Label-Free Isolation and mRNA Detection of Circulating Tumor Cells

Nov 4, 2015 - Label-Free Isolation and mRNA Detection of Circulating Tumor Cells from ... Cancer for Disease Diagnosis and Monitoring Therapeutic Effi...
0 downloads 0 Views 4MB Size
Download PDF
Article pubs.acs.org/ac

Label-Free Isolation and mRNA Detection of Circulating Tumor Cells from Patients with Metastatic Lung Cancer for Disease Diagnosis and Monitoring Therapeutic Efficacy Jidong Wang,†,⊥ Wenjing Lu,†,⊥ Chuanhao Tang,‡,⊥ Yi Liu,‡,⊥ Jiashu Sun,*,† Xuan Mu,§ Lin Zhang,§ Bo Dai,∥ Xiaoyan Li,‡ Hailong Zhuo,‡ and Xingyu Jiang† †

Beijing Engineering Research Center for BioNanotechnology & CAS Key Laboratory for Biological Effects of Nanomaterials and Nanosafety, National Center for Nanoscience and Technology, Beijing, 100190, China ‡ Affiliated Hospital of Academy of Military Medical Sciences (307 Hospital), No. 8 Dongdajie, Beijing, 100071, China § Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, 100730, China ∥ Department of Urology, Fudan University Shanghai Cancer Center, Shanghai, 200032, China S Supporting Information *

ABSTRACT: We develop an inertial-based microfluidic cell sorter combined with an integrated membrane filter, allowing for size-based, label-free, and high-efficiency separation and enrichment of circulating tumor cells (CTCs) in whole blood. The cell sorter is composed of a double spiral microchannel that hydrodynamically focuses and separates large CTCs from small blood cells. The focused CTCs with the equilibrium position around the midline of microchannel are further captured and enriched by a membrane filter (pore size of 8 μm) attached at the middle outlet. This integrated microfluidic device can process 1 mL of whole blood containing spiked tumor cells (A549, human lung adenocarcinoma epithelial cell line) within 15 min, with the capture efficiency of 74.4% at the concentration as low as tens of A549 cells per mL of whole blood. This microfluidic cell sorter is further adopted for isolation of CTCs from peripheral blood samples of patients with metastatic lung cancer. The immunostaining and CK-19 mRNA detection are applied for identification of captured CTCs, showing that our method can detect 90% of metastatic lung cancer patients before therapy, whereas the commercially used system can only detect 40% of the same patients. We also use the expression of CK-19 mRNA from captured CTCs as an indicator for monitoring the therapeutic efficiency, which correlates well with X-ray computed tomography (CT) assessment of the disease.

T

biomarker of cancer.7−9 Because of the extremely low number of CTCs (as few as 1 CTC in a background of billions of blood cells),10−12 different technologies have been developed to enrich and detect CTCs by exploiting their physical (including size,13,14 density,15 and deformability16) or biological (specific expression of biomarkers) properties.17−19 The FDA-approval CellSearch system (Veridex) employs anti-EpCAM (epithelial cell adhesion molecule) conjugated magnetic beads for immunomagnetic capture and isolation of CTCs, followed by immunostaining with specific antibodies to confirm and count isolated CTCs.20,21 The CellSearch system is regarded as the gold standard for detecting CTCs in cancer patients with metastatic breast, prostate, or colorectal cancer.22

he spread of cancer, either by metastasis to distant organs or local invasion, could result in death from cancer.1 Early diagnosis and treatment of cancer offers the best hope for cure, as well as significantly improving the survival rates.2−4 For example, as a main cause of cancer-related death worldwide, lung cancer frequently metastasizes to brain, bone, and liver, with a 5-year survival rate of lower than 15%.5 For lung cancer diagnosed at an early stage, surgical resection can improve the 5-year survival rate up to 60−80%.2 However, a major challenge in early diagnosis and screening of cancer is the lack of early clinical symptoms and the predictive biomarkers.6 Circulating tumor cells (CTCs), shed from a primary tumor and circulating in the bloodstream, might be responsible for cancer metastasis and tumor recurrence.1,7 In contrast to invasive tissue biopsies that may impose high risk to patients, CTCs detection from peripheral blood is considered as the realtime “liquid biopsy” in a noninvasive manner. CTCs show great promise for assessing prognosis of cancer, monitoring the therapeutic treatment, as well as serving as a surrogate © XXXX American Chemical Society

