Integrated Analysis of Multiple Biomarkers fromCirculating Tumor Cells Enabled by Exclusion-Based Analyte IsolationJamie M. Sperger1, Lindsay N. Strotman2, Allison Welsh3, Benjamin P. Casavant2,Zachery Chalmers3, Sacha Horn4, Erika Heninger4, Stephanie M. Thiede4, Jacob Tokar2,Benjamin K. Gibbs4, David J. Guckenberger2, Lakeesha Carmichael5, Scott M. Dehm6,Philip J. Stephens3, David J. Beebe2,4, Scott M. Berry2, and Joshua M. Lang1,4
Abstract
Purpose: There is a critical clinical need for new predictive andpharmacodynamic biomarkers that evaluate pathway activity inpatients treated with targeted therapies. A microscale platformknown as VERSA (versatile exclusion-based rare sample analysis)was developed to integrate readouts across protein, mRNA, andDNA in circulating tumor cells (CTC) for a comprehensive anal-ysis of the androgen receptor (AR) signaling pathway.
Experimental Design: Utilizing exclusion-based sample prep-aration principles, a handheld chip was developed to performCTC capture, enumeration, quantification, and subcellular local-ization of proteins and extraction of mRNA and DNA. Thistechnology was validated across integrated endpoints in cell linesand a cohort of patients with castrate-resistant prostate cancer(CRPC) treated with AR-targeted therapies and chemotherapies.
Results: The VERSA was validated in cell lines to analyze ARprotein expression, nuclear localization, and gene expression
targets. When applied to a cohort of patients, radiographicprogression was predicted by the presence of multiple AR splicevariants and activity in the canonical AR signaling pathway.AR protein expression and nuclear localization identified phe-notypic heterogeneity. Next-generation sequencing with theFoundationOne panel detected copy number changes andpoint mutations. Longitudinal analysis of CTCs identifiedacquisition of multiple AR variants during targeted treatmentsand chemotherapy.
Conclusions: Complex mechanisms of resistance to AR-tar-geted therapies, across RNA, DNA, and protein endpoints, existin patients with CRPC and can be quantified in CTCs. Inter-rogation of the AR signaling pathway revealed distinct patternsrelevant to tumor progression and can serve as pharmacody-namic biomarkers for targeted therapies. Clin Cancer Res; 23(3);746–56. �2016 AACR.
IntroductionThe landscape of treatment options for patients with solid
tumors has changed dramatically. In the last 5 years, more than30 new agents have been approved by the FDA, and the need totailor treatment recommendations to each individual has neverbeen greater (1). However, this personalization of cancer thera-pies requires biomarkers that predict therapeutic benefit, identify
emerging mechanisms of resistance, and tailor subsequent treat-ment strategies to continually evolving tumors (2). Successfuldevelopment of predictive and pharmacodynamic biomarkerssuitable for these purposes requires frequent sampling cellsthroughout the course of therapy. This approach is rarely feasiblefor patients with solid tumors given the invasive nature of tumorbiopsies. Circulating tumor cells (CTC) are shed into peripheralcirculation from primary and metastatic tumor sites, and enu-meration of CTCs in prostate cancer is prognostic of patientoutcomes (1, 3–6). However, enumeration does not identify theunderlying mechanisms of resistance that develop during treat-ment with targeted therapies. Preclinical and clinical studies haveidentified a vast range of resistance mechanisms to therapiestargeting the AR signaling pathway that include genomic andfunctional adaptations. For example, Antonarakis and colleagues(7) identified expression of the splice variant AR-V7 that corre-lated with primary resistance to both abiraterone acetate andenzalutamide. This group then examined AR-V7 longitudinally in14 patients and observed conversion of AR-V7 status during thecourse of treatment, suggesting it may be a dynamic biomarker(8). Others have reported heterogeneity in expression and local-ization of AR protein in CTCs frompatients with castrate-resistantprostate cancer (CRPC; refs. 9, 10). However, the molecularanalyses from these CTC technologies are limited to a singlereadout. It is unclear whether these biomarkers reflect driver or
1Department of Medicine, University ofWisconsin-Madison, Madison,Wisconsin.2Department of Biomedical Engineering, University of Wisconsin-Madison,Madison, Wisconsin. 3Foundation Medicine, Cambridge, Massachusetts. 4Car-bone Cancer Center, University of Wisconsin-Madison, Madison, Wisconsin.5Department of Biostatistics and Medical Informatics, University of Wiscon-sin-Madison, Madison, Wisconsin. 6Masonic Cancer Center and Department ofLaboratory Medicine and Pathology, University of Minnesota, Minneapolis,Minnesota.
Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).
Corresponding Author: Joshua M. Lang, Department of Medicine, University ofWisconsin-Madison, 1111 Highland Avenue, WIMR 7151, Madison, WI 53705.Phone: 608-262-0705; Fax: 608-265-0614; E-mail:[email protected]
passenger alterations in resistant prostate cancer, the extent towhich the AR signaling pathway continues to drive disease pro-gression, or whether these alterations can be therapeuticallytargeted. Integrating these distinct endpoints for CTC enumera-tion, genomic, transcriptomic, and protein endpointsmay furtherfacilitate discovery endpoints of novel biomarkers or biologicalalterations in the development of therapeutic resistance.
Integrating CTC capture with multiplexed molecular analyses,such as gene expression or protein analytics, presents multipletechnological challenges (11).One of themost significant hurdlesis analyte loss that occurs during standard cell staining or nucleicacid extractionwhen isolating rare cell populations (12–14).Mostsystems are designed for a single endpoint while other multi-plexed analyses are add-ons that increase sample loss and impairthroughput through transfer of samples or analytes. Over the lastdecade, sample preparation methods based on traversing animmiscible phase barrier have been developed to minimizeanalyteminipulation (15–18).We have recently developed exclu-sion-based sample preparation (ESP) methods that leverage thedominance of surface tension over gravity at the microscale. ESPutilizes microscale constrictions to stabilize the positioning ofimmiscible fluids side-by-side, creating an immiscible barrierbetween two aqueous fluids. By pulling paramagnetic particles(PMP) bound to specific analytes of interest through these immis-cible barriers, we isolate the analytes from a general samplewithout dilution, splitting, or perturbation (13, 14, 19–23). TheVERSA (versatile exclusion-based rare sample analysis) wasdesigned to use ESP technology for CTC isolation and multi-endpoint analyses (13, 14, 19–22, 24–26), thus achieving mod-ularity without compromising each individual assay.
