Document Server@UHasselt >
Research >
Research publications >

Please use this identifier to cite or link to this item: http://hdl.handle.net/1942/16103

Title: Metabolic phenotyping of blood plasma by 1H-NMR spectroscopy to detect lung cancer?
Authors: LOUIS, Evelyne
THOMEER, Michiel
VANHOVE, Karolien
Vandeurzen, K.
Sadowska, A.
Issue Date: 2013
Citation: Young Belgian Magnetic Resonance Scientist, Blankenberge, 2-3/12/2013
Abstract: Introduction. Lung cancer is the leading cause of cancer death worldwide. An effective method which allows to detect lung cancer is urgently needed. Accumulating evidence shows that the metabolism of cancer cells differs from that of normal cells(1). Disturbances in biochemical pathways which occur during the development of cancer provoke changes in the metabolic phenotype(2). Objective. To determine the metabolic phenotype of lung cancer by means of proton nuclear magnetic resonance (1H-NMR) spectroscopy. Methods. Fasting venous blood samples of 77 patients with confirmed lung cancer (before any treatment) and 78 controls are collected and analyzed by 1H-NMR spectroscopy. The metabolic phenotype in blood plasma of lung cancer patients and controls is expressed by the integration values of 110 spectral regions, defined in the 1H-NMR spectrum. Multivariate orthogonal partial least squares discriminant analyses (OPLS-DA) are performed to investigate whether the metabolic composition of blood plasma allows to discriminate between lung cancer patients and controls. Robustness of the model constructed by OPLS-DA is evaluated by means of a receiver operating characteristic (ROC) curve. To identify statistically significant spectral regions, and so potential biomarkers, an S-plot (showing up- and downregulated spectral regions) was constructed, on the basis of their contribution to the OPLS-DA score plot. Results. OPLS-DA multivariate statistics shows that the metabolic composition of blood plasma discriminates between lung cancer patients and controls with a sensitivity of 85% and a specificity of 95%. Area under the curve (AUC) of ROC analysis is 0.95, indicating high predictive accuracy of the constructed model. Moreover, the S-plot shows that the spectral regions which contain signals of lipids, lactate, α-ketoglutarate, fructose, inositol and the amino acids arginine, asparagine, alanine and serine are key-players in the discrimination between lung cancer patients and controls. These metabolites can be defined as potential biomarkers and can help to unravel the disturbed biochemical pathways. Conclusion: Metabolic phenotyping of blood plasma by 1H-NMR spectroscopy detects lung cancer with a high degree of sensitivity and specificity. At present, a second cohort of lung cancer patients and controls is recruited to validate these promising results in a larger population study. 1. Cantor JR. and Sabatini DM. Cancer Discovery 2012, 2(10), 881-98. 2. Armitage EG and Barbas C. Journal of Pharmaceutical and Biomedical Analysis 2013, in press: http://dx.doi.org/10.1016/j.jpba.2013.08.041
URI: http://hdl.handle.net/1942/16103
Category: C2
Type: Conference Material
Appears in Collections: Research publications

Files in This Item:

Description SizeFormat
N/A1.22 MBUnknown

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.