MR Imaging Index for Differentiation of Progressive Supranuclear Palsy from Parkinson Disease and the Parkinson Variant of Multiple System Atrophy

Abstract

PURPOSE: To prospectively assess sensitivity and specificity of magnetic resonance (MR) imaging measurements of midbrain, pons, middle cerebellar peduncles (MCPs), and superior cerebellar peduncles (SCPs) for differentiating progressive supranuclear palsy (PSP) from Parkinson disease (PD) and Parkinson variant of multiple system atrophy (MSA-P), with established consensus criteria as reference standard. MATERIALS AND METHODS: All study participants provided informed consent; study was approved by the institutional review board. Pons area, midbrain area, MCP width, and SCP width were measured in 33 consecutive patients with PSP (16 possible, 17 probable), 108 consecutive patients with PD, 19 consecutive patients with MSA-P, and 50 healthy control participants on T1-weighted MR images. The pons area-midbrain area ratio (P/M) and MCP width-SCP width ratio (MCP/SCP) were also used, and an index termed MR parkinsonism index was calculated [(P/M).(MCP/SCP)]. Differences in MR imaging measurements among groups were evaluated with Kruskal-Wallis test, Mann-Whitney U test, and Bonferroni correction. RESULTS: Midbrain area and SCP width in patients with PSP (23 men, 10 women; mean age, 69.3 years) were significantly (P < .001) smaller than in patients with PD (62 men, 46 women; mean age, 65.8 years), patients with MSA-P (five men, 14 women; mean age, 64.0 years), and control participants (25 men, 25 women; mean age, 66.6 years). P/M and MCP/SCP were significantly larger in patients with PSP than in patients in other groups and control participants. All measurements showed some overlap of values between patients with PSP and patients from other groups and control participants. MR parkinsonism index value was significantly larger in patients with PSP (median, 19.42) than in patients with PD (median, 9.40; P < .001), patients with MSA-P (median, 6.53; P < .001), and control participants (median, 9.21; P < .001), without overlap of values among groups. No patient with PSP received a misdiagnosis when the index was used (sensitivity and specificity, 100%). CONCLUSION: The MR parkinsonism index can help distinguish patients with PSP from those with PD and MSA-P on an individual basis.