Review

Cultured circulating tumor cells and theirderived xenografts for personalizedoncology

Ruoxiang Wang a,*, Gina C.Y. Chu a, Stefan Mrdenovic a,Alagappan A. Annamalai b, Andrew E. Hendifar a,Nicholas N. Nissen b, James S. Tomlinson c, Michael Lewis d,Nallasivam Palanisamy e, Hsian-Rong Tseng f,Edwin M. Posadas a, Michael R. Freeman b, Stephen J. Pandol a,Haiyen E. Zhau a, Leland W.K. Chung a,b

a Uro-Oncology Research, Department of Medicine, Samuel Oschin Comprehensive Cancer Institute,Cedars-Sinai Medical Center, Los Angeles, CA, USAb Uro-Oncology Research, Department of Surgery, Samuel Oschin Comprehensive Cancer Institute,Cedars-Sinai Medical Center, Los Angeles, CA, USAc Department of Surgery, West Los Angeles VA Hospital, Greater Los Angeles Veterans AffairsHealthcare System, Los Angeles, CA, USAd Department of Pathology, West Los Angeles VA Hospital, Greater Los Angeles Veterans AffairsHealthcare System, Los Angeles, CA, USAe Henry Ford Health System, Detroit, MI, USAf Department of Molecular and Medical Pharmacology, University of California Los Angeles, LosAngeles, CA, USA

Received 16 August 2016; accepted 16 August 2016Available online 25 August 2016

KEYWORDSCancer metastasis;Peripheral blood;Circulating tumorcell;ex vivo culture

Abstract Recent cancer research has demonstrated the existence of circulating tumor cells(CTCs) in cancer patient’s blood. Once identified, CTC biomarkers will be invaluable tools forclinical diagnosis, prognosis and treatment. In this review, we propose ex vivo culture as arational strategy for large scale amplification of the limited numbers of CTCs from a patientsample, to derive enough CTCs for accurate and reproducible characterization of the biophys-ical, biochemical, gene expressional and behavioral properties of the harvested cells. Because

* Corresponding author.E-mail address: Ruoxiang.wang@cshs.org (R. Wang).Peer review under responsibility of Second Military Medical University.

http://dx.doi.org/10.1016/j.ajur.2016.08.0052214-3882/ª 2016 Editorial Office of Asian Journal of Urology. Production and hosting by Elsevier B.V. This is an open access article underthe CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/ajur

Asian Journal of Urology (2016) 3, 240e253

of tumor cell heterogeneity, it is important to amplify all the CTCs in a blood sample for acomprehensive understanding of their role in cancer metastasis. By analyzing critical stepsand technical issues in ex vivo CTC culture, we developed a cost-effective and reproducibleprotocol directly culturing whole peripheral blood mononuclear cells, relying on an assumedsurvival advantage in CTCs and CTC-like cells over the normal cells to amplify this specifiedcluster of cancer cells.ª 2016 Editorial Office of Asian Journal of Urology. Production and hosting by Elsevier B.V. Thisis an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction

Given its life-threatening consequences, cancer metastasiscould be considered the most crucial topic in cancerresearch. The mechanism of metastasis should be intenselyinvestigated to define the critical players for therapeutictargeting to improve patient survival. Based on historicalprecedents, in which once-incurable diseases becamecurable or successfully managed by medical oncologistsafter advances in biomedical and clinical science [1,2],cancer metastasis will be curable in the future. The overallcancer death rate in the United State, for example, hasbeen falling over the last few decades [3], thanks mainly tothe identification of tobacco use as an oncogenic risk factorand a protracted period involving in multistep nature ofcarcinogenesis. In addition, the recognition of extensivetumor cell heterogeneity has led to personalized oncology,which is markedly improving cancer treatment and patientsurvival [4e6]. A full understanding of metastatic tumorpathophysiology, in combination with a careful examinationof the tumor microenvironment for a clear definition ofparticipant cell types and the signaling network in cancer-stromal and cancer-host immune interaction, may provideclinicians with critical mechanistic insights into the treat-ment of cancer metastasis [7,8].

This review discusses one of our ongoing research pro-jects focusing on isolation and characterization of circu-lating tumor cells (CTCs) following ex vivo culture.Expanding CTCs in large quantities will facilitate detailedgenotypical and phenotypical examination, allowingcausal abnormalities driving cancer metastasis to beaccurately identified and exploited as biomarkers forcancer diagnosis, prognosis and treatment. Thus, estab-lishment of a highly reproducible and efficient CTC prop-agation protocol holds the promise for advancing precisionmedicine and personalized oncology. This review summa-rizes the current status of high quality CTC research re-sults with special emphasis on ex vivo culture of thisspecific cancer cell type. Based on published literatureand our own work, we comment on critical issues facingthe development of CTC culture methodologies and pro-pose a rational strategy for culturing CTCs from cancerpatient peripheral blood samples. The goal of this dis-cussion is to attract research interest to the developmentof a cost-effective and reproducible protocol for ex vivoculture of a limited number of CTCs from a given clinicalblood sample into a large population amendable to mo-lecular characterization.

2. Limitations in the research of cancermetastasis

Many confounding factors make the study of cancermetastasis more difficult than the study of oncogenicinitiation, promotion and progression. Firstly, the study ofcancer initiation is based on a unified hypothesis, whichpostulates that cancer initiation is due to genomic mutationor genetic abnormality, while environmental risk factorspromote these perturbations and culminate in cancerdevelopment and local invasion. In contrast, metastasis isshown to be the result of interplay between cancer cellsand resident cells of the tumor microenvironment,involving mesenchymal, endothelial and immune cells atsecondary organ sites [9e11], where many of thesebystander cell types potentially affect cancer cells throughmultiple reciprocity [12,13]. The presence of plural vari-ants both in heterogeneous cancer cells and host micro-environment makes identification of critical effectorsdifficult. Secondly, the study of cancer development has alonger history facilitated by having immortalized cell linesas study subjects and genetically reconstituted animalmodels to investigate how specific genetic mutations affectcancer development. The study of cancer metastasis is lessgranular, due to the difficulty of manipulating the drivers ofcancer metastasis, while animal models often fail to reca-pitulate human metastatic profiles. Thirdly, cancer initia-tion and development can be studied with definedexperimental systems, in which the effect of a suspectedcausal factor can be quantitatively evaluated by multiplemodalities such as cell proliferation, apoptosis, migration,invasion and xenograft tumor formation. The study ofcancer metastasis, by contrast, is often hindered by a lackof optimal model systems to decipher the interaction be-tween cancer cells and the tumor microenvironment at themetastatic site. Finally, from the therapeutic perspective,many primary cancers can be cured either by surgery and/or adjuvant therapy, while there is a general lack of con-fidence in the curability of cancer metastasis [14e16],which often leads to the patient’s demise. It is challengingfor investigators to establish reliable in vitro models todelineate the mechanism of cancer metastasis at anorganismic level. It would be ideal if metastatic cancer celllines could be established routinely from each individualcancer patient.

Human cancers are broadly categorized into two groupsaccording to their lineage origin, tumors from mesenchymallineage and cancers of epithelial cell origin. Though the

Technical issues with ex vivo CTC culture 241

true cause of either cancer type remains to be defined,histopathologic analysis has shown that these two groups ofmalignancies have distinctive natures of development,progression and metastasis [17,18]. In the following dis-cussion, the term “cancer” refers mainly to malignancies ofepithelial origin.

Many cancer theories assume that compared to normalcells, tumor cells acquire survival and proliferation advan-tages through tumor cell evolution in situ. Counterintui-tively, tumor cells of epithelial origin are difficult toamplify by ex vivo culture [19e21]. Only a few successeshave been achieved establishing immortalized cancer celllines. Pancreatic cancer, for example, is a fast progressing,highly metastatic and devastating disease, but only a fewinvasive human pancreatic cancer cell lines have beenestablished to date [22,23]. The same is true of prostatecancer, another highly lethal disease with high incidence inWestern countries [24]. Cell line scarcity makes it espe-cially difficult to study cancer cell heterogeneity, known tocontribute ultimately to cancer progression and to anti-tumor therapeutic resistance. More importantly, theinability to culture cancer cells from patient tumor tissuesmakes it difficult to practice precision oncology at thepersonalized level. It would be ideal to grow tumor cellsex vivo from every cancer patient, and to then subjectthese cells to screening for the most effective regimenspecifically tailored to the individual patient. An improve-ment in primary human cell culture would contribute to theadvancement of cancer research and precision-based can-cer treatment.

Our laboratory’s research focus on the mechanism ofcancer metastasis has revealed the importance of cellularinteractions, which collectively contributing to metastaticdissemination of prostate cancer cells to the mouse skel-eton [7,8,25e29]. We have found that cancer cells can re-cruit and reprogram naı̈ve non-tumorigenic bystander cellsto participate in tumorigenic and metastatic cascade viaactivation of critical transcription factors that confermesenchymal, stemness and neuroendocrine phenotypes tothe bystander cells [25,30,31]. These results exposed apreviously unrecognized facet of cancer metastasis: thoughthe mechanism of cancer initiation and development inprimary tumors may be a tumor cell-centered event, cancermetastasis is a bona fide systemic disease, in which manyinnocent cell types in the tumor microenvironment can be“transdifferentiated” to express malignant phenotypes andbecome active participants in the process of tumor cellspreading. Detection of these newly converted cells mayprovide sensitive biomarkers for metastasis detection, andblockade of the recruitment and reprogramming mecha-nism could be a new approach to delay metastasis andprolong patient survival.

