Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Dissociable Associations of Semantic and Phonemic Fluency with Dual-Phase Amyloid PET in Amnestic Mild Cognitive Impairment
Poster Session F, Friday, October 2, 2:45 - 4:45 pm, Wangari Maathai
Junyoung Shin1, Seo Young Kang2, Jimin Park1, Mijung Park3, Jee Hyang Jeong2, Jee Eun Sung1,4; 1Ewha Womans University, 2Ewha Womans University Seoul Hospital, 3Neurophet Inc., 4Ewha Brain Institute
Across the Alzheimer's disease (AD) continuum, semantic and phonemic verbal fluency appear dissociable in their neural substrates. Semantic fluency shows closer associations with temporolimbic structures preferentially affected in early AD, whereas phonemic fluency is linked to prefrontal regions that remain relatively preserved. The resulting semantic-phonemic discrepancy has been considered a sensitive index of early, regionally selective degradation within temporolimbic semantic networks. Dual-phase positron emission tomography (PET) provides complementary indices of pathological and functional brain status. Whereas late-phase imaging indexes cerebral amyloid-beta accumulation, early-phase imaging reflects cerebral perfusion. However, no studies have examined whether early-phase PET captures functional vulnerability underlying verbal fluency performance. This study investigated how semantic fluency, phonemic fluency, and their relative discrepancy are associated with early- and late-phase PET uptake in individuals with amnestic mild cognitive impairment (aMCI), focusing on Braak-defined regions along the spatiotemporal trajectory of AD pathology. Thirty-two participants with aMCI were recruited. Clinical diagnoses were established using the Seoul Neuropsychological Screening Battery-II. All participants met Petersen's criteria: objective memory impairment, preserved general cognition, and Clinical Dementia Rating score of 0.5. Semantic fluency required generation of animal names within 60s. Phonemic fluency involved generation of words beginning with specified Korean letters (k/o/s) within 60s, averaged across letters. The semantic-phonemic discrepancy (SPD) was calculated as semantic minus phonemic fluency. All participants underwent dual-phase [18F]florbetaben PET imaging. Early-phase images (0-10min post-injection) indexed perfusion-related signals; late-phase images (90-110min) quantified amyloid-beta burden. Region-specific standardized uptake value ratios (SUVRs) were extracted and aggregated by Braak stages (early: I/II; intermediate: III/IV; advanced: V/VI). Pearson correlations examined associations between verbal fluency measures and SUVRs. Early-phase PET showed significant positive associations with semantic fluency at early stages (r=0.65, p<.001) and intermediate stages (r=0.64, p<.001), but not advanced stages (r=0.38, p=.09). Phonemic fluency showed no significant associations at any stage (all ps>.05). The SPD showed significant positive associations across all stages (early: r=0.56, p<.001; intermediate: r=0.58, p<.001; advanced: r=0.35, p<.05). Late-phase uptake showed no significant associations with any fluency measures at early stages (all ps>.05). At intermediate stages, late-phase uptake showed negative associations with semantic fluency (r=-0.56, p<.01) and SPD (r=-0.42, p<.01). These negative associations persisted at advanced stages (semantic: r=-0.51, p<.01; SPD: r=-0.38, p<.05). Phonemic fluency showed no associations with late-phase uptake at any stage (all ps>.05). Within a dual-phase PET framework, verbal fluency associations in aMCI were specific to semantic fluency and the relative semantic advantage, with no relationships for phonemic fluency. Early-phase uptake in entorhinal-hippocampal regions (Braak I/II) increased with semantic fluency and semantic advantage, whereas late-phase amyloid uptake showed no associations in these regions. This dissociation indicates that semantic fluency performance at early disease stages is more closely linked to functional network changes, potentially reflecting compensatory increases in regional perfusion. This pattern aligns with the neuroanatomical organization of semantic memory, which relies on medial temporal structures constituting the earliest cortical entry points of AD pathology. The findings demonstrate that semantic fluency in aMCI is supported by cerebral perfusion within entorhinal-hippocampal networks, detectable before robust associations with amyloid burden emerge.
Topic Areas: Language Production, Disorders: Acquired