comprehensive mutation and copy number ﬁling in archived ... · 2016 meeting (# p3151), at 10th...

Translational Science

Comprehensive Mutation and Copy NumberProfiling in Archived Circulating Breast CancerTumor Cells Documents HeterogeneousResistance MechanismsCostanza Paoletti1,2, Andi K. Cani3,4,5, Jose M. Larios1,2, Daniel H. Hovelson3,6,KimberlyAung1,2, ElizabethP.Darga1,2, EmilyM.Cannell1,2, Paul J.Baratta1,2,Chia-JenLiu3,5,David Chu7, Maryam Yazdani8, Allen R. Blevins8, Valeria Sero8, Nahomi Tokudome1,2,9,DafyddG.Thomas1,2,5, ChristinaGersch1,2, Anne F. Schott1,2,Yi-MiWu2,3, Robert Lonigro2,3,Dan R. Robinson2,3, Arul M. Chinnaiyan2,3,5, Farideh Z. Bischoff8, Michael D. Johnson10,Ben H. Park7, Daniel F. Hayes1,2, James M. Rae1,2, and Scott A. Tomlins2,3,5,11

Abstract

Addressing drug resistance is a core challenge in cancer research,but the degree of heterogeneity in resistance mechanisms in canceris unclear. In this study, we conducted next-generation sequencing(NGS) of circulating tumor cells (CTC) from patients withadvanced cancer to assess mechanisms of resistance to targetedtherapy and reveal opportunities for precision medicine. Compar-ison of the genomic landscapes of CTCs and tissue metastases iscomplicated by challenges in comprehensive CTC genomic profil-ing and paired tissue acquisition, particularly in patients whoprogress after targeted therapy. Thus, we assessed by NGS somaticmutations and copy number alterations (CNA) in archived CTCsisolated from patients with metastatic breast cancer who wereenrolled in concurrent clinical trials that collected and analyzedCTCs and metastatic tissues. In 76 individual and pooled infor-mative CTCs from12 patients, we observed 85% concordance in atleast one or more prioritized somaticmutations and CNA betweenpaired CTCs and tissue metastases. Potentially actionable genomic

alterations were identified in tissue but not CTCs, and vice versa.CTC profiling identified diverse intra- and interpatient molecularmechanisms of endocrine therapy resistance, including loss ofheterozygosity in individual CTCs. For example, in one patient,we observed CTCs that were either wild type for ESR1 (n ¼ 5/32),harbored the known activating ESR1 p.Y537S mutation (n ¼ 26/32), or harbored a novel ESR1 p.A569S (n ¼ 1/32). ESR1 p.A569Swas modestly activating in vitro, consistent with its presence as aminority circulating subclone. Our results demonstrate the feasi-bility and potential clinical utility of comprehensive profiling ofarchived fixed CTCs. Tissue and CTC genomic assessment arecomplementary, and precise combination therapies will likely berequired for effective targeting in advanced breast cancer patients.

Significance: These findings demonstrate the complementarynature of genomic profiling from paired tissue metastasis andcirculating tumor cells frompatientswithmetastatic breast cancer.Cancer Res; 78(4); 1110–22. �2017 AACR.

IntroductionThe promise of precision oncology implies that patients with

advanced cancer might undergo somatic genomic profiling toidentify molecular alterations that match with potentially active

molecularly targeted therapies. Most of the strategies of precisionmedicine to date have used archival or prospectively biopsiedtissues for somatic profiling. The use of liquid biopsies, to obtaincirculating tumor cells (CTC)or circulating cell-freeDNA(cfDNA)

1Breast Oncology Program of the University of Michigan Comprehensive CancerCenter, Ann Arbor, Michigan. 2Comphrehensive Cancer Center, University ofMichigan, Ann Arbor, Michigan. 3Michigan Center for Translational Pathology,Department of Pathology, University of Michigan, Ann Arbor, Michigan. 4Molec-ular and Cellular Pathology Graduate Program, University of Michigan MedicalSchool, Ann Arbor, Michigan. 5Department of Pathology, University of MichiganMedical School, Ann Arbor, Michigan. 6Department of Computational Medicineand Bioinformatics, University of Michigan Medical School, Ann Arbor, Michigan.7The Sidney Kimmel Comprehensive Cancer Center, The Johns Hopkins Uni-versity School of Medicine Department of Oncology, Baltimore, Maryland.8Menarini Silicon Biosystems, Inc., San Diego, California. 9Present address: ThirdDepartment of Internal Medicine, Wakayama Medical University, Wakayama,Japan. 10Georgetown University, Washington, D.C. 11Department of Urology,University of Michigan Medical School, Ann Arbor, Michigan.

Note: Supplementary data for this article are available at Cancer ResearchOnline (http://cancerres.aacrjournals.org/).

C. Paoletti and A.K. Cani contributed equally to this article.

Prior presentation/publications: Data from this work have been partiallyreported by Paoletti and colleagues in the form of a poster at SABCS 2015meeting (# P2-02-19), by Paoletti and colleagues in the formof a poster at AACR2016 meeting (# P3151), at 10th International Symposium on Minimal ResidualCancer Conference (# P74), and Cani and colleagues' poster at the 2017 USCAPmeeting (# 14330).

Corresponding Author: Scott A. Tomlins, Department of Pathology, Universityof Michigan School of Medicine, 1500 E. Medical Center Drive, 7322 CancerCenter, Ann Arbor, MI 48109. Phone: 734-764-1549; Fax: 734-647-7950; E-mail:[email protected]

doi: 10.1158/0008-5472.CAN-17-2686

�2017 American Association for Cancer Research.

CancerResearch

Cancer Res; 78(4) February 15, 20181110

on October 29, 2020. © 2018 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from

Published OnlineFirst December 12, 2017; DOI: 10.1158/0008-5472.CAN-17-2686

http://crossmark.crossref.org/dialog/?doi=10.1158/0008-5472.CAN-17-2686&domain=pdf&date_stamp=2018-1-31

http://cancerres.aacrjournals.org/

for phenotypic or molecular analyses, has the potential to over-come tissue availability as a major barrier to precision oncology(1–4). Such approaches may be particularly valuable in thecontext of patients progressing after targeted therapy, in whoma single tissue biopsy may be unable to capture the diverseresistance mechanisms driving individual clonal populations ofprogressing metastases (5). Importantly, although numerouscfDNA and CTC platforms have been profiled, the only FDA-cleared CTC platform is the CellSearch system (Menarini, SiliconBiosystems), which has regulatory clearance in several cancers forprognostic utility of CTC counts (6, 7). CTC profiling in treat-ment-refractory cancers holds particular promise as a translationalresearch tool to capture the global set of resistant cell populationsin individual patients, as deconvolution of individual resistantpopulations in cfDNA requires massive sequencing depth orbreadth (when cfDNA tumor content is low) and tissue-basedprofiling only provides information on the biopsied metastasis(8–11).

Nearly all patients with hormone receptor–positive (HRþ)metastatic breast cancer (MBC) initially respond to antiestrogentreatments (endocrine therapy), but ultimately nearly all progress.Several endocrine therapy resistance mechanisms in estrogenreceptor–positive (ERþ) MBC have been identified, including ERdownregulation (through deletion or suppression), alterations inER signaling pathway genes, deregulation of growth pathways,downregulation of apoptosis pathways, and unbalanced ER–coregulator activity (12). More recently, we and others havereportedmutations in the ligand-binding domain (LBD) of ESR1,the gene that encodes ERa, in patients with MBC after endocrinetherapy (11, 13–16). These mutations appear to confer absoluteresistance to estrogen depletion, and relative resistance to selective

ER modulators, such as tamoxifen, and selective ER downregu-lators, such as fulvestrant (11).

We and others have investigated concordance between tissue,cfDNA, and CTC-based assessment of ESR1mutations in patientswithMBC (17–19). Importantly, themajority of CTC-based geno-mic profiling approaches inMBC, and other cancers, either rely onplatforms that do not fix CTCs, utilize low pass whole-genomesequencing to identify broad copy number alterations (CNA) andtargeted next-generation sequencing (NGS)/Sanger sequencingapproaches that only assess mutations in a very focused set ofgenes, or rely only on pools of CTCs as opposed to single cells(20–24). To interrogate CTC-based endocrine therapy resistancemechanisms and to determine concordance with tissue biopsy,we performed comprehensive mutation and copy number profil-ing in 130 genes from archived CTCs captured from patientswith endocrine therapy–resistant MBC enrolled on a prospectiveclinical trial. Importantly, these patients were enrolled on a con-current clinical trial in which they underwent metastatic tissuebiopsy [forwhole-exome sequencing (WES)] enabling comparisonof comprehensive CTCs and tissue metastasis genomic profiles.

Materials and MethodsPatient population

Twenty-eight eligible patients with MBC, from the MI-ONCO-SEQ protocol (performing WES of metastatic tissue researchbiopsy) at the University of Michigan Comprehensive CancerCenter (Ann Arbor, MI; ref. 10), were enrolled in the companionMi-CTC-ONCOSEQ protocol after obtaining a separate writteninformed consent approved by the University of Michigan Insti-tutional Review Board in accordance with the Declaration of

Patients with <5 CTC/7.5 mL (n = 12)

Patients enrolled in Mi-CTC-ONCOSEQ (n = 30)

Baseline: CTC collected for evaluable patients (n = 28)

Group of patients with ≥5 CTC/7.5 mL (n = 16)

Patients with ≥ 5 CTC/7.5 mL, with no high quality CTC-DNA (n = 4)

Excluded: Ineligible (n = 1); protocol deviation (n = 1)

Patients with ≥ 5 CTC/7.5 mL, with no high confidence CTC after sequencing (n = 1)

Group of patients with ≥5 CTC/7.5 mL with at least one high quality CTC-DNA (n = 11)

Progression: Patients with ≥5 CTC/7.5 mL with at least one high quality CTC-DNA sequenced (n = 2)

Progression: Patients with ≥ 5 CTC/7.5 with at least one high quality CTC (n = 1)

Progression: Patients with ≥ 5 CTC/7.5 mL with at least one high quality CTC (n = 1)

Figure 1.

REMARK diagram for patient enrollment and distribution. Of 30 enrolled patients, 12 were protocol conforming, had �5 CTC/7.5 mL whole blood by CellSearch,and had at least one CTC with high-quality DNA at baseline or progression.