The VERSA platform integrates CTC capture for any extracel-lular target of interest, extra- and intracellular staining for enu-meration and immunocytochemistry, and mRNA and DNAextraction. We evaluated these integrated endpoints in patientswith advanced prostate cancer for (i) enumeration (27, 28); (ii)gene expression analysis of multiple AR splice variants; (iii)activity in the canonical AR signaling pathway (29, 30); (iv) ARprotein quantification and subcellular localization (31); and (v)AR genomic alterations (32).We identified expression ofmultipleAR variants with correspondingly high activity in the canonical ARpathway in patients progressing on targeted therapies. Longitu-
dinal CTC analysis identifies acquisition of multiple AR variantsand increased activity in the AR signaling pathway despite AR-targeted therapies and chemotherapy.Genomic analysis identifiessimultaneous AR amplifications and point mutations in a subsetof patients, as predicted in preclinical models. These integratedCTC endpoints reveal complex mechanisms of resistance totargeted therapies with corresponding functional alterations inprotein and signaling pathways that would direct a subset ofpatients toward AR-independent therapies.
Materials and MethodsVERSA device manufacturing
The VERSA device was injection molded in two pieces (ProtoLabs). The front and back sides were fabricated separately frompolystyrene to a thickness of 2 and2.5mm.The front side containsthe cell capture well (250 mL), the extracellular staining well (30mL), sieve well (50 mL), oil well (30 mL), and themRNA extractionwell (15mL). Theback side contains the sievewell thatmates to thefront sieve well (50 mL), oil well (30 mL), andDNA extraction well(15 mL). All wells are connected via trapezoid oil wells that have a300-mm depth and height that tapers from 2 to 0.8 mm with achannel above for oil loading. The front and back sides weresolvent bonded using acetonitrile (Sigma-Aldrich; ref. 33)with an8-mm microporous membrane (Part PET8025100, Sterlitech)sandwiched in the sieve well. A pressure-sensitive adhesive filmwas applied to each side of the device (MicroAmp, AppliedBiosystems).
Cell culture. The prostate cancer cell lines 22Rv1 and LNCaPs,were a gift from Dr. Douglas McNeel (University of Wisconsin-Madison) and were authenticated by short tandem repeat profil-ing in March 2014 by DDCMedical. Both cell lines were culturedin Corning Cellgro RPMI 1640 Medium (VWR) containing 10%FBS, 1% pennicilin–streptomycin, 1% sodium pyruvate, and 1%a-MEM. The prostate cancer cell line R1-D567 (from the labora-tory of Dr. Scott Dehm at the University of Minnesota (Mine-apolis, MN; ref. 34) was cultured in Corning Cellgro RPMI 1640Medium and 10% FBS. The R1-D567 cell line was geneticallyengineered to remove exon5, 6, and7of theAR LBD. All cellswerecultured at 37�C and maintained under 5% CO2. Cells werepassaged using a 0.05% trypsin/EDTA solution.
Specificity and sensitivity of RT-PCR TaqMan.Cell dilutions of 100,10, and 1 for R1-AD1, R1-D567, and VCAP cell lines were createdand mRNA extracted as described above. Next, extracted mRNAwas reverse transcribed, amplified using TaqMan PreAmp, andused in TaqMan RT-PCR assays as described above. Thresholdcycle (Ct) values were reported.
Clinical study design and statistical analysisThis was a prospective biomarker study evaluating expression
of AR protein and splice variant expression in CTCs from patientswith CRPC receiving systemic treatment with chemotherapy, AR-targeted therapies, or radium 223. The study was approved by theInstitutional Review Board at the University of Wisconsin (Madi-son, WI), and all patients supplied written informed consent.Patients were required to have histologically confirmed prostateadenocarcinoma, progressive disease despite "castration levels" ofserum testosterone [<50 ng/dL (1.73 nmol/L)] with continuedandrogen deprivation therapy, and documented metastases, as
Translational Relevance
The goal of precisionmedicine is to tailor treatments to eachcancer. Given the complex resistancemechanisms that occur inprostate cancer, it is critical to test protein,DNA, andmRNA forresistance signatures that emerge over the course of treatment.Circulating tumor cells may be an ideal source of tumor cellsfor precision medicine. We report the development of a newhandheld chip that leverages the dominance of surface tensionover gravity at the microscale to integrate cell capture withprotein staining for any target of interest or extraction of bothmRNA and DNA. Complex and emerging resistance mechan-isms were identified in patients with castrate-resistant prostatecancer that canbeused to predict benefit and early resistance totargeted therapies. These assays are now being tested in mul-tiple, prospective clinical trials.
Integrated Analysis of Multiple Biomarkers from CTCs
www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 747
confirmedonCTorbone scanningwith technetium-99m–labeledmethylene diphosphonate. Patients had to have two or morerising serum PSA values obtained 2 or more weeks apart, with thelast value being 2.0 ng/mL or higher, criteria for PSA progressionthat are consistent with Prostate Cancer Clinical Trials WorkingGroup 2 guidelines. Patients with PSA progression underwentrestaging radiographic imaging in the form of bone scan and CTscan of the abdomen/pelvis to evaluate for radiographic progres-sion. All the authors vouch for the completeness and integrity ofthe data and for the fidelity of the study to the clinical protocol.Peripheral blood samples, for analysis of CTCs, were obtainedfrom eligible patients at the time of disease evaluationwith serumPSA and radiographic imaging if documented PSA progression.All the clinical investigators were blinded to AR-V7 status of theparticipants. All the laboratory investigators were blinded toclinical information when determining CTC results. The associ-ation between radiographic progression status with gene expres-sion was evaluated using multivariate logistic regression withtreatment category as a covariate.
VERSA operationBlood specimins were collected in Cellsave (Jansen Diagnos-
tics, fixed) or vacutainer tubes (BD Biosciences, live) with EDTAanticoagulant. Mononuclear cells were isolated with a ficollgradient. EDTA samples were CD45 depleted to improve purityof live cell capture of CTCs. Cellsave samples were fixed using BDCytofix. CTCs were isolated with VERSA using an antibody toEpcam conjugated to paramagnetic particles. Downstream pro-cesses, including immunofluorescent staining and extraction ofmRNA and DNA, are integrated on the VERSA. Details of bloodprocessing, paramagnetic particle preparation, and VERSA oper-ation can be found in the Supplementary Materials andMethods.
AR nuclear localization and AR quantification analysisImages were taken with a 10� objective using a Nikon Eclipse
Ti-E with a ORCA-Flash 4.0 V2 Digital CMOS camera (Hama-matsu) and NIS-Elements AR Microscope Imaging Software(Nikon). Images were processed using Image J. CTCswere definedas having an intact nuclei, EpCAM, or cytokeratin positive andCD45 negative. For AR nuclear localization, a threshold binaryimage was created for both the nuclear and AR stains to establishregions of interest (ROI). The ROIs were overlaid on the AR stainand the mean intensity and area measured. Background wassubtracted from the mean intensity, which was then multipliedby the area to determine the integrated AR intensity. AR nuclearlocalization was determined by dividing the corrected AR inten-sity from the nuclear ROI over the total AR ROI.
