Genomics of AML Prognosis

1. ABSTRACT: Acute Myeloid Leukemia (AML) is a heterogeneous disease with a dismal outcome; fewer than 20% of elderly patients and only 50-65% of pediatric patients are cured and survive more than 3 years following diagnosis. Despite this, the standard therapy for AML treatment has relied primarily on an intensive combination of ara- C, daunorubicin, and etoposide (ADE) for over 40 years. Relapsed and resistant disease following treatment with standard therapy (ADE: ara-C, daunorubicin, and etoposide) are the most common clinical failures that occur in treating this disease. Application co-PIs, Drs. Lamba (pharmacology) and Pounds (biostatistician specializing in cancer genomics) have successfully collaborated for over a decade to develop methods and discover molecular prognostic factors for AML. During our recently completed second funding cycle, we were very productive with 13 scientific publications, 53 conference presentations, and two pending patents. Our second-cycle scientific achievements include the development of an ara-C SNP score that predicts leukemic cell intracellular levels of ara- CTP, the active form of ara-C the development of the innovative integrative analysis procedure; canonical correlation with projection onto the most interesting statistical evidence (CC-PROMISE), that dramatically increases statistical power for meaningful biological discovery in a rare-disease small sample size setting; using CC- PROMISE to discover that reduced methylation and increased expression of the DNMT3B associates with greater genome-wide methylation burden and worse prognosis; translating the DNMT3B discovery into evaluation of demethylating agents in the ongoing AML16 clinical trial; the development and initial validation of the six-gene pediatric leukemia stem cell (pLSC6) score (patent pending) and five-gene ADE resistance score (ADE-RS5) score that predict prognosis. In this renewal application, we propose to accelerate our exciting progress by extensively validating the pLSC6 and ADE-RS5 scores to provide a robust scientific foundation to translate them into clinical practice and further developing our understanding of the biological basis of AML development and prognosis into other molecular domains. In the current proposal we seek to accelerate our exciting progress by extensively validating the pLSC6 and ADE-RS5 scores in ~ 4000 patients across 10 independent national and international AML cohorts treated on intensive chemotherapy, to provide a robust scientific foundation for its translation into clinical practice. In aim 2, we propose to perform a clinical outcome-GWAS (CO-GWAS) and establish prognostic genes and variants with a constellation of genomic, epigenomic, and transcriptomic features that associate with clinical outcomes which will undergo thorough mechanistic validation in aim 2. The successful completion of these studies will be scientifically and clinically significant by preparing a sound scientific rationale to incorporate prognostic gene expression scores into the risk stratification of AML patients and revealing additional layers of the molecular basis of AML prognosis to guide the development of more effective therapies.