rb loss is characteristic of prostatic small cell ... · anti-human p53 antibody (clone do-7; dako;...

Human Cancer Biology

Rb Loss Is Characteristic of Prostatic Small CellNeuroendocrine Carcinoma

Hsueh-Li Tan1, Akshay Sood1, Hameed A. Rahimi1, Wenle Wang1, Nilesh Gupta2, Jessica Hicks1,Stacy Mosier1, Christopher D. Gocke1,3, Jonathan I. Epstein1,3,4, George J. Netto1,3,4, Wennuan Liu5,William B. Isaacs3,4, Angelo M. De Marzo1,3,4, and Tamara L. Lotan1,3

AbstractPurpose: Small cell neuroendocrine carcinoma of the prostate is likely to become increasingly common

with recent advances in pharmacologic androgen suppression. Thus, developingmolecularmarkers of small

cell differentiation in prostate cancer will be important to guide the diagnosis and therapy of this aggressive

tumor.

Experimental Design:We examined the status of RB1, TP53, and PTEN in prostatic small cell and acinar

carcinomas via immunohistochemistry (IHC), copy-number alteration analysis, and sequencing of for-

malin-fixed paraffin-embedded specimens.

Results:We found retinoblastoma (Rb) protein loss in 90%of small cell carcinoma cases (26 of 29) with

RB1 allelic loss in 85% of cases (11 of 13). Of acinar tumors occurring concurrently with prostatic small cell

carcinoma, 43% (3 of 7) showed Rb protein loss. In contrast, only 7% of primary high-grade acinar

carcinomas (10 of 150), 11% of primary acinar carcinomas with neuroendocrine differentiation (4 of 35),

and 15% of metastatic castrate-resistant acinar carcinomas (2 of 13) showed Rb protein loss. Loss of PTEN

proteinwas seen in 63%of small cell carcinomas (17 of 27), with 38%(5of 13) showing allelic loss. By IHC,

accumulation of p53 was observed in 56% of small cell carcinomas (14 of 25), with 60% of cases (6 of 10)

showing TP53 mutation.

Conclusions: Loss of RB1 by deletion is a common event in prostatic small cell carcinoma and can be

detected by a validated IHC assay. As Rb protein loss rarely occurs in high-grade acinar tumors, these data

suggest that Rb loss is a critical event in the development of small cell carcinomas and may be a useful

diagnostic and potential therapeutic target. Clin Cancer Res; 20(4); 890–903. �2013 AACR.

IntroductionSmall cell neuroendocrine carcinoma is an unusual and

aggressive subtype of prostate cancer, accounting for 0.5%to 2% of all untreated primary tumors (1, 2). Small cellcarcinoma represents an extreme example of neuroendo-crine differentiation in prostate cancer. The other end of thespectrum encompassesmore typical acinar primary tumors,in which scattered rare cells expressing neuroendocrinemarkers can be found in 10% to 100% of adenocarcinomas

of the prostate (3, 4). Interestingly, autopsy studies havesuggested that more extensive neuroendocrine differentia-tion (including small cell carcinoma) is present in up to25% of lethal, widely metastatic cases (3–5). This apparentselection for the neuroendocrine phenotype in end-stagedisease is likely due to the fact that tumors with extensiveneuroendocrine differentiation, like small cell carcinoma,are almost invariably castration resistant and typically lackexpression of both androgen receptor (AR) and down-stream targets of the androgen signaling axis (2, 4, 6–12).Indeed, it is anticipated that recent improvements in phar-macologic options for androgen suppression therapy suchas abiraterone and enzalutamide, which are even moreeffective at blocking AR signaling than traditional methods,will lead to a relative upsurge in small cell carcinoma casesin the near future (13), making it especially important tofully understand the molecular and immunophenotypicunderpinnings of this tumor type.

Both de novo and treatment-related small cell carcinomacases commonly occur concurrently with typical acinaradenocarcinoma (1, 2). We and others have shown limitedmolecular evidence of a clonal relationship between the twocomponents based on the concordant presence of ERG generearrangements or TP53mutations (12, 14–17); however, it

Authors' Affiliations: 1Pathology, 2Department of Pathology, Henry FordHealth System, Detroit, Michigan; 3Oncology, and 4Urology, Johns Hop-kins University School of Medicine, Baltimore, Maryland; and 5Center forGenomics and Personalized Medicine Research, Wake Forest School ofMedicine, Winston-Salem, North Carolina

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

H.-L. Tan and A. Sood contributed equally to this work.

CorrespondingAuthor:Tamara L. Lotan, JohnsHopkins University Schoolof Medicine, 454Rangos, 855NWolfe Street, Baltimore, MD 21205. Phone:410-614-9196; Fax: 410-502-9911; E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-13-1982

�2013 American Association for Cancer Research.

ClinicalCancer

Research

Clin Cancer Res; 20(4) February 15, 2014890

on March 20, 2021. © 2014 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 9, 2013; DOI: 10.1158/1078-0432.CCR-13-1982

http://clincancerres.aacrjournals.org/

remains unclear how small cell carcinoma differs at amolecular level from its acinar counterpart. Indeed, delin-eating theuniquemolecular features of small cell carcinomathat distinguish it from acinar carcinoma would be usefulfor both diagnosis and potential therapy of this aggressivesubtype. Accurately distinguishing small cell carcinomafrom high-grade (Gleason pattern 4 or 5) acinar prostatecarcinoma can be difficult based onmorphology alone, andcurrent immunohistochemicalmarkers for neuroendocrinedifferentiation may be nonspecific (18). However, thisdistinction is extremely consequential for clinical patientmanagement as small cell carcinomas are often resistant toandrogen deprivation therapy at the outset, and are at leasttemporarily sensitive to cisplatin-based chemotherapeuticregimens in contrast with acinar carcinoma (2, 6–11). Thus,identifying additional molecular differences between smallcell carcinoma and acinar carcinoma would not only aid inaccurate diagnosis of these lesions, but also provide addi-tional potential therapeutic targets for this currently uni-formly lethal tumor type.Although prior molecular studies of prostatic small cell

carcinomas have been limited, a number of studies of smallcell carcinomas arising in other organs have suggested thatloss of the RB1 and TP53 genes are extremely commonevents in this tumor type in humans, and these changes aresufficient to generate small cell carcinomas in the lung ofmice (19–21). Intriguingly, prostate-specific inactivation ordeletion of these two genes in the mouse is sufficient toresult in highly penetrant metastatic carcinoma with neu-roendocrine features (22, 23). In prostatic acinar carcino-mas, loss of theRB1 andTP53 tumor suppressors does occurin a subset of primary tumors; however, PTEN is typicallythe most commonly lost tumor suppressor in this setting

(24–26). Here, we comprehensively examined the proteinand genomic status of RB1, PTEN, and TP53 in a series ofsmall cell and acinar carcinomas. We report that RB1 lossoccurs virtually uniformly in small cell carcinoma of theprostate, a finding that may be useful in the future for bothdiagnosis and therapy of this tumor.

Materials and MethodsTissue selection

A total of 29 specimens frompatients with prostatic smallcell carcinoma were retrieved from the surgical pathologyand consultation files of the Johns Hopkins Hospitals from1994 to 2008 as previously reported (12). Nineteen cases(66%)were transurethral resections of the prostate (TURP),8 (27%) were bladder, prostate, or rectal biopsies, 1 (3%)was a radical prostatectomy, and 1 (3%) was a liver metas-tasis. (Table 1). Prostatic origin was documented for themajority of cases using the following criteria: in 76% (22 of29) of cases, a concurrent or prior history of prostatic acinarcarcinoma was established, whereas 7% (2 of 29) of caseswere diagnosed by positive PSA (prostate-specific antigen)immunostaining and 3% (1 of 29) had a documentednegative cystoscopy. Thus, in the vast majority of cases, thesmall cell carcinoma arose in the setting of a current or prioracinar prostatic adenocarcinoma. Although 97% (28 of 29)of the tissue samples were from the prostate or contiguoussites (i.e., most likely nonmetastatic specimens), due to thefact that the majority of these cases were consultations,adequate clinical history was not available in 16 (55%)cases to determine which small cell carcinomas arose in thecontext of prior hormonal therapies (treatment-relatedneuroendocrine prostate cancer, t-NEPC). Overall, of thecases in which we did have clinical data, 7 (23%) patientshad no prior history of therapy for prostate cancer andrepresented new diagnoses, 3 (10%) had a history of priorradiation treatment for prostate cancer, and 3 (10%) of thepatients had a documented history of prior treatment withhormonal therapies.

