Disease related metabolic patterns as imaging biomarkers for multiple system atrophy and progressive supranuclear palsy

Besides Parkinson’s disease, parkinsonisms comprises of atypical parkinsonisms - multiple system atrophy (MSA), progressive supranuclear palsy (PSP), and corticobasal degeneration. Distinguishing between them is difficult; therefore a biomarker is needed.

We identified metabolic brain patterns related to MSA (MSARP) and PSP (PSPRP) in Slovenian population (20 healthy controls (HC), 20 MSA/PSP patients), using ¹⁸F-fluorodeoxyglucose positron emission tomography brain images and scaled subprofile model/principal component analysis, which detect the greatest source of in-group variance. To evaluate differences between patients with either MSA or PSP syndrome, we developed a new method to extract the pattern’s regional sub-scores.

Patterns’ expression discriminated between HC and MSA/PSP patients as well as between different parkinsonisms (all at p < 0.001). Both patterns are specific and sensitive (MSARP: AUC = 0.96; PSPRP: AUC = 0.99), and the featured brain regions agree with the pathophysiology of the diseases. Sub-scores reveal, in line with clinical presentation, heterogeneity in MSA group.

The study confirms that both patterns are reliable biomarkers of MSA/PSP disease and could contribute to accurate clinical diagnosis.