Abstract Purpose: Preclinical studies by our lab and others have demonstrated fibroblast growth factor receptor (FGFR) inhibition is a viable therapeutic strategy in tumors with FGFR amplification, such as gastric and breast cancer, as well as FGFR2 mutation positive endometrial cancer (EC). A significant clinical issue that we aim to address is that of intrinsic and acquired resistance to anti-FGFR therapeutics in the context of EC. Experimental Design: Our lab performed a BaF3 screen to identify FGFR2 kinase domain mutations that would provide resistance to dovitinib, a multikinase inhibitor with anti-FGFR activity. BaF3 cells transduced with FGFR2 were treated with TKI258 at 10x and 15x the IC50 for dovitinib in these cells. Following clonal selection of dovitinib-resistant cells, the exons encoding the intracellular domain of FGFR2 were sequenced and mutations identified. Independent BaF3 stable lines were generated with wild-type FGFR2 and each of the dovitinib resistance mutations. Using these cells, sensitivity of the different FGFR2 mutants to dovitinib, ponatinib, BGJ398, AZD4547, and DCC-2036 was measured. To explore nonkinase domain mechanisms of acquired resistance to FGFR inhibition, we have also passaged AN3CA and MFE280 cells (both FGFR2 mutant) in increasing concentrations of dovitinib or PD173074. In preliminary studies, we compared the gene expression profile at baseline and 14 hours after dovitinib/PD173074 treatment of one resistant clone for both AN3CA and MFE280 versus parental clones for each. Results: Mutations in the kinase domain of FGFR2 were found in 26/63 resistant clones treated with 10X and 15X IC50. Mutations observed in individual resistant clones included M536I, M538I, I548V, N550K, N550S, V565I, E566G, L618M, Y770lfsX14 and multiple clones were found to carry the N550H mutation in FGFR2. Modeling of these mutations on FGFR2 crystal structures indicate that the majority either abrogate the “molecular brake” in the kinase hinge region resulting in receptor activation, or strengthen the hydrophobic spine, a series of interactions involved in stabilizing the activated conformation of the kinase. All but one mutation provided resistance to both dovitinib and PD173074 when analyzed in an independent BaF3 stable line. Together this data indicates that dovitinib and PD173074 predominantly bind the inactive conformation of the FGFR kinases. This is clinically significant as it suggests that cancer patients harboring activating FGFR2 mutations, such as N550K, may not respond to the anti-FGFR activity of dovitinib as a first line therapy. Significantly, ponatinib inhibited the growth of BaF3 cells transduced with all resistance mutations, except the gatekeeper mutation V565I. From our gene expression profiling of isogenic resistant EC cell lines, only a small number of significant gene expression changes were noted between resistant clones and their respective parental lines at baseline. With drug treatment, however, we observed a significant activation of pro-survival pathways, most notably BCL-XL and the WNT signaling pathway. Conclusions: In this study we identified FGFR2 mutations that confer resistance to dovitinib and show that these mutations increase FGFR2 ligand dependent and independent activity. Importantly, we found that ponatinib has strong inhibitory activity against FGFR2 activating mutations suggesting that this inhibitor may be more effective as a first-line therapy, as well as in the second line setting to target tumors with resistance to dovitinib. Moreover, it suggests that the active state of the FGFR2 kinase should be targeted for anticancer drug discovery.