MR-Neurography of the sural nerve in patients with hereditary amyloidosis

Background

Sural nerve biopsies are often performed in order to detect the underlying disease in patients suffering from unclear polyneuropathic symptoms. In transthyretin familial amyloid-polyneuropathy (TTR-FAP) the diagnostic value of invasive sural nerve biopsies is controversially discussed as it often lacks to detect amyloid deposits [Simmons et al, J Neurol Sci 1993]. As we recently reported, amyloid related nerve-injury in TTR-FAP can be unambiguously determined in large-caliber nerves (sciatic, tibial and common peroneal nerve) by applying high-resolution MR-Neurography (MRN) [Kollmer et al, Brain 2015]. However, the diagnostic yield of MRN of the small-caliber sural nerve, representing the target nerve specimen for biopsies, is still unclear and was subject to this investigation.

Methods

We prospectively enrolled 25 patients with manifest TTR-FAP, 10 asymptomatic gene-carriers with confirmed mutations in the TTR-gene, and 40 age/gender-matched healthy volunteers. Besides detailed neurological and electrophysiological examinations in all patients, a sural nerve biopsy was obtained in 12/25 manifest TTR-FAP patients. All participants underwent the following high-resolution MRN protocol (3Tesla/Magnetom/TIM-TRIO/Siemens):1) axial 2D-T2-TSE-fs (TR/TE 5970/55ms, voxel-size 0.4x0.3x3.5mm³); 2) axial 2D-dual-echo-TSE-fs (TR 5210ms, TE1/TE2 12/73ms, voxel-size 0.4x0.3x3.5 mm³).

On each axial imaging slice the sural nerve was identified and manually segmented. After signal-normalization (histogram-based, comparison with control population), nerve-voxels were statistically classified as nerve-lesion-voxels by operator-independent, threshold-based segmentation. The apparent-T2-relaxation-time and proton-spin-density were calculated for all nerve-lesion-voxels.

Results

Sural nerve lesion-voxels were found to be significantly higher in manifest TTR-FAP vs. controls (p<0.0001), in asymptomatic gene-carriers vs. controls (p<0.0001) and in manifest TTR-FAP vs. asymptomatic carriers (p=0.0035). Wilcoxon rank-sum-test revealed with high statistical significance that proton-spin-density was higher in severely affected TTR-FAP patients (p<0.0001), in moderate TTR-FAP (p<0.0001) and also in asymptomatic gene-carriers (p=0.0003) compared to healthy controls. The apparent-T2-relaxation-time was significantly increased in symptomatic TTR-FAP (p<0.05) but not in asymptomatic gene-carriers (p=0.4286) compared to controls.

Conclusion

MRN of the sural nerve is a new, non-invasive and highly sensitive diagnostic tool, which can clearly differentiate between symptomatic TTR-FAP, asymptomatic gene-carrier status and healthy controls by evaluating nerve-lesion-voxels and proton-spin-density. Additional analyzes of the apparent-T2-relaxation-time can further confirm symptomatic disease. Results of this evaluation may have a strong impact for a better diagnostic interpretation of negative sural nerve biopsies.