Quantitative RT-PCRThe mRNA elution sample containing PMPs was reverse tran-
scribed using a High-Capacity cDNA Reverse Transcriptase Kit(Life Technologies) according to the manufacturer's directionsusing Bio-Rad C1000 Thermo Cycler (Bio-Rad). The RT reaction(12.5 mL) was then amplified for 10 cycles using TaqMan PreAmp(Life Technologies) according to the manufacturer's directionsand diluted 1:5 in 1� TE (10 mmol/L Tris-HCL pH8, 1 mmol/LEDTA). For TaqMan assays, 5 mL of diluted cDNA template wasmixed with 10 mL iTaq Master Mix (Bio-Rad), 1 mL TaqMan GeneExpression Assay (specified in Supplementary Table S1, LifeTechnologies), and 4 mL nuclease-free (NF) water. Each reactionwas amplified for 45 cycles (denatured at 95�C for 15 seconds,
followed by annealing at 60�C for 1minute) using a CFXConnectReal-Time PCR System (Bio-Rad). A table of primers used isavailable in Supplementary Information (Supplementary TableS1). Threshold cycle (Ct) values were reported.
Whole-genome amplificationCTC specimenswere delivered as extractedDNA.DNAwas split
and amplified in multiple reactions using Phi29 enzyme at 30�Cto allow variant consensus calling across splits. After whole-genome amplification (WGA), the amplified DNA was purifiedusing AMPure XP beads (Agencourt), eluted in EB Buffer (Qia-gen), and quantified using Quant-iT PicoGreen dsDNA Assay(Invitrogen). Specimens with adequate yield (>200 ng) and sizeprofile (200 bp–10 kb), as determined using the Agilent 2200TapeStation system, were allowed to proceed for comprehensivegenomic profiling.
Comprehensive genomic profilingIn samples passing primary quality metrics, amplified DNA
(50–200 ng/sample) underwent whole-genome shotgun libraryconstruction and solution hybridization using methods previ-ously developed and validated (35), to capture full exons from315 cancer-related genes, as well as introns from 28 genes fre-quently rearranged in solid tumors. Hybrid capture librariesmeeting yield (>25 nmol/L, PicoGreen ds DNA assay) and sizespecifications (�300 bp, Agilent 2200 TapeStation system) weresequenced to a minimum of 300� median coverage, with >88%of exons achieving>100� coverage, using IlluminaHiSeq250049� 49 paired-end reads. Sequence data were processed and ana-lyzed using a custom pipeline previously described in detail (35)as well as a secondary pipeline developed specifically for use withthe WGA methodologies described above. Base substitutions,short insertions or deletions (indels), and rearrangements weredetected by statistical analysis and local reassembly of mappedreads. Coverage at each target and SNP allele frequencies wereused to estimate genome-wide tumor purity and ploidy. Onlyknown somatic oncogenic variants were reported.
ResultsVERSA design and operation
The design of the VERSA and workflow are described in Fig. 1.TheVERSAdevice is producedusing twopieces of injectedmoldedpolystyrene bonded together using solvent bonding (Fig. 1A). ForCTC capture and isolation from patients with CRPC, functiona-lized PMPs (3) are added to bind CTCs from the buffy coatfraction and binding occurs on-chip (Fig. 1B). CTCs are capturedfrom the residual nucleated cells by moving an external magnetfrom the input well to the extracellular staining well (Fig. 1C).After staining, cells are transferred to the sieve well using theexternal magnet. The sieve well contains an 8-mm porous mem-brane, dividing the well into a front and back chamber. Themembrane allows low-pressure fluid exchanges to facilitateremoval of released and unbound PMPs (2.8 mm) while prevent-ing larger cells (8–20 mm) from passing through the porousmembrane (21). The sieve well is also used to stain intracellularproteins, enabling cell permeabilization, antibody incubation,and fluid exchanges to wash away unbound antibody. The VERSAcan be positioned horizontally for on-chip image acquisition. Weuse the VERSA to identify CTCs asHoechstþ/CD45�/cytokeratinþ
cells and tomeasure both intensity and localizationof AR. Finally,
Sperger et al.
Clin Cancer Res; 23(3) February 1, 2017 Clinical Cancer Research748
the VERSA incorporates the SNARE, a highly sensitive ESP nucleicacid extraction method that allows efficient extraction of bothmRNA and DNA from a single sample (20) enabling pairedgenomic/transcriptomic analysis of rare cells, such as CTCs (Fig.1B and C).
Cell capture and protein analysisThe VERSA captures tumor cells using any capture antibody of
interest. Given the clinical relevance of EpCAM in CTCs frompatientswithprostate cancer,we validated capture efficiency usingan anti-EpCAM antibody. The VERSA demonstrated an averagecapture efficiency of 79.2� 11.6% when isolating approximately25 LnCap cells (Fig. 2A), similar to prior publications (21).Longitudinal analysis of EpCAM capture demonstrates highreproducibility in capture efficency using VERSA (SupplementaryFig. S1). Through intracellular staining and imaging of the AR inthe VERSA, we can quantify AR nuclear localization and intensity.For initial validation of the image processing to determine ARnuclear localization, AR-transfected COS-7 cell lines were treatedwith an AR agonist or a vehicle control. The treated cells hadsignificantly higher AR nuclear localization as compared withuntreated cells (P < 0.0004; Fig. 2B).
Next, we analyzed CTC number, AR localization, and total ARintensity in 17 patients with CRPC. Results are grouped on thebasis of the PSA response to the patient's current treatment at thetime of the blood draw. Cells were stained using Hoechst, anti-CD45, and anti-Epcam antibodies to distinguish CTCs fromhematopoetic cells, as well as an anti-AR antibody to probeexpression andnuclear localization (Fig. 2C). Patients respondingto AR-targeted or chemotherapy treatments showed lower averagenumbers of CTCs (Supplementary Table S2) and percentages ofARwithin the nucleus (37%� 12%), comparedwith patients that
have progressed on AR-targeted therapies (60% � 8%; Fig. 2D).The low CTC number for patients responding to treatment isconsistent with other studies (36) but limits the number of datapoints available to calculate the average AR localization. There-fore, we combined single-cell data points from multiple patientsin four plots corresponding to different stages in treatment,demonstrating heterogeneity across CTCs with evaluation of ARintensity and AR nuclear localization. (Fig. 2E). Patients respond-ing to treatment generally displayed a low total AR intensity andlow percentage of AR localized to the nucleus. However, patientsprogressing on AR-targeting therapies have more CTCs withhigher AR expression/nuclear localization with a unique popu-lation of cells expressing low levels of AR, but highly localized tothenucleus (Fig. 2E, Pts 15 and16). Preliminaryfindings from thisinitial cohort of patients suggest an emerging mechanism ofresistance in AR activity within a subset of CTCs that allows theAR to translocate to the nucleus, despite antagonist therapy. TheVERSA thus permits molecular interrogation of phenotypic het-erogeneity inCTCs, for prospective evaluation in clinical trials as apharmacodynamic biomarker.
