Circulating Tumor Cells and EpCAM Expression in ... · should not express EpCAM (12, 13). The aims...

Imaging, Diagnosis, Prognosis

Circulating Tumor Cells and EpCAM Expression in Neuroendocrine Tumors

Mohid S. Khan1,2, Theodora Tsigani2, Mohammed Rashid2, Jeremy S. Rabouhans3, Dominic Yu3,Tu Vinh Luong4, Martyn Caplin1, and Tim Meyer2,5

AbstractPurpose: Neuroendocrine tumors (NET) are heterogeneous tumors with widely variable survival. It is

unknown whether they express EpCAM (epithelial cell adhesion molecule) and thus whether NET

circulating tumor cells (CTC) are detectable. We systematically investigated EpCAM expression and

CTC detection in patients with metastatic NETs and evaluated the potential of CTCs to predict radiological

progression.

Experimental Design: EpCAM protein expression was evaluated in 74 samples of formalin-fixed,

paraffin-embedded NET tissue by immunohistochemistry. Seventy-nine patients with metastatic NETs (42

midgut, 5 unknown primary, 19 pancreatic, 13 bronchopulmonary) had blood samples drawn for CTC

isolation and enumeration utilizing the CellSearch platform. Patients were classified as having progressive

or nonprogressive disease on the basis of serial imaging.

Results: Strong homogeneous, membranous EpCAM expression was observed in all ileal (n ¼ 26) and

pancreatic NETs (n ¼ 16), whereas variable EpCAM expression was observed in bronchopulmonary NETs

(n ¼ 13). Forty-three percent of midgut and 21% of pancreatic NETs had CTCs detected with a range of 0–

62 and 0–11, respectively. The absence of CTCs was strongly associated with stable disease (P < 0.001).

There was a moderate correlation between CTC levels and urinary 5-hydroxyindole acetic acid (r¼ 0.5, P¼0.007) and between CTC levels and burden of liver metastases (B ¼ 8.91, P < 0.001). There was no or low

correlation between CTC levels and Ki-67 (r¼ 0.08, P ¼ 0.59) and serum chromogranin A (r ¼ 0.246, P ¼0.03).

Conclusions: This is the first systematic analysis showing EpCAM expression and CTC detection in

NETs. CTCs seem to be associated with progressive disease andmay provide useful prognostic information

given the variable survival rates in these tumors. Clin Cancer Res; 17(2); 337–45. �2011 AACR.

Introduction

Neuroendocrine tumors (NET) are a rare, heterogeneousgroup of tumors arising most commonly from the gastro-intestinal tract, pancreas, and lungs. Their natural history isvariable, with median survival ranging from approximately6 months in aggressive high-grade tumors to up to 20 yearsin more indolent disease (1). Current methods of mon-itoring progression in NETs and response to treatmentinclude serial imaging over many years. Although there

are general NET biomarkers such as plasma chromograninA (CgA) and urinary 5-hydroxyindole acetic acid (5-HIAA),there is a lack of prospectively validated prognostic andpredictive biomarkers.

Recent technological advances have enabled circulatingtumor cell (CTC) enumeration and characterization bydifferent methods. The CellSearch platform, an auto-mated immunomagnetic enrichment and staining sys-tem, has been utilized to detect CTCs with highsensitivity, specificity, and reproducibility (2). Large stu-dies have reported that the number of CTCs in patientswith metastatic breast cancer before commencing a newtherapy to be an independent predictor of progression-free and overall survival (3, 4), with similar resultsreported in metastatic colorectal and prostate cancer (5,6). Therefore, it has been suggested that CTC detectioncan aid appropriate patient stratification and design oftailored treatments, which are the aims of ongoing studies(such as SWOG S0500).

The CellSearch platform requires EpCAM (epithelial celladhesionmolecule) expression to isolate CTCs. EpCAM is a39- to 42-kDa transmembrane epithelial glycoprotein (7),overexpressed in human adenocarcinomas (8). Its exactfunction is yet to be fully elucidated, but its expression

Authors' Affiliations: 1Neuroendocrine Tumour Unit, Centre for Gastro-enterology, Royal Free Hospital, 2UCL Cancer Institute, University CollegeLondon, and Departments of 3Radiology, 4Pathology, and 5Oncology,Royal Free Hospital, London, United Kingdom

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

Part of data presented in abstract: European Neuroendocrine TumourSociety 2010.

Corresponding Author: Tim Meyer, UCL Cancer Institute, UniversityCollege London, 72 Huntley Street, London WC1E 6BT, United Kingdom.Phone: 0207-679-6731; Fax: 0203-108-2025; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-10-1776

�2011 American Association for Cancer Research.

ClinicalCancer

Research

www.aacrjournals.org 337

Cancer Research. on November 20, 2020. © 2011 American Association forclincancerres.aacrjournals.org Downloaded from

Published OnlineFirst January 11, 2011; DOI: 10.1158/1078-0432.CCR-10-1776

http://clincancerres.aacrjournals.org/

enables the CellSearch platform to enrich CTCs via immu-nomagnetic separation with iron particles coupled toEpCAM antibodies. Although series have reported a smallneuroendocrine subset of lung cancers (8) and insulino-mas (9) to overexpress this epithelial marker, the systematicanalysis of EpCAM expression in NETs has not been under-taken to our knowledge. Originally thought to be derivedfrom cells of the neural crest sharing secretory and histo-logic properties with neural cells, it is debated whetherNETs are epithelial in origin (10, 11) and thus it has beenassumed that CTCs cannot be isolated in NETs as theyshould not express EpCAM (12, 13).