Received: September 13, 2015 Accepted: November 4, 2015

A

DOI: 10.1021/acs.analchem.5b03484 Anal. Chem. XXXX, XXX, XXX−XXX

Analytical Chemistry

■

Article

EXPERIMENTAL SECTION Design and Assembly of Integrated Cell Sorter. The inertial-based microfluidic cell sorter is composed of a six-loop double spiral microchannel with one sample inlet and three outlets (Figure 1 and Figure S1). The cross-section of the

Despite its sensitivity and reproducibility, the CellSearch system highly relies on the surface expression of EpCAM on CTCs and requires advanced instrumentation. We should note that a population of CTCs may undergo the epithelialmesenchymal transition (EMT) process to invade surrounding tissues and trigger distant metastasis.23 The EMT process results in the loss of epithelial markers (such as EpCAM and cytokeratins), and CTCs with EMT phenotype could not be enriched or detected efficiently by the CellSearch system.24,25 For example, CTCs from lung cancer patients, especially nonsmall cell lung cancer (NSCLC, approximately 70−80% of incidents), are unlikely to be detected by the Cell-Search system.26 Although aptamers with high affinity and selectivity could be considered as alternative to antibodies,27,28 the isolation of CTCs by aptamers still depends on affinity interactions, so that CTCs with EMT phenotype may not be recognized by aptamers. Therefore, novel enrichment and detection strategies are urgently required for diagnosis of CTCs in lung cancer patients or CTCs undergoing EMT. Size-based separation of CTCs from whole blood is labelfree, rapid, cost-effective, and robust.29−32 Among different approaches for size-based CTCs separation, microfiltration is one of the most straightforward methods, but sometimes is subjected to the complicated device fabrication, membrane clogging, and low capture purity.33 Other size-dependent cell separation methods such as dielectrophoresis (DEP), acoustics, and optics always involve the integration of external force fields and the complicated experimental setup.34,35 To achieve a continuous high-throughput size-based CTCs isolation, inertial microfluidics have been proved to be a powerful tool that solely relies on the hydrodynamic effects.36−39 Our group previously developed a curved microfluidic cell sorter that allows for a rapid and continuous rare tumor cell separation based on inertial effects in a label-free manner.40,41 However, the separated CTCs by inertial microfluidics are still suspended in diluted blood plasma, resulting in a low CTCs enrichment. In this work, we combine the inertial-based microfluidic cell sorter with an integrated membrane filter at the outlet of device, allowing for the size-based, label-free, and highefficiency separation and enrichment of CTCs in whole blood from patients with metastatic lung cancer. The cell sorter is composed of a double spiral microchannel with one inlet and three outlets, in which the large CTCs (most CTCs are larger than 15 μm in diameter) in lysed whole blood are gradually focused into a thin stream and flowing out from the middle outlet under the hydrodynamic effect, whereas the small white blood cells (WBCs, most WBCs are in the range of 7−12 μm) are converged and out from the inner outlet, to achieve a sizebased CTCs separation from lysed blood. To further enrich the separated CTCs, a membrane filter of pore size of 8 μm is glued onto the surface of middle outlet and the large CTCs can be captured by a filter. The isolated and enriched CTCs on the membrane filter could be identified either by immunostaining with specific antibodies or Cytokeratin 19 (CK-19) mRNA amplification. The immunofluorescence staining can differentiate the CTCs or CTC clusters from white blood cells (WBC). The expression of CK-19 mRNA from CTCs is used for clinical monitoring of therapeutic efficacy of lung cancer and compared with X-ray computed tomography (CT) assessment of the disease.