Gene expression analysis of prostate cancer cell linesTo evaluate the sensitivity and specificity of gene expression
usingmRNAextracted in theVERSA,weused cell lines that expressdifferent AR splice variants (Fig. 3A). 22rv1 cells express thevariants AR-V1 and AR-V7, whereas the R1-D567 cell line isengineered to express AR-V567ES (34, 37). mRNA was extractedfrom10or 100 cells reverse transcribed, preamplified, and probedwith the described panel of genes. The AR-V1 and AR-V7 variantsare detected in both cell lines at the 10 and 100 cell level. TheAR-567es variant is observed only in the R1-D567 cell line. Asexpected, no detection of the AR LB domain was observed when
Figure. 1.
The VERSA device. The VERSA integrates efficient cell capture with PMP removal, staining, and isolation of mRNA and DNA without dilutive steps. A, The handheldVERSA is filled with colored dye to differentiate the different chambers. B, The VERSA is pictured with boxes designating the well used for capture (red), staining(blue), and nucleic acid isolation (orange). C, A magnet is used to purify PMP-bound CTCs from the input well (pink) through the oil-filled trapezoid into theextracellular staining well (green). After incubation, CTCs are moved into the sieve well (blue), which contains an 8-mm porous membrane, dividing thewell into a front and back chamber. Themembrane allows low-pressure fluid exchanges to facilitate removal of released and unbound PMPswhile preventing cells ofinterest from passing through. The ability to perform multiple fluid exchanges enables cell permeabilization and incubation with antibodies to intracellularantigens. Cells are imaged in device.mRNA is isolated by lysing cells in device, adding oligo-dT PMPs andmovingRNA to the front elutionwell (orangebox, top right).The subsequent addition of silica PMPs with a nuclear lysis buffer enables coextraction of DNA by magnetic transfer of PMPs to the back well.
Integrated Analysis of Multiple Biomarkers from CTCs
www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 749
AR quantification and nuclear localization in CRPC patients. A, Approximately 25 calcein–labeled LNCaP cells were spiked into the input well and imaged.Following VERSA procedure, input, extracellular staining, and sieve wells were again imaged and LNCaPs counted to determine percentage of cells ineach well. B, AR nuclear localization of AR-transfected cos-7 cells stimulated with and without mibolerone (AR agonist). Stimulated cells (n ¼ 30, mean � SD)showed significantly higher AR nuclear localization as compared with unstimulated cells (P < 0.0004). C, A representative patient's peripheral bloodmononuclear cell (PBMC) and CTC is shown here stained with CD45 (PBMC) or AR and cytokeratin (CTC). BF, Brightfield. D, Percent nuclearlocalization is shown for 17 CRPC patients grouped by patient treatment and response. Box plots show average and spread (minimum to maximum) ofthe localization percentage within CTCs for each patient across different classes of therapy. E, For each individual, CTCs within a patient's total ARintensity and AR nuclear localization percentage were plotted for different patient groups.
Sperger et al.
Clin Cancer Res; 23(3) February 1, 2017 Clinical Cancer Research750
probing the R1-D567 cells with primers targeting the boundary ofexons 4 and 5 of theAR, as this cell line has a complete deletion ofexons 5, 6, and 7.
Gene expression in CTCsA biomarker clinical trial at the University of Wisconsin Car-
bone Cancer Center (Madison, WI) enrolled 26 patients withmetastatic prostate cancer, 19 of whom had received or werecurrently being treated with AR signaling pathway inhibitors(Table 1 and Supplementary Table S3). CTCs were captured withEpCAM and stained for intact nuclei, cytokeratin, and CD45. CTCenumeration in the VERSA showed detectable CTCs in 25 of 26patients with a range of 0 to 1,213 CTCs per 7.5 mL of blood (Fig.3B). Patients with both increased serum PSA levels and radio-graphic evidence of disease progression had detectable expressionof full-length AR and multiple AR splice variants, with corre-
spondingly high expression of downstream targets in the ARsignaling pathway (Fig. 3C, Pts 18–22). Detectable expression oftheAR-V7 splice variantwas significantly different in patientswithradiographic progression compared with patients with only PSAprogression or PSA response (71% vs. 5%, P¼ 0.007). Detectableexpression of downstream targets of the AR pathway, FOLH1(PSMA;P¼0.014) and TMPRSS2 (P¼0.030), was also associatedwith radiographic progression. Importantly, expression of otherAR splice variants is foundwithhigh coincidencewithAR-V7,withdetectable expression of theAR-V1 variant in 4 of these 7 patients.Of the 2 patients without expression of AR variants (Pts 23 and24), one developed visceralmetastases and had no clinical benefitfrom abiraterone, suggesting that these integrated biomarkersmay identify a subset of patients with progressive disease notdependent on canonical AR signaling. Detectable expression ofKLK3 (PSA) and ACPP (PAP) transcripts was not associated with
Radiographic progression No radiographic progressionPSA Progression PSA Stable PSA Improving
Figure 3.
Gene expression analysis of the AR signaling pathway. A, Results from quantitative RT-PCR are presented as Ct values represented as a heatmap. mRNA wasisolated from the indicated number of cells (n¼ 3) from either 22rv1 or R1-D568 cell lines. B, Enumeration of CTCs (defined as the number of cells with intact nuclei,CKþ, CD45�/7.5 mL blood) from a fixed sample run in parallel to gene expression analysis. C, mRNA was isolated from Epcam-positive fraction from 15 mLof EDTA anticoagulated blood. mRNA was reversed transcribed, preamplified, and probed for the AR gene splice junctions, including multiple splice variantsand prostate cancer (PRCA)-specific genes.
Integrated Analysis of Multiple Biomarkers from CTCs
www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 751
disease response, with patients responding to AR-targeted thera-pies also showing expression of these genes. These results areconsistent with recent studies performing single-cell transcrip-tomic analysis, which found similar heterogeneity in expressionof these targets, with many cells lacking detectable expression ofPSMA and PSCA (38, 39). Further prospective clinical trials areneeded to determine whether these signatures are solely due totreatment response or a fundamental biologic difference in thesetumor cells that have entered circulation.
Longitudinal analysisThe ability to followpatients longitudinally usingCTCanalyses
is amajor potential advantage of the liquidbiopsyover traditionalbiopsies of metastatic sites. Figure 4 shows an example of inte-grated longitudinal analysis for 2 patients. Patient 40 (Fig. 4A)presented with metastatic prostate cancer, with lymph node andbone metastases, and a PSA of 35.7. He began chemohormonaltherapy as per the CHAARTED protocol (40), with an almostcomplete resolution in lymphadenopathy and PSA (nadir to0.07) by cycle 3 (Fig. 4A, month 0). However, his serum PSAbegan to rise by cycle 4, with a significant increase in CTC numberand detectable expression of AR-V7 and PSA genes (Fig. 4A,month 1). Bymonth 2 (cycle 6), detection of PSMA and TMPRSS2was identified and associated with continued rise in his serumPSA. Restaging radiographic imaging did not show radiographicprogression, and the patient elected to take a treatment breakgiven toxicities from docetaxel chemotherapy. However, off ther-apy, his PSA rapidly rose with corresponding detection of mul-tiple AR splice variants and increasing activity in the canonical ARsignaling pathway (month 4). An antiandrogen withdrawal wasperformed prior to enrolling on a clinical trial with a PARPinhibitor. AR expression and nuclear localization was highlyheterogeneous, suggesting a subpopulation of tumor cells retain-ing sensitivity to targeted therapy. However, there is clearly aresistant population of tumor cells that fluctuated from highnuclear localization to AR overexpression (Fig. 4A, graphs), coin-ciding with AR variant transcript detection. The second patient(Fig. 4B, Pt 36) rapidly developed disease progression on enza-
lutamidewith acquisition of theAR-V7 variant (month 3).Hewassubsequently treated with cabazitaxel chemotherapy with noevidence of a serum PSA response. Corresponding with his lackof response, no phenotypic alterations in AR nuclear localizationwere identified in this patient, a proposed mechanism by whichtaxane chemotherapymay exert its effect (41). The integrated geneexpression data further support the notion that this patient'sdisease is still driven via the AR signaling pathway, with highexpression of AR variants, TMPRSS2, and PSMA as the patientdeveloped PSA and radiographic progression (Fig. 4B).