3. CTCs as a marker of cancer metastasis

There is a general lack of sensitive biomarkers for diag-nosis, treatment, and prognosis of many metastatic cancers[32]. Conventional cancer research relies on comparativeanalysis of cancer versus normal control cells to identifydifferential biomarkers. Extensive biomarker search hasyielded few bona fide cancer biomarkers, partly because

the survival advantage of cancer cells is achieved bycoopting normal regulatory mechanisms, rather than byactivating cancer-specific “oncogenes”. Shared structureand gene expression between cancer and normal cellswould make it almost impossible to use “cancer-specific”biomarkers to distinguish the presence of cancer from othernonmalignant cells. In addition, published data in theliterature suggest that abnormalities of cancer cells aredynamic, changing from one pathologic state to anotheralong with abnormal cell division [6,33,34]. Clinically, onlythe appearance of certain morphologically distinct cells,indicative of cancer, in the secondary organ sites is taken asa concrete sign of cancer metastasis.

Many forms of patient samples can be used for diseasediagnosis, but the most common and least invasive samplingis liquid biopsy of peripheral blood. In cancer research, ithas long been postulated that the circulation system is athoroughfare for cancer spreading [35e37]. This postula-tion has now been supported by accumulating evidence, ascancer cells have been detected and isolated from bloodsamples both experimentally [38e40] and from clinicalcancer patients [41e43]. CTCs in cancer patient blood forman attractive liquid biopsy compartment that can be sub-jected to molecular interrogation to unveil genomic ab-normality in tumor cells and their fate in the metastaticcascade. With in-depth characterization, CTCs per se orunique features of the CTCs may be used clinically for thediagnosis, prognosis and treatment of cancer metastasis inthe future [44].

4. Argument for the necessity of ex vivo CTCculture

There are two strategies for studying CTCs. The popularstrategy presently is to study CTCs directly from patientblood. In a few cases, CTCs are amplified ex vivo into alarge population before investigation. We believe that forreproducible investigation, especially for behavioral char-acterization, CTCs have to be amplified, preferably intostable cell lines.

4.1. The practical value of ex vivo CTC culture

A manifest disadvantage of directly detecting CTCs forcancer diagnosis is the sparse and sporadic nature of thisabnormal cell type, which is mixed among the vast pop-ulations of peripheral blood mononuclear cells (PBMCs)[42,45,46]. This feature makes it difficult to perform anycomprehensive and reproducible analyses directly, eitherfor CTC detection and enumeration at the cellular level,or for genomic and gene expression studies at the mo-lecular level. Indeed, partly due to CTC heterogeneityand partly due to present-day technological limitations,even a confident counting of CTCs in any given patientblood sample is hard, as is correlation of CTC counts todisease progression or metastasis. Contributing factorsto the conundrum and the anemic understanding ofCTCs as a pathologic and tumor progression entity willbe discussed in detail in the following sections. Itseems that for CTC biomarker identification, an assess-ment of the heterogeneity of genomic and gene

242 R. Wang et al.

expression abnormalities can be made only by acquiring alarge amount of CTCs for concise molecular analyses andreproducible validation [47].

4.2. Unknown fundamental biophysical andbiochemical information of the CTCs

Our marginal knowledge of CTCs is in stark contrast to ourprofound understanding of the normal constituents of theperipheral blood. Whereas each blood cell type has beenmeticulously studied for its biophysical, biochemical, geneexpressional and behavioral characteristics, none of theseaspects is clearly defined for CTCs in the same sample.Individual PBMC constituents, for instance, are well char-acterized for their total counts, differential ratios, lineagerelationships, distinct cell sizes at resting state, cellulardensities, cytoplasmic granularities, nucleus morphologies,surface markers, growth and differentiation capacities,half-lives in circulation, biological functions and possiblepathophysiologic abnormalities. All these features could beexplored for the enumeration and isolation of a specificPBMC fraction. In contrast, none of these features has beenfully explored in CTCs, making it difficult to conduct unbi-ased detection and isolation of this abnormal cell type fromthe vast blood cell population. Ultimately the unique fea-tures of CTCs can only be identified by detailed charac-terization of live CTCs.

The objective of ex vivo CTC culture is to expand thelimited numbers of CTCs in a given blood sample into alarge population, preferably into an immortal cell line, sothat molecular, cellular and behavioral investigations canbe conducted on these cells with high accuracy andreproducibility.

5. Low success rate as the default of ex vivoCTC culture

Many attempts have made to propagate CTCs throughex vivo culture, with only limited success. This can bedemonstrated by the fact that though a huge literature onCTC research is available, only about 50 publications in thelast decade address culturing CTCs from clinical patientblood samples. Most of these cultures are short-term,lasting from 3 days to 14 days [48e61]. Only a few long-term cultures have reported successfully established CTClines with demonstrated immortality that could be culturedstably for many passages [62e65]. These successes aresummarized in Table 1.

Though ex vivo CTC culture currently has low successrates, the successes indicate that some CTCs in patientblood are immortal and could be cultured ex vivo intostable cell lines. In this discussion, long-term CTC culturewill be the main issue of examination, mainly because long-term CTC culture is hindered by technical difficulties [19].

5.1. Rarity of CTCs by detection and enumeration

The main difficulty is the extreme rarity of CTCs. Forinstance, a peripheral blood sample contains huge numbersof red blood cells (around 5 � 109/mL), platelets (around

4 � 108/mL), and PBMCs (around 6 � 106/mL). In contrast,CTC counts in published studies have been detected as two-digit numbers in a 7.5 mL blood [66]. CTCs thus exist as aminute fraction among the vast number of blood cells, andwithout ex vivo expansion it would be difficult to conductany comprehensive characterization on this elusive celltype.

5.2. The caveat of using peripheral blood fortranslational CTC study

CTCs are presumed to be in a dynamic equilibrium withhomeostasis of the tumor in situ, and are considered as asurrogate marker for tumor status. CTCs, however, may notbe uniformly distributed in the blood. Analogous to the non-uniform distribution of cells, proteins, hormones, and gasesin the circulation system, a non-uniform CTC distributionshould be expected, based on anatomy and physiology ofthe circulation system. As depicted in Fig. 1, distribution ofCTCs may be determined by two factors: origin of the CTCs,and size of the CTCs. Individual CTCs will be filteredthrough two serial capillary beds; and in periphery, CTCsare transported in parallel circuits. Structure of the sys-temic circulation implies that peripheral blood may onlycontain the least and smallest CTCs. Nonetheless, it is thisfraction of rare and small CTCs that are mediating tumorspreading. The clinical value of peripheral blood samplingfor CTC detection and enumeration should be criticallyevaluated.

5.3. CTC immortality

Though immortality is ideal for ex vivo CTC culture, in theclinical setting immortality is not essential for CTCs tometastasize from the primary tumor to a distant site. A CTCdoes not have to be immortalized to cause metastaticgrowth, as long as a viable CTC can be transported todistant sites by the circulation system, which is known to beswift and highly efficient for corporal distribution. Thispresumption is in agreement with the finding that CTCsfrom patient blood are vulnerable [67,68], just as they aresusceptible to ex vivo culture conditions [19]. Based onextremely low published success rates, and our own CTCculture experience, it is intriguing to postulate that untillate cancer stages, when heterogeneity becomes so thor-ough that certain tumor cells have extensively evolved,immortal CTCs may be rare in the patient’s blood. Thisobservation predicts that, by default, ex vivo cultureshould be a low-success-rate practice, especially when alarge number of cancer patients are sampled without pre-selection for late-stage cases. Development of an optimalprotocol for ex vivo culture is thus a critical subject forreproducible CTC research.

6. Technical issues in the various CTC isolationmethods

It is habitual to assume that, in order to culture CTCs, thesecells should first be isolated from patient blood samples.Most reported CTC culture protocols employ a two-step

Technical issues with ex vivo CTC culture 243

strategy: isolating the CTCs from whole blood and thensubjecting the isolated cells to ex vivo culture underdefined conditions. A summary of CTC isolation strategies isshown in Fig. 2. Though there are many conventionaltechnologies for isolating a specific cell type from a mixedsample, none is suitable for the isolation of CTCs fromcancer patient blood, because specific traits that can beexplored for a clean isolation of CTCs have not yet beendefined.

