The Kinetics of Reduction of Minimal Residual Disease Impacts on Duration of Survival and Response of Adult Patients with Acute Myeloid Leukemia.

By multiprametric flow cytometry, we have previously showed that MRD negativity, as defined by a number of BM residual leukemic cells (BMRLC) <3.5x10⁻⁴ after the consolidation, was associated with a significant longer relapse free survival (RFS) and overall survival (OS), in adult AML patients. The present study was designed to confirm the prognostic role of a delayed MRD determination (at consolidation check-point) and and to verify whether the kinetics of MRD levels measured after induction and consolidation time-points is associated with outcome. One hundred and 5 patients were enrolled in the EORTC/GIMEMA protocols AML-10/AML-12 (age < 61 yrs) and AML-13 (age > 61 yrs), all consisting in intensive induction and consolidation cycles, and, for patients aged <61 years, autologous or allogeneic stem cell transplantation. Median age was 52 years (range 18–78), all FAB subtypes were represented with the exception of APL cases. The analysis of the quartile distribution of MRD levels identified a threshold of 2.0x10⁻⁴ BMRLC corresponding to the lower quartile, either after induction or consolidation; this value was selected to discriminate MRD negative (MRD^neg; ≤2.0×10⁻⁴ BMRLC) from MRD positive (MRD^pos; >2.0×10⁻⁴ BMRLC) patients. After induction, 30 (29%) of 105 patients were MRD^neg, the remaining 75 (71%) were MRD^pos. Fifthy four of 75 (72%) patients in the MRD^pos group underwent relapse at a median time of 13 months (range 1–120), while in the MRD^neg group 14/30 (47%) had disease relapse at a median time of 21 months (range 4–147) (P=0.014). The 5-year actuarial probability of OS and RFS was 47% and 50%, respectively, for patients in the MRD^neg group, compared to 27% and 25%, respectively, for those in the MRD^pos group (P=0.031 and 0.008, respectively). Seven patients had early relapse after induction therapy and 98 proceeded to receive consolidation; of these, one died of septic shock and 97 were evaluable for successive MRD assessment. At the post-consolidation time point, 26 (27%) of 97 patiens were MRD^neg whereas 71 (73%) were MRD^pos. Six of 26 (23%) and 55 of 71 (77%) of patients in MRD^neg and in MRD^pos group underwent relapse at a median time of 73 months (range 3–147) and 12 months (range 2–120), respectively, (P<0.001). The 5-year actuarial probability of OS and RFS was 67% and 74%, respectively, for patients in the MRD^neg group, compared to 23% and 20%, respectively, for those in the MRD^pos group (P<0.001 for both comparisons). Based on the comparative analysis of MRD levels detected at the induction and consolidation time-points, we segregated the series of 97 patients into three groups: I) 26 patients, MRD^neg after consolidation; II) 12 patients, still MRD^pos after consolidation but showing a reduction by at least 1-log between induction and consolidation (chemosensitive MRD^pos); III) 59 patients, MRD^pos after consolidation with increased levels of MRD between induction and consolidation check-points (chemoresistant MRD^pos). These groups differed significantly in terms of relapse rate (23% vs 50% vs 83%, P<0.001), duration of OS and RFS (P<0.001 for both comparisons). In conclusion, 1) post-consolidation MRD determination provides the best predictive informations on patients outcome; 2) the comparative analysis of MRD levels between induction and consolidation time-points may allow the identification of MRD^pos patients with different prognosis (chemosensitive vs chemoresistant).