The Initial Detection and Partial Characterization of ... › content › ... · Biology of Human Tumors The Initial Detection and Partial Characterization of Circulating Tumor Cells

Biology of Human Tumors

The Initial Detection and Partial Characterizationof Circulating Tumor Cells in NeuroendocrineProstate CancerHimisha Beltran1,2, Adam Jendrisak3, Mark Landers3, Juan Miguel Mosquera2,4,Myriam Kossai2, Jessica Louw3, Rachel Krupa3, Ryon P. Graf3, Nicole A. Schreiber5,DavidM.Nanus1, Scott T.Tagawa1, DenaMarrinucci3, RyanDittamore3, andHoward I. Scher5

Abstract

Purpose: The transition of prostate adenocarcinoma to a pre-dominantly androgen receptor (AR) signaling independent phe-notype can occur in the later stages of the disease and is associatedwith low AR expression þ/� the development of small-cell orneuroendocrine tumor characteristics. As metastatic tumor biop-sies are not always feasible and are difficult to repeat, we sought toevaluate noninvasive methods to identify patients transitioningtoward a neuroendocrine phenotype (NEPC).

Experimental Design: We prospectively studied a metastatictumor biopsy, serum biomarkers, and circulating tumor cells(CTC, Epic Sciences) from patients with castration-resistant pros-tate cancer (CRPC) including those with pure or mixed NEPChistology present on biopsy. CTCs labeled with the patient'sclinical status were used to learn features that discriminate NEPCpatients, which was then applied to an independent cohort.

Results: Twenty-seven patients with CRPC including 12 NEPCand 5 with atypical clinical features suggestive of NEPC transitionwere studied. CTCs from NEPC patients demonstrated frequentclusters, low or absent AR expression, lower cytokeratin expres-sion, and smaller morphology relative to typical CRPC. A mul-tivariate analysis of protein and morphologic variables enableddistinguishing CTCs of NEPC from CRPC. This CTC classifier wasapplied to an independent prospective cohort of 159 metastaticCRPC patients and identified in 17/159 (10.7%) of cases,enriched in patients with high CTC burden (P < 0.01) and visceralmetastases (P ¼ 0.04).

Conclusions: CTCs from patients with NEPC have uniquemorphologic characteristics, which were also identified in a subsetofCRPCpatientswithaggressive clinical featurespotentially under-going NEPC transition. Clin Cancer Res; 22(6); 1510–9. �2015 AACR.

BackgroundNeuroendocrine prostate cancer (NEPC) is an aggressive,

androgen receptor (AR) independent subtype of prostate cancerthat most commonly becomes manifest in the later stages ofcastration-resistant prostate cancer (CRPC) and is associated withtreatment resistance (1–5). The diagnosis of NEPC remains chal-lenging and currently relies on a combination of pathologic andclinical features suggestive of AR signaling independence. BeforeNEPC develops, metastatic tumor biopsies often show mixedfeatures with both adenocarcinoma and neuroendocrine carci-noma cells present. There are no reliable serum markers toconsistently diagnose patients transforming to the NEPC pheno-

type and the incidence of circulating tumor cells (CTC) in thesepatients is unknown.Detection ofNEPChas clinical implications,as NEPC patients would not be expected to respond well tocurrently approved AR-targeted therapies for CRPC and may bebetter served by therapies specifically directed to NEPC.

CTCs provide the potential for noninvasive, real-time molec-ular characterization of cancer in patients withmetastatic disease.To date, the only FDA-cleared test for CTC detection and enu-meration is the CellSearch technology, based on immunomag-netic enrichment of CTCs expressing the epithelial cell adhesionmolecule (EpCAM). Several other platforms have recently beendeveloped to improve sensitivity of CTC detection, most of whichinclude enrichment and/or other physical selection methods(6, 7). There is mounting evidence that nontraditional popula-tions of CTCs also exist, including EpCAM/cytokeratin (CK)-negative CTCs (8) and/or cells smaller in size than traditionalCTCs, some even smaller than neighboring white blood cells(6, 9). The Epic Sciences platform is a non–selection-basedplatform that characterizes all nucleated cells and identifies CTCsbased on a multiparametric digital pathology process identifyingabnormal cells among the normal white blood cells utilizingprotein expression and cell morphology (10–12). This techniquehas demonstrated the ability to identify distinct CTC populationsincluding traditional (CKþ, CD45�), apoptotic, CK-negative, andCTC clusters (12, 13). We aimed to characterize CTCs frompatients with CRPC and NEPC utilizing the Epic platform andcorrelate results with patient-matched tumor biopsy and clinicalfeatures.

1Division of Hematology and Medical Oncology, Department of Med-icine, Weill Cornell Medicine, New York, New York. 2Institute for Pre-cision Medicine, New York Presbyterian-Weill Cornell Medicine, NewYork, NewYork. 3Epic Sciences, San Diego,California. 4Department ofPathologyandLaboratoryMedicine,Weill CornellMedicine, NewYork,NewYork. 5Genitourinary Oncology Service, Department of Medicine,Memorial Sloan Kettering Cancer Center, Weill Cornell Medicine,New York, New York.

