Targeted Metabolomics Approach Identifies Alterations in the Plasma Metabolome of Multiple Myeloma Patients with and without Extramedullary Spread

Introduction: Metabolomics refers to the identification and quantitation of metabolites in cells, tissues and biofluids. The metabolome (complete set of metabolites in a biological sample) reflects the biochemical events occurring in an organism at a given time, thus providing a valuable source to analyse metabolic changes in a variety of diseases, including cancer. Metabolic profiling of blood cancers represents a useful tool for the detection of novel biomarkers and therapeutic targets. Multiple myeloma (MM) accounts for less than 2% of all new cancer cases in the United States. The rare and aggressive sub-entity of MM, extramedullary multiple myeloma (EMM), develops when malignant plasma cells escape the bone marrow microenvironment and colonize distal tissues or organs. The prognosis of EMM is poor, and the incidence of EMM increases at disease progression, occurring in up to 30% of relapsed MM patients. The pathogenic mechanisms of EMM are poorly understood with no targeted therapies currently available to strategize treatment regimens. Here, we use a targeted metabolomics approach to identify metabolic changes in the plasma of MM patients with and without extramedullary spread.

Methods: Targeted metabolomic analysis of age and gender-matched medullary MM (n=8) and EMM (n=9) blood plasma samples was performed using the MxP® Quant 500 kit (Biocrates Life Sciences AG, Innsbruck, Austria) with a SCIEX QTRAP 6500plus mass spectrometer. The MxP® Quant 500 kit is capable of quantifying more than 600 metabolites from 26 compound classes. Quality control (QC) samples were employed to monitor the performance of the analysis with metabolite concentration in each sample normalised based on these QC samples. Isotopically labelled internal standards and seven-point calibration curves were used in the quantitation of amino acids and biogenic amines. Semi-quantitative analysis of other metabolites was performed using internal standards. Data quality was evaluated by checking the accuracy and reproducibility of QC samples. Metabolites were included only when the concentrations of the metabolites were above the limit of detection (LOD) in >75% of plasma samples. Data was imported into MetaboAnalyst 5.0 for further analysis. Feature filtering was performed based on relative standard deviation (RSD) and the resulting data was autoscaled. Metabolites of interest were identified based on p-value < 0.01 and fold-change > 1.5 between experimental groups. Supervised statistical approaches were used to further interrogate the data.

Results: Using a targeted metabolomic technique, we compared the metabolic profile of MM and EMM patient plasma. Univariate analysis using a t-test identified 5 metabolites of interest; HexCer(d18:1/20:0), TG(16:0_34:2), TG(22:4_32:0), TG(18:2_32:0), and Taurine (Figure 1(A)). The supervised clustering technique orthogonal projection to latent structure discriminant analysis (OPLS-DA) was used to determine separation between the two groups (MM and EMM). OPLS-DA scores plot illustrated a distinct separation between MM patients with extramedullary spread (red dots) compared to those without extramedullary spread (green dots) (Figure 1(B)). A permutation test (n=1000) was performed to ensure there was no overfitting of the data. Permutation analysis results (Q2 = 0.478, p = 0.023; R2Y = 0.978, p = 0.033) demonstrated the model was of good predictive quality. Discriminatory variables responsible for the group separation were identified using the OPLS-DA variable importance in projection (VIP) score to identify metabolites with a score greater than 1.5. HexCer(d18:1/20:0), TG(16:0_34:2), TG(22:4_32:0), TG(18:2_32:0), and Taurine had VIP scores of 2.5, 2.3, 2.1, 2.2 and 2.4, respectively. The diagnostic potential of these metabolites as EMM biomarkers was evaluated by receiver operating characteristic (ROC) curve analysis. All 5 metabolites demonstrated high diagnostic potential with area under the curve (AUC) values greater than 0.84.

Conclusion: Investigating disease-associated metabolomes presents an opportunity to identify dysregulated metabolic processes and novel biomarkers. This pilot targeted metabolomic analysis of EMM plasma samples reveals metabolites of interest for further analysis and contributes to our understanding of EMM pathophysiology.

No relevant conflicts of interest to declare.