Novel microfluidic device for measurable residual disease detection in acute leukemia
Ziwei Cai,Shuang Fan,Xiaoyi Sun,Xiaodong Mo,Gen Yang
DOI: https://doi.org/10.1016/j.xinn.2023.100408
2023-01-01
Innovation
Abstract:Relapse is the most important cause of treatment failure in acute leukemia (AL). Thus, how to predict relapse is critical for improving the survival of patients with AL. Measurable residual diseases (MRDs; previously termed minimal residual diseases), referring to the presence of remaining leukemia cells after the declaration of complete remission (CR) detected by morphological analysis, is the most important biomarker for relapse prediction.1Mo X.D. Lv M. Huang X.J. Preventing relapse after haematopoietic stem cell transplantation for acute leukaemia: the role of post-transplantation minimal residual disease (MRD) monitoring and MRD-directed intervention.Br. J. Haematol. 2017; 179: 184-197Crossref PubMed Scopus (34) Google Scholar Several methods, including multiparameter flow cytometry (MFC), real-time quantitative polymerase chain reaction (qPCR), digital PCR (dPCR), and next-generation sequencing (NGS), are used to monitor MRD after treatment, reaching a sensitivity of 10−4 to 10−6. However, all of the methods for MRD monitoring have some limitations. The change in immunophenotype may lead to false negatives in MFC. For qPCR, only one-third of patients have stable, leukemia-specific molecular markers for qPCR. Pan-leukemia molecular markers (e.g., WT1 and immunoglobulin H [IgH]/T cell receptor [TCR] rearrangements) can be used as MRD marker for patients without leukemia-specific molecular markers. However, the increase of WT1 may also be a normal change during immune reconstitution. Besides, dPCR and NGS are too expensive to be used for regular MRD monitoring. Lastly, most of these methods use bone marrow (BM) samples, and monitoring frequently would increase the discomfort of patients.2Heuser M. Freeman S.D. Ossenkoppele G.J. et al.2021 Update on MRD in acute myeloid leukemia: a consensus document from the European LeukemiaNet MRD Working Party.Blood. 2021; 138: 2753-2767Crossref PubMed Scopus (172) Google Scholar Thus, there is an urgent need for establishing a new MRD detection methodology with high sensitivity and specificity, and with a standardized and easy-to-use protocol, without increasing the discomfort of patients. Considering that some physical parameters of leukemia cells are stable and specific, using physical methods to enrich as well as distinguish them from normal peripheral blood cells may provide a new method for MRD monitoring. A previous report indicated that increasing the concentration of tumor cells in a peripheral blood sample (i.e., circulating tumor cells [CTCs]) through enrichment and separation equipment can raise the sensitivity of MRD detection by at least 10 times,3Han J. Lu C. Shen M. et al.Fast, reusable, cell uniformly distributed membrane filtration device for separation of circulating tumor cells.ACS Omega. 2022; 7: 20761-20767Crossref PubMed Scopus (2) Google Scholar which may also help to detect the trace of leukemia cells in peripheral blood. Current CTC capture devices mainly differentiate CTCs from normal cells based on their physical or biological properties, which both have some limitations. Devices relying on the biological property differences mostly use antibodies, such as EpCAM and CK, to capture CTCs specifically. This method requires a long processing time and a high cost and may also miss CTCs that do not express those biomarkers. For the physical property, the most employed method is to separate CTCs by size and deformability using filtering, density gradient centrifugation, and inertial focusing. However, the purity and efficiency are still unsatisfactory.4Pantel K. Alix-Panabières C. Liquid biopsy and minimal residual disease — latest advances and implications for cure.Nat. Rev. Clin. Oncol. 2019; 16: 409-424Crossref PubMed Scopus (559) Google Scholar Among them, microfluidic devices are increasingly used in CTC separation due to their micron size, enabling precise object manipulation and single-cell studies. We designed a microfluidic device (World Intellectual Property Organization patent WO2021013066A1) that can capture CTCs in a specific size range with high efficiency5Lu C. Xu J. Han J. et al.A novel microfluidic device integrating focus-separation speed reduction design and trap arrays for high-throughput capture of circulating tumor cells.Lab Chip. 2020; 20: 4094-4105Crossref PubMed Google Scholar and utilized it to improve the sensitivity of MRD detection. Compared with other CTC separation technologies based on physical properties, this device integrates the principle of filtration and hydrodynamic methods and shows several superiorities: (1) it avoids the problem of easy blockage and the dilemma of inversely correlated flow rate and capture efficiency of simple filtration devices, (2) CTC can be efficiently captured in a wide range of flow rates, and (3) high hemodilution ratios are not required, resulting in higher effective flow rates. The core functional unit of this device is the focus-separation structure, as shown in Figure 1A. It consists of several groups of