Marrow Hypocellularity, But Not Residual Blast Count or Receipt of Reinduction Chemotherapy, Is Prognostic on Day-14 Assessment in Acute Myeloid Leukemia Patients With Morphologic Residual Disease

In our single center retrospective study of day 14 bone marrow assessment in AML, we found cellularity to be a stronger predictor of response than blast percentage or receipt of re-induction chemotherapy. For patients with blast percent and cellularity. Induction chemotherapy for acute myeloid leukemia (AML) is based on the “7+3” cytarabine/anthracycline regimen. A nonhypocellular day 14 (D14) bone marrow sample with a blast count > 5% to 10% is suggestive of residual leukemia, for which a second course of induction chemotherapy has been recommended. Although the prognostic value of D14 bone marrow findings has been established, its use as a decision point is controversial because the benefit of repeat induction has been questioned. In the present single-center retrospective study of 113 patients with newly diagnosed AML, we evaluated the role of cellularity on the clinical outcomes of patients with residual morphologic leukemia (blasts ≥ 5%). Among 64 patients with D14 bone marrow samples, 31 had residual morphologic leukemia. The complete remission (CR) rates were greater for the hypocellular (11 of 16) than for the nonhypocellular (4 of 15) patients (P = .03). The median overall survival (OS) for the hypocellular D14 patients was longer than that for the nonhypocellular patients (17 vs. 8 months; P = .02). No significant difference between the receipt of reinduction therapy and CR or OS was found on logistic or survival model analysis. The specificity for residual leukemia on D14 bone marrow samples was better for cellularity ≥ 20% and blasts ≥ 20% than for blasts ≥ 5%. The results of our study have shown that patients with < 20% cellularity and < 20% blasts on the D14 bone marrow assessment should continue observation until recovery rather than receive additional immediate therapy.