6.1. CTC isolation by density gradient fractionation

In many reports, CTCs are initially isolated from erythro-cytes together with the PBMC population. Cellular density-based isolation with Ficoll polysaccharides is one of thebest-established methods for isolating PBMCs from a bloodsample. This method, nonetheless, only allows the isolationof a lighter fraction of the PBMCs. Ficoll-Paque mediumwith a density of 1.077 g/mL (Ficoll-Paque 1077), forexample, selectively isolates monocytes and lymphocytesthat in comparison to red blood cells, have lighter den-sities [69,70]. The use of Ficoll-Paque with a density of1.084 g/mL may additionally isolate basophils, but leavesbehind heavier neutrophils and eosinophils. In our study ofCTCs from prostate and pancreatic cancer patient bloodsamples [42], some candidate CTCs appeared much heav-ier, and could not be isolated by Ficoll-Paque 1.077 or 1.084banding. Because the density of CTCs is not defined andprobably is not uniform, using density gradient to isolateCTCs in monocyte/lymphocyte banding is not supported bybiophysical data.

Alternative strategies for isolating CTCs directly fromblood sample encounter difficulties as well. Most suchmethods are based on surface marker expression: detectingand isolating CTCs that bear a predetermined surfacemarker absent from normal blood cells. Both positive andnegative selection strategies are commonly used. Neitherof these strategies guarantees a comprehensive isolation ofall the CTCs in a given sample, but may introduce compli-cations to later analyses.

6.2. CTC isolation by positive selection

To isolate by positive selection, CTCs are presumed to haveretained surface marker expression of the primary tumorcells. Since primary cancers are of epithelial origin, andsince most cancer cells carry epithelial surface markers,these markers are often applied for fishing CTCs from theblood. Epithelial marker expression, however, is known tobe altered and may even be lost in cancers. The epithelialcell adhesion molecule (EpCAM), for example, is universallyexpressed on normal epithelial cells. It is involved inmaintenance of intercellular adhesion in an epitheliallayer, clenching epithelial cells into a static and docilestate [71,72]. During cancer development and progression,however, its expression would be reduced or lost [73e75]due to multiple oncogenic signaling, or through epithelialto mesenchymal transition. Significantly, loss of these typesof epithelial markers accounts for the acquired mobility,and invasive and migratory potential, of cancer cells. Theformation and shedding of CTCs into blood circulation,therefore, could imply that these special cancer cell types

Table 1 Summary of successes in long-term CTC culture with clinical cancer patient blood samples.

Disease Method of CTCisolation

CTC culture conditions Groupsize

CTC linesestablished

Morphology Reference

Breastcancer

CD45/CD66 iChipnegative selection

RPMI 1640 medium,serum-free, with EGF, FGF,and B27, in 4% O2,spheroid culture

36 6 Spheroids Yu, et al.,2014 [62]

Coloncancer

CD45 RosetteSepnegative selection

DMEM: F12 medium, 2% FBS,with insulin, EGF, and bFGF,supplemented with N2,in 2% O2 for 2 weeks;then in RPMI 1640 medium,2% FBS, with insulin, EGF,bFGF, supplemented withtransferrin and selenite,in normoxia

71 1 Spheroids Cayrefourcq,et al., 2015 [63]

Prostatecancer

Ficoll-Paque densitygradient and CD45RosetteSep negativeselection

DMEM: F12 medium, with EGF,bFGF, FGF10, supplementedwith R-spondin 1, DHT, B27,nicotinamide, A83-01,SB202190 and Y27632,in Matrigel

17 1 Spheroids Gao, et al.,2014 [64]

Small celllung cancer

Ficoll-Hypaquedensity gradient

RPMI 1640 medium, serum-free,with insulin, IGF-1 and selenite

30 3 Spheroids orattached

Hamilton,et al., 2015 [65]

bFGF, basic fibroblast growth factor; CTC, circulating tumor cells; DHT, dihydrotestosterone; EGF, epidermal growth factor; FBS, fetalbovine serum; FGF, fibroblast growth factor; IGF-1, insulin-like growth factor 1.

244 R. Wang et al.

have lost most of their EpCAM surface expression. Due totumor cell heterogeneity in epithelial marker expression,the use of epithelial proteins as surrogate cancer cellmarkers may detect certain CTCs, but may not be an idealstrategy for comprehensive isolation of all the CTCs in agiven blood sample. Importantly, it should be kept in mindthat the CTCs whose epithelial behavior is completely lostmight be the ones that are more representative of themalignant stage of disease [76e78].

Another drawback to positive selection is that thismethod is inherently harmful to CTCs, as isolated CTCs are

no longer viable for ex vivo culture. Firstly, antibody-binding could be harmful to cell survival. Some antibodiesagainst EpCAM, for example, may directly cause cancer celldeath [79] or indirectly induce antibody-dependent cell-mediated cytotoxicity (ADCC). Indeed, anti-EpCAM medi-ated cancer cell killing is a mechanism sought in studiesinvolving adecatumumab, edrecolomab and catumaxomab.Secondly, positive selection relies on mechanisms such asantibody-antigen ligation, ligand-receptor binding, orintegrin-extracellular matrix (ECM) adhesion for CTCrecognition. To physically isolate CTCs, the recognition partof the ligands in the forms of antibodies, peptides, ororganic molecules, has to be affixed onto a solid or semi-solid surface, either magnetic beads, microfluidic chips,or polysaccharides, so the CTCs can be tethered onto thesurface and then isolated by physical means. In many cases,isolated CTCs remain affixed onto the solid surface and areeasily killed by dehydration. In some cases, live CTCs mayactively attach onto the surface due to close adjacency. Tofree the isolated CTCs from the solid surface, eitherenzymatic digestion to break the affixation or laser capturemicro-dissection (LCM) would most likely kill the CTC aswell.

6.3. CTC isolation by negative selection

Technical difficulties are similarly seen in CTC isolation bynegative selection, which is designed to deplete all normalPBMCs from a blood sample using well-known hematopoi-etic surface markers, leaving only CTCs behind. The majordrawback with this strategy is that not all normal cells in ablood sample can be removed by negative selection,because a fraction of hematopoietic cells does not expressknown hematopoietic surface markers [80]. In addition, notall blood cells are hematopoietic, so a peripheral bloodsample may contain many non-hematopoietic cell typesbesides CTCs [81].

CD45 isoforms, for example, are said to be expressed onall cells of the hematopoietic lineage and are commonlyused in negative selection to deplete hematopoietic cellsfrom blood. Not all blood cells, not even all hematopoieticcells, are CD45-positive [80]. Though the frequency ofCD45-negative cells in human blood has not been system-atically quantitated, this fraction of cells is attracting moreresearch. In different patient groups and with differentdetection settings, CD45-negative cells are found to rangefrom 0.02% [82] to almost 2% [80] of PBMCs. These maytranslate into a CD45-negative count ranging from 1200/mLto 120,000/mL blood sample, still much higher than theassumed two-digit CTC counts in the same volume. Theresult of a CD45-negative selection, therefore, is CTCsmixed with many poorly characterized but otherwisenormal cells of the blood. Similarly, an alternative negativedepletion with human lineage marker (Lin, a cocktail ofantibodies against known cell types of the entire hemato-poietic lineage) would result in CTCs mixed with variousLin-negative normal cells [80,83]. Though negativeselection may have less effect on CTC viability, thismethod fails to yield a pure CTC population and willpotentially be misleading by totaling enriched normal bloodcells as CTCs.

Figure 1 Size-limiting structure of the blood circulationsystem. A diagram of the human blood circulation systemshows the likeliest site to find CTCs. Pulmonary circulation(upper portion) may be considered as a serial perfusion system,transporting CTCs (golden balls) from lung tumors to theoxygenated blood (red), where the CTCs will be filtered byperipheral capillary beds. Systemic circulation (lower portion)is comprised of parallel vascular systems. Two parallel capillarybeds are used to show scenarios of CTC circulation. In onescenario, CTCs (red balls) from tumors of visceral organs maybe shed to the internal vein, and will be filtered by the pul-monary capillary network. In another scenario, blood samplesfrom upper limb superficial veins are commonly used for CTCstudies. Though this is convenient and less invasive for routineclinical phlebotomy, cancer patient blood samples from theseveins may contain only the smallest CTCs in size and the fewestCTCs in number. According to this diagram, CTCs from lungtumors may be easier to detect in arterial blood, whereas CTCsfrom tumors in other organs may be easier to be detected fromthe blood approximate to the right side of the heart.

Technical issues with ex vivo CTC culture 245

6.4. CTC isolation by size-based technology

In addition to negative selection, there are other isolationmethods that may not guarantee a successful CTC isolationand may instead introduce distortions in CTC quality andquantity. Size-based isolation, for example, may not beable to isolate all the CTCs from a given blood sample,simply because the size distribution of CTCs in a given bloodsample is unknown. It is possible that, similar to establishedcancer cell lines, CTCs in a given patient blood sample mayhave varied cell sizes, overlapping the size range of resi-dent hematopoietic cells. This complexity is compoundedby the fact that PBMCs in the same blood sample highly varyin size as well, with diameters ranging from 7 mm for restinglymphocytes up to 30 mm for monocytes.

On the other hand, lumen dimensions of the limitingcapillary micro-vessels range from 7.5 mm to 10 mm indiameter, and pulmonary capillaries could be even nar-rower [84]. To be free to travel throughout the circulatingsystem, a CTC may have to adopt a small size, or not bemarkedly larger than PBMCs; otherwise CTCs may notcirculate through capillary vessels. The relationship of size-limiting capillary beds and CTC distribution is shown inFig. 1.

