LAUSR

Mathew 25:29. Increased understanding of the biology of acute myeloid leukemia (AML) has encouraged development of individualized therapy strategies, which focus on identifying which persons are likely to benefit from which interventions [1]. An example is using data from results of measurable residual disease (MRD) testing to predict outcomes and direct therapy [2]. Most data indicate persons with AML in histological complete remission with a negative MRD test have a markedly lower cumulative incidence of relapse (CIR) compared with persons with a positive MRD test [3–10]. This correlation led the European LeukemiaNet (ELN) to include histological complete remission with a negative MRD test as a new response category [11]. The question arises which persons with AML in 1st histological complete remission are appropriate candidates for an allogeneic haematopoietic cell transplant? The ELN AML working party consensus statement, using a dynamic riskassessment approach including results of MRD testing, favors an allotransplant when the estimated leukemia relapse risk is >35–40% and when the projected improvement in diseasefree survival (DFS) is >10% [12]. This is a complex calculus for several reasons some of which are insoluble. For example, estimating leukemia relapse risk at the subject-level even using data, such as results of cytogenetic testing and mutation analyses is imprecise with wide 95% confidence intervals. In many if not most instances these confidence intervals span the landmark point-estimates for leukemia relapse risk and DFS specified in the ELN recommendation. Many co-variates, often confounded, correlate with transplant outcomes such as age, co-morbidities, interval from remission to contemplated transplant, interval antileukemia therapy(ies), donor–recipient relationship and HLA matching, pretransplant conditioning regimen, posttransplant immune suppression. Also, the DFS endpoint does not account for persons with serious transplantrelated complications such as chronic graft-versus-host disease (GvHD). How might results of MRD testing be used to identify persons with AML in 1st histological complete remission are appropriate candidates for an allotransplant? Viewed simplistically and controlling for all other variables, known, unknown, and unknowable, transplants might be best used in persons with a positive rather than negative MRD test. This strategy assumes a transplant will overcome the high CIR associated with a positive MRD test. However, knowing this is true requires data from a randomized trial, in which persons who are MRD test positive are assigned to receive a transplant or not. There are no data from such a trial supporting this hypothesis. In contrast, data from some retrospective analyses, in which multi-parameter flow cytometry was used to test for MRD reported poor transplant outcomes in subjects who were MRD test positive pretransplant compared with those who were MRD-testnegative pretransplant [13]. This was so regardless of the intensity of pretransplant conditioning and donor–recipient type and HLA-matching. In one study, subjects in 1st histological complete remission who were MRD-test-positive pretransplant had a 3-year CIR of 67% like the 3-year CIR of 65% in subjects not in 1st histological complete remission (95% Confidence Intervals [CIs] not reported). Thol * Adriano Venditti [email protected]