The aims of this study were to systematically investigateEpCAM expression in NETs, to identify and quantify CTCsin patients with NETs, and to evaluate their potential topredict radiological progression.

Materials and Methods

ImmunohistochemistryThe study included 74 patients with a histopathologic

diagnosis of NET in which blocks of formalin-fixed andparaffin-embedded tissue were available. The tumors wereclassified according to the site of origin and criteria of theWorld Health Organization and graded according to theEuropean Neuroendocrine Tumour Society (ENETS) pro-posal for grading and staging of NETs (14, 15).

Briefly, 3-mm sections of tumor tissue were deparaffi-nized in xylene and rehydrated in graded alcohols. Endo-genous peroxidase was blocked with 0.5% H2O2 inmethanol for 10 minutes. Thereafter, sections were sub-jected to antigen retrieval for 10 minutes in 0.1% trypsin at37�C. Immunohistochemical staining was performed withthe NovoLink polymer detection system (Novocastra).Sections were incubated with mouse anti-human EpCAM

monoclonal antibody (ESA, clone VU-1D9; Novocastra) ata dilution of 1:50 for 1 hour at room temperature, post–primary block for 30 minutes, followed by NovoLinkpolymer for 30 minutes. Reaction products were visualizedwith the application of diaminobenzidine substrate chro-mogen solution. Slides were counterstained in hematox-ylin and mounted.

Normal small and large bowel epithelial cells, pancreaticacini, and islet cells and bile duct epithelial cells served asinternal positive controls. Stromal cells were used as inter-nal negative controls. All cases were evaluated indepen-dently by 2 examiners, including 1 pathologist, withoutknowledge of pathologic data. Any discordant results werereviewed together to reach agreement. Scoring was basedon intensity of staining: 0, negative; 1, weakly positive; 2,moderate; and 3, strongly positive. Extent of tumor stainingwas scored, in which 10 random high-power fields wereassessed and the average percentage of positive stainingcells was estimated (1, <25%; 2, 25%–75%; and 3, >75%).The product of staining intensity and extent was used as theoverall score giving final values of 0, 1, 2, 3, 4, 6, and 9.Scores of 0 were classified as negative, 1–2 as weak, 3–4 asmoderate, and 6–9 as strong staining.

Cell culture and immunofluorescenceThe bronchial NET cell line, NCI-H727, was main-

tained in RPMI 1640 (PAA Laboratories) supplementedwith 10% FBS, 2 mmol/L of L-glutamine, penicillin, andstreptomycin. A breast cancer cell line (as a negativecontrol for CD56 expression), MCF-7, was maintainedin MEM (PAA Laboratories) supplemented with 10% FBS,2 mmol/L of L-glutamine, penicillin, and streptomycin at37�C. Sterile 12-mm diameter coverslips were placed inwells of a 24-well plate (BD Falcon). Approximately 2 �105 cells were seeded into each well and incubated at37�C for 24 hours at 5% CO2. After media was removed,cells were washed, and then fixed in 1mL of 1% formal-dehyde for 10 minutes. After fixative was removed, 0.1mol/L of glycine was added to each well to neutralize for 5minutes. To permeabilize the cells, 0.1% Triton X-100was added to each well for 10 minutes before furtherwashing. To prevent nonspecific binding, cells wereblocked with 0.1% BSA for 30 minutes. Cells were incu-bated with the FITC-conjugated mouse anti-human anti-body synaptophysin (Acris Antibodies) at a dilution of1:20 or the Alexa Fluor 488–conjugated mouse anti-human antibody, CD56 (BD PharMingen) at a dilutionof 1:20 for 1 hour at room temperature. After furtherwashing, coverslips were removed from each well andmounted onto slides with Vectashield (Vector Labora-tories Ltd.) and sealed with clear nail varnish beforevisualization.

PatientsThis study was approved by the Local Ethics Committee

and all patients provided written informed consent. Eligi-ble patients had pathologically confirmed NETs categor-ized according to their primary site of origin: midgut,

Translational Relevance

For the first time, we have systematically demon-strated EpCAM (epithelial cell adhesion molecule)expression in the majority of neuroendocrine tumors(NET), suggesting an epithelial origin to these raretumors that were once thought to be derived from cellsof the neural crest. EpCAM expression has allowed us todetect circulating tumor cells (CTC) in patients withNETs, and we show that rates of detection are particu-larly high in midgut tumors. We also report the associa-tion between absence of CTCs and radiologically stabledisease, whereas the presence of CTCs seems to beassociated with progressive disease. CTC enumerationmay therefore be a useful means of stratifying patientsfor therapeutic intervention or clinical trials that isincreasingly important as the number of active agentsfor treating NETs increases. Furthermore, the detectionof CTCs in NETs opens the way for predictive andprognostic studies as currently applied to metastaticbreast and prostate cancer.

