data on 12 NPS assessed via the Neuropsychiatric Inventory‐Questionnaire (NPI‐Q) from their earliest visit. We performed AT‐8 or HLA‐DR immunohistochemistry on amygdala sections to visualize Pick bodies or HLA‐DR+ activated microglia, respectively. We quantified the density of Pick bodies per mm3 via modified unbiased stereology and used HALO software (Indica Labs) to quantify percent area of HLA‐DR immunopositivity. We assessed differences between clinical groups via Welch’s t‐test (Pick bodies) or repeated‐measures two‐way ANOVA (HLA‐DR). Pearson clinicopathologic correlations were examined between Pick bodies or HLA‐DR immunopositivity and initial endorsement of total (out of 12) or behavioral/comportmental (apathy, disinhibition, motor stereotypies, or appetite disturbance) NPI‐Q symptoms.


Amygdalar Pick body densities were similar between PPA (M = 28,822/mm3) and bvFTD (M = 29,114/mm3) cases. Compared to bvFTD, PPA cases trended towards a lower percent area of HLA‐DR immunopositivity (5.54% vs. 2.39%, p<0.10). There was no significant relationship between Pick Body density and initial NPIQ symptoms endorsed across all cases. However, area of HLA‐DR immunopositivity was positively correlated with initial endorsement of total (r = 0.47, p<0.10) and behavioral/comportmental (r = 0.54, p<0.05) NPI‐Q symptoms.


Our findings suggest that microglial pathology in the amygdala contributes to early neuropsychiatric presentations in PiD, particularly symptoms reflective of behavioral/comportmental disturbance. Future study will investigate the impact of neuronal size and density in the amygdala on NPS in these dementia syndromes due to PiD.

More from our Archive