Estudio de los perfiles metabolómicos en muestras de suero para el diagnóstico del cáncer de pulmón

Résumé

Cancer is one of the most important diseases in the world due to its incidence, prevalence and mortality. Amoung all types of tumors, lung cancer (LC) accounts for 15-20% of all neoplasias, so it has become a health problem of great magnitude. The number of cases of lung cancer and related deaths have been increased proportionally, mainly due to increased smoking rates, since according to data provided by the World Health Organization (WHO), smoking is the risk factor which causes by itself. Unfortunately, 75% of patients with this pathology are diagnosed at an advanced stage due to the absence of symptoms at the beginning of the disease, which worsens prognosis greatly. Therefore, prevention based on early diagnosis and reduction of risk factors that predispose to lung cancer are the most useful strategy to reduce mortality related with this disease today. Low-dose computerized tomography (LDCT) has become one of the most used screening methods to prevent and reduce the risk associated to this disease to a 20 %, which has open an good alternative to early diagnosis of this disorder. An alternative to the use of imaging techniques is the study of metabolic changes occurring in the organism on the oncological processes onset. This allows the identification of biomarkers associated to the early detection of oncological processes; nevertheless the results obtained today are not conclusive. In the present study it has been optimized and applied a metabolomics approach based in direct infusion of blood serum extracts into a triple quadrupole time of fight mass spectrometer (DI-ESI-QqQ-TOF-MS) provided with a electrospray source, using samples from diagnosed lung cancer patients (LC) and health ones (HS) in order to get samples classification into two clusters on the basis of their metabolomics profiles. In a second step, metabolites causing this discrimination were identified, which can be used as potential biomarkers of lung cancer. For this purpose, a prospective observational study, with the inclusion of 60 serum samples (30 patients with lung cancer and 30 healthy subjects) was performed, using the metabolic high resolution mass spectrometry techniques previously described, and subsequent discriminant analysis partial least squares (PLS- DA), in order to classify the groups to study and find the variables (metabolites) that determine this classification. Initial study presented biased results caused by the high rate of hemolysis in a number of samples, which represents a disturbing factor in the metabolic interpretation, especially in the ability of discrimination between groups, which decreases with increasing hemoglobin concentrations. Therefore, new samples with haemolytic index lower than 20 g/dl were analyzed, including 15 CP samples, 16 CS and 15 samples of a third group of patients with respiratory diseases different to lung cancer. The results showed a clear differentiation between the three groups, also getting the identification of the following metabolites: phospholipids, glutamine and threonine, with overexpressed levels in the group of CP; and L- ornithine, urea, lysophospholipids, and triacylglycerols, inhibited with respect to the CS group levels. All these metabolites are related to well established metabolic pathways in cancer pathology. In addition, the influence of tobacco burden on metabolic profiles of lung cancer patients has been assessed in comparison with healthy subjects. Differences between certain metabolites were ascertained especially associated to patients with high smoking load, suggesting the influence of smoking habit on the occurrence of these metabolic processes in CP. Finally, validity of metabolomic high resolution techniques for detection of metabolites associated to certain lung cancer processes was confirmed, which may provide potential biomarkers for early diagnosis of this disease.