Khan et al.

Clin Cancer Res; 17(2) January 15, 2011 Clinical Cancer Research338




pancreas, bronchopulmonary, or of unknown primary;withmetastatic diseasemeasurable by Response EvaluationCriteria in Solid Tumors (RECIST). Although some patientshad undergone resection of the primary, all patients hadmetastatic disease. Patients that had undergone chemother-apy, biological therapy, interferon, receptor-targeted radi-olabeled therapy, or embolization within the previous2 months were excluded. To assess whether a patienthad progressive or nonprogressive disease, imaging withcomputed tomography (CT) or magnetic resonance ima-ging (MRI) within 6 weeks of sampling, was compared withthe previous imaging. Comparisons were performed by anindependent radiologist who recorded absolute changes intumor dimension and classified them according to RECIST.Plasma CgA and 24-hour urinary 5-HIAA, where applic-able, were evaluated. Radiological burden was assessed byquantification of hepatic tumor load from 4 to 6 slices of aCT/MRI scan with the most amount of disease by a semi-quantitative approach. Hepatic tumor burden was categor-ized as 25% or less, more than 25% but 50% or less, morethan 50% but 75% or less, or more than 75%. Data werecollected on primary site, duration of diagnoses, any pre-vious treatment received, and whether the primary tumorhad been resected.

CTC enumerationBlood samples (7.5 mL) were drawn into CellSave tubes

(Veridex LLC) and maintained at room temperature. TheCellSearch (Veridex LLC) platform was used for isolationand enumeration of CTCs as previously described (3). Inbrief, the platform consists of a semiautomated system thatenriches the sample for cells expressing EpCAM with anti-body-coated ferroparticles. The identification of CTCsby immunofluorescence staining was performed withCellSearch Analyzer II (Veridex LLC), a semiautomatedfluorescence-based microscopy system. The fluorescentlylabeled monoclonal antibodies, anti-CD45–allophycocya-nin (APC) and anti–pan-cytokeratin (CK)-phycoerythrin(PE) are used to distinguish epithelial cells from leukocytes.CTCs were defined as 4,2-diamidino-2-phenylindole-dihy-drochloride (DAPI)-stained cells lacking CD45 and expres-sing cytokeratin. NETs have been found to express lowmolecular weight cytokeratins in previous studies (16–18). All evaluations were performed by 2 independentoperators without knowledge of clinical status. Technicaldetails of the CellSearch platform including accuracy, pre-cision, linearity, and reproducibility have been describedelsewhere (2).Analysis of synaptophysin and CD56 expression on

CTCs was performed on the platform by using FITC-con-jugated mouse anti-human synaptophysin antibody (AcrisAntibodies) and Alexa Fluor 488–conjugated mouse anti-human CD56 antibody (BD PharMingen) as additionalmarkers, respectively. A selected patient’s sample was pro-cessed alongside samples of healthy control blood spikedwith 103 NCI-H727 and 103 MCF-7 cells as well asunspiked healthy blood. The antibodies were substitutedfor PBS for antibody omission controls. For the CD56 run,

NCI-H727 was used as a positive control and MCF-7 as anegative control. For the synaptophysin run, NCI-H727was used as a positive control and leucocytes as an internalnegative control. The antibody concentrations were 100mg/mL for the synaptophysin antibody and 12 mg/mL forthe CD56 antibody. The positivity of synaptophysin andCD56 of neuroendocrine CTCs was evaluated by 2 inde-pendent operators by using the research mode of Cell-Search Analyzer II with an integration time of 0.8 seconds.

Statistical analysisStudy analysis was performed using SPSS for Windows

(SPSS Inc.) and GraphPad Prism (GraphPad Software),with values of P < 0.05 were considered significant. Asso-ciation between the presence of CTCs and radiologicalstatus was assessed with Fisher’s exact test. Differences inbaseline characteristics between progressors and nonpro-gressors were analyzed with Fisher’s exact, Mann–Whitney,and t tests. Assuming the presence of CTCs in 70% ofprogressors and 30% of nonprogressors with unequalgroups (ratio ¼ 0.4), a sample size of 40 and 16 wasrequired, respectively, with 80% power and significancelevel of 0.05. Correlation between CTCs and CgA, Ki-67,24-hour urinary 5-HIAA, or resection of primary wasassessed using Spearman’s rank or Mann–Whitney tests.Correlation between burden and CTCs was assessed usinglinear regression.

Results

Immunohistochemical analysis of EpCAM expressionExpression of EpCAM in normal tissue. EpCAM was

expressedmostly on the basal or basolateral cell membraneof small and large intestinal, appendiceal, and bile ductepithelia. Variable EpCAM expression, including somecytoplasmic staining, was seen in pancreatic islets, pan-creatic acini, and gastric oxyntic glands. EpCAM was nega-tive in hepatocytes, mesothelium, and gastric foveolarepithelium. Variable staining, including some cytoplasmicstaining, was seen in pancreatic islets. These results areconsistent with previous published work (8).