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

Corresponding Author: Himisha Beltran, Weill Cornell Medicine, 413 East 69thStreet, #1412, New York, NY 10065. Phone: 646-962-2072; Fax: 646-962-1603;E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-15-0137

�2015 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 22(6) March 15, 20161510

on July 25, 2020. © 2016 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 15, 2015; DOI: 10.1158/1078-0432.CCR-15-0137

http://clincancerres.aacrjournals.org/

Materials and MethodsCTC collection

Under Institutional Review Board approved protocols atthe Weill Cornell Medicine and Memorial Sloan KetteringCancer Center, patients with metastatic CRPC including thosewith pure or mixed NEPC were prospectively enrolled. NEPCwas defined by the presence of either a pure or mixed small-cellhigh-grade neuroendocrine carcinoma histology in a metastatictumor biopsy and confirmed by at least 20% positive immu-nohistochemical (IHC) staining for the neuroendocrine markerchromogranin and/or synaptophysin. CRPC was defined clin-ically, with or without a metastatic biopsy confirming prostateadenocarcinoma. CRPC patients were subclassified as atypicalCRPC if a biopsy showed adenocarcinoma and the patient hadclinical features suggestive of an AR-independent transition,which included radiographic progression in the setting of a lowserum prostate specific antigen (PSA) <1 ng/mL, visceral pro-gression in the absence of PSA progression [defined by theProstate Cancer Working Group 2 criteria (14)] and/or elevatedserum chromogranin A > 3� upper limit of normal.

Clinical demographics including prior therapies, sites ofmetastases, PSA, serum neuroendocrine marker levels, and CTCnumber (CellSearch) were collected. Blood (10 mL) from eachsubject was shipped to Epic Sciences within 48 hours andprocessed immediately on arrival. Red blood cells were lysed,approximately 3 million nucleated blood cells dispensed onto10 to 16 glass slides as previously described (10–13) and storedat �80�C.

CTC identificationTwo slides from each patient were evaluated by immunofluo-

rescence (IF) (refs. 12, 13, 15; Fig. 1A) using antibodies targetingcytokeratins (CK), CD45, AR, and 40,6-diamidino-2-phenylin-dole (DAPI) counterstain. Slides were imaged using a platformthat captures all 3 million cells per slide in less than 15 minutes,and analyzed by a proprietary software that characterizes each cellby parameters including cell size, shape, nuclear area, presence ofmacronucleoli, CK and AR expression, uniformity, and cellularlocalization. CTC candidates were identified in an interactivereport, reviewed by trained technicians. CKþ/CD45� cells withintact, DAPIþ nuclei exhibiting tumor-associated morphologies

were classified as traditional CTCs. CTCs with nontraditionalcharacteristics were recorded, such as CK�/CD45� cells withmorphologic distinction and/or AR positivity, CKþ/CD45� smallcells, CTC clusters, CTCs with multiple marconucleoli, and apo-ptotic CTCs (with nuclear or cytoplasmic fragmentation).

Pathologic evaluationPatient-matched metastatic tumor biopsies were reviewed by

two anatomic genitourinary pathologists and classified as ade-nocarcinoma or NEPC based on presence of either pure ormixed small-cell high-grade neuroendocrine carcinoma histol-ogy in a metastatic tumor biopsy and confirmed by at least 20%positive IHC staining for the neuroendocrine marker chromo-granin and/or synaptophysin (16). IHC was quantified on scaleof 0 to 3 and positive IHC was defined as any staining intensityseen in target cells above background. To assess AURKA ampli-fication, we used a locus-specific probe plus reference probefluorescence in situ hybridization (FISH) assay, as previouslydescribed (17).

Statistical analysisCTC morphologic/molecular data and clinical information

were compiled into patient datasets (NEPC, CRPC, and atypicalCRPC) using KNIME, where cytokeratin expression, AR expres-sion, presence of clusters, and various nuclear and cytoplasmicmorphologic features were analyzed with single cell resolution(Supplementary Table S1). Kernel density estimates (KDE) ofeach CTC characteristic were performed to provide univariatedistributions across each aggregate subtype. Patient samples wereanalyzed for frequency of cell types at calculated cell counts permilliliter of blood, and univariate distributions of CTC biomar-kers were compared at the patient level for each diagnosticcategory. Supervised learning was performed using the RandomForest classifier algorithm (R package "randomForest") built with1,001 decision trees and configured to provide a probabilityoutput (18).

Leave-one-out cross-validationTo evaluate the robustness of the Random Forest classifier,

leave-one-out cross-validation was performed; CTCs frompatients with atypical CRPC were removed from analysis, andCTCs from NEPC were labeled NEPCþ and CRPC were labeledNEPC�. Leave-one-out cross-validation at the blood samplelevel with the dataset partitioned into training and test sets isshown in Supplementary Figs. S1–S3. For each blood tube,CTCs from every other sample were used to train a classifier,and CTCs from the blood tube being evaluated were held out asa test set. CTCs from the test set were analyzed by the trainedclassifier, where the output is an estimated probability of classmembership to NEPCþ and NEPC� for each CTC belonging tothe held-out sample. This cycle was repeated iteratively for eachsample, and the classifier output was collected at the end ofeach iteration. The criteria for patient-level class membershipwas established as at least three CTCs with a p(NEPC) scoregreater than 0.95.

Atypical CRPC and contemporary cohort analysisA classifier was first trained on NEPC and CRPC samples,

without atypical CRPC samples. This classifier was then used toclassify the atypical CRPC sample CTCs, as well as CTCs from a

Translational Relevance

NEPC is an aggressive variant of prostate cancer that mostcommonly arises in later stages of prostate cancer as a mech-anism of treatment resistance. Diagnosis of NEPC typicallyrelies on metastatic tumor biopsy as serum markers are unre-liable. We show that the CTC populations from patients withNEPC characterized with the Epic platform demonstrate aunique morphology, lower AR expression, and lower cytoker-atin expression compared with CTCs from other patients withcastration-resistant prostate cancer (CRPC). The presence ofNEPC-like CTC subpopulations occurs in approximately 10%of unselected patients with CRPC and is associated withaggressive clinical features. CTCs may provide utility for earlydiagnosis of NEPC-associated acquired resistance and war-rants larger clinical studies.

