Circulating Tumor Cells in Esophageal Cancer

Introduction

Esophageal cancer (EC) is an increasing entity with an incidence of 7,200 newly diagnosed patients prognosticated in 2016 in Germany [1]. The effective 5-year overall survival (OS) is 20% [1]. Most of the cases of EC are either squamous cell carcinomas (ESCC) arising from the squamous cell epithelial lining of the esophagus, or adenocarcinomas (EAC). In EAC, it is thought that the squamous cells are replaced by metaplastic columnar epithelium from which the adenocarcinoma arises. Over the last 20 years the incidence of ESCC has been stable, while the number of patients presenting with EAC has dramatically risen in Europe and North America.

Although multidisciplinary treatment of patients with clinically non-metastasized disease offers the chance for cure in EC, nearly 50% of these curatively treated patients develop recurrence due to distant metastasis within 5 years [2^,3^,4]. The most significant transit mechanism of this distant tumor recurrence is hematogenous distribution of circulating tumor cells (CTC) from the primary tumor to the location of distant metastasis. These CTC are rare cells found in the peripheral and central blood of cancer patients [5]. These cells are genetically and phenotypically heterogeneous [5]. In the past, their presence and quantity during treatment regimens has been correlated with poor oncological outcome in patients with various solid tumors including EC [6^,7]. CTC are also referred to as circulating epithelial cells (CEC) when phenotypically benign and found in patients without a known tumor diagnosis, and circulating tumor microemboli (CTM) when seen in clusters [8^,9]. These cells in transit, obtained by a simple blood draw, not only offer the hope of early tumor diagnosis and possible treatment decisions , they may also help to improve our understanding of tumor biology and ultimately to improve treatment options. This review focuses on the biology of CTC in ESCC and EAC and highlights the potential for CTC to serve as liquid tumor biopsies. It also discusses the problem of heterogeneous and incomparable data due to inconsistent isolation techniques. An English language literature search in PubMed/Medline was performed using the search terms circulating tumor cells, CTCs, CTC detection methods, esophageal cancer, esophageal adenocarcinoma , esophageal squamous cell carcinoma (up to November 2016).

Epidemiology and Histological Types

Malignant tumors of the esophagus are constituted by the main subtypes ESCC and EAC. With a prognosticated current yearly incidence of 7,200 newly diagnosed patients, EC is a proportionally rare tumor in Germany [1]. Over the last 2 decades the incidence especially of EAC has been rapidly rising in Europe and North America. In the USA the incidence of EAC has exceeded ESCC since 1997 [10]. EC is especially rare in young people and increases in incidence with age, peaking in the seventh and eighth decades of life. EC is 3-5 times more common in men than in women with a median age at onset of 67 years in men and 73 years in women.

Current Staging and Management

Clinically, EC presents with symptoms of progressive dysphagia, anemia or weight loss. An increasing percentage of diagnoses is discovered in screening and surveillance endoscopy programs especially for Barret's metaplasia of the esophagus. Macroscopic and histological confirmation of the diagnosis of EC is performed by endoscopy with biopsy. Current management of EC is stage dependent and often carried out multidisciplinarily by surgical, interventional endoscopic, chemotherapeutic and radio-chemotherapeutic means. Pre-therapeutically it is essential to generate an exact picture of the local and systemic extent of the tumor disease. Computed tomography (CT) provides information not only on the local extent of the tumor but also on local and distant nodal involvement and distant metastasis in liver, lung and bone. Endoscopic ultrasound offers the highest preoperative clinical accuracy concerning the local tumor infiltration of the esophageal wall (UICC T-Stage) and the local nodal involvement (UICC N-Stage) [11]. Positron emission tomography (PET) and PET-CT scans are complementary diagnostic modalities in EC staging, which provide the possibility to detect previously unseen distant metastasis in 5% of cases [12]. PET-CT is generally considered an effective method for post-neoadjuvant assessment [13]. As tumor recurrence especially in the distant compartment is frequent in approximately 50% of curatively treated patients, it has to be assumed that small metastatic or micro-metastatic disease is missed by current staging modalities [2^,3^,4].

Novel tools and staging systems for early tumor detection, adequate prognostic staging, and accurate treatment monitoring and selection, in particular in the neo-adjuvant setting, are also needed. The seventh edition of the AJCC Cancer Staging Manual introduced the new category M0(i+) for breast cancer [14]. It is defined by the presence of disseminated tumor cells not exceeding 0.2 mm detectable in bone marrow, circulating blood or other non-regional tissues of the non-(macro)metastatic patient. The development and utilization of reliable methods for the detection of this minimal metastatic disease, like CTC detection techniques, and the introduction of new staging categories, such as M0 (i+), could significantly improve prognostic and predictive value of clinical cancer staging not only in breast cancer but also in gastrointestinal cancers such as EC.