As shown in Fig. 1, a size-limiting structure of the bloodcirculation system additionally indicates that, unless bloodhas to be sampled from large internal veins of the patient,it would be difficult to expect that a size-based methodcould isolate out any pure CTC population. In this regard, amore invasive sampling from large visceral organ-drainingveins could indeed catch large cancer cells and even largetumor cell clusters [85e87], while our group has identified atype of CTCs with a small nucleus from peripheral bloodsamples, which correlated with the state of castration-resistant progression and metastasis of prostate cancer[45].

The distribution of CTCs in blood circulation is not uni-form, but is limited to certain segments of the vasculature.As pointed out in Fig. 1, any CTCs will be filtered againsttwo in-serial capillary beds, which must have enormousanti-tumor capacities for purging all the trapped CTCs.Also, the anatomical site of blood draw has to be consid-ered in CTC research. Arterial blood may be CTC-rich forcancers of the pulmonary circulation, and vena cava bloodshould be rich for CTCs from other organs perfused bysystemic circulation. In any case, because of capillaryfiltration and the parallel pattern of systemic circulation,peripheral blood always represents a CTC-poor sample

Figure 2 Current methodologies for ex vivo CTC expansion. Current strategies, methods, and technologies employed in CTCculture are summarized in this flowchart. Arrowed lines denote blood sample manipulative steps. Thin blue lines in the flowchartindicate the unique experimental steps developed in our laboratory. CDX, CTC-derived xenograft; CTC, circulating tumor cells;LCM, laser capture micro-dissection; 2-D, two-dimension; 3-D, three-dimension.

246 R. Wang et al.

(Fig. 1), although peripheral phlebotomy is the mostconvenient and least invasive sampling route.

As discussed above, the term “CTC” is presently anoperational concept, simply denoting any cancer cellsbeing shed from a tumor into blood circulation. Cancer cellsare highly plastic and are in dynamic evolution with pro-cesses mimetic to stem cell differentiation [88], trans-differentiation [89], or somatic cell fusion [26]. Cancer cellevolution ultimately results in a highly heterogeneous cellpopulation, with individual cells in a given patient dis-playing a wide spectrum of biophysical, biochemical, geneexpressional and behavioral alterations, none of which willbe representative of the entire tumor cell population.Similar scenarios of cellular heterogeneity may be presentin the CTCs in a patient’s blood sample, and any CTCstudies based on a predetermined parameter would onlydetect and isolate a fraction of the whole CTC population.

7. Current status of ex vivo CTC culture

Presently, isolated CTCs are difficult to maintain in a viablestate, and are rarely cultured ex vivo into a large popula-tion, least of all into stable CTC cell lines. Furthermore,CTCs have been said to be an inherently vulnerable celltype, possibly because the circulating blood may not be anamiable environment for CTC survival [67,68]. All theseconfounding factors make the isolation of viable CTCsdifficult.

7.1. Studying CTCs without amplification

Because isolated CTCs are not viable and cannot beexpanded in numbers by ex vivo culture, most ensuingmolecular and genetic investigations on CTCs have to beperformed on very limited numbers of CTCs, which yieldtoo few and too little starting materials for extensive andcomprehensive examinations. Presently, only a fewanalytical methods may be adopted for CTC examination,most as nucleotide-based assays, while the starting mate-rial has to be pre-amplified with polymerase chain reaction(PCR). Popular CTC assay methods include microarrayexpression profiling, single-cell sequencing, and quantita-tive PCR (qPCR) detection.

While single-cell analysis is a cutting-edge advance-ment, accumulating results suggest that many technolog-ical shortcomings of single-cell analyses remain to beovercome. A single cell, for instance, can now be used asstarting material for whole exon sequencing with the next-generation sequencing platform, but the result is oftensuboptimal. As a single run on the platform routinely putout more than one million sequencing reads, the reads froma single-cell analysis are found with frequent base ambi-guities, while computational analysis can assign only a smallfraction of these reads to a small fraction of encoding genesin the human genome.

The scenario becomes even more problematic forcomparative studies, because many assigned reads in onesingle CTC are not detectable in another single CTC, due torandomness and the sporadic nature of whole exonsequencing with limited starting material [90,91]. For

“profiling” expression in a single CTC, the potential powerof next-generation sequencing is severely limited in single-cell sequencing. It would be ideal if limited numbers ofCTCs could be amplified into a larger population by ex vivoculture, so comprehensive genomic and genetic analysescould be performed with high reproducibility and accuracy,and then compared confidently among different CTC cul-tures from individual patients against the stages of diseasedevelopment and progression.

7.2. CTC culture by short-term cultivation

Several laboratories have subjected isolated CTCs frompatient blood samples to short term ex vivo culture.Enriched CTCs from PBMCs in free form or on solid surfaceswere cultured ex vivo for 3e14 days. In these studies,choice of culture medium was mostly based on each labo-ratory’s experience, as with long-term culture studies(Table 1).

Reports with short-term culture demonstrated success-ful ex vivo expansion of isolated cells, as determined byamplification of cell numbers and colony formation[48e61]. In many cases, the expanded cells were shown tobe suitable subjects for immunohistological analysis,karyotypic examination, and gene expressional profiling.

As discussed above, short-term CTC culture may becompromised by impurity of the CTC isolation or by adistortion of the mixed blood cells. As the limited numbersof colonies grown in short-term culture may alleviate thelimitations seen in single-CTC analysis, results of CTCcharacterization from these studies have to be furthervalidated by additional means, because of the limits on CTCreproducibility. Nonetheless, these results provide sup-porting evidence that CTCs in patient blood samples can becultured. With careful optimization, these culture condi-tions may be applied to successful long-term culture.

7.3. Long-term culture to establish stable CTC lines

It is difficult in general to culture human cells in vitro, andeven harder to culture human cells into cell lines. It issimilarly difficult to culture CTCs from cancer patient bloodsamples, even though cancer cells are believed to haveevolved survival and proliferation advantages. As summa-rized in Table 1, there are nonetheless a few reports oflong-term culture success, in which CTCs in cancer patientperipheral blood samples have been expanded ex vivo intolarge population, and a few of these CTCs even showed thecapability of continuous passage for numbers of generations[62e65], suggesting that similar to established cancer celllines, these CTCs may harbor immortality.

The capability for independent survival and growth issaid to be a watershed feature of cancer progression.Tumor cells at advanced stages are thought to be able tosurvive adversities in their surroundings and proliferateindifferent to regulatory cues in the microenvironment.Paradoxically, many ex vivo CTC cultures resort to the helpof cytokines, growth factors, tissue organ extracts, hor-mones, and small molecules to facilitate CTC survival andgrowth in vitro [62e65]. The optimal conditions for ex vivoCTC culture remain to be defined.

Technical issues with ex vivo CTC culture 247

Current CTC culture recipes are mostly varied adoptionsof stem cell culture conditions. Besides cocktails of con-ventional cell culture mediums, proprietary mediums areoften used in CTC culture, but may not guarantee CTCamplification. In addition to alternative choices of animalserum, different recipes are found using epidermal growthfactor (EGF) and basic fibroblast growth factor (bFGF, orFGF2), which are often seen in recipes for stem cell cul-ture. Additional growth factors used include fibroblastgrowth factor-10 (FGF10), granulocyte macrophage colony-stimulating factor (GM-CSF), insulin and insulin-like growthfactor 1 (IGF-1). In some cases, R-spondin 1 protein is addedin an attempt to sustain survival and growth of epithelialcells through WNT/b-catenin signaling [92]. In cases whenserum-free CTC culture media are used, the recipes mostlyinclude proprietary B27 and N2 supplements, together withinsulin, transferrin and selenium (ITS) supplementation.

Other factors that may promote stem cell growth arealso used, including dihydrotestosterone (DHT) and nico-tinamide. In addition, small molecule inhibitors for specificsignaling pathways have been tested for promoting CTCgrowth. Inhibitors of A83-01 and ALX-270-445, for example,may specifically inhibit the function of TGFb type 1 re-ceptor kinase (ALK5) and so prevent epithelial to mesen-chymal transition and promote CTC survival andproliferation. SB202190, a specific P38 inhibitor, may pre-vent apoptosis due to P38 activation. Y27632, a cellpermeable inhibitor of Rho-associated coiled-coil contain-ing protein kinase (ROCK), is used in CTC culture to promoteorganoid growth and formation.

The most interesting phenomena in all the published CTCculture studies is how few of the cultured CTCs display anyexpectedmorphologies or behaviors of an epithelial lineage.Instead, most of the successful cultures yielded CTCsgrowing in suspension in 3-dimensions (3-D), aggregating to aspheroid or organoid organization. Intriguingly, thesemorphologic and behavioral features are commonly seenwith stem cell culture, and are taken as a sign of stemness[93,94]. Though it remains to be determined whether CTCsamplified by ex vivo culture are cancer stem cells, theseresults confirmed that CTCs represent cancer cells that havecompletely lost epithelial marker expression.

