Please use this identifier to cite or link to this item:
Title: Validation of 1H-NMR spectroscopy based-metabolomics as a tool to detect lung cancer via a simple blood sample
Authors: LOUIS, Evelyne 
MESOTTEN, Liesbet 
Thomeer, Michiel 
Vanhove, Karolien 
Vandeurzen, Kurt
Sadowska, Anna
Wauters, Els
Dooms, Christophe
Vansteenkiste, Johan
Reekmans, Gunter 
Adriaensens, Peter 
Issue Date: 2013
Source: Markers in Cancer, Brussels - Belgium, 7 - 9 November
Abstract: Background: Lung cancer is the leading cause of cancer death worldwide. Until today no effective methods permit the early detection of lung cancer. Therefore, detection methods with an improved specificity and sensitivity are urgently needed. Over the past decade, accumulating evidence has shown that the metabolism of cancer cells differs from that of normal cells. Metabolites are the end products of cellular metabolism and disturbances in biochemical pathways which occur during the development of cancer consequently provoke changes in the metabolic phenotype. Recently, our research group has build a statistical classifier (i.e. a metabolic phenotype) using multivariate orthogonal partial least squares-discriminant analysis (OPLS-DA). After removing the outliers (original dataset: 78 lung cancer patients and 78 control subjects), this classifier allows to discriminate between lung cancer patients and control subjects with a sensitivity of 83% (62 out of 75) and a specificity of 96% (71 out of 74). The present study aims to validate these promising results in an independent study population of 80 patients with anatomopathologically confirmed lung cancer (before any treatment) and 80 control subjects. Material and methods: Fasting venous blood samples are collected and analyzed by 1H-NMR spectroscopy. Subsequently, the constructed classifier is used to predict this independent group of lung cancer patients and control subjects. Results: By using the constructed classifier, 64 out of 80 (80%) lung cancer patients and 55 out of 80 (69%) control subjects are correctly classified. Conclusions: The constructed classifier allows to classify the majority of the lung cancer patients and control subjects correctly. Once we have collected sufficient samples to validate this method, we want to investigate at random whether it allows to detect lung cancer in a population with a low prevalence, actually whether it can be used as a valid screening tool.
Document URI:
Category: C2
Type: Conference Material
Appears in Collections:Research publications

Files in This Item:
File Description SizeFormat 
Poster-EvelyneLouis-MarkersInCancer.pdf572.99 kBAdobe PDFView/Open
Show full item record

Page view(s)

checked on May 18, 2022


checked on May 18, 2022

Google ScholarTM


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