Hypomethylating agents (HMAs) (azacitidine (aza), decitabine (dec)) and lenalidomide (len) are approved agents and used to treat patients with myelodysplastic syndromes (MDS) or acute myeloid leukemia (AML). Despite their widespread use, HMAs fail in the majority of these patients, and len fails in 75% of non-del(5q) MDS. Currently, no method exists to predict disease response, thus the management of MDS and AML patients is challenging.

Methods: Patients with AML or MDS were recruited to a clinical trial (NCT02435550) designed to assess predictive values by comparing computer predictions of drug response to actual clinical response. Genomic profiling was conducted by cytogenetics, whole exome sequencing, and array CGH. Genomic results were inputted into a computational software (Cellworks), which generates disease-specific protein network maps using PubMed and other resources. Digital drug simulations were conducted by quantitatively measuring drug effect on a cell growth score (proliferation + viability + apoptosis). Each patient-specific protein network was screened for the extent by which aza, dec or len reduced disease growth in a dose-respondent manner. Treatment was physician’s choice of SOC. Clinical outcomes were prospectively recorded. IWG criteria were used to define response. Western blot assays were performed to validate the predicted protein network perturbations. Fisher’s exact test was used to compare prediction values of the genomics-informed computer method versus empiric drug administration.

Results: 88 patients have had all molecular tests and computational modeling performed. Lab validation of computer-predicted, activated protein networks in 19 samples from 13 different patients showed correct prediction of 5 activated networks (Akt2, Akt3, PIK3CA, p38, Erk1/2) in 17 samples, with 89% accuracy. At the time of this report, 26/88 patients were eligible for efficacy evaluation. 8/26 patients showed clinical response to SOC therapy, 18/26 did not. 24/26 outcome predictions were correctly matched to their clinical outcomes, and 2/20 were incorrectly matched, resulting in 92% prediction accuracy, 80% PPV, 100% NPV, 100% sensitivity, and 89% specificity. The accuracy of the genomics-informed computer method was significantly greater than empiric drug administration (p=1.664e-05). New genomic signature rules were discovered to correlate with clinical response after aza, dec or len.

Summary: A computational method that models multiple genomic abnormalities simultaneously showed high predictive value of protein network aberrations and clinical outcomes after SOC treatments. The network method uncovered molecular reasons for drug failure and highlighted resistance pathways that could be targeted to recover chemosensitivity. This technology could also be used to establish eligibility criteria for precision enrollment in drug development trials

Citation Format: Leylah Drusbosky, Kimberly E. Hawkins, Shireen Vali, Taher Abbasi, Ansu Kumar, Neeraj Kumar Singh, Kabya Basu, Chandan Kumar, Amjad Husain, Caitlin Tucker, Randy A. Brown, Maxim Norkin, John Hiemenz, Jack Hsu, John Wingard, Christopher R. Cogle. iCare 1: A prospective clinical trial to predict treatment response based on mutanome-informed computational biology in patients with AML and MDS [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr CT085. doi:10.1158/1538-7445.AM2017-CT085