Circulating Hybrid Cells Join the Fray of Circulating Cellular Biomarkers

Abstract

Gastrointestinal cancers account for more cancer-related deaths than any other organ system, owing in part to difficulties in early detection, treatment response assessment, and post-treatment surveillance. Circulating biomarkers hold the promise for noninvasive liquid biopsy platforms to overcome these obstacles. Although tumors shed detectable levels of degraded genetic material and cellular debris into peripheral blood, identifying reproducible and clinically relevant information from these analytes (eg, cell-free nucleotides, exosomes, proteins) has proven difficult. Cell-based circulating biomarkers also present challenges, but have multiple advantages including allowing for a more comprehensive tumor analysis, and communicating the risk of metastatic spread. Circulating tumor cells have dominated the cancer cell biomarker field with robust evidence in extraintestinal cancers; however, establishing their clinical utility beyond that of prognostication in colorectal and pancreatic cancers has remained elusive. Recently identified novel populations of tumor-derived cells bring renewed potential to this area of investigation. Cancer-associated macrophage-like cells, immune cells with phagocytosed tumor material, also show utility in prognostication and assessing treatment responsiveness. In addition, circulating hybrid cells are the result of tumor–macrophage fusion, with mounting evidence for a role in the metastatic cascade. Because of their relative abundance in circulation, circulating hybrid cells have great potential as a liquid biomarker for early detection, prognostication, and surveillance. In all, the power of the cell reaches beyond enumeration by providing a cellular source of tumor DNA, RNA, and protein, which can be harnessed to impact overall survival.

Summary.

Circulating cell-based biomarkers, a source of tumor DNA, RNA, and proteins, can be enumerated or provide in-depth tumor analyses to aid in cancer detection and disease monitoring. Here, we review the progress toward this goal and highlight future directions.

Cancers of the gastrointestinal (GI) tract account for more cancer-related deaths in the United States than any other organ site, including pulmonary.¹ Each GI cancer has unique challenges in early diagnosis, staging, and treatment that could benefit from improved noninvasive biomarkers to diagnose and track disease evolution. Specifically, as the second leading cause of cancer-related deaths in the United States, colorectal cancer (CRC) accounts for more than 150,000 cancer diagnoses and more than 51,000 deaths annually.¹ Despite advances in screening regimens for adults older than age 50 years, new CRC diagnoses in younger adults has increased 1.4% annually since 2004.² CRC diagnosed after a symptom-initiated work-up often portends an advanced burden of disease and a dramatic decrease in expected survival; Surveillance, Epidemiology and End Results data report 5-year survival for CRC diagnosed as locoregional disease at 80%–90%, compared with 14% in distantly metastatic disease.1, 3, 4 Late-stage diagnosis is even more common in pancreatic ductal adenocarcinoma (PDAC) owing to absent or nonspecific symptoms during the early stages of disease and contributes to its dismal prognosis. Although CRC has multiple effective screening regimens, PDAC currently lacks effective early detection modalities or validated biologic biomarkers,⁵ however, both cancers would benefit from additional noninvasive modalities for early detection and surveillance.

Noncellular Circulating Biomarkers

The holy grail of early cancer detection is the development of noninvasive biomarkers that elucidate both the presence of cancer and tumor progression. Current screening methods fall short of this goal. Screening colonoscopies for CRC are recommended for average-risk adults aged 50–75 years and are effective at detecting cancer with the added benefit of removing premalignant adenomas.⁶ However, colonoscopy is not universally accessible owing to high cost and the need for trained staff with specialized equipment. The fecal occult blood test and fecal immunochemical test are Food and Drug Administration–approved stool assays that expand accessibility but reduce the specificity of CRC detection.⁷ In addition, gold standard serum biomarkers available for PDAC and CRC, including carcinoembryonic antigen and cancer antigen 19-9, fall far short of reliable usage for diagnosis. Today, these tests are used primarily for surveillance and to monitor disease response during treatment.5, 8 To improve the sensitivity and specificity of cancer detection through noninvasive methods, a new generation of blood-based analytes with correlative or biologic value are in development, including exosomes, cell-free tumor DNA (ctDNA) or nucleic acids, and proteins (Figure 1).9, 10

Figure 1.

