Therapy Biosimulation Using the Cellworks Computational Omics Biology Model (CBM) Is Predictive of Individual Acute Myeloid Leukemia (AML) Patient Probability of Clinical Response (CR) and Overall Survival (OS): Mycare-023

Background: Although some genomic biomarkers have been integrated into therapeutic decision-making for the management of AML, the complete remission and cure rates have significant margin for improvement. Except for a few targeted therapies, genomic assessments offer limited guidance on treatment. Nevertheless, comprehensive molecular profiling of AML discloses a complex and heterogeneous disease network that impacts the efficacy of individual chemotherapeutics differently in individual patients. The Cellworks Computational Omics Biology Model (CBM) was developed using artificial intelligence heuristics and literature sourced from PubMed to generate a patient-specific protein network map. The Cellworks Biosimulation Platform uses the CBM to model each patient's unique cancer and predict personalized responses to standard AML drugs, identify novel drug combinations for treatment-refractory patients and optimize treatment selection to improve outcomes. Methods: A prospectively designed study involving observational data from 416 de novo AML patients was used to test the hypothesis that biosimulation using the Cellworks Biosimulation Platform predicts clinical response to individual drugs and estimates likelihood of response and survival better than physician prescribed treatment (PPT) alone. Cytogenetic and molecular data obtained from clinical trials including AMLSG 07-04, Beat AML, TCGA and PubMed publications was used to create personalized in silico models of each patient's AML and generate a Singula™ biosimulation report with a Therapy Response Index (TRI) to determine the efficacy of specific chemotherapeutic agents. The impact of specific AML agents on each patient's disease network was biosimulated to determine a treatment efficacy score by estimating the effect of chemotherapy on the cell growth score, a composite of cell proliferation, viability, apoptosis, metastasis, DNA damage and other cancer hallmarks. The mechanism of action of each drug was mapped to each patient's genome and biological consequences determined response. Multivariate logistic regression models for clinical response and likelihood ratio tests were used to assess the contribution of the Cellworks Biosimulation Platform beyond PPT. Similarly, multivariate Cox proportional hazards models were used to test the hypothesis that the Cellworks Biosimulation Platform is predictive of overall survival (OS) and provides predictive information beyond PPT alone. Scoring quantifies the bene