MSA seems to be a more heterogeneous diagnosis than previously thought.

Eur J Nucl Med Mol Imaging

. 2025 Aug 27.

doi: 10.1007/s00259-025-07509-3. Online ahead of print.

Molecular imaging based spatiotemporal dynamics progression of brain glucose metabolism in multiple system atrophy

Xiaofeng Dou ^{# 1 2 3}, Jing Wang ^{# 1 2 3}, Daoyan Hu ^{# 1 2 3 4}, Haotian Wang ^{# 5}, Congcong Yu ^{1 2 3}, Rui Zhou ^{1 2 3}, Xiaohui Zhang ^{1 2 3}, Qiong Yao ^{1 2 3}, Mei Tian ⁶, Hong Zhang ^{7 8 9 10 11 12}, Yan Zhong ^{13 14 15 16 17}, Chentao Jin ^{18 19 20 21 22}

Affiliations Expand

• PMID: 40859026

DOI: 10.1007/s00259-025-07509-3

Abstract

Objective: Multiple system atrophy (MSA) is a progressive neurodegenerative disorder with complex clinical manifestations, which is essential for patient management and mechanistic understanding of MSA. In this study, we aimed to use disease progression modeling (SuStaIn model) to elucidate the in vivo spatiotemporal progression patterns of brain glucose metabolism in MSA patients, and investigate the differential profiles of clinical characteristics and dopaminergic function among the identified progression-related subtypes.

Methods: A total of 192 participants (117 MSA patients [70 MSA-P, 47 MSA-C] and 75 healthy controls) who underwent [¹⁸F]FDG PET scans, with 82 MSA patients additionally receiving [¹⁸F]FP-CIT PET imaging were retrospectively enrolled. [¹⁸F]FDG PET-based SuStaIn model was established to illustrate spatiotemporal evolutionary patterns of brain glucose metabolism using the cross-sectional data, and identified distinct metabolic subtypes. Metabolic subtypes and stages were correlated with motor function (UPDRS-III), cognitive function (MMSE, MoCA), autonomic symptoms, and dopamine transporter (DAT) activity.

Results: The [¹⁸F]FDG PET-based SuStaIn model identified two robust spatiotemporal metabolic progression subtypes, with Subtype 1 enriched in MSA-C (52.0%, 39/75) and Subtype 2 in MSA-P (78.8%, 26/33). Subtype 1 was characterized by initial hypometabolism in the cerebellum, sequentially progressing to brainstem, striatum, and cortical regions. Subtype 2 demostrated a striatal-onset pattern, progressing sequentially to the brainstem, cerebellum, and frontal lobe. Despite comparable disease duration, subtype 1 patients exhibited significantly poorer cognitive performance (MMSE, FDR q = 0.013; MoCA, FDR q = 0.032) and reduced anterior-to-posterior putamen DAT ratios (FDR q < 0.001) compared to subtype 2. Conversely, subtype 2 patients showed more obvious motor deficits (UPDRS-III, FDR q = 0.042). Significant correlations were observed between SuStaIn progression stages and clinical features across all patients, including UPDRS-III (r = 0.322, p = 0.001), MMSE (r = -0.263, p = 0.009), and MoCA scores (r = -0.292, p = 0.004). These results were confirmed in an independent validation cohort.

Conclusion: This study for the first time used [¹⁸F]FDG PET-based SuStaIn model to elucidate spatiotemporal dynamic progression of MSA, and identified novel metabolic subtypes. These findings provided metabolic evidence of the biological heterogeneity in MSA, which maybe helpful for patients managment and the understanding of mechanisms.