Circulating biomarkers in cancer. Summary of analytes detectible in peripheral blood, including cell-free nucleic acids (both DNA and RNA), proteins, membrane-bound structures, and cells. Functional use of each analyte, as well as their suitability for use in early detection and treatment responsiveness, is reported.

ctDNA is hypothesized to arise from tumor cell death, whether by necrosis, cell lysis, or apoptosis, resulting in the release of naked DNA into circulation and creating a residual fingerprint. ctDNA was first detected in healthy individuals in the late 1940s. However, it was not until the 1970s–1980s that neoplastic characteristics were discovered, leading to the realization that cancer patients had higher concentrations of ctDNA relative to healthy controls.11, 12 Quantification of ctDNA is useful in some disease states when used alongside more established blood assays,¹³ but is better suited for detecting the mutational evolution of cancer. However, despite major technological advances, ctDNA is not readily detectable in all cancers and is scarce in early disease.¹⁴ In addition, ctDNA does not always reflect tumor cell biology¹⁵ and further complicates its utility as a noninvasive biomarker.

Exosomes are another noncellular analyte sparking excitement in cancer research as an emerging biomarker with the potential to forecast the presence of malignancy, treatment response, and tumor progression. First described in the late 1980s, these small membrane-bound vesicles range in size between 50 and 140 nm and carry cargo that include proteins, DNA, RNA, and various lipid types.¹⁶ Detected in a myriad of cancers, exosomes are described to mediate angiogenesis,¹⁷ establish a premetastatic niche, and contribute to tumor progression.¹⁸ As a liquid biomarker, they remain a promising prognostic and diagnostic analyte, carrying an array of microRNAs that differ significantly between healthy controls and patients with various different cancer types, including glioblastoma multiforme,19, 20 pancreatic,21, 22 colorectal,²³ lung,²⁴ and breast cancer.²⁵ The wide variety of cargo carried by exosomes points to their functional relevance in intercellular communication, with the potential to inform tumor state and response to treatment.

Cellular Circulating Biomarkers

Beyond noncellular markers, circulating tumor cells (CTCs) were first identified by Ashworth²⁶ in 1869 in a metastatic cancer patient. CTCs are cells shed into peripheral blood directly from tumors. Although extremely rare in circulation,²⁷ evidence that CTCs correlate with poor prognoses exists in a number of disease sites, including CRC28, 29, 30 and PDAC.31, 32 Conventionally defined CTCs are identified based on the presence of epithelial or tumor markers, typically cytokeratin (CK) or epithelial cellular adhesion molecule (EpCAM), and the absence of the panleukocyte marker CD45. Platforms for CTC detection leverage various features of these cells, including density,33, 34 charge,35, 36, 37 size,38, 39, 40 and associated antigens41, 42, 43, 44, 45 (Table 1), which allow for investigations in early46, 47 and late-stage cancers.48, 49 More recently, stem cell identities in CTCs were shown to correlate with the presence of brain metastases in breast cancer patients,⁴⁸ supporting that circulating cells harbor important biologic information. This and other evidence suggest that CTCs may be useful in defining discrete differentiation states⁵⁰ and drug susceptibility.⁵¹

Table 1.