A tissue microarray (TMA) was manually constructedfrom these cases as previously described (12). In each case,a minimum of three 1.0-mm cores were punched from thesmall cell carcinoma component, the acinar carcinomacomponent (whenpresent), and the paired benignprostatictissue, with 3 to 18 cores from each patient represented onthe array. Of note, 24% (7 of 29) of cases had a concurrentacinar carcinoma component present on the TMA.

For assessment of retinoblastoma (Rb) protein expres-sion in high-grade acinar prostatic carcinoma cases unas-sociated with a small cell carcinoma component, 170 casesfrompreviously described TMAs comprising several cohortsof high grade, high-stage prostate tumors were used (27–29). Rb protein expression was also assessed on standardsections of an additional 15 cases of non–small cell primaryprostatic carcinoma with neuroendocrine differentiation(including 6 cases of large cell neuroendocrine carcinoma)from the consultation files of the Johns Hopkins Hospital.Finally, Rb immunostaining was performed on standard

Translational RelevanceSmall cell carcinoma of the prostate represents an

extreme example of neuroendocrine differentiation inprostate cancer. However, with the advent of potentandrogen suppressive therapies such as abiraterone andenzalutamide, neuroendocrine differentiation is likelyto become more common, representing an importantmode of drug resistance. Thus, defining molecular andimmunohistochemical markers for prostatic small cellcarcinoma could be helpful in predicting treatmentresponse in this setting. We report that loss of theretinoblastoma (Rb) tumor suppressor occurs nearlyuniversally in prostatic small cell carcinomas. In con-trast, loss of this tumor suppressor occurs rarely inconventional high-grade acinar prostate tumors, acinarcarcinomas with neuroendocrine differentiation, andcastrate-resistant prostate carcinomas. These data sug-gest that Rb loss is a critical event in the development ofsmall cell carcinomas of the prostate andmay be a usefuldiagnostic and potential therapeutic target in the settingof neuroendocrine differentiation in prostate cancer.

RB1, TP53, and PTEN Loss in Prostatic Small Cell Carcinoma

www.aacrjournals.org Clin Cancer Res; 20(4) February 15, 2014 891




sections of 13 samples of metastatic castrate-resistant pros-tate cancer that were previously characterized for RB1 copy-number alteration using a high-density single-nucleotidepolymorphism array (Affymetrix 6.0 SNP microarray; Sup-plementary Table S1; ref. 30). For validation of the nCoun-ter Cancer Copy Number Variation panel at the PTEN andRB1 loci (described below),DNAderived from frozen tissuesamples from 5 of these cases of metastatic acinar prostaticcarcinoma in this series was used (30).

ImmunohistochemistryTissue sections were deparaffinized, rehydrated, and

briefly equilibrated in water. Antigen unmasking was per-formed by either steaming in high-temperature targetretrieval solution (Target Retrieval Solution; Dako) for 50minutes (p53, Rb) or in EDTA for 45 minutes (PTEN).Endogenous peroxidase activity was quenched by incuba-tion with peroxidase block for 5 minutes at room temper-ature. Slides were incubated with a mouse monoclonal

anti-human p53 antibody (Clone DO-7; Dako; 1:400 dilu-tion), a mouse monoclonal anti-human Rb antibody (CellSignaling Technology; 1:100 dilution), or a rabbit anti-human PTEN antibody (Clone D4.3 XP; Cell SignalingTechnology; 1:100 dilution). A horseradish peroxidase–labeled polymer (PowerVision; Leica Microsystems) wasapplied for 30 minutes at room temperature. Signal detec-tion was performed using 3,30-diaminobenzidine tetrahy-drochloride (DAB) as the chromagen. Slides were counter-stained with hematoxylin, dehydrated, and mounted.

Validation and scoring of immunohistochemistryRb, p53, and PTEN immunostains were scored in small

cell carcinoma and the associated acinar carcinoma com-ponent by a urologic pathologist (T.L. Lotan). On the basisof the controls described in Results below, a case wasconsidered to have lost Rb protein if any TMA spot showedloss (0þ staining) in >95% of the tumor nuclei. Positivenuclear staining in surrounding endothelial cells provided

Table 1. Rb, PTEN, and p53 IHC results in small cell carcinoma (SCC) and associated acinar carcinoma(ACa) component where present

Rb PTEN p53

Case # Specimen type Prostatic SCC confirmed by SCC ACa SCC ACa SCC ACa

1 TURP Concurrent ACa n p n n n p4 TURP PSA n n p5 TURP Concurrent ACa n n p6 Rectal biopsy History of ACa n n p7 TURP Unknown n p n8 TURP Concurrent ACa n p p p p p9 Rectal biopsy History of ACa n

10 TURP Concurrent ACa n p p11 Prostate biopsy Unknown n n p13 TURP Unknown n n p14 TURP Concurrent ACa n n n15 TURP PSA p n n16 TURP History of ACa n p n17 RP Concurrent ACa n p p n p p18 Prostate biopsy Concurrent ACa n n19 TURP Negative cystoscopy n n n20 Bladder biopsy Concurrent ACa p p n21 TURP Concurrent ACa n p n23 TURP Unknown p p24 Bladder biopsy Concurrent ACa n n n n n n25 TURP Concurrent ACa n p n26 TURP Concurrent ACa n p n n p p28 TURP Concurrent ACa n n p p p29 TURP Concurrent ACa n n n n p n30 Bladder biopsy History of ACa n n p31 Bladder biopsy History of ACa n n p32 TURP History of ACa n n n33 TURP History of ACa n n p34 Liver History of ACa n

Abbreviations: n, negative; p, positive; RP, radical prostatectomy; TURP, transurethal resection of the prostate.

Tan et al.

Clin Cancer Res; 20(4) February 15, 2014 Clinical Cancer Research892




Tab

le2.

Assoc

iatio

nof

protein

status

,cop

y-nu

mber

status

,and

mutationstatus

forRb,PTE

N,a

ndp53

,byca

se

Cas

e#

Disse

ction

metho

dRb

IHC

RB1

copy

#RB1

seque

nce

PTEN

IHC

PTEN

copy

#PTEN

seque

nce

Mutation

type

Mutation

effect

p53

IHC

TP53

seque

nce

Mutationtype

Mutation

effect

5LC

Mn

1NA

n1

NA

pex

7:c.69

5_69

6TC

>AA(SS)

Misse

nse;

DNA

bind

ing

Predicted

deleterio

us7

LCM

n1

WT(NGS)

p2

WT(NGS)

nex

5:c.49

7C>A

(NGSon

ly)

Non

sens

eDeleterious

10MACRO

n1

WT(NGS)

p3

WT(NGS)

pex

5:c.52

4G>A

(NGS,S

S)

Misse

nse;

DNA

bind

ing

Deleterious

14MACRO

n0

WT(NGS)

n0

WT(NGS)

nin

8:c.91

9þ1G

>A(NGS,S

S)

Splicesite

brok

enDeleterious

15MACRO

p2

NA

n2

NA

nNA

17MACRO

n0

WT(NGS)

p1

WT(NGS)

pex

8:c.81

7C>T

(NGS)

Misse

nse;

Deleterious

19MACRO

n0

WT(NGS)

n2

WT(NGS)

nWT(NGS)

21MACRO

n1

WT(NGS)

p2

WT(NGS)

nWT(NGS)

24MACRO

n1

NA

n1

NA

nNA

28MACRO

n1

NA

p2

NA

pNA

31MACRO

n2

WT(NGS)

n2

ex8: c.86

3delA

(NGS)

fram

eshift;

C2do

main

predicted

deleterious

pWT(NGS,S

S)

32MACRO

n1

WT(NGS)

n1

WT(NGS)

nWT(NGS,S

S)

33MACRO

n1

WT(NGS)

n2

WT(NGS)

pex

5:c.52

4G>A

(NGS,S

S)

Misse

nse;

DNA

bind

ing

Deleterious

Abbreviations

:LC

M,lase

rca

pturemicrodisse

ction;

MACRO,mac

rodisse

ction;

n,ne

gativ

e;NA,no

tas

sessed

;NGS,AmpliS

eqHotsp

otne

xt-gen

erationse

que

ncingas

say;

p,p

ositive

;SS,S

ange

rse

que

ncing;

WT,

wild

-typ

e.






an internal control inmost cases and a casewas excluded if itlacked this endothelial staining.

The immunohistochemical protocol and interpretation ofstaining for cytoplasmic PTEN protein detection were exten-sively validatedusingmultiple genetic controls (cell lines andtissues) as described previously (31). A case was consideredto have lost PTENprotein if the intensity of staining in tumorcell cytoplasm was markedly decreased or entirely negativeacross all tumor cells comparedwith the surrounding benignglands and/or stroma in any given TMA spot. A given spotwas dropped from the analysis if these benign areas lackedPTEN staining (internal positive control).