CTCs in Neuroendocrine Prostate Cancer

www.aacrjournals.org Clin Cancer Res; 22(6) March 15, 2016 1511




159 patient validation cohort. In the validation cohort, the samecriteria for patient positivity [at least three CTCs with p(NEPC)greater than 0.95] was applied to generate patient-level predic-tions from the classifier's single-cell output. KDE curves were usedto plot the distribution of NEPCþ class membership values forindividual CTCs for each patient.

ResultsCTCs from 27 patients with metastatic prostate cancer were

evaluated. The patients identified either pathologically as NEPC(n ¼ 12) or clinically as atypical CRPC (n ¼ 5) as defined abovedemonstrated a higher frequency of liver metastases and lowerPSA compared with other CRPC patients (Table 1, Supplemen-tary Table S2). Overall, bone metastases were present in 24/27(88.9%) of patients, and liver metastases were present in 8/12(66.7%) of NEPC and 5/15 (33.3%) of CRPC of whom 4 hadatypical clinical features (Supplementary Table S3). Medianserum PSA level was 1.9 ng/mL in NEPC, 2.8 ng/mL in atypical

CRPC, and 53.4 ng/mL in other CRPC patients. Serum neuro-endocrine marker levels varied considerably within the NEPCsubgroup and were also elevated in cases of CRPC (Supple-mentary Table 2).

CTCs in NEPC versus CRPCEnumeration of CTCs using both the CellSearch and Epic

platforms was performed. Of note, 6/13 evaluated NEPCand atypical CRPC patients had CellSearch CTC count of lessthan 5 CTC/7.5 mL (range 0–384, with 5 of these 13 patientshaving a CellSearch CTC count of 0). In contrast, all 17NEPC andatypical CRPC patients had CTCs greater than or equal to 5 CTC/7.5 mL using the Epic platform. Further characterization of thedetected CTCs revealed heterogeneity of CK and AR expression inboth NEPC and CRPC, with a significantly greater proportion ofCK-negative and AR-negative CTCs in NEPC compared withCRPC (Figs. 1 and 2; Supplementary Table S4). CTCs in NEPCpatients overall had lower AR expression, higher cytoplasmiccircularity, and higher nuclear to cytoplasmic ratio. The

Figure 1.A, Epic platform workflow starting from a single blood tube to the identification and characterization of all nucleated cells. Steps include (1) blood lysed, nucleatedcells from blood sample placed onto slides; (2) slides stored in �80�C biorepository; (3) slides stained with CK, CD45, DAPI, and AR; (4) slides scanned; (5)multiparametric digital pathology algorithms run; (6) software and human reader confirmation of CTCs and quantitation of biomarker expression; B, observed CKexpression, and C, AR expression of each CTC subtype: traditional CTC (blue), clustered CTCs (green), apoptotic CTCs (red), and CK� CTCs (purple) from eachpatient sample organized by their clinical diagnosis from biopsy.

Beltran et al.

