Marieke van Rosmalen

Chapter 8 132 not MMN). 33–35 Two of these studies measured nerve root sizes in the ganglion in patients with CIDP (n = 14 in both studies) and healthy controls (n = 10 and n = 9 respectively). 33,34 They obtained a cut-off value of 5.0 mm for nerve root C6-C8 and 6.5 mm for C5-T1 respectively. The ganglion often appears larger on MRI and measurements in the ganglion may therefore may overestimate nerve root thickness. The latter study also measured nerve thickness at random, not predefined, sites distal from the ganglion (C5-T1), and reported a cut-off value of 4.0 mm with an area under the curve of 0.76. 34 We found a comparable area under the curve for measurements of nerve root size, although comparison is difficult as measurement sites were not standardized in the latter study. 34 Interrater reliability was higher (0.95) compared to our study but method and analysis of interrater reliability are not well-explained in the paper. For example, it is unknown if raters scored the scans at the same time or with a time interval and whether exact measurement sites were known prior to assessment. Also, details are missing on reliability per nerve root, as they calculated reliability for all nerve roots together which further complicates comparison with our data. The third study measured nerve root diameter in the ganglion and did not find any differences between patients with CIDP (n = 15) and disease controls, most of them with cervical spondylosis (n = 19). 35 A major limitation in all three studies is sample size which hampers interpretation and translation of the results in clinical practice. Also, the control groups are less relevant for clinical practice in all three studies. Our study in chapter 4 included a relatively large sample of patients with CIDP and MNN (n = 81). We included disease controls (i.e. patients with a clinical phenotype that mimics CIDP or MMN) instead of healthy controls as we tried to represent clinical practice as much as possible. This resulted in the inclusion of a homogeneous group of patients with MND with a lower motor neuron syndrome at onset of symptoms. As earlier described, differentiation with this patient group is of great importance as patients with MND have a much less favorable prognosis and have no symptomatic or curative options for treatment compared to patients with CIDP and MMN. 36 Patients with MND mimic patients with MMN, but also 11% of patients with a clinical suspicion of CIDP are initially diagnosed with MND. 37 The majority of patients with CIDP present with slowly progressive motor deficits (approximately 90%). 9 The homogeneity of our control group facilitates the translation to clinical practice, especially to neuromuscular centers that focus on diagnosis of MND. Inclusion of a more variated control group, e.g. inclusion of patients with axonal neuropathies, hereditary demyelinating polyneuropathies and paraproteinemic polyneuropathies may further improve the representativity of our results. However, the hereditary demyelinating polyneuropathies and paraproteinemic polyneuropathies are a heterogeneous group of neuropathies and could also be differentiated from CIDP and MMN through their characteristic clinical phenotypes and laboratory findings. MRI is normally not needed as part of the diagnostic work-up of these neuropathies. The prevalence of hereditary demyelinating polyneuropathies is approximately 16-30/100.000 citizens and diagnosis is based on characteristic clinical features as pes cavus and a symmetric distal motor neuropathy with pronounced atrophy of lower leg muscles, combined with a positive family history or confirmative genetic defect. 38–41 Diagnosis of paraproteinemic polyneuropathies relies on the