Accumulation of p53 protein is frequently associatedwith deleterious mutations in this tumor suppressor gene,which extend the protein’s half-life. The immunohisto-chemical staining protocol for p53 was validated usingprostate cancer cell lines with known TP53 mutations orlines that are known to be null for TP53 (32). Using thisprotocol, DU145 cells that have two TP53 mutations stainpositively, and PC3 cells that are TP53-null and stainnegatively. Although there is no strict cutoff for the amountof p53 immunostaining that provides definitive evidence ofan underlying mutation in TP53, a number of studies,predominantly in bladder cancer, used 10% as a cutoff. Inthe current study, we used an even more stringent cutoff:positive p53 overexpression was scored if any spot in a caseshowed strong expression (3–4þ) in >50% of tumor cells.

DNA preparationFor 13 of the 29 prostatic small cell carcinoma samples,

there was adequate tissue available for DNA preparation.Between one and three 10-mm sections of formalin-fixedparaffin-embedded (FFPE) tumor tissue were deparaffi-nized and briefly rehydrated. Eleven tumor samples (11/13) were enriched for tumor bymacrodissection using a 26-gauge needle under stereomicroscopic visualization guidedby a contiguous hematoxylin and eosin (H&E) section. Twotumor samples (2/13) were subjected to laser capturemicrodissection using a Leica P.A.L.M. laser microdissec-tion system (Leica Microsystems) according to the manu-facturer’s directions. All tumor samples were at least 80%pure by visual estimation using a contiguous H&E section.For each sample, genomic DNA was extracted by using theQIAamp DNA Micro Kit (Qiagen) following the manufac-turer’s directions.

Copy-number alteration analysisOf note, 300 to 500 ng of genomic DNA from each small

cell carcinoma sample was used according to the manufac-turer’s directions for NanoString nCounter Cancer CopyNumber Assay (NanoString Technologies). This assay, opti-mized for DNA prepared from FFPE tissue samples, detectscopy-number changes in 86 genes that are commonlyamplified or deleted in cancer using 255 gene-specificprobes and an additional 54 probes for invariant regionsof the genome. DNA from five additional samples of met-astatic acinar prostatic carcinoma (for which copy numberalterations have previously been reported by high density

SNP microarray) was also used for a validation set. Dataanalysis was performed using the nSolver Analysis Software(NanoString). Raw counts were first normalized to theaverage counts across probes to invariant regions to nor-malize for any differences in DNA input. The copy numberfor each probe was then normalized to the reference sample(Human blood genomic DNA; Roche) and the averagecopy-number call reported for each gene with a standarddeviation (0–0.5¼ 0 copies; 0.6–1.4¼ 1 copy, 1.6–2.4¼ 2copies; etc.). Probes for TP53 were unintentionally omittedfrom the version 1 NanoString assay, thus TP53 copynumber was not available for the present analyses.

Next-generation sequencingIon Torrent Ion AmpliSeq Cancer Hotspot Panel v2 (Life

Technologies) was used for sequencing hotspot regions in50 frequently mutated tumor suppressor and oncogenes,covering approximately 2,800 COSMIC mutations in total.For this article, we focused on exonic regions of RB1 (thepanel included 10 amplicons covering roughly 10 out of 27exons: exon 4, 6, 10, 11, 14, 17, 18, 20, 21, and 22), TP53 (8amplicons covering roughly 7 out of 11 exons: exon 2, 4, 5,6, 7, 8, and 10), and PTEN (8 amplicons covering roughly 6out of 9 exons: exon 1, 3, 5, 6, 7, and 8). Genomic DNA (10ng input) from 13 out of the 30 prostatic small cell carci-noma cases was selected for the assay and Ion 318 chipswere used for greater sample multiplexing. Ion TorrentVariant Caller, Ion Reporter, and Integrated Genomic View-er (IGV; Broad Institute) were used for analyzing targetmutations. Variant frequency was determined by the ratioof variant allele reads to total reads.

Sanger sequencingDirect sequencing for exons 4, 5, 7, and 8 of TP53

was performed on PCR-amplified DNA from 13 paraffin-embedded prostatic small cell carcinoma samples usingestablished primers and protocols available at the IARC(InternationalAgency forResearchonCancer) p53Databasewebsite (http://p53.iarc.fr/Download/TP53_DirectSequen-cing_IARC.pdf). Both forward and reverse strands of thePCR product were subjected to sequencing. Sequence var-iants were verified on both strands as well as by sequencingan independent PCR reaction from the same sample.Sequence analysis was conducted using Mutation Surveyor(SoftGenetics). The predicted effect of the mutation (dele-terious, nondeleterious)wasdeterminedusing the IARCp53Database (http://p53.iarc.fr/TP53GeneVariations.aspx).

ResultsValidation of Rb immunohistochemistry assay

Immunohistochemical staining for Rb protein was vali-dated by blindly scoring Rb protein status on a TMA con-structed from 72 FFPE cell lines, predominantly from theNCI-60 panel (31). Cell lines and samples were scored asnegative for Rbprotein if >95%of cells in all spots lackedRb(0þ staining). In all, 5 (7%) cell lines completely lacked Rbprotein by immunohistochemistry (IHC; Fig. 1). These

Tan et al.





Figure 1. A, immunostaining for Rbwas validated using aTMAmade from theNCI-60cell linepanel. PC3cells showhighRbexpressionwithwild-typeRB1. SF-539 cells are negative for Rb protein and are known to have a 4 base-pair deletion resulting in a frameshift mutation. SNB-75 cells are not known to have ahomozygous deletion in RB1; however, numerous previous studies have shown absence of the Rb protein by Western blotting in this cell line. DU-145cells haveahomozygousdeletion resulting in a truncationof theprotein at exon21andshowvery lowRbstainingby IHC.B, representativeRb immunostainingresults from case 7 show negative staining in tumor cells with retained endothelial staining (arrow) as an internal positive control. Case 20 showsstrongly positive staining. Case 8 shows negative staining in the small cell carcinoma (SCC) component and positive staining in the adjacent acinar carcinomacomponent (ACa). Images in this panelwereassembled fromamosaic of higher-power images.C, representative positiveRb immunostaining results in ahigh-grade acinar carcinoma unassociated with small cell carcinoma. D, representative negative Rb immunostaining in a high-grade acinar carcinomaunassociated with small cell carcinoma. Endothelial cells (arrow) provide an internal positive control. E, graphical representation of percentage of caseswith Rb protein loss comparing small cell carcinomas, acinar carcinomas associated with small cell carcinoma, and high-grade acinar carcinomasunassociated with small cell carcinoma. Rb loss is most frequent in small cell carcinomas and rarely seen in high-grade acinar carcinomas unassociated withsmall cell carcinoma, whereas acinar carcinomas occurring concurrently with small cell carcinoma show an intermediate rate of Rb protein loss. Acinarcarcinomas with neuroendocrine differentiation and metastatic castrate-resistant prostate carcinomas rarely show Rb loss.






included the only two cell lines on the TMA known to havecomplete homozygousdeletions involvingRB1: BT549 cellshave a homozygous 343 base-pair deletion in the gene, andSF-539 cells have a 4 base-pair deletion, resulting in aframeshift mutation (33). Of the three other cell lines thatlacked Rbprotein by immunohistochemical assay (SNB-75,OVCAR-8, and NCI-adr-Res), none have known mutationsin theRB1 gene; however, all have previously been shown tobe negative for Rb protein by Western blotting or reversephase protein arrays (33–35). In addition, seven cell linesshowed reduced, but not absent, immunohistochemicalstaining for Rb protein (CAKI-1, DU-145, HCC-2998,Hep3B, MDA-PCA2b, NCI-H522, PrSc, and RWPE cells)defined as 1þ weak staining in >5% of cells. Of these, DU-145 cells are known tohavehomozygous deletion, resultingin a truncation of the protein at exon 21, and HCC-2998

cells have a heterozygous nonsense mutation in the gene(32, 33).

