Single-subject SPM FDG-PET patterns predict risk of dementia progression in Parkinson disease

A Pilotto, E Premi, SP Caminiti, L Presotto, Rosanna Turrone, Antonella Alberici, Barbara Paghera, B Borroni, A. Padovani, D Perani

Research output: Contribution to journalArticlepeer-review

Abstract

OBJECTIVE: To evaluate the statistical parametric mapping (SPM) procedure for fluorodeoxyglucose (FDG)-PET imaging as a possible single-subject marker of progression to dementia in Parkinson disease (PD). METHODS: Fifty-four consecutive patients with PD without dementia (age at onset of 59.9 ± 10.1 years, disease duration of 5.3 ± 3.4 years) entered the study. The patients underwent an extensive motor and cognitive assessment and a single-subject FDG-PET SPM evaluation at baseline. A 4-year follow-up provided disease progression and dementia diagnosis. RESULTS: The FDG-PET SPM was evaluated by 2 expert raters allowing the identification of a "typical PD pattern" in 29 patients, whereas 25 patients presented with "atypical patterns," namely, dementia with Lewy bodies (DLB)-like (n = 12), Alzheimer disease (AD)-like (n = 6), corticobasal syndrome (CBS)-like (n = 5), and frontotemporal dementia (FTD)-like (n = 2). At 4-year follow-up, 13 patients, all showing atypical brain metabolic patterns at baseline, progressed to dementia (PD dementia). The DLB- and AD-like SPM patterns were the best predictor for incident dementia (p <0.005, sensitivity 85%, specificity 88%), independently from demographics or cognitive baseline classification. CONCLUSIONS: This study suggests that FDG-PET SPM at the single-subject level might help in identifying patients with PD at risk of developing dementia. © 2018 American Academy of Neurology.
Original languageEnglish
Pages (from-to)e1029-e1037
Number of pages9
JournalNeurology
Volume90
Issue number12
DOIs
Publication statusPublished - 2018

Fingerprint

Dive into the research topics of 'Single-subject SPM FDG-PET patterns predict risk of dementia progression in Parkinson disease'. Together they form a unique fingerprint.

Cite this