Purpose: The use of circulating tumor cells (CTC) as “liquid biopsy” is limited by the very low yield of CTCs available for subsequent analyses. Most in vitro approaches rely on small sample volumes (5–10 mL).

Experimental Design: Here, we used a novel approach, the GILUPI CellCollector, which enables an in vivo isolation of CTCs from peripheral blood. In total, 50 lung cancer patients were screened in two subsequent device applications before and after therapy (n = 185 applications).

Results: By in vivo isolation, 58% (108/185) of the patients were positive for ≥1 CTC (median, 5 CTCs; range, 1–56 cells) as compared with 27% (23/84; range, 1–300 cells) using the FDA-cleared CellSearch system. Furthermore, we could show that treatment response during therapy was associated with significant decreases in CTC counts (P = 0.001). By dPCR, mutations in the KRAS and EGFR genes relevant for treatment decisions could be detected in CTCs captured by in vivo isolation and confirmed in the primary tumors of the same patients.

Conclusions:In vivo isolation of CTCs overcomes blood volume limitations of other approaches, which might help to implement CTC-based “liquid biopsies” into clinical decision making. Clin Cancer Res; 22(9); 2197–206. ©2015 AACR.

Translational Relevance

The CellCollector used in this study enables an in vivo isolation of circulating tumor cells (CTC) directly from the arm vein of cancer patients. In this study, 50 lung cancer patients were screened in two subsequent device applications before and after therapy (n = 185 applications). This device captured CTCs from lung cancer patients more frequently than the FDA-cleared CellSearch system. Our findings indicate that in vivo isolation of CTCs is a promising approach to isolate CTCs in sufficient amounts for treatment monitoring and further downstream molecular analyses in lung cancer patients. This new approach can now be tested in interventional clinical studies where the number and/or characteristics of CTCs will be used to stratify patients to particular treatments and to assess established endpoints, including survival.

Although tumor cells detached from primary tumors or metastases into the blood (circulating tumor cells; CTC) have been discovered almost 150 years ago, specific detection and molecular examination have only recently gained heightened attention. This may be explained by the fact that technology improvements have only recently become available to enable more consistent CTC isolation. However, despite encouraging research progress, it is still a tremendous challenge to enrich sufficient amounts of CTCs from peripheral blood of cancer patients (1, 2). Detection and molecular characterization of CTCs is challenging because CTCs are rare events, with a frequency of approximately 1 to 10 CTCs in the background of 6 × 106 leukocytes, 2 × 108 platelets and 4 × 109 erythrocytes per mL of patient-derived blood sample (3). Thus, despite promising results demonstrating that enumeration and characterization of CTCs might provide important information on the individual risk of cancer patients, there is still an urgent demand to increase the yield of capturing CTCs.

Lung cancer is one of the most frequently occurring and deadly diseases (4). A sensitive CTC isolation method holds great promise for early detection of minimal residual disease (MRD). Furthermore, CTCs are being evaluated as predictive biomarkers to guide individual cancer treatment strategies (personalized medicine). Hence, sensitive isolation and profound molecular characterization of CTCs could serve as a “liquid biopsy” and facilitate improved individual treatment regimens for cancer patients. So far, several reports have shown that CTC counts are much higher in small cell lung cancer (SCLC) than in non–small cell lung cancer (NSCLC; refs. 4, 5). Using the FDA-cleared CellSearch, CTC values in NSCLC seem to be even lower than in other cancer entities (4, 6, 7), pointing to the need for improved CTC detection in NSCLC. Moreover, liquid biopsies would be particularly desirable in NSCLC, where an emerging number of new molecular targets have been recently discovered that would benefit from CTC-based patient selection (8).

To date, more than 40 different approaches for CTC detection have been developed (9, 10). Various approaches have also been used to increase the yield of CTCs in NSCLC patient blood samples (8); to our best knowledge, none of these assays have been sufficiently tested to support FDA approval. In the current study, we validated a new strategy. The CellCollector (GILUPI CellCollector, GILUPI) enables an in vivo isolation of CTCs directly from the arm vein of cancer patients. During the 30-minute incubation in the vein, the device is exposed to approximately 1 L of blood, which increases the chance to capture CTCs. Our present investigation includes (i) the analysis of a large cohort of lung cancer patients to assess the distribution of CTCs, (ii) consecutive applications of the CTC device to demonstrate reproducibility of blood sampling for CTCs (which has rarely been shown in any of the >16,000 CTC publications in PubMed), (iii) monitoring of lung cancer patients during treatment for clinical response correlated to CTC counts, (iv) comparison of CellCollector (catheter-based) results to the FDA-cleared CellSearch system (blood sample-based), and (v) molecular characterization of CTCs captured by the medical device for mutations relevant to cancer therapy and comparison to the primary tumor of the same patients. This in-depth validation will support the use of in vivo CTC capture as a new diagnostic principle for future clinical trials testing the relevance of “liquid biopsies” to direct cancer therapies.

In vitro application of the CellCollector

The GILUPI CellCollector (termed as catheter-based or wire-based enrichment of EpCAM-positive cells) is based on a sterile stainless steel medical wire, covered with 2-μm gold and a hydrogel layer at its functionalized tip. This hydrogel layer is covalently coupled with antibodies against the EpCAM protein (humanized HEA 125, GILUPI GmbH; Fig. 1A). For our in vitro studies, the wire was incubated together with unauthenticated lung cancer cells (NCI-H332, NCI-H3122, NCI-H1650, 32M1, NCI-H1975, and A549) spiked into blood (EDTA) from healthy individuals for 30 minutes on a rotator. After the incubation, the wire was washed twice and fixed with acetone (10 minutes, room temperature). Captured cells were permeabilized (0.1% Triton X-100/PBS, 10 minutes), blocked (3% BSA/PBS, 30 minutes), and stained against EpCAM (Acris, clone HEA125-FITC) and pan-keratins (Exbio, clone C11-Alexa488; Millipore, clone LP5K-FITC; Exbio, clone A53-B/A2-Alexa488). Pan-keratins were stained to improve the detection rate of clinically relevant CTCs (11). CD45 staining was performed to exclude unspecific leucocytes (Exbio, clone MEM-28-Alexa647). Nuclear counterstaining was done using Hoechst33342 (Invitrogen) at a concentration of 1 μg/mL for 2 minutes. Next, the medical wire was placed into a special holder and subsequently screened for CTCs under the microscope (Fig. 5A). Images were obtained using the Axio Imager A1 microscope equipped with an AxioCam digital camera and the AxioVision 4.6 software (Zeiss). A tumor cell was counted as positive when its morphological features (≥4 μm, intact nucleus) and staining patterns were consistent with those of an epithelial cell (Hoechst33342-/pan-keratin- and/or EpCAM-positive/CD45-negative). The cell line A549 was purchased from CLS (Cat. No. 300114) and cultured in 45% Dulbecco's MEM (Biochrom; Cat. No. F0425) + 45% HAM's F-12 Medium (Biochrom; Cat. No. FG0815) + 10% FBS (Biochrom; Cat. No. S0615) + 2 mmol/L l-Glutamin (Biochrom; Cat. No. K0282) at 37°C with a 5% CO2 atmosphere. The cell line NCI-H1975 was purchased from the ATCC (Cat. No. CRL-5908) and cultured in 90% RPMI-1640 (Biochrom; Cat. No. FG1235) + 10% FBS (Biochrom; Cat. No. S0615) at 37°C with a 5% CO2 atmosphere. Cell line cells (NCI-H1650; Cat. No. CRL-5883 and NCI-H322; Cat. No. CRL-5806) were bought from ATCC. The cell line NCI-H3122 was kindly provided by C. Costa, PhD (Pangaea Biotech), and Dr. E. Weber (University of Halle) kindly provided the cell line 32M1. The cell lines NCI-H1650, NCI-H322, NCI-H3122, and 32M1 were cultured in 90% RPMI-1640 (Biochrom; Cat. No. FG1235) + 10% FBS (Biochrom; Cat. No. S0615) at 37°C with a 5% CO2 atmosphere. Initially to the experiments, all cell lines were tested negative for Mycoplasma using the Venor GeM Mycoplasma Detection Kit (Minerva Biolabs; Cat. No. 11-1100).