CTC Genomic Profiling in Metastatic Breast Cancer Patients

www.aacrjournals.org Cancer Res; 78(4) February 15, 2018 1111




R9

C

R2

B

R3

C

R1

3 B

D6

A

R1

4 B

R1

2 C

D1

A

R1

1 B

A1

2 A

D3

A

A9

A

R1

C

R3

B

R7

B

R1

1 C

R1

B

R2

0 B

D2

A

R4

C

R7

C

R6

C

R5

B

R1

7 B

R1

8 C

R1

0 C

7P

3b

A

6P

R7

- 20

B

5P

R1

-7 C

6P

A3

-D5

A

6P

R1

-6 B

5P

R9

-18

C

cfDNA

ddPCR

Tissue

WES (VF or

log2CNV)

Tumor

cont.

CDH1 p.Q641X 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.91 0.5 NC NC NC NC NC NC NC NC NC 1.00 1.00 1.00 1.00 0.93 0.71 NA YES (0.75)

CDH1 p.S70F 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NC NC NC 1.00 0.5 1.00 1.00 1.00 1.00 1.00 1.00 1.00 NC 0.8 1.00 1.00 1.00 0.99 0.33 0.44 NA YES (0.73)

ESR1 p.Y537S 1.00 1.00 0.87 0.7 0.65 0.5 0.5 0.34 0.10 0.5 0.4 0.5 1.00 0.5 0.5 0.5 0.4 0.3 1.00 0.83 0.47 0.46 0.41 0.35. 0.15 0.40 0.87 YES (0.45)

ESR1 p.A569S 0.56 NO * NO

ESR1 p.D538G 0.01 NO

TP53 CNV -2.0 -1.3 -1.2 -1.8 -1.2 -3.6 -1.2 -2.2 -1.4 -1.0 -1.9 -1.1 -1.4 NA YES (-0.8)

CDH1 CNV -1.0 1.4 -1.4 -1.1 -1.3 NA YES (-0.8)

ChrX CNV -1.3 -1.1 -1.8 -1.5 -1.1 -1.4 -1.1 -1.6 -1.0 -1.1 -1.2 -3.7 -1.0 -1.0 NA YES (-0.8)

ATM CNV -1.1 1.1 -2.8 1.9 NA NO

SMARCB1 CNV -1.1 -1.2 -1.4 -1.3 -8.1 -2.2 -1.3 NA NO

80%

High -

ly

aneu

-plo -

id

Pa

tie

nt

2

No

r ma

l CK

+ c

ell

Pt0

4 R

3

Hi g

hly

A neu

ploi

d

Pt 0

4 R

2

Pt0

4 R

1

Pt0

4 (3

P)

Pt0

7 A

2

Pt0

7 (3

P)

Pt1

2 A

9

Pt1

4 (3

P)

Pt1

7 (4

P)

Pt1

9 ( 3

P)

Pt2

4B

(5P

)

Pt2

4B

( 4P

1)

Pt2

4B

(4P

2)

Pt 2

4P

R2

Pt2

4P

R3

Pt2

4P

(3P

)

Pt2

5 R

4

Pt2

5 R

12

Pt2

5 R

13

Pt2

5 (3

P)

Pt2

5 (2

P)

Pt2

8 R

3

Pt2

8 ( 2

P)

Pt2

9 R

1

Pt2

9 R

2

Pt2

9 R

3

Pt2

9 R

5

Pt2

9 (4

P1

)

Pt2

9 (4

P2

)

Pt3

0 R

1

Pt3

0 R

2

Pt 3

0 R

3

Pt3

0 R

5

Pt3

0 R

10

Pt3

0 (4

P)

Tissue

WES (VF

or log2CNV)

Tumor

cont.

PIK3CA p.E545K 1.00 0.52 YES (0.27)

TP53 p.D259H 1.00 0.43 YES (0.29)

TP53 p.K139N YES (0.15)

TP53 p.S116F YES (0.16)

TP53 p.R280T 0.65 0.16 NO

DNMT3A CNV -1.6 -1.6 NO

KRAS CNV 2.2 1.5 NO

PIK3CA p.E542K YES (0.30

ESR1 p.Y537S YES (0.30)

CDK4 CNV 2.0 NO

BAP1 CNV -2.3 NO

CCND1 CNV YES (1.4)

PIK3CA p.H1047R 0.85 YES (0.63)

TP53 p.R248Q 0.73 YES (0.86)

WT1 CNV 3.0 YES (3.0)

TSC2 CNV 1.7 YES (1.3)

MYC CNV 1.6 YES (1.4)

NF1 CNV YES (1.4)

HNF1A p.W206C 0.2 NO

CCND1 CNV YES (2.7)

PTEN CNV YES (-1.8)

PIK3CA p.H1047R YES (0.37)

PIK3CA p.E418K NC YES (0.36)

NF1 p.W2494X YES (0.55)

BRCA2 p.Q1931X 0.10 NO

FGFR1 CNV YES (1.4)

MDM2 CNV YES (1.1)

Patient 19 PTCH1 p.E1242X 0.28 NO 65%

CDH1 p.I584fsdel 0.79 0.68 0.77 1.00 1.00 0.45 YES (0.59)

TP53 p.152_156fsdel 0.94 0.30 0.88 1.00 0.98 0.51 YES (0.79)

CDH1 p.E841X 0.14 NO

TP53 p.V216L 0.36 NC NC NO

NOTCH1 p.S2492X 0.18 NO

TET2 p.V1718L 1.00 0.58 0.75 NO

TET2 p.E1199Q YES (0.82)

NF1 p.S1030X YES (0.22)

SOX2 CNV 2.5 2.1 2.2 1.1 2.0 1.9 YES (2.2)

PIK3CA CNV 1.8 1.7 1.6 1.3 1.6 YES (1.6)

MYCN CNV 2.8 3.2 NO

TP53 p.H179R 1.00 NC 0.61 YES (0.67)

CCND1 CNV 2.5 2.0 YES (3.0)

MYC CNV 1.6 -7.4 1.4 -1.4 YES (2.1)

BIRC3 CNV 2.1 1.1 1.1 1.3 NO

BAP1 CNV -2.8 NO

PIK3R1 CNV -3.1 NO

TP53 p.G245V 1.00 1.00 YES (0.79)

PIK3CA p.E542K 0.20 0.55 YES (0.40)

ESR1 CNV 3.9 2.5 YES (2.8)

MYC CNV 3.4 2.5 YES (2.6)

ERBB2 CNV YES (3.7)

CDKN2A CNV -1.9 NO

Chr1 arm q CNV YES

CDKN2A CNV -1.6 -1.8 NO

PIK3CA p.H1047R 0.52 0.36 0.54 0.65 0.49 0.48 YES (0.30)

ESR1 p.D538G 1.00 0.29 1.00 1.00 NC 1.00 YES (0.60

FGFR1 CNV 3.0 3.4 3.4 3.0 3.4 3.4 YES (3.5)

CCND1 CNV 1.3 2.3 1.9 1.8 2.2 1.4 YES (1.6)

ESR1 CNV -1.2 -1.6 -2.6 -1.3 YES (-0.7)

Chr4 q12 CNV -1.2 -1.3 -1.1 YES (-0.6)

Chr1 arm q CNV YES

Chr1 arm p CNV YES

Patient 25

Patient 28

Patient 29

Patient 30 70%

67%

70%

67%

Patient 14

Patient 24

75%

Patient 4

Patient 7

Patient 12

59%

67%

58%

66%

88%

Patient 17

No

rma

l C

K+

ce

ll

No

r ma

l C

K+

ce

ll

A

B

Paoletti et al.

Cancer Res; 78(4) February 15, 2018 Cancer Research1112




Helsinki. It allowed blood collection for cfDNA (data partiallyreported elsewhere; ref. 17) and CTC analyses (SupplementaryMaterials and Methods).

CTC enrichment and enumerationCTCs were enriched from whole blood using EpCAM anti-

body–coated ferrofluid particles and enumerated using the Cell-Search System, according to themanufacturer's instructions (Jans-sen Diagnostics, LLC), and as described previously (Supplemen-tary Materials and Methods; ref. 25).

Single cell purification and DNA isolationEach CellSearch cartridge was processed to recover single

CTC using the DEPArray system (Menarini Silicon Biosystems,S.p.A.) per the manufacturer's instructions. After being flushedfrom the cartridges, cells were placed into the DEPArray 300kchip, which separates individual cells based on immunofluo-rescent staining criteria and cell morphology [CTC: CK-FITCpositive, DAPI positive, and CD45-APC negative; white bloodcells (WBC): CK-FITC negative, DAPI positive, and CD45-APCpositive]. After imaging, individual selected cells were routedfor isolation and recovery (26). DNA from individual CTCs orWBCs was isolated and subjected to whole-genome amplifica-tion (WGA) using Ampli1 WGA Kit (Menarini Silicon Biosys-tems, S.p.A.) with MseI digestion per the manufacturer'sinstructions (27). Subsequent DNA quality control was per-formed using Ampli1 QC Kit (27), and low-quality DNA cellswere excluded from downstream sequencing (SupplementaryMaterials and Methods).

CTC genomic profiling and data analysisNGS was performed essentially as previously described using

20 ng of WGA DNA for each CTC/WBC sample for targeted,multiplexedPCR-basedNGS (Ampliseq, IonTorrent; refs. 28, 29).Libraries were constructed using the DNA component of theOncomine Comprehensive Assay (OCP), a panel of 2,531 ampli-cons targeting 130 genes covering 260,717 bases (Thermo FisherScientific), and selected mutations were confirmed by SangerSequencing (SupplementaryMaterials andMethods; Supplemen-tary Table S1).

Tissue biopsy WESSequencing of clinical samples was performed as described

previously (11). Genomic DNA from frozen needle biopsies wasused to generate exome libraries of matched tumor/normal pairsusing the Illumina TruSeq DNA Sample Prep Kit. WES wasperformed on Illumina HiSeq 2000 or HiSeq 2500 (paired-end)and analyzed as described previously (Supplementary Materialsand Methods; ref. 11).