Rbprotein expression in small cell carcinomaandhigh-grade acinar carcinoma cases

Ninety percent (26 of 29) of prostatic small cell carcino-ma cases showed Rb protein loss, compared with only 43%(3 of 7) of concurrent acinar carcinoma cases (Figs. 1and 2, Table 1). In cases with concurrent small cell andacinar carcinoma components, 57% (4 of 7) showed con-cordance of the small cell carcinoma and acinar carcinomacomponents for Rb status, whereas 43% (3 of 7) showedpresence of Rb protein in the acinar carcinoma componentwith loss of Rb in the small cell carcinoma component.Interestingly,where present, the acinar component typicallyshowed weak or reduced staining similar to DU-145 cells

Figure 2. Immunohistochemicalassessment of PTEN and p53tumor suppressor loss in small cellcarcinomas and their associatedacinar carcinoma component. A,case 24 shows PTEN loss in boththe small cell and associated acinarcarcinoma, with retainedendothelial/stromal staining forPTEN providing an internal positivecontrol. Case 29 shows p53accumulation in the small cellcomponent only. B, IHC for Rb,PTEN, and p53 in case 5demonstrates loss of Rb andPTEN,with overexpression for p53.C, copy number by NanoStringnCounter Cancer Copy NumberVariation Panel across 86 probedgenes for case 5. c-MYCamplification (5 copies), andhemizygous PTEN and RB1 lossare present. D, Sanger sequencingchromatogram for TP53 fromcase 5 shows a 2-base pairmissense mutation (exon 7:c.695_696TC>AA) predicted todeleteriously affect the DNA-binding domain.

Tan et al.





(Fig. 1).Cases inwhich the acinar component had loss of Rbprotein and the small cell carcinoma component retainedthe protein were not identified.To assess whether the very high frequency of Rb protein

loss is unique to small cell differentiation in prostate cancer,we assessed Rb expression in a spectrum of high-gradeprimary acinar carcinomas unassociated with small cellcarcinoma, primary carcinomas with neuroendocrine dif-ferentiation, and metastatic castrate-resistant acinar carci-nomas. First, we scored Rb expression in an additional 150acinar carcinoma cases enriched for high-grade disease andoccurring without concurrent small cell carcinoma thathave been described elsewhere (27–29). Gleason score wasavailable in 108 (72%)of these cases: 11%(n¼12)Gleason6, 26%(n¼28)Gleason 7, 32%(n¼35)Gleason 8, 29%(n¼ 31) Gleason 9, and 2% (n ¼ 2) Gleason 10. Of theseacinar tumors, only 7% (10 of 150) showed Rb protein loss(Fig. 1).In addition, we examined primary non–small cell acinar

carcinomas with evidence of neuroendocrine differentia-tion. Because the pathologic classification of non–small cellneuroendocrine differentiation in prostate carcinoma wasthe topic of a recent consensus meeting and is currentlyundergoing revision (Mahul Amin, Mark Rubin, HimishaBeltran, Tamara Lotan, Juan-Miguel Mosquera, Victor Reu-ter, Brian Robinson, Patricia Troncoso, and JonathanEpstein, in preparation), we used several definitions forthis entity. First, we examined the frequency of Rb proteinloss in primary prostate tumors diagnosed as "high-gradeprostatic adenocarcinoma with neuroendocrine differenti-ation."Morphologically, these cases fell short of a diagnosisof small cell carcinoma, but showed focal architectural orcytologic features suggestive of neuroendocrine differenti-ation, combined with nonfocal (generally >20%) expres-sion of either chromogranin or synaptophysin. Of thesecases, 22% (2 of 9) showed Rb protein loss (SupplementaryFig. S1). Next, we examined cases of large cell neuroendo-crine prostate carcinoma, characterized by large nests oftumor cells with peripheral palisading and often geographicnecrosis, in which the cytology is that of non–small cellcarcinoma (prominent nucleoli, vesicular clumpy chroma-tin and/or large cell size, and abundant cytoplasm). Thesetumors generally express at least one neuroendocrinemark-er and frequently lack prostatic/AR axis signaling markers.Of these rare cases, 17% (1 of 6) showed Rb protein loss(Supplementary Fig. S1).Finally, we examined Rb protein status in usual prostatic

adenocarcinomas with neuroendocrine differentiationestablished solely by nonfocal immunohistochemicalexpression of chromogranin and/or synaptophysin.Because as many as 100% of prostate acinar carcinomascan show focal expression of neuroendocrine markers, wearbitrarily defined this category as cases in which >20% ofthe cells expressed chromogranin and/or synaptophysin.Ofthe 170 high-grade acinar prostatic adenocarcinomasassayed for nonfocal chromogranin and synaptophysinpositivity, 12% (20 of 170) met the criterion of neuroen-docrine marker staining in >20% of the cells in at least one

TMA spot. Of these cases, only 5% (1 of 20) showed Rbprotein loss.

To determine whether Rb protein loss is specific to smallcell carcinoma or simply a result of progression of disease,we also examined Rb expression in a series of metastaticcastrate-resistant acinar prostatic carcinoma samples. Mul-tiple metastases from these patients have been previouslycharacterized for copy-number alterations at the RB1 locususing high-density SNPmicroarray (30). Importantly, how-ever, for Rb IHC studies, we did not have access to theidentical tissue specimen characterized by SNP microarray,although in 7 cases (54%), we were able to analyze tissuesfrom the same metastatic site as had been characterized bySNPmicroarray. Overall, 15% (2 of 13) of cases showed Rbprotein loss (Supplementary Table S1). By SNPmicroarray,23% (3 of 13) of the castrate-resistant cases had homozy-gous loss of RB1, although in two out of three of these cases,homozygous RB1 loss was heterogeneous among the fivemetastatic sites examined, strongly suggesting that completeloss of RB1 is a late event in metastatic progression. Con-cordant with this, Rb protein was lost only in the single casewith homozygous deletion in all the four metastases exam-ined, and was actually only focally lost in this metastasis,suggesting that both inter- and intrametastatic heterogene-ity in Rb status is common in advanced prostate cancer(Supplementary Fig. S1). An additional 38% (5of 13) of thecastrate-resistant metastases had hemizygous deletion ofRB1. Of these cases, 100% (5 of 5) had hemizygous loss inall metastatic sites examined by SNP array and 20% (1 of 5)had Rb protein loss, albeit in a metastasis from a separatesite as the ones characterized by SNP microarray. Theremaining five cases (38%) were copy-number neutral forRB1, including two cases that had deletion of one allele,with amplification of the remaining allele (loss of hetero-zygosity). Of these, all were positive for Rb protein.

PTEN and p53 protein expression in small cellcarcinoma cases

Sixty-three percent (17 of 27) of small cell carcinomacases and 71% (5 of 7) of concurrent acinar carcinoma casesdisplayed PTEN protein loss (Fig. 2, Table 1). Of note, 86%(6 of 7) of the cases showed concordance in small cellcarcinoma and acinar carcinoma components for PTENstatus. One case showed retained PTEN protein in thesmall-cell component and loss in the acinar component.Fifty-six percent (14 of 25) of small cell carcinoma casesand 66% (4 of 6) of concurrent acinar carcinoma casesshowedp53 accumulation (Fig. 2, Table 1). Sixty-six percent(4 of 6) of cases showed concordance in the small cellcarcinoma and acinar carcinoma components for p53 pro-tein status.

Assessment of RB1 and PTEN copy-number alterationTo determine the molecular mechanism of the high rate

of Rb and PTEN protein loss among small cell carcinomacases, we next assessed RB1 and PTEN gene copy number in13 of the 29 small cell carcinoma cases with adequate tissueavailable for DNA isolation. To validate the NanoString






nCounter Cancer Copy Number Assay for this purpose, wefirst tested the assay on the DNA purified from five previ-ously described samples of metastatic prostatic adenocar-cinoma for which copy-number alterations across thegenome have been reported by high-density SNP micro-array (AffymetrixGenomeWideHumanSNPArray 6.0; 30).For four of these samples, the correlation between the PTENcopy number as assessed by NanoString assay and theAffymetrix SNP array was high (R2 ¼ 0.85; SupplementaryFig. S2). For the fifth sample, a partial homozygous deletionencompassing exons 2 to 9 of the gene was detected by bothassays, with only themost 50 of the threeNanoString probesto PTEN hybridizing to this sample (data not shown).Similarly, at the RB1 locus, the copy-number data from theNanoString Copy Number Variation assay and the Affyme-trix SNP array was highly correlated across all the fivesamples (R2 ¼ 0.92).

Of the prostatic small cell carcinoma samples, RB1 allelicloss was detected in 85% (11 of 13), of which 73% (8 of 11)showedhemizygous loss and 27%(3of 11) of cases showedhomozygous loss at RB1 (Table 2). Of the cases with RB1allelic loss, 100% (11 of 11) showed Rb protein loss by IHC(Table 3 and Fig. 2). Of the cases without RB1 loss, 50% (1of 2) showed Rb protein loss. At the PTEN locus, 38% (5 of13) of prostatic small cell carcinoma cases showed allelicloss, with 80% (4 of 5) showing hemizygous loss and 20%(1 of 5) showing homozygous loss. Of the cases with allelicloss, 80% (4 of 5) showed PTEN protein loss, whereas only50% (4 of 8) of cases without PTEN allelic loss showedPTEN protein loss (Table 4).