Integrated genomic–transcriptomic–phenotypic analysis of theAR pathway
Preclinical and clinical evidence has identified multiple geno-mic alterations in the AR signaling pathway that contribute toresistance to AR inhibitors, extending beyond AR splice variants(42–45). Using the capability of the VERSA platform to sequen-tially extract DNA after mRNA isolation, we performed next-generation sequencing (NGS) using a custom genomic assaydeveloped and validated for compatibility with the Foundatio-nOne panel. This analysis was performed in a subset of patientswith more than 50 CTCs at greater than 20% purity and includedWGA, followed by comprehensive sequencing of full exons from315 cancer-related genes (including AR), plus introns from 28additional genes, allowing simultaneous detection of all classes ofknown oncogenic genomic alterations (base substitutions, shortinsertions and deletions, copy number changes, and rearrange-ments). We utilized this assay with the goal of identifying coin-cident mechanisms of resistance in paired genomic–transcrip-tomic analyses. CTCs isolated from patient 19 (Fig. 5A) not onlyexpressed multiple AR variants (via mRNA) at the time of pro-gression on enzalutamide but also showed genomic evidence forAR amplification [as well as known oncogenic alterations, includ-ing a hotspot TP53mutation (46)] and copy number evidence forthe 8q gain that is a hallmark of the prostate cancer genome (47).This contrasts with the CTCs from patient 28 (Supplementary Fig.S2), which contained the AR T878A point mutation, a well-characterized alteration resulting in a progesterone-activated AR,
Table 1. Patient characteristics stratified by patients progressing on their current treatment versus patients responding
predicting resistance to AR-targeted therapies (44). Patient 18(Fig. 5B) did not have detectable genomic alterations in AR,although again displayed the hallmark 8q gain, as well as a TP53truncating mutation in the tetramerization domain (48). ARprotein analysis revealed broad phenotypic distributions withAR overexpression and AR hypernuclear localization, cooccuringin patients with AR splice variant expression and detected geno-mic AR pathway alterations.
DiscussionThere is a critical need for predictive and pharmacodynamic
biomarkers to guide therapy while simultaneously evaluatingemerging mechanisms of resistance. We achieve these multi-endpoint analytics in a microscale platform utilizing ESP con-cepts that integrate CTC capture with analyte extraction.Termed the VERSA, this handheld device leverages the domi-nance of surface tension over gravity at the microscale toalternate aqueous and immiscible solutions across distinctoperational paths. This enables use of standard reagents andworkflows in the VERSA to manipulate and extract capturedanalytes by hand, as all reagents and samples are loaded withstandard micropipetters. The flexibility and modularity of theworkflow allows collection of live or fixed samples enabling awide variety of assay endpoints, including visualization ofintracellular antigens and extraction of high-quality nucleicacids for use in gene expression and comprehensive NGS-basedanalyses. In this study, magnetic beads conjugated to EpCAMwere used for capture and define CTCs as Hoechst positive,
cytokeratin and/or EpCAM positive, and CD45 negative. Thisplatform further permits capture and staining of CTCs for anymolecular target of interest, including factors that may con-tribute to tumor invasion, proliferation, and treatment resis-tance (e.g., epithelial–mesenchymal transitions). The simplicityand cost effectiveness of this platform allow translation toresearch laboratories with minimal upfront investment, asVERSA operation is performed in a standard biosafety cabinetand does not require large, expensive machinery.
In this report, we use the VERSA to create a comprehensive CTCassay to evaluate preexisting and emerging resistancemechanismsto AR-targeted therapies, including AR amplification, AR pointmutations, and splice variant expression (49). To better under-stand the complexity of these resistance mechanisms, we need toevaluatemultiple analytes in the same patient. We developed twocomplementary workflows that leverage the inherent flexibility ofthe VERSA. First, fixed cell processing allows simultaneous CTCenumeration and assessment of AR protein expression and local-ization. Second, the extraction of mRNA and genomic DNA froma matched live sample allows paired analysis of gene expressionand NGS-based analysis from the same cells. We have shown thatlongitudinal analysis across these endpoints can identify emerg-ing biomarkers of resistance, including evidence for persistentactivity in the canonical AR signaling pathway as well as expres-sion of multiple AR splice variants. The biomarker evaluationsperformed in this study are now being evaluated prospectively inclinical trials with abiraterone acetate (NCT02025010), enzalu-tamide (NCT01942837 and NCT02384382), and VT-464(NCT02445976). The inherent flexibility of the VERSA creates
CAB + Docetaxel AAWD
AR L
ocal
izat
ion
AR Intensity
AR V1
AR V7
AR567es
KRT8
TMPRSS2
KLK3 (PSA)
ACPP (PAP)
FOLH1 (PSMA)
PSCA
RPLP0
00.07
10.42
21.19
441.7
581.4
Month:PSA:
CabazitaxelEnzalutamide
016.3
358.2
4125
A B
AR Intensity
AR
Loc
aliz
atio
n
AR V1
AR V7
AR567es
KRT8
TMPRSS2
KLK3 (PSA)
ACPP (PAP)
FOLH1 (PSMA)
PSCA
RPLP0
AR (Total)AR (LB Domain)
AR (Total)AR (LB Domain)
Figure 4.
Longitudinal analysis of patients with CRPC. Patient 40 (A) and patient 36 (B) were monitored as they progressed through the indicated treatments. Treatment isindicated at the top. The initial blood draw (month 0) corresponds to cycle 3 of chemohormonal therapy for patient 40 (A) and cycle 2 of enzalutamidetreatment for patient 36 (B). For each patient, and at each time point, we present AR nuclear localization versus intensity (each point representing a singleCTC). Below each plot, we show panels with gene expression data (represented as a heatmap of Ct values). CAB, combined androgen blockade; AAWD,antiandrogen withdrawal.
