Introduction: AML patients with relapsed/refractory (R/R) disease have few effective treatment options. LEN+AZA may be an active and better tolerated regimen compared with conventional chemotherapy. This combination has been tested in a phase 2 pilot study of LEN+AZA in 37 R/R older patients with a 49% overall response rate (4 complete remission (CR) / CR with incomplete recovery (CRi) and 14 with morphologic leukemia free state (MLFS). The objective of this study was to utilize the genomics of these patients to retrospectively predict response to AZA+LEN treatment using CBM workflow and identify key genomic characteristics of the responder profiles.
Methods: We analyzed the clinical and genomic (NGS WES, cytogenetics) data from pre-treatment sample for a cohort of R/R AML patients. (Blood 2017 130:1337) All available genomic data was input into the CBM predictive workflow (Cellworks Group). Customized, individualized AML patient models - patient specific disease network maps of activated and inactivated dysregulated pathways were created.
Digital drug models of AZA and LEN were created for CBM by programming their mechanism of action and effects on specific protein targets and pathways determined from published literature. The digital drug models were simulated individually and in combination on every patient's disease at varying doses. Drug impact was assessed by quantitatively measuring a disease inhibition score (DIS), the degree to which cancer phenotypes (proliferation, viability, apoptosis) were repressed, along with impact on the identified patient specific biomarkers. CBM predictions were compared with clinical outcomes in a retrospective blinded manner. Clinical response was defined as partial response or better as per International Working Group 2006 criteria.
Results: 32 of 37 patients were analyzed by CBM based on availability of genomic profiling. 24 of the 32 patients were clinically evaluated, of which 4 had CR or CRi and 9 had MLFS response. CBM predictions matched for 20 of 24 profiles. The statistics of prediction accuracy against clinical outcome are presented in Table 1.
A common aberration associating with sensitivity discovered in 7 profiles was a synonymous WT1 SNP rs16754. It was found in 1 CR and 5 MLFS cases and in the 1 non-responder case that along with WT1 SNP also had mutations in EZH2 and NOTCH1 that caused drug resistance. This WT1 SNP rs16754 correlated with higher mRNA expression of Wt1 and was associated with significantly improved outcome in pAML (PMID: 21189390) WT1 directly transactivates DNMT3A expression, thus improving response to hypomethylating agent. (PMID: 23042785)
Another surprising find was the LOF mutation in BCORL1 (BCL6 Corepressor Like 1) L622P, that was the only common aberration found in all the 4 CR profiles. BCORL1 is a transcriptional corepressor that can also bind and regulate class II HDACs. LOF mutations in BCORL1 have been found in AML patients with abnormal karyotypes (PMID:22210327) In aplastic anemia, clones with BCORL1 mutations correlated with better response to immunosuppressive therapy and higher OS (PMID: 26132940), but no such association has been established in AML and may be worth exploring further in larger datasets.
Conclusion: The CBM prediction of AZA+LEN in R/R AML patients based on genomic abnormalities and clinical data showed high predictive accuracy of 83.33%. The study validates the approach to a priori predict response and identify the right therapy option for the patient and could be used to establish criteria for precision enrollment in drug development trials. In addition, analysis uncovered novel predictors of response to AZA+LEN therapy that include presence of WT1 SNP rs16754 and possibly BCORL1 LOF, that were found in the responder profiles and linked to enhancing the drug efficacy network. This clinical decision support technology and these biomarkers may improve patient selection for AZA+LEN treatment in R/R AML.
Abbasi:Cell Works Group Inc.: Employment. Singh:Cellworks Research India Private Limited: Employment. Ullal:Cellworks Research India Private Limited: Employment. Kumari:Cellworks Research India Private Limited: Employment. Tyagi:Cellworks Research India Private Limited: Employment. Alam:Cellworks Research India Private Limited: Employment. Lunkad:Cellworks Research India Private Limited: Employment. Cogle:Celgene: Other: Steering Committee Member of Connect MDS/AML Registry. Vali:Cell Works Group Inc.: Employment. Pollyea:AbbVie: Consultancy, Research Funding; Celgene: Membership on an entity's Board of Directors or advisory committees; Karyopharm: Membership on an entity's Board of Directors or advisory committees; Curis: Membership on an entity's Board of Directors or advisory committees; Gilead: Consultancy; Pfizer: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding; Celyad: Consultancy, Membership on an entity's Board of Directors or advisory committees; Argenx: Consultancy, Membership on an entity's Board of Directors or advisory committees; Agios: Consultancy, Membership on an entity's Board of Directors or advisory committees, Research Funding.
Asterisk with author names denotes non-ASH members.