Diagnostic and Prognostic Sarcoma Signatures

U01 CA114757

Timothy J. Triche, M.D., Ph.D.
Children’s Hospital Los Angeles

Dr. Triche’s program focused on refining and validating molecular signatures to provide a more accurate diagnosis of the common childhood sarcomas, and signatures that more accurately predict clinical behavior of these tumors. The project built on signatures developed during the Director’s Challenge program.

Specifically, the program aimed to investigate the following: the accuracy of molecularly defined diagnostic classes versus traditional histopathologic classes of rhabdomyosarcoma; inclusion and exclusion criteria for entry on rhabdomyosarcoma protocols based on myogenic gene expression; distinction of treatment-resistant versus metastatic profiles in osteosarcoma; the role of genomic features, particularly fusion gene type and expression level, P53 mutation, and p16 loss, on expression profile and outcome in Ewing’s sarcoma; and gene clusters that accurately identify existing and new molecularly defined classes of non-myogenic soft tissue sarcomas.

The program also evaluated the relative accuracy and potential superiority of ‘gene’ expression analysis at the exon level as opposed to whole-transcript analysis, in order to detect and evaluate the potential role of splice variants and other RNAs as independent predictors of class and outcome. This approach was chosen for its potential to allow use of scant amounts of tissue, as often encountered clinically, as well as the possible use of formalin fixed, paraffin embedded tissue, available from all patients, which would in turn facilitate translation of these ‘sarcoma signatures’ to clinical practice.

Following refinement, these signatures were integrated with standard diagnostic and prognostic criteria to create more accurate predictors for these tumors. The predictors were then prospectively validated in the uniformly treated patient populations available from the Children’s Oncology Group (COG), which enrolls virtually all the childhood sarcoma cases in the North America. This program also defined profiles that predict response to specific therapies and that identify potential new therapeutic targets, with the goal of incorporating these signatures into the standard of care for sarcoma patients treated on COG clinical therapeutic trials.

Collaborators:

• The project included investigators from Children’s Hospital Los Angeles, Baylor College of Medicine, Children’s Memorial Hospital Chicago, Northwestern University, University of Southern California, the Children’s Oncology Group (COG) and the National Childhood Cancer Foundation (NCCF).
• Statistical, analytical and bioinformatics expertise was provided by the COG Data and Statistical Center and the individual collaborators. COG provides specimens and data from clinical trials.

Projects:

• Evaluate and validate diagnostic and prognostic signatures in rhabdomyosarcomas and in non-rhabdomyosarcoma soft-tissue sarcomas.
• Evaluate and validate prognostic signatures in osteosarcomas.
• Evaluate and validate prognostic signatures in Ewing’s sarcomas.
• Develop signatures that identify diagnostic class and predict response to therapy in all types of sarcoma.

Featured Publications:

Davicioni E et al (2009) Molecular classification of rhabdomyosarcoma genotypic and phenotypic determinants of diagnosis: a report from the Children’s Oncology Group. Am J Pathol 174: 550-564. PMID: 19147825

A gene expression signature determined by the Triche SPECS project to distinguish the alveolar and embryonal subtypes of rhabdomyosarcoma out-performs histology at predicting patient outcome.

Abdueva, D., M. Wing, et al. (2010) Quantitative expression profiling in formalin-fixed paraffin-embedded samples by affymetrix microarrays. J Mol Diagn 12(4): 409-417. PMCID: 2893624

Davicioni, E., J. R. Anderson, et al. (2010) Gene expression profiling for survival prediction in pediatric rhabdomyosarcomas: a report from the children’s oncology group. J Clin Oncol 28(7): 1240-1246. PMCID: 3040045