Figure 1. Schematic of the microfluidic cell sorter with an integrated membrane filter for size-based CTCs isolation and detection. The lysed whole blood sample containing rare CTCs is injected into the double-spiral cell sorter at 25 mL/h. The large CTCs are focused and deflected into the middle outlet, while small lysed red blood cells (RBCs) and white blood cells (WBCs) are removed from the side outlets. A membrane filter with pore size of 8 μm attached on the middle outlet is used to capture and enrich the separated CTCs. The immunostaining and CK-19 mRNA detection are applied for identification of captured CTCs.

double spiral channel is 300 μm wide × 40 μm high, and the space between two adjacent loops is 450 μm. The diameter of the outermost circle is 18 mm, and the total length of the curved channel is 334 mm. The trifurcated outlets are termed as the inner outlet (the outlet closest to double spiral channel), the middle outlet, and the outer outlet (the outlet farthest to the double spiral channel). The widths of the inner, middle, and outer outlets are 60 μm, 115 μm, and 125 μm, respectively. The details for device fabrication can be found in the Supporting Information. Sample Preparation. A human cell line (A549, human lung adenocarcinoma epithelial cell) is grown in Dulbecco’s modified Eagle’s medium (DMEM) with 10% fetal bovine serum (FBS) and 1% penicillin/streptomycin at 37 °C, 5% CO2 inside an incubator (Thermo Scientific). The human whole blood samples are drawn from healthy volunteers and collected into 3 mL ethylenediaminetetraacetic acid (EDTA) vacuum tubes. To mimic the low CTCs event in blood, we spike a small number of A549 cells into healthy blood samples (More details in the Supporting Information). The blood samples from patients with metastatic lung cancer are received from the Affiliated Hospital of Academy of Military Medical Sciences (307 hospitals) and Peking Union Medical College Hospital after obtaining the patient-informed consent. For patient samples, 2−4 mL of whole blood is lysed by 10−20 mL of lysis buffer (Figure 2a) and directly processed by the microfluidic cell sorter (more details in the Supporting Information). A total of 34 blood samples from lung cancer patients are analyzed by our method, while CTCs from 13 patient samples are also detected by the CellSearch system for a side-by-side comparison. To demonstrate the possibility of CTCs as a biomarker for cancer prognosis, we monitor the B

DOI: 10.1021/acs.analchem.5b03484 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

4′,6-diamidino-2-phenylindole (DAPI). After immunostaining, the membrane with attached cells is imaged using the laser scanning confocal microscope (Zeiss LSM710). LAMP of CK-19 mRNA from Captured CTCs. Loopmediated isothermal amplification (LAMP) is adopted for detection of CK-19 mRNA from captured CTCs.42,43 The intermediate filament protein CK-19 is a specific epithelial marker, and the detection of CK-19 mRNA can severe as an index of CTCs in blood, bone marrow, and epithelial cancers.44 LAMP primers targeting CK-19 gene are designed using Primer Explorer version 3 (Table S1).45 The reagents and procedures for LAMP are in the Supporting Information.

