Detection of minimal residual disease in acute leukemia: possibilities and limitations.

Studies are presented on the detection of 'minimal residual leukemia' using a monoclonal antibody (MCA) and fluorescence-activated cell sorting (FACS). As a preclinical model the BN rat acute myelocytic leukemia was used (BNML). The MCA Rm124 (IgM) strongly binds to BNML cells as measured by fluorescence intensity of a second-layer antibody (goat anti-mouse IgM fluorescein isothiocyanate). Only weak cross-reactivity occurred with normal mature granulocytes. It appeared possible to detect as low as 1 leukemic cell per 10,000 normal marrow cells, both in artificial mixtures and in marrows obtained after in vivo remission-induction chemotherapy with cyclophosphamide. Furthermore, an example is given of describing the kinetics of growth of leukemia in the liver, based on serial determinations of leukemic cells with the MCA technique. The population doubling time (Td) in the liver calculated in this way did not significantly differ from that derived from time-consuming and expensive bioassays. Finally, the extrapolation of these techniques developed in a preclinical model to studies in human acute leukemia is discussed.