Marieke van Rosmalen

Chapter 8 146 and fat fraction changed over time in patients with CIDP, but not in MMN. Unfortunately, we could not establish high correlations between these alterations and clinical data (i.e. age, disease duration, time since last treatment, sex, receiving maintenance treatment and treatment response). Despite the lack of clinical correlations, we established that the quantitative MRI parameters after one-year follow-up differed between patients with CIDPandMMN, which is similar to the results we found in chapter 6 .We concluded that the differences in longitudinal data between CIDP andMMN are probably based on nerve changes that are part of the natural history of CIDP, and for example cannot be attributed to treatment response as extensively explained in the discussion section of chapter 7 . Previous longitudinal quantitative MRI studies mainly focused on DTI of the brain and only one study correlated clinical outcome measures (i.e. a decline in executive function) to an increase of RD after radiation therapy in patients with a primary brain tumor (n = 22). 90 Only one study focused on the peripheral nervous system (i.e. sciatic nerve injury) but this study was performed in rats (n = 63) and only evaluated FA. 91 The authors reported a correlation between an increase of FA and nerve regeneration. FA is a summary measure of AD and RD and changes in FA, without knowledge of changes in AD or RD, are unspecific. Although our study is one of the first longitudinal studies of the peripheral nervous system, the results indicate that DTI, T2 mapping and fat fraction analyses of the brachial plexus are unlikely to serve as a biomarker to predict prognosis or to monitor treatment response. Future research into reliable and robust biomarkers for (early) treatment response and prognosis could focus on other tissues as target for imaging (e.g. muscle) or on other techniques than MRI, such as laboratory findings or electrophysiology. The use of other techniques is preferred as imaging of other tissues, such as the muscles or the peripheral nerves, has considerable limitations. Quantitative muscle MRI has been studied as a potential biomarker for neuromuscular diseases, such as Duchenne muscular dystrophy, Becker muscular dystrophy, limb-girdle muscular dystrophy and spinal muscular atrophy (SMA). 92–95 However, these studies are performed in diseases that affect the large muscles of the arms and legs. In the majority of patients with CIDP and MMN symptoms are most pronounced in the smaller distal muscles of the hands and feet. Smaller structures may be less suited for imaging studies. This might lead to lower data quality, more artifacts and loss of data and this compromise feasibility for use as a biomarker. Imaging of the peripheral nerves is an alternative but these relatively thin structures might also lead to artifacts and loss of data. 70 Furthermore, to image the peripheral nerves in the arms with MRI a patient has to be positioned with the arms above the head, which is an uncomfortable position, particularly when scan duration is long. Also other imaging modalities, such as nerve ultrasound, failed to show correlations between nerve size and clinical disease activity or treatment response in a recent longitudinal study. 96