Identification of a PAX-FKHR gene expression signature that defines molecular classes and determines the prognosis of alveolar rhabdomyosarcomas.

5635

Rhabdomyosarcoma (RMS) is a genetically and clinically heterogeneous family of soft tissue sarcomas that occur in children, adolescents and young adults. The more aggressive alveolar RMS (ARMS) tumors are known to specifically express the PAX3-FKHR or PAX7-FKHR (PAX-FKHR) fusion genes resulting from the t(2;13) or t(1;13) chromosomal translocations, respectively. However, up to 25% of ARMS tumors are fusion-negative, making it unclear whether ARMS represent a single disease or multiple clinical and biological entities with a common phenotype. To test to what extent PAX-FKHR determine class and behavior of ARMS and in order to better define subgroups within the ARMS tumor group, we used oligonucleotide microarray expression profiling on 139 primary RMS tumors and an in vitro model. We found that ARMS tumors expressing either PAX-FKHR gene share a common expression profile distinct from fusion-negative ARMS and from the other RMS variants. In addition, we observed that PAX-FKHR expression above a minimum level, as determined by quantitative RT-PCR, is necessary for the presence of this expression profile. Using transduced embryonal RMS cells as an ectopic PAX3-FKHR and PAX7-FKHR expression model we identified a gene expression signature regulated by PAX-FKHR, which is also detected specifically in PAX-FKHR positive ARMS tumors. Data mining for functional annotations of signature genes suggested a role for PAX-FKHR in regulating ARMS proliferation and differentiation. Cox regression modeling identified a subset of genes within this PAX-FKHR expression signature that segregated ARMS patients into good and poor prognosis subgroups with 5-year overall survival of 85% and 23%, respectively. These prognostic classes were independent of conventional clinical risk factors. Our results demonstrate that PAX-FKHR dictate a specific expression signature that helps define the molecular phenotype of PAX-FKHR positive ARMS tumors and, since it is linked with disease outcome in ARMS patients, determine tumor behavior. Our analysis of the largest reported dataset of rhabdomyosarcoma tumors shows that expression profiling, especially when supported by data from in vitro tumor models, can significantly improve classification. Furthermore, this approach identifies relevant molecules that may lead to targeted and therefore less toxic therapy options.