Magnetic Resonance Parkinsonism Index and midbrain to pons ratio: Which index better distinguishes Progressive Supranuclear Palsy patients with a low degree of diagnostic certainty from patients with Parkinson Disease?

Abstract

IntroductionSeveral studies have compared the performances of midbrain to pons area ratio (M/P) and the Magnetic Resonance Parkinsonism Index (MRPI) in distinguishing patients with Progressive Supranuclear Palsy (PSP) from those with Parkinson's disease (PD) with conflicting results. The current study aimed to compare the performance of these indexes in a well-characterized sample of PSP patients using either a manual or a fully automated approach to measure the brainstem structures involved in M/P and MRPI calculation. MethodsThis study involved 179 patients affected by idiopathic PD, 35 patients affected by PSP (15 probable and 20 possible) and 87 healthy controls. Sensitivity, specificity, positive predictive value (PPV) and area under the curve (AUC) of MRPI and M/P in distinguishing possible and probable PSP from PD and controls were calculated. ResultsNo significant difference was found between manual and automated values for both MRPI and M/P. MRPI and M/P differentiated probable PSP from PD with similar performance. By contrast, MRPI showed higher sensitivity and specificity than M/P when patients with possible PSP were compared with PD (MRPI, sensitivity 100%, specificity 98.88%; M/P, sensitivity 85%, specificity 93.85%). A significant difference was also observed in AUC between MRPI and M/P in distinguishing possible PSP from PD. ConclusionOur study demonstrates that MRPI was more accurate than M/P, in differentiating patients with possible PSP from those with PD. In patients suspected of having PSP with a low level of clinic diagnostic accuracy, MRPI should be preferred to M/P for distinguishing these patients from PD.