Marieke van Rosmalen

MRI of the intraspinal nerve roots in chronic inflammatory neuropathies 85 5 Table 5.5 Mean intraspinal nerve root sizes per clinical phenotype and controls Ventral Dorsal MD 95% CI p 1 OVERALL 0.64 (0.14) 0.67 (0.16) 0.03 (SE 0.01) 0.02 – 0.04 < 0.001 Sensorimotor 0.64 (0.13) 0.67 (0.15) 0.02 (SE 0.01) 0.01 – 0.04 0.018 Motor 0.67 (0.16) 0.69 (0.18) 0.02 (SE 0.01) -0.04 – 0.11 0.248 Sensory/ataxic 0.68 (0.13) 0.73 (0.17) 0.53 (SE 0.02) 0.01 – 0.09 0.011 MND 0.63 (0.11) 0.67 (0.14) 0.04 (SE 0.01) 0.02 – 0.06 <0.001 Healthy 0.53 (0.12) 0.56 (0.12) 0.03 (SE 0.02) -0.01 – 0.07 0.149 p 2 < 0.001 a,b,c < 0.001 a,d,e Nerve root sizes are mean in millimeters with standard deviation. Mean differences (MD) are shown with standard error (SE) and 95% confidence intervals (CI). p 1 = p values concerning ventral vs. dorsal roots p 2 = p values concerning ventral roots between phenotypes, and dorsal roots between phenotypes. a p < 0.001 for healthy vs. sensorimotor, motor, and sensory or ataxic phenotype, and MND b p = 0.018 for motor vs. sensorimotor phenotype c p = 0.002 for motor phenotype vs. MND d p = 0.001 for sensory or ataxic phenotype vs. sensorimotor e p = 0.006 for sensory or ataxic phenotype vs. MND Abbreviations: MND = motor neuron disease. DISCUSSION This study shows that MRI of the intraspinal nerve roots shows changes in patients with an inflammatory neuropathy and that the location of these changes corresponds with the nature of neurological deficits. Ventral intraspinal nerve roots were thicker in patients with a pure motor phenotype (i.e. MMN and pure motor CIDP) compared to patients with a sensorimotor phenotype and patients with MND. Dorsal intraspinal nerve roots were thicker in patients with a pure sensory or ataxic phenotype (i.e. sensory CIDP). These findings for the first time show that anatomical abnormalities correspond with clinical deficits in inflammatory neuropathies. Previously used imaging techniques for chronic inflammatory neuropathies focused on the peripheral nerves and nerve roots in arms or legs outside the vertebral foramen. 2,3,11 Since these nerves almost always have a mixed composition (i.e. motor and sensory), it was impossible to detect preferential involvement of motor or sensory nerves. Nerve conduction studies often show normal sensory conduction in pure motor neuropathies and vice versa, but assessing involvement of the most proximal parts is challenging. F-waves could provide information on the most proximal parts of the peripheral nervous system, but this information is unspecific and does not localize the exact site of injury. Previous NCS studies did not show a clear relation between nerve function and nerve