Immunophenotypes for 272 patients with acute myelogenous leukemia (AML) were analyzed using a panel of 22 antibodies. Numerical evidence for unusual coexpressions (present in normal marrow at < or = 0.1%) of surface markers on = or = 10% of the blast cells was found in 85% of all cases. Asynchronous expression of myeloid differentiation antigens occurred in 70% of the cases. Unusual coexpression of T-lymphoid, B- lymphoid, or natural killer (NK) markers with myeloid markers occurred in 38%, 13%, and 21%, respectively, of all AML cases. Two- and three- color analyses confirmed coexpression in 15 of 15 cases, and indicated that these percentages are an underestimate, because coexpression can be demonstrated in cases without numerical overlap. These data indicate that the unusual coexpression of normal differentiation antigens is a common occurrence in AML. Markers in 12 of 13 patients were similar between presentation and relapse, and in two patients, unusual phenotypes detected at first relapse were shown at second relapse, indicating these immunophenotypes are stable in the majority of AML patients. Significant correlations were found between t(8;21) cytogenetics and coexpression of CD19 with CD15 or CD34, t(9;22) and coexpression of CD19 and CD34, and t(15;17) and coexpression of CD2 and myeloid antigens. Multiparameter fluorescence analysis allows detection of unusual phenotypes when the blast counts are < 5% (classical remission). Analysis of 16 patients in remission indicated the presence of presentation phenotypes in 0.2% to 7.9% of the lymphocyte + blast light scatter region, representing 0.03% to 1.4% of the total nucleated marrow cells. Of the 16 patients with = or = 4 months follow-up after detection of these cells, 6 of 6 patients with = or = 0.2% unusual presentation phenotypic marrow cells have relapsed, while 9 of 10 patients with < 0.2% remain in remission. The detection of cells with the unusual presentation phenotype may reflect residual AML cells, and their increase may predict relapse.

This content is only available as a PDF.
Sign in via your Institution