CTC Isolation Methods

Isolation method	Platforms	Advantages	Disadvantages
Density gradient33, 34	Accucyte (RareCyte, Inc, Seattle, WA) OncoQuick (Greiner Bio-One, Monroe, NC)	Relatively fast and inexpensive Independent of cellular antigens	Low specificity
Electrophoresis35, 36, 37	DEPArray (Menarini-Silicon Biosystems, Castel Maggiore, Italy) ApoStream (Precision for Medicine, Frederick, MD)	Independent of cellular antigens	Pre-enrichment required High cell loss during sample preparation
Size-based38, 39, 40	ISET (Rarecells Diagnostics, Paris France) ScreenCell (Westford, MA) CellSieve	Relatively fast and inexpensive Captures CTC aggregates Independent of cellular antigens	Small CTCs not captured Large leukocytes captured
Immunoaffinity41, 42, 43, 44, 45	CellSearch MagSweeper (Stanford University, Stanford, CA) Isoflux (IsoFlux, Inc, Pittsford, NY) Microfluidic chips	FDA approved (CellSearch) Some platforms are semi-automated Good reproducibility	Cost Specialized equipment required Dependent on cellular antigens

FDA, Food and Drug Administration; ISET, isolation-by-size of epithelial tumor cells.

In theory, a cell-based assay for early cancer detection would provide the greatest possible diversity of information, including DNA mutation status, tissue of origin, protein expression for signaling pathway activation, stem cell phenotypes, and gene expression. CTCs are the only cell population with commercially available assays approved for use in cancer-related treatment decisions. Unfortunately, CTCs are rare entities in circulation, even in patients with metastatic cancer. This rarity impedes their utility as a basis for routine transcriptomic or robust protein assessment. To date CTCs have not been shown to provide biologic insights to inform therapeutic decision making, despite initially promising results.⁵² However, CTCs represent the first identified cell population in an exciting new field, specifically that of circulating cells in cancer patients that have either tumor identity or characteristics that may have utility in a cell-based assay. This review recapitulates the advances made in the field of circulating cellular biomarkers, including a review of CTCs and the discovery of cancer-associated macrophage-like cells (CAMLs),⁵³ culminating in the newly described tumor-derived circulating hybrid cell (CHC).⁵⁴

CTCs: Prognostic Tumor Dandruff

Since their discovery in the mid-19th century, convincing correlation of CTC detection with disease burden has led to validated commercially available assays (Table 1). The Food and Drug Administration–approved CellSearch system (Menarini Silicon Biosystems Inc, Huntingdon Valley, PA) enriches cells using magnetic ferrofluid-coated antibodies targeting EpCAM for initial separation. Cells screened for the expression of CK8/18/19 and lack of CD45 expression identify EpCAM⁺/CK⁺/CD45^- cells as CTCs.⁵⁵ However, it should be noted that the CellSearch approach misses CTC subpopulations that express CD45 (ie, CHCs),⁵⁴ or that lack EpCAM expression, which can result from an epithelial-to-mesenchymal transition (EMT) and has important prognostic implications.⁵⁶ Platforms such as the isolation-by-size of epithelial tumor cells technique uses size-based filtration to capture a range of CTC types for downstream analysis and enumeration.⁵⁷ Although the clinical utility of CTC assays is undisputed in certain organ sites, there remains controversy over the causative role of CTCs in cancer metastasis, as well as their practical applications in early cancer diagnosis and treatment guidance. To this end, CTCs have been studied in a variety of malignancies; the subsequent sections review advances in CTC utilization.

Advances in CTC Utilization

Although CTCs are of interest in GI-derived cancers, they are perhaps best studied in the setting of breast and prostate cancers. Higher levels of CTCs portend poorer survival in locally advanced breast cancer patients undergoing adjuvant treatment, as well as in patients with distant metastatic disease.29, 58, 59 Similarly, CTCs are a valuable biomarker in metastatic, castrate-resistant prostate cancer because levels correlate with overall survival (OS)⁶⁰ and outperforms prostate-specific antigen (PSA) as a marker of early response after the initiation of chemotherapy.⁶¹ In addition, patients with a decrease in CTCs after systemic therapy have improved OS.⁶² The integration of CTCs into decision making during cancer treatment has remained an elusive goal. To date, the only clinical trial investigating CTC enumeration to measure the effectiveness of chemotherapy and inform a switch to alternative regimens was the SWOG S0500 trial, which did not show a survival difference when using CTC levels to guide chemotherapy regimens in breast cancer patients.⁶³ Upcoming trials investigating the utility of CTCs for this purpose, including the CirCe01 trial (NCT01349842) and the STIC CTC trial (NCT 01710605), may provide additional insight.