In addition to conventional suspension culture setups,more ex vivo CTC cultures are adopting various 3-D formats.Soft agar colony formation, non-adherent culture, andspheroid or organoid culture on growth factor-reducedMatrigel matrix are popular CTC culture systems. Toinduce survival responses in CTCs, hypoxic culture condi-tions or consecutive culture in hypoxic and normoxic envi-ronment are used.

It should be noted that, because optimal CTC culturecondition has yet to be defined, current research relies onstem cell culture methods to ensure a maximal CTCexpansion. Each protocol is designed based on the re-searcher’s own expertise and empirical results, not by arationale based on CTC biological features, which arepresently not defined. Published protocols yielded very lowsuccess rates, generally less than 5% of the patient bloodsamples cultured. Rationale-based modifications should betested to improve the success rate.

One of the future modifications probably should be areduction or complete avoidance of exogenous growth

factors and small molecules in CTC culture, because CTCsgrown with the support of these factors may have limitedapplications for the mechanistic investigation of cancerprogression and metastasis. Cancer cells already have ac-quired growth and survival autonomy. On the other hand,cancer metastasis, homing, and colonization are charac-terized by cancer cell interaction with resident cells of themesenchymal lineage in the tumor microenvironment, andthe interaction has been shown to be mediated by growthfactors and can be modulated by small molecule signalingpathway inhibitors. CTCs cultured in the presence ofgrowth factors and small molecule inhibitors thus may notbe ideal subjects for modeling cancer and stromalinteraction.

8. Our experience with ex vivo CTC culture

Our laboratories study the mechanism of cancer progressionand metastasis. As CTCs are plausible mediators in distanttumor cell spreading, we used mouse xenograft models totrack CTCs during metastasis [39,95]. In recent years, wehave established close collaborations with clinical oncolo-gists and surgeons in translational studies with clinical bloodsamples from cancer patients. These intercalated researchinterests led to a platform for translational research oncritical issues relevant to the welfare of cancer patients.

8.1. Modeling CTC culture with experimentalanimals

We initially carried out studies in tumor-bearing mice toconfirm that there indeed were CTCs in animal blood. Tofacilitate tracking live CTCs, we tagged the ARCaPE humanprostate cancer cells with a stable expression of red fluo-rescence protein (RFP). Selected ARCaPE-RFP clones wereinoculated orthotopically to the prostates of athymic micefor xenograft tumor formation, which caused a wide spreadof cancer cells to internal organs in addition to skeleton[39,95]. In one of the experiments, intra-cardiac bloodsamples were observed to contain red fluorescent cells,indicating that metastasis of the xenograft tumor wasmediated by these cells.

We then drew blood from the tumor-bearing mice andtested a basic protocol for expanding CTCs through ex vivoculture [39]. This series of experiments led to the devel-opment of a simple and cost-effective protocol, with whichCTCs in a drop of blood (about 75 mL) from either a tail cutor from the left ventricle were enough to form multiplecolonies of red fluorescent cells. Interestingly, these col-onies were markedly different from parental ARCaPE-RFPcells in both morphological, behavioral and gene expres-sional senses; and the changes were irreversible, with CTC-derived clones losing most epithelial features and acquiringconspicuous mesenchymal cell traits. We subsequentlyrepeated the ex vivo CTC culture with a different xenograftmodel, in which RANKL overexpression converted androgen-dependent non-tumorigenic LNCaP prostate cancer cells toandrogen-independent, highly tumorigenic, metastatic cells[25]. From these mice, blood samples from retro-orbitalsinus, submandibular vein, tail cut and left ventricle couldall be cultured to form cancer cell colonies (unpublished

248 R. Wang et al.

data). These preliminary results prompted us to test CTCex vivo culture from clinical cancer patient blood samples.

8.2. Modification of PBMC isolation

We obtained cancer patient samples in the form of packedblood cells, the cellular fraction after plasma was taken forclinical use. In some cases, both blood and surgical tumorfrom the same patient were sampled and subjected toidentical ex vivo culture, in order to validate the culturemethod.

We first tested if CTCs were present in cancer patientperipheral blood as well. PBMCs were isolated from pros-tate cancer patient peripheral blood samples with twomethods. With the first method, density gradient banding inFicoll-Paque 1077 or 1084 was used to isolate the monocyteand lymphocyte fraction. Candidate CTCs in this fractionwere detected with fluorescein-conjugated anti-EpCAMantibody, together with NIR-783, a prototype heptamethinecarbocyanine near infrared organic dye, which we havedetermined to have tumor cell specificity [96]. This exam-ination demonstrated CTC-like signals in the patient blood,and the count of these cells correlated well with diseaseprogression in general trends [42]. These results confirmedthat there were detectable numbers of CTCs in bloodsamples of prostate cancer patients, especially in thosewith advanced disease.

On the other hand, this series of experiments revealedthat density gradient centrifugation was not an idealmethod for isolating all CTCs from a given patient bloodsample, because after removal of the monocyte andlymphocyte band, additional EpCAMþNIRþ signals could bedetected from the remaining heavier Ficoll-Paque frac-tions. The signal was even detected from the heaviesterythrocyte pellet. This finding indicated that in a givencancer patient blood sample, heterogeneous CTCs mightvary in density, making gradient fractionation a less thanoptimal method for comprehensive isolation of all CTCsfrom the sample.

With the Ficoll-Paque isolated PBMCs, we additionallytested whether CTCs could be sorted from PBMCs by flowcytometry. Though anti-EpCAM antibody and NIR-783stained cells could be isolated by flow cytometric sortingin all the patient blood samples tested, the number ofsorted cells was mostly too low and too diluted in sortingfluid for successful ex vivo culture.

The second method we tested was whole PBMC isolationby ammonium chloride hemolysis. Because it recovers allthe nucleated cells from a blood sample in a cost-effectiveand highly reproducible manner, ammonium chloride he-molysis is now becoming the mainstay of PBMC isolation. Toassess whether it could be a favorable method for ex vivoCTC culture, we found that some hemolysis formulas hadmarked time-dependent toxicity to PBMCs and even toestablished human prostate cancer cell lines. Accordingly,we evaluated a series of formulas to identify a modifiedrecipe, which was potent for hemolysis but carried littlecytotoxicity to nucleated cells. In subsequent CTC cultures,this modified PBMC isolation protocol has been found to bemarkedly cost-effective, less labor intensive, and highlyreproducible.

8.3. Culturing entire PBMC populations for ex vivoCTC expansion

As discussed above, the process of isolation either affectedCTC viability or yielded a mixed cell population. In addi-tion, CTC isolation is a costly practice not suitable to largescale studies or routine CTC culture. We reasoned that theisolation step may not be essential to ex vivo culture ofCTCs, if these specified cancer cells indeed had survival andproliferation advantages over other constituent cells inPBMCs. We explored the possibility of culturing CTCs byskipping the isolation step.

It is well known that most blood cells of hematopoieticlineage are terminally differentiated and have lost prolif-eration capability. Though the T and B lymphocytes harborproliferation potential, proliferation has to be triggered byspecific antigen receptor ligation [97], by a Ca2þ influxtogether with a PKC activation [98], or by super-antigencross-linking [99]. Moreover, the activated lymphocyteproliferation is transient, ending with a programmedactivation-induced cell death [100e102]. In general,without additional stimulation, PBMCs in resting statewould perish in ex vivo culture within days, each cell typedying in a programmed manner. Macrophages are the onlycell type in PBMCs capable of survival for weeks in culture.Macrophages, however, are also a terminally differentiatedcell type incapable of ex vivo proliferation. The constantpresence of some macrophages may not affect the survivaland proliferation of non-hematopoietic cells, includingCTCs. These observations suggested that CTCs could beamplified naturally by culturing whole PBMCs, should CTCsindeed have survival and proliferation advantages.

8.4. Choosing a culture medium

Different from culturing immortalized cell lines, whichwould survive and proliferate in vitro under varied cultureconditions, culturing primary human cells is thought torequire specific culture conditions. For ex vivo CTC culture,however, an optimal culture medium has yet to be defined,because we do not know the preferred conditions for CTCsurvival and proliferation, and because we are not evensure whether the determinants for CTC survival and growthreside within the cell or outside the cell.

To choose a practical medium formula for ex vivo CTCculture, we tested various culture mediums and conditions,with reference to published methods for various primaryhuman cells. We concluded that, though nutrition-rich orstem cell culture mediums may prolong the survival ofprimary tumor cells, some PBMCs, and probably CTCs, nonepromoted cellular proliferation. Rather, the subject cellsadopted a “blast” morphology but undertook little cell di-vision. In contrast, some regular recipes for routine main-tenance of commonly used cancer cell lines were amiablefor the proliferation of CTC-like cells, which could surviveand grow into large population, some even showingimmortality behaviors, capable of proliferation beyond 60continuous passages [103,104].

