Search

Optimization of therapy for childhood acute lymphoblastic leukemia (ALL) requires a greater understanding of the cells that proliferate to maintain this malignancy because a significant number of cases relapse, resulting from failure to eradicate the disease. Putative ALL stem cells may be resistant to therapy and subsequent relapses may arise from these cells. We investigated expression of CD133, CD19, and CD38 in pediatric B-ALL. Cytogenetic and molecular analyses demonstrated that karyotypically aberrant cells were present in both CD133(+)/CD19(+) and CD133(+)/CD19(-) subfractions, as were most of the antigen receptor gene rearrangements. However, ALL cells capable of long-term proliferation in vitro and in vivo were derived from the CD133(+)/CD19(-) subfraction. Moreover, these CD133(+)/CD19(-) cells could self-renew to engraft serial nonobese diabetic-severe combined immunodeficient recipients and differentiate in vivo to produce leukemias with similar immunophenotypes and karyotypes to the diagnostic samples. Furthermore, these CD133(+)/CD19(-) ALL cells were more resistant to treatment with dexamethasone and vincristine, key components in childhood ALL therapy, than the bulk leukemia population. Similar results were obtained using cells sorted for CD133 and CD38, with only the CD133(+)/CD38(-) subfraction demonstrating xenograft repopulating capacity. These data suggest that leukemia-initiating cells in childhood B-ALL have a primitive CD133(+)/CD19(-) and CD38(-) phenotype.

A significant proportion of children with T-cell acute lymphoblastic leukemia (T-ALL) continue to fail therapy. Consequently, characterization of the cells that proliferate to maintain the disease should provide valuable information on the most relevant therapeutic targets. We have used in vitro suspension culture (SC) and nonobese diabetic-severe combined immune deficient (NOD/SCID) mouse assays to phenotypically characterize and purify T-ALL progenitor cells. Cells from 13 pediatric cases were maintained in vitro for at least 4 weeks and expanded in 8 cases. To characterize the progenitors, cells were sorted for expression of CD34 and CD4 or CD7 and the subfractions were evaluated in vitro and in vivo. The majority of cells capable of long-term proliferation in vitro were derived from the CD34+/CD4- and CD34+/CD7- subfractions. Moreover, the CD34+/CD4- or CD7- cells were the only subfractions capable of NOD/SCID engraftment. These T-ALL cells successfully repopulated secondary and tertiary recipients with equivalent levels of engraftment, demonstrating self-renewal ability. The immunophenotype and genotype of the original leukemia cells were preserved with serial passage in the NOD/SCID mice. These data demonstrate the long-term repopulating ability of the CD34+/CD4- and CD34+/CD7- subfractions in T-ALL and suggest that a cell with a more primitive phenotype was the target for leukemic transformation in these cases.

A preliminary study has linked raised blast glutathione levels with chemoresistance in acute myeloid and lymphoblastic leukemia in adults and children. In this study, therefore, the relationship between leukemic blast glutathione levels and prognosis in childhood acute lymphoblastic leukemia (ALL) was investigated. A total of 77 childhood ALL samples were analyzed, 62 at initial presentation and 15 at relapse. A 20-fold interindividual variation in glutathione levels at presentation (median, 6.54 nmol/mg protein; range, 1.37 to 27.9) was demonstrated. The median level in T-lineage ALL was 2. 3-fold higher than in B-lineage ALL (Mann-Whitney test, P <.0001). There was a significant correlation between presenting white cell count (WBC) and glutathione level (Spearman rank correlation coefficient, rho = 0.45, P =.001). A high DNA index correlated with low glutathione levels (Mann-Whitney test, P =.013). There was no significant relationship between glutathione levels and in vitro drug sensitivity. Patients with glutathione levels above the median had a significantly greater risk of relapse (log-rank test statistic, 5.55; P =.018), and the overall survival rate was significantly reduced (log-rank test statistic, 4.38; P =.04). Multivariate analysis demonstrated that glutathione concentration was of independent prognostic value when assessed in conjunction with age, gender, WBC, and immunophenotype. The association of elevated blast glutathione levels with an increased risk of relapse suggests that glutathione-depleting agents may be of therapeutic value in patients who present with a high WBC.