Abstract 5636: Gene expression signature discriminates sporadic from post-radiotherapy radiation-induced sarcomas

Abstract

Abstract The purpose of the study is to identify markers that could define a robust signature discriminating case by case radiation-induced tumors from their sporadic counterparts. The rare occurrence of human radiation-induced tumors increases the difficulty of their identification. Nevertheless, in case of sarcomas occurring in the irradiation field after radiotherapy for a primary neoplasm, stringent criteria allow the radiation-induced etiology to be established with a high level of confidence. However, the available series remains limited. Global transcriptome studies are particularly affected by this problem since the methods used for data analysis are generally efficient only for long series. In order to solve this problem, we have developped new methods of classification, based on transcriptome analysis, for the case by case tumor diagnosis. To find a signature we first select a set of candidate genes, using Expectation-Maximisation algorithm, on a learning set of tumors. Second, using a training set of tumors, we select the genes with the higher potential for classifying the tumors. Then, the robustness of the signature was tested by blindly classifying an independent set of tumors. Using a learning/training set of 12 radiation-induced (RI) and 12 sporadic (SP) sarcomas including angiosarcomas, leiomyosarcomas and osteosarcomas, we identified a minimal signature of 143 genes discriminating the sarcomas according to the etiology. A larger signature of 1100 genes, using less stringent criteria of classification was also established. The robustness of these signatures was tested by the blindly case by case classification of an independent set of 22 RI and 14 SP sarcomas of various histologies. After the histology code-break, 28/36 sarcomas were well-classified using the minimal signature, in 6 cases (1 RI and 5 SP) the etiology could not be established whereas 2 SP sarcomas were classified as RI. Using the large signature, the etiology was not determined in 4 cases (1 RI and 3 SP) and the 32 other sarcomas were well-classified. Several pathway were found differentially expressed according to the etiology. Notably, the NRF2-mediated oxidative stress response pathway, the ubiquitination-deubiquitination pathways and several genes coding subunits of the 26S proteasome and heat shock proteins underwent a general up regulation in RI sarcomas. These data suggest that RI sarcomas are submitted to oxidative stress that could generate the formation of mis-folded and oxidated proteins further degradated by the ubiquitination-deubiquitination-proteasome pathway. In this study, we identify a robust signature discriminating radiation-induced from sporadic sarcomas, opening the possibility to recognise, case by case, the implication of ionizing radiation in a tumor formation. Note: This abstract was not presented at the AACR 101st Annual Meeting 2010 because the presenter was unable to attend. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 101st Annual Meeting of the American Association for Cancer Research; 2010 Apr 17-21; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2010;70(8 Suppl):Abstract nr 5636.