MR neurography of lumbosacral nerve roots for differentiating chronic inflammatory demyelinating polyneuropathy from acquired axonal polyneuropathies: a cross-sectional study

BACKGROUND: Magnetic resonance (MR) neurography is an imaging technique focused on the peripheral nerves. Its role in the diagnosis and differential diagnosis of chronic inflammatory demyelinating polyneuropathy (CIDP) has yet to be investigated. This study explored the value of MR neurography in identifying CIDP and differentiating it from acquired axonal polyneuropathies. METHODS: In this study, 20 patients with CIDP, 10 patients with acquired axonal polyneuropathies, and 20 healthy controls were prospectively enrolled. Three-dimensional T2-weighted image fat-suppressed and diffusion tensor imaging sequences of the lumbosacral plexus were completed in all participants. The cross-sectional area (CSA) and diffusion parameters, including the fractional anisotropy (FA) and apparent diffusion coefficient (ADC) of the L3 to S1 nerve roots, were measured and compared across the 3 groups using Kruskal-Wallis 1-way analysis of variance. Receiver operating characteristic (ROC) curves were plotted to determine the value of CSA and diffusion parameters in the diagnosis and differential diagnosis of CIDP. RESULTS: CSA and ADC increased in CIDP patients but didn’t differ between patients with axonal polyneuropathies and healthy controls [CAS: 45.35±23.889, 22.25±3.878, 22.81±4.079 mm(2), ADC: (1.64±0.269)×10(−3), (1.37±0.204)×10(−3) and (1.39±0.156)×10(−3) mm(2)/s, in CIDP, axonal polyneuropathies and healthy controls, respectively, both P<0.001]. Compared with healthy controls, FA reduced in patients with CIDP and axonal polyneuropathies but no difference was observed in the two groups (FA: 0.24±0.053, 0.27±0.014 and 0.32±0.045, in CIDP, axonal polyneuropathies and healthy controls, respectively, P<0.001). To identify CIDP, ROC analysis showed that FA had better efficiency with cut-off value of 0.278 and sensitivity and specificity of 85% and 90% respectively. To differentiate CIDP from axonal polyneuropathies, CSA had better diagnostic accuracy with cut-off value of 29.46 mm(2 )and sensitivity and specificity of 75% and 100% respectively. CONCLUSIONS: CSA and ADC values of lumbosacral nerve roots can help to identify patients with CIDP and further distinguish them from patients with axonal polyneuropathies. FA decreased in both types of polyneuropathies and may thus have limited value in the discrimination of the 2 types of neuropathies.