Exploring platelet metabolomics and fatty acid profiles for ALS prognosis and diagnosis

Abstract

Amyotrophic lateral sclerosis (ALS) is a fatal neurodegenerative disease with heterogeneous clinical progression, reflecting distinct underlying pathological mechanisms. Early and accurate diagnosis and prognosis require reliable biomarkers to improve clinical management and therapeutic stratification. The present study explores the potential of platelet global metabolomics and fatty acid (FA) profiling as potential sources of diagnostic and prognostic biomarkers for ALS. We analysed platelets from 15 recently diagnosed ALS patients and 21 healthy controls (CTLs) using liquid chromatography-mass spectrometry (LC-MS) for metabolomics and gas chromatography-flame ionization detection (GC-FID) for FA profiling. ALS patients were classified as fast or slow progressors based on the median ALS Functional Rating Scale-Revised (ALSFRS-R) slope. While global metabolomic and FA profiles have shown limited potential for distinguishing ALS from CTL, preliminary molecular annotation based on mass and retention times disclosed specific metabolites with potential diagnostic value. Importantly, both global metabolomic and FA analyses demonstrated a marked capacity to differentiate fast progressors from slow progressors (receiver operating characteristic (ROC) curves of approximately 1), revealing distinct metabolic signatures associated with disease progression. Our findings demonstrate that platelet global metabolomics and FA profiling hold promise as prognostic biomarkers in ALS.

Fig. 1

Multivariate analyses of platelet metabolome profile for ALS and CTL comparison. (a) PCA and (b) hierarchal clustering analyses of the 25 top differential metabolites show a limited effect of global metabolomic data for disease status prediction. (c) ROC analysis of a potential PI(34:0) and (d) a potential LysoSM(d18:0) shows an AUC > 0.7 and a p value < 0.05 for ALS diagnostic classification.

Fig. 2

Multivariate analyses of fatty acid profile for ALS and CTL comparison. (a) PCA and (b) hierarchal clustering analyses of the 25 top differential fatty acid species, indices and enzymatic activity estimations do not result in accurate separation.

Fig. 3

Multivariate analyses of the platelet metabolome for slow (S) and fast (F) progressors. (a) PCA did not result in good group separation. However, (b) hierarchal clustering analyses of the 25 top differential analyses show a good separation of both populations. (c) Three metabolites correlated with the ALSFRS-R slope according to Pearson correlation analysis. (d) Potential DG(28:2), (e) PC(38:3), (d) P-Nme(42:7), and (f) unidentified metabolites with an AUC > 0.9 and a p value > 0.05 for ALS prognosis. *p value < 0.05, **p value < 0.01.

Fig. 4

Multivariate analyses of the platelet metabolome for slow (S) and fast (F) progressors. (a) Although PCA did not yield good separation, (b) hierarchal clustering analyses of the 25 top differential metabolites revealed good separation of both populations. (c) The n-6 pathway and f22:5n6 were positively correlated with the ALSFRS-R slope according to Pearson correlation analysis, whereas the f18:2n-6 pathway was negatively correlated. ROC analysis of the FA analysis revealed an AUC > 0.9 and p value < 0.05 for the following measurements: (d) Δ8 n-6, (e) the n-6 pathway, (f) f22:5n6 fatty acid, and (g) f20:2n6 fatty acid. * p value < 0.05, ** p value < 0.01.