Resumen de Heterogeneity of biomarker expression in non-small cell lung cancer /
The success of precision medicine in oncology is dependent to a large extent on an adequate selection of patients who will receive targeted therapies aimed at specific molecular traits of their tumor. In order to be able conduct such patient selection, predictive biomarkers that can inform therapeutic decisions are essential.
MET and PD-L1 are two relevant membrane receptors for non-small cell lung cancer (NSCLC) biology. MET is an oncogene the activation of which is involved in multiple pro-tumorigenic processes such as cell proliferation, motility and invasion. PD-L1 is a key molecule that acts during the immune response, and its overexpression in tumors is thought to mediate the ability of tumor cells to avoid immune cell recognition and destruction. Currently, there are specific therapies directed against these molecules. The most commonly used strategy to select the patients that will benefit from such drugs is the analysis of the expression of both molecules in tumor tissue. However, the value of MET and PD-L1 as predictive biomarkers and the method by which it should be determined is a subject of debate.
Recent studies have detected a high degree of genomic heterogeneity in NSCLC tumor samples. This heterogeneity could significantly affect biomarker-based patient classification especially in the case of NSCLC, since biomarker studies are usually performed in small biopsies or cytology samples obtained through minimally invasive techniques. The main objective of the work presented in this thesis is to study the heterogeneity of the expression of MET and PD-L1 in NSCLC samples.
For this purpose, we have analyzed tumor samples from NSCLC patients that had undergone surgical treatment at Hospital del Mar. Of each tumor, we have selected multiple geographically separate areas, which we analyzed independently. In the study evaluating MET, we selected four tumor areas per patient, while in the study evaluating PD-L1 we selected two areas. In each tumor area, we measured the expression MET and PD-L1 using immunohistochemical and fluorescence in situ hybridization methods (FISH). Finally, we compared the expression of MET and PD-L1 in different tumor areas.
Regarding MET, we have found discordances between different tumor areas in 20-40% of cases using immunohistochemistry and in 25-50% of cases using FISH. Regarding PD-L1, this discrepancy was greater if we evaluated PD-L1 expression in tumor infiltrating lymphocytes (17-27%) than if we did so only in tumor cells (10-19%). Moreover, 36% of the cases with amplification of the gene coding for PD-L1 determined by FISH presented gene amplification only in one of the two areas analyzed.
Overall, our results suggest that the expression of both biomarkers is heterogeneous, whether measured by immunohistochemistry or by FISH. This heterogeneity can have a potential impact on the classification of tumors based on the expression of biomarkers and, therefore, could represent a hurdle for the development of targeted therapies for NSCLC patients.
