A panel of commercially available monoclonal antibodies and five heteroantisera were used to distinguish and subtype 138 cases of acute leukemia (AL). The immunophenotype was compared with the French- American-British (FAB) classification obtained on the cases. The immunophenotype discriminated acute myelogenous leukemia (AML) from acute lymphoblastic leukemia (ALL) and recognized cases not distinguished by cytochemistry (22% of cases), mixed lineage phenotypes (13% of cases), and cases with separate populations of lymphoblasts and myeloblasts (one case). Using the immunologic panel and derived criteria to subtype AML, correspondence of the immunophenotype to the FAB subtypes M1, M2, M4, and M5 was possible in greater than 80% of cases. A combined classification of the immunophenotype and FAB morphology/cytochemistry was devised for AML subtyping. It is recommended that immunophenotyping should be done at least in all cases with negative orinconclusive cytochemistry. At present, we suggest that until a “gold standard” for identifying leukemic subtypes is developed, the best method for typing acute leukemia is by using a combination of morphology, cytochemistry and immunophenotyping.

Acute mixed myeloid-lymphoid leukemia is uncommon. We report four cases in which myeloid and lymphoid cell markers were observed simultaneously or sequentially when 94 patients with acute leukemia were phenotyped according to the French-American-British (FAB) classification system, with cytochemical stains, and with immunologically defined differentiation markers (identified by monoclonal antibodies and antiterminal deoxynucleotidyl transferase [TdT]). In one case, conversion from acute lymphoblastic leukemia to acute myeloid leukemia was noted (FAB L1, TdT+ to FAB M4, Auer rods, TdT-). In another patient, two distinct populations of myeloid and lymphoid blast cells were observed simultaneously (TdT-, LeuM1+/TdT+, LeuM1-). In two additional patients, acute leukemia was characterized by the expression of both lymphoid and myeloid markers on the same cell (TdT+/Leu M1+, B4+/Leu M1+ and greater than or equal to 70% TdT+, T11+, My9+). The Philadelphia (Ph1) chromosome was negative in all cases, though other chromosomal abnormalities were noted in three out of four cases. Malignant transformation of a pluripotential stem cell for both lymphoid and myeloid lineages, with or without the Ph1 chromosome marker, could explain the coexistence of distinct populations of lymphoblasts and myeloblasts in acute leukemia. Acute leukemia with a biphenotypic profile may reflect genome depression accompanying neoplasia.