Expression of EpCAM in NET. All ileal (n ¼ 26),pancreatic (n ¼ 16), unknown primary (n ¼ 2), andgastric (n ¼ 4) NETs showed strong (score 6–9) homo-geneous membranous staining for EpCAM (Fig. 1). Mod-erate to strong staining was seen in appendiceal (n ¼ 7)NETs. EpCAM expression was not affected by grade.Bronchopulmonary NETs (n ¼ 13) showed variableEpCAM expression from negative, weak to strong staining.EpCAM distribution was also variable in bronchopul-monary NETs: EpCAM was observed in cell membranesand in cytoplasm. EpCAM was not expressed in paragan-glioma (n ¼ 1). Details can be found in SupplementaryTable S1.

Patients for CTC evaluationEighty patients with metastatic NETs were recruited for

CTC evaluation. One sample was excluded because of

CTCs in NETs

www.aacrjournals.org Clin Cancer Res; 17(2) January 15, 2011 339




postsampling hemolysis. Clinical characteristics are shownin Table 1.

CTCs in NETsOf patients with metastatic midgut tumors, 43% had

CTCs detected with a mean � SEM of 8.9 � 2.8 CTCs per7.5 mL of blood (range¼ 0–62; Fig. 2). Fewer patients withpancreatic (21%) and bronchopulmonary NETs (31%)

had detectable CTCs, although the 2 patients with thegreatest number of CTCs both had bronchopulmonaryNETs (Fig. 2). In addition, 4 of 5 patients with NETs ofunknown primary had CTCs present with amean� SEM of4.2� 1.2 (range¼ 0–7). There was no significant differencein the presence (P ¼ 0.61) or amount (P ¼ 0.21) of CTCsbetween those on somatostatin analogues and those whowere not.

A

B

C

D

E

F

G

H

Figure 1. Immunohistochemistrywith EpCAM. A, gastric mucosawith absent expression, apartfrom, in oxyntic cells. B, ileal NETresection with intensemembranous staining of NET andof normal mucosa (short arrow)with negative stroma (long arrow).C, EpCAM positivity in ileal NET athigh power. D, EpCAM-positiveappendiceal NET (short arrow)with positive normal appendixmucosa (long arrow). E, normalpancreas with positive acini, islet,and duct. F, pancreatic NET withmembranous staining at highpower. G, poorly differentiatedgastric NETwithmembranous andcytoplasmic staining. H,bronchopulmonary NET withcytoplasmic staining (short arrow)and negative lung parenchyma/alveoli (long arrow).

Khan et al.





To confirm the neuroendocrine origin of CTCs, theexpression of synaptophysin and CD56 was evaluatedon CTCs from a patient with metastatic midgut NET andon cell line controls. By immunofluorescence, synaptophy-sin was expressed on NCI-H727 and MCF-7, and CD56 onNCI-H727 but not on MCF-7. Spiked NCI-H727 processedon the CellSearch platform was positive for synaptophysinand CD56; MCF-7 was positive for synaptophysin butnegative for CD56. Eighty-two percent of the patient’sCTCs were positive for synaptophysin and 21% forCD56 (Fig. 3).Twenty-six of the 79 patients who had blood samples

enumerated for CTCs had archival histopathologic tissueavailable for assessment of EpCAM expression (7 midgut, 2liver metastases with unknown primary, 8 pancreatic, 6bronchopulmonary resections, 3 bronchial biopsies). All 7midgut NETs and 2 liver metastases from an unknownprimary had strong staining for EpCAM. Five and 2 of thesehad CTCs present, respectively. All 8 pancreatic NETs (4pancreatic resections, 4 biopsies) had strong EpCAM stain-ing but only 1 had CTCs present. Four of the broncho-pulmonary NET patients had moderate or strong EpCAMexpression; 3 of these had CTCs present. Five bronchopul-monary NETs had weak or no EpCAM staining; none hadCTCs. CTCs were detected only in those patients withbronchopulmonary NETs that had moderate to strongexpression of EpCAM.

CTC correlation with existing markersIn midgut and unknown NETs, there was a moderate

correlation between CTC levels and urinary 5-HIAA, whereavailable (r¼ 0.5, P¼ 0.007, n¼ 28). For all NETs, becauseof small category size, the liver burden categories 50%–75% and 75% were combined. A significant associationbetween CTC levels and burden of liver metastases wasfound (B ¼ 8.91, 95% CI ¼ 4.3–13.5, P < 0.001; Supple-mentary Fig. S1).There was no or low correlation betweenCTC levels and Ki-67 (r ¼ 0.08, P ¼ 0.59; SupplementaryFig. S2), CgA (r ¼ 0.246, P ¼ 0.03; Supplementary Fig. S3),or resection of primary (U ¼ 698.5, P ¼ 0.68).