Sequencing of TP53, PTEN, and RB1Slightly more than half of the small cell carcinoma cases

showed accumulation of p53 protein, indicating the pos-sible presence of an inactivatingmutation in TP53. To assessfor TP53 mutations in a subset of 10 small cell carcinomacases with adequate DNA available for analysis, we used a

combination of next-generation sequencing via the Ampli-Seq Cancer Hotspot Panel on the Ion Torrent (whichincludes 8 amplicons covering the most commonly mutat-ed exons in the gene: exons 2, 4, 5, 6, 7, 8, and 10) and directSanger sequencing of exons 4, 5, 7, and 8. Because of limitedDNA availability for some samples, six samples had bothAmpliSeq and Sanger sequencing data available, whereasthree samples had only AmpliSeq data available and onesample had only Sanger sequencing data available. Overall,60% (6 of 10) of the small cell carcinoma samples showedmutations in TP53 (Table 2), including one deleteriousmissense mutation in the DNA-binding domain in exon5 (c.524G>A) that was seen in two separate samples andhasbeen reported in multiple other tumor types and severalfamilies with Li-Fraumeni syndrome; one nonsense muta-tion in exon 5 (c.497C>A) that has been reported in anumber of other tumors as well as tumor cell lines; adeleterious splice site mutation in intron 8 (c.919þ1G>A),and a missense mutation in exon 8 (c.817C>T), both ofwhich have been reported in other tumors and in familieswith Li-Fraumeni syndrome. In addition, we found a novelmissense mutation in exon 7 (c.695_696TC>AA) predictedto deleteriously affect the DNA-binding domain. Of the sixsamples subjected to both next generation and Sangersequencing, five (83%) showed identical results by bothassays, whereas one (17%) showed an additional mutationdetected by next-generation sequencing that was notdetected in the Sanger sequence of the same sample (thenonsense mutation in exon 5 of sample 7: c.497C>A; Table2). Of the small cell carcinoma samples with a TP53 muta-tion detected, 66% (4 of 6) showed accumulation of p53protein, compared with 25% (1 of 4) of the cases with wild-type TP53 (Table 5).

As part of the AmpliSeq Cancer Hotspot Panel, we alsoexamined the sequence of six of the nine exons in PTEN thataremost frequentlymutated (exons 1, 3, 5, 6, 7, and 8). Outof nine small cell carcinoma samples with available DNA,we foundonly one frameshiftmutation in theC2domainofexon 8 of PTEN (c.863delA) in a single sample (Table 2).This mutation has been previously reported in severalendometrial carcinoma samples as well as in a colon car-cinoma sample; however, the functional significance isunclear (36, 37). Interestingly, this sample had no detect-able PTEN protein by immunohistochemical assay, despitetwo intact copies of the PTEN gene by NanoString analysis.Finally, we examined the sequence of 10 out of 26 exonsof RB1 (exons 4, 6, 10, 11, 14, 17, 18, 20, 21, and 22), and

Table 3. Association between Rb protein levelsby IHC and RB1 copy number by NanoString

Rb IHC

RB1 copy # Present Absent

Normal 1 1LOH 0 11

Table 4. Association between PTEN proteinlevels by IHC and PTEN copy number byNanoString

PTEN IHC

PTEN copy # Present Absent

Normal 4 4LOH 1 4

Table 5. Associationbetweenp53protein levelsby IHC and TP53 copy number by NanoString

p53 IHC

TP53 status Absent Present

WT 3 1Mut 2 4

Tan et al.





did not find any mutations in the nine samples analyzed(Table 2).

DiscussionAlthough de novo small cell carcinoma of the prostate is a

rare disease, t-NEPC is expected to become increasinglycommon with the recent availability of more potent anti-androgen therapeutics for prostate cancer treatment (aber-aterone, enzalutamide, and TAK700; ref. 13). Because bothde novo small cell carcinomas and t-NEPCs are resistant toandrogen deprivation therapies and transiently responsiveto platinum-based chemotherapeutics (2, 6–11), accuratediagnosis of these entities and distinction from ordinaryhigh-grade acinar prostate cancer is essential. Currently, themainstay of pathologic diagnosis of small cell carcinoma ismorphology (1, 18). Small cell carcinomas typically possessthe classic "oat-cell" morphology first described for theircounterparts in the lung, accompanied by a high mitoticindex and frequent apoptosis. However, diagnosis by H&Estain can be challenging and given the clinical conse-quences, use of an immunohistochemical panel to identifysmall cell carcinoma differentiation is frequently helpful.Unfortunately, neuroendocrine markers such as synapto-physin, chromogranin, and CD56 do not comprise anoptimal marker panel for small-cell differentiation in pros-tate cancer, as they can be seen in up to 100% of acinarcarcinomas in some series (38). Other markers that areuseful in clinical practice include absence of PSA or otherprostatic marker positivity in approximately 80% of thecases (due to absence of AR expression), high ki-67 labelingindex, and TTF-1 expression (18, 39, 40). Recently, CD44was also proposed as a potential marker (41). However,given the rising prevalence of the disease and the lack ofsensitivity and specificity of currently used IHC panels,identifying additional markers of small cell carcinoma andt-NEPC currently represents an important area of unmetclinical need.A number of studies in human samples as well as trans-

genicmousemodels have suggested that loss of theRB1 andTP53 tumor suppressorsmaybe critical for the developmentof neuroendocrine carcinoma in multiple organ systems.Human lung small cell carcinomas were first shown to havefrequent TP53 mutations as well as nearly universal loss ofRB1 (19, 20). Subsequent studies inmouse models showedthat germline RB1 heterozygosity is associated with thedevelopment of neuroendocrine tumors in multiple organsystems (42). Furthermore, conditional RB1 and TP53 lossare sufficient to cause highly penetrant small cell carcinomain the mouse lung (21). Recent studies have suggested thatthe cell of origin is likely neuroendocrine in this tumormodel (43). Similarly, in theprostate, activationof the SV40large T-cell antigen (which inactivates all Rb family proteinsand p53) has long been known to result in aggressivetumors with neuroendocrine features (22). More recentstudies have shown that although conditional RB1 lossalone is not sufficient to cause invasive prostatic carcinomain the mouse (44, 45), concurrent loss of RB1 and TP53

leads to highly metastatic neuroendocrine carcinomas witha small cell-like morphology, likely originating from theproximal prostatic ducts (23, 46).

Despite the wealth of data for human lung and murineprostate tumors, ours is the first study to comprehensivelyevaluate the status of the RB1, TP53, and PTEN tumorsuppressors in a large series of human prostatic small cellcarcinomas. This is likely in part because these specimensare quite rare and almost uniformly FFPE. Most small cellcarcinomas are first discovered in biopsies or TURP speci-mens (due to the aggressive nature of the disease, patientscommonly present with urinary retention), and radicalprostatectomy is contraindicated in these patients, makingit difficult to collect frozen tissue specimens for analysis.However, recent improvements in genomic analysis fromFFPE specimens have expanded our ability to cull genomicinformation from these small samples.

In the current studies, we took advantage of both Nano-String and targeted next-generation sequencing technolo-gies to examine the status of theRB1,TP53, andPTEN tumorsuppressors.We found that Rb loss occurs nearly universallyat the protein level in small cell carcinomas, suggesting thatabsence of this protein may provide a useful marker ofsmall-cell differentiation in the prostate. In contrast, inhigh-grade primary or metastatic acinar carcinomas, eventhosewith immunohistochemical ormorphologic evidenceof (non–small cell) neuroendocrine differentiation, Rbprotein loss occurred in only a minority of cases. Becausenegative markers always run the risk of producing falsenegative results due to technical failures, and because it isnot 100% sensitive for detecting small cell carcinomas, RbIHC will likely be most useful as part of a larger panel ofimmunohistochemical markers for neuroendocrine differ-entiation, including synaptophysin, chromogranin, andmarkers of the AR signaling axis. As such, this panel ofimmunohistochemical markers would provide additionalverification of small-cell differentiation in histologicallyindefinite cases. In addition, it is tempting to speculate thatinclusion of Rb in this panel may also help to identify aclinically important subgroup of high-grade metastaticprostate tumors that show clinical resemblance to smallcell carcinomas (including response to platinum-basedchemotherapeutics), but lack apparent small cell or, insome cases, even neuroendocrine differentiation by con-ventional markers (47). Although studies are currentlyunderway to test this hypothesis, the results of our currentwork suggest the possibility that Rb loss in the context of anotherwise unremarkable non–small cell adenocarcinomamay presage the later development of small cell carcinomaor castrate-resistant prostate carcinoma with neuroendo-crine differentiation, thus, serving as apredictive biomarker.