CTC Enrichment and Detection Techniques in EC

CTC were first described in 1869 by Thomas Ashwoth [15]. Since then several techniques for detection, isolation and enumeration of CTC have been developed. As CTC are present in the blood of patients with solid tumors such as EC in concentrations of 1 cell per 5-10 × 10⁶ white blood cells, detecting them is like looking for a needle in a haystack [16]. For this reason, very sensitive and specific analytical techniques are needed first to ensure reliable enrichment and second to allow sure detection of CTC [17]. The methods that have been developed for enrichment of CTC from other blood cells can be grouped into technologies that select CTC based on their antigenic surface structure and methods that enrich CTC by physical properties of the cells. For antigenic surface structure-based enrichment procedures, immune-magnetic cell selection procedures are mostly applied [6]. For enrichment of CTC based on physical properties, different approaches can be used: CTC cell size for membrane filter-based enrichment procedures, specific density for gradient centrifugation enrichment, cellular electric charge properties in dielectrophoretic procedures and the deformation ability of the cell in chip-based microfluidic CTC enrichment methods [18^,19^,20^,21]. After enrichment of CTC, detection is carried out either by nucleic acid-based methods or by methods that apply different techniques of whole cell recognition.

For nucleic acid-based CTC recognition, reverse transcription polymerase chain reaction (RT-PCR) is primarily applied and markers, such as cytokeratin (CK) and carcinoembryonic antigen (CEA), and mRNA are used to detect CTC in EC patients (table 1). New biomarkers such as p63 and survivin are now also being used in RT-PCR detection techniques (table 1). Since at least the conventional biomarkers like CK and CEA are also expressed in physiological peripheral blood cells and non-malignant epithelial cells, the specificity of nucleic acid-based CTC recognition is impaired by significant false-positive results caused by biomarker expression by physiological and non-malignant circulating cells [22].

Open multimedia modal

In cytometric-based CTC detection, immunocytometric techniques like immunocytochemical visualization, immunofluorescent visualization or flow cytometry are utilized [23]. Cytometric detection methods isolate individual cells based on their surface antigen expression using, for example, monoclonal antibodies directed against epithelium-specific antigens. The clear advantage of cytometric methods over nucleic acid-based methods is the possibility to further characterize the cells as the target cells are not lysed in the detection procedure. Subsequent morphological identification and molecular characterization of CTCs are, therefore, possible. The major limitation is the current lack of tumor-specific antibodies for most solid tumors including EC. The commonly used CK antigens are also expressed on white blood cells and nucleated hematopoietic precursor cells [23].

Up to now different methods for enrichment and consecutive cytometric detection of CTC have been developed, including immunocytochemistry, flow cytometry, and the isolation by size technique (ISET) [24^,25^,26]. However, most of these methods are predisposed to technical variation and have not been widely available commercially. More recently, technical developments in the methods of CTC capture and enumeration have led to the development of an FDA-approved CTC detection platform: the established CellSearch® system (Janssen Diagnostics, Raritan, NJ USA). Through immunomagnetic cell separation with the epithelial surface antigen EpCAM, this system enables consistent isolation and quantification of CTC from the blood of patients with different solid tumors [2^,6]. Studies evaluating the CellSearch® platform have found wide variation in the detection of CTC across different tumor types [27^,28]. In these studies, CTC are usually defined as EpCAM⁺/CK⁺/CD45^- cells with a round or oval intracellular nucleus. A large study including patients with cancer, patients with non-malignant diseases, and healthy volunteers found that CTC are present in 11-57% of patients with cancer but cannot be found in healthy individuals [16]. The CellSearch® platform has also been used in prospective studies of CTC in EC [6]. Although several studies have shown the utility of the CellSearch® method, there is emerging evidence that there are cells in transit that escape the immunomagnetic mechanism of CTC enrichment due to different, more mesenchymal surface antigens [26^,29^,30]. Thus, alternative, technically simple, less costly, and also commercially available methods such as the ISET (e.g. ScreenCell®, Paris, France) have been developed. Our study group recently reported a study on CTC isolated with the ScreenCell® method in pancreatic cancer [5^,31], which found CTC in 73% of patients, with a combination of cytology and genetic evaluation. Several studies have identified size-based CTC isolation as a valid accompanying technique to epithelial marker-dependent CTC isolation [32^,33^,34^,35].

CTC in EC

Compared to other solid tumor entities, only few studies have explored the meaning of CTC for progression and prognosis in EC. Investigations performed in the first decade of this century, which were mainly carried out in Asian EC patients, used nucleic-acid based CTC detection by RT-PCR. These studies revealed varying CTC detection rates ranging from 2 to 57% in patients with ESCC subtype (table 2). The relatively broad range may be explained by inconsistent definition of positive CTC status in RT-PCR, different nucleic acid-extraction protocols and the utilization of different molecular markers. A recent meta-analysis examined 13 eligible studies with a total of 979 ESCC patients. In this meta-analysis, 11 of the 13 studies (821 of 979 patients) used RT-PCR for CTC detection. The meta-analysis revealed positive CTC in patients' blood to be associated with both worse progression-free/disease-free survival (PFS/DFS) with a hazard ratio (HR) of 2.32 (95% confidence interval (CI) 1.57-3.43, p < 0.001) and poorer OS with an HR of 2.64 (95% CI 1.69-4.14, p < 0.001) [36].

