Summary Acute myeloid leukemia (AML) is one of the most intensely studied of human malignancies. In recent years there have been major advances in defining the molecular pathogenesis of AML and introduction of new therapies. But the three-year survival rate remains below 50% indicating failures to make therapeutic advances for the disease. In part, we believe that this reflects a failure to adequately use genetically defined PDX models to develop new therapies targeted at molecularly defined subsets of the disease. As discussed below, although the overall genetics of AML is complex, we and others have found that there are dominant oncogenes in AML and that targeting of these oncogenes can induce therapeutic responses although not cures. Examples of this approach include the use of FLT3 inhibitors for de novo and relapsed FLT3 ITD mutant AML and development of IDH1 and IDH2 inhibitors. In each of these cases, current therapy includes a choice of targeted therapies which are active but not curative. This likely reflects the multi-variate molecular pathogenesis. To advance the field, we have collected thousands of independent AML collections and characterized dozens of PDX models of AML. These models have been widely used for studies of basic biology but here we propose to advance the use of AML PDX modeling for therapy development. There are several obstacles to progress. First, there are not widely available and well characterized AML PDX models defined by genetic alterations. To address this concern, we have chosen to separate AML into 7 subsets defined by dominant oncogenes as recently done by the NIH MyeloMatch study. Leukemia’s will be sub-classified based on mutations in FLT3, DNMT3A, NPM1c, IDH1, IDH2, TP53 or KMT2A fusion protein (FP). Here we will characterize xenotransplantation of three AML models for each of these seven sub-groups and specifically characterize the biology of serial transplant of AML in the NSG mouse strain. Initial documentation of engraftment has already been completed for 33 of the 35 samples. In SA2, we will focus on performance of an XP2 study to direct ongoing clinical efforts. Two of these sub-groups are defined by the presence of fusion proteins involving KMT2A (previously mixed lineage leukemia or MLL) or Nucleophosmin (NPM) mutations (that cause cytosolic re-localization of the protein) (NPMc). Both of these sub-types of AML require menin, an epigenetic co-regulator for pathogenesis and AML’s with either class of mutations responding in pre-clinical models to Menin inhibition. Menin inhibitors are currently in Phase 1 and Phase 2 human studies and early results suggest that these drugs, like other targeted therapies in AML, are active but not curative. Recent results have suggested that enhanced differentiation of AML samples containing KMT2A fusion proteins (FP) can be achieved by combining a Menin inhibitor with a KAT6A inhibitor. Here we will use our characterized KMT2A FP and NPMc mutant AML PDX models to ...