Early detection of lung cancer: role of biomarkers.

Early detection of lung cancer requires none or few invasive techniques. Distal lung cancer (40% of the cases in most European countries) can be sensitively detected by spiral computed tomography. Theoretically, in 60% of cases, the proximal lesions (main to segmental bronchi, accessible by bronchoscopy) should be able to be detected by sputum cytology. Unfortunately, this very specific technique has a low sensitivity and is time consuming. Fluorescent bronchoscopy increases the detection rate of early or micro-invasive lesions and may be proposed in highly selected populations, but not as a screening test. Biomarkers in blood and sputum have not yet been clinically validated. However, the amount of data generated from studies first on resected tumours, then on early bronchial lesions and more recently on blood and sputum offer a wide field for investigation. Lung carcinogenesis is a multistep process characterised by the accumulation of successive molecular genetic and epigenetic abnormalities, resulting in selection of clonal cells with uncontrolled growth capacities throughout the whole respiratory tract (field cancerisation). Molecular lesions far precede morphological transformation of preneoplastic bronchial lesions (dysplasia) or alveolar lesions (atypical alveolar hyperplasia). Genetic and epigenetic abnormalities in the genes involved in cell cycle, senescence, apoptosis, repair, differentiation and cell migration control may be detected on bronchial biopsies, on respiratory cells from the sputum and even in the circulating deoxyribonucleic acid (DNA). The key genes involved include those in the P53-retinoblastoma (Rb) pathways. The balance between cyclin-dependent kinases and their inhibitors regulates the level of Rb phosphorylation and its function at G1-S transition; P53 plays at least two functions (cell cycle and apoptosis control). The balance of bax-bcl2 is important in the control of apoptosis as well as loss of fragile histidine triad expression. O(6)-methylguanine-DNA methyltransferase seems to be important in DNA repair control, the RARbeta receptor in differentiation, and cadherin H and E and different metalloproteases genes in cell migration. The demonstration of hyperexpression or silencing of these genes needs different validated techniques: immunohistochemistry on biopsies or cytological preparations, molecular biology techniques for mutations, loss of heterozygosity and aberrant methylation abnormalities. Automation and miniaturisation of these techniques will allow early detection and may be widely applied once clinically validated.