Open multimedia modal

Recent technical advancements in both CTC enrichment and detection methods have been centered on cytometric-based CTC detection assays rather than nucleic acid-based CTC detection methods, not only in EC but also in many other solid cancer entities. In 2014, a first pilot study on immunomagnetic CTC detection with the FDA-approved CellSearch® platform in 18 patients with either metastasized gastric cancer or EC showed ≥ 2 CTC/7.5 ml blood (EpCAM⁺/CK⁺/CD45^-) in 8 of the patients [37]. After systemic chemotherapy, the tumor response rate was 60% versus 38% in patients with < 2 CTC/7.5 ml blood versus ≥ 2 CTC/7.5 ml blood. Median PFS and OS were 6.1/10.5 months in the groups of patients with < 2 CTC and 5.2/6.1 months in those with ≥ 2 CTC.

Using the same platform, another study showed that 28% of the patients (25/90) with advanced non-treated ESCC were positive for CTC (EpCAM⁺/CK⁺/CD45^-). OS was significantly shorter in CTC-positive patients. After chemotherapy or chemoradiotherapy, CTC follow-up was obtained from 71 patients. After treatment, CTC positivity was significantly higher in patients with clinical progressive disease (n = 27) compared to patients with clinical partial response to therapy (n = 32; 40.7 vs. 6.3%). Patients with no CTC at baseline and patients with a change in CTC status from positive to negative showed a favorable prognosis (fig. 1) [38].

Open multimedia modal

A larger prospective examination of CTC in ESCC (n = 29) and EAC (n = 68) was carried out in 100 chemotherapy-naive patients with resectable EC and pre-operative immunomagnetic CTC (EpCAM⁺/CK⁺/CD45^-) detection. Notably, 10% (3/29) of the ESSC patients and 21% (14/68) with EAC were CTC positive. The study revealed that CTC-positive patients with non-metastatic disease had a significantly shorter OS and PFS than CTC-negative patients [6].

As the EpCAM⁺/CK⁺/CD45^- definition of CTC used by the CellSearch® platform only refers to epithelial surface markers, it is likely to miss many CTC subpopulations especially if they have undergone epithelial-mesenchymal transition. There is emerging evidence that there are cells in transit that escape the immunomagnetic mechanism of CTC enrichment due to a different, mainly mesenchymal surface antigenic profile [26^,29^,30]. Thus, alternative, technically simple and also commercially available methods such as ISET have been developed.

In 2015, Li et al. [39] compared the marker-dependent immunomagnetic CellSearch method and the epithelial marker-independent ISET method in 61 non-treated ESCC patients. CTC were detected in 33% (20/61) by ISET and in less than 2% (1/61) by the immunomagnetic method. Moreover, circulating tumor microemboli (CTM) were observed in 5% (3/61) patients using ISET, but were undetectable in any of the patients by immunomagnetic method. Therefore, further studies are certainly needed to assess the potential clinical relevance of the different CTC detection technologies in EC. With this goal, the immunomagnetic CellSearch method and the epithelial marker-independent ISET method are currently being compared in a multicenter clinical trial of multimodal treatment of non-metastasized EAC in more than 400 patients [40].

Future Clinical Implications of CTC in EAC

One main clinical aim of CTC research is to assess CTC as a prognostic and predictive biomarker for EAC to stratify patients into defined prognostic and therapeutic subgroups. The key aim of future studies is to determine whether pre- and intra-therapeutic CTC detection can accurately indicate prognosis in EAC patients and may improve preoperative staging, differential indication for specific neoadjuvant therapeutic modalities and indication for adjuvant therapy [40]. With regard to multimodal treatment protocols, CTC detection could improve selection for either protocols with stronger systemic effects such as perioperative chemotherapy or protocols with a predominantly locoregional effect on the tumor disease such as neoadjuvant chemoradiation. The addition of CTC detection to intra-therapeutic re-staging procedures and the comparison of the results to pre-therapeutic CTC detection values could also possibly identify patients with either strong or limited response to neoadjuvant treatment. If the presence of CTC following neoadjuvant chemotherapy or chemoradiation could contribute to the currently available prognostic information, it will help to identify non-metastatic EAC patients at high risk for disease progression who may derive benefit from additional adjuvant therapies or inclusion into clinical trials of novel therapies for high-risk patients [47].

Conclusion

Appropriate staging systems are essential for determining treatment strategies, especially those involving multimodal treatments in EC. Despite the availability of several diagnostic techniques, current pre-treatment staging remains inconsistent. Novel tools for adequate prognostic staging and accurate therapy monitoring in EC are urgently needed. CTC have been found to be independent risk factors for worse prognosis in patients with EC. Pre- and intra-therapeutic CTC enrichment and detection has the potential to improve preoperative staging, differential indication for neoadjuvant therapeutic modality and indication for adjuvant therapy.

Disclosure Statement

The authors report no competing financial interests.