In vitro functional studies of ESR1 A569S mutationpCDH-ESR1 plasmid was mutated at alanine 569 of ER to serine

withQuick Change Lightning Kit (Agilent Technologies). Cells wereobtained from the Tissue Culture Shared Resource at the LombardiComprehensive Cancer Center (Georgetown University, Washing-ton, DC) in 2001. Cells were never passaged more than 15 times,were free of mycoplasma contamination (most recent testing, May2015), and identity was confirmed by short tandem repeat (mostrecent testing, July 2014). Lentivirus containing the p.A569S ESR1transgene was packaged in 293T cells. Twenty-four hours afterplating, cells were transfected with 8 mg of pCDH-ESR1-A569Splasmid, 5 mg psPAX2, and 2 mg pMD2.G plasmids. MCF-7 cellswere virally transduced with 1 mL of viral supernatant supplemen-ted with 4 mg/mL polybrene for 8 hours. Steroid-depleted parentaland ESR1 p.A569SMCF-7 cells were seeded into 96-well plates andtreated with 17b-estradiol (Sigma Aldrich) or ethanol control aloneor in combinationwith tamoxifen, 4-hydroxytamoxifen, endoxifen,or fulvestrant. Cell numberwas assessed by crystal violet stain 5daysafter hormone treatment as described previously (SupplementaryMaterials and Methods; ref. 30).

ResultsPreclinical proof-of-concept validation of targetedNGSofCTCsfrom archived CellSearch cartridges

We performed a pilot, preclinical study to determine whethercells stored for long periods of time in CellSearch cartridges afterenrichment fromwhole blood could be purified using theDEPAr-ray technology for subsequent high-quality genomic profiling.Cultured BT-474 human breast cancer cells spiked into normalhuman blood and preenriched by CellSearch were archived in thecollection cartridges for several months-years. After separationfrom leucocytes to purity, single-cell DNA derived from spikedcells underwent WGA using Ampli1 WGA Kit and NGS with anapproximately 300 amplicon multiplexed PCR-based panel(CHPv2). This approach successfully identified the expected TP53p.E285K mutation in cells from all seven samples processed withan elapsed time between archiving of the CellSearch cartridge (in4�C) and cell isolation using DEPArray of 3 days to 6 months;cartridges archived for 2 years at room temperature did not yieldassessable genomic DNA (Supplementary Table S2).

Trial cohort for CTC assessment and comparison withmatchedtissue metastases

Genomic endocrine therapy resistance mechanisms in individ-ual CTCs and concordance with biopsy-obtained fresh tissue wereexamined in 30 patients with MBC. These patients participatedin the Michigan Oncology Sequencing Center (MI-ONCOSEQ)trial, in which a biopsy of metastatic tissue was subjected togenomic profiling. They were also enrolled in a companion trial

Figure 2.Integrative heatmapof somaticmolecular alterations identified in archivedCTCs and comparisonwithmetastatic tissue in endocrine therapy–resistant MBCpatients.Next-generation targeted sequencing for CTC, WES for tissue biopsy, and ddPCR for cfDNA identified high confidence mutations (top half of each patienttable) andCNAs (bottomhalf) in patient #2 (A) and patients #4–30 (B). Colored boxes, presence of alteration; empty boxes, absence of alteration (despite adequatesequencing coverage for the position). NC, no adequate sequencing coverage to evaluate mutation presence; NA, not assayed. Numbers inside mutationboxes show VF, with dark and light green boxes indicating homozygous (>0.8 VF) and heterozygous/complex (<0.8 VF) mutations, respectively. Orange boxes,presence of mutation and corresponding VF for CTC pools, tissue WES, and cfDNA. Numbers inside CNA boxes show the log2 copy number ratio (CNR), withgains and losses shown in red and blue, respectively. CTC IDs are shown at the top of heatmap with pooled sample IDs shaded. Note: tumor cont., tumor content; forpatient #24, "B" and "P" represent CTC samples at baseline and progression, respectively; � , ddPCR droplets for this mutation were detected, but below thepredetermined threshold.






(Mi-CTC-ONCOSEQ) to collect whole blood for CTC enrichmentand purification using the CellSearch and DEPArray systems.Two patients were deemed ineligible for regulatory reasons. Ofthe 28 remaining patients, 16 (57%) had �5 CTC/7.5 mL wholeblood by CellSearch at baseline (Fig. 1). Approximately 15% ofthe enriched CTCs were purified using DEPArray and had high-quality DNA (Supplementary Table S3). Eleven patients had atleast one CTC with sufficiently high-quality DNA for genomicanalysis.

Two patients had a blood sample drawn containing �5 CTC/7.5 mL whole blood at the time of disease progression, so that atotal of 12 patients had a sufficient number of CTCs for genomicanalysis at either baseline or progression. Eleven of the 12 patientshad been diagnosed with HRþ breast cancer (either from primaryor metastatic tissue); the remaining patient (#19) was diagnosedwith triple-negative [ER�/progesterone receptor (PR)�/HER2(ERBB2)�] breast cancer in both primary and metastatic tissue.Clinical histories are shown in Supplementary Fig. S1.

Simultaneous assessment of somatic mutations and CNAs inarchived CTCs from patients with MBC

From the 12 patients with evaluable CTCs, individual CTCswere isolated from archived CellSearch cartridges by DEPArray,followed by genomic DNA extraction and WGA. DNA from 53individual CTCs, 23 pooled CTC samples (containing equal DNAamounts from 2 to 7 individual CTCs), and 16 individual orpooled WBCs was subjected to comprehensive multiplexed PCR-based NGS using the DNA component of the OCP (Supplemen-tary Table S4). This assay, which is also being used in the NCIMATCH trial (31), interrogates activating and deleterious muta-tions and CNAs in 130 genes (29). All patients had WES per-formed on near synchronous metastatic tissue biopsies as part ofthe MI-ONCOSEQ platform. After OCP profiling, CTC doublypositive (CD45þ/CKþ) or pooled CTC samples lacking all molec-ular alterations (any somatic CNAs or prioritizedmutations)wereexcluded, resulting in 67 evaluable CTC samples (both single andpooled cells) from 12 patients.

In these 67 evaluable CTC samples, we identified a total of23 high-confidence, prioritized, somatic, nonsynonymous pointmutations and short insertions/deletions (median¼ 1.5, range¼0–6/patient), and31high-confidencehigh-level CNAs (median¼2.5, range ¼ 0–6/patient), as shown in an integrative heatmap(Fig. 2A and B). Importantly, no high-confidence mutations werefound in the matched WBCs in any of the 12 patients, and high-confidence CNAs inWBCs were exceedingly rare and were limitedto occasional copy losses, consistent with high-fidelity purifica-tion, WGA, and NGS (Fig. 3A and B).

Comparison of somatic mutations and CNAs in matched CTCsand tissue metastases

Our study design provided the opportunity to compare prior-itized somatic mutations and CNAs in CTCs versus synchronous/near-synchronous metastatic tissue biopsies. Critically, weobserved highly concordant somatic alterations from CTCs sub-jected to targetedNGS andmatchedWESof freshmetastatic tissuebiopsy. Specifically, 57 of 67CTC samples (85%) andCTCs in 8 of12 patients (67%) showed at least one, but usually multiple,prioritized genomic alterations detected in the correspondingtissue biopsy (Fig. 2A and B; Supplementary Table S5). Of 23point mutations and short indels detected in CTCs across allpatients, 14 (61%) were also found in the WES of corresponding

tissue biopsies, which additionally harbored 9 mutations/indelsthat were assayed, but not detected by the targeted panel in any ofthe corresponding CTCs (Supplementary Table S6A). Of note, thefraction of sequencing reads containing the variant [variant fre-quency (VF)] in individual CTCs was in the vast majority of caseseither 1.0 or approximately 0.5, consistent with homo- or het-erozygous status of mutations in individual cells, and was highlyconcordant with tumor content–corrected VFs in tissue samples(Fig. 2A and B).

Although previous studies have assessed mutations or CNAsfrom CellSearch isolated CTCs and other fixative-based CTCplatforms, we are unaware of simultaneous assessment of bothcategories of somatic alterations in fixed CTCs. We have previ-ously validated the ability of our multiplexed PCR-based NGSapproach to assess gene-level CNAs in fresh tissue, FFPE tissues,and cfDNA (28, 29, 32, 33). Herein, we adapted our approach toinclude only high-performance amplicons in WGA CTC genomicDNA and assessed CNAs (both high-level amplifications anddeletions) in 71 robustly assessedOCP target genes (seeMaterialsand Methods). Critically, we observed high concordance in pri-oritized CNAs detected in CTCs and matched tissue biopsiesacross our cohort. Specifically, 19 of 31 (61%) total CNAsdetected in CTCs were also present in fresh tissues subjected toWES, while WES identified 7 alterations assessable but notdetected in matched CTCs (Supplementary Table S6B). Further-more, similarly to mutations, the approximate copy number ofconcordant CNAs was highly consistent between CTCs andmatched tissue biopsies, as shown in Figs. 2A and B and 3A andB. For example, the tissue metastasis from patient #12 harboredprioritized PIK3CA p.H1047R and TP53 p.R248Q mutations, aswell asWT1, TSC2,MYC, andNF1 amplifications. The single high-quality CTC from this patient similarly harbored bothmutations,as well as the WT1, TSC2, and MYC amplifications, with WT1showing the greatest estimated copy number in both the tissueand CTC samples (Fig. 2B and B). Further supporting our tech-nique, patient #2 CTC CNA data showed no batch effects byunsupervised clustering despite being processed from three dif-ferentCellSearch cartridges (A; B; C) stored for varying amounts oftime before single-cell isolation (Fig. 3A; Supplementary TableS3).

Discordant genomic alterations between CTCs and corre-sponding tissue metastases were found in several patients,even if they only had a few high-quality CTCs. For example,patients #7, 14, 17, and 19 had complete discordance ingenomic alterations between CTCs and tumor tissues (Figs.2B and 3B). Of note, potentially targetable BRCA2 p.Q1931Xand PTCH1 p.E1242X mutations were found exclusively inCTCs (and not tissues) from patients #17 and 19, respectively.Whether such events represent bona fide mutations or WGA/NGS technical artifacts cannot be reliably determined. Tofurther investigate these possibilities, we performed Sangersequencing for the BRCA2 p.Q1931X mutation, which con-firmed its presence in one of the four individual cells pooledfor NGS analysis, consistent with the subclonal VF (0.10) inthat pooled NGS sample (Supplementary Fig. S2A). Important-ly, however, this CTC pool did not contain the PIK3CA (p.H1047R) or the NF1 (p.W2494X) mutations clearly presentin this patient's tissue, suggesting that these cells are not ofthe same origin as the tumor tissue that underwent WES.Similarly, none of the CTCs from patient #17 harbored thePIK3CA (p.E542K) and ESR1 (p.Y537S) hotspot mutations

Paoletti et al.





present in tissue (Fig. 2B). In addition, CTCs and tissue fromthese patients harbored several discordant copy numberchanges, again suggesting that these CTCs were from clonesthat were entirely different from that of the biopsied tissue(Figs. 2B and 3B).