CTCs and progressive diseaseOf the 66 patients with metastatic disease from midgut,

pancreatic, or unknown primary who had CTCs evaluated,63 had imaging within 6 weeks and a previous scan forcomparison. The 3 remaining patients had been diagnosedrecently and only imaged once. Bronchopulmonary NETswere excluded from this analysis because of variableEpCAM expression and thus CTC detection. Eighteen of19 (95%) patients who had progressive disease accordingto RECIST had CTCs detected compared with 9 of 44 (20%)patients who had nonprogressive disease (SupplementaryFig. S4). This was a significant difference (c2 ¼ 31.4, P <0.001), with no statistical difference in other factorsbetween progressors and nonprogressors (Table 2). When

Table 1. Clinical characteristics of NET patient sample

Pancreatic(n ¼ 19)

Midgut(n ¼ 42)

Bronchopulmonary(n ¼ 13)

Unknown primary(n ¼ 5)

Age, median (range), y 60 (38–87) 61.5 (36–79) 57 (30–80) 62 (30–64)Sex, n (%)

Male 10 20 8 0Female 9 22 5 5

GradeLow 8 24 6 0Intermediate 7 16 5 5High 4 2 2 0

Burden of liver metastases, %�25 10 20 10 125 � 50 7 15 2 450 � 75 2 4 1 0>75 0 1 0 0Unknown – 2 – –

Ongoing somatostatin analogue therapyYes 8 31 0 1No 11 11 13 4

Duration of diagnosis, median (range), mo 58.5 (2–166) 60 (2–278) 22 (8–287) 16 (9–63)Previous treatments

Resection of primary 4/19 23/42 6/13 0Chemotherapy 6 2 2 2Embolization 1 5 0 1Radionuclides 1 10 1 1Liver resection 2 4 1 0

CTCs in NETs





absolute changes in target lesions were assessed, 23 of 27(85%) patients with detectable CTCs had growth of tumorlesions, and 31 of 36 (86%) patients without CTCs had nogrowth or spontaneous shrinkage (Fig. 4). There was noassociation between the number of CTCs and absoluteincrease in tumor size.

Discussion

To our knowledge, this is the first systematic analysis ofEpCAM expression and CTCs in NETs.

All gastroenteropancreatic NET tissue showed strongmembranous EpCAM expression regardless of grade. Theorigin of NETs is still debated, but the expression of thiscarcinoma-associated antigen adds evidence to an epithe-lial origin rather than being derived from the neural crest aswas once originally thought (10, 19). EpCAM upregulatesc-myc and cyclins, promoting cell cycling and enhancingproliferation (20–24). Anti-EpCAM therapy has beentrialed in metastatic breast cancer, colorectal cancer, andin malignant ascites (25–30) and EpCAM expression inNETs presents an opportunity for EpCAM directed therapy(9). However, caution should be taken as normal pancrea-tic and intestinal tissue express EpCAM. The CellSearchmethod of CTC enrichment has been validated previouslyand only 0.3% of healthy controls and benign cases have 2

Figure 2. Levels of CTCs in NETs.

Fig. 3. Synaptophysin and CD56expression of CTC patients.

Khan et al.





or more CTCs per 7.5 mL of blood (2), although the rate ofCTCs in healthy controls depends on criteria used foranalysis of the images provided by the CellSearch system(31). We have shown that CTCs in our patient group areneuroendocrine in origin by using the NET unique immu-nohistochemical profile. In those with midgut andunknown primary NETs, 47% had CTCs detectable, 68%

of whom had greater than 5 CTCs/7.5 mL. These levels arecomparable with other tumor types, although not as highas in metastatic breast and prostate cancers, perhaps reflect-ing the indolent nature of most NETs (2). Although prog-nostic cutoff levels of CTCs have been determined in othercancers (3, 5, 6), further studies are required to determinethese in NETs.

Table 2. Characteristics of cases with nonprogressive versus progressive disease in patients withmetastatic midgut, pancreatic, or unknown primary

Characteristic Nonprogressivedisease (n ¼ 44)

Progressivedisease (n ¼ 19)

Test/P value

Age, median (range), y 59.5 (38–79) 56.1 (30–77) U ¼ 0.36/P ¼ 0.37Grade Fisher's exact/P ¼ 0.57

Low 22 8Intermediate 19 9High 3 2

Burden of liver metastases, % Fisher's exact/P ¼ 0.39�25 24 725 � 50 16 10>50 4 2

CgA, mean (range), pmol/L 297.8 (33–1,000) 359.4 (34–1,000) t ¼ 0.46/P ¼ 0.65Resection of primary Fisher's exact/P ¼ 0.99

Yes 16 8No 28 11

Duration of diagnosis, median (range), mo 58.5 (5–278) 55 (10–108) U ¼ 343/P ¼ 0.99Ongoing somatostatin analogue therapy Fisher's exact/P ¼ 0.77

Yes 22 11No 12 8

Interval to last scan, median (range), mo 18 (12–56) 18 (10–30) U ¼ 347/P ¼ 0.76CTC Fisher's exact/P < 0.001

0 35 1�1 9 18

Figure 4. Percentage increase inlesions on imaging grouped by thepresence of CTCs. Each barrepresents an individual case.Cases without CTCs are groupedon the left-hand side of the chart(CTC ¼ 0); cases with CTCsdetected on the right-hand side(CTC � 1).

CTCs in NETs





Despite widespread EpCAM positivity in pancreaticNETs, only a small proportion had CTCs detected com-pared with midgut NETs, consistent with findings in pan-creatic adenocarcinomas (2). Explanations include loss ofEpCAM expression, slow shedding of CTCs from pancreaticNETs, or unidentified factors particular to this pancreaticsample. Absent EpCAM expression has been cited as areason for the lack of CTCs in a substantial number ofcases of metastatic breast cancer (32), but until now, therehas been no study dealing with synchronous CTC detectionand EpCAM expression in patients.