Our choice of conventional medium for CTC culture wasalso supported by our results culturing surgical tumorspecimens from cancer patients. We have established new

Technical issues with ex vivo CTC culture 249

prostate cancer and pancreatic cancer cell lines byculturing clinical surgical tumor specimens (unpublisheddata), and these cell lines are currently being character-ized. As both nutrition-rich stem cell culture mediums andconventional mediums facilitated outgrowth of tumor cells,we found that immortalized cancer cell lines could only beestablished from cultures in conventional medium, albeit atlow success rates. These observations strongly suggestedthat, for primary human cell culture, survival and prolif-eration is determined by mechanisms inside the cell, ratherthan by the conditions of the ex vivo culture.

8.5. Establishing stable cultures of CTC-like cellpreparations

Using our ex vivo culture protocol, we have accumulated apanel of immortalized cells from clinical cancer patientblood samples. While these cells have yet to be fullycharacterized, results from xenograft tumor formationhave confirmed their malignancy to be similar to the pa-tients’ primary disease (unpublished data). With a cost-effective and straightforward method established, furtherfine tuning of culture conditions and standardizing phle-botomy and sample preparation protocols may markedlyimprove the success rates of CTC culture. Additionally, weare identifying biomarkers from the established CTC-likecell lines. Through CTC biomarker detection from pro-spective patient blood samples, these CTC-like cell lineswill be validated as true CTCs. It is rational to predict thatin the future these markers will serve well as surrogates forclinical CTC detection during cancer diagnosis, treatmentand prognosis, so time-consuming CTC culture from everycancer patient would not always be needed for precisiononcology and personalized cancer therapy.

Though ex vivo CTC survival and proliferation aredetermined by endogenous mechanisms, our preliminarystudies suggested that these abilities could be modulatedby exogenous growth factors and specific small moleculepathway inhibitors, similar to published observations byother laboratories. The application of exogenous factors,however, could not increase the success rate of CTC cultureand might introduce undesired complications to furtherinvestigation. For instance, upon intra-cardiac inoculation,tissue-specificity of homing and colonization of the inocu-lated CTCs in host mice would be determined by exogenouscues in the tumor microenvironment [8]. We reasoned thataddition of growth factors in the ex vivo CTC expansion stepmay pre-condition these cells into a peculiar state, andthus complicate data interpretation in studies particularlyrelevant to CTC colonization at secondary organ sites.

8.6. Establishment of CTC-derived xenograft tumormodels

Our ex vivo culture has resulted in the establishment of apanel of CTC cultures from peripheral blood samples ofboth pre-surgical and non-resectable pancreatic cancercases, and also from both primary pre-surgical andcastration-resistant and metastatic prostate cancer pa-tients. Besides determining the malignant characteristics ofthese CTCs in vitro, we assessed the tumorigenicity of each

of these preparations with a CTC-derived xenograft (CDX)model. CTCs in mixture with Matrigel were inoculatedsubcutaneously, intracardially or orthotopically to thepancreas or the prostate of athymic mice. Prevalent andrapid xenograft tumor formation and lethality wereobserved in host mice, confirming the malignant nature ofthese cells [103,104]. Importantly, the tumor-bearing micedisplayed frequent and multiple metastases in distant or-gans. As it is well known that human cancer cell xenografttumors rarely cause metastasis in experimental mice, theseCDX models may have inherited the highly invasive andmetastatic nature of the patients’ in situ tumor. Charac-terizations of the CDX models, in the context of correlationwith the original clinical specimens, and evaluation of theirapplication in prostate and pancreatic cancer translationalresearch are underway.

9. Conclusion

In this review, we critically examined the major steps ofex vivo culture protocols, to identify the technical chal-lenges to successful CTC amplification from cancer patientblood samples. We think that the most critical issue facingCTC research is the lack of information on the fundamentalcharacteristics of CTCs in terms of their biophysical,biochemical, gene expressional, and behavioral features.Due to this lack of information, no reliable methods can yetbe designed for isolating CTCs from vast numbers of bloodcells. A full illustration of the fundamental characteristicsof CTCs is an urgent topic in cancer research.

In keeping with overall cancer cell heterogeneity, CTCsmay be highly heterogeneous as well. Because it is chal-lenging to isolate all the heterogeneous CTCs from a givenpatient peripheral blood, in our laboratory we skipped theCTC isolation step by culturing whole PBMCs in order toamplify CTCs, relying on their survival and proliferationadvantages. Using this strategy, we have established apanel of CTC-like cell lines from pancreatic, prostate,kidney, breast and gastric cancer patient blood samples.Full characterizations of these cells are underway. Thesemolecular interrogations could lead to the identification ofCTC biomarkers and molecular signaling network underlyingtherapeutic responses and resistance observed in patients.

Conflicts of interest

The authors declare no conflict of interest.

Acknowledgments

This work is supported by US NIH/NCI research grant(2PO1CA098912), Cedars-Sinai Medical Center Board ofGovernors Cancer Research Chair (LWKC), and US NIH/NCI(UO1CA198900) (HRT/EMP).

References

[1] Dubey AK, Gupta U, Jain S. Epidemiology of lung cancer andapproaches for its prediction: a systematic review andanalysis. Chin J Cancer 2016;35:71.

250 R. Wang et al.

[2] Hanna N, Einhorn LH. Testicular cancer: a reflection on 50years of discovery. J Clin Oncol 2014;32:3085e92.

[3] Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CACancer J Clin 2016;66:7e30.

[4] Cresswell GD, Apps JR, Chagtai T, Mifsud B, Bentley CC,Maschietto M, et al. Intra-tumor genetic heterogeneity inwilms tumor: clonal evolution and clinical implications.EBioMedicine 2016;9:120e9.

[5] Parker JS, Perou CM. Tumor heterogeneity: focus on theleaves, the trees, or the forest? Cancer Cell 2015;28:149e50.

[6] Sun XX, Yu Q. Intra-tumor heterogeneity of cancer cells andits implications for cancer treatment. Acta Pharmacol Sin2015;36:1219e27.

[7] Chung LW, Huang WC, Sung SY, Wu D, Odero-Marah V,Nomura T, et al. Stromal-epithelial interaction in prostatecancer progression. Clin Genitourin Cancer 2006;5:162e70.

[8] Sung SY, Hsieh CL, Law A, Zhau HE, Pathak S, Multani AS,et al. Coevolution of prostate cancer and bone stroma inthree-dimensional coculture: implications for cancer growthand metastasis. Cancer Res 2008;68:9996e10003.

[9] Chappard D, Bouvard B, Basle MF, Legrand E, Audran M. Bonemetastasis: histological changes and pathophysiologicalmechanisms in osteolytic or osteosclerotic localizations. Areview. Morphologie 2011;95:65e75.

[10] Miles FL, Sikes RA. Insidious changes in stromal matrix fuelcancer progression. Mol Cancer Res 2014;12:297e312.

[11] Yuan Y. Spatial heterogeneity in the tumor microenviron-ment. Cold Spring Harb Perspect Med 2016;6.

[12] Liotta LA, Kohn EC. The microenvironment of the tumour-host interface. Nature 2001;411:375e9.

[13] Place AE, Jin Huh S, Polyak K. The microenvironment inbreast cancer progression: biology and implications fortreatment. Breast Cancer Res 2011;13:227.

[14] Schofield PE, Stockler MR, Zannino D, Tebbutt NC, Price TJ,Simes RJ, et al. Hope, optimism and survival in a randomisedtrial of chemotherapy for metastatic colorectal cancer.Support Care Cancer 2016;24:401e8.

[15] Seah DS, Scott S, Guo H, Najita J, Lederman R, Frank E, et al.Variation in the attitudes of medical oncologists towardresearch biopsies in patients with metastatic breast cancer.Oncologist 2015;20:992e1000.

[16] Tanriverdi O, Yavuzsen T, Akman T, Senler FC, Taskoylu BY,Turhal S, et al. The perspective of non-oncologist physicianson patients with metastatic cancer and palliative care(ALONE study): a study of the palliative care working com-mittee of the Turkish oncology group (TOG). J Cancer Educ2015;30:253e9.

[17] Mohseny AB, Hogendoorn PC. Concise review: mesenchymaltumors: when stem cells go mad. Stem Cells 2011;29:397e403.

[18] Pinchuk IV, Mifflin RC, Saada JI, Powell DW. Intestinalmesenchymal cells. Curr Gastroenterol Rep 2010;12:310e8.

[19] Maheswaran S, Haber DA. Ex vivo culture of CTCs: anemerging resource to guide cancer therapy. Cancer Res 2015;75:2411e5.

[20] Centenera MM, Raj GV, Knudsen KE, Tilley WD, Butler LM. Exvivo culture of human prostate tissue and drug development.Nat Rev Urol 2013;10:483e7.

[21] Mitra A, Mishra L, Li S. Technologies for deriving primarytumor cells for use in personalized cancer therapy. TrendsBiotechnol 2013;31:347e54.

[22] Hwang CI, Boj SF, Clevers H, Tuveson DA. Preclinical modelsof pancreatic ductal adenocarcinoma. J Pathol 2016;238:197e204.

[23] Deer EL, Gonzalez-Hernandez J, Coursen JD, Shea JE,Ngatia J, Scaife CL, et al. Phenotype and genotype ofpancreatic cancer cell lines. Pancreas 2010;39:425e35.

