Kinetics of Bone Marrow Blasts during Remission Induction Course in Acute Myeloid Leukemia: Effect on Complete Remission and Relapse-Free Survival.

Abstract

Achievement of complete remission (CR) is crucial to prolong survival in acute myeloid leukemia (AML). The definition of CR has been well established; however, there are no objective measures for deciding when the probability of achieving CR has become so low that a patient's disease can be considered resistant to therapy. In particular, it can be difficult to distinguish patients with resistant disease from those with persistent disease but who subsequently will enter CR, a distinction that underlies the decision to start a second course or change therapy. We attempted to facilitate this decision by examining the relation between the % marrow blasts 21 days, and later, after start of course 1 of initial induction therapy and the subsequent probability of CR on course 1. Our database consisted of the 593 adults with AML (≥20% blasts, acute promyelocytic leukemia excepted) who had bone marrow examined 21 days after beginning induction therapy including cytarabine at cumulative dose of 5–6 g/m2 at M. D. Anderson Cancer Center from 1995 to 2004. 340 of the 593 patients had an additional bone marrow examination between day 22 and day 28 (day 22–28) of course 1; similarly, day 29–35 marrows were done in 185 patients, day 36–42 marrows in 89 patients and so on. Bone marrows were categorized as morphologic leukemia-free state (MLFS; <5% blasts), persistent disease (PD; ≥5% blasts), or too few cells to count (TFTC). 197 of the 593 patients (33%) had an MLFS on day 21. This conferred a 94% probability of CR on course 1, independent of cytogenetic group. 275 patients (47%) had PD on day 21 and 57% of these 275 entered CR on course 1, with the probability of subsequent CR being predictable from the combination of cytogenetics, day 21 bone marrow blasts, and day 21 platelet and neutrophil counts. Patients with PD on day 21, but who achieved an MLFS on day 28 were highly likely to enter CR (40/47). However, those with PD beyond day 28 were very unlikely to enter CR on course 1, and no CR was observed in patients with PD after day 43. 121 patients (20%) had a TFTC marrow on day 21, with this finding associated with a CR rate of 72% on course 1 (p<0.001 vs MLFS, and p<0.001 vs PD). Not surprisingly given the respective CR rates, patients with MLFS on day 21 had significantly longer survival than patients with PD and patients with TFTC marrow (p<0.001). However, Relapse-free survival was not different among the 3 groups (p=0.109), which was also confirmed by multivariate analysis accounting for cytogenetics, antecedent hematologic disorder, and age. These results appear useful in management of AML, and recommend that bone marrow be examined on day 28, in patients with PD on day 21 and a <50% probability of subsequent CR and in patients with TFTC on day 21. Should PD persist on day 28, and especially on day 35, a second course should be started or new therapy instituted.