■

RESULTS AND DISCUSSION Size-Based Separation and Enrichment of CTCs. The cell separation inside the double spiral microfluidic device is based on the competition between inertial lift and Dean drag force acting on cells, giving rise to the distinct equilibrium positions for different sized cells.46,47 Our previous studies demonstrate that the double spiral microchannel allows for the focusing of CTCs and blood cells into narrow bands in the first spiral, while the focused large tumor cells and small blood cells are separated in the second spiral.40,41 We should note that this size-based cell sorter can only separate diluted blood samples (20−50× diluted blood) to avoid the clogging of the device, resulting in the prolonged processing time and a relatively low enrichment factor of CTCs. To significantly enhance the enrichment factor and to reduce the processing time, we propose a membrane filter-integrated microfluidic sorter to isolate CTCs from lysed whole blood. To investigate the performance of the integrated cell sorter, we first monitor the trajectories of fluorescently labeled A549 cells spiked into the lysed blood or 30× diluted blood at a flow rate of 25 mL/h (Figure 2). Figure 2b displays the superimposed trajectories of A549 cells (red stream), WBCs (green stream), and lysed RBCs at the trifurcated outlet with the membrane filter at the middle outlet. The large A549 cells tend to remain at the midline of the double spiral microchannel, while the relatively small WBCs are more easily influenced by the Dean flow inside the curved channel and migrate toward the inner wall, leading to a nearly complete separation of A549 cells and WBCs (Figure 2b). We cannot visualize the trajectories of lysed RBCs because the small fragments of lysed RBCs may get entrained in the Dean vortices and spread over a wide region of the cross-section. For separation of A549 cells and diluted blood, we find that most of the focused 30× diluted blood (dark stream under optical microscope) is reflected to the inner outlet, whereas the A549 cells are still ordered and out from the middle outlet (Figure 2c). The characterization of the position and distribution of A549 cells, lysed blood, and 30× diluted blood indicate that the large tumor cells could be rapidly focused into a thin stream and oscillate around the midline of microchannel at the flow rate of 25 mL/h, regardless in the lysed blood or diluted blood (Figure 2d,e). To characterize the capture efficiency of the integrated microfluidic device, we spike an accurate number of fluorescently labeled A549 cells into 1 mL of whole blood from healthy donors. After mixing the 1 mL blood with 5 mL of lysis buffer, the lysed blood sample of 6 mL is introduced into the cell sorter at 25 mL/h. By counting the number of isolated A549 cells from lysed blood on the membrane filter at the middle outlet under a microscope, we can evaluate the capture

Figure 2. (a) Photography of the whole blood, lysed blood and 30 times (30×) diluted blood. (b) Trajectories of spiked A549 cells in the lysed blood at 25 mL/h. The fluorescent red band in the middle outlet is the focused A549 cells, and the fluorescent green band in the inner outlet is the focused WBCs. (c) Trajectories of spiked A549 cells in 30× diluted blood. The dark band in the inner outlet is the focused 30× diluted blood. (d) Trajectories of A549 cells (fluorescent red) and lysed blood (bright field) in each loop along the 6-loop double spiral microchannel. Scale bar is 300 μm. (e) Trajectories of A549 cells (fluorescent red) and 30× diluted blood (bright field) inside double spiral microchannel. The dashed blue box indicates the outlet of the double spiral channel.

gene expression and number of CTCs from 3 lung cancer patients before and after receiving the therapy using our method and the CellSearch system. CTCs Separation and Enrichment. For separation of CTCs or A549 cells from diluted or lysed blood, a syringe containing 6 mL of sample is loaded into a syringe pump (Harvard Apparatus) and connected to the integrated microfluidic cell sorter using the plastic tube. The syringe pump injects the sample into the microfluidic device at a constant flow rate of 25 mL/h. To avoid cell sedimentation, the syringe pump is placed vertically. A maximum throughput of 10 samples is achieved using a pump module (Figure S2). The separated CTCs or A549 cells are enriched and captured by the membrane filter glued at the middle outlet. The filter with cells attached on the surface can be carefully peeled off from the outlet using a sharp-tipped tweezer, allowing for the further analysis of enriched CTCs by immunostaining or gene detection. To visualize the cell trajectories inside the microchannel, the device is mounted onto the stage of a Leica DMI 6000 microscope (Leica Microsystems). Optical and fluorescent microscopic images are captured by a CCD camera. The superimposed images are obtained using Leica Application Suite (Leica Microsystems). Immunostaining of Captured CTCs. Immunofluorescence staining is performed to confirm and distinguish the CTCs from similar sized WBCs, based on the different expression of specific proteins between CTCs and WBCs. For cell immunostaining, we first remove the filter membrane from the middle outlet of microfluidic device. The cells attached on the flat membrane are fixed with 4% paraformaldehyde and subsequently permeabilized with 0.2% Triton X-100 in PBS. The fixed cells are immunofluorescently stained with monoclonal fluorescein isothiocyanate (FITC)labeled antibody against cytokeratin 19 (CK-19, ab7754, abcam), rabbit antihumananti-CD45 (ab10558, abcam), and C