Integrated Analysis of Multiple Biomarkers from CTCs
www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 753
opportunities to evaluate complex mechanisms of resistance inprotein, DNA, andmRNA in any solid tumor of interest and is thesubject of ongoing studies.
The importance of orthogonal endpoints cannot be under-stated in diseases with complex resistance patterns as occurringin prostate cancer (32). Although CTC enumeration or proteinsignatures identify emerging tumor clones resistant to thecurrent therapy, they do not inform on the underlying mechan-isms driving tumor resistance and proliferation. The clearstrength of the AR-V7 splice variant analysis pioneered inAntonarakis and colleagues (7) is the potential link of abiomarker that also acts as the driver of treatment resistance.However, this marker alone does not capture the many complexresistance mechanisms that can lead to progression of CRPC.For example, we have identified patients with radiographicprogression that show minimal expression of the AR and othergenes in the canonical AR pathway. Genomic evaluation in a
subset of patients in this report identified AR amplifications, ARpoint mutations, and other alterations that can contribute toresistance to AR-targeted therapies. This integration of orthog-onal endpoints, across protein, RNA, and DNA readouts, cre-ates opportunities to evaluate the extent that new drugs mod-ulate the AR signaling pathway in the setting of the complexresistance signatures identified in this report.
The last decade has shown a dramatic increase in the numberof therapeutic options for patients with advanced cancer. How-ever, the discriminatory power of our available diagnostics doesnot inform on the relevance or activity of a given pathway todrive cancer progression. One of the most common questionson the relevance of genomic alterations is whether they aredriver or passenger events. Integrating transcriptomic and pro-tein analytics in this context will permit pathway-specific eval-uation that informs on the functional alterations driving dis-ease progression at a given timepoint. It is within the dynamic
AR (Total)
AR (LB Domain)
AR V1
AR V7
AR567es
KRT8
TMPRSS2
KLK3 (PSA)
ACPP (PAP)
FOLH1 (PSMA)
PSCA
RPLP0
AR (Total)
AR (LB Domain)
AR V1
AR V7
AR567es
KRT8
TMPRSS2
KLK3 (PSA)
ACPP (PAP)
FOLH1 (PSMA)
PSCA
RPLP0
Median coverage
% Exons with >100¥ coverage
Known oncogenic short variants (MAF)
254× 88% TP53 V173L (10%)
Median coverage
% Exons with >100¥ coverage
Known oncogenic short variants (MAF)
420x 96% TP53 R342* (8%)
A
B
8q Gain AR amp
8q Gain
AR Intensity104 105 1060
20406080
100
AR
Loc
aliz
atio
n
AR Intensity104 105 1060
20406080
100
AR
Loc
aliz
atio
n
Figure 5.
Multi-parametric analysis of gene expression, genomic profiling, and AR protein analysis from captured CTCs.A,Gene expression and AR protein analysis in a singleblood draw from patient 19, and genomic profiling from a subsequent blood draw from this patient. B, Results shown are from a single blood draw frompatient 18. Gene expression data is shown as a heatmap, in which red indicates high expression, black median expression, green low expression, and gray indicatesno detectable expression. AR protein analysis of CTCs is reported as a graph of AR nuclear localization versus AR intensity. Sequencing coverage and knownoncogenic short variants detected using FoundationOne comprehensive genomic profiling are shown along with corresponding copy number plots (each dotrepresents coverage ratio compared with normal for all exon and intron targets as well as SNPs). Prostate cancer–specific copy number changes are indicated withred arrows (gain of 8q in both patients and focal amplification of AR in patient 19; A).
Sperger et al.
Clin Cancer Res; 23(3) February 1, 2017 Clinical Cancer Research754
environment in which resistance develops that longitudinalassessment of pharmacodynamic biomarkers is needed to notonly identify tumor clones with emerging resistance, but alsoevaluate the extent to which these emerging clones impactpatient outcomes. For example, in lung carcinoma, Sequistand colleagues (50) identified new histologic biomarkers onserial biopsies that mandated a change in treatment away fromtargeted therapies to chemotherapy and radiation. However,disease progression in a subset of these patients was due to theprior lung cancer histology, and further benefit was obtained byrechallenging with the prior targeted therapy. It is within thisprecision therapeutic paradigm that these predictive and phar-macodynamic CTC biomarkers have the greatest potential toimprove patient outcomes. In prostate cancer, the presence oftumor cells with low AR nuclear localization suggests a pop-ulation of tumor cells that retains sensitivity to AR-targetedtherapies despite PSA or radiographic progression. Thus, amultifaceted approach for these patients might improve clinicaloutcomes by combining AR-targeted therapies with chemother-apy, immunotherapy, radiation, or an investigational agent.These CTC biomarkers may have their greatest clinical benefit aspharmacodynamic biomarkers for targeted therapies, suitableto identify pathway activity in the setting of emerging drugresistance to provide greater precision for clinical decisionmaking.
Development of pharmacodynamic biomarkers may alsoimprove evaluation of pharmacokinetic analyses employed inearly-phase clinical trials. These trials perform a critical anal-ysis of drug absorption, exposure, and half-life among others.Linking these pharmacokinetic evaluations with pathway-spe-cific pharmacodynamic biomarkers would enable paired anal-ysis on the drug exposure needed to modulate relevant path-ways. This integration of pharmacokinetics and pharmacody-namics is a clear route to establishing optimal biologicaldosing strategies beyond the current maximally tolerateddosing strategies. For example, evaluation of TMPRSS2 andPSMA expression in CTCs correlates with radiographic pro-gression and reflects increasing activity in the AR signalingpathway despite targeted therapies. It is unknown whetherincreasing drug exposure through dose escalation would besufficient to regain disease control. However, this therapeuticapproach is commonly employed in clinical practice forpatients who were previously dose reduced due to excessivetoxicity. Clinical trials incorporating pharmacodynamic bio-markers with dose escalation have intriguing potential todrive precision medicine beyond matching a drug to a patientbut rather matching drug dose to the individual. It is withinthis context that novel CTC technologies evaluating orthog-onal analytes can drive cancer therapeutics toward precisionmedical care.
Disclosure of Potential Conflicts of InterestL.N. Strotman, B.P. Casavant, and J.M. Lang have ownership interest (includ-
ing patents) in Salus Discovery. A. Welsh, Z. Chalmers, and P.J. Stephens haveownership interest (including patents) in Foundation Medicine. D.J. Gucken-berger is an employee of Salus Discovery and has ownership interest (includingpatents) in Salus Discovery and Tasso. S.M. Dehm is a consultant/advisoryboardmember forMedivation/Astellas.D.J. Beebe and S.M. Berry are employeesof and have ownership interest (including patents) in Salus Discovery. J.M.Lang, D.J. Beebe, S.M. Berry, L.N. Strotman, and B.P. Casavant are listed as co-inventors on a provisional patent application on the circulating tumor celltechnology in this article that is owned by the Wisconsin Alumni ResearchFoundation (WARF) and licensed to Salus Discovery. No potential conflicts ofinterest were disclosed by the other authors.
