Rapid decline of chronic myeloid leukemic cells in long-term culture due to a defect at the leukemic stem cell level.

In this report we describe a quantitative in vitro assay for the most primitive type of leukemic precursors yet defined in patients with chronic myeloid leukemia (CML). This assay is based on the recently described "long-term culture-initiating cell" (LTC-IC) assay for primitive normal human hematopoietic cells. Such cells, when cocultured with competent fibroblast feeder layers, give rise after a minimum of 5 weeks to multiple single and multilineage clonogenic progenitors detectable in secondary semisolid assay cultures. Similar cultures initiated by seeding a highly enriched source of leukemic cells from patients onto normal feeders showed the clonogenic cell output after 5 weeks to be linearly related to the input innoculum over a wide range down to limiting numbers of input cells, thus allowing absolute frequencies of leukemic LTC-ICs to be determined using standard limiting dilution analysis techniques. Leukemic LTC-IC concentrations in CML marrow were found to be decreased, on average to less than 10% of the normal LTC-IC concentration in normal marrow, but were greatly increased (up to greater than 10(5) times) in CML blood. Assessment of the number of clonogenic cells produced per leukemic LTC-IC by comparison to normal blood or marrow LTC-IC values showed this function to be unchanged in leukemic LTC-ICs [i.e., 3.1 +/- 0.4 clonogenic cells per CML LTC-IC (mean +/- SEM, n = 6) versus 3.7 +/- 1.2 (n = 3) and 4.3 +/- 0.4 (n = 5), respectively, for normal blood and marrow LTC-ICs]. In contrast, leukemic LTC-IC maintenance in LTC proved to be highly defective by comparison to normal LTC-IC of either blood or marrow origin. Thus, when cells from primary LTC were subcultured into secondary LTC-IC assays, leukemic LTC-IC rapidly declined (greater than 30-fold) within the first 10 days of culture, whereas normal LTC-IC numbers remained unchanged during this period. These findings illustrate how self-maintenance and differentiation events in primitive human hematopoietic cells can be differentially modulated by an oncogenic process and provide a framework for further studies of their manipulation, analysis, and therapeutic exploitation.