Integrative mutation and CNA assessment of resistancemechanisms and clinically relevant intratumoral heterogeneityin CTCs from individual patients

Our ability to assess mutations and CNAs in individual CTCsfrom patients with matched tissue metastases profiled by WES

MYCL1

BCL9M

CL1D

DR2

MD

M4

MYCN

DN

MT3A

ALKM

SH2

IDH

1ERBB4PP ARGCTN

NB1

BAP1PIK3CASO

X2PD

GFRA

KITKD

RTET2FBXW

7TE RTPIK3R1APCESR1IL6EG

FRCD

K6M

ETSM

OFG

FR1M

YCCD

274CD

KN2A

PTCH1

ABL1TSC1PTENFG

FR2W

T1CD

44CCN

D1

BIRC3ATMKRASA CVRL1CD

K4M

DM

2PTPN

11FLT3BRCA2G

AS6APEX1PN

PIG

F1RTSC2CD

H1

TP53TIAF1N

F1ERBB2BRCA1RPS6KB1JAK3CCN

E1CSN

K2A1BCL2L1SM

ARCB1N

F2ARM

ED12

Pt30_WBCPt29_WBCPt28_WBCPt25_WBC_P1Pt24_WBC_P2Pt24_WBC_P1Pt19_WBCPt17_WBCPt14_WBCPt07_WBC_PPt04_WBCPt14_(6P)Pt07_A2Pt04_R1

Pt30_(4P)Pt30_R10Pt30_R5Pt30_R3Pt30_R2Pt30_R1Pt29_(4P2)Pt29_(4P1)Pt29_R5Pt29_R3Pt29_R2Pt29_R1Pt28_(2P)Pt28_R3Pt25_(2P)Pt25_(3P)Pt25_R13Pt25_R12Pt25_R4Pt24P_(3P)Pt24P_R3Pt24P_R2Pt24B_(4P2)Pt24B_(4P1)Pt24B_(5P)Pt19_(3P)Pt17_(4P)Pt14_(3P)Pt12_A9Pt07_(3P)Pt04_(3P)Pt04_R2Pt04_R3

WBCWBC_P2WBC_P1R10_C

Patient24712

14171924

25282930

SampleCTC

Norm. Epi.

WBC

>1.51.0 - 1.5-1.0 - 1.0< -1.0

Log2(CN Ratio)

Patient

Sample

Patient 2

Patients 4–30

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

Sample

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

MYCL1

BCL9M

CL1D

DR2

MD

M4

MYCN

DN

MT3A

ALKM

SH2

IDH

1ERBB4PP ARGCTN

NB1

BAP1PIK3CASO

X2PD

GFRA

KITKD

RTET2FBXW

7TERTPIK3R1APCESR1IL6EG

FRCD

K6M

ETSM

OFG

FR1M

YCCD

274CD

KN2A

PTCH1

ABL1TSC1PTENFG

FR2W

T1CD

44CCN

D1

BIRC3A TMKRASA CVRL1CD

K4M

DM

2PTPN

11F LT3BRCA2G

AS6APEX1PN

PIG

F1RTSC2CD

H1

TP53TIAF1N

F1ERBB2BRCA1RPS6KB1JAK3CCN

E1CSN

K2A1BCL2L1SM

ARCB1N

F2ARM

ED12

CTC

CTC

Normal

Normal

Patient

D6_AR14_BA9_A6P_A3−D5_AA12_A7P_3bR3_C5P_R9−18_C5P_R1−7_C6P_R1−6_BR11_CR4_CR13_BD2_AR1_BR6_CR9_CR5_BR3_BR2_BR12_CR1_CD1_AD3_AR7_CR7_B6P_R7−20_BR11_BR17_BR20_BR18_C

A

B

Figure 3.

Heterogeneity of CNA detected in CTCs across patients with MBC. Next-generation targeted sequencing generated gene-level log2 copy number ratios (CNR), withCNA losses in blue (negative log2 CNR) and gains in red (positive CNR). A and B, Patient #2 (A) and patients #4–30 (B). Greater absolute value of log2 CNRindicates a more pronounced CNA. A dendrogram from unsupervised hierarchical clustering of CNAs in patient #2 CTC is shown on the left.






provides a unique cohort to assess intra- and interpatient diversityin endocrine therapy resistance/progression mechanisms, includ-ing ESR1mutations andCNAs, aswell as comprehensive genomicprofiles. For example, patient #30 had HRþ MBC and developedendocrine therapy resistance between her tissue biopsy and CTCcollection. In the tissue biopsy subjected to WES, we detected aheterozygous PIK3CA p.H1047R hotspot mutation, a homozy-gous ESR1 p.D538G hotspot mutation [with loss of heterozygos-ity (LOH)due to single copyESR1 loss], andhigh-level FGFR1 andCCND1 amplifications. Consistent with the presumed contribu-tion of ESR1 hotspot mutations in resistance to some, but not all,types of endocrine therapy, she progressed on endocrine therapyafter her tissue biopsy. Critically, all informative individual (n ¼5) and pooled (n ¼ 1) CTC samples harbored the same hetero-zygous PIK3CA p.H1047R mutations that were detected in hertissue biopsy. Similarly, although all informative individual (n¼4) and pooled (n ¼ 1) CTC samples also harbored the ESR1 p.D538G mutation, our integrative mutation and copy numberprofiling allowed us to clearly distinguish CTCs harboring homo-zygous ESR1 mutations with LOH (n ¼ 3 individual and n ¼ 1pooled CTC samples; VF of 1.0 and one copy ESR1 loss) fromCTCs harboring ESR1 mutation with no CNA (n ¼ 1 individualCTC; VF of 0.29 CTC and no ESR1 copy alteration). Finally, boththe FGFR1 and CCND1 amplifications were detected in all indi-vidual and pooled CTCs at estimated copy numbers that wereconcordantwith those in the tissuebiopsy (FGFR1>CCND1; Figs.2B, 3B, and4A). Taken together, these results support the ability ofour single (and pooled) CTC profiling approach to identifyknown clinically relevant mutations and CNAs from tissue sam-ples, as well as integrate the CTC sequencing results to identifyLOH mechanisms.

Likewise, in patient #28, who had endocrine therapy–resistantHRþ MBC, we detected homozygous TP53 p.G245V and hetero-zygous PIK3CA p.E542K mutations in both CTC (one individualand one pooled sample) and in a liver tissue biopsy sequenced byWES (Figs. 2B and 4B). In addition, this patient harbored high-level ESR1, MYC, and ERBB2 amplifications in the tissue biopsy(Figs. 3B and 4B). Importantly, although both the ESR1 andMYCamplifications were clearly detected in the CTC samples, neitherof the CTC samples harbored the ERBB2 amplification (Fig. 4B).Of note, although this patient was initially considered HER2/ERBB2 negative based on clinical IHC (1þ expression in primaryand metastatic bone lesion), the ERBB2 amplification in the liverbiopsy prompted repeat bone metastasis IHC that showed het-erogeneous HER2/ERBB2 expression, with 10% to 15% of cellshaving 3þ staining. In addition to further validating the utility ofour approach, these results highlight the importance of discordantsubclonal alterations, as previously reported at the mRNA level(34), that may only be present in individual metastases resultingin low levels/frequencies or absence in the circulation.

Comprehensive profiling of single CTC identifies potentialalterations driving progressive disease

Amajor potential advantage of "liquid" biopsies is the ability tononinvasively monitor driving genomic alterations during dis-ease progression. For patient #24, who had endocrine therapy–resistant lobular MBC, CTCs were isolated from blood specimensconcurrently with the tissue biopsy (baseline) as well as atprogression (465 days later) after three lines of chemotherapy(Fig. 4C). As expected for lobular carcinoma, we identified aCDH1 frameshift mutation (p.I584fsdel) in all informative base-

line CTC (n ¼ 3 pooled CTC, B in Fig. 2B) as well as the tissuebiopsy. We similarly detected a TP53 frameshift mutation(p.152_156fsdel), as well as PIK3CA and SOX2 amplificationsin all baseline CTC samples and the tissue biopsy. Of note,although all four of these alterations were also present in theCTC samples at progression (P in Figs. 2B and 4C), two of theindividual progression CTCs also harbored high-level MYCNamplifications (estimated at 16 copies) that were not present inany baseline specimen (CTC or tissue), demonstrating the utilityof this approach to identify somatic alterations conferring treat-ment resistance during disease monitoring.

Comprehensive CTC profiling in a single patientidentifies multiple endocrine therapy resistance mechanismsin circulation

Tissue- and cfDNA-based studies havedemonstratednumerousresistance mechanisms to targeted therapies in the same patient(35), including studies in breast cancer demonstrating multipleESR1 mutations detectable in cfDNA from patients with endo-crine therapy–resistant MBC (35, 36). Hence, we performeddetailed profiling of numerous CTCs from patient #2, who hadendocrine therapy–refractory lobular breast carcinoma at the timeof research biopsy (for WES) and CTC collection (189 dayslater; Fig. 4D). Tissue WES detected two E-cadherin (CDH1)mutations (p.Q641Xandp.S70F) at homozygousVFdue to singlecopy CDH1 loss (Figs. 2A and 4D) as well as a heterozygousESR1p.Y537Smutation, which is presumed to be onemechanismcontributing to endocrine therapy resistance (11). No high-levelamplifications or deletions (in genes targeted in CTCs) werepresent, although TP53 and chromosome (chr) X one copy losseswere present by WES.

