POMD01 Blinded analysis of conventional MRI in a cohort of pathologically confirmed parkinsonian illnesses. Issue 11 (22nd October 2010)

• Record Type: Journal Article
• Authors: Massey, L; Paviour, D C; Fox, N C; Micallef, C; Burn, D J; Holton, J L; Lees, A J; Jager, H R; Revesz, T; O'Sullivan, S S
• Is Part Of: Journal of neurology, neurosurgery and psychiatry. Volume 81:Issue 11(2010)
• Page Start: e57
• Page End: e57
• Publication Date: 2010-10-22
• Subjects: Neurology -- Periodicals; Nervous system -- Surgery -- Periodicals; Psychiatry -- Periodicals; 616.8
• DOI: 10.1136/jnnp.2010.226340.161
• Languages: English
• ISSNs: 0022-3050

Abstract

Abstract : Background: Although conventional MRI (cMRI) has been reported to show specific abnormalities in parkinsonism, the diagnosis of the underlying pathology can be challenging. We report the radiological diagnostic accuracy, and the sensitivity, specificity, positive (PPV) and negative predictive values (NPV) of cMRI abnormalities in a pathologically confirmed cohort of Parkinson's disease (PD), progressive supranuclear palsy (PSP), multiple system atrophy (MSA) and corticobasal degeneration (CBD). Method: cMRI of donors to a neurological brain bank was reviewed by two experienced neuroradiologists blinded to clinical details. Images were systematically reviewed for the presence of specific reported abnormalities. Results: 48 pathologically confirmed cases (23 PSP, 6 PD, 13 MSA and 6 CBD) and nine controls were studied. The final clinical diagnosis was supported in 68% and the radiological diagnosis in 60%. Using cMRI 55% of PSP, 67% of PD, and 77% of MSA were correctly identified. In this study the Hummingbird sign has 68% sensitivity, 100% specificity and PPV for PSP and the Hot Cross Bun has 62% sensitivity, 100% specificity and PPV for MSA. Conclusion: The confirmation that using cMRI key diagnostic features with very high specificity and PPV have low sensitivity endorses the development of newer MR techniques (e.g., high field diffusion tensor imaging) as useful biomarkers early in disease.