The diagnostic value of nerve ultrasound in inflammatory neuropathies

Abstract

Distinction of chronic inflammatory neuropathies (CIN) from the more common axonal neuropathies or motor neuron disease is important since treatment can improve outcome. Nerve conduction studies (NCS) play a key role in diagnosing CIN. However, NCS are not a flawless technique that allows identification of all patients who would respond to treatment. New and better diagnostic tools for identification of CIN are therefore required and nerve ultrasound (NU) is a promising candidate, because of its relatively low-cost, time-efficient imaging of multiple nerves and the lack of burden for patients. In this thesis we examined the diagnostic and prognostic value of NU and the natural history of multifocal motor neuropathy (MMN). Moreover, we examined a new treatment in MMN. In chapter 2 we investigated interobserver variability of NU and found variability between investigators primarily in leg nerves and C6 and C7 nerve roots. The multilevel mixed model showed that different devices and different hospitals had no influence on interobserver variability. These findings indicate that NU of arm nerves and the C5 root can be reliably used in clinical practice. In chapter 3 we examined the sonographic pattern in Wartenberg’s migrant sensory neuritis and found mild multifocal enlargement in clinically affected and non-affected nerves at both entrapment sites and non-entrapment sites. In chapter 4 we compared NCS and NU results in CIN and found no correlation between NCS features of demyelination and sonographic nerve enlargement. This suggests that NU and NCS detect distinctive pathophysiological mechanisms. In chapter 5 we visualized the entire tract of both the median and ulnar nerve to complement previous findings at specific nerve sites. The distinct sonographic patterns between the different forms of CIN seem to be of additional diagnostic value and support the hypothesis of different disease entities. In chapter 6-8 we examined the diagnostic value of NU in CIN. In our studies, compared to NCS, sensitivity of NU was higher and specificity much lower. Therefore, these investigations are complementary rather than comparable techniques. The added value of NU in the detection of treatment-responsive CIN was 25%. In chapter 10 we performed a multicenter study on the prognostic value of NU and found that the distribution and development over time of nerve enlargement in CIN was very heterogeneous and thus the prognostic value of NU seems limited. In chapter 9 we performed a longitudinal study in MMN and found that almost all clinical outcome measures significantly deteriorated. This confirms that MMN is a progressive disorder. In chapter 11 we investigated Human Immune Globulin 10% with Recombinant Hyaluronidase, a new subcutaneous immunoglobulin therapy, and found comparable safety, efficacy and treatment satisfaction to intravenous immunoglobulin therapy. In conclusion, we found that NU could be reliably implemented in clinical practice and that addition of NU to routine diagnostic work-up improves identification of patients who may benefit from treatment by 25%. Therefore, our results indicate that NU deserves a prominent place in future revisions of diagnostic consensus criteria and that this promising diagnostic tool should be implemented in general neurologic practices.