Multifocal fiber loss in proximal sciatic nerve in symmetric distal diabetic neuropathy

Magnetic resonance neurography detects diabetic neuropathy early and with Proximal Predominance

“…Furthermore, nerve fascicle imaging in FAP revealed an entirely diffuse pattern of proximal nerve fascicle signal increase [Fig 4 in Kollmer et al11], instead of the focal/multifocal pattern observed here in DPN. These differences between FAP and DPN observed in‐vivo by MR neurography are in close agreement with similar histopathological differences between DPN4, 5, 6, 13 and FAP. In FAP focal/multifocal patterns of fiber loss have never been reported.…”

“…These imaging findings differentiate the nerve lesion gradient from the longitudinal gradient of symptoms in DPN, which typically follow the opposite direction (strong distal focus). Interestingly, nerve lesion voxels formed clusters corresponding to nerve fascicles and exhibited focal/multifocal distributions at the thigh level (Fig 1) that showed striking similarity to the historical histopathological observations in this region 4, 6, 13. The extensive anatomical coverage allowed us to map the burden of nerve lesion voxels from far proximal to ankle levels.…”

“…Through very extensive and meticulous histological analyses, a few studies have reported results for the distal levels of the sural, peroneal, and tibial nerves and, importantly, also for defined proximal levels, including several serial sections of spinal nerves, the plexus, and the sciatic nerve at the thigh level 4, 6, 13. In these studies, Dyck et al,4 Sugimura and Dyck,13 and Johnson et al6 consistently showed that fiber loss at more proximal levels (eg, in the sciatic nerve at the thigh level) occurs not diffusely, but instead within focal and multifocal zones that appear “punched out” from the sciatic nerve fascicles, increasing in severity at more distal levels. This focal/multifocal lesion pattern on sciatic nerve cross‐sections at the thigh level closely resembled the focal/multifocal fiber loss observed after experimentally induced ischemic insults, for example, by microsphere embolization into vessels supplying the sciatic nerve18 or in human necrotizing vasculitic neuropathy 19.…”

“…Because biopsies of sciatic, peroneal, and tibial nerve trunks at the thigh level are not possible for scientific purposes, any interpretation of our imaging data cannot be based on direct correlation with histology, but instead must rely on (1) historical histological specimens of distal and proximal nerve segments,4, 5, 6, 13 as noted above, and (2) the following physical considerations regarding how the MR signal is influenced by microstructural sources on a quantitative level.…”

“…Because of the considerable time needed for the acquisition of the extensive imaging protocol for coverage of large anatomical areas and simultaneous high spatial resolution, signal quantification was not possible over the entire area covered and was possible only within this anatomically restricted imaging slab. We chose the thigh level for signal quantification because preliminary measurements in a few patients with DPN and the results from a pilot study12 indicated that the lesion burden would be highest at the thigh level and because historical histological data reported proximal nerve lesions at the thigh level 4, 13. Comprehensive statistical analyses of both quantitative measures were performed by ordered logistic regression on group .…”