Xenotransplantation of human acute myeloid leukemia (AML) in immunocompromised animals has been critical for the definition of leukemic stem cells. However, existing immunodeficient strains such as NOD/SCID and NOD/SCID/b2mnull have short life spans, age dependent leakiness of humoral immunity and low levels of AML cell engraftment making long-term evaluation of primary human AML biology difficult. A recent study suggested that the nonobese diabetic/severe combined immondeficient/IL2Rgnull (NOG) mouse has enhanced ability to engraft AML cells but this study relied on neonatal injections that are technically challenging. We performed an extensive analysis of AML engraftment in adult NOG mice using intravenous tail vein injection. Thirty-six different AML samples were analyzed including 2 samples of acute promyelocytic leukemia (APML). We used a threshold for AML engraftment of >0.5% human CD45+33+ cells in the murine bone marrow. Based on this threshold, 22 samples (61%) showed engraftment in NOG mice. Of these samples, 14 (64%) showed high levels of engraftment (greater than 10% of murine marrow replaced with human CD45+CD33+ cells). Engraftment did not correlate with FAB subtype or cytogenetic abnormalities to a statistically significant degree, however we noted that one sample with an 11q23 translocation and several samples with Flt3 ITD mutations showed consistent high level engraftment. Several samples demonstrated engraftment as high as 95% of the murine marrow with total AML cell expansion of 2-30 fold. Evaluation of AML stem cell frequency and expansion is ongoing. Engraftment in spleen was variable and in general significantly lower than in bone marrow. For most samples, peripheral blood engraftment was barely detectable. In contrast to NOD/SCID mice, both APML samples engrafted well in the NOG mouse with high levels of peripheral blood involvement. Some samples occasionally showed engraftment of a population of cells expressing CD2 and other T cell associated markers by flow cytometry, however this observation was inconsistent even between mice injected with the same sample. All samples tested (n=5) showed consistent engraftment in secondary and tertiary recipients with most samples tested showing further expansion of total AML cells in subsequent transplants. Importantly, a number of animals developed organomegaly and a wasting illness consistent with advanced leukemic disease. Several such animals showed extramedullary leukemic infiltration into non-hematopoietic tissues. Etoposide monotherapy (40 mg/kg in divided doses) of heavily engrafted mice did not induce a significant response in terms of leukemia regression. Studies of other chemotherapeutic agents are ongoing. We conclude that the NOG xenotransplantation model is a robust model for studying human AML cell engraftment which will allow for better characterization of AML biology and testing of new therapies

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author