CTCs in GI Cancers

There is significant interest in the development of a CTC-based liquid biopsy platform for GI cancers; however, efforts have been stymied by lower CTC detection rates in many GI-derived cancers compared with breast and prostate cancers.⁶⁴ CellSearch and other platforms detect as low as a single CTC in 7.5 mL of blood, reflecting poor sensitivity, especially at early cancer stages, and have limited the clinical utility of CTCs to impact outcomes based on earlier detection.⁶⁵ However, improved detection of PDAC has been shown when CTCs are used in combination with other circulating biomarkers, such as cancer antigen 19-9 and doublecortin-like kinase 1.18, 66, 67 Although CTC enumeration has been shown to correlate with disease stage32, 68, 69, 70 and prognosticate future development of metastasis in PDAC,32, 71 some studies have found no correlation.72, 73, 74 Advances in isolation techniques in more recent studies may account for this discrepancy; however, the clinical role of CTC enumeration in PDAC staging remains unclear.

Beyond CTC Enumeration

Although poorer survival has been associated with detectable CTC levels, the reliance on a minute number of cells has led many investigators to look beyond enumeration to molecular and phenotypic characteristics of CTCs for an enhanced prognostic readout in CRC and PDAC. In addition, as with the CellSearch system, the sole reliance on epithelial markers fail to capture subpopulations, which provide prognostic insight.18, 75, 76, 77 By expanding marker profiling, associations between cyclooxygenase-2 (COX2), Leucine-rich repeat-containing G-protein coupled receptor 5 (LGR5), or caudal-type homeobox-2 (CDX2), expression in CTCs have been reported to convey poor OS in CRC patients.78, 79, 80 However, detection of mesenchymal CTCs through the co-expression of CK and mesenchymal markers, such as vimentin or Twist family BHLH Transcription Factor 1 (TWIST1), predict aggressive tumor biology and earlier cancer recurrence in PDAC.69, 70, 75, 77 This supports the theory that EMT facilitates distant metastasis through tumor cell acquisition of mesenchymal properties that are essential for migration and motility.81, 82 Furthermore, the Circulating Tumor Cells in Pancreatic Cancer (CLUSTER) trial (NCT02974764) showed that alterations in total CTC levels and the ratio of mesenchymal CTCs may provide a marker of treatment response and disease progression to help guide therapy decisions.⁷⁷ There is also evidence that EMT encourages the formation of CTC clusters,⁸³ which have been shown to possess greater metastatic potential than individual CTCs⁸⁴ and predict worse survival in patients with PDAC,⁸⁵ as well as metastatic breast cancer patients.⁸⁶ Although the mechanisms are not fully understood, multicellular clustering in the portal vein may contribute to liver metastasis in PDAC patients by promoting CTC growth and immune cell evasion.⁸⁷ In addition, a recent study by Szczerba et al⁸⁸ showed that CTC clusters with associated neutrophils in the peripheral blood of breast cancer patients drives cell-cycle progression, and may be a potential therapeutic target.

A major strength of cellular-based circulating biomarkers is their potential to harness the entire tumor genome and assess treatment-related changes without the need for multiple invasive biopsies. Although KRAS mutations have been detected in CTCs using polymerase chain reaction,73, 89, 90, 91 in-depth genetic analysis at the single-cell level has been challenging.⁹² Newer technologies that leverage unique CTC features, such as glycoprotein surface expression, can facilitate their isolation for downstream analyses of DNA mutations and gene expression.⁹³ These advances could be developed for the monitoring of CTC subtypes across therapeutic treatment for a real-time readout of tumor burden.

