Supplementary material

Table e-1. Metabolomic signature in patients with multiple sclerosis identified in the time-series analysis of the prospective cohort. Table shows the untargeted UPLC ID peaks (indicating the metabolite when identification was achieved) significantly different in the serum of patients of the prospective cohort (n=61) compared to healthy controls, along the 9 time points (months 0-3-6-9-12-15-18-21-24). Analysis was done with the ANOVA for repeated measurements and smoothing splines mixed (SME) analysis. Table shows the mass/charge relationship (m/z) and retention time (Rt) for each peak.

ID / formula / m/z / Rt / SME
stat / p
ID090 / NA / 882.5505 / 6.6651 / 7.149774 / 0.000237
ID096 / NA / 367.212 / 2.8831 / 8.911585 / 0.002754
ID014 / NA / 197.0439 / 2.6454 / 8.614677 / 0.002512
ID016 / C39H77N2O6P
(Sphingomyelin) / 745.548 / 6.6304 / 6.095418 / 0.001115
ID093 / NA / 526.3485 / 4.207 / 21.92018 / 0.003249
ID365 / NA / 813.5387 / 6.6276 / 8.460216 / 0.000148
ID411 / NA / 335.2199 / 3.5495 / 11.82004 / 0.009905
ID440 / NA / 846.5871 / 6.9589 / 12.56641 / 0.005358
ID309 / NA / 482.3603 / 4.3891 / 27.09837 / 0.008106
ID312 / C21H44NO6P
(Lysophosphatidyl-ethanolamine) / 438.299 / 4.4296 / 9.389545 / 0.008292
ID514 / NA / 798.5648 / 6.3488 / 7.595514 / 0.006185
ID542 / NA / 378.7813 / 6.603 / 4.632788 / 0.003865
ID553 / NA / 464.2797 / 6.6408 / 5.964487 / 6.21E-05
ID558 / NA / 858.5863 / 6.6566 / 8.785001 / 0.007056
ID575 / NA / 790.5618 / 6.7823 / 25.23018 / 0.007505
ID618 / NA / 826.6029 / 7.3137 / 12.11069 / 0.006357
ID190 / NA / 351.2171 / 3.0278 / 8.212357 / 0.001625
ID224 / NA / 359.2206 / 3.5686 / 9.90853 / 0.004377
ID232 / NA / 319.2275 / 3.7275 / 10.82425 / 0.008962
ID487 / NA / 808.5851 / 6.2084 / 7.517106 / 0.007041
ID541 / NA / 774.5682 / 6.5995 / 11.58308 / 0.006104
ID531 / NA / 866.5915 / 6.5143 / 5.04847 / 0.001775
ID569 / NA / 358.7929 / 6.7481 / 4.755392 / 0.000202
ID590 / NA / 351.2813 / 6.8715 / 6.372994 / 0.000559
ID591 / NA / 371.799 / 6.8718 / 6.187009 / 0.001406
ID593 / NA / 342.28 / 6.8722 / 5.493876 / 0.001543
ID595 / NA / 614.3844 / 6.9279 / 7.585249 / 0.008158
ID599 / NA / 860.6137 / 6.9515 / 5.30034 / 0.002205
ID615 / NA / 840.5728 / 7.1502 / 7.905559 / 0.008768

Table e-2. Serum metabolomic profiling (see excel file Table e-2)

Descriptive and statistical data obtained for the complete metabolomics analysis in serum samples. Raw: metabolite class, metabolite subclass, metabolite identification detail [“Individual notation” refers to the confirmed identification of the metabolites. Overlapping of two or more metabolites or non-confirmed identification is indicated in “Individual composition (or probable ID)”].

Increase of the EDSS (ΔEDSS) in the 2 year follow-up statistical analysis: Mean and standard deviation (SD) per group (patients with ΔEDSS <3, between 3.5 and 4.5, and >4.5), ANOVA analysis and Tukey’s HSD post hoc test.

Relapse-free statistical analysis: Mean and standard deviation (SD) per group (relapse and stable), fold-change (relapse/stable) and Student t-test relapse vs. stable.

Figure e-1

Figure e-1. Metabolomic pipeline. We made use of an targeted pipeline applying three separate UPLC-MS based platforms, each of them optimal for the profiling of different metabolites: Platform 1) MeOH extract: fatty acyls, bile acids, steroids and lysoglycerophospholipids; Platform 2) Derivatised MeOH extract: amino acids and amino acids derivatives; Platform 3) CHCl3/MeOH extract: glycerolipids, cholesterol esters, sphingolipids, and glycerophospholipids. Data pre-processing included peak identification and integration. Data normalization was performed with internal and external controls.

Figure e-2

Figure e-2. Time-series analysis of changes in serum metabolites in patients with multiple sclerosis and controls from the prospective cohortover 24 months. Significant p value < 0.01 using rm ANOVA and smoothing splines mixed (SME) analysis for MS patients vs healthy controls for two significant ID peaks: ID016 (sphingomyelin) and ID312 (phosphatidylethanolamine). Changes along time were associated with the use of DMDs in the case of phosphatidylethanolamine (p=0.01) but not for sphingomyelin (p=0.18). Left column - mean intensities over time, right column - approximated splines. Red color: MS patients, green color: healthy controls.