Naval G. Daver, MD, from the MD Anderson Cancer Center, Houston, TX, discusses biomarkers which may help to predict the response to immune checkpoint inhibition in acute myeloid leukemia (AML) at the International Symposium on Acute Leukemias (ISAL) 2017 in Munich, Germany. He explains that the best data for baseline biomarkers has been seen in the study of azacitidine and nivolumab treatment in AML salvage patients who did not respond to standard treatment (NCT02397720). In all patients who responded to the combination immune checkpoint inhibitor treatment, a higher T-cell infiltrate into the bone marrow was seen at baseline. Particularly, the patients with a higher ratio at baseline of CD4 and CD8 effector T-cells to regulatory T-cells, a favorable immune profile usually associated with a good tumor outcome, responded much better than those who had few total T-cell or an unfavorable ratio of CD8 effector T-cells to regulatory T-cells. Dr Daver explains that it has been shown in lung cancer, kidney cancer and bladder cancer that patients who already have a favorable infiltrate in the tumor tissue more readily respond to manipulations, as the treatment only needs to activate T-cells, and not recruit them as well, and suggests that the presence and immune profile of T-cells in the bone marrow could be a useful biomarker in predicting the response to immune checkpoint inhibitor treatment in AML. Other biomarkers being studied include PD-1/PD-L1 and CTLA-4 at baseline, however no clear correlation has been found between T-cell PD-1/PD-L1 expression and response to immune checkpoint inhibitor treatment. However, the ratio of CD-8 T-cells expressing PD-1/PD-L1 to regulatory T-cells expressing PD-1/PD-L1, indicative of a greater number of activated T-cells expressing PD-1/PD-L1, significantly corresponded to a positive response. Dr Daver speculates that in the next few years, as currently ongoing clinical trials develop into larger multi-center studies, the scenario will be similar to that currently seen in solid tumors, with 30 – 35% of patients showing a durable response. He hopes that in future, patients can be selected for immune checkpoint inhibitor treatment by the addition of simple markers such as CD3, CD4, CD8 and FOXP3 into flow cytometry panels, which will be studied in more detail in upcoming studies.