A high complete remission rate is currently achieved in patients with acute myeloid leukemia (AML). However, many patients eventually relapse due to the persistence of low numbers of residual leukemic cells that are undetectable by conventional cytomorphologic criteria (minimal residual disease [MRD]). Using immunophenotypic multiparametric flow cytometry, we have investigated in sequential studies (diagnosis and follow-up) the impact of MRD detection on the outcome of 53 AML patients that had achieved morphologic remission with standard AML protocols and displayed at diagnosis an aberrant phenotype. Patients were studied at diagnosis with a panel of 35 monoclonal antibodies in triple staining combinations for detection of aberrant or uncommon phenotypic features. According to these features, a patient's probe was custom-built at diagnosis for the identification of possible residual leukemic cells during follow-up. The level of MRD at the end of induction and intensification therapy correlated with the number of relapses and relapse-free survival (RFS). Thus, patients with more than 5 x 10(-3) residual cells (5 residual cells among 1,000 normal bone marrow [BM] cells) identified as leukemic by immunophenotyping in the first remission BM showed a significant higher rate of relapse (67% v 20% for patients with less than 5 x 10(-3) residual cells; P = .002) and a lower median RFS (17 months v not reached; P = .01). At the end of intensification, with a cut-off value of 2 x 10(-3) leukemic cells, AML patients also separated into two distinct groups with relapse rates of 69% versus 32% (P = .02), respectively, and median RFS of 16 months versus not reached (P = .04). In addition, overall survival was also significantly related to the level of residual cells in the marrow obtained at the end of induction and particularly after intensification therapy (P = .008). Furthermore, we have explored whether residual disease was related with the functional expression of multidrug resistance (MDR-1) at diagnosis as assessed by the rhodamine123 assay. Patients with > or =5 x 10(-3) residual leukemic cells at the end of induction therapy had a significantly higher rhodamine-123 efflux (mean, 56% +/- 24%) than those with less than 5 x 10(-3) residual cells (mean, 32% +/- 31%; P = .04). Finally, multivariate analysis showed that the number of residual cells at the end of induction or intensification therapy was the most important prognostic factor for prediction of RFS. Overall, our results show that immunophenotypical investigation of MRD strongly predicts outcome in patients with AML and that the number of residual leukemic cells correlates with multidrug resistance.