DOI: 10.1021/acs.analchem.5b03484 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

μm, and previous study indicates that the CTC clusters show higher metastatic propensity than single CTCs.48 Our sizebased cell sorter can thus isolate single CTCs as well as CTC clusters with much larger sizes from clinical blood samples. CK-19 mRNA Detection. To precisely identify the captured CTCs by a microfluidic device, loop mediated isothermal amplification (LAMP) is employed to quantify CK-19 mRNA inside CTCs. Although the commonly used immunofluorescence staining can identify and quantify the CTCs after cell isolation, this method suffers from several issues: (1) complicated and time-consuming procedures. The immunostaining requires several steps of incubation and washing. The whole procedure takes 4 to 5 h to yield the final immunostaining result. (2) Elaborate enumeration of CTCs. The count and differentiation of the immunostained CTCs may be subjective and laborious. Different pathologists may get different counts of CTCs for the same clinical sample. Cytokeratin 19 (CK-19) is a marker of epithelial cells and has been used in the FDA-approval CellSearch system to stain the isolated CTCs. Although the polymerase chain reaction (PCR) and other methods have been applied to detect CK-19 mRNA from CTCs in blood, the false negative rate is high, probably due to the excess WBCs in blood and the relatively low sensitivity of PCR. In comparison, LAMP is a promising method for CK-19 mRNA detection because of its high specificity and sensitivity. As our integrated microfluidic device is capable of removing WBCs from CTCs, the combination of LAMP and size-based CTCs separation may provide an attractive method for detecting CK-19 mRNA from captured CTCs in clinical samples. To determine the detection sensitivity of the LAMP reaction, we prepare five samples with different plasmid concentrations from 2.3 × 102 to 2.3 × 106 copies by serial dilutions of constructed plasmids containing the target nucleic acid fragments of CK-19. The detailed sequences of CK-19 are shown in Figure S5, and the sets of primers for LAMP are in Table S1. The amplification time of plasmids containing target CK-19 is linearly proportional to the log of concentrations of plasmids (Figure S6). The amplification time decreases from 56 to 20 min, with the increased plasmid concentrations from 2.3 × 102 to 2.3 × 106 copies. To demonstrate the CK-19 mRNA detection in tumor cells from blood samples, we spike different amounts of A549 cells into 1 mL whole blood from healthy donors and perform the size-based cell separation by integrated microfluidic chip and gene detection by LAMP. The real time LAMP results indicate that we can detect as few as 34 A549 cells from 1 mL of whole blood, with the amplification time (also termed as the time over threshold) of 52 min (Figure 4). As one cancer cell contains approximately 100 to 1000 copies of CK-19 mRNA, the limit of detection is around 3 × 103 to 3 × 104 copies of mRNA from captured cells by our method. With the number of spiked A549 cells increasing to 84 and 1000, the amplification time decreases to 43 and 38.5 min (Figure 4). In comparison, CK-19 mRNA remains undetectable in 1 mL of healthy whole blood without spiked A549 cells (Figure 4). We note that there is no linear relationship between the amplification time and the number of spiked A549 cells because of the complex constitutions of tumor cells and the loss of RNA during extraction. However, the tendency of increased amplification time with decreased numbers of spiked A549 cells is clearly observed. CK-19 mRNA Detection from Patients with Metastatic Lung Cancer. To demonstrate the utility of the present

