BACKGROUND. Optimal management of acute myeloid leukemia (AML) requires monitoring of treatment response, but minimal residual disease (MRD) may escape detection. We sought to identify distinctive features of AML cells for universal MRD monitoring. METHODS. We compared genome-wide gene expression of AML cells from 157 patients with that of normal myeloblasts. Markers encoded by aberrantly expressed genes, including some previously associated with leukemia stem cells, were studied by flow cytometry in 240 patients with AML and in nonleukemic myeloblasts from 63 bone marrow samples. RESULTS. Twenty-two (CD9, CD18, CD25, CD32, CD44, CD47, CD52, CD54, CD59, CD64, CD68, CD86, CD93, CD96, CD97, CD99, CD123, CD200, CD300a/c, CD366, CD371, and CX3CR1) markers were aberrantly expressed in AML. Leukemia-associated profiles defined by these markers extended to immature CD34+CD38– AML cells; expression remained stable during treatment. The markers yielded MRD measurements matching those of standard methods in 208 samples from 52 patients undergoing chemotherapy and revealed otherwise undetectable MRD. They allowed MRD monitoring in 129 consecutive patients, yielding prognostically significant results. Using a machine-learning algorithm to reduce high-dimensional data sets to 2-dimensional data, the markers allowed a clear visualization of MRD and could detect 1 leukemic cell among more than 100,000 normal cells. CONCLUSION. The markers uncovered in this study allow universal and sensitive monitoring of MRD in AML. In combination with contemporary analytical tools, the markers improve the discrimination between leukemic and normal cells, thus facilitating data interpretation and, hence, the reliability of MRD results. FUNDING. National Cancer Institute (CA60419 and CA21765); American Lebanese Syrian Associated Charities; National Medical Research Council of Singapore (1299/2011); Viva Foundation for Children with Cancer, Children’s Cancer Foundation, Tote Board & Turf Club, and Lee Foundation of Singapore.
Elaine Coustan-Smith, Guangchun Song, Sheila Shurtleff, Allen Eng-Juh Yeoh, Wee Joo Chng, Siew Peng Chen, Jeffrey E. Rubnitz, Ching-Hon Pui, James R. Downing, Dario Campana