Abstract
Background:
Acute myeloid leukemia (AML) is most common in people over the age of 60 where it remains an almost incurable malignancy with a grim prognosis. Evaluation of new therapeutic agents in vitro and in vivo is critical for drug development, yet there are few in-vivo models for studying human leukemia and its therapy. The best model used is the high cost immune-deficient mice and that require several weeks to assess drug response. To complicate matters, AML almost certainly relapses with cells that are not necessarily exactly identical to the original malignant clone, often limiting therapeutic options. The development of anti-leukemia therapies could be facilitated by a rapid and cost effective in vivo system for evaluating response to new drugs. Additionally, decisions regarding personalized treatment for relapsed or refractory leukemia patients must be rapid, and produce results within several days, as longer time periods can be detrimental. Herein, we describe a fast, economical, in ovo turkey embryo model, which provides a unique system to meet these requirements. The model can be used for the assessment of human leukemia infiltration in medullary and extramedullary tissues and more importantly for rapid testing of anti- leukemic agents within the bone marrow (BM). This system can be applied for new drug development and for personalized real time response of patient cells to potential leukemia therapies.
Methods:
BCR/Abl+ AML lines K562 and LAMA-84 , c-Kit+ CHRF-4288, fresh AML patient and Raji Burkit lymphoma cells (5x106) were injected into turkey egg chorioallantoic membrane (CAM) veins on embryonic day E11 previously optimized (1). Engraftment in BM was detected by flow cytometry (FC) using anti-human CD71 or anti- human CD33 for AML and anti human CD45 for Raji cells, or by Quantitative real time PCR (Q-PCR) comparing the amount of genomic human to the amount of avian DNA and number of human /avian cells in BM. Drug response was tested by IV injection of therapeutic range doses of Imatinib (Glivec ®), Doxorubicin or dexamethasone, 48H after grafting cells. Drug levels were precalibrated to be non-toxic to the developing embryo by LD50 and BM cell viability compared to control (Taizi M et al Exp Hem 34:1698,2006, Grinberg I et al, Leuk Res. 33:1417, 2009). Six days later (E19) the embryos were sacrificed and the BM collected for FC and hematopoietic and non-hematopoietic tissues for molecular analysis.
Results:
The kinetics of leukemia cell engraftment in the BM on E15, E18, and E23 in BM and liver after cell injection on day E11 was assessed to determine the optimal treatment and readout times. The highest engraftment level in BM and liver was detected at E18 by Q-PCR, and FC in more than 90% of the injected embryos. We quantitatively compared the engraftment of AML cells at E18-20 without and with drug treatment that was administered IV 48 hours after cell injection. The average engraftment (±SD.) in the BM after one week was 4.5%+1.7 K562, 5.83% +0.88 LAMA-84, 11.2%+3.5 CHRF-4288, 8.9% +1.6 Raji (n=7-15 per group) and 2.5% fresh leukemia cells detected by FC and confirmed by Q-PCR. A single dose of either 0.75 mg Imatinib or 50 mg Doxorubicin /embryo previously calibrated to be non toxic to the embryos reduced engraftment of AML cells in BM and in several other organs by more than ten fold. A similar effect was also obtained by a single dose of 5mg dexamethasone in Raji injected embryos. Treatment with 0.5 mg Imatinib on injected ARH-77 (multiple myeloma) or Raji cells had no effect on cell engraftment, while treatment with a single non toxic dose of Revlimid as previously described eliminated engraftment of ARH77 cells (1), clearly demonstrating the specificity of the drugs and utility of the system.
Conclusion:
Our study demonstrates the potential utility of a practical and unique avian embryo model for testing drug activity on human AML cells in vivo. This system, under preclinical development, is expected to provide a new xenograft platform for real time affordable testing leukemia therapies. More importantly, this may open a new venue for individualized screening for response or resistance to specific therapeutic agents for the relapsed or refractive patient and may lead to better optimization of practical and applicable therapeutic strategies.
Disclosures
No relevant conflicts of interest to declare.
Detection of infusate leakage in the brain using real-time imaging of convection-enhanced delivery