Figure 1.

A, schematic overview of the wire and the in vivo application. The wire is coated with anti-EpCAM antibodies and placed into the arm vein of cancer patients for 30 minutes. During the application EpCAM-positive cells bind to the device. Afterward, the tumor cells are stained for EpCAM and keratins. Nuclear counterstain is done using Hoechst33342. CD45 staining is necessary to classify false positive events (leukocytes). B, schematic overview of the study design. The schedule of the study was designed with the first visit directly before therapy initiation and the second visit 12 weeks after treatment started. Patients had the opportunity to have additional CTC investigations at a third and fourth visit (always after a 12-week period in between). Patients underwent the same CTC investigation procedures at each visit (7.5 mL blood draw for CellSearch analysis and two wire applications).

Figure 1.

A, schematic overview of the wire and the in vivo application. The wire is coated with anti-EpCAM antibodies and placed into the arm vein of cancer patients for 30 minutes. During the application EpCAM-positive cells bind to the device. Afterward, the tumor cells are stained for EpCAM and keratins. Nuclear counterstain is done using Hoechst33342. CD45 staining is necessary to classify false positive events (leukocytes). B, schematic overview of the study design. The schedule of the study was designed with the first visit directly before therapy initiation and the second visit 12 weeks after treatment started. Patients had the opportunity to have additional CTC investigations at a third and fourth visit (always after a 12-week period in between). Patients underwent the same CTC investigation procedures at each visit (7.5 mL blood draw for CellSearch analysis and two wire applications).

Close modal

Study design and clinical information

This prospective, blinded, single-center clinical study (in accordance with article 23b MDA) investigated the isolation and molecular characterization of CTCs from the peripheral blood of lung cancer patients using the CE-certified medical wire. The study was conducted at the University Medical Center Ulm and approved by the Ethics Committee of the University of Ulm in 2012. The study included 50 patients with newly diagnosed and locally advanced or metastatic lung cancer. The most predominant histology was NSCLC (n = 48, 96%), among them the adenocarcinoma subtype (n = 30, 60%). All patients were able to receive systemic anticancer treatment according current recommendations (12). A detailed summary of all patient data, including the CTC counts, is disclosed in Table 1 and Supplementary Table S1.

Table 1.