Our finding that Rb loss is common in prostatic small cellcarcinomas is in agreement with a recent study that exam-ined prostatic xenografts with varying degrees of neuroen-docrine differentiation (48). In this study, Rb protein losswas almost invariably seen in xenografts with evidence ofhigh-grade neuroendocrine differentiation, and microdele-tions at the RB1 locus were detected by microarray-based






comparative genomic hybridization (aCGH) in many ofthese specimens. In 14 additional small cell carcinomas ofthe prostate, Rb protein levels were also markedlydecreased. Adding to this study, we found that in contrastwith small cell carcinomas, Rb loss at the protein leveloccurs in only 10% of high-risk acinar primary carcinomas,despite the fact thatRB1 allelic loss occurs in 18% to 40%ofthese cases (24–26, 49). Even in prostatic primaries withevidence of neuroendocrine differentiation (based onmor-phologic features and/or immunohistochemical expressionof neuroendocrine markers), Rb protein was only lost in11% of cases, overall. In part, this is likely because homo-zygous deletion is extremely uncommon in the setting ofacinar primary tumors. In a recent study of high-risk pri-mary tumors, only 1.6% or 2 of 125 of primary acinar casesshowed homozygous deletion compared with 41.6% or 52of 125 cases with hemizygous deletion (ref. 49 and Wen-nuan Liu, unpublished data).

Indeed, complete inactivation of RB1 generally occursquite late in acinar tumor progression. In castrate-resistantprostate tumors, the gene expression signature for Rb func-tional loss is relatively enriched, compared with that inprimary tumors (50) and up to 60% show RB1 hemizygousdeletion (Supplementary Fig. S1; refs. 24, 26, 30). However,the rate of homozygous RB1 deletion in castrate-resistanttumors does not exceed 20% in the current series (Supple-mentary Fig. S1; refs. 24, 26, 30) and we confirmed that Rbprotein loss is seen in a similar minority of cases. Interest-ingly, in our series, of the cases with homozygous RB1 loss,only 33% (1 of 3) had homozygous loss documented inmore than one of the multiple metastatic sites sampledwithin the case. This provides further support for theconcept that homozygous deletion of RB1 is a relativelylate-stage genomic alteration in acinar prostate cancerprogression.

Although Rb protein loss is far more common in smallcell carcinomas than acinar castrate-resistant prostatetumors, the mechanism of Rb protein loss in small cellcarcinomas remains unclear. Despite complete loss of theprotein in almost every case, homozygous loss of RB1 wasalso relatively rare in our series of small cell carcinomas,suggesting that alternative mechanisms of RB1 loss may becommon in this tumor type. Importantly, the relativelylarge amounts of DNA required for the NanoString assay(500 ng) preclude microdissection of tumor samples foranalysis, thus we cannot entirely rule out the possibility thatcontamination by surrounding benign glands and stromaconfounded the copy-number call by the NanoString assay.However, all tumor samples were estimated to be >80%pure by macrodissection and complete loss of Rb proteinexpression was also frequently seen with low transcriptlevels despite the presence of one intact RB1 allele by CGHin the LuCaP castrate-resistant prostate xenograft series(50). Another possible explanation for the apparent dis-crepancybetweenprotein and copy-number assays for Rb inour study and others could be that the second allele isfrequently inactivated by mutation in prostate tumors. Yet,in primary and metastatic acinar prostate tumors, RB1 is

inactivated by mutation in less than 5% of cases in mostseries (24, 25). In agreement with these data, we performedlimited sequencing of RB1 using the AmpliSeq CancerHotSpot Panel and did not find any mutations in the genein small cell carcinomas. However, it is important to notethat this panel covers only roughly 10 out of 27 exons fromthis large gene (exons 4, 6, 10, 11, 14, 17, 18, 20, 21, and 22)and it might be possible that we missed significant muta-tions in the second allele because of this limitation. Anotheralternative mechanism for RB1 inactivation could be epi-genetic gene silencing. Although differential deoxycytidinemethylation within CpG dinucleotides near the regulatoryregions of the RB1 locus has been reported previously incastration-resistant prostate tumors, a recent study of neu-roendocrine xenografts did not confirm this finding(48, 51). Thus, it remains unclear how the second allele ofRB1 is inactivated in small cell carcinomas.

In addition to RB1 loss, we also found that TP53 loss iscommon in small cell carcinomas, with protein accumula-tion and underlying detrimental mutation of the geneoccurring in well over half of all specimens examined.Overall, our IHC results for p53 protein overexpressioncorrelated quite well with themutational status of the gene,aswouldbe expected, given thatTP53mutations commonlyextend the half-life of the protein. In some cases, TP53mutations can also decrease the expression of the protein(52), and this may account for a few cases in which weobserved a mutation in the absence of protein accumula-tion. In the only previous study of p53 in a series of prostaticsmall cell carcinomas, Chen and colleagues found that p53protein was overexpressed in 74% (23 of 31) of prostaticsmall cell carcinomas (53). Interestingly, on sequencing, 5of 7 of these tumors showed the same novel G747A mis-sense mutation in the gene. Similarly, a case study of smallcell carcinoma and associated acinar carcinoma showed anidentical TP53mutation in both components, supporting acommonorigin for the two tumors (17).TP53mutation canbe seen in acinar prostate carcinomas, but it is less frequent-ly present in primary tumors, occurring in less than 10% ofspecimens examined in recent reports (24–26).However, incastrate-resistant tumors, it is substantially more commonand canbe seen in 30% to 40%of themetastases (24). Thus,p53 accumulation at the protein level or evidence of TP53mutations are not specific to small cell carcinoma, do notcorrelate with AR or neuroendocrinemarker IHC and occurrelatively commonly in advanced prostate cancer. Accord-ingly, it is unclear whether evidence of p53 inactivation orPTEN loss adds much information about potential smallcell differentiation compared with Rb loss, which is seennearly universally in this tumor type. Future studies willspecifically examine whether p53 and PTEN are usefuladjunct markers in combination with Rb.

A high proliferative rate is a key feature of small cellcarcinoma in any organ, thus it may not be surprising thatloss of critical cell-cycle regulators such as RB1 and TP53 iscommon in this tumor type in multiple organ systems.Active, hypophosphorylated Rb binds to and inhibits thefunction of E2F family transcription factors that promote S-

Tan et al.





phase entry. p53 similarly inhibits the G1–S cell-cycle tran-sition upon DNA damage recognition. Along these lines,recent studies have suggested that other cell-cycle regulators,suchasAuroraAkinase andUBE2Care frequently amplifiedor overexpressed in small cell neuroendocrine tumors(16, 48). Of note, however, we did not find evidence ofAurora kinase A amplifications in our samples (n ¼ 13) byNanoString assay (data not shown). It is also important topoint out that Rb protein can be inactivated by a number ofnongenomicmeans, such as hyperphosphorylation follow-ing overexpression of cyclin D1, CDK4/6 or loss of p16,events that occur commonly in other tumor types (54).Although we could not reliably examine Rb phosphoryla-tion status in our FFPE samples, hyperphosphorylation ofRbmost commonly results in overexpression of the protein,a finding we did not see in the small cell carcinoma cases(55). Accordingly, we did not see evidence of high levelCDK4/6 or cyclin D1 amplification in our small cell carci-noma cases by the NanoString assay, with only a fewsamples showing three copies of these genes (data notshown). This finding was not mutually exclusive withhomozygous RB1 loss, suggesting it is of unclear signifi-cance (data not shown). In addition, recent reports suggestthat cyclinD1protein levels are low, rather thanhigh, in thistumor type (48). We did find hemizygous copy loss ofCDKN2A (p16) in four small cell carcinoma samples, noneof which showed homozygous RB1 loss; however, all ofthese cases were negative for Rb protein (data not shown),again making the significance of this finding unclear. Over-all, our data suggest that Rb protein loss is overwhelminglythe preferential mechanism of Rb inactivation in thesetumors.The extraordinarily high frequency of Rb loss in prostatic

small cell carcinomas suggests that loss of this tumorsuppressor, perhaps in combination with inactivation ofp53, may be an essential event in the development of thistumor type. However, other than increasing the prolifer-ative potential of the tumor cells, it remains unclear whatsignals downstream of Rb and p53 loss mediate small celldifferentiation. Because Rb protein loss occurs in the over-whelming majority of small cell carcinomas and is rarelyseen in primary acinar carcinomas unassociated with asmall-cell tumor, it is tempting to speculate that Rb lossitself may be a critical mediator of neuroendocrine trans-differentiation, as has been suggested by some studies in thelung (56). However, it remains equally plausible that Rb issimply required to prevent the proliferative expansion ofneuroendocrine cells in the prostate, and its functions inthis capacity are less critical for nonneuroendocrine epithe-lial cells. If Rb loss is a prerequisite for the development ofsmall cell carcinoma that is frequently AR-negative, this also

raises the interesting question of how Rb function mayinterface with AR expression. Interestingly, studies per-formed in usual-type prostatic adenocarcinomas withoutsmall cell differentiation have supported the idea that Rblossmay actually increaseAR levels andARaxis activity (50).The fact that small cell carcinomas are most commonly AR-and Rb negative strongly suggests that the effects of Rb lossmay be highly context dependent, with differing outcomesin tumors with and without small cell neuroendocrinedifferentiation, as has been suggested by others (54).

