RESUMO
BACKGROUND: Enumeration of blasts in the bone marrow is an essential component in the diagnosis and treatment of acute leukemia. The current gold standard method is based on a morphologic counting of 500 marrow nucleated cells despite its operator dependence and inter-observer variability. OBJECTIVES: To compare the percentages of marrow blasts derived from two different approaches comprising routine morphology-based manual counting and flow cytometric analysis. MATERIAL AND METHOD: Fifty-five marrow samples were collected from 38 acute leukemia patients (36 AML and 19 ALL) after hematologic recovery from chemotherapy. The blast percentages were enumerated manually and by flow cytometer using CD45 and side scatter gates. RESULTS: A good correlation was found in the overall 55 samples (r = 0.829) and 36 AML samples (r = 0.86). The blast percentages derived from flow cytometer were higher than from morphologic counting in 46 samples (83.6%). Using a cut-off point of < 5% blasts to define complete remission (CR), 48 cases (87%) were classified as morphological CR (83% CR in AML and 95% CR in ALL). By flow cytometry, only 24 cases (44%) were in CR (28% CR in AML and 74% CR in ALL). The results from each method were concordant in determining CR in 27 samples (49%), with a kappa value of 0.07 for overall samples, 0.057 for AML and -0.096 for ALL samples. CONCLUSION: A good correlation between the percentages of blasts achieved by either method was demonstrated, particularly in AML samples. Discordant results occurred when <5% blasts were used as a cut-offpoint to determine CR. Both methods should be complementarily performed to ensure a truly complete response to chemotherapy. The method discrepancy should be further investigated in order to increase the level of confidence in CR status.