Dielectrophoretic Microfluidic Chip Enables Single-Cell Measurements for Multidrug Resistance in Heterogeneous Acute Myeloid Leukemia Patient Samples

The front-line treatment for adult acute myeloid leukemia (AML) is anthracycline-based combination chemotherapy. However, treatment outcomes remain suboptimal with relapses frequently observed. Among the mechanisms of treatment failure is multidrug resistance (MDR) mediated by the ABCB1, ABCC1, and ABCG2 drug-efflux transporters. Although genetic and phenotypic heterogeneity between leukemic blast cells is a well-recognized phenomenon, there remains minimal data on differences in MDR activity at the individual cell level. Specifically, functional assays that can distinguish the variability in MDR activity between individual leukemic blasts are lacking. Here, we outline a new dielectrophoretic (DEP) chip-based assay. This assay permits measurement of drug accumulation in single cells, termed same-single-cell analysis in the accumulation mode (SASCA-A). Initially, the assay was optimized in pretherapy samples from 20 adults with AML whose leukemic blasts had MDR activity against the anthracyline daunorubicin (DNR) tested using multiple MDR inhibitors. Parameters tested were initial drug accumulation, time to achieve signal saturation, fold-increase of DNR accumulation with MDR inhibition, ease of cell trapping, and ease of maintaining the trapped cells stationary. This enabled categorization into leukemic blast cells with MDR activity (MDR+) and leukemic blast cells without MDR activity (MDR–ve). Leukemic blasts could also be distinguished from benign white blood cells (notably these also lacked MDR activity). MDR–ve blasts were observed to be enriched in samples taken from patients who went on to enter complete remission (CR), whereas MDR+ blasts were frequently observed in patients who failed to achieve CR following front-line chemotherapy. However, pronounced variability in functional MDR activity between leukemic blasts was observed, with MDR+ cells not infrequently seen in some patients that went on to achieve CR. Next, we tested MDR activity in two paired AML patient samples. Pretherapy samples taken from patients that achieved CR to front-line chemotherapy were compared with samples taken at time of subsequent relapse. MDR+ cells were frequently observed in leukemic blast cells in both pretherapy and relapsed samples, consistent with MDR as a mechanism of relapse in these patients. We demonstrate the ability of a new DEP microfluidic chip-based assay to identify heterogeneity in MDR activity in leukemic blasts. The test provides a platform for future studies to characterize the mechanistic basis for heterogeneity in MDR activity at the individual cell level.