Across this patient's 32 assessed CTC samples (26 individualand6pools of 5–7 total individual CTCs),we identified theCDH1p.Q641X and p.S70F mutations in 22 of 23 (96%) and 27 of 28(96%) CTC samples informative for those genomic positions,respectively. Of note, 12 of 13 individual CTCs informative forboth CDH1 mutations showed both at homozygous VF, consis-tent with tissue metastasis WES and the known early role ofdeleterious alterations in CDH1 in lobular breast carcinoma.Although our CNA profiling is not optimal for single copyalterations, we observed the CDH1, TP53, and chr X lossesvariably across individual and pooled CTCs (n ¼ 13, 14, and 4of 32 CTC samples, respectively). Finally, one of this patient'spurified CKþ/CD45� circulating cells was wild type for all thetested genes, suggesting the presence of a minority of circulatingepithelial cells of nontumor origin.

As expected, in this patient, greater genomic heterogeneity wasobserved for the ESR1 p.Y537S mutation, which presumablyarose after endocrine therapy treatment (whether through selec-tion for a rare preexisting or acquired mutated clone). ESR1 p.Y537Smutations were detected in 26 of 32 individual and pooledCTC samples. However, five of the six CTCs harboring one ormore CDH1 mutations (strongly suggesting these are true cancercells) lacked theESR1mutation. Likewise, both heterozygous (n¼14/20 ESR1 mutation-harboring individual CTCs) and homozy-gous (n¼ 6/20 ESR1mutation-harboring individual CTCs) ESR1p.Y537S mutations were observed in CTCs. Of particular interest,one individual CTC from this patient (A12A) harbored bothhomozygous CDH1 mutations and lacked the ESR1 p.Y537Smutation, but instead harbored a unique, previously undescribedESR1 p.A569S mutation at heterozygous VF (Fig. 2A). Although

Paoletti et al.





A

Surgery RTTAMLetrozoleAnastrozoleExemestaneFulvestrant Other therapyChemo OophorectomyEstrace

* Everolimus# Palbociclib

N.E.D

Patient 30

Patient 28

#*

*30 mo 34 mo

12 mo 20 mo 15 mo 11 mo

PIK3CA p.H1047R, ESR1 p.D538G−4

−2

0

2

4 1 3 5 7 9 11 13 15 17 19 21 222018161412108642 X

−4

−2

0

2

4 1 3 5 7 9 11 13 15 17 19 21 222018161412108642 X

Met. Disease

WES Biopsy

CTC

TP53 p.G245V, PIK3CA p.E542K

ESR1

FGFR1 CCND1

ESR1 MYC ERBB2

5 mo

5 mo

−2−1

0123

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

PIK3CA p.H1047R, ESR1 p.D538G (VF=1.0)

PIK3CA p.H1047R, ESR1 p.D538G (VF=0.3)

ESR1 FGFR1 CCND1

−2−1

0123

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

TP53 p.G245V, PIK3CA p.E542K

ESR1 MYC ERBB2

Patient 24

Patient 2*

60 mo 60 mo 17 mo 24 mo

42 mo

Log2

Cop

y N

umbe

r Rat

io

−4

−2

0

2

4 1 3 5 7 9 11 13 15 17 19 21 222018161412108642 X

−4

−2

0

2

4 1 3 5 7 9 11 13 15 17 19 21 222018161412108642 X

CDH1 p.I584fs, TP53 p.152_156fs

CDH1 p.Q641X, ESR1 p.Y537S

MYCN PIK3CA, SOX2

CDH1 TP53

5 mo

5 mo

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

−2−1

0123

CDH1 p.I584fs, TP53 p.152_156fs1 3 5 7 9 11 13 15 17 19 2220161412108642 X

−2−1

0123

CDH1 p.I584fs, TP53 p.152_156fs

MYCN

MYCN

PIK3CA, SOX2

PIK3CA, SOX2

B

C

D

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

−2−1

0123

CDH1 p.Q641X, ESR1 p.Y537SCDH1 TP53

Tissue WES

CTC R2

CTC R3

Tissue WES

CTC 4P1

CTC R3

Tissue WES

CTC 5P R1-R7 C

Tissue WES

−2−1

0123

1 3 5 7 9 11 13 15 17 19 2220161412108642 X

ESR1 FGFR1 CCND1

CTC R3

Figure 4.

Integrative CNA andmutational profiling ofCTCs in comparison with tissue metastasesidentifies intra- and interpatientheterogeneity in resistance/progressionalterations. A–D, Clinical time lines, copynumber plots, and mutations from tissuemetastases subjected to WES and CTCssubjected to targeted NGS from patient#30 (A), patient #28 (B), patient #24 (C),and patient #2 (D). For timelines, treatmentcourses, development of metastasis, andresearch biopsy (for exome sequencing)and CTC collection, time points areindicated according to the legend.Treatment courses shown with a brokenbox are given in months. For WES, log2copy number ratios per segment areplotted and prioritized mutations are givenbelow the plot. For CTCs, each individualamplicon is represented by a single graydot, and gene-level copy number estimatesare shown in red bars (blue bars for genesof interest). Selected high-level CNAs areindicated. RT, radiotherapy; TAM,tamoxifen; N.E.D., no evidence of disease.






the ESR1 p.A569S mutation was not observed in any otherindividual or pooled CTCs, or in the patient's tissue metastasis(Fig. 2A), it, as well as ESR1 p.Y537S and CDH1 mutations inother CTCs, was confirmed by Sanger sequencing of WGA DNA(Supplementary Fig. S2A).

Genomicprofiling of tissuebiopsymaterial and afinite numberof CTCs, as in our study, is still likely to underestimate the fullrepertoire of minor subclonal driving mutations in an individualpatient with targeted therapy resistance. To further investigate thisconcept, droplet digital PCR (ddPCR) was performed on variousspecimen types at several time points during this patient's course.An additional ESR1 p.D538G hotspot mutation was detectedalbeit at extremely low level. Interestingly, the novel ESR1 p.A569S mutation detected in one CTC was also only present in afew droplets, but below the predetermined detection threshold(17) in post–endocrine therapy specimens (Supplementary TableS7; Supplementary Table S8).

Taken together, these results are consistent with a largecirculating pool of heterogeneous subclones, which may eachharbor differing mechanisms of resistance to one or more typesof endocrine therapy in patients with initially HRþ MBC. Theyfurther support complementary approaches to characterize thisdiversity of resistance mechanisms.

In vitro functional characterization of the novel ESR1 A569Smutation

As mentioned, the observed ESR1 p.A569S mutation has notbeen described previously in endocrine therapy–resistant MBCand was only observed in 1 of 32 total CTC samples (pooled andindividual) from patient #2 (Figs. 2A and 5A). Therefore, wehypothesized that it conferred amodest selective advantage as thecell in which it was detected was clearly a CTC (it harbored bothCDH1mutations present in the tissuemetastasis andother CTCs);it lacked the ESR1 p.Y537S mutation present in 26 of 32other CTC samples from the patient, and no other prioritizedmutations were observed in this or any other CTCs from thepatient. Hence, we used lentivirus-mediated infection to stablyoverexpress the ESR1 p.A569S mutation in the ERþ breast cancercell line MCF-7. Expression of the p.A569S mutation was con-firmed by Sanger sequencing and Western blot analysis (Supple-mentary Fig. S2B and S2C). In vitro growth assays, which com-pared MCF-7 ESR1 p.A569S with parental MCF-7, showed thatthe former was estrogen dependent, unlike cells expressing ESR1p.Y537S or p.D538G mutations, which confer estrogen-indepen-dent growth (Supplementary Fig. S3). Notably, however, MCF-7ESR1 p.A569S were more responsive to estradiol-induced growthcompared with parental MCF-7 (EC50 of 2.97 pmol/L vs. 5.38pmol/L; P ¼ 0.0001, F test; Fig. 5B).

Tamoxifen is a partial ER agonist. Therefore, we tested its abilityto stimulate growth in parental MCF-7 and MCF-7 ESR1 p.A569Scells in the absence of estradiol. Although tamoxifen stimulatedgrowth in both MCF-7 and MCF-7 ESR1 p.A569S cells in theabsence of estradiol, MCF-7 ESR1 p.A569S cells showed a signif-icantly greater growth increase than parental MCF-7 cells (130%vs. 180% over vehicle control, P < 0.0001, F test; Fig. 5C). Asexpected, the tamoxifen metabolites 4-hydroxytamoxifen andendoxifen (4-hydroxy-N-desmethyltamoxifen), both potent ERantagonists, did not stimulate growth in either parental MCF-7 orMCF-7 ESR1 p.A569S cells in the absence of estradiol (Supple-mentary Fig. S4A and S4B). Finally, the antiestrogens tamoxifen,4-hydroxytamoxifen, endoxifen, and fulvestrant were all able to

similarly attenuate growth induced by 50 pmol/L estradiol inboth parental MCF-7 and MCF-7 ESR1 p.A569S (SupplementaryFig. S5A–S5D). Taken together, these results demonstrate that thisESR1 mutation showed increased estradiol and tamoxifen-induced growth. Presumably, this mutation provides a modestselective growth advantage, consistent with the rarity of thismutation in circulation (1/32 CTC samples) compared with theESR1 p.Y537S mutation known to confer estrogen-independentgrowth and present at higher frequency in the patient's CTCs (26/32 samples) and tissue metastasis.

DiscussionIn these proof-of-concept studies, we have demonstrated that

individual CTCs from archived cartridges of the FDA-clearedCellSearch system can be isolated, whole-genome amplified, andcomprehensively profiled for somatic mutations and CNAs usingmultiplexed PCR-based NGS. Importantly, although comprehen-sive genomic profiling of single nonfixed CTCs has been reported(37), and CTCs isolated by CellSearch (or other fixative-basedapproaches) have been used for assessment of genome-wide CNAprofiling and mutations in a limited set of genes (19, 21, 24, 38),we are unaware of prior concurrent somatic mutation and precisegene-level CNA assessment in individual fixed CTCs from NGS.Leveraging near-synchronous matched tissue biopsies subjectedto WES as part of the MI-ONCOSEQ clinical trial, we demon-strated high concordance between driving somaticmutations andCNAs identified by our approach in CTC compared with singletissue metastases, including CNAs, mutations, and LOH assess-ment in both oncogenes (e.g., ESR1 and PIK3CA) and tumorsuppressors (e.g., CDH1 and TP53).