Lack of EpCAM expression may be relevant in bronch-opulmonary NETs, as CTCs can only be detected on thisplatform when the primary tumor expresses EpCAM. Thosebronchopulmonary NETs that were EpCAM positive hadthe highest CTC counts perhaps due to the absence ofportal filtration that has recently been shown to decreaseCTCs (33). Our sample of bronchopulmonary NETs wassmall, with only 13 cases, and the significance of EpCAMexpression in bronchopulmonary NETs requires furtherinvestigation.

The absence of CTCs in NETs is associated with stabledisease as defined by RECIST, whereas the presence of CTCsseems to be associated with progressive disease. Smallerincreases in tumor size not attaining RECIST also seems tobe associated with CTCs. Radiological methods of mon-itoring for progression in NETs and response to treatmentmay be confounded by interobserver variability and thefibrotic reaction often seen in NETs (34, 35). Given thevaried survival with NETs compared with other tumors, thisdependency on serial imaging is costly and exposes patientsto radiation. In NETs, CTCs may be of prognostic value indiscriminating progressing from stable tumors, which mayassist stratification for aggressive therapy at time of diag-nosis. Given the delayed response seen in NETs withchemotherapy and radionuclides (36, 37), CTCs may offerpredictive information early during therapy and may beuseful in monitoring response to therapy without repeatedexposure to radiation. We recognize limitations in thisanalysis whereby tumors were classified as progressivemaking retrospective comparisons with previous scans inheterogeneous groups. However, the median intervalbetween scans was similar between patients who hadtumor progression and those with no tumor progression.In addition, the sample size was sufficient to detect thisassociation between CTC presence and progressive disease.We also note the similar histologic grades between patients

with and without tumor progression, which may partlyreflect small sample size with few cases of high-gradeNETs. The number of CTCs did not correlate with theexisting prognostic markers Ki-67 proliferation index orplasma CgA. Ki-67 is evaluated in tumor specimensobtained at diagnosis, which may occur several yearspreviously and may not reflect dynamic changes duringthe course of the disease. The association between CTCsand 24-hour urinary 5-HIAA in midgut NETs could beexplained by tumors being more metabolically activewhen CTCs are present or by CTCs secreting metabolicallyactive compounds into the circulation. The associationbetween CTCs and burden conflicts with some studies inother cancers in which such an association was not found(38). The confirmation of CTCs in NETs presents anexciting opportunity to perform molecular characteriza-tion of NETs without invasive biopsies, and this mayaccelerate development of new therapies. Given the vari-able survival of NETs and increasing armory of treatmentoptions (36, 37, 39–42), CTCs could be utilized as aprognostic marker in NETs to stratify therapy, and in real-time monitoring of tumor growth or treatment response.Because of the heterogeneity of our patient populationwith respect to therapy, we were not able to distinguishbetween the role of CTCs as a prognostic or predictivemarker and large collaborative prospective studies areneeded to address this important question.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Acknowledgments

We thank the National Institute of Health Research (NIHR), Experi-mental Cancer Medicine Centre Network (ECMC), and "Bottoms Up"charity for their financial support.

Grant Support

The study was funded by Cancer ResearchUK, Bottoms UpCharity, QuietCancer Appeal. T. Meyer is partly funded by UCLH/UCL Department ofHealth’s NIHR Biomedical Research Centres funding scheme.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received July 5, 2010; revised September 14, 2010; accepted September29, 2010; published OnlineFirst January 11, 2011.

References1. Yao JC, Hassan M, Phan A, Dagohoy C, Leary C, Mares JE, et al.

One hundred years after "carcinoid": epidemiology of and prog-nostic factors for neuroendocrine tumors in 35,825 cases in theUnited States. J Clin Oncol 2008;26: 3063–72.

2. Allard WJ, Matera J, Miller MC, Repollet M, Connelly MC, Rao C, et al.Tumor cells circulate in the peripheral blood of all major carcinomasbut not in healthy subjects or patients with nonmalignant diseases.Clin Cancer Res 2004;10:6897–904.

3. Cristofanilli M, Hayes DF, Budd GT, Ellis MJ, Stopeck A, Reuben JM,et al. Circulating tumor cells: a novel prognostic factor for newlydiagnosed metastatic breast cancer. J Clin Oncol 2005;23:1420–30.

4. Cristofanilli M, Budd GT, Ellis MJ, Stopeck A, Matera J, Miller MC,et al. Circulating tumor cells, disease progression, and survival inmetastatic breast cancer. N Engl J Med 2004;351:781–91.

5. Cohen SJ, Punt CJ, Iannotti N, Saidman BH, Sabbath KD, Gabrail NY,et al. Relationship of circulating tumor cells to tumor response,

Khan et al.





progression-free survival, and overall survival in patients with meta-static colorectal cancer.[Erratum in J Clin Oncol. 2009;27:1923]. J ClinOncol 2008;26:3213–21.

