Peripheral blood blast clearance during induction therapy in acute myeloid leukemia

To the editor:

Elliott and colleagues have reported in a group of 73 patients with acute myeloid leukemia (AML) that the time to clearance of peripheral blood blasts (PBB) during standard induction therapy is a strong predictor of both overall (OS) and relapse-free survival (RFS).¹  We have previously shown²  in 30 AML patients that the kinetics of PBB clearance is a predictor of complete remission (CR). Thus, the 2 studies have in common the objective to obtain the maximum predictive information from the analysis of peripheral blood (ie, a much less invasive procedure than bone marrow aspiration); however, their results differ in several respects.

(1) The study reported by Elliott et al was retrospective whereas ours was prospective. (2) Elliott et al assumed that PBB clearance is a surrogate of in vivo chemosensitivity, but their study was carried out only on responder patients whose leukemic cells are, by definition, at least sufficiently chemosensitive for the patients to achieve CR; our study, instead, was carried out on unselected consecutive patients. (3) In the study by Elliott et al, PBB clearance was evaluated by differential count; in our study we identified by flow cytometry for each patient at the time of diagnosis a population of leukemic cells with aberrant immunophenotype (LAIP), and then determined absolute LAIP-positive blast counts on each of the first 5 days of treatment. Approval was obtained from Careggi Hospital institutional review board for this study. Informed consent was obtained in accordance with the Declaration of Helsinki.

By our approach (having doubled our series²  to 61 patients), we have observed from day 2 (ie, within 24 hours from starting therapy) a clear dichotomy between responders and nonresponders (Figure 1A): indeed, the difference between the medians in the 2 groups is statistically significant from day 2. CR took place in 31 of 41 (76%) patients who had a reduction greater than 2 logs on day 5; but in only 1 of 20 (5%) patients who had a lesser reduction.

Unlike Elliott et al, in our series we do not yet have long-term follow-up data. However, because we found that peripheral blood LAIP-positive cell clearance correlates with bone marrow LAIP-positive residual disease (LD14: see Figure 1B), and residual disease in turn is known to correlate with RFS,^3-5  it is reasonable to assume that PBB clearance will correlate with RFS. Thus, the combined data provided by Elliott et al and by our study demonstrate that from peripheral blood analysis it is possible to obtain strong predictors of both CR and RFS: in this respect the 2 studies are complementary. We concur therefore in the hope that PBB clearance could be of help also with respect to tailoring treatment.