Functional and Phenotypic Characterization Of Acute Myeloid Leukemia By Analysis Of Diagnostic/Relapse Paired Samples

Treatment of relapsed acute myeloid leukemia (AML) results in lower complete remission rates compared to treatment of AML at diagnosis. Immunophenotypic changes in leukemic blasts are common from diagnosis to relapse (Baer et al. Blood. 2001), suggesting that the underlying biology of AML changes with disease progression. A better understanding of the biologic properties of AML cells at different stages of disease will facilitate the development of biomarker tools and more effective therapies. We therefore studied paired diagnostic/relapse samples obtained from AML patients. Cells from 11 paired samples were transplanted into immune deficient mice (NOD.SCID IL2Rg null) over a range of cell doses. In 10 of 11 patients, a leukemic graft could be generated after transplantation of lower cell doses from the relapse sample compared to the paired diagnostic sample. By limiting dilution analysis, leukemia stem cell (LSC) frequency was higher in relapse samples (1 in 5.8×102 to 1 in 2.4×106, median 1 in 2.0×103) compared to diagnostic samples (1 in 5.0×103 to 1 in 6.1×106, median 1 in 5.5×104); the fold increase in LSC frequency ranged from 2.2 to 745 (median 8.6). Multiparameter flow cytometric analysis carried out on 13 paired diagnostic/relapse samples demonstrated an increase in 2 known stem cell markers, CD34 and CD117, from diagnosis to relapse: CD34 was gained or increased at relapse in 7/13 (54%) of paired samples, while CD117 was gained or increased at relapse in 9/13 (69%) of paired samples. We plan to take a comprehensive approach to examine the surface marker expression of paired diagnostic/relapse samples by a high throughput flow cytometric screen (HTS) of both bulk and LSC-containing AML populations to identify markers that are altered at relapse. As a first step, we have performed HTS of 373 surface markers on 10 AML patient samples, including a relapse sample from a diagnostic/relapse pair. A significant proportion of markers (155/373, 42%) were expressed on less that 5% of cells in all 10 AML samples analyzed. We will therefore focus on the remaining markers in our HTS analysis of paired samples. Surface markers that are differentially expressed from diagnosis to relapse will be further characterized in order to gain insight into disease progression and identify potential therapeutic targets. Disclosures: No relevant conflicts of interest to declare.