14800-DvRappard

114 Chapter 7 Scatterplots for age-corrected concentrations of metabolites that were different between patient outcome groups (i.e. NAA, Cr, Glx and Lac) are shown in Figure 3. This figure illustrates that patients with poor outcome were completely separated from moderate and good outcome based on NAA and also on Cr. Vice versa, patients with good outcome were completely separated from moderate and poor outcome based on NAA and also on Glx. Despite prominent differences in mean Lac between groups, Lac could not uniquely distinguish the groups. Scatterplots for MRI score and lesion volume (Figure 3E, F) show that these parameters had less power to distinguish good and moderate outcome. Figure 3. Scatterplots per clinical outcome. A-D: age-correctedWMmetabolite concentrations. E: MRI score. F: lesion volume. Lines indicate mean and standard deviation. The natural course of the disease inevitably results in poor outcome after a longer follow- up. To achieve insight into the pace of disease progression, we also evaluated outcome for non-transplanted patients two years after diagnosis, which altered outcome for only one patient (good 2 years after diagnosis, moderate at end of follow-up). In this case, the moderate outcome group contained only 2 patients, and was marginally separated from the good outcome group. Spearman’s rank correlations betweenmetabolite concentrations at baseline andmotor performance at latest follow-up showed a strong correlation between MLD-GMF and NAA (r=-0.78, p<0.001, Figure 4), as well as with age-corrected NAA (r=-0.75, p=0.001).