6. de Bono JS, Scher HI, Montgomery RB, Parker C, Miller MC, TissingH, et al. Circulating tumor cells predict survival benefit from treatmentin metastatic castration-resistant prostate cancer.[Erratum in ClinCancer Res 2009;15:1506]. Clin Cancer Res 2008;14:6302–9.

7. Litvinov SV, Velders MP, Bakker HA, Fleuren GJ, Warnaar SO. Ep-CAM: a human epithelial antigen is a homophilic cell-cell adhesionmolecule. J Cell Biol 1994;125:437–46.

8. Went PT, Lugli A, Meier S, Bundi M, Mirlacher M, Sauter G, et al.Frequent EpCam protein expression in human carcinomas. HumPathol 2004;35:122–8.

9. Raffel A, Eisenberger CF, Cupisti K, Schott M, Baldus SE, Hoffmann I,et al. Increased EpCAM expression in malignant insulinoma: potentialclinical implications. Eur J Endocrinol 2010;162:391–8.

10. Pictet RL, Rall LB, Phelps P, Rutter WJ. The neural crest and the originof the insulin-producing and other gastrointestinal hormone-produ-cing cells. Science 1976;191:191–2.

11. Rindi G, Leiter AB, Kopin AS, Bordi C, Solcia E. The "normal" endo-crine cell of the gut: changing concepts and new evidences. Ann N YAcad Sci 2004;1014:1–12.

12. Bevilacqua S, Gallo M, Franco R, Rossi A, De LA, Rocco G, et al. A"live" biopsy in a small-cell lung cancer patient by detection ofcirculating tumor cells. Lung Cancer 2009;65:123–5.

13. Jiao LR, Apostolopoulos C, Jacob J, Szydlo R, Johnson N, Tsim N,et al. Unique localization of circulating tumor cells in patients withhepatic metastases. J Clin Oncol 2009;27:6160–5.

14. Rindi G, Kloppel G, Alhman H, Caplin M, Couvelard A, de Herder WW,et al. TNM staging of foregut (neuro)endocrine tumors: a consensusproposal including a grading system. Virchows Arch 2006;449:395–401.

15. Rindi G, Kloppel G, Couvelard A, Komminoth P, Korner M, Lopes JM,et al. TNM staging of midgut and hindgut (neuro)endocrine tumors: aconsensus proposal including a grading system. Virchows Arch2007;451:757–62.

16. Raju GC. Occurrence and expression of cytokeratins in carcinoidtumours of the gastrointestinal tract and their probable precursorcells. Ann Acad Med Singapore 1989;18:298–301.

17. Valli M, Fabris GA, Dewar A, Hornall D, Sheppard MN. Atypicalcarcinoid tumour of the lung: a study of 33 cases with prognosticfeatures. Histopathology 1994;24:363–9.

18. Wilander E, Scheibenpflug L. Cytokeratin expression in small intest-inal and appendiceal carcinoids. A basis for classification. Acta Oncol1993;32:131–4.

19. Rindi G, Leiter AB, Kopin AS, Bordi C, Solcia E. The "normal" endo-crine cell of the gut: changing concepts and new evidences. Ann N YAcad Sci 2004;1014:1–12.

20. Trzpis M, McLaughlin PM, de Leij LM, Harmsen MC. Epithelial celladhesion molecule: more than a carcinoma marker and adhesionmolecule. Am J Pathol 2007;171:386–95.

21. Nubel T, Preobraschenski J, Tuncay H, Weiss T, Kuhn S, Ladwein M,et al. Claudin-7 regulates EpCAM-mediated functions in tumor pro-gression. Mol Cancer Res 2009;7:285–99.

22. Sankpal NV, Willman MW, Fleming TP, Mayfield JD, Gillanders WE.Transcriptional repression of epithelial cell adhesion molecule con-tributes to p53 control of breast cancer invasion. Cancer Res2009;69:753–7.

23. Maaser K, Borlak J. A genome-wide expression analysis identifies anetwork of EpCAM-induced cell cycle regulators. Br J Cancer2008;99:1635–43.

24. Maetzel D, Denzel S, Mack B, Canis M, Went P, Benk M, et al. Nuclearsignalling by tumour-associated antigen EpCAM. Nat Cell Biol2009;11:162–71.

25. Heiss MM, Murawa P, Koralewski P, Kutarska E, Kolesnik OO, Ivan-chenko VV, et al. The trifunctional antibody catumaxomab for the treat-

ment of malignant ascites due to epithelial cancer: results of a prospec-tive randomized phase II/III trial. Int J Cancer 2010;127:2209–21.

26. Kurtz JE, Dufour P. Adecatumumab: an anti-EpCAM monoclonalantibody, from the bench to the bedside. Expert Opin Biol Ther2010;10:951–8.

27. Schmidt M, Scheulen ME, Dittrich C, Obrist P, Marschner N, Dirix L,et al. An open-label, randomized phase II study of adecatumumab, afully human anti-EpCAM antibody, as monotherapy in patients withmetastatic breast cancer. Ann Oncol 2010;21:275–82.

28. Sebastian M, Kiewe P, Schuette W, Brust D, Peschel C, Schneller F,et al. Treatment of malignant pleural effusion with the trifunctionalantibody catumaxomab (Removab) (anti-EpCAM x Anti-CD3): resultsof a phase 1/2 study. J Immunother 2009;32:195–202.