Clin Cancer Res; 22(6) March 15, 2016 Clinical Cancer Research1512




Table

1.Clinical

datainclud

ingdiagno

sis,sitesofmetastases,priortherap

ies,biopsy

site,p

atho

logy,

andIHCresults

Metastaticbiopsy

Sitesofmetastases

Immun

ohistoch

emistry

Patient

#Diagno

sis

LILU

LNBO

AD

PL

PPI

Num

ber

of

therap

ies

forCRPC

Priortherap

ies

Biopsy

site

Patho

logy

SYP

CgA

AR

2NEPC

xx

0ADT

Lymphno

de

Neu

roen

docrinecarcinoma

þþ

NA

3NEPC

xx

xx

1ADT,carboplatin-etoposide

Live

r,TURP

Poorlydifferentiatedcarcinoma

w/neu

roen

docrinedifferentiation

þ�

�

7NEPC

xx

0ADT

Prostate

Highgradeprostatecancer

with

extensivene

uroen

docrine

differentiation

þþ

NA

8NEPC

xx

1ADT,cisplatin-etoposide

Lymphno

de

Neu

roen

docrinecarcinoma

þþ

NA

9NEPC

xx

1ADT,a

biraterone

Lymphno

de

Highgradeprostatecancer

with

neuroen

docrinedifferentiation

þþ

NA

10NEPC

xx

x1

ADT,cisplatin-etoposide

Bone

Small-cellcarcinoma

þþ

�12

NEPC

xx

xx

1ADT,a

biraterone

Live

rSmall-cellcarcinoma

þ�

�13

NEPC

xx

1ADT,cisplatin-docetaxel

Brain

Small-cellcarcinoma

þþ

�14

NEPC

xx

Prostate

Prostatecarcinomawithove

rlap

features

ofsm

all-cellcarcinoma

þ�

þ

24NEPC

xx

x1

ADT,carboplatin-etoposide

Live


þþ

NA

16NEPC

xx

xx

xLy

mphno

de

Small-cellcarcinoma

þ�

�26

NEPC

xx

2ADT,cisplatin-etoposide,

topotecan

Live


NA

NA

NA

6atyp

ical

CRPC

xx

x0

ADT

TURP

Poorlydifferentiatedad

enocarcinoma

��

þ1

atyp

ical

CRPC

xx

2ADT,d

ocetaxel,ab

iraterone

Lymphno

de


enocarcinoma

��

þ4

atyp

ical

CRPC

xx

3ADT,d

ocetaxel,en

zalutamide,

cabazitaxel

Live

rPoorlydifferentiatedAden

ocarcinoma

��

þ5

atyp

ical

CRPC

xx

x2

ADT,cisplatin-etoposide,

abiraterone

Pleura

PoorlydifferentiatedAden

ocarcinoma

��

NA

11atyp

ical

CRPC

xx

x2

ADT,d

ocetaxel,ab

iraterone

Live

rPoorlydifferentiatedad

enocarcinoma

��

þ17

CRPC

xx

x1

ADT,d

ocetaxel,ab

iraterone

Live

rPoorlydifferentiatedad

enocarcinoma

��

þ18

CRPC

x3

ADT,e

nzalutam

ide,

radium-223

,J59

1Bone


enocarcinoma

NA

NA

NA

19CRPC

xx

2ADT,sipuleu

cel-T,a

biraterone

Notperform

edNA

NA

NA

NA

20CRPC

xx

1ADT,sipuleu

cel-T

Notperform

edNA

NA

NA

NA

21CRPC

xx

4ADT,sipuleu

cel-T,J59

1,docetaxel,

abiraterone

Notperform

edNA

NA

NA

NA

22CRPC

xx

5ADT,a

biraterone

,enzalutam

ide,

J591,

docetaxel,carboplatin-paclitaxel

Notperform

edNA

NA

NA

NA

23CRPC

xx

6ADT,J59

1,docetaxel-lena

lidam

ide,

ipilumim

ab,sipieucel-T,a

biraterone

,cabozantinib

Bone


enocarcinoma

NA

NA

NA

25CRPC

xx

1ADT,a

biraterone

TURP


enocarcinoma

NA

�NA

27CRPC

x1

ADT,a

biraterone

Notperform

edNA

NA

NA

NA

28CRPC

xx

x2

ADT,d

ocetaxel,ab

iraterone

þGDC0068/

placebo

Pleura


enocarcinoma

��

NA

Abbreviations:A

D,adrena

l;ADT,and

rogen

dep

riva

tiontherap

y;AR,and

rogen

receptor;BO,b

one

;CgA,chromograninA;LI,liver;LN,lym

phno

de;LU

,lun

g;N

/A,notperform

ed;P

L,pleural;P

PI,pelvicperitone

alim

plants;SYP,

syna

ptophy

sin;

TURP,transurethral

resectionofprostate.






Figure 2.A, Kernel density estimate (KDE) curves for cytokeratin expression (left) and androgen receptor expression (right) for CTCs aggregated from all NEPC (red), CRPC(blue), and atypical CRPC (green) patient samples. B, Representative CTC images from patients with NEPC and CRPC, shown with the classifier output, whichis the estimated probability of the CTC's class membership as NEPCþ.

Beltran et al.





A a

b

c

d

e

f

g

h

i

BCRPC

2.5 0.8

0.6

0.4

0.2

0.0

0.8

0.6

0.4

0.2

0.0

16 32 64 128 256 512 16 32 64 128 256 1024512

2.0

1.5

1.0

0.5

0.0

0.01 0.1 1 10(log10) N/C ratio (log2) Nuclear area (mm2) (log2) Max cell area (mm2)

(log2) Nuclear entropy(log2) Cytoplasmic circularity(log2) Nuclear circularity0.5 1 2 0.5 2

2

2

3

1

0

1

0

Den

sity

Den

sity

Den

sity

Den

sity

Den

sity

2

3

1

0

Den

sity

1

84 2 84

NEPC

Figure 3.A, representative images of CTCs from CRPC and NEPC patients evaluated by IF for CD56 expression. All CTCs evaluated from CRPC patients were CD56negative. Heterogeneous expression of CD56 was observed within and among NEPC patient samples. CTCs including CKþ/CD56� (a–d, and f), CK�/CD56þ

(e), CKþ/CD56þ (g and h), small CKþ/CD56þ (g), and CKþ/CD56þ clusters (i) were identified in patient samples. a, patient 21, CD56 negative;b, patient 23, small CD56 negative; c, patient 23, CD56 negative; d, patient 28, CD56 negative, NEPC; e, patient 24, small CK� CD56 positive; f, patient 12, CD56negative; g, patient 12, CD56 positive; h, patient 10, CD56 positive; i, patient 26, cluster CD56 positive. B, Kernel density estimate (KDE) curvesare shown for CTC morphologic features ranging from nuclear/cytoplasmic ratio (top left), nuclear area (top center), maximum cell area (right top),nuclear circularity (bottom left), cytoplasmic circularity (bottom center), and nuclear entropy (bottom right) for CTCs aggregated from all NEPC (red),CRPC (blue), and atypical CRPC (green) patient samples.






prevalence of CK-negative CTC subpopulations in NEPC patientsis potentially consistent with epithelial–mesenchymal transition(EMT; refs. 19, 20).

