[en] Lung cancer is the most prevalent cancer in term of mortality in developed countries. This is due to the quietness of symptoms at early stage. Thus, the majority of patients are only diagnosed at an advanced stage, resulting in a poor prognosis. As the curability of lung cancer is highly dependent on early diagnosis, there is, therefore, an urgent need to develop earlier diagnostic screening tests allowing detection of lung cancer at a more curable stage.
Human exhaled breath contains several hundreds of volatile organic compounds (VOCs) that can be seen as a fingerprint that could possibly be used to differentiate between individuals exhibiting various health statuses. Breath analysis has been shown to be usable to highlight possible markers of specific diseases in these individuals. Such an approach is particularly adapted to potential early diagnosis of cancer because its low level of invasiveness and relative eases of implementation on a large scale basis. The implementation of an early diagnostic procedure for cancer screening by means of breath analysis could thus contribute to increase the survival rate of diagnosed patients.
Comprehensive two-dimensional gas chromatography coupled to time-of-flight mass spectrometry (GC×GC-TOFMS) has been reported to be able to isolate more than a thousand VOCs from one single human breath. Such an approach is, however, still far from clinical use as it still has to go through full clinical validation. Additionally, the routine use of GC×GC-TOFMS in hospitals for cancer screening is probably not the way to go, as other more simple techniques can be implemented to screen for markers of illness (e.g. e-noses, selected ion flow-tube mass spectrometry, ...). Nevertheless, GC×GC-(HR)TOFMS is a key step in extracting a list of reliable markers from the complex mixture made by breath VOCs prior clinical use. We are using GC×GC-(HR)TOFMS for the analysis of exhaled air samples taken from lung cancer patients, as well as from the headspace of cancer cell cultures. Solid-phase micro extraction (SPME) and thermal desorption (TD) are used for sampling. Several data mining approaches and statistical tools (Fisher ratio, random forest, principal component analysis, clustering...) have been implemented to digest the large amount of data generated. A short list of potential markers has been extracted among the large number of features detected initially to bring breath analysis closer to clinical use.
