Abstract 3123: Modeling anti-leukemic therapy by patient derived AML xenografts with distinct phenotypes/geneotypes

Proceedings: AACR Annual Meeting 2014; April 5-9, 2014; San Diego, CA We have recently successfully engrafted leukemic cells from bone marrow of 3 AML patients, subtypes of M5 (AM7577), 2 (AM8096) and 6 (AM8070), into immunocompromised NOD/SCID mice. All three displayed aggressive AML diseases but with distinct characteristics. AM7577 displays typical M5 disease subtype, which starts at bone marrow and gradually expands to peripheral (spleen, lymph nodes and peripheral blood). AM7577 also harbors interesting genotype, including mutations of IDH2-R140Q, FLT3-ITD, DNMT3A R882H and NPM1. AM8096, derived from a recurrent AML-M2 patient sample, also displayed aggressive disease with 100% mortality, but many distinct pathological phenotypes: first, although with abundant leukemic cells in bone, the peripheral symptom is significant minor, characterized by lower leukemic counts in peripheral blood and only slightly enlarged spleen and smaller lymph nodes, as compared to AM7577; second, the leukemic cell morphology is also rather different where AM8096 demonstrates less differentiated phenotype; third, the genotype, in contrast to AM7577, is wild-types for all the above oncogenes. AM8070, on the extreme contrary, has severe bone marrow disease, but with few peripheral leukemia. While we have not identified the likely leukemogenesis driver gene for these models, we are currently performing RNAseq for all three models in order to explore the underlying molecular pathogenesis. In addition, we are investigating the model response to standard of care (SOC, AraC). We found that the treatment resulted in rapid control of the disease as shown by reduction (disappearance) of leukemic cells in peripheral blood, significantly relieved disease symptoms and extended life. However, when drug administration was stopped, the disease rapidly re-emerged. In order to investigate the underlying mechanism of the rapid disease relapse without obvious development of resistance, we examined the drug effect on leukemia at different relevant leukemic organs: bone, spleen, lymph nodes and blood. Surprisingly, we found that the drug has no effect on bone leukemic load, while completely suppressed leukemia in blood, with partial effect in spleen and lymph nodes. Our current working hypothesis is that the failure to suppress leukemia in bone, including LICs, are likely responsible for the observed rapid relapse of the disease immediately after drug withdrawal. In summary, the established PDX AML models could serve as useful experimental models to investigate the diverse leukemogenesis and to model anti-leukemia therapies. Citation Format: Jinping Liu, Xiaoyu An, Na Wang, Di Wang, Liang Huang, Ran Wu, Jie Cai, Jean-Pierre Wery, Henry Li. Modeling anti-leukemic therapy by patient derived AML xenografts with distinct phenotypes/geneotypes. [abstract]. In: Proceedings of the 105th Annual Meeting of the American Association for Cancer Research; 2014 Apr 5-9; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2014;74(19 Suppl):Abstract nr 3123. doi:10.1158/1538-7445.AM2014-3123