Within the NEPC subgroup, there was a greater proportion ofsmall-cell CTCs in patients with metastatic biopsy confirmingsmall-cell carcinoma highlighting phenotypic similaritiesbetween tumor and CTCs. CTCs were tested by IF for the presenceof the neuroendocrine marker CD56 (Fig. 3A). Of those sampleswith matched with metastatic biopsy showing neuroendocrinefeatures and detectable CTCs, 7/12 (58%) had at least oneCD56þCTCwhereas 0/8 (0%)non-neuroendocrine sampleswithdetectable CTCs had �1 CD56þ CTC. Of patient samples withsmall-cell carcinoma pathology on tumor biopsy, 5/7 (71%) had�1CD56þCTC. A confusionmatrix demonstrates high specificityfor small-cell NEPC patients, demonstrating concordance totumor tissue (Supplementary Table S5). Additional molecularcharacterization of these CTCs using FISH for AURKA, a genecommonly amplified in NEPC (21), showed concordance withmatched metastatic biopsies in selected cases (an example ishighlighted in Supplementary Fig. S4) but was not present in allcases or all cells in positive cases.

Based on the observed differences in CTCs between groups, wesought to identify CTC characteristics specific to NEPC. Cell-levelfeatures were utilized to train Random Forest cell-level classifiersfor both leave-one-out cross-validation and for the classificationofCTCs in the test cohort, shown in Supplementary Table S1. KDEanalysis of the patient groups' CTCs in aggregate revealed signif-

icant differences in CK, AR, and morphologic characteristicsbetween patients with NEPC and CRPC (Figs. 2 and 3B).

Identification of NEPC CTCsTo demonstrate the diagnostic potential of CTC characteristics

in distinguishing NEPC, the observed differences between NEPCand CRPC were used to train a Random Forest classifier. Resultsfrom leave-one-out cross-validation of NEPC and CRPC samplesare shown in Supplementary Figs. S5 and S6, where the outputfrom the classifier is a p(NEPCþ) value and a p(NEPC�) value foreach CTC, corresponding to the estimated probability of the cell'sclass membership as NEPCþ and NEPC�.

From the density curve in Supplementary Fig. S4A, the samplesfrom patients with NEPC demonstrated a spike in the curves nearthe high end of the p(NEPCþ) spectrum, with many curvespeaking near a p(NEPCþ) score of 95%. In Supplementary Fig.S4B, the number of CTCs/mL with a p(NEPCþ)�� score greaterthan or equal to 95% is presented as a bar chart for each patientsample, where each column is colored by the actual clinicaldiagnosis that the classifier is trying to predict.

Obtaining positive signals at theCTC level fromsamples that theclassifier does not encounter during training demonstrates theclassifier's ability to detect NEPC from CRPC in a robust mannerthat mitigates the risk of overfitting. These conditions simulate theenvironment that the classifier would face in practice, in the sensethat any future blood sample sent in for NEPC analysis is presentedto the algorithm as a series of CTCs that it has not encountered

A B

C D

H&E H&E

SYP AR SYP ARPatient 6 Patient 12

Patient 6 Patient12

Atypical CRPC

6

4

2

0

0.00 0.25 0.50 0.75 1.00 0.00

0.0

0.5

1.0

1.5

2.0

NEPC

Den

sity

0.25 0.50 0.75 1.00p(NEPC+) p(NEPC+)

Den

sity

Figure 4.Metastatic biopsy showingmorphologic characteristics and IHCfor synaptophysin (SYP) andandrogen receptor (AR) of metastaticbiopsies frompatients 6 (A) and 12 (B).Atypical CRPC patient 6 tumor wascharacterized as poorly differentiatedadenocarcinoma, synaptophysinnegative, AR positive; NEPC patient 12tumor was characterized as small-cellcarcinoma, synaptophysin positive,AR negative. C, Patient 6 distributionbased on NEPC probability classmemberships. D, Patient 12distribution based on NEPCprobability class membership.


Beltran et al.




during training, and the classifier then estimates the probability ofclass membership for each CTC from the new sample.

Atypical CRPCThe clinical significance of patients with castration-resistant

adenocarcinoma that develop progressive disease in the setting oflow serum PSA <1 ng/mL, visceral metastases in the absence ofPSA progression, or elevated serum chromogranin is not wellestablished.One hypothesis is that these tumors are less androgenresponsive and may be in transition toward an AR low or NEPCphenotype and/or demonstrate intratumoral heterogeneity withboth adenocarcinoma and NEPC present within or betweenmetastases. We applied the NEPC classification model trained todistinguish NEPC versus CRPC CTCs to the five atypical CRPCpatients and found that atypical CRPC is associated with anincrease in heterogeneity of CRPC cells and a higher burden ofNEPC-like cells compared with CRPC patients (SupplementaryFig. S4, Supplementary Table S6).

Patient case studiesAtypicalCRPCpatient 6, for example, harboredCTCsof various

morphologies with a predominance of NEPCþ CTCs (Fig. 4).Patient 6 is a 64-year-old man who presented with metastatichormone naive prostate cancer, developed clinical progressionwithin 6 months on primary hormonal therapy, had short dura-tion of response to subsequent abiraterone, radium-223, anddocetaxel, and developed progressive bone metastases and newliver metastases in the setting of a non-rising PSA. Despite hisbone biopsy at progression showing adenocarcinoma withoutneuroendocrine features (Fig. 4A), his clinical history and CTCcharacteristics obtained at the time of bone biopsy supported ARindependence.

Another example of how CTCs can be used to understanddisease heterogeneity is illustrated in the case of patient 12, a

68-year-old gentleman with CRPC who had a bone biopsy at thetime of castration resistance for research, which showed prostateadenocarcinoma. He was treated with abiraterone and predni-sone. Despite PSA stability, follow-up imaging at 3 months onabiraterone revealed new liver and lungmetastases and his serumchromogranin was markedly elevated at 17,340 ng/mL (ULN 95ng/mL). Liver biopsy was consistent with NEPC (small-cellcarcinoma; Fig. 4B). Similar to patient 6, CTCs at time of liverbiopsy showed heterogeneous CTC populations including bothNEPC and CRPC cell characteristics, suggesting intrapatient het-erogeneity. These cases support CTCs as potentially useful incapturing tumor heterogeneity that might not be assessed onmetastatic biopsy.