The prognostic value of CTCs is established in a number of malignancies, robust data demonstrating their ability to inform management decisions or detect early disease is lacking. These limitations have prompted investigations for cellular alternatives, leading to the discovery and characterization of new populations of circulating cells in the past decade, including CAMLs and CHCs.

CAMLs: A Circulating Sampler Platter of the Tumor Microenvironment

Adams et al⁵³ reported the isolation of large cells with atypical nuclei and vacuoles of tumor material from blood samples of patients with breast, pancreatic, and prostate cancers in 2014 using low-flow microfiltration. These circulating cells predominantly express CD14, a typical macrophage antigen, thus the name cancer-associated macrophage-like cells. Macrophage and other monocyte-derivative cells are a multifaceted immune population with key roles in maintaining tissue homeostasis through direct cellular functions, such as phagocytosis, and the transmission of immune regulatory cell signals.⁹⁴ Macrophages are recruited to the tumor microenvironment (TME) through cancer cell–derived cytokines,⁹⁵ where they can account for up to 50% of a tumor’s mass and become known as tumor-associated macrophages (TAMs).⁹⁶ It is hypothesized that CAMLs originate from TAMs that have undergone macrophage–macrophage fusion and phagocytose dying neoplastic cells before ultimately disseminating back into circulation with internalized tumor fragments. Although cellular fusion of monocyte-derived giant cells has been well-described in inflammatory diseases, the exact mechanism of CAML formation is largely unknown. However, a portion of their life cycle may derive from ongoing interaction with CTCs in circulation because 10% of studied patients with metastatic cancer have CAMLs bound to CTCs.⁵³

CAMLs are enlarged, highly differentiated cells with atypical or multiple nuclei and phagocytosed tumor matter. As an immune cell population, they are differentiated from CTCs by their CD45 positivity and detection of internalized tumor markers (CK or EpCAM) within the cytoplasm. Notably, CAMLs are large cells that range from 25 to 300 μm and have highly variable morphology,⁵³ although CHCs and CTCs generally are round, with sizes ranging from 5 to 20 μm and 12 to 25 μm, respectively (Table 2).54, 97 In addition, as immune cells, CAMLs are not thought to be directly tumorigenic but may play a role in facilitating metastatic seeding when bound to CTCs by providing a mechanism of immune evasion.⁵³

Table 2.

Characteristics of Circulating Cellular Biomarkers

	Identity	Function	Relative abundancea		Marker expression		Morphology	Size, μm
			Early stage	Late stage	Cell surface	Cytoplasmic
CTC52	Disseminated tumor cell in circulation	Metastatic seeding	↑	↑↑	EpCAM	Cytokeratin	Shape: round Nuclei: ≥50% of cytoplasm	15–25
CAML53	Macrophage with phagocytosed tumor material	Immune cell functions	↑↑	↑↑↑	CD45	Cytokeratin EpCAM	Shape: variable (amorphous, round, oblong) Nuclei: large or multiple	25–300
CHC54	Product of tumor–macrophage fusion Characteristics of both macrophage and tumor cells	Metastatic seeding	↑↑↑	↑↑↑↑	CD45 EpCAM MUC4	Cytokeratin	Shape: round Nuclei: round or binucleated	5–15

MUC4, mucin 4.

^aRelative abundance based on pancreatic adenocarcinoma data.