channels called focus structure and two separation channels called separation structures. In the focus structure, branch channels of different lengths create different flow resistance and separation ratios, which keep cells above a certain size (i.e., CTCs) in the center channel and cause them to be constantly pushed toward the center of the central channel by the side walls. After several focus structures, CTCs are concentrated in the middle of the channel, while only cells below a threshold size can be found in the edge region. Then, the separation channels with a specific flow resistance separate the liquid without the target cells and thus achieve the effect of increasing CTC concentration. A chip includes 4–5 focus-separation structures, each one separating about 50% of the liquid, so the processed liquid volume approximately drops by two orders of magnitude compared with the initial sample, while the number of CTCs remains the same. At the rear of the focus-separation structure, a filter structure is optional, and the flow rate at the filter structure will decrease to 10−2 of the inlet flow rate, reducing the probability of CTCs flowing away due to excessive flow rate. As a large proportion of blood cells are separated by the separation channel, the possibility of filter blockage is greatly reduced as well. To verify the function of the chip, uniformly sized fluorescent microspheres were mixed into phosphate buffer before being injected into the chip at a flow rate of 20 mL/h, and the streamlines in Figure 1B were observed under a microscope. The capture efficiencies were calculated to be more than 90% in the flow rate range of 6 to 40 mL/h. Experiments have also been carried out, showing that the removal ratio of cells with sizes just outside the critical range is higher than 90%. The chip has been used in the CTC detection of solid tumors, and the results of the double-blind experiments showed that the concentration of CTCs in the blood of a patient with cancer ranged from 6 to 117 cells/mL, which was significantly different from that of healthy volunteers (2–3 cells/mL).5Lu C. Xu J. Han J. et al.A novel microfluidic device integrating focus-separation speed reduction design and trap arrays for high-throughput capture of circulating tumor cells.Lab Chip. 2020; 20: 4094-4105Crossref PubMed Google Scholar Since the size distribution range of leukemia cells and mature blood cells in peripheral blood has a large overlap, the premise of using this device for residual leukemia cell detection is to select an appropriate critical value to screen cells in the blood: the upper and lower limits, denoted as rmax and rmin, respectively. Having determined rmax (18 μm) and rmin (16 μm) by experiment and calculation, the diluted patient sample is injected into the chip to separate cells larger than rmax, and then the collected liquid from the waste port is injected into another chip to separate cells larger than rmin. After these two steps, the concentration of cells with a specific size increased by two orders of magnitude, and other blood cells were removed at levels of 4–5 orders of magnitude, as shown in the schematics of Figures 1C and 1D. If the sensitivity of simple morphological detection is 1:20, then the sensitivity after the CTC enrichment process could promisingly reach 1:20,000, which is comparable with, or even better than, MFC and qPCR. As for the expense of this MRD detection method, it cost about 30 yuan in chips and about 10–20 yuan in pretreatment and staining for each patient blood experiment. The overall cost of a test is less than 50 yuan, which is far lower than the cost of MFC or PCR. In conclusion, our CTC enrichment separation device provides a new option for MRD detection. Based on the physical characteristics of leukemia cells, it is unaffected by the heterogeneity of the immunophenotype or other molecular biomarkers and could be used in nearly all patients with AL. Due to the sensitivity up to 10−5, our method seemed to be as sensitive as qPCR and could be considered as the potential alternative biological marker for WT1 or IgH/TCR rearrangements in patients without the leukemia fusion gene. It has a simple operation and a low cost and could replace BM with peripheral blood samples, which reduces the patient’s discomfort. All these advantages make this method have great potential for popularization. In the future, combined with automated sample injection, hyperspectral imaging, and artificial intelligence recognition, an automatic and standardized MRD detection method may come true, which can be popularized in primary hospitals easily. This work was supported by the National Key Research and Development Plan of China (grant numbers 2019YFF01014402 and 2022YFC2502606); the CAMS Innovation Fund for Medical Sciences (CIFMS) (grant numbers 2022-I2M-C&T-B-121 and 2019-I2M-5-034); the Program of the National Natural Science Foundation of China (grant numbers 11875079 and 82170208); Peking University People’s Hospital Scientific Research Development Funds (grant number RZ2022-02); and the State Key Laboratory of Nuclear Physics and Technology, PKU (grant numbers NPT2020KFY19 and NPT2020KFJ04). The authors declare no competing interests.