Prognostic significance of drug-regulated genes in high-grade osteosarcoma

Abstract

About 25-45% of patients with high-grade osteosarcoma poorly respond to chemotherapy with an increased risk of relapse and the development of metastasis. Therefore, the aim of this study was the evaluation of the prognostic value of eight previously identified drug-regulated candidate genes on osteosarcoma therapy outcome. Gene expression of 8 candidate genes was analyzed in 35 formalin-fixed, paraffin-embedded, laser-microdissected osteosarcoma biopsies. The prognostic value of these genes was evaluated by the correlation of gene expression with therapy outcome, overall survival and event-free survival in univariate and multivariate analysis. Upon univariate analysis, the expression of MALAT-1, IMPDH2, FTL and RHOA significantly correlated with response to chemotherapy. Expression of all four genes was increased in the poor responder group. Upon multivariate analysis, IMPDH2 maintained its independent prognostic value (P=0.025). Concerning the overall survival of the patients, we observed a significant association with the expression of FTL, PHB, ATAD2, ACTN1 and RRM2 as well as lactate dehydrogenase serum levels. In the subgroups of patients with high expression of these genes and those with elevated lactate dehydrogenase levels, the mean overall survival was decreased 1.7-, 1.9-, 2.2-, 2.4-, 1.5- and 4.5-fold, respectively. Except RRM2, all genes and lactate dehydrogenase serum levels remained significant in the multivariate analysis. In addition, the event-free survival was significantly decreased in the subgroups of patients with high FTL, ATAD2 and IMPDH2 expression (1.8-, 6.3- and 2.4-fold, respectively). These data demonstrate that among the identified genes are valuable markers for the prediction of osteosarcoma therapy outcome. Especially IMPDH2 and FTL are promising candidates for the stratification of osteosarcoma patients into low- and high-risk groups. Owing to their involvement in drug action these genes may further be potential targets for the modulation of drug sensitivity.