efficiency. After performing three individual experiments, the numbers of captured A549 cells on the membrane filter are 72, 101, and 144 with the initial spiked numbers of 103, 140, and 177, respectively. The capture efficiency of A549 cells by the integrated device is thus determined to be 74.5 ± 6.1%. We further estimate the enrichment factor of A549 cells by the microfluidic device. The initial tumor-to-blood cells ratio is approximately of 2 × 10−8, and the ratio increases to ∼7.5 × 10−2 after separation and enrichment, indicating the enrichment factor of 3.75 × 106. The isolated A549 cells captured by the membrane filter after separation could maintain intact morphology (Figure S3), which proliferate well under standard culture conditions (Figure S4). The high viability of separated A549 cells suggests that we can perform immunostaining or gene analysis to further interrogate the captured CTCs in whole blood from lung cancer patients. Immunofluorescence Staining. After the separation and enrichment of CTCs from lung cancer patients, we use the immunostaining method for CTCs analysis. Since the inertialbased cell separation is solely size-dependent, the large WBCs with sizes similar to CTCs may be captured by a membrane filter as well. The immunostaining could differentiate WBCs and CTCs. The criteria of CTCs determination is the DAPI positive, cytokeratin (CK) positive, and CD45 (a leukocyte marker) negative after immunofluorescence staining (Figure 3e−h). The characteristics of WBCs is DAPI positive, CK

Figure 3. Immunofluorescence staining of the captured cells attached on the membrane filter from the lung cancer patient samples. Threecolor immunostaining allows for the identification of the WBCs (a−d), CTCs (e−h), and CTC clusters (i−l) using DAPI (nuclear specific, blue), anticytokeratin antibody (marker for CTCs, green), and antiCD45 antibody (marker for WBCs, red).

negative, and CD45 positive (Figure 3a−d). In addition, the morphology of CTCs is consistent with the malignant cancer cells, indicated by the large size and the high nuclear to cytoplasmic ratio (Figure 3e−h). We also find the CTC clusters in blood samples from the advanced lung cancer patient after size-based CTCs separation and immunofluorescence staining (Figure 3i−l). The size of CTC clusters is about 30 μm × 50 D

DOI: 10.1021/acs.analchem.5b03484 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

fluidic cell separation and gene amplification exhibits an excellent specificity and sensitivity for CTCs detection from patients with metastatic lung cancer. We also compare our method with the CellSearch system for the detection of CTCs from patients diagnosed with NSCLC and small cell lung cancer (SCLC). For four patients with SCLC before therapy, the CellSearch system can detect 75% of patients, with the optimal cutoff value of 5 CTCs in 7.5 mL of blood sample, whereas the combination method of size-based CTCs separation and CK-19 mRNA amplification can detect 100% of patients (Table 1). For six patients with NSCLC before therapy, the detection efficiency by the CellSearch system is only 16.7% (Table 1), mainly because the CTCs in NSCLC tend to lose the epithelial markers during the EMT process. The CellSearch system highly replies on these surface markers to capture the CTCs. In contrast, our method can detect 83.3% of NSCLC patients (Table 1), as a result of efficient size-based separation of large CTCs from blood cells and highly-sensitive detection of CK-19 mRNA from captured CTCs. In general, the detection efficiency of the CellSearch system is 40%, whereas that of our method is 90%. Therefore, our method shows a better efficacy in detecting CTCs from lung cancer samples, especially NSCLC, compared with the CellSearch system. CTCs Monitoring during the Therapy. CTCs could be employed as a surrogate biomarker to monitor the therapy efficacy of lung cancer. Here we quantify the expression of CK19 mRNA from captured CTCs before and after therapy using the integrated microfluidic cell sorter and the following gene amplification. The detection of CTCs by our method is compared with the results from the CellSearch system. The correlation between the CTCs and the tumor burden before and after the therapy is evaluated by comparing the quantity of CTCs with X-ray computed tomography (CT) scans. Three patients diagnosed with SCLC or NSCLC, and treated by tyrosine kinase inhibitors (TKIs) or chemotherapy, are subjected to the CTC enumeration. Patient 5 in Table 1 is a 62 year-old male NSCLC patient that received TKI treatment, whose epidermal growth factor receptor (EGFR) shows deletion in exon 19. Before TKI treatment, the amplification time of CK-19 mRNA from

