Abstract
Background/Aim: Circulating tumor cells (CTCs) may be affected by the environment encountered during blood circulation. We aimed to explore the association between the molecular phenotype of CTCs and systemic inflammatory markers. Patients and Methods: CTCs isolated from patients with recurrent/metastatic head and neck squamous cell carcinoma by CD45-negative selection were analyzed for the expression of multiple genes. The correlations between gene expression levels in CTCs and systemic inflammation markers were examined. Results: Thirty-five (83.3%) of the 42 patients were positive for CTCs. No significant differences in systemic inflammatory markers were observed between CTC-positive and CTC-negative patients. Notably, VIM or ZEB2 expression was strongly correlated with that of CD44 or ALDH1. PIK3CA, CD44, ALDH1A1, and PDCD1LG2 expression in CTCs was correlated with lymphocyte- and/or albumin-related systemic inflammatory markers. Conclusion: CTCs acquire a survival advantage through phenotypic alterations in the hostile blood environment, and evade circulatory immune surveillance.
- Circulating tumor cell
- head and neck squamous cell carcinoma; blood microenvironment
- molecular phenotype
- systemic inflammatory marker
Similar to primary tumors and metastatic lesions, circulating tumor cells (CTCs) that have detached from a primary tumor into the bloodstream are also heterogeneous tumor cell populations (1, 2). CTC migration to the bloodstream exposes them to a hostile environment; in response, some subpopulations may undergo necrosis, anoikis, or apoptosis by physical forces and/or immune surveillance, while others survive by acquiring resistance mechanisms. Travelling into the blood circulation may affect CTC characteristics through direct and/or indirect interactions with various blood cells and their secreted humoral factors (3). For instance, the association between neutrophils and CTCs drives cell cycle progression within the bloodstream and expands the metastatic potential of CTCs (4). Similarly, platelets can protect CTCs against not only physical stress but also antitumor immune responses (3, 5). Moreover, platelet-CTC interactions induce an epithelial-mesenchymal-like transition and promote metastasis by platelet activation and the release of soluble mediators (6). Thus, in the blood environment, the phenotypic heterogeneity and plasticity of CTCs could be shaped by complex interaction networks between CTCs and circulating blood cells.
Cancer-related inflammation is a hallmark of cancer (7). With disease progression and/or selective pressure due to cancer treatment, the blood environment, specifically the systemic inflammation status, is drastically altered. Tumor cells can activate or inhibit a variety of blood cells and form complicated patient-specific inflammatory states in both the tumor microenvironment and systemic circulation, which are related to clinical outcomes including disease progression, treatment response, and prognosis (8-10).
The clinical significance of systemic inflammation markers based on counts, ratios, or the scores of circulating blood cells or acute-phase proteins, such as neutrophil-to-lymphocyte ratio (NLR), lymphocyte-to-monocyte ratio (LMR), and a measure based on elevated serum C-reactive protein (CRP) level and decreased serum albumin level, has been extensively investigated (10). Similar to other types of cancers, inflammatory markers may be useful markers for predicting the clinical outcomes of patients with head and neck squamous cell carcinoma (HNSCC). Rassouli et al. showed that the platelet-to-lymphocyte ratio (PLR) and NLR are independent predictors of mortality and recurrence, respectively (11). Similarly, LMR is an independent prognostic factor for HNSCC (12).
We previously reported that 24 (80.0%) of 30 patients with recurrent/metastatic (R/M) HNSCC were positive for CTCs and measured the expression of immune-regulatory molecules in the 24 CTC-positive patients (13). In the present study, we molecularly characterized the CTCs in 42 patients including those of 12 who were newly diagnosed with R/M HNSCC. Finally, we investigated the association between the CTC molecular phenotype and systemic inflammation in patients with R/M HNSCC.
Patients and Methods
Patients. This study enrolled a total of 42 patients with R/M HNSCC, including 30 patients described previously (13). Their characteristics are shown in Table I. The median age of the patients was 69 years (range=48-86 years). The tumor origins included the oral cavity (n=5), nasopharynx (n=1), oropharynx [n=5 (2 were p16 positive; 3 were p16 negative)], hypopharynx (n=19), larynx (n=5), paranasal sinuses (n=6), and parotid gland (n=1). Recurrences in all patients, including the overlapping cases, were observed in 17 (40.5%), 25 (59.5%), and 23 (54.8%) local, regional, and distant sites, respectively. This study was approved by the Ethical Committee of the Gunma University Hospital (No.12-12), and written informed consent was obtained from each patient.