[24] Gunderson K, Wang CY, Wang R. Global prostate cancerincidence and the migration, settlement, and admixturehistory of the Northern Europeans. Cancer Epidemiol 2011;35:320e7.

[25] Chu GC, Zhau HE, Wang R, Rogatko A, Feng X, Zayzafoon M,et al. RANK- and c-Met-mediated signal network promotesprostate cancer metastatic colonization. Endocr Relat Can-cer 2014;21:311e26.

[26] Wang R, Sun X, Wang CY, Hu P, Chu CY, Liu S, et al. Spon-taneous cancer-stromal cell fusion as a mechanism of pros-tate cancer androgen-independent progression. PLoS One2012;7:e42653.

[27] Wang R, Xu J, Juliette L, Castilleja A, Love J, Sung SY, et al.Three-dimensional co-culture models to study prostate can-cer growth, progression, and metastasis to bone. SeminCancer Biol 2005;15:353e64.

[28] Zhau HE, Goodwin TJ, Chang SM, Baker TL, Chung LW.Establishment of a three-dimensional human prostate orga-noid coculture under microgravity-simulated conditions:evaluation of androgen-induced growth and PSA expression.In Vitro Cell Dev Biol Anim 1997;33:375e80.

[29] Zhau HE, He H, Wang CY, Zayzafoon M, Morrissey C,Vessella RL, et al. Human prostate cancer harbors the stemcell properties of bone marrow mesenchymal stem cells. ClinCancer Res 2011;17:2159e69.

[30] Pathak S, Nemeth MA, Multani AS, Thalmann GN, vonEschenbach AC, Chung LW. Can cancer cells transformnormal host cells into malignant cells? Br J Cancer 1997;76:1134e8.

[31] Shiao SL, Chu GC, Chung LW. Regulation of prostate cancerprogression by the tumor microenvironment. Cancer Lett2016;380:340e8.

[32] Beketic-Oreskovic L, Maric P, Ozretic P, Oreskovic D,Ajdukovic M, Levanat S. Assessing the clinical significance oftumor markers in common neoplasms. Front Biosci Elite Ed2012;4:2558e78.

[33] Laughney AM, Elizalde S, Genovese G, Bakhoum SF. Dynamicsof tumor heterogeneity derived from clonal karyotypic evo-lution. Cell Rep 2015;12:809e20.

[34] Masramon L, Vendrell E, Tarafa G, Capella G, Miro R, Ribas M,et al. Genetic instability and divergence of clonal populationsin colon cancer cells in vitro. J Cell Sci 2006;119:1477e82.

[35] Fidler IJ, Gersten DM, Hart IR. The biology of cancer invasionand metastasis. Adv Cancer Res 1978;28:149e250.

[36] Warner TF, Krueger RG. Circulating lymphocytes and thespread of myeloma. Review of the evidence. Lancet 1978;1:1174e6.

[37] Weiss L, Ward PM. Cell detachment and metastasis. CancerMetastasis Rev 1983;2:111e27.

[38] Feo-Zuppardi FJ, Taylor CW, Iwato K, Lopez MH, Grogan TM,Odeleye A, et al. Long-term engraftment of fresh humanmyeloma cells in SCID mice. Blood 1992;80:2843e50.

[39] He H, Yang X, Davidson AJ, Wu D, Marshall FF, Chung LW,et al. Progressive epithelial to mesenchymal transitions inARCaPE prostate cancer cells during xenograft tumor for-mation and metastasis. Prostate 2010;70:518e28.

[40] Yang X, Shao C, Wang R, Chu CY, Hu P, Master V, et al. Opticalimaging of kidney cancer with novel near infrared hepta-methine carbocyanine fluorescent dyes. J Urol 2013;189:702e10.

[41] Wang X, Qian X, Beitler JJ, Chen ZG, Khuri FR, Lewis MM,et al. Detection of circulating tumor cells in human periph-eral blood using surface-enhanced Raman scattering nano-particles. Cancer Res 2011;71:1526e32.

[42] Shao C, Liao CP, Hu P, Chu CY, Zhang L, Bui MH, et al.Detection of live circulating tumor cells by a class of near-infrared heptamethine carbocyanine dyes in patients with

Technical issues with ex vivo CTC culture 251

localized and metastatic prostate cancer. PLoS One 2014;9:e88967.

[43] Liu S, Tian Z, Zhang L, Hou S, Hu S, Wu J, et al. Combined cellsurface carbonic anhydrase 9 and CD147 antigens enablehigh-efficiency capture of circulating tumor cells in clear cellrenal cell carcinoma patients. Oncotarget August 1 2016.http://dx.doi.org/10.18632/oncotarget.10979 [Epub aheadof print].

[44] Ankeny JS, Court CM, Hou S, Li Q, Song M, Wu D, et al.Circulating tumour cells as a biomarker for diagnosis andstaging in pancreatic cancer. Br J Cancer 2016;114:1367e75.

[45] Chen JF, Ho H, Lichterman J, Lu YT, Zhang Y, Garcia MA,et al. Subclassification of prostate cancer circulating tumorcells by nuclear size reveals very small nuclear circulatingtumor cells in patients with visceral metastases. Cancer2015;121:3240e51.

[46] Ignatiadis M, Riethdorf S, Bidard FC, Vaucher I, Khazour M,Rothe F, et al. International study on inter-reader variabilityfor circulating tumor cells in breast cancer. Breast CancerRes 2014;16:R43.

[47] Court CM, Ankeny JS, Sho S, Hou S, Li Q, Hsieh C, et al. Re-ality of single circulating tumor cell sequencing for moleculardiagnostics in pancreatic cancer. J Mol Diagn 2016;18:688e96.

[48] Bobek V, Gurlich R, Eliasova P, Kolostova K. Circulating tumorcells in pancreatic cancer patients: enrichment and cultiva-tion. World J Gastroenterol 2014;20:17163e70.

[49] Bobek V, Kacprzak G, Rzechonek A, Kolostova K. Detectionand cultivation of circulating tumor cells in malignant pleuralmesothelioma. Anticancer Res 2014;34:2565e9.

[50] Bobek V, Matkowski R, Gurlich R, Grabowski K,Szelachowska J, Lischke R, et al. Cultivation of circulatingtumor cells in esophageal cancer. Folia Histochem Cytobiol2014;52:171e7.

[51] Cegan M, Kolostova K, Matkowski R, Broul M, Schraml J,Fiutowski M, et al. In vitro culturing of viable circulatingtumor cells of urinary bladder cancer. Int J Clin Exp Pathol2014;7:7164e71.

[52] Kolostova K, Broul M, Schraml J, Cegan M, Matkowski R,Fiutowski M, et al. Circulating tumor cells in localizedprostate cancer: isolation, cultivation in vitro and relation-ship to T-stage and Gleason score. Anticancer Res 2014;34:3641e6.

[53] Kolostova K, Cegan M, Bobek V. Circulating tumour cells inpatients with urothelial tumours: enrichment and in vitroculture. Can Urol Assoc J 2014;8:E715e20.

[54] Kolostova K, Matkowski R, Gurlich R, Grabowski K, Soter K,Lischke R, et al. Detection and cultivation of circulatingtumor cells in gastric cancer. Cytotechnology 2016;68:1095e102.

[55] Kolostova K, Spicka J, Matkowski R, Bobek V. Isolation, pri-mary culture, morphological and molecular characterizationof circulating tumor cells in gynecological cancers. Am JTransl Res 2015;7:1203e13.

[56] Kolostova K, Zhang Y, Hoffman RM, Bobek V. In vitro cultureand characterization of human lung cancer circulating tumorcells isolated by size exclusion from an orthotopic nude-mouse model expressing fluorescent protein. J Fluoresc2014;24:1531e6.

[57] Malara N, Trunzo V, Foresta U, Amodio N, De Vitis S,Roveda L, et al. Ex-vivo characterization of circulating coloncancer cells distinguished in stem and differentiated subsetprovides useful biomarker for personalized metastatic riskassessment. J Transl Med 2016;14:133.

[58] Sheng W, Ogunwobi OO, Chen T, Zhang J, George TJ, Liu C,et al. Capture, release and culture of circulating tumor cellsfrom pancreatic cancer patients using an enhanced mixingchip. Lab Chip 2014;14:89e98.

[59] Chen JY, Tsai WS, Shao HJ, Wu JC, Lai JM, Lu SH, et al.Sensitive and specific biomimetic lipid coated microfluidicsto isolate viable circulating tumor cells and microemboli forcancer detection. PLoS One 2016;11:e0149633.

[60] Zhang Z, Shiratsuchi H, Lin J, Chen G, Reddy RM, Azizi E,et al. Expansion of CTCs from early stage lung cancer pa-tients using a microfluidic co-culture model. Oncotarget2014;5:12383e97.

[61] Paris PL, Kobayashi Y, Zhao Q, Zeng W, Sridharan S, Fan T,et al. Functional phenotyping and genotyping of circulatingtumor cells from patients with castration resistant prostatecancer. Cancer Lett 2009;277:164e73.

