Diagnostic accuracy of MRI and ultrasound in chronic immune-mediated neuropathies.

Abstract

To assess and compare the diagnostic performance of qualitative and (semi-)quantitative MRI and ultrasound for distinguishing chronic inflammatory demyelinating polyneuropathy (CIDP) and multifocal motor neuropathy (MMN) from segmental spinal muscular atrophy (sSMA). Patients with CIDP (n = 13), MMN (n = 10), or sSMA (n = 12) and healthy volunteers (n = 30) were included. MRI of the brachial plexus, using short tau inversion recovery (STIR), nerve-specific T2-weighted (magnetic resonance neurography [MRN]), and diffusion tensor imaging (DTI) sequences, was evaluated. Furthermore, with ultrasound, cross-sectional areas of the nerves were evaluated. Three radiologists blinded for diagnosis qualitatively scored hypertrophy and increased signal intensity (STIR and MRN), and intraobserver and interobserver agreement was assessed. For the (semi-)quantitative modalities, group differences and receiver operator characteristics were calculated. Hypertrophy and increased signal intensity were found in all groups including healthy controls. Intraobserver and interobserver agreements varied considerably (intraclass correlation coefficients 0.00-0.811 and 0.101-0.491, respectively). DTI showed significant differences (p < 0.05) among CIDP, MMN, sSMA, and controls for fractional anisotropy, axial diffusivity, and radial diffusivity in the brachial plexus. Ultrasound showed significant differences in cross-sectional area (p < 0.05) among CIDP, MMN, and sSMA in upper arm and brachial plexus. For distinguishing immune-mediated neuropathies (CIDP and MMN) from sSMA, ultrasound yielded the highest area under the curve (0.870). Qualitative assessment of hypertrophy and signal hyperintensity on STIR or MRN is of limited value. DTI measures may discriminate among CIDP, MMN, and sSMA. Currently, ultrasound may be the most appropriate diagnostic imaging aid in the clinical setting.