Circulating tumor cell (CTC) isolation and gene expression analysis. CTC isolation and gene expression analysis were performed as described previously (13). In brief, peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Hypaque density gradient centrifugation of blood samples (7.5 ml) obtained from patients. The contaminating erythrocytes were further lysed with red blood cell lysis buffer (Roche Diagnostic GmbH, Mannheim, Germany). The cell suspension was incubated with a human CD45 depletion cocktail for 15 min and subsequently with magnetic particles for 10 min (EasySep™ Human CD45 Depletion Kit, Stemcell Technologies). Tubes containing the PBMCs were placed in a magnet for 10 min twice, and the unbound cells (CTCs) were transferred to new tubes.
Total RNA from the CTCs was extracted using a RNeasy Micro kit (QIAGEN, Hilden, Germany) according to the manufacturer’s instructions. cDNA synthesis was performed using the QuantiTect Reverse Transcription kit (QIAGEN) with 14 cycles of pre-amplification using the TaqMan™ PreAmp Master Mix kit (Applied Biosystems). The pre-amplified products were subsequently analyzed by real-time quantitative polymerase chain reaction (Applied Biosystems, Waltham, MA, USA) for the 16 target genes. Sixteen primers for the 15 targets [epithelial cell adhesion molecule (EPCAM), MET, keratin 19 (KRT19), epidermal growth factor receptor (EGFR), phosphatidylinositol-4,5-bisphosphate 3-kinase catalytic subunit alpha (PIK3CA), cyclin D1 (CCND1), snail family transcriptional repressor 1 (SNAI1), vimentin (VIM), zinc finger E-box binding homeobox 2 (ZEB2), CD44, nanog homeobox (NANOG), aldehyde dehydrogenase 1 family member A1 (ALDH1A1), CD47, CD274, and programmed cell death 1 ligand 2 (PDCD1LG2)] and ACTB (β-actin) as control were purchased from Applied Biosystems (TaqMan™ Gene Expression Assays). All samples were analyzed in triplicate. The primer sequences for the 16 genes included in this study are shown in Table II. Target gene expression in CTCs was determined using a relative quantification method. Detection of at least one of the four epithelial-related genes (EPCAM, MET, KRT19, and EGFR) was defined as CTC positivity. Since samples obtained by negative selection are invariably contaminated with leukocytes, the average threshold cycle (Ct) value of the CTC-negative samples was used as the baseline for the control group. The Ct values of the target genes were normalized to a reference gene (ACTB), and the expression levels of the target genes in CTCs were estimated as fold changes compared to those in the CTC-negative samples by the relative quantification 2-delta-delta Ct method.
Data acquisition and systemic inflammatory markers. Laboratory data including platelet, neutrophil, lymphocyte, and monocyte counts and serum CRP and albumin levels were collected from patients’ clinical records within 2 weeks of blood collection for CTC isolation. NLR, PLR, and LMR were calculated by dividing the absolute values of the corresponding hematological parameters. The systemic immune-inflammation index (SII), prognostic nutrition index (PNI), and CRP/albumin ratio (CAR) were calculated as follows: SII=platelet count × neutrophil/lymphocyte count (14), PNI=10× serum albumin+ 0.005× lymphocyte count (15), and CAR=CRP/albumin (16).
Statistical analyses. GraphPad Prism version 8.0 for Windows (GraphPad Software, San Diego, CA, USA) was used to perform the statistical analyses. Mann–Whitney U-tests were used to examine the differences in continuous variables. The correlations of two continuous variables, gene expression levels in CTC and systemic inflammatory markers, were determined by Spearman’s rank tests. The Kaplan–Meier curves were plotted and compared using the log-rank tests to compare survival curves between subgroups. The optimal cut-off values of systemic inflammatory markers for overall survival were determined based on receiver operating characteristics curve analysis. Two-sided p-values <0.05 were considered statistically significant.