Figure 4. CK-19 mRNA detection of different amounts of A549 cells spiked into 1 mL of whole blood from healthy donors, after size-based separation inside the integrated microfluidic chip. (a) Real time LAMP curves of different amounts of spiked A549 cells from 34 to 1000. NC is the negative control with no A549 cells spiked into whole blood. (b) The amplification time of CK-19 mRNA from different amounts of A549 cells.

method in clinical diagnosis, we combine the microfluidic sizebased cell separation and LAMP-based CK-19 mRNA detection for CTCs identification in whole blood from patients with metastatic lung cancer. A total of 21 nonsmall cell lung cancer (NSCLC) samples and 12 healthy samples are tested in a single-blind manner (Figure 5). A 2 to 4 mL of blood sample is first portioned into 2 to 4 tubes, with each tube containing 1 mL of blood. The 1 mL of whole blood is lysed by 5 mL of lysis buffer, with the final sample volume of 6 mL, which is directly introduced into the integrated microfluidic device at 25 mL/h. For processing 2 to 4 mL of whole blood, we need 2 to 4 individual microfluidic devices, and the processing time for each device is 14.4 min to efficiently avoid cell sedimentation. Since each sample has 2 to 4 amplification curves for CK-19 mRNA after CTCs separation and enrichment, we choose the curve of the fastest amplification time (also termed as time over threshold) to represent the amplification time for CK-19 mRNA detection. For 21 samples from lung cancer, the amplification time for CK-19 mRNA in captured CTCs ranges from 34 to 51 min, and the average time is 41.3 ± 5.2 min (Figure 5). In comparison, none of the 12 healthy samples shows positive signal of CK-19 mRNA after 60 min amplification. Therefore, our method combined the micro-

Figure 5. Statistics of amplification time (time over threshold) of CK-19 mRNA from captured CTCs in NSCLC patient samples and healthy samples. *** represents P < 0.001. E

DOI: 10.1021/acs.analchem.5b03484 Anal. Chem. XXXX, XXX, XXX−XXX

Article

Analytical Chemistry

Table 1. Comparison of CTCs Detection from Lung Cancer Patient Samples by CellSearch System and Our Method Combined the Microfluidic Size-Based CTCs Separation and LAMP-Based Gene Detection patient number 1 2 3 4 5(BT)a 5(AT)b 6(BT)a 6(AT)b 7 8 9 10(BT)a 10(AT)b

age

sex

64 65 79 46 62

c

histologic features

staging

M Fd Fd Mc Mc

e

NSCLC (Adeno) NSCLCe(Adeno) NSCLCe(Adeno) NSCLCe(Squa) NSCLCe(Adeno)

IV IV IV Ia IV

61

Fd

NSCLCe(Adeno)

IV

73 75 55 42

Mc Mc Mc Fd

SCLCf SCLCf SCLCf SCLCf

extensive limited stage extensive limited stage

CellSearch no. per 7.5 mL

CTCs-LAMP (min)

0 4 3 9 1 0 2 0 132 29 20 0 1

36 51 NDg 39 51 NDg 40 ND 37.5 41 38 36.5 42

a

BT: before therapy. bAT: after therapy. cM: male. dF: female. eNSCLC: nonsmall cell lung cancer. fSCLC: small cell lung cancer. gND: not detected.

Figure 6. Monitoring therapeutic effects of patients with metastatic lung cancer. (a) X-ray computed tomography (CT) scans and CK-19 mRNA amplification from captured CTCs by our method for a NSCLC patient that received TKI treatment. The tumor burden on the lung is decreased by 90% after treatment. (b) CT and CK-19 mRNA detection for another NSCLC patient that received TKI treatment. Some of the brain metastases disappear after treatment. (c) CT and CK-19 mRNA detection for a SCLC patient that received chemotherapy. After therapy, a small decrease of the tumor burden (