Authors' ContributionsConception and design: J.M. Sperger, L.N. Strotman, B.P. Casavant,E. Heninger, S.M. Berry, J.M. LangDevelopment of methodology: J.M. Sperger, L.N. Strotman, A. Welsh,B.P. Casavant, Z. Chalmers, E. Heninger, S.M. Thiede, D.J. Guckenberger,D.J. Beebe, J.M. LangAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): J.M. Sperger, L.N. Strotman, A. Welsh, B.P. Casavant,Z. Chalmers, S. Horn, S.M. Thiede, J. Tokar, B.K. Gibbs, J.M. LangAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): J.M. Sperger, L.N. Strotman, A. Welsh, B.P. Casavant,Z. Chalmers, B.K. Gibbs, L. Carmichael, S.M. Dehm, J.M. LangWriting, review, and/or revision of the manuscript: J.M. Sperger,L.N. Strotman, A. Welsh, B.P. Casavant, Z. Chalmers, B.K. Gibbs,L. Carmichael, S.M. Dehm, P.J. Stephens, D.J. Beebe, S.M. Berry, J.M. LangAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A. Welsh, Z. Chalmers, E. Heninger,D.J. Guckenberger, J.M. LangStudy supervision: D.J. Beebe, J.M. Lang
AcknowledgmentsWe would like to thank all patients who participated in this study. We
are also grateful for the help of the UWCCC GU clinical research group,especially Jamie Wiepz, John Cress, Kelly Bush, Jill Kubiak, Mulusew Yaye-hyirad, Dorothea Horvath, Jane Straus, Mary Jane Staab, Dr. Glenn Liu,Dr. Douglas McNeel, Dr. Christos Kyriakopolous and Dr. George Wilding.
Grant SupportThis work was supported by a Movember-Prostate Cancer Foundation
Challenge Award (to J.M. Lang and D.J. Beebe), a Prostate Cancer Founda-tion Young Investigator award (to J.M. Lang), the Bill & Melinda GatesFoundation through the Grand Challenges in Global Health initiative (to S.M.Berry and D.J. Beebe), NIH grant #1R01CA181648 (to S.M. Berry andJ.M. Lang), DOD PCRP grant #W81XWH-12-1-0052 (to J.M. Lang), NIH grant#5R33CA137673 (toD.J. Beebe), andNSFGRFPDGE-0718123 (toD.J. Beebe).
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 May 5, 2016; revised May 26, 2016; accepted June 25, 2016;published OnlineFirst July 11, 2016.
References1. Lang JM, Casavant BP, BeebeDJ. Circulating tumor cells: gettingmore from
less. Sci Transl Med 2012;4:141ps13.2. Armstrong AJ, Tannock IF, de Wit R, George DJ, Eisenberger M, Halabi S.
The development of risk groups inmenwithmetastatic castration-resistantprostate cancer based on risk factors for PSA decline and survival. Eur JCancer 2010;46:517–25.
3. de Bono JS, Scher HI, Montgomery RB, Parker C, Miller MC, Tissing H,et al. Circulating tumor cells predict survival benefit from treatment in
metastatic castration-resistant prostate cancer. Clin Cancer Res 2008;14:6302–9.
4. Danila DC, Anand A, Sung CC, Heller G, Leversha MA, Cao L, et al.TMPRSS2-ERG status in circulating tumor cells as a predictive biomarkerof sensitivity in castration-resistant prostate cancer patients treated withabiraterone acetate. Eur Urol 2011;60:897–904.
5. Scher HI, Morris MJ, Larson S, Heller G. Validation and clinical utility ofprostate cancer biomarkers. Nat Rev Clin Oncol 2013;10:225–34.
www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 755
Integrated Analysis of Multiple Biomarkers from CTCs
6. Armstrong AJ, Marengo MS, Oltean S, Kemeny G, Bitting RL, Turnbull JD,et al. Circulating tumor cells from patients with advanced prostate andbreast cancer display both epithelial and mesenchymal markers. MolCancer Res 2011;9:997–1007.
7. Antonarakis ES, Lu C, Wang H, Luber B, Nakazawa M, Roeser JC, et al. AR-V7 and resistance to enzalutamide and abiraterone in prostate cancer. NEngl J Med 2014;371:1028–38.
8. Nakazawa M, Lu C, Chen Y, Paller CJ, Carducci MA, Eisenberger MA, et al.Serial blood-based analysis ofAR-V7 inmenwith advancedprostate cancer.Ann Oncol 2015;26:1859–65.
9. Reyes EE, VanderWeeleDJ, IsikbayM,DugganR,Campanile A, StadlerWM,et al.Quantitative characterization of androgen receptor protein expressionand cellular localization in circulating tumor cells from patients withmetastatic castration-resistant prostate cancer. J Transl Med 2014;12:313.
10. Crespo M, van Dalum G, Ferraldeschi R, Zafeiriou Z, Sideris S, Lorente D,et al. Androgen receptor expression in circulating tumour cells fromcastration-resistant prostate cancer patients treated with novel endocrineagents. Br J Cancer 2015;112:1166–74.
12. Miller MC, Doyle GV, Terstappen LW. Significance of circulating tumorcells detected by the CellSearch System in patients with metastatic breastcolorectal and prostate cancer. J Oncol 2010;2010:617421.
13. Berry SM,Alarid ET, BeebeDJ.One-step purificationof nucleic acid for geneexpression analysis via immiscible filtration assisted by surface tension(IFAST). Lab Chip 2011;11:1747–53.
14. Berry SM, Strotman LN, Kueck JD, Alarid ET, Beebe DJ. Purification of cellsubpopulations via immiscible filtration assisted by surface tension(IFAST). Biomed Microdevices 2011;13:1033–42.
15. ShikidaMTK, InouchiK,HondaH, SatoK.Usingwettability and interfacialtension to handle droplets of magnetic beads in amicro-chemical-analysissystem. Sens Actuators B 2006;113:563–9.
16. Sur K, McFall SM, Yeh ET, Jangam SR, Hayden MA, Stroupe SD, et al.Immiscible phase nucleic acid purification eliminates PCR inhibitors witha single pass of paramagnetic particles through a hydrophobic liquid. JMolDiagn 2010;12:620–8.
17. Chen H, Stanley E, Jin S, Zirkin BR. Stem Leydig cells: from fetal to agedanimals. Birth Defects Res C Embryo Today 2010;90:272–83.
18. BordelonH, AdamsNM, KlemmAS, Russ PK,Williams JV, Talbot HK, et al.Development of a low-resource RNA extraction cassette based on surfacetension valves. ACS Appl Mater Interfaces 2011;3:2161–8.
19. Strotman LN, LinG, Berry SM, Johnson EA, BeebeDJ. Facile and rapidDNAextraction and purification from food matrices using IFAST (immisciblefiltration assisted by surface tension). Analyst 2012;137:4023–8.
