Recapitulation of the biologic differences between blastsnaive and blastsresist in NOD/SCID mice. (A) When transplanted at equal cell doses, blastsresist exhibited superior engraftment than blastsnaive, which was significant at 9 and 12 weeks after transplantation. The engrafting cells were all CD33+, consistent with the immunophenotype expected of AML. (B) Limiting dilution assays were performed, showing that the LIC frequencies in blastsresist were 12.0-, 12.2-, and 13.3-fold higher than those of blastsnaive in AML1, AML3, and AML10, respectively. (C) NOD/SCID mice transplanted with blastsresist exhibited inferior survivals than those of blastsnaive. All mice that died after 6 weeks after transplantation were confirmed to have > 90% human engraftment in the BM. Twenty-eight mice were transplanted with blastsnaive and another 28 with blastsresist. All mice were killed at 12 weeks after transplantation. (D) Two of 28 mice transplanted with blastsresist developed myeloid sarcoma in the body and the eye, with tumor cells being CD45+/CD33+/CD19−. (E) Sorafenib significantly reduced leukemia engraftment by blastsnaive but not blastsresist in NOD/SCID mice. Each point represented the result from one mouse (5 different experiments from paired samples from 3 patients AML4, 7, 8). (F-G) Sorafenib but not vehicle (DMSO in carboxymethylcellulose) treatment induced differentiation of engrafting cells, as indicated immunophenotypically (F) and morphologically (G). Nikon Eclipse E800M; eyepiece 10×; objective 40×; numerical aperture of objective lens 0.95; camera Nikon DS cooled camera head DS-5MC and DC camera control unit DS-L2; magnification of picture 400×; software for imaging processing Adobe Photoshop CS5 Extended Version 12.0.4x32. (H) Sorafenib had no effect on CD11b expression in mice engrafted with blastsresist. (I) PCR for FLT3-ITD in marrow cells from mice engrafted with blastsnaive. In all 3 cases, despite a reduced level of engraftment as shown in panel B after sorafenib treatment, the predominant FLT3-ITD clone persisted, which was also confirmed by quantitative analysis (supplemental Figure 4). Patient AML14 was not included in the clinical trial and has not been treated with sorafenib.