Patient characteristics

StagingCTC count: CellCollector A-1/A-2 (sum)CTC count: CellSearch
PatientHistologyTNMUI CCV1V2V1V2Treatment response
001 Squamous cell carcinoma 1b IV 2/0 (2) 0/3 (3) SD 
002 Adenocarcinoma 2b 1b IV 0/0 (0) 0/0 (0) n.a. n.a. n.a. 
003 SCLC 1b IV 3/3 (6) 6/2 (8) SD 
004 Squamous cell carcinoma 1a IV 10/0 (10) 0/0 (0) n.a. n.a. 
005 Adenocarcinoma IIIB 3/0 (3) 0/0 (0) n.a. n.a. 
006 Adenocarcinoma IIIA 1/6 (7) 4/9 (13) PR 
007 Adenocarcinoma 1b 1b IV 1/0 (1) 45/7 (52) n.a. SD 
008 Large cell carcinoma 1b IV 0/5 (5) 27/6 (33) 300 PD 
009 Adenocarcinoma 1b IV 1/0 (1) 5/5 (10) PD 
010 Adenocarcinoma 1b IV 0/0 (0) 1/11 (12) SD 
011 Squamous cell carcinoma 1b IV 0/0 (0) 11/6 (17) SD 
012 Adenocarcinoma 2a 1b IV 0/0 (0) 2/14 (16) SD 
013 Adenocarcinoma 1b IV 3/5 (8) 1/1 (2) 24 PD 
014 Adenocarcinoma 1b IV 3/11 (14) 8/18 (26) SD 
015 Squamous cell carcinoma IIIA 0/2 (2) 4/8 (12) SD 
016 Adenocarcinoma 1b IV 42/29 (71) 0/0 (0) n.a. PR 
017 Squamous cell carcinoma 1b IV 0/0 (0) 0/0 (0) n.a. n.a. 
018 Squamous cell carcinoma IIIA 5/13 (18) 0/0 (0) n.a. 
019 Squamous cell carcinoma IIIB 6/25 (31) 0/6 (6) n.a. PR 
020 Adenocarcinoma 1b IV 26/24 (50) 0/0 (0) n.a. SD 
021 Adenocarcinoma IIIB 10/46 (56) 5/8 (13) PR 
022 Adenocarcinoma 1b IV 20/22 (42) 0/0 (0) n.a. SD 
023 Squamous cell carcinoma 1b IV 35/13 (48) 0/3 (3) CR 
024 Adenocarcinoma 1b IV 41/56 (97) 0/0 (0) n.a. n.a. 
025 Adenocarcinoma 1a IV 17/1 (18) 1/0 (1) PR 
026 Adenocarcinoma 1b IV 2/0 (2) 0/0 (0) SD 
027 Adenocarcinoma 2a 1b IV 2/1 (3) 0/0 (0) PD 
028 Adenocarcinoma 1b IV 0/4 (4) 0/0 (0) n.a. n.a. 
029 Adenocarcinoma 1b IV 1/0 (1) 0/0 (0) n.a. n.a. 
030 Squamous cell carcinoma 1b IV 2/1 (3) 0/0 (0) SD 
031 Adenocarcinoma 1b IV 14/4 (18) 0/0 (0) n.a. n.a. 
032 Squamous cell carcinoma 1b IV 6/0 (6) 0/0 (0) SD 
033 Squamous cell carcinoma IIIB 6/6 (12) 3/4 (7) SD 
034 Adenocarcinoma IIIB 0/30 (30) 0/0 (0) SD 
035 Adenocarcinoma 1b IV 1/2 (3) 0/0 (0) SD 
036 Not otherwise specified 1b IV 1/6 (7) 0/0 (0) n.a. n.a. 
037 Squamous cell carcinoma IIIB 0/2 (2) 0/0 (0) SD 
038 Adenocarcinoma 1b IV 0/0 (0) 0/0 (0) n.a. n.a. 
039 Adenocarcinoma 1a IV 0/0 (0) 6/0 (6) n.a. PR 
040 Squamous cell carcinoma IIIB 0/0 (0) 0/0 (0) SD 
041 Adenocarcinoma 1a 1b IV 2/0 (2) 0/0 (0) n.a. n.a. 
042 Adenocarcinoma 1b IV 2/2 (4) 1/3 (4) SD 
043 Not otherwise specified 1b IV 0/0 (0) 2/0 (2) SD 
044 Squamous cell carcinoma 1b IV 0/0 (0) 0/0 (0) PD 
045 SCLC n.a. III 1/8 (9) 0/0 (0) n.a. n.a. n.a. 
046 Adenocarcinoma 1a IV 0/8 (8) n.a./0 (0) SD 
047 Adenocarcinoma 1b IV 0/1 (1) 0/12 (12) SD 
048 Adenocarcinoma 1a IV 0/0 (0) 1/2 (3) PR 
049 Adenocarcinoma 1a IV 16/0 (16) 1/1 (2) SD 
050 Squamous cell carcinoma 1a IV 0/29 (29) 0/4 (4) n.a. SD 
StagingCTC count: CellCollector A-1/A-2 (sum)CTC count: CellSearch
PatientHistologyTNMUI CCV1V2V1V2Treatment response
001 Squamous cell carcinoma 1b IV 2/0 (2) 0/3 (3) SD 
002 Adenocarcinoma 2b 1b IV 0/0 (0) 0/0 (0) n.a. n.a. n.a. 
003 SCLC 1b IV 3/3 (6) 6/2 (8) SD 
004 Squamous cell carcinoma 1a IV 10/0 (10) 0/0 (0) n.a. n.a. 
005 Adenocarcinoma IIIB 3/0 (3) 0/0 (0) n.a. n.a. 
006 Adenocarcinoma IIIA 1/6 (7) 4/9 (13) PR 
007 Adenocarcinoma 1b 1b IV 1/0 (1) 45/7 (52) n.a. SD 
008 Large cell carcinoma 1b IV 0/5 (5) 27/6 (33) 300 PD 
009 Adenocarcinoma 1b IV 1/0 (1) 5/5 (10) PD 
010 Adenocarcinoma 1b IV 0/0 (0) 1/11 (12) SD 
011 Squamous cell carcinoma 1b IV 0/0 (0) 11/6 (17) SD 
012 Adenocarcinoma 2a 1b IV 0/0 (0) 2/14 (16) SD 
013 Adenocarcinoma 1b IV 3/5 (8) 1/1 (2) 24 PD 
014 Adenocarcinoma 1b IV 3/11 (14) 8/18 (26) SD 
015 Squamous cell carcinoma IIIA 0/2 (2) 4/8 (12) SD 
016 Adenocarcinoma 1b IV 42/29 (71) 0/0 (0) n.a. PR 
017 Squamous cell carcinoma 1b IV 0/0 (0) 0/0 (0) n.a. n.a. 
018 Squamous cell carcinoma IIIA 5/13 (18) 0/0 (0) n.a. 
019 Squamous cell carcinoma IIIB 6/25 (31) 0/6 (6) n.a. PR 
020 Adenocarcinoma 1b IV 26/24 (50) 0/0 (0) n.a. SD 
021 Adenocarcinoma IIIB 10/46 (56) 5/8 (13) PR 
022 Adenocarcinoma 1b IV 20/22 (42) 0/0 (0) n.a. SD 
023 Squamous cell carcinoma 1b IV 35/13 (48) 0/3 (3) CR 
024 Adenocarcinoma 1b IV 41/56 (97) 0/0 (0) n.a. n.a. 
025 Adenocarcinoma 1a IV 17/1 (18) 1/0 (1) PR 
026 Adenocarcinoma 1b IV 2/0 (2) 0/0 (0) SD 
027 Adenocarcinoma 2a 1b IV 2/1 (3) 0/0 (0) PD 
028 Adenocarcinoma 1b IV 0/4 (4) 0/0 (0) n.a. n.a. 
029 Adenocarcinoma 1b IV 1/0 (1) 0/0 (0) n.a. n.a. 
030 Squamous cell carcinoma 1b IV 2/1 (3) 0/0 (0) SD 
031 Adenocarcinoma 1b IV 14/4 (18) 0/0 (0) n.a. n.a. 
032 Squamous cell carcinoma 1b IV 6/0 (6) 0/0 (0) SD 
033 Squamous cell carcinoma IIIB 6/6 (12) 3/4 (7) SD 
034 Adenocarcinoma IIIB 0/30 (30) 0/0 (0) SD 
035 Adenocarcinoma 1b IV 1/2 (3) 0/0 (0) SD 
036 Not otherwise specified 1b IV 1/6 (7) 0/0 (0) n.a. n.a. 
037 Squamous cell carcinoma IIIB 0/2 (2) 0/0 (0) SD 
038 Adenocarcinoma 1b IV 0/0 (0) 0/0 (0) n.a. n.a. 
039 Adenocarcinoma 1a IV 0/0 (0) 6/0 (6) n.a. PR 
040 Squamous cell carcinoma IIIB 0/0 (0) 0/0 (0) SD 
041 Adenocarcinoma 1a 1b IV 2/0 (2) 0/0 (0) n.a. n.a. 
042 Adenocarcinoma 1b IV 2/2 (4) 1/3 (4) SD 
043 Not otherwise specified 1b IV 0/0 (0) 2/0 (2) SD 
044 Squamous cell carcinoma 1b IV 0/0 (0) 0/0 (0) PD 
045 SCLC n.a. III 1/8 (9) 0/0 (0) n.a. n.a. n.a. 
046 Adenocarcinoma 1a IV 0/8 (8) n.a./0 (0) SD 
047 Adenocarcinoma 1b IV 0/1 (1) 0/12 (12) SD 
048 Adenocarcinoma 1a IV 0/0 (0) 1/2 (3) PR 
049 Adenocarcinoma 1a IV 16/0 (16) 1/1 (2) SD 
050 Squamous cell carcinoma 1a IV 0/29 (29) 0/4 (4) n.a. SD 

NOTE: Patient data and CTC counts from individual patients per visit (A-1/A-2 = wire application 1 or 2, V = visit, SD = stable disease, PR = partial remission, PD = progressive disease, n.a. = not analyzed).

In vivo application of the wire

Prior to the use of the device, a 20G peripheral venous catheter was placed into the median cubital vein of the patient. The wire was carefully inserted into the vein through the catheter until the tip of the device was extended 2 cm into the vein. To assess the reproducibility of measurements with the medical wire, two subsequent applications were performed during the same day (visit) for every cancer patient participating into the study. Patients underwent two subsequent device applications both before and after 12 weeks of therapy. Fourteen patients underwent only the pretherapy but not the following applications. Exceeding the initial study design, 7 patients had an additional third visit and one of them a fourth visit, each visit with double applications. This led to 16 additional device applications and a total of 185 analyzable applications. After the in vivo application, the wire was treated as described above. Cells captured by the device were stained and counted by different operators who were blinded to the patient's history and the CellSearch results.

CellSearch analysis

For method comparison, 7.5 mL of venous blood were collected into CellSave tubes immediately before the application of the wire. Blood was sent to the University Medical Center Hamburg-Eppendorf and processed within 96 hours as recommended by Janssen Diagnostics. The operator was blinded to the catheter-based results.

Statistical analysis

Clinical factors (T, N, M, and UICC stage) were tested for association with CTC count by multivariate analysis using ordinal logistic regression. Paired analyses between replicate measurements were tested for (i) difference in median using Wilcoxon signed rank test, (ii) difference for negative (0 CTCs) or positive (≥1 CTC) using Liddell's exact test, and (iii) correlation between ordinal values of CTCs using Kendall's tau. The difference between median number of CTCs as detected by the wire and CellSearch was assessed by the Wilcoxon signed rank test, and the agreement and correlation between the results of the two systems was assessed using Cohen's kappa and Kendall's tau, respectively. The differences in CTC count (visit 2 − visit 1) for catheter-based and blood sample–based system results were tested for association with therapy response using the Kruskal–Wallis test.

Mutational analysis of tumor cells captured by in vivo isolation

CTC-positive wires from 2 patients (from different applications and visits) with known mutational status of the primary tumor were used for further molecular characterization of the CTCs by chip-based digital PCR (dPCR). Briefly, the wire was fragmented into small pieces into a 0.2 mL PCR-tube filled with 4.8 μL PBS (Qiagen). Whole genome amplification (WGA) was performed using the REPLI-g Single Cell Kit (Qiagen) using a 1.2× reaction volume. Amplified gDNA was quantified with the QuantiFlour ONE dsDNA System (Promega) and a fluorescence reader (FluoroskanAscent; Thermo Fisher) at 485 nmEx/538 nmEm. For chip-based dPCR the ThermoFisher QuantStudio 3D system was used. Fifty nanograms of WGA DNA (2 μL) was combined with 7.5 μL of 2 × dPCR Mastermix and 0.75 μL of 20 × SNP assay for the appropriate allele. The total volume was brought to 15 μL with water, applied to the chip and sealed according to the manufacturer's instructions. The chip was placed on a flat-block PCR machine and cycled: 45°C 1:00 minute, 98°C 2:30 minutes (95°C 1:00 minute, 53.5°C 0:1 minute with 30% ramp, 57.5°C 0:30 minutes) for 40 cycles, and held at either 10°C or 20°C until read. Chips were read on the QuantStudio 3D reader and analyzed by QuantStudio 3D AnalysisSuite and Microsoft Excel. SNP assays were designed using the ThermoFisher Custom SNP design pipeline (Life Technologies) without modifications. PBMCs (1,000) from a healthy donor subjected to the same WGA protocol were used as a negative control to prove the specificity of the assay.

Mutational analysis of the primary tumor

Mutational analysis of the primary tumor tissue was performed in the institute of pathology at the University Medical Center Ulm. EGFR (exons 18, 19, 20, and 21) mutational analysis was done with Sanger sequencing with a lower limit of 10% mutated DNA. KRAS (exons 2, 3, and 4) mutational analysis was done with pyro sequencing with a lower limit of 5% mutated DNA.

Study design and patient characteristics

Fifty patients with newly diagnosed and locally advanced or metastatic lung cancer (stage III, n = 11; stage IV, n = 39) with a performance status of ECOG 0 or 1 and who were able to receive a systemic antitumor treatment (platinum-based doublet) according to current clinical guidelines (12) were enrolled into this study. Tumor histology included squamous cell carcinoma (n = 15), adenocarcinoma (n = 30), small cell carcinoma (n = 2), and large cell carcinoma (n = 1), as well as 2 carcinoma patients with unknown histology (Table 1). Staging for distant overt metastases was available from 49 patients; 39 (78%) patients had distant overt metastases (stage M1), whereas 10 patients (20%) had no clinical or radiological signs of overt metastases (stage M0). Local tumor extension (T stage) ranged from T1 to T4 with the majority of patients showing large T3 or T4 tumors (80%). No significant differences were seen with regard to CTC numbers and clinicopathologic parameters such as TNM or UICC stage (Supplementary Table S2).

A schematic overview of the wire during in vivo application is presented in Fig. 1A. In total, 185 in vivo applications were performed and screened for CTCs (Fig. 1B). Briefly, the wire—coated with anti-EpCAM antibodies—was placed into the arm vein of cancer patients for 30 minutes where EpCAM-positive cells can bind to the device. Afterward, the wire was placed into a special holder under the microscope and tumor cells were identified based on epithelial markers (EpCAM- and/or pan-keratin-positive) and nuclear counterstain (Hoechst33342-positive) by fluorescence microscopy. CD45 staining is necessary to discriminate possible leukocytes bound to the wire.

CTC counts and correlations to clinicopathologic characteristics

In total, 58% (108/185) of the analyzed wires were positive for ≥1 CTC (range, 1–56 cells; Fig. 2A). Most of the identified cells were single CTCs. However, in 20 of 185 analyzed wires CTC clusters (range, 3–7 cells) could also be detected. Representative image galleries of tumor cells from spiking experiments, CTCs, CTC clusters, and leukocytes (range, 0–≥1,000 per wire) are disclosed in Fig. 2B and C. No “CTC-like” events were found on the wires placed into the arm vein of healthy individuals (n = 10; data not shown).

Figure 2.

A, CTC counts of all patients from all visits. Overall, 185 wire samples were taken from a total of 50 patients, resulting in CTC detection in 108 samples, as determined by immunofluorescence analysis. B, representative images of 6 different lung cancer cell lines (NCI-H332, NCI-H3122, NCI H1650, 32M1, A549, and NCI-H1975) isolated by the CellCollector in spiking experiments (in vitro). Tumor cells were identified as EpCAM- and/or pan-keratin-positive events (green). C, representative images of CTCs and CTC clusters isolated with the wire and 2 leukocytes. Tumor cells were identified as EpCAM- and/or pan-keratin-positive (green) and CD45-negative (red) events. Hoechst33342 (blue) was used for nuclear counterstain. Scale bars, 20 μm.

Figure 2.

A, CTC counts of all patients from all visits. Overall, 185 wire samples were taken from a total of 50 patients, resulting in CTC detection in 108 samples, as determined by immunofluorescence analysis. B, representative images of 6 different lung cancer cell lines (NCI-H332, NCI-H3122, NCI H1650, 32M1, A549, and NCI-H1975) isolated by the CellCollector in spiking experiments (in vitro). Tumor cells were identified as EpCAM- and/or pan-keratin-positive events (green). C, representative images of CTCs and CTC clusters isolated with the wire and 2 leukocytes. Tumor cells were identified as EpCAM- and/or pan-keratin-positive (green) and CD45-negative (red) events. Hoechst33342 (blue) was used for nuclear counterstain. Scale bars, 20 μm.

Close modal

For assessing clinical correlation to CTC counts, the sum of the CTC counts of the double applications were used in our multivariate analysis. First, an ordinal scale with four categories for CTC values was created (negative, 0 CTCs; low, 1–4 CTCs; intermediate; 5–17 CTCs, high, ≥18 CTCs) such that every category had approximately the same number of cases. The resulting regression revealed that none of the clinical factors (T, N, or M stage as well as UICC) was significantly correlated with the CTC categories, although the tumor stage was close to significant (P = 0.09).

Reproducibility of wire applications

To assess the reproducibility of measurements with the wire, the results of the two subsequent CTC analyses performed at each visit from each cancer patient (Table 1) were compared and statistically evaluated for potential differences.

At the first (pre-therapeutic) visit, CTC counts of ≥1 were detected in 78% of the study participants (combined analysis). As expected, the paired analyses resulted in some differences in CTC counts; however, these differences were not significant (P = 0.33, Wilcoxon signed rank test). Furthermore, the number of CTC-positive (≥1 CTCs) and -negative (0 CTCs) cases was also not significantly different between the two analyses (P = 0.59, relative risk 95% CI, 0.35–2.84, Liddell exact test). And finally, a positive and significant correlation could be detected between CTC values of the paired analyses of the first visit (P = 0.003, τ = 0.36, Kendall rank correlation).

At the second visit, 60% (20/33) and 56% (20/36) of the patients were identified to be CTC positive with repeated wire analyses (combined positivity rate 72%). Comparing the results of the two analyses, no significant difference between the number of identified tumor cells could be detected (P = 0.42, Wilcoxon signed rank test). Furthermore, the number of CTC negative and positive patients was comparable (P = 0.64, Liddell's exact test). Finally, the correlation between the number of CTCs detected with the two analyses was also significantly correlated (P = 0.0015, τ = 0.50, Kendall rank correlation).

Performing the same statistical analyses on the two wire analyses of visit 3, the number of CTCs and the number of CTC-positive and -negative patients were not significantly different (P = 0.88, Wilcoxon signed rank test; P = 0.25, Liddell exact test); however, the correlation between the CTC values was not significant anymore because almost all cases were CTC negative (P = 0.59, τ = 0.27, Kendall rank correlation). Visit 4 was excluded from this analysis because only one case was analyzed.

Comparison of CTC counts obtained with the wire and CellSearch

For a direct comparison of the two CTC assays, 84 analyzable blood samples (7.5 mL) were taken prior to the application of the wire and screened ex vivo for CTCs by the FDA-cleared CellSearch system. As shown in Fig. 3A, the incidence of CTC-positive lung cancer patients was higher if determined by the wire compared with the FDA-cleared technology. Direct comparison of paired patient data showed that the wire isolated a median of 3 CTCs more per patient (P < 0.0001, Wilcoxon signed rank test). Only in 11 measurements were more CTCs found with the CellSearch system than with the wire.

Figure 3.

A, direct method comparison of CTC detection with the wire and the CellSearch system. Only complete data sets that include results from CellSearch and both wire applications were used. The respective detection rates are shown below the graph. B, direct comparison (weighted scatter plot) between the wire and CellSearch.

Figure 3.

A, direct method comparison of CTC detection with the wire and the CellSearch system. Only complete data sets that include results from CellSearch and both wire applications were used. The respective detection rates are shown below the graph. B, direct comparison (weighted scatter plot) between the wire and CellSearch.

Close modal

Taken together, these results indicated that in vivo isolation has a higher sensitivity in capturing CTCs from lung cancer patients. It should be noted that there was no correlation between the positive and negative cases as determined between in vivo isolation and CellSearch results (P = 0.76, κ = 0.0598, Cohen's kappa) or between the ordinal CTC values (P = 0.99, τ = 0, Kendall rank correlation). A direct comparison (weighted scatter plot) between the wire and CellSearch is visualized in Fig. 3B.

Monitoring of CTCs during systemic therapy

To evaluate changes in CTC counts as a potential marker for monitoring treatment response, patients were grouped into responsive, stable, and progressive disease, according to current RECIST criteria, 1.1 (13). Cancer patients were screened for CTCs using two subsequent applications of the medical wire, both before and after 12 weeks of any systemic treatment. Patients were grouped according to their therapy responses assessed after initiation of therapy—responder, stable disease, or progressive disease. We could show that patients responding to the therapy [complete responders (CR) or partial responders (PR)] showed reduced CTC counts with a median decrease of 8.25 CTCs after 12 weeks initial first-line therapy (Fig. 4). Furthermore, patients with stable disease had a median difference of 0 CTCs compared with their previous cell number investigation, whereas patients with progressive disease were diagnosed with a median increase of 4.5 CTCs (P = 0.0013, Kruskal–Wallis test).

Figure 4.

Reduction of CTC counts under therapy determined with the wire. Box-and-whisker plot of CTC count difference between different visits (V=visit) for patients with responsive, stable, and progressive disease.

Figure 4.

Reduction of CTC counts under therapy determined with the wire. Box-and-whisker plot of CTC count difference between different visits (V=visit) for patients with responsive, stable, and progressive disease.

Close modal

Such significant differences were not seen by CellSearch (Fig. 4), where the median CTC counts between subsequent visits were similar in all response groups (P = 0.13, Kruskal–Wallis test).

Mutational analysis of CTCs captured by the medical wire

CTCs could be used for “liquid biopsy” to guide individual treatment decisions by mutational analysis of genes expressing targets or relevant pathway proteins. Therefore, we performed a proof-of-principle investigation whether CTCs captured by the wire were suitable for further molecular characterization. For this purpose, captured cells (after staining) from CTC-positive patients with known mutational status of the primary tumor were analyzed (Fig. 5A–C).

Figure 5.

Mutational analysis of the KRAS and EGFR genes in captured CTCs. A, cells isolated with the CellCollector were analyzed microscopically by means of immunofluorescence staining according to the staining procedure in the clinical setup. The wire was fragmented and subjected to WGA. Patient material examined by dPCR, CTC count, and primary tumor mutation status (NTC = none template control, WT = wild-type, and V = visit). B, amplified gDNA of patient 13 was analyzed by dPCR for the known primary tumor mutation KRAS G12D and KRAS G12C as specificity control. C, amplified gDNA of patient 47 was analyzed by dPCR for the known primary tumor mutation EGFR H773R and the other EGFR mutations L858R and T790M to control for specificity.

Figure 5.

Mutational analysis of the KRAS and EGFR genes in captured CTCs. A, cells isolated with the CellCollector were analyzed microscopically by means of immunofluorescence staining according to the staining procedure in the clinical setup. The wire was fragmented and subjected to WGA. Patient material examined by dPCR, CTC count, and primary tumor mutation status (NTC = none template control, WT = wild-type, and V = visit). B, amplified gDNA of patient 13 was analyzed by dPCR for the known primary tumor mutation KRAS G12D and KRAS G12C as specificity control. C, amplified gDNA of patient 47 was analyzed by dPCR for the known primary tumor mutation EGFR H773R and the other EGFR mutations L858R and T790M to control for specificity.

Close modal

CTC-positive wires from the two patients (patients 013 and 047) with confirmed primary tumor mutations in the KRAS (KRAS p.G12D, c.35A>D) or EGFR (EGFR p.H773R, c.2318A>G) gene were used for molecular analysis applying the QuantStudio 3D Digital PCR system (Life Technologies). Captured CTCs were detected by immunofluorescence microscopy followed by cell lysis, including unspecific adhering blood cells, and WGA. The amplified genomic DNA was then analyzed for specific KRAS and EGFR mutations. The same KRAS and EGFR mutations found in the primary tumors were clearly detectable when 1 to 5 CTCs were present on the wire. To ensure specificity of the approach, the KRAS G12D-mutated sample was also tested for G12C mutation. Furthermore, the EGFR H773-mutated sample was tested for the alternative activating mutation L858R and the resistance mutation T790M. The analyses clearly demonstrated the specificity of our dPCR approach because no signals were seen when analyzing the G12C, L858R, or T790M mutations (Fig. 5B and C).

In this study, we demonstrated that in vivo isolation of CTCs is a promising approach for sensitive capture of CTCs. In vivo isolation demonstrated higher CTC capture from lung cancer patients compared with the FDA-cleared CellSearch system. Both devices capture and identify EpCAM-/keratin-positive CTCs; thus, the observed differences are most likely due to the new principle of capturing CTCs in vivo. Although taking a blood sample is certainly easier than placing the wire into the arm vein for 30 minutes, all cancer patients undergo much more time consuming and unpleasant diagnostic approaches, including CT, MRI, and even invasive biopsies with the risk of severe bleeding or infection.

Combined wire applications during the same visit (pre- or post-therapy) revealed CTC positivity rates of 72% to 78%. Similar detection rates could be observed in advanced NSCLC patients using an EpCAM-independent filter technology for CTC detection based on cell size selection (ISET 80% vs. CellSearch 23%; ref. 6). However, the ISET filter has a pore size of 8 μm (14) and smaller CTCs exist (15) that can be lost. Nevertheless, filter systems as well as other EpCAM-independent assays can be helpful in clarifying the relevance of “mesenchymal CTCs” with loss of epithelial markers. Experimental studies indicated that tumor cells with a mesenchymal phenotype have a higher propensity to disseminate through the blood stream, but appear to lack the ability to form metastases at distant sites (16, 17). This would explain why CTCs expressing (at least some) epithelial markers, such as EpCAM and keratins, are excellent indicators of prognosis in cancer patients (8).

In this current investigation, we could further demonstrate that subsequent applications of the wire during the same visit are feasible and increase the capture rate of CTCs. Multiple collector wires can be used for different types of downstream analyses (e.g., immunostainings and PCR-based mutation analyses) at the same time point, which increases the applicability of the wire. Moreover, double applications provide an opportunity to analyze the reproducibility of subsequent samples performed with the wire. Although the paired in vivo analyses at the same visit resulted in differences regarding exact CTC quantification, no significant differences between repeated CTC measurements were observable, indicating that the CTC results are reproducible. This was also shown by the reproducible and specific detection of a KRAS mutation using four samples from 1 patient. Regarding the low number of CTCs detected, this aspect is of utmost importance and has been rarely addressed on real clinical specimens using other CTC assays.

It has been shown that patients who harbor clusters of CTCs have a worse clinical outcome (18). In this study, CTC clusters were detectable in 20 of 185 wires after in vivo application. Future studies are needed to clarify the clinical relevance of CTC clusters found on the wire. The cellular composition of clusters can be heterogeneous and their biology is still under investigation. It can be speculated that clusters might contain CTCs from different tumor clones that are required in a cooperative manner to establish metastases at distant sites (5, 6).

Monitoring of treatment responses is one of the key applications of CTC assays (19–21). In the present study, we observed significant correlation between changes in CTC counts and response to therapy. Decreased CTC counts after therapy initiation were frequently seen in lung cancer patients who responded to systemic treatment, whereas increased CTC numbers were observed in most patients with progressive disease. Although this finding needs to be validated in larger studies, it suggests that in vivo isolation of CTCs might be useful for monitoring therapies in NSCLC. In contrast, using the ex vivo blood sample-based approach to monitor changes in CTC counts did not correlate to treatment response in our present cohort of patients, which seems to be different from another report demonstrating that a change in CellSearch-based CTC counts after treatment was predictive of progression-free survival in NSCLC (22). This discrepancy may be explained by a different selection of patients, and larger multicenter studies should be analyzed to define the value of both assays in NSCLC.

Besides monitoring of response to therapy, the molecular characterization of CTCs can provide important information on therapeutic targets or resistance mechanisms (23–28). Thus, we evaluated whether the in vivo isolated CTCs also allow subsequent molecular characterization of genes relevant to therapy of cancer patients. We selected KRAS and EGFR as the two prominent examples already used in clinical routine for decision making (28). Although only two patients have been analyzed, we could demonstrate that the DNA from CTCs captured by the wire could be isolated, amplified, and analyzed for mutations in these genes, and—most importantly—the results matched the gene status in the primary tumor of the same patients. Control assays for mutually exclusive mutations in the same genes were negative, proving the specificity of the analysis. Although this analysis could only be done for 2 patients due to the limited clinical data set available, it clearly indicated the technical feasibility of mutational analysis on (or with) isolated CTCs. Such results may lead to treatment decisions for or against EGFR TKIs based on the presence or absence of activating or resistance mutations in EGFR or against cetuximab therapy due to the presence of a KRAS mutation. The feasibility of mutational analysis based on device-isolated CTCs is an important prerequisite for future use of the wire for “liquid biopsy” analyses in cancer patients. Taking biopsies from lung cancer patients for the assessment of molecular alterations relevant to therapy is an invasive, time-consuming, painful, and risk-associated procedure (e.g., severe bleeding events). The present investigation suggests that CTCs may offer solutions to this problem in the near future. Genomic information can also be obtained by the analysis of circulating cell-free DNA released by dying cancer cells into the blood samples, which has recently achieved great attention (29). However, CTC analysis can be performed on the DNA, RNA, protein level and capture of sufficient amounts of CTCs to also allow functional studies (25, 30, 31).