In sum, our data suggest that loss of the RB1 and TP53tumor suppressors is common in prostatic small cell carci-nomas, similar to their counterparts in the lung and otherorgans. Overall, genetic inactivation of these loci by allelicloss or mutation, respectively, was well correlated withvalidated immunohistochemical assays for these tumorsuppressor proteins. Indeed, we found total loss of Rb inthe vast majority of prostatic small cell carcinomas, withonly rare loss in high-grade acinar carcinoma primarytumors, suggesting that as part of a panel, Rb protein lossis a potentially clinically useful marker of small-cell differ-entiation in the prostate.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

Authors' ContributionsConception and design: J.I. Epstein, G.J. Netto, A.M. De Marzo, T.L. LotanDevelopmentofmethodology:H.-L. Tan, A. Sood, J.Hicks, A.M.DeMarzo,T.L. LotanAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): A. Sood, H.A. Rahimi, W. Wang, N. Gupta, S.Mosier, C.D. Gocke, G.J. Netto, W. Liu, W.B. Isaacs, T.L. LotanAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis): H.-L. Tan, A. Sood, H.A. Rahimi, S.Mosier, C.D. Gocke, J.I. Epstein, W. Liu, T.L. LotanWriting, review, and/or revision of the manuscript: H.-L. Tan, A. Sood,W. Wang, C.D. Gocke, J.I. Epstein, G.J. Netto, W. Liu, A.M. De Marzo, T.L.LotanAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): A. Sood, W. Wang, J. Hicks, S.Mosier, T.L. LotanStudy supervision: J.I. Epstein, T.L. Lotan

AcknowledgmentsThe authors thank Jianfeng Xu for SNP array copy-number data used to

validate the NanoString assay.

Grant SupportThis research was supported in part by a Prostate Cancer Foundation

Young Investigator Award anda grant by thePatrickC.Walsh ProstateCancerResearch Fund (to T.L. Lotan).

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 18, 2013; revised October 16, 2013; accepted November 15,2013; published OnlineFirst December 9, 2013.

References1. Epstein JI, Netto GN. Biopsy interpretation of the prostate. 2nd ed.

Philadelphia, PA: Lippincott, Williams & Wilkins; 2008.2. Palmgren JS, Karavadia SS, Wakefield MR. Unusual and underappre-

ciated: small cell carcinoma of the prostate. Semin Oncol 2007;34:22–9.

3. Mucci NR, Akdas G, Manely S, RubinMA. Neuroendocrine expressionin metastatic prostate cancer: evaluation of high throughput tissuemicroarrays to detect heterogeneous protein expression. Hum Pathol2000;31:406–14.






4. Zhang X, Coleman I, Coleman R, Doan K, Roudier M, Ch�ery L, et al.Characterizing the molecular features of the neuroendocrine pheno-type in castration resistant prostate cancer [abstract]. In: Proceedingsof the 104th Annual Meeting of the American Association for CancerResearch; 2013 Apr 6-10; Washington, DC. Philadelphia (PA): AACR;2013.Abstract nr 406.

5. Shah RB, Mehra R, Chinnaiyan AM, Shen R, Ghosh D, Zhou M, et al.Androgen-independent prostate cancer is a heterogeneous group ofdiseases: lessons from a rapid autopsy program. Cancer Res 2004;64:9209–16.

6. Hindson DA, Knight LL, Ocker JM. Small-cell carcinoma of prostate.transient complete remission with chemotherapy. Urology 1985;26:182–4.

7. Amato RJ, Logothetis CJ, Hallinan R, Ro JY, Sella A, Dexeus FH.Chemotherapy for small cell carcinoma of prostatic origin. J Urol1992;147(3 Pt 2):935–7.

8. Moore SR, Reinberg Y, Zhang G. Small cell carcinoma of prostate:effectiveness of hormonal versus chemotherapy. Urology 1992;39:411–6.

9. Rubenstein JH, Katin MJ, Mangano MM, Dauphin J, Salenius SA,Dosoretz DE, et al. Small cell anaplastic carcinoma of the prostate:seven new cases, review of the literature, and discussion of a thera-peutic strategy. Am J Clin Oncol 1997;20:376–80.

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

11. Spiess PE, Pettaway CA, Vakar-Lopez F, Kassouf W, Wang X, BusbyJE, et al. Treatment outcomes of small cell carcinoma of the prostate: asingle-center study. Cancer 2007;110:1729–37.

12. Lotan TL, Gupta NS, Wang W, Toubaji A, Haffner MC, Chaux A, et al.ERG gene rearrangements are common in prostatic small cell carci-nomas. Mod Pathol 2011;24:820–8.

13. Mosquera JM, Beltran H, Park K, MacDonald TY, Robinson BD,Tagawa ST, et al. Concurrent AURKA and MYCN gene amplificationsare harbingers of lethal treatment-related neuroendocrine prostatecancer. Neoplasia 2013;15:1–10.

14. Williamson SR, Zhang S, Yao JL, Huang J, Lopez-Beltran A, Shen S,et al. ERG-TMPRSS2 rearrangement is shared by concurrent prostaticadenocarcinoma and prostatic small cell carcinoma and absent insmall cell carcinoma of the urinary bladder: evidence supportingmonoclonal origin. Mod Pathol 2011;24:1120–7.

15. Guo CC, Dancer JY, Wang Y, Aparicio A, Navone NM, Troncoso P,et al. TMPRSS2-ERG gene fusion in small cell carcinoma of theprostate. Hum Pathol 2011;42:11–7.

16. Beltran H, Rickman DS, Park K, Chae SS, Sboner A, MacDonald TY,et al. Molecular characterization of neuroendocrine prostate cancerand identification of new drug targets. Cancer Discov 2011;1:487–95.

17. Hansel DE, Nakayama M, Luo J, Abukhdeir AM, Park BH, BieberichCJ, et al. Shared TP53 gene mutation in morphologically andphenotypically distinct concurrent primary small cell neuroendo-crine carcinoma and adenocarcinoma of the prostate. Prostate2009;69:603–9.

18. Wang W, Epstein JI. Small cell carcinoma of the prostate. A morpho-logic and immunohistochemical study of 95 cases. Am J Surg Pathol2008;32:65–71.

19. Yuan J, Knorr J, Altmannsberger M, Goeckenjan G, Ahr A, Scharl A,et al. Expression of p16 and lack of pRB in primary small cell lungcancer. J Pathol 1999;189:358–62.

20. PeiferM, Fernandez-Cuesta L, SosML,George J, Seidel D, Kasper LH,et al. Integrative genome analyses identify key somatic driver muta-tions of small-cell lung cancer. Nat Genet 2012;44:1104–10.

21. Meuwissen R, Linn SC, Linnoila RI, Zevenhoven J, Mooi WJ, Berns A.Induction of small cell lung cancer by somatic inactivation of bothTrp53 and Rb1 in a conditional mouse model. Cancer Cell 2003;4:181–9.

22. Greenberg NM, DeMayo F, FinegoldMJ,Medina D, TilleyWD, AspinallJO, et al. Prostate cancer in a transgenic mouse. Proc Natl Acad SciU S A 1995;92:3439–43.

23. Zhou Z, Flesken-Nikitin A, Corney DC, Wang W, Goodrich DW, Roy-Burman P, et al. Synergy of p53 and rb deficiency in a conditionalmouse model for metastatic prostate cancer. Cancer Res 2006;66:7889–98.

24. GrassoCS,WuYM,RobinsonDR,CaoX,DhanasekaranSM,KhanAP,et al. The mutational landscape of lethal castration-resistant prostatecancer. Nature 2012;487:239–43.

