Investigating the Expression of the MRD Marker CD58 on Leukaemia Initiating Cells in Childhood Acute Lymphoblastic Leukaemia

Abstract 1887

Further improvements in outcome for childhood acute lymphoblastic leukaemia (ALL) will require a better understanding of the underlying biology of this disease and the fundamental mechanisms of drug resistance. The discoveries that a few populations can initiate leukemia in mouse models and that new populations of leukaemia initiating cells (LIC) can be detected following an initial round of transplantation in these models raises important questions about the biology of the leukaemias. If several cell populations have LIC properties, what are the relationships of these populations to each other and which populations are most important to target with therapy? It will also be important to determine whether there is any correlation in the biological properties of LIC identified in the model systems with the response of the patients to therapy. Assessment of minimal residual disease (MRD) levels provides a sensitive measurement of early treatment response and permits detection of the in vivo selected drug resistant population. CD58 (leucocyte function-associated antigen 3; LFA-3) is a useful marker in MRD tracking of B cell precursor (BCP) ALL. CD58 is over expressed in these cases permitting discrimination of leukaemia blasts from normal B cells. In this study we investigated whether CD58 is expressed on LIC populations in childhood ALL. Expression of CD58 and CD34 was assessed in a cohort of 12 diagnostic samples with mixed prognoses and compared to levels detected in 11 normal bone marrow (NBM) samples. Levels of CD58 were significantly higher in the ALL cases (57.4±37.7%) than on NBM cells (21.1±12.2%; p=0.007). Likewise, the CD34⁺/CD58⁺ population was larger in ALL cases than in normal cells (22.2±34.7% and 0.25±0.25%, respectively; p=0.05). Cells from eight of the 12 patients, were sorted on the basis of expression or lack of expression of these markers and the functional ability of the sorted subpopulations was assessed in vitro and in vivo. On sorting, the majority of cells were CD34⁻/CD58⁻ (43.7±39.2%), 20.7±30.7% were CD34⁻/CD58⁺, 19±14.3% were CD34⁺/CD58⁺ and the CD34⁺/CD58⁻ population accounted for 16.6±35.3%. Unsorted cells and all 4 sorted populations were set up in long-term culture to assess proliferative capability and the in vivo propagating potential was assessed in NSG mice. All 4 sorted subpopulations proliferated over the 6 week period but the highest levels of expansion were observed in the cultures of CD34⁺/CD58⁺ (6–420 fold) and CD34⁺/CD58⁻ (3–24 fold) cells. Cytogenetic analyses confirmed that leukaemia cells were maintained in the culture system. Results from the in vivo analyses on 5 cases to date indicate that all 4 subpopulations contain LIC. In these cases, higher levels of engraftment were observed with CD34⁺/CD58⁺ (up to 20%) and with CD34⁻/CD58⁻ subpopulations (6.1-98%). Serial transplantation studies will determine whether there are differences in the repopulating and self-renewal abilities of these LIC. These findings suggest that using CD58 alone or in combination with CD34 would be insufficient to track disease progression in ALL. Incorporating additional markers that are commonly used in MRD panels will provide valuable information on LIC populations and facilitate development of improved disease monitoring.