In addition to the high concordance between prioritized altera-tions identified in CTCs and matched tissue metastases, severallines of evidence support the validity of our findings. Process-matched WBCs assessed in parallel for each of the patients werewild type for all assessed genomic alterations. Likewise, at thepatient level, in addition to high concordance for somatic muta-tion and CNA presence/absence in CTC versus matched tissuemetastases, VFs (for mutations) and estimated copy numberratios (for CNAs) were similarly concordant. Finally, Sangersequencing of WGA DNA validated the selected mutations testedin 2 patients, one of which had highly heterogeneous findingsamong tissue, CTCs, and cfDNA.

Several patients had somatic alterations present in tissuemetas-tases that were not present in all matched CTCs, such as ESR1 p.Y537Smutation for patient #2. Likewise, in patient #28, the high-level ERBB2 amplification present in the liver metastasis was notidentified in the CTCs. In fact, only approximately 15% of tumorcells in a femur metastasis had 3þ HER2 expression by IHC,consistent with ERBB2 amplification being a subclonal eventwithin an individual tissue metastasis as well as in tissue versuscirculation.

A purported advantage of liquid biopsies is the theoreticalability to capture the global collection of molecular alterationsharboredby apatientwithmetastatic cancer. Inour study,wewereable to simultaneously assess mutations and copy number inindividual cells. Thus, we observed numerous examples of intra-patient somatic alteration variability between individual CTCsand tissue metastases, including assessment of heterozygosity/homozygosity supported by copy number state. For example,although all CTCs from patient #30 harbored PIK3CA p.H1047R

Paoletti et al.





mutations, we detected individual CTCs that harbored eitherheterozygous or homozygous (with one copy loss) ESR1 p.D538G mutations. Similarly, we observed several CTCs thatharbored somatic mutations and CNAs not present in tissue, butalso detected alterations in tissue metastases not observed inCTCs. Importantly, we identifiedpotentially targetable alterations(in PTCH1 andNOTCH1) in subsets of CTC from 2 patients (#19and #24) but not in matched tissue, a finding that supports theseas subclonal alterations. In this case, knowledge of the subclonalpopulations might enable more specific, and even tailored com-bination targeted therapy, either at initiation of treatment or foremerging subclones during cancer progression short of any deepsequencing of primary/metastatic tissue. The latter has shown thatmost discordant "private" CTC alterations can be detected in thetissue at low frequencies (24).

As an additional mechanism of discordance between CTCs andtissuemetastases, patient #4 harbored PIK3CA p.E545K and TP53

p.D259H mutations in both CTCs and tissue metastases. Impor-tantly, however, although both these mutations were heterozy-gous in the tissuemetastasis, we observed both as homozygous ina single CTC (without evidence of copy loss), but wild type in theother two cells. Finally, in patient #24, although CTCs and theconcurrent tissue metastasis shared several alterations, CTCstaken subsequently after progression on several lines of chemo-therapynot only harbored these alterations, but alsouniquely hadhigh-level MYCN amplifications. These findings further suggestthat serial "liquid profiling" to monitor molecular alterationsmediating resistance might permit specific selection of combina-tion targeted therapies during a patient's clinical course.

Emerging genomic tumor heterogeneity due to accumulatedsubclones with different mutations as well as plastic epigenetic/transcriptomic heterogeneity as a function of cellular stress andenvironmental changes are drivers of resistance to specific treat-ments. Indeed, we and others have demonstrated inter- and

A

180 263 302 553 595

Y537S

D538G

A569S

LBD

A B C E F

B

100

50

0No

rma

lize

d g

row

th v

s. 0

.01

% E

tOH

-14 -13 -12 -11 -10 -9

Log [M] estradiol

Parental

A569S

Parental A5369S

EC50 (pmol/L) 5.38 2.97

Log [M] tamoxifen

-11 -10 -9 -8 -7

100

50

0

150

Parental

A569S

% G

row

th v

s. 0

.1%

EtO

H

Parental A569S

M aximal Growth at 100 nmol/L

130% 180%

C

P < 0.0001

P = 0.0001

Figure 5.

Functional validation of novel ESR1p.A569S mutation demonstrates modestestradiol sensitivity and increasedtamoxifen agonist activity. A, Summaryof ESR1 mutations detected in patient #2from tissue, CTC-DNA, and cfDNA inschematic representation of the encodedERa with the LBD and F domainsindicated. B, MCF-7 ESR1 A569S are moresensitive to estradiol after 5 days whenassessed by crystal violet assay. Insettable indicates EC50 of estradiol in pmol/L.C, Tamoxifen has increased agonistactivity in MCF-7 cells overexpressingER-A569S. Inset table indicates maximalgrowth at 100 nmol/L tamoxifen. Assayswere performed in triplicates, with Pvalues error bars (SE) indicated.






intrapatient heterogeneity of CTC protein expression, includingER, BCL2, HER2, Ki-67, and PI3K from patients with MBC(25, 39–42). These data, coupled with the results of the currentreport, demonstrate similar heterogeneity in both somatic muta-tions and CNAs through simultaneous profiling and suggest thatboth genetic and protein data should be monitored to truly tailorprecision therapy.

Of note, 3 of 46 individual CTCs included in our cohortmatched the IHC definition of CTCs (CKþ, CD45�), but did notcontain any high-confidence mutations, indels, or high-levelCNAs. These cells were present in patients where other CTCs wereconcordant with tissue metastases. These observations suggestthat benign, nonhematopoietic circulating cells of epithelialorigin may be captured by platforms using antiepithelial enrich-ment/purification methods. Because the CellSearch enrichmentmethod is based on epithelial cell capture by EpCAM expressionand characterized by the presence of CK and lack of CD45expression, it is possible that these cells represent normal circu-lating epithelial cells that were in the patient's blood either as afunction of shedding or during the blood draw in patient withoutcancer (43, 44). Likewise, we confirmed (by Sanger sequencing)the presence of a subset of mutations that were detected exclu-sively in several other CTCs and not matched tissue samples.However, due to the complete discordance between alterations inthese CTCs (some of which were pooled) and tissue samples, wecannot exclude that these cells are either of nontumor origin orthat these findings may be technical artifacts. These findingsemphasize the need for CTC platforms to document that detectedCTCs are malignant cells through the identification of pathogno-monic molecular alterations or orthogonal tissue-based valida-tion. We cannot be certain that CTCs captured by the CellSearchassay are the source of tissue metastases. As CellSearch requiresfixation of the captured cells, culture of those cells is not possible.Other investigators have demonstrated presumed CTCs (CKþ/DAPIþ/CD45�) captured using different platforms can, indeed,be cultured ex vivo and in vivo (45, 46). Importantly, althoughnumerous CTC platforms report the ability to detect CTCs"missed" by capture and/or expression-based CTC platforms(3), the CellSearch platform has been clinically validated for itsprognostic role in breast, colon, prostate, and lung cancer (6, 7,47). Combined with the high concordance of known oncogenicgenomic alterations between CellSearch- identified CTCs andtissuemetastases, our results support these CTCs as havingmalig-nant potential, at least in part. In addition, our results hereinsupport the vast majority of identified CTCs as bona fide tumorcells based on somatic molecular alteration concordance withtissue metastases.

In an effort to comprehensively profile the CTC genomiclandscape in an endocrine therapy–resistant patient, we profiled32 individual and pooled CTC samples from patient #2, who hadendocrine therapy–resistant metastatic lobular breast cancer.Both CTCs and tissue metastases showed concordant CDH1mutations, as expected. However, although 20 of 26 single CTCsharbored ESR1 p.Y537S mutations consistent with that detectedin the tissue metastasis, a single CTC instead harbored a novelunreported ESR1 p.A569S mutation confirmed by Sangersequencing of the amplified genomic DNA. Intriguingly, A569is localized to the F domain of the ER receptor, which differs fromthe more common mutations, such as Y537S and D538G, iden-tified in the LBD(Fig. 5A). LBDmissensemutations lead to ligand-independent constitutive ER activation (11–14, 48) and confer

relative resistance to tamoxifen and fulvestrant. In contrast, ESR1p.A569S-expressingMCF-7 cells were estrogen dependent andnottamoxifen nor fulvestrant resistant. However, the ESR1 p.A569Smutation conferred a modest, but statistically significant,increased agonist activity to tamoxifen in the absence of estrogenin MCF-7 cells compared with wild-type ESR1-expressing cells.Interestingly, a mutation in the adjacent amino acid (T570I) wasrecently reported (without functional characterization) in a CTCof an endocrine therapy–treated MBC patient (21). In addition, amutation in the same domain in another steroid hormone recep-tor family member, the androgen receptor (F876L), in prostatecancer cell lines confers agonist activity to the androgen antago-nist enzalutamide (49). It is well established that in breast,tamoxifen is not a pure ER antagonist, but rather serves as aselective estrogen receptor modulator, with mixed ER agonismand antagonism, including in estrogen-deprived ERþ culturedhuman breast cancer cells. Multiple mechanisms of this duality intamoxifen effect have been proposed, including upregulation of avariety of other genes such as NFkB and HER2 (50, 51). Of note,the F domain, which harbors our mutation, was shown to beimportant in the agonist and antagonist balance of antiestrogens(52). We speculate that the ESR1 p.A569S mutation may alsocontribute to this paradoxical effect of tamoxifen on ER.

Notably, this modest activity of the ESR1 p.A569S mutation isconsistent with the rarity of this mutation among the patient'sCTC burden compared with the highly active p.Y537S. However,given the interest in developing therapeutic approaches for ESR1LBD mutations, including ongoing clinical trials based on ESR1LBD mutation detection (e.g., NCT03079011), our data demon-strate that such patients likely have rare tumor subclones capableof expanding upon selective pressure. Of note, ddPCR profiling ofprevious tissue samples and cfDNA demonstrated even morecomplexity in regards to ESR1 status in this patient. Taken togeth-er, these results highlight the diverse repertoire of endocrinetherapy resistance mechanisms in patients with advanced, endo-crine-treated ERþ breast cancer, analogous to those observed inrapid autopsy series of hormonally treated prostate cancer (53).