Validation cohortWe evaluated baseline CTCs from 159 CRPC patients prospec-

tively enrolled in an independent patient cohort atMSKCC for thepresence of NEPCþ CTCs (Fig. 5A). NEPCþ CTC subpopulationswere identified in 17 of 159 (10.7%) cases. A significantly higherproportion of CRPC patients with visceral metastases harboredNEPCþCTCs compared with those that were NEPC� (35% versus15%, respectively; P¼ 0.04). Patients with NEPCþ CTCs also hadanoverall higher CTCburden (medianCTC count 64.6 versus 4.2;P < 0.01). To address whether the CTC classifier was a reflection ofan overall higher CTC count, linearity was assessedwith a Pearsoncoefficient showing a weak relationship between frequency ofNEPC CTCs and total cell count (Supplementary Fig. S6). Rep-resentative images of NEPCþ CTC characteristics observed in thevalidation cohort are shown in Fig. 5B.

DiscussionHistologic and molecular subtyping of cancer often influences

clinical decision making, and tissue confirmation is typicallyrequired at cancer diagnosis before treatment recommendations

Liver mets

Traditional CTC(CK+, CD45-)

CTC clusters(≥2 CTCs in a network)

CK-CTC(CK-, CD45-, biomarker+)

Apop CTC(CK+, CD45-, nuclear

hotspots, pulling apartor condensing)

Lung mets

Visceralmets

Bone mets

Age

Median MedianRange MedianRange Range

69

39

4.2

48−87

<0.01

<0.01

0.1−3728.2

0.9−376.714.1−897.7

0.7−2589.9

52−9166

190.7

64.60.9−897.7

0.1−3728.2

48−91

5

68

43.9

138

27

17

15 9% 18%

24%

35%

100%

11%

17%

87% 17

6

4

3

NS

NS

NS

NS

NEPC−NEPC+AllA

B

N = 142N = 17N = 159

8%

9%

15% 0.04

P value

85%

21

121

13

12

PSA (ng/mL)

CTC (mL)

Continuousparameters

Categoricalparameters

Figure 5.A, clinical data of validation cohort(n ¼ 159) including sites ofmetastases, age, serum PSA, and CTCcount. NS, not statistically significantat 0.05 level. P values from two-sidedtests comparing NEPCþ to NEPC� arebased on the Fisher exact test forcategorical parameters and Wilcoxonrank sum for continuous parameters.B, representative images showingCTC characteristics from patients invalidation cohort that are classifierpositive.






are offered. Prostate cancer is the most common cancer in men inthe United States and Europe (22, 23), and in nearly all casesdiagnostic biopsies reveal adenocarcinoma upon initial diagno-sis. Prostate adenocarcinomas are characterized by AR expressionand activation, and therefore hormonal therapies targeting the ARare the mainstay of systemic therapy (24). Small-cell neuroen-docrine carcinoma of the prostate is a rare histologic subtype atdiagnosis, representing less than 1% of all new prostate cancerdiagnoses (25). However, in a subset of patients with metastaticprostate adenocarcinoma treated with AR-targeted therapies,prostate adenocarcinomas can develop histologic transformationtoward a predominantly neuroendocrine carcinoma likely as amechanism of acquired resistance (1–5). The NEPC phenotype isassociated with aggressive disease, frequent visceral metastases,and low or absent AR expression onmetastatic tumor biopsy (4).In this setting, patients are often offered platinum-based chemo-therapy with regimens similar to small-cell neuroendocrine car-cinoma of the lung (26, 27). Therefore, identification of advancedprostate cancer patients that have acquired NEPC has potentialclinical implications.

However, the diagnosis of NEPC can be complex as there is aspectrum of morphologies seen in advanced prostate cancer withAR-positive adenocarcinoma and AR-negative small-cell carcino-ma representing the extreme phenotypes. Metastatic biopsiesoften reveal mixed features of both adenocarcinoma and neuro-endocrine carcinoma, with variable AR or neuroendocrinemarkerprotein expression (16). The clinical significance ofmixed tumorsis less clear and treatment decisions are often individualizedbasedon a combination of pathologic and clinical features. Further-more, for patients with an atypical clinical presentation such asrapid radiographic progression in setting of a low or modestlyelevated PSA, platinum-based therapies are sometimes consid-ered even in the absence of neuroendocrine morphology onbiopsy. Another challenge in the diagnosis of NEPC is thatmetastatic biopsies are not always feasible for patients sufferingfrom advanced prostate cancer, may carry additional risks for thepatient including complications from biopsy procedure or delayof initiation of appropriate systemic therapy, and does not alwayscapture disease heterogeneity. Therefore, a noninvasive marker todetect NEPC progression and simultaneously capture intrapatientheterogeneity is an unmet need.