29. Fields AL, Keller A, Schwartzberg L, Bernard S, Kardinal C, CohenA, et al. Adjuvant therapy with the monoclonal antibody Edreco-lomab plus fluorouracil-based therapy does not improve overallsurvival of patients with stage III colon cancer. J Clin Oncol2009;27:1941–7.

30. Gires O, Bauerle PA. EpCAM as a target in cancer therapy. J ClinOncol 2010;28:e239–40.

31. Riethdorf S, Fritsche H, Muller V, Rau T, Schindlbeck C, Rack B, et al.Detection of circulating tumor cells in peripheral blood of patients withmetastatic breast cancer: a validation study of the CellSearch system.Clin Cancer Res 2007;13:920–8.

32. Sieuwerts AM, Kraan J, Bolt J, van der Spoel P, Elstrodt F, Schutte M,et al. Anti-epithelial cell adhesion molecule antibodies and the detec-tion of circulating normal-like breast tumor cells. J Natl Cancer Inst2009;101:61–6.

33. Jiao LR, Apostolopoulos C, Jacob J, Szydlo R, Johnson N, Tsim N,et al. Unique localization of circulating tumor cells in patients withhepatic metastases. J Clin Oncol 2009;27:6160–5.

34. Ford R, Schwartz L, Dancey J, Dodd LE, Eisenhauer EA, Gwyther S,et al. Lessons learned from independent central review. Eur J Cancer2009;45:268–74.

35. Karrison TG, Maitland ML, Stadler WM, Ratain MJ. Design of phase IIcancer trials using a continuous endpoint of change in tumor size:application to a study of sorafenib and erlotinib in non small-cell lungcancer. J Natl Cancer Inst 2007;99:1455–61.

36. Turner NC, Strauss SJ, Sarker D, Gillmore R, Kirkwood A, HackshawA, et al. Chemotherapy with 5-fluorouracil, cisplatin and streptozocinfor neuroendocrine tumours. Br J Cancer 2010;102:1106–12.

37. Cwikla JB, Sankowski A, Seklecka N, Buscombe JR, Nasierowska-Guttmejer A, Jeziorski KG, et al. Efficacy of radionuclide treatmentDOTATATE Y-90 in patients with progressive metastatic gastroenter-opancreatic neuroendocrine carcinomas (GEP-NETs): a phase IIstudy. Ann Oncol 2010;21:787–94.

38. Budd GT, Cristofanilli M, Ellis MJ, Stopeck A, Borden E, Miller MC,et al. Circulating tumor cells versus imaging—predicting overall sur-vival in metastatic breast cancer. Clin Cancer Res 2006;12:6403–9.

39. Kulke MH, Lenz HJ, Meropol NJ, Posey J, Ryan DP, Picus J, et al.Activity of sunitinib in patients with advanced neuroendocrine tumors.J Clin Oncol 2008;26:3403–10.

40. Yao JC, Lombard-Bohas C, Baudin E, Kvols LK, Rougier P, Rusz-niewski P, et al. Daily oral everolimus activity in patients with meta-static pancreatic neuroendocrine tumors after failure of cytotoxicchemotherapy: a phase II trial. J Clin Oncol 2010;28:69–76.

41. Rinke A, Muller HH, Schade-Brittinger C, Klose KJ, Barth P, WiedM, et al. Placebo-controlled, double-blind, prospective, rando-mized study on the effect of octreotide LAR in the control of tumorgrowth in patients with metastatic neuroendocrine midgut tumors:a report from the PROMID Study Group. J Clin Oncol 2009;27:4656–63.

42. Petersenn S, Schopohl J, Barkan A, Mohideen P, Colao A, Abs R, et al.Pasireotide (SOM230) demonstrates efficacy and safety in patientswith acromegaly: a randomized, multicenter, phase II trial. J ClinEndocrinol Metab 2010;95:2781–9.

CTCs in NETs





2011;17:337-345. Published OnlineFirst January 11, 2011.Clin Cancer Res Mohid S. Khan, Theodora Tsigani, Mohammed Rashid, et al. Neuroendocrine TumorsCirculating Tumor Cells and EpCAM Expression in

Updated version

10.1158/1078-0432.CCR-10-1776doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2011/03/21/1078-0432.CCR-10-1776.DC1Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/17/2/337.full#ref-list-1

This article cites 42 articles, 18 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/17/2/337.full#related-urls

This article has been cited by 12 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

SubscriptionsReprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. (CCC)Click on "Request Permissions" which will take you to the Copyright Clearance Center's

.http://clincancerres.aacrjournals.org/content/17/2/337To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-10-1776

http://clincancerres.aacrjournals.org/content/suppl/2011/03/21/1078-0432.CCR-10-1776.DC1

http://clincancerres.aacrjournals.org/content/17/2/337.full#ref-list-1

http://clincancerres.aacrjournals.org/content/17/2/337.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/17/2/337


Circulating Tumor Cells and EpCAM Expression in ... · should not express EpCAM (12, 13). The aims...

Documents

Transcript of Circulating Tumor Cells and EpCAM Expression in ... · should not express EpCAM (12, 13). The aims...