A New In Vivo Platform for Studying the Growth of Hematopoietic Cancer Initiating Blood Cells and Response to Therapy.

Abstract 4142

There are few in-vivo models for studying human leukemia and its therapy. These include the high cost immune-deficient NOD/SCID mice and large mammalian fetuses, which both require weeks to assess treatment response. We described a rapid and low-cost alternative in-vivo system for human leukemia in the preimmune chick embryo (Taizi et al, Exp Hem 34;1698-708,2006). We recently demonstrated that the turkey embryo provides a more robust model for the preclinical assessment of human leukemia infiltration (Grinberg et al, Leukemia Res 33;1417-26 2009). Here we describe the application of this powerful and inexpensive model for rapid preclinical assessment of anti-cancer therapies and basic research of blood malignancies. BM engraftment was robust occurring in 95% of the embryos using cells lines and 40% using fresh samples. Leukemia cells homed to the BM where they were already detected at 20 hours after injection and reached highest levels on days E19-24, using FACS and RT-PCR. Serial engraftment in secondary recipients of all three human stem cell leukemia lines and one fresh sample was detected in embryos injected with BM harvested 8-10 days after the first inoculation, validating the engraftment of cancer initiating cells. Human stem cell leukemia lines K562 and LAMA, (both BCR/Abl+) and CHRF (c-Kit+), and myeloma cell lines ARH-77 and CAG and fresh patient samples were injected IV into turkeys, on embryonic day E 11, using 5×10⁶ cells from lines or 10⁷ fresh patient cells. Engraftment of human leukemia and myeloma cells (cell lines and fresh samples) was detected 8-14 days later (E19-24), in the BM and in several hematopoietic organs at a frequency of 0.5->20%, by real-time PCR, immunohistochemistry and flow cytometry. CAG and ARH-77 myeloma cells engraftment was also detected by the presence of human monoclonal free light chain (6-10 mg/L) in blood collected from vessels of the chorioallantoic membrane, one week after cell injection. The growth of leukemias treated with doxorubicin or the tyrosine-kinase inhibitor Imatinib and myeloma with Velcade, at levels that were not toxic to the developing embryonic BM, was dramatically inhibited in vivo when the drug was injected together with cells on E11 or 48-72 hours after injection of the cells and homing to the BM. Using flow cytometry analysis the frequency of CHRF cells (detected with anti-human CD33) was reduced from 8% to 0.01% and K562 and LAMA (detected with anti-human CD71) from 1%-3% engraftment to <0.17% following treatment with Imatinib. Q-PCR analysis supported these results showing an average 8 fold reduction of CHRF and a 2-5 fold reduction of K562 and LAMA cells in the Imatinib treated turkey embryos. The ARH-77 cells (detected with anti-human CD38 and CD138) were inhibited from 8.5% to 0.72% after Velcade treatment, with a 16.5 fold reduction determined by Q-PCR analysis compare to untreated embryos. These results prove the efficacy and demonstrate the utility of the turkey embryo as a new complementary in-vivo model for studying cancer initiating cells and the growth of human blood malignancies and their response to treatment. With further improvements, it may provide an affordable, rapid in vivo system for studying the growth blood malignancies and help reduce time and cost of drug development.