Diffusion tensor imaging MR Neurography detects polyneuropathy in type 2 diabetes

Abstract

Aim: To evaluate if diffusion-tensor-imaging MR-Neurography (DTI-MRN) can detect lesions of peripheral nerves due to polyneuropathy in patients with type 2 diabetes.

Methods: Ten patients with type 2 diabetes with polyneuropathy (DPN), 10 patients with type 2 diabetes without polyneuropathy (nDPN) as well as 20 healthy controls (HC) were included. DTI-MRN covered proximal (sciatic nerve) and distal regions (tibial nerve) of the lower extremity. Fractional-anisotropy (FA) and diffusivity (mean (MD), axial (AD) and radial (RD)) were calculated and compared to neuropathy severity. Conventional T2-relaxation-time and proton-spin-density data were obtained from a multi-echo SE sequence. Furthermore, we evaluated sensitivity and specificity of DTI-MRN from receiver operating characteristics (ROC).

Results: The proximal and distal FA was lowest in patients with DPN compared with nDPN and HC (p < 0.01). Likewise, proximal and distal RD was highest in patients with DPN (p < 0.01). MD and AD were also significantly different though less pronounced. ROC curve analyses of DTI separated nDPN and DPN with area-under-the-curve values ranging from 0.65 to 0.98. T2-relaxation-time and proton-spin-density could not differentiate between nDPN and DPN.

Conclusion: DTI-MRN accurately detects DPN by lower nerve FA and higher RD. These alterations are likely to reflect both proximal and distal nerve fiber pathology in patients with type 2 diabetes.

Keywords: Diabetes; Diffusion tensor imaging; Magnetic resonance neurography; Neuropathy; Receiver operating characteristics; Type 2 diabetes.