Results
Detection of CTCs and systemic inflammatory markers in patients with R/M HNSCC. Thirty-five (83.3%) of the 42 patients with R/M HNSCC, including 30 patients reported in our previous study (13), were positive for at least one epithelial-related gene. The expression levels of epithelial-related markers in 35 CTC-positive patients are summarized in Figure 1. Among the 35 CTC-positive patients, EPCAM was detected in 15 (35.7%), KRT19 in 29 (69.0%), EGFR in 14 (33.3%), and MET in 21 (50.0%) patients, respectively. Next, we compared systemic inflammatory markers including blood cell counts between CTC-positive and CTC-negative patients and observed no significant differences between the two groups (data not shown).
Molecular characterization of CTCs. The CTCs obtained from 35 R/M HNSCC patients were subsequently investigated for the expression of 11 genes grouped into four gene signatures: cell growth (PIK3CA and CCND1), epithelial-mesenchymal transition (EMT) (SNAI1, VIM, and ZEB2), cancer stemness (CD44, NANOG, and ALDH1A1), and immune regulation (CD47, CD274, and PDCD1LG2). The expression of each gene in CTCs varied according to R/M HNSCC patients (Figure 2), confirming our previous report of phenotypic heterogeneity of CTCs among patients (17). Furthermore, we also analyzed the correlations between gene expression levels and observed a significant correlation between the same gene signatures (Figure 3). VIM or ZEB2 expression was strongly correlated with that of CD44 or ALDH1, suggesting a significant role of these CTC subpopulations in the metastatic process (VIM and CD44, r=0.9328, p<0.001; VIM and ALDH1, r=0.8602, p<0.001; ZEB2 and CD44, r=0.8098, p<0.001; ZEB2 and ALDH1, r=0.7062, p<0.001).
Prognostic significance of systemic inflammatory markers in patients with R/M HNSCC. The Kaplan–Meier survival analyses were performed to evaluate the prognostic significance of systemic inflammatory markers in patients with R/M HNSCC (Figure 4). R/M HNSCC patients with higher platelet counts, higher CRP levels, lower albumin levels, higher NLR, lower PNI, or higher CAR had significantly shorter survival than those in the pair groups for each systemic inflammatory marker (platelet count, p=0.036; CRP level, p=0.001; albumin level, p=0.041; NLR, p=0.038; PNI, p=0.020; CAR, p=0.002).
Correlation between gene expression levels of CTCs and systemic inflammatory markers. Finally, we examined the correlations between gene expression levels in CTCs and systemic inflammation markers in patients with R/M HNSCC (Figure 5A and B). PIK3CA expression in CTCs was correlated with lymphocyte-related systemic inflammatory markers including lymphocyte count, NLR, and LMR, while CD44 and ALDH1A1 expression in CTCs was correlated with lymphocyte- and albumin-related markers (lymphocyte count, albumin level, and PNI) and lymphocyte-related markers (lymphocyte count, NLR, PLR, and PNI), respectively. Moreover, PDCD1LG2 expression in CTCs was correlated with nine of the 12 markers tested (positive correlation: lymphocyte count, albumin level, LMR, and PNI; negative correlation, neutrophil count, NLR, PLR, SII, and CAR).
Discussion
Accumulating evidence has shown an association between the existence or enumeration of CTCs and various clinical factors in HNSCC (18-20); however, CTC heterogeneity and longitudinal changes within the same patient are substantial hurdles to understanding the biology and clinical potential of CTCs. This study aimed to molecularly characterize CTCs and their association with systemic inflammation in HNSCC. Since CTCs are more frequently detected and show extensive heterogeneity in patients with more advanced disease compared to those in patients with early disease (21), we enrolled patients with R/M HNSCC. As expected, CTCs were detected in approximately 80% of patients. We performed molecular characterization of CTCs based on four gene signatures, cell growth, EMT, cancer stemness, and immune regulation. Although the pattern of gene expression levels in CTCs varied among patients, the overall gene expression levels in CTCs revealed a strong positive correlation with those in the same gene signature group, except for genes regulating cell growth. Of note, the expression levels of EMT-related genes were highly correlated with those of cancer stemness. Previous reports have indicated that two CTC phenotypic characteristics, EMT and cancer stemness, are closely related and essential for metastasis. Most CTCs of metastatic breast cancer possess EMT and stem cell characteristics (22). Similarly, Papadaki et al. reported that high CTC expression levels of EMT and stemness markers were more frequently detected in patients with metastatic breast cancer (23). As the subjects in the present study had already developed a locoregional recurrence and/or distant metastasis, they may have had an enriched subpopulation of CTCs with both EMT and stemness phenotypes. During tumor blood dissemination, CTCs in the blood not only encounter various host cells, including leukocytes, platelets, and other CTCs but also interact with the blood environment. Thus, CTCs may acquire survival and metastatic potential through phenotypic alterations due to environmental exposure.
