In acute myeloid leukemia (AML) patients achieving complete remission, the levels of minimal residual disease (MRD) as determined by flow cytometry have been shown to impact on remission duration and survival. However, some issues such as the most suitable source (BM or PB) or the most appropriate timing (early or delayed evaluation) for MRD determination are still a subject of debate. In our experience, we observed that MRD negativity, as defined by a number of bone marrow residual leukemic cells (BMRLC) <3.5x10−4, after consolidation cycle was associated with a significantly longer disease free survival (DFS) and overall survival (OS). Based on this, the present study was designed to analyze the kinetic of MRD reduction during the post-induction phase, and therefore to determine to what extent post-induction chemotherapy might impact on the outcome of patients with AML. Eighty-nine adult patients with AML were entered into the EORTC/GIMEMA protocols AML10/AML12 (age <61 yrs) or AML13/AML15 (age >61 yrs), all consisting in intensive induction and consolidation cycles, and allogeneic or autologous stem cell transplantation for patients aged < 61 years. Median age was 53 years (range 17–78), all FAB subtypes were represented with the exception of APL cases which were not included. Eighty-one of 89 patients maintained complete remission after consolidation (8/89 had early relapse after induction) and were suitable for the analysis. After consolidation cycle, 32 of 81 (39%) patients had <3.5x10−4 BMRLC and were considered MRD neg. The remaining 49 (61%), were MRD pos since the measured levels of MRD were ≥3.5x10−4 BMRLC. Among these MRD pos patients, in 16 the levels of MRD, although still above the value of 3.5x10−4 BMRLC, were significantly reduced as compared to the post-induction levels (median reduction 6x10−4 BMRLC, range 0–643). Therefore, these 16 patients were considered as having a “chemosensitive MRD” and in fact, within this MRD pos category, they had a superior duration of DFS and OS (P=.010 and .004, respectively) as compared to MRD pos patients with “chemoresistant MRD”, namely those not showing improvement in the level of MRD between the induction and consolidation course. Therefore, we identified 3 discrete categories of patients: 1) 32 patients MRD neg at the end of consolidation therapy (BMRLC <3.5x10−4); 2) 16 patients MRD pos at the end of consolidation therapy (BMRLC ≥3.5x10−4) but with “chemosensitive MRD”; 3) 33 patients MRD pos at the end of consolidation therapy (BMRLC ≥3.5x10−4) but with “chemoresistant MRD”. These 3 groups differed significantly in terms of relapse rate (84% vs. 75% vs. 28%, respectively) both in univariate and multivariate analysis (P<.001). Accordingly, DFS and OS (at 5-years, 66%, 50 % and 12%, respectively) (P<.001) duration also differed; the multivariate analysis confirmed the independent prognostic role of MRD status at the end of consolidation (P<.001). In conclusion, 1) at variance with previous reports emphasizing the prognostic value of an early flow-cytometric determination of MRD, we have found that a delayed (post-consolidation) MRD evaluation provides the best predictive information on patients outcome; 2) the quantitative determination of MRD at specific time-points may also allow the identification of MRD pos patients with variable prognosis (chemosensitive vs chemoresistant MRD).

Author notes

Corresponding author

Sign in via your Institution