CAMLs as a Diagnostic and Prognostic Tool

By using the CellSieve (Creatv MicroTech, Inc., Potomac, MD) microfiltration system, CAMLs have been isolated from peripheral blood in patients with a wide range of cancer types, including esophageal, liver, pancreatic, colorectal, breast, and prostate.98, 99 CAMLs are not detected in healthy patients, but are identified in patients with noninvasive conditions, including benign breast conditions.¹⁰⁰ Similar to CTCs, CAMLs have high detection rates in late-stage disease, but lower sensitivity in stage I patients, which ultimately limits their utility for early detection.⁹⁸ However, CAML enumeration does show promise as an indicator of treatment responsiveness, shown by increased levels in patients after initiating chemotherapy.⁵³ In addition, both increased CAML size and higher levels in untreated breast cancer patients correlated with shorter progression-free survival and worse OS.98, 101

Ultimately, CAMLs provide a promising method for cancer detection, prognostication, and treatment response. However, more research is needed to understand the mechanisms of their formation, function in circulation, and contribution to the metastatic progression of cancer. Interestingly, CAMLs have some similarities with CHCs, another newly described circulating cell population with tumor characteristics. Despite their shared CD45 positivity and tumor marker expression, the large size, variable morphology, and cytoplasmic staining of EpCAM distinguish CAMLs from CHCs. In addition, although both CHCs and CAMLs derive from macrophages, CAMLs retain their immune cell identity, while CHCs are a distinct product of macrophage fusion with tumor cells that imbues properties essential for the initiation of the metastatic cascade.⁵⁴

CHCs: Chimeras of Groundbreaking Significance

Cellular fusion is a phenomenon by which cells from identical (homotypic) or distinct (heterotypic) lineage combine into a single cell with shared nuclear and cytoplasmic contents, and is known to occur during both homeostatic and inflammatory states.102, 103, 104 Recent investigations in both murine and human models have described fusion in malignant states, where cell fusion hybrids may play an important role in cancer progression and metastasis.105, 106, 107, 108, 109, 110, 111, 112, 113 The concept that fusion between leukocytes and tumor cells may promote cancer metastasis was first postulated by Aichel¹¹⁴ more than a century ago. However, evidence directly linking cellular fusion to phenotypic diversity and cancer metastasis only recently has come to light through the discovery of tumor-derived hybrid cells within the TME and in circulation.⁵⁴ CHCs represent a possible mechanism of cancer metastasis, and CHC enumeration and analysis may have utility as a diagnostic and prognostic biomarker in human malignancies.

Cellular Fusion in Solid Tumors

Direct observation through in vitro live imaging provides the most compelling evidence of cancer–macrophage fusion. By co-culturing MC-38 colorectal cancer cells expressing red fluorescent protein (RFP) with macrophages expressing green fluorescent protein (GFP), cellular fusion and the formation of hybrid cells harboring cytoplasmic GFP and nuclear RFP has been observed.⁵⁴ These GFP⁺/RFP⁺ hybrid cells retain mitotic activity, producing daughter cells with identical fluorescence expression. Fusion hybrids are identified using similar methods in a murine breast cancer model and in vitro human breast cancer cell lines.115, 116 Together, these data distinguish cellular fusion from other immune cell functions such as phagocytosis and trogocytosis, a process by which leukocytes can extract and express surface antigens from antigen-presenting cells.¹¹⁷

Although all leukocytes may be capable of cellular fusion, the macrophage is the principal fusogenic leukocyte in solid tumors, as evident in mouse models of bone marrow transplants (BMTs) and mouse models of breast cancer.106, 115 Notably, the fusion process involves the up-regulation of genes in pathways linked to metastatic spread, including activated leukocyte cell adhesion molecule (ALCAM) FMS related tyrosine kinase 4 (FLT4), and runt-related transcription factor 1 (RUNX1).¹⁰⁶ Further, fusion hybrids show enhanced migratory and invasive properties relative to unfused cancer cells.⁵⁴ In addition to the acquisition of macrophage properties, hybrids retain tumorigenicity, as evidenced by tumor growth after injection into recipient mice in models of colon and ovarian carcinoma.54, 118