[62] Yu M, Bardia A, Aceto N, Bersani F, Madden MW,Donaldson MC, et al. Cancer therapy. Ex vivo culture ofcirculating breast tumor cells for individualized testing ofdrug susceptibility. Science 2014;345:216e20.

[63] Cayrefourcq L, Mazard T, Joosse S, Solassol J, Ramos J,Assenat E, et al. Establishment and characterization of a cellline from human circulating colon cancer cells. Cancer Res2015;75:892e901.

[64] Gao D, Vela I, Sboner A, Iaquinta PJ, Karthaus WR,Gopalan A, et al. Organoid cultures derived from patientswith advanced prostate cancer. Cell 2014;159:176e87.

[65] Hamilton G, Burghuber O, Zeillinger R. Circulating tumorcells in small cell lung cancer: ex vivo expansion. Lung 2015;193:451e2.

[66] Court CM, Ankeny JS, Sho S, Tomlinson JS. Circulating tumorcells in gastrointestinal cancer: current practices and futuredirections. In: Bentrem D, Benson AB, editors. CancerTreatment and Research. Gastrointentional malignancies.Springer International Publishing; 2016. p. 345e76.

[67] den Toonder J. Circulating tumor cells: the grand challenge.Lab Chip 2011;11:375e7.

[68] van de Stolpe A, Pantel K, Sleijfer S, Terstappen LW, denToonder JM. Circulating tumor cell isolation and diagnostics:toward routine clinical use. Cancer Res 2011;71:5955e60.

[69] Kelbaek H. Sterile isolation of polymorphonuclear leukocytesfrom large blood volumes. J Clin Chem Clin Biochem 1985;23:17e20.

[70] Zipursky A, Bow E, Seshadri RS, Brown EJ. Leukocyte densityand volume in normal subjects and in patients with acutelymphoblastic leukemia. Blood 1976;48:361e71.

[71] Martowicz A, Seeber A, Untergasser G. The role of EpCAM inphysiology and pathology of the epithelium. Histol Histo-pathol 2016;31:349e55.

[72] Schnell U, Cirulli V, Giepmans BN. EpCAM: structure andfunction in health and disease. Biochim Biophys Acta 2013;1828:1989e2001.

[73] Fong D, Moser P, Kasal A, Seeber A, Gastl G, Martowicz A,et al. Loss of membranous expression of the intracellulardomain of EpCAM is a frequent event and predicts poor sur-vival in patients with pancreatic cancer. Histopathology2014;64:683e92.

[74] Gosens MJ, van Kempen LC, van de Velde CJ, van Krieken JH,Nagtegaal ID. Loss of membranous Ep-CAM in budding colo-rectal carcinoma cells. Mod Pathol 2007;20:221e32.

[75] Songun I, Litvinov SV, van de Velde CJ, Pals ST, Hermans J,van Krieken JH. Loss of Ep-CAM (CO17-1A) expression pre-dicts survival in patients with gastric cancer. Br J Cancer2005;92:1767e72.

[76] Beuran M, Negoi I, Paun S, Ion AD, Bleotu C, Negoi RI, et al.The epithelial to mesenchymal transition in pancreaticcancer: a systematic review. Pancreatology 2015;15:217e25.

[77] May CD, Sphyris N, Evans KW, Werden SJ, Guo W, Mani SA.Epithelial-mesenchymal transition and cancer stem cells: adangerously dynamic duo in breast cancer progression.Breast Cancer Res 2011;13:202.

252 R. Wang et al.

[78] May M, Brookman-May S, Burger M, Koch S, Otto W, Brundl J,et al. A switch from epithelial to mesenchymal propertiescorrelates with lymphovascular invasion in squamous cellcarcinoma of the penis. Pathol Res Pract 2015;211:641e5.

[79] Liao MY, Lai JK, Kuo MY, Lu RM, Lin CW, Cheng PC, et al. Ananti-EpCAM antibody EpAb2-6 for the treatment of coloncancer. Oncotarget 2015;6:24947e68.

[80] Ciraci E, Della Bella S, Salvucci O, Rofani C, Segarra M, BasonC,et al. Adult human circulating CD34eLineCD45eCD133e cellscan differentiate into hematopoietic and endothelial cells.Blood 2011;118:2105e15.

[81] Anversa P, Kajstura J, Leri A. Circulating progenitor cells:search for an identity. Circulation 2004;110:3158e60.

[82] Williams ES, Rodriguez-Bravo V, Chippada-Venkata U, De IaIglesia-Vicente J, Gong Y, Galsky M, et al. Generation ofprostate cancer patient derived xenograft models fromcirculating tumor cells. J Vis Exp 2015;105:53182.

[83] Preffer FI, Dombkowski D, Sykes M, Scadden D, Yang YG.Lineage-negative side-population (SP) cells with restrictedhematopoietic capacity circulate in normal human adultblood: immunophenotypic and functional characterization.Stem Cells 2002;20:417e27.

[84] Townsley MI. Structure and composition of pulmonary ar-teries, capillaries, and veins. Compr Physiol 2012;2:675e709.

[85] Hart IR. New evidence for tumour embolism as a mode ofmetastasis. J Pathol 2009;219:275e6.

[86] Tien YW,KuoHC,HoBI, ChangMC,ChangYT, ChengMF, et al. Ahigh circulating tumor cell count in portal vein predicts livermetastasis from periampullary or pancreatic cancer: a highportal venous CTC count predicts livermetastases. Med Baltim2016;95:e3407.

[87] Catenacci DV, Chapman CG, Xu P, Koons A, Konda VJ,SiddiquiUD,etal.Acquisitionofportal venouscirculating tumorcells from patients with pancreaticobiliary cancers by endo-scopic ultrasound. Gastroenterology 2015;149: 1794e803.e4.

[88] Marjanovic ND, Weinberg RA, Chaffer CL. Cell plasticity andheterogeneity in cancer. Clin Chem 2013;59:168e79.

[89] Ansieau S. EMT in breast cancer stem cell generation. CancerLett 2013;338:63e8.

[90] Hou Y, Wu K, Shi X, Li F, Song L, Wu H, et al. Comparison ofvariations detection between whole-genome amplificationmethods used in single-cell resequencing. Gigascience 2015;4:37.

[91] Liang J, Cai W, Sun Z. Single-cell sequencing technologies:current and future. J Genet Genomics 2014;41:513e28.

[92] de Lau WB, Snel B, Clevers HC. The R-spondin protein family.Genome Biol 2012;13:242.

[93] Mueller-Klieser W. Multicellular spheroids. A review oncellular aggregates in cancer research. J Cancer Res ClinOncol 1987;113:101e22.

[94] Sutherland RM, Durand RE. Growth and cellular characteris-tics of multicell spheroids. Recent Results Cancer Res 1984;95:24e49.

[95] Xu J, Wang R, Xie ZH, Odero-Marah V, Pathak S, Multani A,et al. Prostate cancer metastasis: role of the host microen-vironment in promoting epithelial to mesenchymal transitionand increased bone and adrenal gland metastasis. Prostate2006;66:1664e73.

[96] Yang X, Shi C, Tong R, Qian W, Zhau HE, Wang R, et al. NearIR heptamethine cyanine dye-mediated cancer imaging. ClinCancer Res 2010;16:2833e44.

[97] Berridge MJ. Lymphocyte activation in health and disease.Crit Rev Immunol 1997;17:155e78.

[98] Gupta S. Mechanisms of transmembrane signalling in humanT cell activation. Mol Cell Biochem 1989;91:45e50.

[99] Kilpatrick DC. Mechanisms and assessment of lectin-mediated mitogenesis. Mol Biotechnol 1999;11:55e65.

[100] Facchinetti A, PanozzoM, Pertile P, Tessarollo L, BiasiG. In vivoand in vitro death ofmature Tcells induced by separate signalsto CD4 and alpha beta TCR. Immunobiology 1992;185:380e9.

[101] Sharma K, Wang RX, Zhang LY, Yin DL, Luo XY, Solomon JC,et al. Death the Fas way: regulation and pathophysiology ofCD95 and its ligand. Pharmacol Ther 2000;88:333e47.

[102] Green DR, Scott DW. Activation-induced apoptosis in lym-phocytes. Curr Opin Immunol 1994;6:476e87.

[103] Wang RX, Chu CY, Nissen NN, Lewis MS, Palanisamy N,Edderkaoui M, et al. Novel patient-derived CTC-xenograftmodels for the study of pancreati cancer biology, metastasisand therapy. In: American Pancreatic Association 2015 AnnualMeeting, LoewsCoronadoBay, SanDiego, November 4e7; 2015.

[104] Wang RX, Posadas EM, Hu P, Chu CY, Sievert M, Bhowmick I,et al. Expansion of circulating tumor cells from prostateancer patients by culturing circulating peripheral bloodmononuclear cells. In: 2012 Society for Basic UrologicalResearch Fall Symposium, Trump International Beach Resort,Miami Beach, FL, November 15e18; 2012.

Technical issues with ex vivo CTC culture 253