25. Barbieri CE, Baca SC, LawrenceMS, Demichelis F, Blattner M, Theur-illat JP, et al. Exome sequencing identifies recurrent SPOP, FOXA1 andMED12 mutations in prostate cancer. Nat Genet 2012;44:685–9.

26. BeltranH,YelenskyR, FramptonGM,ParkK,DowningSR,MacDonaldTY, et al. Targeted next-generation sequencing of advanced prostatecancer identifies potential therapeutic targets and disease heteroge-neity. Eur Urol 2013;63:920–6.

27. Antonarakis ES, KeizmanD, ZhangZ,Gurel B, Lotan TL,Hicks JL, et al.An immunohistochemical signature comprising PTEN, MYC, and Ki67predicts progression in prostate cancer patients receiving adjuvantdocetaxel after prostatectomy. Cancer 2012;118:6063–71.

28. Chuang AY, DeMarzo AM, Veltri RW, Sharma RB, Bieberich CJ,Epstein JI. Immunohistochemical differentiation of high-grade pros-tate carcinoma from urothelial carcinoma. Am J Surg Pathol 2007;31:1246–55.

29. Herawi M, Epstein JI. Immunohistochemical antibody cocktail staining(p63/HMWCK/AMACR) of ductal adenocarcinoma and Gleason pat-tern 4 cribriform and noncribriform acinar adenocarcinomas of theprostate. Am J Surg Pathol 2007;31:889–94.

30. Liu W, Laitinen S, Khan S, Vihinen M, Kowalski J, Yu G, et al. Copynumber analysis indicates monoclonal origin of lethal metastaticprostate cancer. Nat Med 2009;15:559–65.

31. Lotan TL, Gurel B, Sutcliffe S, Esopi D, Liu W, Xu J, et al. PTEN proteinloss by immunostaining: analytic validation and prognostic indicatorfor a high risk surgical cohort of prostate cancer patients. Clin CancerRes 2011;17:6563–73.

32. Rubin SJ, Hallahan DE, Ashman CR, Brachman DG, Beckett MA,Virudachalam S, et al. Two prostate carcinoma cell lines demonstrateabnormalities in tumor suppressor genes. J Surg Oncol 1991;46:31–6.

33. Ikediobi ON,DaviesH,Bignell G, Edkins S,StevensC,O'Meara S, et al.Mutation analysis of 24 known cancer genes in the NCI-60 cell line set.Mol Cancer Ther 2006;5:2606–12.

34. Ha TU, Segev DL, Barbie D, Masiakos PT, Tran TT, Dombkowski D,et al.Mullerian inhibiting substance inhibits ovarian cell growth throughan rb-independent mechanism. J Biol Chem 2000;275:37101–9.

35. Nishizuka S, Charboneau L, Young L, Major S, ReinholdWC,WalthamM, et al. Proteomic profiling of the NCI-60 cancer cell lines using newhigh-density reverse-phase lysatemicroarrays. ProcNatl Acad Sci USA 2003;100:14229–34.

36. RuddML, Price JC, Fogoros S, Godwin AK, Sgroi DC,MerinoMJ, et al.A unique spectrum of somatic PIK3CA (p110alpha) mutations withinprimary endometrial carcinomas. Clin Cancer Res 2011;17:1331–40.

37. Berg M, Danielsen SA, Ahlquist T, Merok MA, Agesen TH, Vatn MH,et al. DNA sequence profiles of the colorectal cancer critical gene setKRAS-BRAF-PIK3CA-PTEN-TP53 related to age at disease onset.PLoS ONE 2010;5:e13978.

38. Di Sant Agnese PA, Cockett AT. The prostatic endocrine-paracrine(neuroendocrine) regulatory system and neuroendocrine differentia-tion in prostatic carcinoma: a review and future directions in basicresearch. J Urol 1994;152(5 Pt 2):1927–31.

39. Helpap B, Kollermann J. Undifferentiated carcinoma of the prostatewith small cell features: immunohistochemical subtyping and reflec-tions on histogenesis. Virchows Arch 1999;434:385–91.

40. Yao JL, Madeb R, Bourne P, Lei J, Yang X, Tickoo S, et al. Small cellcarcinoma of the prostate: an immunohistochemical study. Am J SurgPathol 2006;30:705–12.

41. Simon RA, di Sant'Agnese PA, Huang LS, Xu H, Yao JL, Yang Q, et al.CD44 expression is a feature of prostatic small cell carcinoma anddistinguishes it from its mimickers. Hum Pathol 2009;40:252–8.

42. Nikitin AY, Juarez-Perez MI, Li S, Huang L, Lee WH. RB-mediatedsuppression of spontaneous multiple neuroendocrine neoplasia andlung metastases in rbþ/� mice. Proc Natl Acad Sci U S A 1999;96:3916–21.

Tan et al.





43. Sutherland KD, Proost N, Brouns I, Adriaensen D, Song JY, Berns A.Cell of origin of small cell lung cancer: inactivation of Trp53 and Rb1in distinct cell types of adult mouse lung. Cancer Cell 2011;19:754–64.

44. Wang Y, Hayward SW, Donjacour AA, Young P, Jacks T, Sage J, et al.Sex hormone-induced carcinogenesis in rb-deficient prostate tissue.Cancer Res 2000;60:6008–17.

45. Hill R, SongY, Cardiff RD, VanDyke T. Heterogeneous tumor evolutioninitiated by loss of pRb function in a preclinical prostate cancer model.Cancer Res 2005;65:10243–54.

46. Zhou Z, Flesken-Nikitin A, Nikitin AY. Prostate cancer associatedwith p53 and rb deficiency arises from the stem/progenitor cell-enriched proximal region of prostatic ducts. Cancer Res 2007;67:5683–90.

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

48. Tzelepi V, Zhang J, Lu JF, Kleb B,WuG,Wan X, et al. Modeling a lethalprostate cancer variantwith small-cell carcinoma features.ClinCancerRes 2012;18:666–77.

49. LiuW, XieCC, ThomasCY,KimST, Lindberg J, EgevadL, et al. Geneticmarkers associated with early cancer-specific mortality followingprostatectomy. Cancer 2013;119:2405–12.

50. Sharma A, Yeow WS, Ertel A, Coleman I, Clegg N, Thangavel C, et al.The retinoblastoma tumor suppressor controls androgen signaling andhuman prostate cancer progression. J Clin Invest 2010;120:4478–92.

51. Friedlander TW, Roy R, Tomlins SA, Ngo VT, Kobayashi Y, AzameeraA, et al. Common structural and epigenetic changes in the genome ofcastration-resistant prostate cancer. Cancer Res 2012;72:616–25.

52. Kuhn E, Kurman RJ, Vang R, Sehdev AS, Han G, Soslow R, et al. TP53mutations in serous tubal intraepithelial carcinoma and concurrentpelvic high-grade serous carcinoma–evidence supporting the clonalrelationship of the two lesions. J Pathol 2012;226:421–6.

53. ChenH, Sun Y,WuC,Magyar CE, Li X, Cheng L, et al. Pathogenesis ofprostatic small cell carcinoma involves the inactivation of the P53pathway. Endocr Relat Cancer 2012;19:321–31.

54. Aparicio A, Den RB, Knudsen KE. Time to stratify? the retinoblastomaprotein in castrate-resistant prostate cancer. Nat Rev Urol 2011;8:562–8.

55. Chatterjee SJ, George B, Goebell PJ, Alavi-Tafreshi M, Shi SR, FungYK, et al. Hyperphosphorylation of pRb: a mechanism for RB tumoursuppressor pathway inactivation in bladder cancer. J Pathol 2004;203:762–70.

56. Wikenheiser-Brokamp KA. Rb family proteins differentially regulatedistinct cell lineages during epithelial development. Development2004;131:4299–310.






2014;20:890-903. Published OnlineFirst December 9, 2013.Clin Cancer Res Hsueh-Li Tan, Akshay Sood, Hameed A. Rahimi, et al. CarcinomaRb Loss Is Characteristic of Prostatic Small Cell Neuroendocrine

Updated version

10.1158/1078-0432.CCR-13-1982doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2013/12/09/1078-0432.CCR-13-1982.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/20/4/890.full#ref-list-1

This article cites 54 articles, 19 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/20/4/890.full#related-urls

This article has been cited by 39 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/20/4/890To 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-13-1982

http://clincancerres.aacrjournals.org/content/suppl/2013/12/09/1078-0432.CCR-13-1982.DC1

http://clincancerres.aacrjournals.org/content/20/4/890.full#ref-list-1

http://clincancerres.aacrjournals.org/content/20/4/890.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/20/4/890


rb loss is characteristic of prostatic small cell ... · anti-human p53 antibody (clone do-7; dako;...

Documents