We cannot entirely exclude that the observed ESR1 (A569S)mutation may be a technical artifact from CTC fixation, WGA, orNGS (54). However, this uniquemutationwas found at uniallelicvariant fraction of 0.56 in a cell with the same CDH1 mutationsobserved in other CTCs and the tissue metastasis. Likewise,detection by Sanger sequencing in pre-NGS WGA DNA stronglyargues against this being an NGS error. In addition, the fact thatthe other most common ESR1mutation (Y537S) is not present inthis CTC underlines that these mutations are mutually exclusive,and multiple circulating ESR1 mutations in the same patient arewell described inMBC, with 17%of patients having >2 detectableESR1 LBD mutations in cfDNA (55). Finally, the modest, butobservable, functional activity of the mutant supports this muta-tion as a bona fide mutation present as a minor circulatingsubclone rather than a technical artifact.

Our study is limited due to relatively small sample size. Indeed,the heterogeneity that we have detected has effectively reducedeach patient to an "n of 1." Furthermore, this investigation wasperformed as a pilot, principally to determine whether DNA fromCTCs that have been fixed and archived in CellSearch cartridgesfor some period of time could be harvested, purified, andsequenced with analytic fidelity, and subsequently, whether thesedata can be reliably compared with genomic analysis of tissuecollected in roughly the same timeframe. Importantly, our data

Paoletti et al.





provide proof-of-concept evidence that this strategy is viable, andstudies to determine whether our approach can be used to predictpatient outcome or guide therapy are ongoing.

In conclusion, we have demonstrated the ability to purify andcomprehensively sequence archived single CTC from an FDA-cleared CTC detection platform (CellSearch) coupled with anautomated technique for single-cell purification and analysis(DEPArray). Furthermore, we have reproducibly performedsimultaneous somatic mutation and CNA profiling of these cells.We observed high, but not absolute, concordance between somat-ic alterations in CTCs and matched tissue metastases subjected toWES. The resultant discordancemay identify potentially clinicallyinformative/actionable alterations exclusively present either inCTCs or tissue metastases, supporting the complementary natureof these approaches. Through sequencing >20 CTCs in a singlepatient with endocrine therapy–resistant lobular breast carcino-ma, we identified distinct ESR1 mutations in individual CTCs,including a novel, modestly active ESR1mutation present in onlya single cell compared with the strongly activating ESR1mutationpresent in the vast majority of the patient's CTCs and a tissuemetastasis. Taken together, our results support the feasibility ofsimultaneous somatic mutation and copy number profiling fromarchivedCTCs,whichmaybe used to track resistancemechanismsunder selective pressure of targeted therapy. We speculate thatthese findings could lead to identification of both CTC proteinexpression and genomic abnormalities that could serve as poten-tial therapeutic targets, and complement tissue or cfDNA-basedprecision oncology approaches.

Disclosure of Potential Conflicts of InterestJ.M. Larios is the resident physician at ProvidenceHospital. D.G. Thomas is a

consultant/advisory board member for Resonant Therapeutics. A.M. Chinnai-yan is a consultant/advisory board member for Tempus. D.F. Hayes reportsreceiving commercial research grants from AstraZeneca, Menarini/Silicon Bio-Systems, Pfizer, and Veridex (Johnson & Johnson) and has provided experttestimony for a patent licensed to Janssen but now Menarini. S.A. Tomlins is aconsultant/lab director at Strata Oncology and reports receiving a commercialresearch grant from Life Technologies/Thermo Fisher Scientific. No potentialconflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: C. Paoletti, A.K. Cani, J.M. Larios, M.D. Johnson,B.H. Park, D.F. Hayes, J.M. Rae, S.A. TomlinsDevelopment of methodology: C. Paoletti, A.K. Cani, J.M. Larios, D.H. Hovel-son, M. Yazdani, V. Sero, N. Tokudome, D.F. Hayes, J.M. Rae, S.A. TomlinsAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.):C. Paoletti, A.K. Cani, J.M. Larios, D. Chu, M. Yazdani,A.R. Blevins, V. Sero, N. Tokudome, D.G. Thomas, A.F. Schott, Y.-M. Wu,D.R. Robinson, F.Z. Bischoff, M.D. Johnson, D.F. Hayes, J.M. RaeAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): C. Paoletti, A.K. Cani, J.M. Larios, D.H. Hovelson,E.P. Darga, E.M. Cannell, P.J. Baratta, D. Chu,M. Yazdani, V. Sero,D.G. Thomas,R. Lonigro, D.R. Robinson, F.Z. Bischoff, M.D. Johnson, B.H. Park, D.F. Hayes,J.M. RaeWriting, review, and/or revision of the manuscript: C. Paoletti, A.K. Cani,J.M. Larios, D.H. Hovelson, K. Aung, E.P. Darga, E.M. Cannell, D. Chu,D.G. Thomas, A.F. Schott, Y.-M. Wu, A.M. Chinnaiyan, F.Z. Bischoff,M.D. Johnson, B.H. Park, D.F. Hayes, J.M. Rae, S.A. TomlinsAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): C. Paoletti, A.K. Cani, J.M. Larios, D.H. Hovel-son, K. Aung, E.P. Darga, E.M. Cannell, C.-J. Liu, C. Gersch, Y.-M. Wu,A.M. Chinnaiyan, D.F. HayesStudy supervision: A.K. Cani, D.F. Hayes

AcknowledgmentsThis work was supported by Veridex/Janssen, LLC, Menarini Silicon Biosys-

tems, Inc., Fashion Footwear Charitable Foundation of NewYork/QVCPresentsShoes on Sale (to D.F. Hayes). D. Chu and B.H. Park were supported by theCommonwealth Foundation, NIH CA194024, and the Breast Cancer ResearchFoundation. A.K. Cani was supported by the NIH Training Program in Trans-lational Research T32-GM113900. S.A. Tomlins was supported by the A. AlfredTaubman Medical Research Institute. We are grateful to all the patients whogenerously volunteered to participate in the study. We thank the research nursesand study coordinators for their efforts on behalf of the patients. We would liketo thank theMI-ONCOSEQ team (MichiganOncology Sequencing Center). Wewould also like to acknowledge thoughtful suggestions from Dr. BenitaKatzenellenbogen.

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Received September 6, 2017; revisedOctober 19, 2017; acceptedDecember 7,2017; published OnlineFirst December 12, 2017.

References1. Alix-Panabieres C, Pantel K. Clinical applications of circulating tumor

cells and circulating tumor DNA as liquid biopsy. Cancer Discov2016;6:479–91.

2. Paoletti C, Smerage J, Hayes DF. Circulating tumor cells as a marker ofprognosis. Princip Prac Oncol 2012;26:1–8.

3. Paoletti C, Hayes DF. Circulating tumor cells. Adv Exp Med Biol 2016;882:235–58.

4. Parsons HA, Beaver JA, Park BH. Circulating plasma tumor DNA. Adv ExpMed Biol 2016;882:259–76.

5. Jankowitz RC, Oesterreich S, Lee AV, Davidson NE. New strategies inmetastatic hormone receptor-positive breast cancer: searching for biomar-kers to tailor endocrine and other targeted therapies. Clin Cancer Res2017;23:1126–31.

6. Cristofanilli M. Circulating tumor cells, disease progression, and survivalin metastatic breast cancer. N Engl J Med 2004;351:781–91.

7. Smerage JB, Barlow WE, Hortobagyi GN, Winer EP, Leyland-Jones B,Srkalovic G, et al. Circulating tumor cells and response to chemo-therapy in metastatic breast cancer: SWOG S0500. J Clin Oncol 2014;32:3483–9.

8. Jamal-Hanjani M, Wilson GA, McGranahan N, Birkbak NJ, Watkins TBK,Veeriah S, et al. Tracking the evolutionof non-small-cell lung cancer.NEnglJ Med 2017;376:2109–21.

9. AbboshC,BirkbakNJ,WilsonGA, Jamal-HanjaniM,Constantin T, Salari R,et al. Phylogenetic ctDNA analysis depicts early-stage lung cancer evolu-tion. Nature 2017;545:446–51.

10. Roychowdhury S, Iyer MK, Robinson DR, Lonigro RJ, Wu YM, Cao X, et al.Personalized oncology through integrative high-throughput sequencing: apilot study. Sci Transl Med 2011;3:111ra21.

11. RobinsonDR,WuYM, Vats P, Su F, Lonigro RJ, Cao X, et al. Activating ESR1mutations in hormone-resistant metastatic breast cancer. Nat Genet2013;45:1446–51.

12. Nardone A, De Angelis C, Trivedi MV, Osborne CK, Schiff R. The changingrole of ER in endocrine resistance. Breast 2015;24:S60–6.

13. ToyW, Shen Y,WonH,Green B, Sakr RA,WillM, et al. ESR1 ligand-bindingdomain mutations in hormone-resistant breast cancer. Nat Genet2013;45:1439–45.

14. Li S, Shen D, Shao J, Crowder R, Liu W, Prat A, et al. Endocrine-therapy-resistant ESR1 variants revealed by genomic characterization of breast-cancer-derived xenografts. Cell Rep 2013;4:1116–30.

15. ZhangQX, Borg A,Wolf DM,Oesterreich S, Fuqua SA. An estrogen receptormutantwith strong hormone-independent activity fromametastatic breastcancer. Cancer Res 1997;57:1244–9.

16. Merenbakh-Lamin K, Ben-Baruch N, Yeheskel A, Dvir A, Soussan-GutmanL, Jeselsohn R, et al. D538G mutation in estrogen receptor-alpha: a novel






mechanism for acquired endocrine resistance in breast cancer. Cancer Res2013;73:6856–64.

17. Chu D, Paoletti C, Gersch C, VanDenBerg DA, Zabransky DJ, Cochran RL,et al. ESR1 mutations in circulating plasma tumor DNA from metastaticbreast cancer patients. Clin Cancer Res 2016;22:993–9.

18. Shaw JA,GutteryDS,Hills A, Fernandez-GarciaD, Page K, Rosales BM, et al.Mutation analysis of cell-free DNA and single circulating tumor cells inmetastatic breast cancer patients with high circulating tumor cell counts.Clin Cancer Res 2017;23:88–96.

19. MuZ, Benali-FuretN,UzanG, Znaty A, Ye Z, PaolilloC, et al. Detection andcharacterization of circulating tumor associated cells in metastatic breastcancer. Int J Mol Sci 2016;17:88.