The blood environment comprises several factors. Systemic inflammatory markers are associated with disease progression, treatment responses, and prognosis in various cancers, including HNSCC (11, 12, 24-26). Our results indicated that platelet count, CRP level, albumin level, NLR, PNI, and CAR were prognostic markers in patients with R/M HNSCC, suggesting that these markers might be useful for evaluating life expectancy in R/M HNSCC.
In this study, lymphocyte- and albumin-related markers were associated with PIK3CA, CD44, ALDH1A1, and PDCD1LG2 expression in CTCs. Both peripheral blood lymphocytes and serum levels of albumin, a parameter of nutritional status, influence immune cell function (27, 28); therefore, high lymphocyte counts and high serum levels of albumin may contribute to antitumor immune responses. As they have left the immune-suppressive tumor microenvironment, CTCs in the bloodstream are more vulnerable to antitumor immune responses in the peripheral blood compared to primary tumors and metastatic lesions. Thus, the circulatory immune system might select for CTCs with immune-resistant phenotypes. The immune escape mechanisms employed by CTCs in the peripheral blood include human leukocyte antigen down-regulation and Fas ligand, CD47, and programmed death-ligand 1 (PD-L1) expression (29-31). Additionally, our findings suggest that molecular alterations in CTCs allow them to escape immune surveillance. Growing evidence supports the association between immunosuppressive phenotypes and the gene signatures tested in this study. Using a model of pancreatic cancer, Sivaram et al. showed that PIK3CA-AKT signaling in tumors reduced major histocompatibility complex class I and CD80 expression on the cell surface to promote immune evasion (32). In contrast, CD44+ tumor-initiating cells in HNSCC not only have an EMT phenotype but are also less immunogenic, and selectively express PD-L1 compared to CD44 cells (33). Similarly, EMT is strongly associated with an inflammatory tumor microenvironment in lung cancer, with increased levels of multiple immune checkpoint molecules (34). Although these findings were based on tumor tissues or xenografts and these phenomena might be mirrored within the tumor microenvironments, CTCs in the blood likely exhibit similar biological behaviors to acquire immune-suppressive properties.
The main limitations of the present study are its small sample size and heterogeneous population comprising different primary sites and recurrence patterns. We are currently conducting a large-scale study in pretreated HNSCC patients to assess the molecular phenotypic alterations of CTCs under blood environmental stress including antitumor immunity. Taken together, the present study investigated CTCs from patients who had received various treatments and had recurrent and/or metastatic lesions; therefore, it was difficult to determine the molecular mechanism of CTCs within the cascade of metastasis. However, our results suggest the potent survival advantage of CTCs through various phenotypic alterations in the hostile blood environment, specifically against immune functions defined by lymphocyte counts and serum albumin levels.
Acknowledgements
This work was supported in part by KAKENHI (Grants-in-Aid for Scientific Research) from the Ministry of Education, Culture, Sports, Science and Technology, Japan. The Authors would like to thank Editage (www.editage.com) for English language editing.
Footnotes
Authors’ Contributions
Kazuaki Chikamatsu conceived and designed the study. Hiroe Tada, Shota Ida, Ikko Mito, and Toshiyuki Matsuyama performed the experiments, and collected data. Hideyuki Takahashi and Kazuaki Chikamatsu analyzed and interpreted the data. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no conflicts of interest in relation to this study.
- Received December 21, 2020.
- Revision received January 6, 2021.
- Accepted January 7, 2021.
- Copyright © 2021 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.