Candidate imaging biomarkers for PMP22-related inherited neuropathies

Objective: Charcot–Marie–Tooth type 1A (CMT1A) and hereditary neuropathy with liability to pressure palsy (HNPP) are caused by mutations to the peripheral myelin protein 22 (PMP22) gene. A need exists for sensitive and reliable biomarkers of progression and treatment response. Magnetic resonance imaging (MRI) metrics of nerve pathology and morphology were investigated for this purpose. Methods: MRI was performed at 3.0 T in the thigh of CMT1A (N = 11) and HNPP patients (N = 12) and controls (N = 23). Three potential imaging biomarkers of the sciatic nerve were investigated: 1) magnetization transfer ratio (MTR), which assays myelin content, and 2) cross-sectional area (CSA) and 3) circularity, which assay morphological changes. Potential imaging biomarkers were compared across cohorts and assessed for relationships with disability in the legs (CMTES_L), compound motor action potentials (CMAP), and motor conduction velocities (MCV). Inter-rater reliability and test–retest repeatability were established for each imaging metric. Results: Significant differences in MTR, CSA, and circularity were observed in CMT1A relative to controls (p = 0.02, p < 0.001, and p = 0.003, respectively, via Wilcoxon rank-sum tests). Differences were not observed in the HNPP cohort. Significant relationships were observed between MTR and clinical metrics (CMTES_L: p = 0.003, CMAP: p = 0.03, MCV: p = 0.01); and between CSA and electrophysiology (CMAP: p = 0.002, MCV: p < 0.001). All metrics were reliable and repeatable with MTR the most reliable (intraclass correlation coefficient [ICC] >0.999, CV = 0.30%) and repeatable (ICC = 0.84, CV = 3.16%). Interpretation: MTR, CSA, and circularity showed promise as reliable and sensitive biomarkers of CMT1A, but not HNPP. These warrant longitudinal investigation as response biomarkers in upcoming clinical trials of CMT1A, while other methods should be considered for HNPP.