A Pharmaco-Pheno-Multiomic Integration Analysis of Primary Acute Myeloid Leukemia (AML) Reveals Biomarkers of Cytarabine Resistance

The overall survival of patients diagnosed with Acute Myeloid Leukemia (AML) remains low.  While initial responses to therapy are favorable, the duration of response is short and overcoming therapeutic resistance has proven difficult.  A better understanding of the tumor cell biology and resistance mechanisms may shed light onto novel therapeutic targets that improve long-term outcome.  In this study, we performed an exhaustive analysis to include deep tumor phenotyping, drug sensitivity profiling, and comprehensive-omic characterization.  These datasets were included in integrative pharmaco-phenotypic-multiomic (PPMO) integration analyses to identify targets and biomarkers associated with cellular phenotype and drug response.  Our results reveal that the major cellular discriminant within the cellular phenotype is CD34 expression, which associates with a high PDK-mediated metabolic profile and cytarabine sensitivity.  Tumors exhibiting cytarabine resistance associate with a CD34-negative cellular phenotype and molecular characteristics such as MYC copy number gain, and increased expression of SAMDH1, METTL7B and SERPINB2. Prospective application of PPMO integration analyses was able to predict response profiles for 6/7 additional AML samples, therefore suggesting a potential application as a predictive diagnostic tool.