20. Strotman L,O'Connell R, Casavant BP, Berry SM, Sperger JM, Lang JM, et al.Selective nucleic acid removal via exclusion (SNARE): capturingmRNAandDNA from a single sample. Anal Chem 2013;85:9764–70.
21. Casavant BP, Guckenberger DJ, Berry SM, Tokar JT, Lang JM, Beebe DJ. TheVerIFAST: an integrated method for cell isolation and extracellular/intra-cellular staining. Lab Chip 2013;13:391–6.
22. Casavant BP, Guckenberger DJ, Beebe DJ, Berry SM. Efficient samplepreparation from complex biological samples using a sliding lid forimmobilized droplet extractions. Anal Chem 2014;86:6355–62.
23. Beebe DJ, Mensing GA, Walker GM. Physics and applications of micro-fluidics in biology. Annu Rev Biomed Eng 2002;4:261–86.
24. Casavant BP, Strotman LN, Tokar JJ, Thiede SM, Traynor AM, Ferguson JS,et al. Paired diagnostic and pharmacodynamic analysis of rare non-smallcell lung cancer cells enabled by the VerIFAST platform. Lab Chip 2013;14:99–105.
28. Ramiah V, George DJ, Armstrong AJ. Clinical endpoints for drug devel-opment in prostate cancer. Curr Opin Urol 2008;18:303–8.
29. Page ST, Lin DW, Mostaghel EA, Marck BT, Wright JL, Wu J, et al. Dihy-drotestosterone administration does not increase intraprostatic androgen
concentrations or alter prostate androgen action in healthy men: a ran-domized-controlled trial. J Clin Endocrinol Metab 2011;96:430–7.
30. Sun S, Sprenger CC, Vessella RL, Haugk K, Soriano K, Mostaghel EA,et al. Castration resistance in human prostate cancer is conferred by afrequently occurring androgen receptor splice variant. J Clin Invest2010;120:2715–30.
31. Heinlein CA, Chang C. Androgen receptor in prostate cancer. Endocr Rev2004;25:276–308.
32. Robinson D, Van Allen EM, Wu YM, Schultz N, Lonigro RJ, Mosquera JM,et al. Integrative clinical genomics of advanced prostate cancer. Cell2015;161:1215–28.
33. Zhou P, Young L, Chen Z. Weak solvent based chip lamination andcharacterization of on-chip valve and pump. Biomed Microdevices 2010;12:821–32.
34. NyquistMD, Li Y, Hwang TH,Manlove LS, Vessella RL, Silverstein KA, et al.TALEN-engineered AR gene rearrangements reveal endocrine uncouplingof androgen receptor in prostate cancer. Proc Natl Acad Sci U S A 2013;110:17492–7.
35. Frampton GM, Fichtenholtz A, Otto GA, Wang K, Downing SR, He J, et al.Development and validation of a clinical cancer genomic profiling testbased on massively parallel DNA sequencing. Nat Biotechnol 2013;31:1023–31.
36. Scher HI, Heller G, Molina A, Attard G, Danila DC, Jia X, et al.Circulating tumor cell biomarker panel as an individual-level surrogatefor survival in metastatic castration-resistant prostate cancer. J ClinOncol 2015;33:1348–55.
37. Li Y, Chan SC, Brand LJ, Hwang TH, Silverstein KA, Dehm SM. Androgenreceptor splice variants mediate enzalutamide resistance in castration-resistant prostate cancer cell lines. Cancer Res 2013;73:483–9.
38. MiyamotoDT, ZhengY,Wittner BS, Lee RJ, ZhuH, Broderick KT, et al. RNA-Seq of single prostate CTCs implicates noncanonical Wnt signaling inantiandrogen resistance. Science 2015;349:1351–6.
39. Ozkumur E, Shah AM,Ciciliano JC, Emmink BL,MiyamotoDT, Brachtel E,et al. Inertial focusing for tumor antigen-dependent and -independentsorting of rare circulating tumor cells. Sci Transl Med 2013;5:179ra47.
40. Sweeney CJ, Chen YH, Carducci M, Liu G, Jarrard DF, Eisenberger M, et al.Chemohormonal therapy inmetastatic hormone-sensitive prostate cancer.N Engl J Med 2015;373:737–46.
41. Darshan MS, Loftus MS, Thadani-Mulero M, Levy BP, Escuin D, Zhou XK,et al. Taxane-induced blockade to nuclear accumulation of the androgenreceptor predicts clinical responses in metastatic prostate cancer. CancerRes 2011;71:6019–29.
42. Korpal M, Korn JM, Gao X, Rakiec DP, Ruddy DA, Doshi S, et al. An F876Lmutation in androgen receptor confers genetic and phenotypic resistanceto MDV3100 (enzalutamide). Cancer Discov 2013;3:1030–43.
43. Joseph JD, LuN,Qian J, Sensintaffar J, ShaoG, BrighamD, et al. A clinicallyrelevant androgen receptor mutation confers resistance to second-gener-ation antiandrogens enzalutamide and ARN-509. Cancer Discov 2013;3:1020–9.
44. Chen EJ, Sowalsky AG, Gao S, Cai C, Voznesensky O, Schaefer R, et al.Abiraterone treatment in castration-resistant prostate cancer selects forprogesterone responsive mutant androgen receptors. Clin Cancer Res2015;21:1273–80.
45. Romanel A, Tandefelt DG, Conteduca V, Jayaram A, Casiraghi N, Wetters-kogD, et al. Plasma AR and abiraterone-resistant prostate cancer. Sci TranslMed 2015;7:312re10.
46. Walker DR, Bond JP, Tarone RE, Harris CC, Makalowski W, Boguski MS,et al. Evolutionary conservation and somatic mutation hotspot maps ofp53: correlation with p53 protein structural and functional features.Oncogene 1999;18:211–8.
47. Taylor BS, Schultz N, Hieronymus H, Gopalan A, Xiao Y, Carver BS, et al.Integrative genomic profiling of human prostate cancer. Cancer Cell 2010;18:11–22.
48. Chene P. The role of tetramerization in p53 function. Oncogene 2001;20:2611–7.
49. Karantanos T, Evans CP, Tombal B, Thompson TC,Montironi R, IsaacsWB.Understanding the mechanisms of androgen deprivation resistance inprostate cancer at the molecular level. Eur Urol 2015;67:470–9.
50. Sequist LV,Waltman BA,Dias-SantagataD,Digumarthy S, Turke AB, FidiasP, et al. Genotypic and histological evolution of lung cancers acquiringresistance to EGFR inhibitors. Sci Transl Med 2011;3:75ra26.
Clin Cancer Res; 23(3) February 1, 2017 Clinical Cancer Research756
2017;23:746-756. Published OnlineFirst July 11, 2016.Clin Cancer Res Jamie M. Sperger, Lindsay N. Strotman, Allison Welsh, et al. Cells Enabled by Exclusion-Based Analyte IsolationIntegrated Analysis of Multiple Biomarkers from Circulating Tumor
Updated version
10.1158/1078-0432.CCR-16-1021doi:
Access the most recent version of this article at:
