Utility of CK8/18 in identifying circulating tumor cells derived from lesions in patients with non-small cell lung cancer

Introduction

Circulating tumor cells (CTCs) are solid cancer-derived cells that circulate in the blood (1,2). The number of CTCs is a prognostic indicator of survival and treatment response across various cancer types (3,4). However, given the limited number of tumor cells in the blood, estimated to be 1–10² cells for every 10⁷ leukocytes, detecting CTCs in blood samples is a significant challenge. Conventional separation techniques often depend on the epithelial cell adhesion molecule (EpCAM) for enrichment, which is a prevalent surface antigen in CTCs (5). However, a subset of CTCs undergoes epithelial-mesenchymal transition (EMT), resulting in the downregulation of epithelial markers, and techniques based on the EpCAM for capturing CTCs may fail to enrich CTCs effectively (6). Particularly in NSCLC, EpCAM expression has been reported to be lower than that in breast and prostate cancers, resulting in reduced detection sensitivity (6,7). Therefore, several alternative separation methods have emerged using physical parameters, such as cell size, density, and electric charge, instead of EpCAM expression to overcome these limitations (8).

Our group developed a novel CTC enrichment technique based on cell size rather than surface antigen expression (9). ImageStream Mark II (Cytek Biosciences, Bethesda, MD, USA) has the advantages of high-throughput flow cytometry and high-resolution imaging (10). ImageStream Mark II was combined with our enrichment technology to develop a CTC detection system. This innovative system has demonstrated successful CTC enrichment in the peripheral blood of patients with breast and prostate cancers, thereby demonstrating the efficacy of this system in CTC detection (9). This CTC detection technique identifies CTCs based on the absence of pan-leukocyte markers and positive staining for cytokeratin (CK). However, the conventional anti-panCK antibody (AE1/AE3), which is a widely used standard CK marker, exhibits reduced sensitivity in NSCLC because of the diminished expression caused by EMT (11). To overcome this limitation, the anti-CK antibody CK8/18 was used, which has demonstrated potential for detecting CTCs undergoing EMT, thereby improving the CTC detection rate.

This study aimed to evaluate the efficacy of our new CTC detection system using CK8/18 in patients with NSCLC. Furthermore, the lung cancer origin of CK8/18-positive CTCs was confirmed using single-cell sorting and subsequent gene analysis. We present this article in accordance with the STARD reporting checklist (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-155/rc).

Methods

Evaluation of anti-CK antibodies in cell lines

Lung cancer cell lines A549 (CVCL_0023), HCC827 (CVCL_2063), H1299 (CVCL_0060), H1975 (CVCL_1511), H358 (CVCL_1559), H820 (CVCL_1592), H1650 (CVCL_1483), and H2228 (CVCL_1543) were obtained from the American Type Culture Collection (Manassas, VA, USA). Human microvascular endothelial cells were obtained from LONZA (Basel, Switzerland). Human stem cells (Human Cord Blood CD34⁺ cell) were obtained from Stem Cell Technologies (Vancouver, Canada). A549 cells were cultured in Dulbecco’s modified Eagle’s medium (Wako, 043-30085), whereas other cell lines were cultured in RPMI medium supplemented with 10% fetal bovine serum (Sigma10270106) and 1% antibiotic-antimycotic (gibco, 15240062). The cells were cultured at 37 ℃ in a humidified atmosphere containing 5% CO₂ and harvested at 80% confluency. The cells were treated with 5 ng/mL transforming growth factor-β (TGF-β) (PeproTech, 100-21) for one month to induce EMT. The medium was changed once every three days. The cells were immunostained with anti-CK antibodies [anti-panCK eFluor615 antibody (clone: AE1/AE3) (# 42-9003-82, RRID: AB_10804663) and anti-CK8/18 FITC antibody (clone: CK3-6H5) (130-118-964, RRID: AB_2784333)]. The intensity of each antibody was evaluated. The quantum yield is 0.65 and 0.92, and the molar extinction coefficient is 190,000 M-1cm-1 and 75,000 M-1cm-1 for eF617 and FITC, respectively. The brightness of fluorescence is expressed by multiplying the quantum yield by the molar absorption coefficient. The Fluorescence of each antibody was confirmed with an isotype control antibody. The fluorescence intensity of immunostaining was quantified using a well-established method previously employed by our group (12).

Clinical evaluation of patients with NSCLC

A total of 20 patients who were diagnosed with advanced NSCLC at the National Cancer Center Hospital East and three healthy donors were included in this study. The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the committee boards of Sysmex (No. 2021-042) and National Cancer Center Hospital East (No. 2018-180), and informed consent was obtained from all individual participants. Whole blood samples were collected in Cell-Free DNA BCT tubes (Streck, STM-230570) and delivered to the Sysmex laboratory in Kobe, Japan.

CTC enrichment

Blood samples were introduced into a microfluidic chip using a syringe pump (Harvard, PHD ULTRA-IW) and a Medallion syringe (Merit Medical, MSS111E) at a flow rate of 500 or 700 µL·min⁻¹. The sheath buffer was introduced using glass syringes (Hamilton, 86020) at a flow rate of 900–1,500 µL·min⁻¹. All solutions were introduced via Tygon tubing, which was inserted directly into the input and output ports of the chip. The syringe pump was placed vertically with the syringe outlet in the direction of gravity during blood injection to reduce the effect of blood cell sedimentation. Solution collection from all outlets in tubes and bottles was started immediately after flow stabilization. At the end of each run, all syringe pumps were manually stopped after confirming that the sample syringe was empty.

CTC detection using an imaging flow cytometer

After CTC enrichment, the isolated cells were fixed in 2% paraformaldehyde for 15 min and permeabilized with 70% methanol on ice for 10 min. Cells were blocked with 5% bovine serum albumin and Fc blocker (BD Biosciences, 564220) for 15 min to prevent nonspecific staining. Cells were stained with Hoechst, antibodies against CD45 (BioLegend Cat. No. 304048, RRID: AB_2563129), CD16 (BioLegend Cat. No. 302046, RRID: AB_2563803), CD34 (BioLegend Cat. No. 343626, RRID: AB_2632727), CK (including panCK and CK8/18), programmed death ligand 1 (PD-L1) (BioLegend Cat. No. 374514, RRID: AB_2734442), and vimentin (sc-6260) for 30 min. The primary antibodies used and conditions are listed in Table S1. Staining analysis was performed using an imaging flow cytometer under the following conditions: magnification of the objective lens, 40×; flow rate, middle; and extended depth filter, on. The laser power was adjusted to the following settings: 100 mW for a wavelength of 405 nm, 20 mW for 488 nm, 250 mW for 592 nm, and 130 mW for 642 nm. To establish appropriate cutoff values for CK8/18 or panCK, we utilized cells derived from three healthy donors and determined threshold values that ensured no positivity for CK8/18 or panCK. These established cutoff values enabled specific detection while minimizing false-positive signals (Figure S1). CTCs were defined as CK-positive, Hoechst-positive, and CD45/CD16/CD34-negative cells. Specifically, panCK-positive CTCs were defined as cells positive for panCK and Hoechst and negative for CD45/CD16/CD34, whereas CK8/18-positive CTCs were defined as cells positive for CK8/18 and Hoechst and negative for CD45/CD16/CD34.

Detection of single-cell gene mutations

After CTC enrichment, the isolated cells were fixed, permeabilized, and blocked. Single-cell sorting into individual wells of a 384-well plate was performed using FACS Aria II (BD Biosciences, San Jose, CA, USA) with specific operational parameters set as follows: flow rate, 1; ND filter, 1.5; and sort precision, single-cell. Then, whole genome amplification was performed on the sorted single cells using a PicoPLEX Single-Cell WGA kit version 3 (Takara Bio, R300722). Whole genome amplification extracts were purified using Agencourt AMPure XP (Beckman Coulter, A63881). Polymerase chain reaction (PCR) was performed, targeting EGFR exon 19 deletion or L858R mutation with specific primers. PCR was performed under the following conditions: initial denaturation at 98 ℃ for 30 s, denaturation at 98 ℃ for 10 s, annealing at 60 ℃ for 10 s, extension at 72 ℃ for 15 s, and final extension at 72 ℃ for 5 min. The denaturation, annealing, and extension cycles were repeated 40 times. The PCR products were purified using ExoSAP-IT Express PCR Product Cleanup Reagent (Thermo Fisher Scientific). The sequences were analyzed using the Sanger method. Sequence analysis was performed using GENETYX version 15.0.0 (Genetyx, Tokyo, Japan).

Statistical analysis

Data acquired from the imaging flow cytometer were analyzed using IDEAS software (Amnis, Seattle, WA, USA). Resolution was calculated by dividing the difference between the median fluorescence intensity of healthy non-blood cells (Hoechst-positive, and CD45/CD16/CD34-negative cells) such as mesenchymal stem cell (13), or the negative and positive controls by the mean value of each peak width. Peak width was calculated as four times the standard deviation. Statistical analyses were performed using specific tests. The numbers of panCK-positive and CK8/18-positive CTCs were compared using the paired t-test, whereas the numbers of CK8/18-positive CTCs were compared between patients with EGFR mutations and those with EGFR wild-type using the two-sample t-test.

Results

Study workflow

This workflow outlines the sequential steps undertaken in the study, beginning with the evaluation of antibodies in cell lines, then the comparison of CK8/18 and panCK in patients with NSCLC, and finally, the assessment of genetic mutations in patients with EGFR-mutated NSCLC.

Evaluation of anti-CK antibodies in cell lines before and after EMT

In the initial phase, various anti-CK antibodies were evaluated using established cell lines to identify antibodies capable of detecting CTCs undergoing EMT. CK8/18 emerged as a promising candidate, demonstrating its potential for detecting CTCs even in EMT states (Figure 1A).

Figure 1 Study workflow. (A) Lung cancer cell lines were treated with 5 ng/mL TGF-β for 1 month to induce EMT. Harvested cells were stained with antibodies and measured with an imaging flow cytometer. (B) CTCs were enriched by hybrid double-spiral chip from 4.5 mL blood in NSCLC patients. CTCs were stained with antibodies and measured with an imaging flow cytometer. (C) CTCs were enriched by hybrid double-spiral chip from 13.5 mL blood in NSCLC patients and stained with antibodies. Single cells were sorted by FACS Aria II and confirmed EGFR mutation. CK, cytokeratin; CTC, circulating tumor cell; EGFR, epidermal growth factor receptor; EMT, epithelial-mesenchymal transition; NSCLC, non-small cell lung cancer; PCR, polymerase chain reaction; TGF-β, transforming growth factor β; WGA, whole genome amplification.

Comparison of the utility of panCK and CK8/18 in patients with NSCLC

A total of 20 patients with advanced NSCLC (sample Nos. 1–20) were included to evaluate the utility of CK8/18 compared with panCK using our CTC detection system. The aim was to assess and compare the efficacy of CK8/18 and panCK in detecting CTCs in blood samples from patients with advanced NSCLC (Figure 1B).

Single-cell gene mutation analysis of CK8/18-positive CTCs in patients with EGFR-mutated NSCLC

A subset of patients with NSCLC harboring EGFR mutations (sample Nos. 11–20) was selected to further verify the origin of CK8/18-positive CTCs. Genetic mutations in CTCs were evaluated, with a focus on EGFR mutation, to establish an identity of origin between CK8/18-positive CTCs and cancer lesions (Figure 1C).

Evaluation of anti-CK antibodies in cell lines before and after EMT

The staining patterns of anti-panCK and anti-CK8/18 antibodies in EMT-induced lung cancer cell lines were compared. EMT was induced in HCC827 with TGFβ treatment. EMT induction was confirmed by increased vimentin expression (Figure S2). Cells were stained with anti-panCK and anti-CK8/18 antibodies. Cutoff thresholds were established, ensuring that cells from healthy donors exhibited no positivity for CK8/18 or panCK. The cutoff values were 2,530 and 2,000 for panCK and CK8/18, respectively (Figure 2A). The panCK-positive rate was 98.8% before EMT induction and decreased to 42.9% after induction. Conversely, the CK8/18 positivity rate remained constant (99.9% before EMT induction and 99.7% after induction) (Figure 2A). Notably, after EMT induction, over 50% of cells exhibited a lack of panCK positivity, whereas the CK8/18 positivity rate remained relatively stable despite a decrease in CK8/18 expression. Representative cases are presented in Figure 2B. These findings indicate the efficacy of CK8/18 in detecting cells after EMT.

Figure 2 Comparison of panCK with CK8/18 staining in cell lines. (A) Fluorescence intensity of panCK (left) and CK8/18 (right) staining before and after induction of EMT with TGF-β (5 ng/mL for 1 month) in HCC827. Green line is the cutoff threshold at which no positive cells are detected in healthy donors. Blue and red drawing shows before and after EMT induction, respectively. (B) Brightfield and fluorescence images are shown. (C) Fluorescence intensity of panCK (left) and CK8/18 (right) in lung cancer cell lines (H1975, HCC827, A549, H1299, H358, H820, H1650, and H2228), primary cells (HMVEC, HSC), and peripheral marker negative non-blood cells from healthy humans (HD). (+) means after EMT induction and (−) means before EMT induction. Boxplots represent the median (dark horizontal line), with the box representing the 25^th and 75^th percentiles, the whiskers the 5^th and 95^th percentiles. a.u., arbitrary unit; CK, cytokeratin; EMT, epithelial-mesenchymal transition; HD, healthy donor; HMVEC, human microvascular endothelial cell; HSC, human stem cell; TGF-β, transforming growth factor β.

CK8/18 staining was evaluated in various lung cancer cell lines (H1975, HCC827, A549, H1299, H358, H820, H1650, and H2228), primary cells (human mammary vascular endothelial cells derived from the umbilical cord and hematopoietic stem cells), and nonblood cells from healthy donors. The intensities of panCK and CK8/18 in cell lines, primary cells, and nonblood cells of healthy donors are shown in Figure 2C. A comparison of the intensity of lung cancer cell lines and primary cells showed that the resolution of panCK was 0.20, whereas that of CK8/18 was 0.38. Similarly, the comparison between lung cancer cell lines and cells from healthy donors showed that the resolution of panCK was 0.27, whereas that of CK8/18 was 0.38. These findings indicate that CK8/18 had a superior resolution for identifying lung cancer cells. These results showed the utility of CK8/18 in detecting CTCs, regardless of the presence or absence of EMT in patients with lung cancer.

Clinical evaluation of patients with NSCLC

The usefulness of our novel CTC detection system using the anti-CK8/18 antibody

The baseline characteristics of the 20 patients included in this study are shown in Table 1. Of the 20 patients, 18 had stage IV NSCLC, and 2 (10%) had recurrence after surgery. Among the enrolled patients, 10 were EGFR mutation-negative, whereas the remaining 10 were EGFR mutation-positive. Additionally, PD-L1 expression was evaluated in 14 patients, accounting for 70% of the total cohort.

PanCK-positive CTCs were detected in only 8 patients (40%) (1 CTC in 4.5 mL of the whole blood), whereas CK8/18-positive CTCs were detected in 17 patients (85%) (1–66 CTCs in 4.5 mL of the whole blood). The detection rates of CK8/18-positive CTCs were significantly higher than those of panCK-positive CTCs (P<0.01). The number of CK8/18-positive CTCs was notably higher than that of panCK-positive CTCs in all patients (P<0.01) (Figure 3A). Table 2 shows the number of CTCs. Representative cases are presented in Figure 3B. The number of CTCs was significantly lower in patients with EGFR mutations than in those with EGFR wild-type (median number of CTCs, 3 vs. 22, P<0.01). All detected panCK-positive CTCs also exhibited CK8/18 positivity. However, CK8/18-positive CTCs were not detected in three patients (Nos. 8, 13, and 20). All of them had adenocarcinoma pathology. However, no consistent characteristics were observed regarding the tumor-node-metastasis (TNM) stage or smoking history between the patients. These findings indicate the utility of CK8/18 in detecting CTCs in patients with lung cancer.

Figure 3 The number of panCK-positive CTCs and CK8/18-positive CTCs detected among 20 lung cancer patients. (A) The number of CTCs detected using anti-panCK or anti-CK8/18 antibodies. The dots connected by lines indicate CTCs detected in the same patient (**, P<0.01). (B) Brightfield and fluorescence images of WBCs and CTCs acquired by an imaging flow cytometry. CK, cytokeratin; CTC, circulating tumor cell; WBC, white blood cell.

Identification of CTC gene mutations by single-cell analysis

Single-cell gene mutation analysis was performed on the 10 patients with lung cancer with EGFR mutations (sample Nos. 11–20) to confirm the lung cancer origin of the isolated CK8/18-positive CTCs. In the preliminary study using H1650 lung cancer cells, the sorting of single cells accounted for 20%, and the success rate of PCR in single cells was 60%, culminating in a cumulative total of 12%. CK8/18-positive and CD45/CD16/CD34-negative cells were sorted using this system, and the EGFR gene was amplified by PCR and subsequently sequenced. A total of 15 bases were deleted in patients with exon 19 deletions. Thymine was replaced with guanine in patients with the L858R mutation (Figure 4A). Among the 10 patients with EGFR mutations, CTCs carrying EGFR mutations were detected in 6 (60%), all of which matched the mutations identified in the corresponding primary tumors (Figure 4B). However, EGFR mutations were not detected in the remaining four patients. These results support the conclusion that CK8/18-positive CTCs are derived from lung cancer lesions.

Figure 4 EGFR mutation analysis in tissue and single CTC. (A) Sequence of EGFR mutation sites (red boxes) and their surrounding areas detected in CK8/18-positive CTCs sorted from lung cancer patient samples. (B) Mutation analysis results in 10 patients (#11–20) with known to be harboring EGFR mutations. CK, cytokeratin; CTC, circulating tumor cell; EGFR, epidermal growth factor receptor; Fw, forward; MT, mutation; Rv, reverse; WT, wildtype.

Comparison analysis between PD-L1 expression in CTCs and tissue samples

PD-L1 expression in tissues is a known biomarker of immune checkpoint inhibitor therapy. A comparison analysis was performed between PD-L1 expression levels in CTCs and tissue samples to determine the consistency between PD-L1 expression in CTCs and tissue samples (Table S2). The data obtained from the CTC detection system included the number of PD-L1-positive CTCs, PD-L1 positivity within CTCs, and median PD-L1 fluorescence intensity of the CTCs. A threshold of >1% PD-L1 expression in tissues was classified as positive, whereas a threshold of <1% was deemed negative. The tissue samples that were PD-L1-negative demonstrated PD-L1 positivity in CTCs and high median PD-L1 intensity in CTCs. However, an association between PD-L1-positive CTCs and PD-L1 expression in tissue samples could not be conclusively established because of the small number of samples.

Discussion

To the best of our knowledge, this is the first study to compare panCK with CK8/18 and demonstrate the superior efficacy of CK8/18 over panCK in detecting CTCs in patients with NSCLC and establishing an unequivocal correlation between genetic abnormalities found in CTCs and those present in tissue samples, thereby confirming their origin within the tumor. Although previous study has investigated the utility of CK18 for detecting CTCs in various malignancies, no studies, to our knowledge, have focused exclusively on the utility of CK8/18 in lung cancer. Cell line investigations showed that CK8/18 exhibited a higher positivity rate than panCK, regardless of the presence or absence of EMT. Furthermore, analysis of blood samples from patients with NSCLC revealed that panCK-positive CTCs were detected in only 40% of the patients, whereas CK8/18-positive CTCs were identified in 85%. Additionally, single-cell gene mutation analysis corroborated the origin of CK8/18-positive CTCs in lung cancer tissues.

The anti-panCK antibody, a cocktail antibody, can detect multiple CKs, including CK1, 2, 3, 4, 5, 6, 7, 8, 10, 14, 15, 16, and 19, and the anti-CK8/18 antibody can identify CK7, 8, 18, and 19. However, the anti-panCK antibody probably has a lower affinity for CK7, 8, and 19 than the anti-CK8/18 antibody, and CK18 expression should be dominant. Since all panCK-positive CTCs were also CK8/18-positive, targeting CK8/18 did not overlook panCK-positive CTCs. Hence, CK8/18 is a valuable alternative to panCK for CTC detection in NSCLC. This study revealed a notable reduction in CK expression upon EMT induction (Figure 1). However, the degree of CK expression reduction varied depending on the CK type. The expression of panCK was strongly reduced and became negative after EMT induction, whereas that of CK8/18 exhibited only a slight reduction and maintained high positivity after EMT induction. These findings are similar to those of a previous study (11). The detection rates of CK8/18-positive and panCK-positive CTCs were 85% and 40%, respectively. These notable differences might have occurred because of the higher prevalence of EMT in NSCLC. This result affirms that the use of the anti-CK8/18 antibody notably enhanced the CTC detection rate in NSCLC. However, for a more comprehensive CTC detection approach in the future, the validation of other epithelial or mesenchymal markers and combining markers for CTC detection may be useful. In this study, vimentin was used as a mesenchymal marker. However, no difference in expression was observed between healthy donors and patients with lung cancer. Therefore, vimentin is not useful for CTC detection in NSCLC (Figure S3).

Our system demonstrated a detection sensitivity of 85% for CK8/18 positive CTC in patients with NSCLC. This performance represents a marked improvement compared to the approximately 20% sensitivities typically observed for NSCLC CTC detection using conventional technologies based on the EpCAM (5). We hypothesize that this significant disparity in sensitivity stems from the inherent limitation of EpCAM-dependent methods, which often fail to capture CTCs that have undergone EMT and consequently exhibit reduced or lost EpCAM expression (5). In contrast, our system utilizes a size-based separation technique, enabling the isolation of CTCs irrespective of their EMT status. This capability to capture EMT-phenotype CTCs likely contributes to the enhanced sensitivity observed with our approach (9). Supporting this hypothesis, a previous study employing this system in patients with stage IV colorectal cancer successfully detected vimentin-positive CTCs, which are considered indicative of an EMT phenotype, with 100% sensitivity (14). Furthermore, our subsequent identification employs immunostaining for CK8/18. This marker is recognized for their relatively stable expression even post-EMT; importantly, previous reports have documented CK8/18 expression even in tumor cells exhibiting low or negative EpCAM levels (11). It is important to acknowledge, however, that the field of CTC detection is continuously evolving, with the development of alternative strategies that utilize neither EpCAM nor size-based selection. For example, detection methods targeting Heat Shock Protein 70 (Hsp70) have also emerged as a promising approach (15). Therefore, comprehensive comparative studies are warranted in the future to rigorously evaluate the relative strengths and weaknesses of these diverse CTC detection platforms across various clinical contexts. Previous studies have revealed that EpCAM-positive CTCs are derived from tissue in bulk systems (16,17). However, evidence that CK8/18-positive cells are derived from lung cancer tissue is lacking. This study demonstrated that gene mutations identified in CK8/18-positive CTCs identified through single-cell sorting precisely aligned with those identified in tissue samples from patients with confirmed EGFR mutations. This robust result confirmed that CK8/18-positive cells were derived from tumor tissues. However, it is noteworthy that not all patients exhibited genetic mutations in their CTCs, with a detection rate of 60%. This discrepancy may be attributed to limitations in the recovery rate of single-cell gene mutation detection systems. The rarity of the specific cell population being sorted may contribute to the challenge. In this study, the single-cell sorting method demonstrated a relatively low recovery rate of 20%. The diminished recovery rate may be due to the complexity of charging and sorting cells into designated wells. During this process, it is conceivable that CTC loss and potential contamination with non-target cells occurred, leading to decreased detection of aligned EGFR mutations.

In this study, no significant differences in the number of CTCs were observed based on patient clinical stage (IVA vs. IVB), sex, or age. There are conflicting reports on the relationship between CTC number and clinical stage. Some studies have reported a linear correlation between CTC number and clinical stage, whereas others have not. This issue is still in dispute. Thus, further investigations are needed. However, in this study, notably fewer CTCs were detected in patients without EGFR mutation. Furthermore, an association between the number of CTCs and EGFR mutations was reported in a previous study (4). Wei et al. reported an increased number of CTCs in patients with NSCLC harboring EGFR mutations compared with those without EGFR mutations, but the reason was not mentioned. This discrepancy between our results and those of previous studies may be due to differences in detection systems and markers. Moreover, no clear relationship between PD-L1 expression in CTCs and tissue samples was established in this study. This could be attributed to the limited measurable PD-L1 expression observed in only 14 out of the total number of patients in tissue and four out of all patients in CTCs. Similar to our findings, several studies investigating the correlation between PD-L1 expression in CTCs and tissue samples have reported a lack of significant correlations (18-22). These findings indicate that CTCs more accurately reflect the heterogeneity of PD-L1 expression in tissues. Several studies have reported an association between CTCs and treatment efficacy or survival benefit, indicating their potential use in predicting prognosis (23,24). Notably, not all patients received any chemotherapy, the treatment regimens varied among patients, and the number of patients was small. Consequently, we could not effectively assess the relationship between the number of CTCs and treatment outcomes in our cohort.

This study has some limitations. First, a limited number of patients were included in this study, which might affect the generalizability and broader applicability of the findings. Second, genetic testing was performed only in EGFR-positive patients, which might have introduced a bias in the sensitivity of the CTC detection rate. Third, this study exclusively used a CK8/18-based system and did not incorporate an EpCAM-based system, which may restrict the direct comparison of our findings with conventional CTC detection methodologies that rely on EpCAM. A direct comparative analysis using EpCAM-based systems can provide valuable insights into the performance and efficacy of our CK8/18-based approach for detecting CTCs in patients with NSCLC. We are considering evaluating CTCs detected by our system using EpCAM in future studies. Fourthly, our novel system, utilizing CK8/18 was tested exclusively on specimens from patients with NSCLC. Consequently, its utility for detecting CTCs in other types of malignancies could not be evaluated. Given that CK expression patterns are known to vary among different cancer types, the universal applicability of targeting CK8/18 for CTC detection requires further investigation

Conclusions

In conclusion, this study successfully identified and validated suitable biomarkers for the identification of lung cancer CTCs. Particularly, the CK8/18 marker proved more effective in detecting CTCs than panCK. Furthermore, the presence of CK8/18-positive CTCs was confirmed to originate from lung cancer lesions.

Acknowledgments

The authors would like to thank Enago (www.enago.jp) for the English language review. This study was presented in part at the American Association of Cancer Research 2023.

Footnote

Reporting Checklist: The authors have completed the STARD reporting checklist. Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-155/rc

Data Sharing Statement: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-155/dss

Peer Review File: Available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-155/prf

Funding: None.

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tlcr.amegroups.com/article/view/10.21037/tlcr-2025-155/coif). T.S. received a research grant from Amgen, Daiichi Sankyo, GlaxoSmithKline K.K and NEC; and payment or honoraria from AstraZeneca, Chugai, Novartis, MSD, Thermo Fisher Scientific, Takeda, OLYMPUS, Taiho, Merck, Daiichi Sankyo, RIKEN GENESIS and Amco. Yoshitaka Zenke received a research grant from AstraZeneca, Daiichi-Sankyo, Amgen, GSK, Roche, Chugai, MSD, Merck and consulting fees from AstraZeneca, Boheringer-Ingelheim and honoraria from AstraZenca, Lilly, Chugai, Ono, Bristol-Meyers Squibb, Takeda, Boheringer-Ingelheim, Taiho, MSD, Novartis, Pfizer, Nihon Kayaku, Kyowa Kirin, Amgen. E.M.Y., Y.I., K.K., T.Y., S.I. and M.Y. are employees of Sysmex Corporation and have a company stock. M.O., F.K., T.S. and B.D.M. are employees of Sysmex Corporation. A.A.S.B. is an employee of Biolidics Ltd. and an employee of National University of Singapore and advisor of Biolidics Ltd., and received consulting fee from Biolidics Ltd. K.G. received a research grant from Amgen Inc., Amgen K.K., AstraZeneca K.K., AbbVie GK, AnHeart Therapeutics Inc., Bayer Yakuhin, Ltd., Nippon Boehringer Ingelheim Co., Ltd., Bristol-Myers Squibb K.K., Blueprint Medicines Corporation., CHUGAI PHARMACEUTICAL Co., Ltd., Craif Inc., DAIICHI SANKYO COMPANY, LIMITED, Eisai Co., Ltd., Eli Lilly Japan K.K., Guardant Health Asia, Middle East & Africa, Inc., Haihe Biopharma Co., Ltd., Ignyta, Inc., Janssen Pharmaceutical K.K., Kyowa Kirin Co., Ltd., Life Technologies Japan Ltd., Loxo Oncology, Inc., Lunit Inc., MEDICAL& BIOLOGICAL LABORATORIES Co., Ltd., Merck Biopharma Co., Ltd., Merus N.V., MSD K.K., Novartis Pharma K.K., ONO PHARMACEUTICAL Co., Ltd., Pfizer R&D Japan G.K., Precision Medicine Asia Co., Ltd., RIKEN GENESIS Co., Ltd., Sumitomo Pharma Co., Ltd., Spectrum Pharmaceuticals, Inc., Sysmex Corporation., Taiho Pharmaceutical Co., Ltd., Takeda Pharmaceutical Co., Ltd., Turning Point Therapeutics, Inc., and honoraria from Amgen K.K., Amoy Diagnosties Co., Ltd., AstraZeneca K.K., Bayer U.S., Bristol-Myers Squibb K.K., CHUGAI PHARMACEUTICAL Co., Ltd., DAIICHI SANKYO COMPANY, LIMITED, Eisai Co., Ltd., Eli Lilly Japan K.K., Guardant Health Japan Corp., iTeos Therapeutics Inc., Janssen Pharmaceutical K.K., Thermo Fisher Scientific K.K., Merck Biopharma Co., Ltd., Nippon Kayaku Co.,Ltd., Novartis Pharma K.K., ONO PHARMACEUTICAL Co., Ltd., Pharma Mar, S.A., RIKEN GENESIS Co., Ltd., Taiho Pharmaceutical Co., Ltd., Takeda Pharmaceutical Co., Ltd. and participation on a Data Safety Monitoring Board or Advisory Board from Amgen Inc., Amgen K.K., AstraZeneca K.K., Bayer HealthCare Pharmaceuticals Inc., Bristol-Myers Squibb K.K., DAIICHI SANKYO COMPANY, LIMITED, Eli Lilly Japan K.K., GlaxoSmithKline K.K., Haihe Biopharma Co., Ltd., Janssen Pharmaceutical K.K. and SYNEOS HEALTH CLINICAL K.K. The authors have no other conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.The study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The study was approved by the committee boards of Sysmex (No. 2021-042) and National Cancer Center Hospital East (No. 2018-180), and informed consent was obtained from all individual participants.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Ghossein RA, Bhattacharya S, Rosai J. Molecular Detection of Micrometastases and Circulating Tumor Cells in Solid Tumors. Available online: http://aacrjournals.org/clincancerres/article-pdf/5/8/1950/2301073/df089901950.pdf
2. Chaffer CL, Weinberg RA. A Perspective on Cancer Cell Metastasis. Available online: www.sciencemag.org
3. Cohen SJ, Punt CJ, Iannotti N, et al. Relationship of circulating tumor cells to tumor response, progression-free survival, and overall survival in patients with metastatic colorectal cancer. J Clin Oncol 2008;26:3213-21. [Crossref] [PubMed]
4. Wei T, Zhu D, Yang Y, et al. The application of nano-enrichment in CTC detection and the clinical significance of CTCs in non-small cell lung cancer (NSCLC) treatment. PLoS One 2019;14:e0219129. [Crossref] [PubMed]
5. de Wit S, van Dalum G, Lenferink AT, et al. The detection of EpCAM(+) and EpCAM(-) circulating tumor cells. Sci Rep 2015;5:12270. [Crossref] [PubMed]
6. Allard WJ, Matera J, Miller MC, et al. Tumor Cells Circulate in the Peripheral Blood of All Major Carcinomas but not in Healthy Subjects or Patients With Nonmalignant Diseases. Available online: http://clincancerres.aacrjournals.org
7. Hanssen A, Wagner J, Gorges TM, et al. Characterization of different CTC subpopulations in non-small cell lung cancer. Sci Rep 2016;6:28010. [Crossref] [PubMed]
8. Descamps L, Le Roy D, Deman AL. Microfluidic-Based Technologies for CTC Isolation: A Review of 10 Years of Intense Efforts towards Liquid Biopsy. Int J Mol Sci 2022;23:1981. [Crossref] [PubMed]
9. Shirai K, Guan G, Meihui T, et al. Hybrid double-spiral microfluidic chip for RBC-lysis-free enrichment of rare cells from whole blood. Lab Chip 2022;22:4418-29. [Crossref] [PubMed]
10. Dent BM, Ogle LF, O'Donnell RL, et al. High-resolution imaging for the detection and characterisation of circulating tumour cells from patients with oesophageal, hepatocellular, thyroid and ovarian cancers. Int J Cancer 2016;138:206-16. [Crossref] [PubMed]
11. Kodack DP, Farago AF, Dastur A, et al. Primary Patient-Derived Cancer Cells and Their Potential for Personalized Cancer Patient Care. Cell Rep 2017;21:3298-309. [Crossref] [PubMed]
12. Isozaki A, Kita K, Tiffany Ishii N, et al. Investigating T-Cell Receptor Dynamics Under In Vitro Antibody-Based Stimulation Using Imaging Flow Cytometry. Cytometry A 2025;107:88-97. [Crossref] [PubMed]
13. Nauta AJ, Kruisselbrink AB, Lurvink E, et al. Mesenchymal Stem Cells Inhibit Generation and Function of Both CD34-Derived and Monocyte-Derived Dendritic Cells. J Immunol 2006;177:2080-7. [Crossref] [PubMed]
14. Takahashi Y, Ijiri Y, Fujino S, et al. Detection and Characterization of Circulating Tumor Cells in Colorectal Cancer Patients via Epithelial-Mesenchymal Transition Markers. Cancers (Basel) 2025;17:303. [Crossref] [PubMed]
15. Xanthopoulos A, Samt AK, Guder C, et al. Hsp70-A Universal Biomarker for Predicting Therapeutic Failure in Human Female Cancers and a Target for CTC Isolation in Advanced Cancers. Biomedicines 2023;11:2276. [Crossref] [PubMed]
16. Maheswaran S, Sequist LV, Nagrath S, et al. Detection of mutations in EGFR in circulating lung-cancer cells. N Engl J Med 2008;359:366-77. [Crossref] [PubMed]
17. Marchetti A, Del Grammastro M, Felicioni L, et al. Assessment of EGFR mutations in circulating tumor cell preparations from NSCLC patients by next generation sequencing: toward a real-time liquid biopsy for treatment. PLoS One 2014;9:e103883. [Crossref] [PubMed]
18. Janning M, Kobus F, Babayan A, et al. Determination of PD-L1 Expression in Circulating Tumor Cells of NSCLC Patients and Correlation with Response to PD-1/PD-L1 Inhibitors. Cancers (Basel) 2019;11:835. [Crossref] [PubMed]
19. Guibert N, Delaunay M, Lusque A, et al. PD-L1 expression in circulating tumor cells of advanced non-small cell lung cancer patients treated with nivolumab. Lung Cancer 2018;120:108-12. [Crossref] [PubMed]
20. Zhou Q, Liu X, Li J, et al. Circulating tumor cells PD-L1 expression detection and correlation of therapeutic efficacy of immune checkpoint inhibition in advanced non-small-cell lung cancer. Thorac Cancer 2023;14:470-8. [Crossref] [PubMed]
21. Sinoquet L, Jacot W, Gauthier L, et al. Programmed Cell Death Ligand 1-Expressing Circulating Tumor Cells: A New Prognostic Biomarker in Non-Small Cell Lung Cancer. Clin Chem 2021;67:1503-12. [Crossref] [PubMed]
22. Dall'Olio FG, Gelsomino F, Conci N, et al. PD-L1 Expression in Circulating Tumor Cells as a Promising Prognostic Biomarker in Advanced Non-small-cell Lung Cancer Treated with Immune Checkpoint Inhibitors. Clin Lung Cancer 2021;22:423-31. [Crossref] [PubMed]
23. Wang Z, Zhang XC, Feng WN, et al. Circulating tumor cells dynamics during chemotherapy predict survival and response in advanced non-small-cell lung cancer patients. Ther Adv Med Oncol 2023;15:17588359231167818. [Crossref] [PubMed]
24. Tong B, Xu Y, Zhao J, et al. Prognostic role of circulating tumor cells in patients with EGFR-mutated or ALK-rearranged non-small cell lung cancer. Thorac Cancer 2018;9:640-5. [Crossref] [PubMed]