CLINICAL Flow Cytometric Analysis of Acute Leukemias Ricardo E. Duque Norwood Clinic, Birmingham, Alabama Eric T. Everett Medical University of South Carolina, Charleston, South Carolina Jose Iturraspe University of Florida College of Medicine, Gainesville, Florida A cute leukemias are currently classi- fied by the well known French- American-British (FAB) clas- sification.4 Lymphoblastic leukemias are separated into three morphologically defined subgroups: L1, L2, and L3 (the last group representing Burkitt's lym- phoma in leukemic phase). Nonnymphoid leukemias are separated into seven mor- phologically and cytochemically defined subgroups: M1 (myeloblastic), M2 (my- eloblastic with differentiation), M3 (pro- granulocytic), M4 (myelomonocytic), M5a (monoblastic), M5b (monocytic or pro- In This |ssu~' monocytic), M6 (erythroid) and M7 (megakaryoblastic). Myelodysplastic syn- dromes, which involve the progressive ex- pansion of pluripotential hematopoietic clones, eventually terminate in acute leu- kemia in approximately 30% of pa- tients. 17 Their classification is not always straightforward. The FAB group has also established a classification system for myelodysplastic syndromes s and, more re- cently, revised criteria to more clearly separate these processes from acute leu- kemias.6 In the diagnosis of acute leukemias, the interobserver and intraobserver reproduc- ibility is in the range of 60-70% for mor- phologic classification alone, 89% for morphologic and cytochemical classifi- cation, and 99% for when immunopheno- typic information is added. 7 In general, morphologic reproducibility is poor be- cause it is subjective. In the classification of non-Hodgkin's lymphomas, for ex- ample, certain categories have a consensus diagnosis as low as 43% (consensus meaning four out of seven expert patholo- gists). ~ Therefore, this poor reproduci- continued on page 44 Basic Phenotyplng of Lymphocytes: Selection and Testing of Reagents and Interpretation of Data Anne Jackson Becton Dickinson Immunocytometry Systems, San Jose, California A s little as 15 years ago, a clinical immunology laboratory performed only a small number of fluorescent antibody tests (FTA-abs for syphilis, anti- nuclear antibodies [ANA], or surface im- munoglobulin [Ig] for B cells). Polyclonal antibodies specific for the immunoglobulin heavy chains were often found to have unacceptable cross reactivity due to traces of anti-light chain contamination, and those which had specificity for cell surface antigens had very low titers. The devel- opment of monoclonal antibodies by Kohler and Milstein6 revolutionized the field of human immunology, and anti- bodies with specificities previously only four-I in research laboratories are now widely available. Of human leukocyte surface antigens alone, 78 clusters of dif- continued on page 50 CIMNDC 10(4)43-62,1990 Elsevier 0197-1859/90/$0.00 + 2.20