In human malignancies, the most compelling evidence of fusion in solid tumors is reported in female patients who previously received a sex-mismatched BMT and subsequently developed PDAC. When tumor specimens were interrogated with pan-CK antibodies and Y chromosome fluorescence in situ hybridization (FISH) probes, Y chromosomes were discovered within the nuclei of CK⁺ cancer cells, indicating cellular fusion between male donor leukocytes and recipient pancreatic cells.⁵⁴ In this way, fusion hybrid cells are identified in several other solid tumors of sex-mismatched BMT patients, including renal cell carcinoma, head and neck squamous cell carcinoma, lung adenocarcinoma, and ovarian carcinoma.54, 118, 119 Fusion hybrids also have been discovered in non–small-cell lung cancer, melanoma, and prostate cancers using other methodologies.120, 121, 122, 123 The full extent to which cellular fusion plays a role in the metastatic cascade remains unknown, but it has been theorized to contribute to the development of chemotherapy resistance, and may induce the Warburg effect.124, 125

Fusion Hybrids: From Tumor to Circulation

In addition to their role in CAML formation, TAMs also serve as the reservoir for intratumoral cellular fusion, and fusion-derived hybrid cells retain properties of both TAMs and tumor cells.⁵⁴ Subsequently, fusion hybrids are hypothesized to disseminate into circulation, where they are detectable as CHCs. The prevalence of CHCs in peripheral blood of murine cancer models and human cancer patients, combined with their robust tumorigenic capacity, underline their potential role as effectors of cancer metastasis.⁵⁴ Further supporting this, experimental assays of metastasis using in vitro–derived hybrid cells resulted in pulmonary metastases with notably higher seeding and growth when compared with unfused-MC-38 cells (analogous to CTCs).⁵⁴ Furthermore, in a spontaneous metastasis model, detectible tumor cells at both the primary and distant metastatic sites were of fusion origin,⁵⁴ thus providing compelling evidence for the role of CHCs in disease progression.

CHCs appear to be more prevalent in circulation than conventional CTCs. In a murine model of melanoma, CHCs derived from orthotopic injection of RFP⁺ B16F10 melanoma cells into GFP⁺ mice were the predominant tumor-derived cell in circulation (identified as RFP⁺/GFP⁺).⁵⁴ Conventionally defined CTCs (RFP⁺/GFP^-) comprised only 10% of all RFP⁺ circulating cells. Furthermore, of key importance is the fact that the majority of CHCs were shown to express the pan-leukocyte antigen CD45, indicating that dual expression of CD45 and a tumor marker could be translated to human patients for CHC detection.

Reports of atypical circulating macrophage-like cells are surprisingly common in the literature, and it is likely that CHCs have gone unrecognized and unappreciated for some time (Table 3).18, 126, 127, 128, 129, 130, 131, 132, 133, 134, 135 Toyoshima et al¹²⁷ described atypical cells in immunodeficient mice injected with CD45⁺/EpCAM⁺ or CD45^-/EpCAM⁺ tumor cells isolated from patients with advanced gastric cancer. The CD45⁺ fraction had significantly enhanced tumorigenicity compared with the CD45^- fraction, an unanticipated finding.¹²⁷ Similarly, a CD45⁺/CK⁺ cell population isolated from breast cancer patients was reported to prognosticate worse OS, although their significance was not investigated.¹³⁰

Table 3.

Unappreciated CD45⁺/CK⁺ Cells in Prior Publications

Study	Cancer site	Explanation for CD45+/CK+ cells	Relevant findings
Zhang et al18	PDAC	Unknown	ND
Toyoshima et al127	Gastric	Unknown	Increased tumorigenicity
de Wit et al126	NSCLC	Unknown	ND
Nel et al129	NSCLC	Unknown	ND
Stott et al133	NSCLC, prostate	Trogocytosis	CD45+/CK+ more prevalent than CTCs
Sajay et al128	NSCLC, breast	Unknown	ND
Takao and Takeda132	NSCLC, breast	False positive	ND
Lustberg et al131	HNSCC, breast	Unknown	ND
Lustberg et al130	Breast	Artifact	ND
Riethdorf et al134	Breast	Artifact	ND
Allan et al135	Breast	Artifact	ND

HNSCC, head and neck squamous cell carcinoma; ND, not determined; NSCLC, non–small-cell lung cancer.

