Abstract
Abstract 1430
Acute myeloid leukemia (AML) is a heterogeneous disease, originating from leukemic stem cells (LSCs). This relatively rare sub-population shows mostly a CD34+CD38- immunophenotype and has been proposed to be responsible for disease relapse and chemotherapy resistance. In order to identify effective drug(s) for clinical use that can eliminate LSCs, we have screened a library of 2000 compounds from the SPECTRUM collection (MicroSource Discovery Systems, Inc.) using a multi-parameter flow cytometry high-throughput screening. The drugs compiled in the library are 40% US FDA approved drugs, 30% natural products, 10% approved drugs for use in Europe and Japan, and the remainder consistent of bioactive compounds.
The initial screen aimed to identify potential compounds that can decrease viability or increase differentiation of AML cells at 24 and 72 hours of treatment. Compounds demonstrating autofluorescence were eliminated as false positive hits.
We found 137 compounds that decreased viability in all populations. Furthermore, 37 drugs demonstrated ability to increase cell death (more than 50%) in phenotypically described LSCs, progenitor and blasts populations, but with viability over 50% for normal residual lymphocytes within the same AML specimen. Interestingly, only 4 drugs displayed higher specificity to LSCs when compared to total blast populations. Surprisingly, two of these drugs are different forms of retinoic acid (isotretinoin and tretinoin).
The selectivity of isotretinoin and tretinoin was confirmed in different primary AML samples. We observed that phenotypically described stem cell subpopulations were affected when exposed to retinoic acid in 4 out of 5 AML samples. LSC populations display increased apoptosis and differentiation assessed by flow cytometry. Characterization of what confers sensitivity to these types of drugs is under investigation. Together, a multiparameter flow cytometry screen revealed specific anti-LSC activity for retinoid acid, which was not detectable when analyzing total populations and would have been missed using traditional screening. These studies underscore the utility of LSC-based assays for defining novel therapies and suggest that retinoid-based therapies merit further investigation in AML.
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.