Molecular Signatures for Outcome Prediction and Therapeutic Targeting in ALL

Cheryl L. Willman, MD
University of New Mexico, Albuquerque, NM

Dr. Willman’s program will expand on the work accomplished under SPECS I, in which microarray-based gene-expression signatures that are highly predictive of relapse-free survival in high-risk acute lymphoblastic leukemia (ALL) were developed. Newly-discovered mutations and potential therapeutic targets measured on multiple genomic platforms will now be incorporated into the design and testing of new molecular classification algorithms. These will be used in concert with the expression classifiers to further refine diagnosis and treatment determination. Despite significant recent improvements, overall outcomes in pediatric ALL remain poor, and the development of accurate predictive molecular risk classification schemes could have significant clinical impact applicable not only to childhood, but to adolescent-young adult (AYA) and adult ALL as well.

Collaborators:

• The project team includes investigators from the University of New Mexico, Nationwide Children’s Medical Center/Ohio State University, University of California San Francisco, Fred Hutchinson Cancer Research Center (FHCRC), and two NCI Cooperative Groups, the Childrens’ Oncology Group (COG) and SWOG.

• Data management and statistical support is provided by the UNM Center for Advanced Research Computing as well as by each of the collaborators and the Cooperative Groups.

• Specimens will be obtained primarily from three sources: the SWOG ALL bank at FHCRC, the AYA ALL Intergroup study C10403, and the COG Tissue Repository at Nationwide Children’s Medical Center.

Projects:

• Refine microarray-based gene expression classifiers to the best and final set of predictive genes and translate the quantitative measurement of these multi-analyte signatures to a robust clinical diagnostic platform.

• Confirm and validate the multi-analyte gene expression classifier in a new prospective cohort of 500 high-risk ALL cases already accrued to COG trials and assess its predictive power relative to newly discovered molecular abnormalities (IKZF1, JAK) and other known risk factors (such as assays to determine the functional status of signaling pathways in patients with kinase mutations to predict who will respond to inhibitor therapies).

• Prospectively test the predictive power of the new molecular signatures and risk classification algorithms in the next generation of COG high-risk ALL clinical trials that will accrue 3000 newly diagnosed children over 5 years.