Clawson et al¹³⁶ suggested that CHCs occur in human patients. Cells cultured from peripheral blood of melanoma patients expressed macrophage (CD204, CD206, CD163), epithelial (CK, EpCAM), and melanocyte (MLANA, ALCAM) markers and grew tumors when injected into immunodeficient mice. These cells were presumed to be fusion-derived because they were also found in the primary tumors of the melanoma patients.¹³⁶ Similar findings were reported in PDAC patients¹³⁷ and by Gast et al,⁵⁴ who identified CHCs in peripheral blood from the female PDAC patient with a prior sex-mismatched BMT by the expression of CD45, EpCAM, and donor-derived Y chromosome that were also carrying macrophage-specific epitopes and tumor-specific mucin 4 expression. Studies that have reported, but not characterized, the CD45⁺ fraction are listed in Table 3.

Implications of CHCs on Cancer Diagnosis and Treatment

The existence of CHCs in human malignancies invites questions as to their diagnostic and prognostic potential as a liquid biopsy. Promising early data have shown potential for CHC enumeration to discriminate between PDAC disease stage, with high levels significantly correlating with advanced disease states, and correlating with OS regardless of cancer stage.⁵⁴ In this patient population, CHC level showed improved performance compared with CTCs, which were not correlated significantly with either stage or OS.⁵⁴

In summary, CHCs are a newly described circulating cell population in cancer patients, and compelling evidence from multiple investigators has characterized their tumorigenicity, spontaneous formation in murine and human cancers, as well as their relative abundance in circulation. Additionally, CHCs correlate with disease presence, stage, and prognosis. Despite promising early results, CHCs are still a newly described circulating neoplastic cell population with only a few descriptions in the literature54, 136 and therefore further validation of these cells is needed. Furthermore, the utility of CHCs in predicting advanced disease states and in prognostication has yet to be established in larger cohorts of patients with PDAC or across a wider range of GI malignancies. Similarly, the relative prognostic and diagnostic utility of CHCs compared with CTCs across multiple malignancies is untested and requires further study. Nonetheless, there is clearly great promise for this novel cell type.

Discussion

Prognostic biomarkers with enhanced sensitivity and specificity have promise to transform survival from cancer. Thus, the ideal effective biomarker will provide precise and actionable information about tumor location, stage, mutational status, therapeutic vulnerability, and extent of tumor heterogeneity. This is a tall order for a single biomarker. Certainly, dual-analyte biomarkers have enhanced potential for providing the most comprehensive information. For example, a newly developed platform, CancerSEEK (Johns Hopkins Kimmel Cancer Center, Baltimore, MD), combines ctDNA and protein biomarkers to impart tissue localization and a cancer diagnoses in 8 different cancers. Although promising, the ctDNA-based assay showed poor sensitivity for PDAC and other organ sites, and the study lacked inclusion of high-risk cohorts with premalignant pathology.¹³⁸ Novel multi-analyte platforms could use biomarkers that deliver on multiple fronts: DNA, protein, and cellular information.

Assays built around cells that originate directly from the tumor or the TME have the potential to provide genomic information, mutational analysis, tumor-associated protein identity, the ability to survey epigenetic alterations, and cellular heterogeneity that all derive from the same source. The Achilles’ heel of CTCs is their low prevalence in circulation, which presents a challenge to extend analyses beyond enumeration or protein expression. CHCs show immense promise as a plentiful biomarker for the early detection, diagnosis, and surveillance of a wide range of cancers, and have potential applicability beyond that of CTCs and CAMLs. Further study of this cell population may show an untapped resource for the development of effective biomarkers to impact cancer treatment and survival.