We found that CTCs from metastatic prostate cancer patientsare often phenotypically heterogeneous. CTCs from patients withpathologically confirmed NEPC were predominantly of smallersize compared with other CRPC patients and demonstrated lowerAR expression and abnormal nuclear and cytoplasmic features.There was also a higher prevalence of low cytokeratin expressingCTCs in NEPC, possibly related to EMT changes that can occurduring metastatic transit and treatment resistance (28, 29). Whenapplied to an independent cohort, we found that up to 10% ofCRPCpatients also harbored similar NEPCþCTC subpopulationsand their presence was associated with aggressive clinical features(i.e., visceral metastases and high CTC burden). These datasupport the possible detection of circulating neuroendocrine

cancer cells in patients with metastatic CRPC; however, the lackof definitivemarkers andmixed cellular subpopulations observedreinforce the biologic and clinical complexity underlying diseaseprogression andNEPC transformation. Differences in nuclear sizemay also be contributed by the presence of visceral metastases ashas been recently observed by Chen and colleagues (30). Whatremains unclear are the dynamics by which these CTCs arise andhow the classifier performs as a predictive or prognostic biomark-er. Serial monitoring of CTCs in larger cohorts could help eluci-date how consistently the classifier emerges during the course oftherapy and during the CRPC-to-NEPC transition. Future studiesincluding single-cell sequencing of CTCswill also be important tomolecularly characterize these heterogeneous populations andmay improve our understanding of this complex resistancephenotype.

In this proof of principle study, we demonstrate that CTCs frompatients with NEPC have distinct characteristics. The results pre-sented here indicate the feasibility of analyzing CTCs using theEpic platform and support the development of further studies tovalidate the clinical utility of CTCs for the early detection ofpatients transforming toward NEPC and the prognostic andpotential predictive impact of CTC characteristics in predictingresponse to AR-directed therapies in CRPC.

Disclosure of Potential Conflicts of InterestH.I. Scher is a consultant/advisory board member for BIND Therapeutics,

Exelexis, and Janssen, and reports receiving commercial research grants fromBIND Therapeutics, Epic Sciences, Exelexis, and Janssen. No potential conflictsof interest were disclosed by the other authors.

Authors' ContributionsConception and design: H. Beltran, D. Marrinucci, R. Dittamore, H.I. ScherDevelopment of methodology: A. Jendrisak, M. Landers, D. Marrinucci, R.Dittamore, H.I. ScherAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): H. Beltran, M. Landers, J. Louw, R. Krupa, D.M.Nanus, S.T. Tagawa, R. DittamoreAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): H. Beltran, A. Jendrisak, M. Landers, J.M. Mosquera,M. Kossai, R. Dittamore, H.I. ScherWriting, review, and/or revision of themanuscript:H. Beltran, A. Jendrisak,M.Landers, J.M. Mosquera, J. Louw, R. Krupa, R.P. Graf, D.M. Nanus, S.T. Tagawa,D. Marrinucci, R. Dittamore, H.I. ScherAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases):H. Beltran, A. Jendrisak, J.M.Mosquera, J. Louw,N. SchreiberStudy supervision: H. Beltran, R. Dittamore

Grant SupportDepartment of Defense W81XWH-13-1-0275 (to H. Beltran), and Damon

Runyon-Gordon Family Clinical Investigator Award CI-67-13 (to H. Beltran).The costs of publication of this articlewere defrayed inpart by the payment of

page charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received January 16, 2015; revised October 20, 2015; accepted October 21,2015; published OnlineFirst December 15, 2015.

References1. Aggarwal R, Zhang T, Small EJ, Armstrong AJ. Neuroendocrine prostate

cancer: subtypes, biology, and clinical outcomes. J Natl Compr Canc Netw2014;12:719–26.

2. Tagawa ST. Neuroendocrine prostate cancer after hormonaltherapy: knowing is half the battle. J Clin Oncol 2014;32:3360–4.

Beltran et al.





3. Beltran H, Tagawa ST, Park K, MacDonald T, Milowsky MI, Mosquera JM,et al. Challenges in recognizing treatment-related neuroendocrine prostatecancer. J Clin Oncol 2012;30:e386–9.

4. Beltran H, Tomlins S, Aparicio A, Arora V, Rickman D, Ayala G, et al.Aggressive variants of castration-resistant prostate cancer. Clin Cancer Res2014;20:2846–50.

5. Aparicio A, Tzelepi V. Neuroendocrine (Small-Cell) carcinomas: why theyteach us essential lessons about prostate cancer. Oncology (Williston Park)2014;28:831–8.

6. Ozkumur E, Shah AM, Ciciliano JC, Emmink BL, Miyamoto DT, BrachtelE, et al. Inertial focusing for tumor antigen-dependent and -indepen-dent sorting of rare circulating tumor cells. Sci Transl Med 2013;5:179ra47.

7. Yap TA, Lorente D, Omlin A, Olmos D, de Bono JS. Circulating tumor cells:a multifunctional biomarker. Clin Cancer Res 2014;20:2553–68.

8. Yu M, Bardia A, Wittner BS, Stott SL, Smas ME, Ting DT, et al. Circulatingbreast tumor cells exhibit dynamic changes in epithelial andmesenchymalcomposition. Science 2013;339:580–4.

9. Phillips KG,KuhnP,McCartyOJ. Physical biology in cancer. 2. The physicalbiology of circulating tumor cells. Am J Physiol Cell Physiol 2014;306:C80–8.

10. Hsieh HB, Marrinucci D, Bethel K, Curry DN, Humphrey M, Krivacic RT,et al. High speed detection of circulating tumor cells. Biosens Bioelectron2006;21:1893–9.

11. Marrinucci D, Bethel K, Bruce RH, Curry DN, Hsieh B, Humphrey M, et al.Case study of the morphologic variation of circulating tumor cells. HumPathol 2007;38:514–9.

