Abstract
Backgrounds: Transthyretin familial amyloid polyneuropathy (TTR-FAP) is frequently misdiagnosed as chronic inflammatory demyelinating polyneuropathy (CIDP) because of similar phenotypes in the two diseases. This study was intended to identify the role of nerve ultrasonography in evaluating TTR-FAP and CIDP. Methods: Eighteen patients with TTR-FAP, 13 patients with CIDP, and 14 healthy controls (HC) were enrolled in this study. Consecutive ultrasonography scanning was performed in six pairs of nerves of bilateral limbs with 30 sites. The cross-sectional areas (CSAs) and CSA variability data of different groups were calculated and compared. Results: Both TTR-FAP and CIDP showed larger CSAs at most sites of both upper and lower limbs than in HC groups. CIDP patients had larger CSAs than TTR-FAP patients at 8/15 of these sites, especially at U1-3, Sci2 sites (p < 0.01). However, the CSAs at above sites were not a credible index to differentiate TTR-FAP from CIDP with a low area under the curve (<0.8). The CSA variability of median nerves was significantly higher in CIDP than in TTR-FAP and HC groups, with high sensitivity (0.692) and specificity (0.833) to differentiate CIDP from TTR-FAP. The CSA variability of ulnar nerves was not significantly different between the three groups. For the TTR-FAP group, mean CSAs at each site were not correlated with different Coutinho stages, modified polyneuropathy disability, course of sensory motor peripheral neuropathy, Neuropathy Impairment Score, or Norfolk Quality of life-diabetic neuropathy score. The mean compound muscle action potential of ulnar nerves was negatively correlated with the mean CSAs of ulnar nerves. Interpretation: TTR-FAP patients had milder nerve enlargement with less variability in CSAs of median nerves than those with CIDP, suggesting that nerve ultrasound can be a potential useful auxiliary tool to help differentiate the two neuropathies.