Nuklearmedizin - NuclearMedicine · Apr 2026
Semantic verbal fluency and FDG-PET predict survival time in memory clinic patients
By Frings L, Brumberg J, Meyer PT
OBJECTIVE: Semantic verbal fluency (sWFL) is frequently tested as part of the neuropsychological examination in dementia evaluation. We investigated the predictive value of sWFL as well as cerebral glucose metabolism and global amyloid burden in patients at a memory clinic. METHODS: Retrospective analysis of data from patients who had been examined with a neuropsychological test battery (CERAD-NP) as part of a dementia evaluation and whose vital status (07/2024) could be determined. We assessed the predictive value of sWFL (z-score, adjusted for age, sex, and education) for survival time after examination and tested the influence of amyloid status and global cognitive impairment (MMST score) as covariates. In a subgroup with available FDG- and PIB-PET, the predictive value of regional glucose metabolism and global amyloid burden (Centiloids) was investigated (age-adjusted Cox models). RESULTS: 280/583 included patients died over 8.0 years [7.7-8.4]. Better performance in sWFL was significantly associated with a lower mortality risk (HR=0.71 [0.63-0.8] per sWFL z-score, p<0.001). sWFL was the strongest predictor among the CERAD-NP subtests. Amyloid burden and global cognitive impairment did not substantially change the predictive value of sWFL (HR=0.70 [0.58-0.84] and HR=0.75 [0.66-0.85], respectively, both p<0.001). In patients with PET (N=218), FDG uptake in the left IFG (pars opercularis) was a significant predictor (HR=0.68 [0.55–0.84], FWEc p<0.05). The predictive value of sWFL was improved by adding FDG uptake in the right anterior cingulate cortex (FWEc p=0.078). Global amyloid burden was not a predictor of survival. CONCLUSIONS: sWFL can contribute to the prediction of survival time in memory clinic patients, independently of amyloid status and global cognitive impairment. Anterior cortical FDG uptake is itself a predictor of survival and slightly improves the accuracy of prediction by sWFL. Publication History Article published online: 02 April 2026 © 2026. Thieme. All rights reserved. Georg Thieme Verlag KG Oswald-Hesse-Straße 50, 70469 Stuttgart, Germany