12. Marrinucci D, Bethel K, Kolatkar A, Luttgen MS, Malchiodi M, Baehring F,et al. Fluid biopsy in patients with metastatic prostate, pancreatic andbreast cancers. Phys Biol 2012;9:016003.

13. Werner SL, Graf RP, LandersML, ValentaDT, SchroederM, Greene SB, et al.Analytical validation and capabilities of the epic CTC platform: enrich-ment-free circulating tumour cell detection and characterization. J Circu-lating Biomarkers 2015;4:1–13.

14. Scher HI, Halabi S, Tannock I, Morris M, Sternberg CN, Carducci MA, et al.Design and end points of clinical trials for patients with progressiveprostate cancer and castrate levels of testosterone: recommendations ofthe Prostate Cancer Clinical Trials Working Group. J Clin Oncol2008;26:1148–59.

15. Marrinucci D, Bethel K, Lazar D, Fisher J, Huynh E, Clark P, et al. Cyto-morphology of circulating colorectal tumor cells: a small case series. JOncol 2010;2010:861341.

16. Epstein JI, Amin MB, Beltran H, Lotan TL, Mosquera JM, Reuter VE, et al.Proposed morphologic classification of prostate cancer with neuroendo-crine differentiation. Am J Surg Pathol 2014;38:756–67.

17. Mosquera JM, Beltran H, Park K,MacDonald TY, Robinson BD, Tagawa ST,et al. Concurrent AURKA andMYCN gene amplifications are harbingers oflethal treatment-related neuroendocrine prostate cancer. Neoplasia2013;15:1–10.

18. Breiman L. Random forests. Mach Learn 2001;45:5–32.19. Tam WL, Weinberg RA. The epigenetics of epithelial-mesenchymal plas-

ticity in cancer. Nat Med 2013;19:1438–49.20. Scheel C, Weinberg RA. Cancer stem cells and epithelial-mesenchymal

transition: concepts andmolecular links. Semin Cancer Biol 2012;22:396–403.

21. Beltran H, Rickman DS, Park K, Chae SS, Sboner A, MacDonald TY, et al.Molecular characterization of neuroendocrine prostate cancer and identi-fication of new drug targets. Cancer Discov 2011;1:487–95.

22. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2015. CA Cancer J Clin2015;65:5–29.

23. Malvezzi M, Bertuccio P, Levi F, La Vecchia C, Negri E. Europeancancer mortality predictions for the year 2014. Ann Oncol 2014;25:1650–6.

24. Cooper CS, Eeles R, Wedge DC, Van Loo P, Gundem G, Alexandrov LB,et al. Corrigendum: analysis of the genetic phylogeny of multifocalprostate cancer identifies multiple independent clonal expansions inneoplastic and morphologically normal prostate tissue. Nat Genet2015;47:689.

25. Wang W, Epstein JI. Small cell carcinoma of the prostate. A morphologicand immunohistochemical study of 95 cases. Am J Surg Pathol 2008;32:65–71.

26. Papandreou CN, Daliani DD, Thall PF, Tu SM, Wang X, Reyes A, et al.Results of a phase II study with doxorubicin, etoposide, and cisplatin inpatients with fully characterized small-cell carcinoma of the prostate. J ClinOncol 2002;20:3072–80.

27. Aparicio AM, Harzstark AL, Corn PG, Wen S, Araujo JC, Tu SM, et al.Platinum-based chemotherapy for variant castrate-resistant prostate can-cer. Clin Cancer Res 2013;19:3621–30.

28. Armstrong AJ, Marengo MS, Oltean S, Kemeny G, Bitting RL, Turnbull JD,et al. Circulating tumor cells from patients with advanced prostate andbreast cancer display both epithelial and mesenchymal markers. MolCancer Res 2011;9:997–1007.

29. Tanaka H, Kono E, Tran CP, Miyazaki H, Yamashiro J, Shimomura T,et al. Monoclonal antibody targeting of N-cadherin inhibits prostatecancer growth, metastasis and castration resistance. Nat Med2010;16:1414–20.

30. Chen J, Ho H, Lichterman J, Lu Y, Zhang Y, Garcia MA, et al. Subclassi-fication of prostate cancer circulating tumor cells by nuclear size revealsvery small nuclear circulating tumor cells in patients with visceral metas-tases. Cancer 2015;121:3240–51.






2016;22:1510-1519. Published OnlineFirst December 15, 2015.Clin Cancer Res Himisha Beltran, Adam Jendrisak, Mark Landers, et al. Tumor Cells in Neuroendocrine Prostate CancerThe Initial Detection and Partial Characterization of Circulating

Updated version

10.1158/1078-0432.CCR-15-0137doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2015/12/15/1078-0432.CCR-15-0137.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/22/6/1510.full#ref-list-1

This article cites 30 articles, 12 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/22/6/1510.full#related-urls

This article has been cited by 8 HighWire-hosted articles. Access the articles at:

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

Subscriptions

Reprints and

[email protected]

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

Permissions

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

.http://clincancerres.aacrjournals.org/content/22/6/1510To 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-15-0137

http://clincancerres.aacrjournals.org/content/suppl/2015/12/15/1078-0432.CCR-15-0137.DC1

http://clincancerres.aacrjournals.org/content/22/6/1510.full#ref-list-1

http://clincancerres.aacrjournals.org/content/22/6/1510.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/22/6/1510


Documents

The Initial Detection and Partial Characterization of ... › content › ... · Biology of Human Tumors The Initial Detection and Partial Characterization of Circulating Tumor Cells