MR planimetry in neurodegenerative parkinsonism yields high diagnostic accuracy for PSP. (January 2018)

• Record Type: Journal Article
• Authors: Mangesius, Stephanie; Hussl, Anna; Krismer, Florian; Mahlknecht, Philipp; Reiter, Eva; Tagwercher, Susanne; Djamshidian, Atbin; Schocke, Michael; Esterhammer, Regina; Wenning, Gregor; Müller, Christoph; Scherfler, Christoph; Gizewski, Elke R.; Poewe, Werner; Seppi, Klaus
• Is Part Of: Parkinsonism & related disorders. Volume 46(2018)
• Page Start: 47
• Page End: 55
• Publication Date: 2018-01
• Subjects: Magnetic resonance imaging (MRI) -- Parkinsonism -- Planimetry -- Differential diagnosis -- Diagnostic accuracy; Parkinson's disease -- Periodicals; Movement disorders -- Periodicals; Movement Disorders -- Periodicals; Nerve Degeneration -- Periodicals; Nervous System Diseases -- Periodicals; Parkinson Disease -- Periodicals; Tremor -- Periodicals; Parkinson, Maladie de -- Périodiques; Parkinson's disease; 616.833
• DOI: 10.1016/j.parkreldis.2017.10.020
• Languages: English
• ISSNs: 1353-8020

Abstract

Abstract: Introduction: Several previous studies examined different brainstem-derived MR planimetric measures with regards to their diagnostic accuracy in separating patients with neurodegenerative parkinsonian disorders and reported conflicting results. The current study aimed to compare their performance in a well-characterized sample of patients with neurodegenerative parkinsonian disorders. Methods: MR planimetric measurements were assessed in a large retrospective cohort of 55 progressive supranuclear palsy (PSP), 194 Parkinson's disease (PD) and 63 multiple system atrophy (MSA) patients. This cohort served as a training set used to build C4.5 decision tree models to discriminate PSP, PD and MSA. The models were validated in two independent test sets. The first test set comprised 84 patients with early, clinically unclassifiable parkinsonism (CUP). A prospective cohort of patients with PSP (n = 23), PD (n = 40) and MSA (n = 22) was exploited as a second test-set. Results: The pons-to-midbrain diameter ratio, the midbrain diameter, the middle cerebellar peduncle width and the pons area were identified as the most predictive parameters to separate PSP, MSA and PD in C4.5 decision tree models derived from the training set. Using these decision models, AUCs in discriminating PSP, MSA and PD were 0.90, 0.57 and 0.73 in the CUP-cohort and 0.95, 0.61 and 0.87 in the prospective cohort, respectively. Conclusion: We were able to demonstrate that brainstem-derived MR planimetric measures yield high diagnostic accuracy for the discrimination of PSP from related disorders when decision tree algorithms are applied, even at early, clinically uncertain stages. However, their diagnostic accuracy in discriminating PD and MSA was suboptimal. Highlights: Results: Brainstem planimetry has high diagnostic accuracy for parkinsonian syndromes. The defined decision tree algorithms yield highly accurate differential diagnoses. Even early, clinically uncertain stages can reliably be differentiated.