Our first validation study of the wire has produced encouraging results. However, there are still some limitations and open issues that need to be discussed. The wire application does not permit reporting of CTC counts in a given (defined) blood volume (i.e., CTCs in 10 mL of blood). Using this device, the value can only be reported as CTCs collected in 30 minutes. Because the amount of blood that flows through the vein can be variable from patient to patient, or even at different time points in the same patient, this may make it more difficult to interpret quantitative differences after treatment. However, we could show that subsequent applications of the wire during the same visit are feasible and increase the capture rates for CTCs at which time no significant differences between repeated CTC measurements were observable, indicating that the CTC results are reproducible. Besides, we had only a limited number of patients without metastases, which may also explain why no significant differences in CTC rates were observed. Larger studies should be performed including more early-stage lung cancer patients that are screened and analyzed for CTCs. Finally, the current anti-EpCAM antibodies used for capturing of CTCs have demonstrated proof-of-principle, but future generations of the device should include a mixture of different capturing antibodies to isolate subsets of EpCAM-negative cancer cells that do not display the epithelial phenotype, e.g., cells that underwent an epithelial to mesenchymal transition (32). This development will further increase the capture rate of CTCs via the wire.

In conclusion, our present findings indicate that in vivo isolation of CTCs is a promising approach to capture sufficient amounts of CTCs amenable to downstream molecular analyses in lung cancer. This new diagnostic tool can now be tested in interventional clinical studies where the number and/or characteristics of CTCs will be used to stratify patients to particular treatments and established endpoints such as survival will be assessed. There is an increasing number of molecular targets that are only present in a small fraction of NSCLC patients (33), which requires stratification based on the molecular characteristics of tumor cells in individual patients. Furthermore, CTCs have recently been detected in patients with chronic obstructive pulmonary disease without clinically detectable lung cancer (34). The implementation of a sensitive in vivo device may now also allow early diagnosis of lung cancer. Thus, future studies are needed to explore the entire clinical potential of the wire.

N. Brychter and O. von Ahsen report receiving commercial research grants from Bayer Pharma AG. T. Krahn is a consultant/advisory board member for Gilupi UK. K. Pantel is a consultant/advisory board member for Gilupi GmbH. No potential conflicts of interest were disclosed by the other authors.

Conception and design: T. Schalk, S. Riethdorf, K. Lücke, C. Schumann, T. Krahn, K. Pantel

Development of methodology: K. Lücke, S. Jackson, O. von Ahsen, C. Schumann, T. Krahn, K. Pantel

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): N. Penkalla, T. Schalk, S. Riethdorf, S. Jackson, N. Brychta, C. Schumann, T. Krahn, K. Pantel

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): T.M. Gorges, T. Schalk, S.A. Joosse, S. Riethdorf, J. Tucholski, K. Lücke, H. Wikman, S. Jackson, N. Brychta, O. von Ahsen, C. Schumann, K. Pantel

Writing, review, and/or revision of the manuscript: T.M. Gorges, N. Penkalla, T. Schalk, S.A. Joosse, S. Riethdorf, K. Lücke, H. Wikman, O. von Ahsen, C. Schumann, T. Krahn, K. Pantel

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): N. Penkalla, T. Schalk, K. Lücke, T. Krahn

Study supervision: K. Pantel

Other (conception of the manuscript design): T.M. Gorges

Other (using the CellCollectors on the patients directly (in vivo) and their further processing): N. Penkalla

Tobias M. Gorges, Sabine Riethdorf, and Klaus Pantel thank the City of Hamburg, Landesexzellenzinitiative Hamburg (LEXI 2012; Tumor targeting via cell-surface molecules essential in cancer progression and dissemination).

Part of this work is supported by the Bundesministerium für Bildung und Forschung (BMBF) (Förderkennzeichen 01EZ 0863) and the Ministerium für Wirtschaft des Landes Brandenburg und der EU (Förderkennzeichen 80135447). T.M. Gorges, S. Riethdorf, and K. Pantel supported by the ERC Advanced Investigator Grant DISSECT, TRANSCAN ERA-Network: Grant CTC-SCAN and from the Innovative Medicines Initiative Joint Undertaking under grant agreement no. 115749, resources of which are composed of financial contribution from the European Union's Seventh Framework Programme (FP7/2007-2013) and EFPIA companies' in kind contribution.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Alix-Panabieres
C
,
Pantel
K
. 
Challenges in circulating tumour cell research
.
Nat Rev Cancer
2014
;
14
:
623
31
.
2.
Parkinson
DR
,
Dracopoli
N
,
Petty
BG
,
Compton
C
,
Cristofanilli
M
,
Deisseroth
A
, et al
Considerations in the development of circulating tumor cell technology for clinical use
.
J Transl Med
2012
;
10
:
138
.
3.
Barradas
AM
,
Terstappen
LW
. 
Towards the biological understanding of CTC: capture technologies, definitions and potential to create metastasis
.
Cancers
2013
;
5
:
1619
42
.
4.
Truini
A
,
Alama
A
,
Dal Bello
MG
,
Coco
S
,
Vanni
I
,
Rijavec
E
, et al
Clinical applications of circulating tumor cells in lung cancer patients by cellsearch system
.
Front Oncol
2014
;
4
:
242
.
5.
Hou
JM
,
Krebs
MG
,
Lancashire
L
,
Sloane
R
,
Backen
A
,
Swain
RK
, et al
Clinical significance and molecular characteristics of circulating tumor cells and circulating tumor microemboli in patients with small-cell lung cancer
.
J Clin Oncol
2012
;
30
:
525
32
.
6.
Krebs
MG
,
Hou
JM
,
Sloane
R
,
Lancashire
L
,
Priest
L
,
Nonaka
D
, et al
Analysis of circulating tumor cells in patients with non-small cell lung cancer using epithelial marker-dependent and -independent approaches
.
J Thorac Oncol
2012
;
7
:
306
15
.
7.
Isobe
K
,
Hata
Y
,
Kobayashi
K
,
Hirota
N
,
Sato
K
,
Sano
G
, et al
Clinical significance of circulating tumor cells and free DNA in non-small cell lung cancer
.
Anticancer Res
2012
;
32
:
3339
44
.
8.
Krebs
MG
,
Metcalf
RL
,
Carter
L
,
Brady
G
,
Blackhall
FH
,
Dive
C
. 
Molecular analysis of circulating tumour cells-biology and biomarkers
.
Nat Rev Clin Oncol
2014
;
11
:
129
44
.
9.
Gorges
TM
,
Pantel
K
. 
Circulating tumor cells as therapy-related biomarkers in cancer patients
.
Cancer Immunol Immunother
2013
;
62
:
931
9
.
10.
Joosse
SA
,
Gorges
TM
,
Pantel
K
. 
Biology, detection, and clinical implications of circulating tumor cells
.
EMBO Mol Med
2015
;
7
:
1
11
.
11.
Joosse
SA
,
Hannemann
J
,
Spotter
J
,
Bauche
A
,
Andreas
A
,
Muller
V
, et al
Changes in keratin expression during metastatic progression of breast cancer: impact on the detection of circulating tumor cells
.
Clin Cancer Res
2012
;
18
:
993
1003
.
12.
Goeckenjan
G
,
Sitter
H
,
Thomas
M
,
Branscheid
D
,
Flentje
M
,
Griesinger
F
, et al
Prevention, diagnosis, therapy, and follow-up of lung cancer: interdisciplinary guideline of the German Respiratory Society and the German Cancer Society
.
Pneumologie (Stuttgart, Germany)
2011
;
65
:
39
59
.
13.
Eisenhauer
EA
,
Therasse
P
,
Bogaerts
J
,
Schwartz
LH
,
Sargent
D
,
Ford
R
, et al
New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1)
.
Eur J Cancer (Oxford, England: 1990)
2009
;
45
:
228
47
.
14.
Vona
G
,
Sabile
A
,
Louha
M
,
Sitruk
V
,
Romana
S
,
Schutze
K
, et al
Isolation by size of epithelial tumor cells: a new method for the immunomorphological and molecular characterization of circulating tumor cells
.
Am J Pathol
2000
;
156
:
57
63
.
15.
de Wit
S
,
van Dalum
G
. 
Detection of circulating tumor cells
Scientifica
2014
;
2014
:
819362
.
16.
Kang
Y
,
Pantel
K
. 
Tumor cell dissemination: emerging biological insights from animal models and cancer patients
.
Cancer Cell
2013
;
23
:
573
81
.
17.
Tam
WL
,
Weinberg
RA
. 
The epigenetics of epithelial-mesenchymal plasticity in cancer
.
Nat Med
2013
;
19
:
1438
49
.
18.
Aceto
N
,
Bardia
A
,
Miyamoto
DT
,
Donaldson
MC
,
Wittner
BS
,
Spencer
JA
, et al
Circulating tumor cell clusters are oligoclonal precursors of breast cancer metastasis
.
Cell
2014
;
158
:
1110
22
.
19.
Bidard
FC
,
Peeters
DJ
,
Fehm
T
,
Nole
F
,
Gisbert-Criado
R
,
Mavroudis
D
, et al
Clinical validity of circulating tumour cells in patients with metastatic breast cancer: a pooled analysis of individual patient data
.
Lancet Oncol
2014
;
15
:
406
14
.
20.
Krebs
MG
,
Renehan
AG
,
Backen
A
,
Gollins
S
,
Chau
I
,
Hasan
J
, et al
Circulating tumor cell enumeration in a phase II trial of a four-drug regimen in advanced colorectal cancer
.
Clin Colorectal Cancer
2014
.
14
:
115
22
21.
Scher
HI
,
Heller
G
,
Molina
A
,
Attard
G
,
Danila
DC
,
Jia
X
, et al
Circulating tumor cell biomarker panel as an individual-level surrogate for survival in metastatic castration-resistant prostate cancer
.
J Clin Oncol
2015
;
33
:
1348
55
22.
Krebs
MG
,
Sloane
R
,
Priest
L
,
Lancashire
L
,
Hou
JM
,
Greystoke
A
, et al
Evaluation and prognostic significance of circulating tumor cells in patients with non-small-cell lung cancer
.
J Clin Oncol
2011
;
29
:
1556
63
.
23.
Gasch
C
,
Bauernhofer
T
,
Pichler
M
,
Langer-Freitag
S
,
Reeh
M
,
Seifert
AM
, et al
Heterogeneity of epidermal growth factor receptor status and mutations of KRAS/PIK3CA in circulating tumor cells of patients with colorectal cancer
.
Clin Chem
2013
;
59
:
252
60
.
24.
Heitzer
E
,
Auer
M
,
Gasch
C
,
Pichler
M
,
Ulz
P
,
Hoffmann
EM
, et al
Complex tumor genomes inferred from single circulating tumor cells by array-CGH and next-generation sequencing
.
Cancer Res
2013
;
73
:
2965
75
.
25.
Hodgkinson
CL
,
Morrow
CJ
,
Li
Y
,
Metcalf
RL
,
Rothwell
DG
,
Trapani
F
, et al
Tumorigenicity and genetic profiling of circulating tumor cells in small-cell lung cancer
.
Nat Med
2014
;
20
:
897
903
.
26.
Riethdorf
S
,
Muller
V
,
Zhang
L
,
Rau
T
,
Loibl
S
,
Komor
M
, et al
Detection and HER2 expression of circulating tumor cells: prospective monitoring in breast cancer patients treated in the neoadjuvant GeparQuattro trial
.
Clin Cancer Res
2010
;
16
:
2634
45
.
27.
Stinchcombe
TE
. 
Novel agents in development for advanced non-small cell lung cancer
.
Ther Adv Med Oncol
2014
;
6
:
240
53
.
28.
Maheswaran
S
,
Sequist
LV
,
Nagrath
S
,
Ulkus
L
,
Brannigan
B
,
Collura
CV
, et al
Detection of mutations in EGFR in circulating lung-cancer cells
.
N Engl J Med
2008
;
359
:
366
77
.
29.
Ilie
M
,
Hofman
V
,
Long
E
,
Bordone
O
,
Selva
E
,
Washetine
K
, et al
Current challenges for detection of circulating tumor cells and cell-free circulating nucleic acids, and their characterization in non-small cell lung carcinoma patients. What is the best blood substrate for personalized medicine?
Ann Transl Med
2014
;
2
:
107
.
30.
Baccelli
I
,
Schneeweiss
A
,
Riethdorf
S
,
Stenzinger
A
,
Schillert
A
,
Vogel
V
, et al
Identification of a population of blood circulating tumor cells from breast cancer patients that initiates metastasis in a xenograft assay
.
Nat Biotechnol
2013
;
31
:
539
44
.
31.
Cayrefourcq
L
,
Mazard
T
,
Joosse
S
,
Solassol
J
,
Ramos
J
,
Assenat
E
, et al
Establishment and characterization of a cell line from human circulating colon cancer cells
.
Cancer Res
2015
;
75
:
892
901
.
32.
Yu
M
,
Bardia
A
,
Wittner
BS
,
Stott
SL
,
Smas
ME
,
Ting
DT
, et al
Circulating breast tumor cells exhibit dynamic changes in epithelial and mesenchymal composition
.
Science (New York, NY)
2013
;
339
:
580
4
.
33.
Camidge
DR
,
Pao
W
,
Sequist
LV
. 
Acquired resistance to TKIs in solid tumours: learning from lung cancer
.
Nat Rev Clin Oncol
2014
;
11
:
473
81
.
34.
Ilie
M
,
Hofman
V
,
Long-Mira
E
,
Selva
E
,
Vignaud
JM
,
Padovani
B
, et al
“Sentinel” circulating tumor cells allow early diagnosis of lung cancer in patients with chronic obstructive pulmonary disease
.
PLoS One
2014
;
9
:
e111597
.