Prognostic Value of Day 14 Bone Marrow Biopsy (d14BM) in Acute Myeloid Leukemia (AML)

Introduction: because early assessment of response to intensive induction for AML is believed to be an important predictor of complete remission (CR) and a useful tool to trigger reinduction decisions, d14BM is a widely accepted practice. However, its predictive value is controversial, with some studies suggesting that the risks of a redundant reinduction given to patients whose residual d14 disease is of unclear significance may outweigh its benefits. To clarify this issue we analyzed the influence of d14BM in consecutive newly diagnosed younger (≤60y) patients (pts) receiving first-line intensive induction chemotherapy in our center in the last 18 years.

Methods: induction was idarrubicin + cytarabine (3+7), with 2 cycles allowed if no CR with the first. D14BM was mandatory, and blasts were enumerated by morphology and immunohistochemistry. We evaluated the impact on CR rate and on cumulative incidence of relapse (CIR) of d14BM findings (no blasts, <10% or ≥10%) and of other potentially relevant factors: genetic risk, myelodysplasia- or therapy-related disease (secAML), age (< vs ≥40y) and WBC count at entry (< vs ≥ 50000/ul).

Results: in 374 pts treated from May/98 to May/16, median age 48 (14-60), 55% male, 21% secAML, CR rate was 74%, 14% of CRs being achieved after a 2nd cycle (22% in secAML). Most 2nd cycles were given after d28 or count recovery. CR rate was 80% in pts with no blasts in d14BM, 62% for pts with <10% blasts e 23% for pts with >=10% blasts (p<0.001 for all comparisons). Only 2 other factors were significant for CR: genetic risk (favorable 98% CR, intermediate 78%, unfavorable 33%, p<0.001) and de novo vs secAML (79% vs 51%, p<0.001). In multivariate analysis all 3 factors retained independent significance. There was no significant difference in CIR between pts whose CR was obtained after a d14BM with no blasts, with <10% ou >=10% blasts (p=ns).

Conclusion: in our study d14BM is an independent factor for CR. Our data also clarify 2 other issues: a d14BM with few blasts is associated with a high likelihood of achieving CR, suggesting there may be no benefit in immediate reinduction; and CR duration after a d14BM with blasts is not shorter, contrary to a widespread notion of "cosmetic" CRs in the absence of aplasia.