20. Carter L, Rothwell DG, Mesquita B, Smowton C, Leong HS, Fernandez-Gutierrez F, et al. Molecular analysis of circulating tumor cells identifiesdistinct copy-number profiles in patients with chemosensitive and che-morefractory small-cell lung cancer. Nat Med 2017;23:114–19.

21. Paolillo C,Mu Z, Rossi G, SchiewerMJ, Nguyen T, Austin L, et al. Detectionof activating estrogen receptor gene (ESR1) mutations in single circulatingtumor cells. Clin Cancer Res 2017;23:6086–93.

22. Gulbahce N, Magbanua MJM, Chin R, Agarwal MR, Luo X, Liu J, et al.Quantitative whole genome sequencing of circulating tumor cells enablespersonalized combination therapy of metastatic cancer. Cancer Res2017;77:4530–41.

23. Kidess-Sigal E, Liu HE, Triboulet MM, Che J, Ramani VC, Visser BC, et al.Enumeration and targeted analysis of KRAS, BRAF and PIK3CA mutationsin CTCs captured by a label-free platform: comparison to ctDNA and tissuein metastatic colorectal cancer. Oncotarget 2016;7:85349–64.

24. Heitzer E, AuerM,GaschC, PichlerM,Ulz P,Hoffmann EM, et al. Complextumor genomes inferred from single circulating tumor cells by array-CGHand next-generation sequencing. Cancer Res 2013;73:2965–75.

25. Paoletti C,MunizMC, ThomasDG,Griffith KA, Kidwell KM, TokudomeN,et al. Development of circulating tumor cell-endocrine therapy index inpatients with hormone receptor-positive breast cancer. Clin Cancer Res2015;21:2487–98.

26. Peeters DJ, De Laere B, Van den EyndenGG, Van Laere SJ, Rothe F, IgnatiadisM, et al. Semiautomated isolation and molecular characterisation of singleor highly purified tumour cells from CellSearch enriched blood samplesusing dielectrophoretic cell sorting. Br J Cancer 2013;108:1358–67.

27. Polzer B, Medoro G, Pasch S, Fontana F, Zorzino L, Pestka A, et al.Molecular profiling of single circulating tumor cells with diagnostic inten-tion. EMBO Mol Med 2014;6:1371–86.

28. Cani AK, Hovelson DH, McDaniel AS, Sadis S, Haller MJ, Yadati V, et al.Next-gen sequencing exposes frequent MED12 mutations and actionabletherapeutic targets in phyllodes tumors. Mol Cancer Res 2015;13:613–9.

29. Hovelson DH, McDaniel AS, Cani AK, Johnson B, Rhodes K, Williams PD,et al. Development and validation of a scalable next-generation sequencingsystem for assessing relevant somatic variants in solid tumors. Neoplasia2015;17:385–99.

30. Johnson MD, Zuo H, Lee KH, Trebley JP, Rae JM, Weatherman RV, et al.Pharmacological characterization of 4-hydroxy-N-desmethyl tamoxifen, anovel activemetaboliteof tamoxifen.BreastCancerResTreat2004;85:151–9.

31. Conley BA, Doroshow JH. Molecular analysis for therapy choice: NCIMATCH. Semin Oncol 2014;41:297–9.

32. HovelsonDH, Liu CJ,Wang Y, KangQ,Henderson J, Gursky A, et al. Rapid,ultra low coverage copy number profiling of cell-free DNA as a precisiononcology screening strategy. Oncotarget 2017;8:89848–66.

33. Grasso C, Butler T, Rhodes K, Quist M, Neff TL, Moore S, et al. Assessingcopy number alterations in targeted, amplicon-based next-generationsequencing data. J Mol Diagn 2015;17:53–63.

34. Gorges TM, Kuske A, Rock K, Mauermann O, Muller V, Peine S, et al.Accession of tumor heterogeneity by multiplex transcriptome profiling ofsingle circulating tumor cells. Clin Chem 2016;62:1504–15.

35. Schiavon G, Hrebien S, Garcia-Murillas I, Cutts RJ, Pearson A, Tarazona N,et al. Analysis of ESR1 mutation in circulating tumor DNA demonstratesevolution during therapy for metastatic breast cancer. Sci Transl Med2015;7:313ra182.

36. Fribbens C, O'Leary B, Kilburn L, Hrebien S, Garcia-Murillas I, Beaney M,et al. Plasma ESR1 mutations and the treatment of estrogen receptor-positive advanced breast cancer. J Clin Oncol 2016;34:2961–8.

37. Ellsworth DL, Blackburn HL, Shriver CD, Rabizadeh S, Soon-Shiong P,Ellsworth RE. Single-cell sequencing and tumorigenesis: improvedunderstanding of tumor evolution and metastasis. Clin Transl Med2017;6:15.

38. Pestrin M, Salvianti F, Galardi F, De Luca F, Turner N, Malorni L, et al.Heterogeneity of PIK3CA mutational status at the single cell level incirculating tumor cells from metastatic breast cancer patients. Mol Oncol2015;9:749–57.

39. Paoletti C, Larios JM, Muniz MC, Aung K, Cannell EM, Darga EP, et al.Heterogeneous estrogen receptor expression in circulating tumor cellssuggests diverse mechanisms of fulvestrant resistance. Mol Oncol 2016;10:1078–85.

40. Babayan A, Hannemann J, Spotter J, Muller V, Pantel K, Joosse SA.Heterogeneity of estrogen receptor expression in circulating tumorcells from metastatic breast cancer patients. PLoS One 2013;8:e75038.

41. Riethdorf S, Muller V, Zhang L, Rau T, Loibl S, Komor M, et al. Detectionand HER2 expression of circulating tumor cells: prospective monitoring inbreast cancer patients treated in the neoadjuvant GeparQuattro trial. ClinCancer Res 2010;16:2634–45.

42. Gasch C, Oldopp T, Mauermann O, Gorges TM, Andreas A, Coith C, et al.Frequent detectionof PIK3CAmutations in single circulating tumor cells ofpatients suffering fromHER2-negativemetastatic breast cancer. MolOncol2016;10:1330–43.

43. Allard WJ, Matera J, Miller MC, Repollet M, Connelly MC, Rao C, et al.Tumor cells circulate in the peripheral blood of all major carcinomas butnot in healthy subjects or patientswith nonmalignant diseases. ClinCancerRes 2004;10:6897–904.

44. Pantel K, Deneve E, Nocca D, Coffy A, Vendrell JP, Maudelonde T, et al.Circulating epithelial cells in patients with benign colon diseases. ClinChem 2012;58:936–40.

45. HodgkinsonCL,MorrowCJ, Li Y,Metcalf RL, Rothwell DG, Trapani F, et al.Tumorigenicity and genetic profiling of circulating tumor cells in small-celllung cancer. Nat Med 2014;20:897–903.

46. Wang Z, Wu W, Wang Z, Tang Y, Deng Y, Xu L, et al. Ex vivo expansion ofcirculating lung tumor cells based on one-step microfluidics-based immu-nomagnetic isolation. Analyst 2016;141:3621–5.

47. Bidard FC, PeetersDJ, FehmT,Nole F,Gisbert-CriadoR,MavroudisD, et al.Clinical validity of circulating tumour cells in patients with metastaticbreast cancer: a pooled analysis of individual patient data. Lancet Oncol2014;15:406–14.

48. Jeselsohn R, Buchwalter G, De Angelis C, Brown M, Schiff R. ESR1 muta-tions-a mechanism for acquired endocrine resistance in breast cancer. NatRev Clin Oncol 2015;12:573–83.

49. BalbasMD, EvansMJ,HosfieldDJ,Wongvipat J, Arora VK,Watson PA, et al.Overcoming mutation-based resistance to antiandrogens with rationaldrug design. Elife 2013;2:e00499.

50. Osipo C, Gajdos C, Liu H, Chen B, Jordan VC. Paradoxical action offulvestrant in estradiol-induced regression of tamoxifen-stimulated breastcancer. J Natl Cancer Inst 2003;95:1597–608.

51. Massarweh S, Osborne CK, Creighton CJ, Qin L, Tsimelzon A, Huang S,et al. Tamoxifen resistance in breast tumors is driven by growth factorreceptor signaling with repression of classic estrogen receptor genomicfunction. Cancer Res 2008;68:826–33.

52. Montano MM, Muller V, Trobaugh A, Katzenellenbogen BS. The carboxy-terminal F domain of the human estrogen receptor: role in the transcrip-tional activity of the receptor and the effectiveness of antiestrogens asestrogen antagonists. Mol Endocrinol 1995;9:814–25.

53. Gundem G, Van Loo P, Kremeyer B, Alexandrov LB, Tubio JM, Papaem-manuil E, et al. The evolutionary history of lethal metastatic prostatecancer. Nature 2015;520:353–7.

54. Babayan A, Alawi M, Gormley M, Muller V, Wikman H, McMullin RP,et al. Comparative study of whole genome amplification and nextgeneration sequencing performance of single cancer cells. Oncotarget2017;8:56066–80.

55. Spoerke JM, Gendreau S, Walter K, Qiu J, Wilson TR, Savage H, et al.Heterogeneity and clinical significance of ESR1 mutations in ER-positivemetastatic breast cancer patients receiving fulvestrant. Nat Commun2016;7:11579.


Paoletti et al.




2018;78:1110-1122. Published OnlineFirst December 12, 2017.Cancer Res Costanza Paoletti, Andi K. Cani, Jose M. Larios, et al. Resistance MechanismsCirculating Breast Cancer Tumor Cells Documents Heterogeneous Comprehensive Mutation and Copy Number Profiling in Archived

Updated version

10.1158/0008-5472.CAN-17-2686doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2018/03/24/0008-5472.CAN-17-2686.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/78/4/1110.full#ref-list-1

This article cites 55 articles, 20 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/78/4/1110.full#related-urls

This article has been cited by 1 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://cancerres.aacrjournals.org/content/78/4/1110To request permission to re-use all or part of this article, use this link



http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-17-2686

http://cancerres.aacrjournals.org/content/suppl/2018/03/24/0008-5472.CAN-17-2686.DC1

http://cancerres.aacrjournals.org/content/78/4/1110.full#ref-list-1

http://cancerres.aacrjournals.org/content/78/4/1110.full#related-urls

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

mailto:[email protected]

http://cancerres.aacrjournals.org/content/78/4/1110
