RESUMO
Background: With acute myeloid leukemia (AML), there are limited data about the accuracy of day 14 bone marrow (BM) biopsies for predicting complete remission as compared to day 28 BM biopsy results. We here aimed to estimate the correlation between, and the diagnostic accuracy of, both approaches. Materials and Methods: We reviewed 84 patients with AML treated with standard induction chemotherapy to evaluate the remission rate and treatment decisions based on day 14 BM biopsy from 2000-2012. Results: Sixty five patients (77%) demonstrated remission (CR) with less than 5% blasts on their day 14 BM. Thirteen patients (16%) had residual disease (RD), and 6 (7%) were classified as indeterminate response (IR) i.e., blasts 5-20%. Two patients with RD on day 14 underwent re-induction. Out of the 17 remaining cases with RD+IR, 14 (all 6 with IR and 8 out of 11 with residual disease with no re-induction) demonstrated a morphologic complete remission (CR) on day 28 BM. The percentages for complete remissions on days 28 and 14 were significantly different [94% versus 79.3%, respectively; p=0.004, (OR= 0.143, 95% CI: 0.032-0.63)]. Day 14 BM had 82% sensitivity in predicting CR on Day 28; however, it had insufficient specificity (60%) in predicting failure of CR. Conclusions: Induction treatment response assessment based on day 14 BM does not accurately predict the response rate on day 28 and the use of day 14 BM as a sole marker of response to therapy might expose patients to unnecessary interventions.