The role of interim bone marrow assessments in acute myeloid leukemia - A systematic review and meta-analysis.

[BACKGROUND] The current standard of care for acute myeloid leukemia (AML) patients undergoing intensive induction chemotherapy includes an interim bone marrow biopsy around day 14 (D14BM), and reinduction for patients with more than 5 % residual blasts on D14BM (Positive D14BM). However, this approach has become increasingly controversial. This systematic review and meta-analysis assess the sensitivity, specificity, and predictive value of D14BM in patients treated with one cycle of induction chemotherapy and evaluate the efficacy of reinduction versus observation in patients with Positive D14BM.

[METHODS] A systematic literature search was conducted using PubMed, Embase, Web of Science, and Cochrane databases. A bivariate model was used for pooled sensitivity and specificity. The positive predictive value was estimated based on pooled mean sensitivity, specificity, and historical refractory leukemia prevalence. Risk ratios (RR) for complete remission (CR) were meta-analyzed to compare reinduction versus observation in Positive D14BM patients.

[RESULT] Among 1044 identified articles, 12 met inclusion criteria. Ten studies (1683 patients) evaluated the predictive value of D14BM, with a sensitivity of 49.7 % (proportion of refractory cases correctly identified by Positive D14BM) and specificity of 86.2 %. The estimated positive predictive value was 38.9 %, assuming a 15 % prevalence of true refractory leukemia. Eleven studies (832 patients) evaluated efficacy of reinduction, with a pooled RR of 1.00 (61 % vs. 60 %, 95 % CI: 0.76-1.31) for CR.

[CONCLUSION] Positive D14BM has limited predictive power for refractory leukemia. Reinduction based on Positive D14BM does not improve CR rates compared to observation alone only and may expose patients to undue toxicity.