Subtype Diagnosis of Sporadic Creutzfeldt–Jakob Disease with Diffusion Magnetic Resonance Imaging

Alberto Bizzi, R. Pascuzzo, J. Blevins, M. E. M. Moscatelli, M. Grisoli, R. Lodi, F. M. Doniselli, G. Castelli, M. L. Cohen, A. Stamm, L. B. Schonberger, B. S. Appleby, and P. Gambetti

Ann Neurol. 2020 Dec 3. doi: 10.1002/ana.25983. Epub ahead of print. PMID: 33274461.

Selezionato dal Lettore: Angelo Torrente, Università degli Studi di Palermo - AOUP “Paolo Giaccone” di Palermo

Motivation: 
Originality of the work: The researchers propose two procedures for the antemortem diagnosis of sCJD subtype using diffusion magnetic resonance imaging (MRI).

Methodology: researchers selected a large cohort of 487 patients with autopsy-confirmed diagnosis of “pure” sCJD subtype and at least one positive diffusion MRI examination from the National Prion Disease Pathology Surveillance Center (NPDPSC) in Cleveland, Ohio, and their MRIs were collected. An expert neuroradiologist made a huge work, prospectively scoring all diffusion MRIs, blind to the clinical data and preliminary diagnosis. He generated a lesion profile for each MRI study by evaluating the presence of DWI signal hyperintensities in 12 brain regions. Then, a first MRI data driven procedure for subtype diagnosis was designed (PriSCA_MRI), while a second one included even knowledge of the prion protein codon 129 genotype (PriSCA_MRI + Gen).

Data presentation: they adopted a machine learning method known as Classification and Regression Tree (CART) to classify subtypes of sCJD using MRI findings alone (PriSCA_MRI) or combined with PRNP129 findings (PriSCA_MRI + Gen). The first method classified the 3 most prevalent subtypes with 82% accuracy, while the second one raised the accuracy to 89% and identified all five subtypes. They included important data about clinical findings related to the different sCJD subtypes and showed in a clear way the accuracy of PriSCA_MRI and PriSCA_MRI + Gen diagnosis compared to anatomopathological one. 

Clinical relevance: early sCJD subtype identification gives important prognostic information and may help to identify particular subsets of patients eligible for future clinical trials.