Possibilities of Liquid Biopsy in Clinical Practice

REVIEW Libri Oncol. 2017;45(1):23–30

POSSIBILITIES OF LIQUID BIOPSY IN CLINICAL PRACTICE

JOSIPA PERIŠA, PETRA BULIĆ, ZVJEZDANA ŠPACIR PRSKALO, MIHAELA GAĆE and LJILJANA MAYER

Department of Medical Biochemistry in Oncology, University Hospital for Tumors, Sestre milosrdnice University Hospital Center, Zagreb, Croatia

Summary Liquid biopsy represents innovation in the fi eld of oncology diagnosis. It refers to non-invasive blood collection method with isolation of circulating tumor cells, cell-free DNA (with circulating tumor cell free DNA fraction), circulating exosomes or micro RNA. Although requires specifi c methods for isolation and analysis of targets of interest, liquid biopsy could be of benefi t in preoperative prediction od overall survival of patients with carcinoma, for monitoring of disease and progression, along with surveillance of therapy response and early recurrence of cancer. The review describes methods for liquid biopsy analysis, together with concerns and achievements in implementing liquid biopsy in clinical practice. KEY WORDS: liquid biopsy, circulating tumor DNA - ctDNA, circulating tumor cells – CTCs, circulating microRNA – cmiRNA, exosomes, clinical practice

MOGUĆNOSTI TEKUĆE BIOPSIJE U KLINIČKOJ PRAKSI Sažetak Tekuća biopsija predstavlja inovaciju na području onkološke dijagnostike. Neinvazivna je metoda kojom se iz uzorka krvi izoliraju cirkulirajuće tumorske stanice, slobodna cirkulirajuća DNA (uz frakciju tumorske DNA), cirkulirajući exosomi ili mikroRNA. Iako zahtjeva speciﬁ čne metode izolacije i analize pojedinih elemenata od interesa, tekuća biopsija se može primijeniti u predoperativnoj predikciji preživljenja pacijenata s karcinomom, za praćenje tijeka i progresije bolesti, kao i za praćenje odgovora na terapiju ili ranog povratka bolesti. Članak se bavi opisom do danas poznatih metoda analize kompo- nenata tekuće biopsije, kao i problematikom i postignućima u uvođenju tekuće biopsije u kliničku praksu. KLJUČNE RIJEČI: tekuća biopsija, cirkulirajuća tumorska DNA - ctDNA, cirkulirajuće tumorske stanice – CTCs, cirkulirajuća mikroRNA - cmiRNA, egzosomi, klinička praksa

Cancer incidence worldwide counted 14,1 residual tumor diagnosis. Consequently, interest million new cases in 2012, and the number is pre- in tumor molecular diagnostics is rapidly increas- dicted to overgrow 23 million cases each year by ing. Up to these days tissue biopsy has been the the 2030 (1). Early-stage tumor diagnosis repre- gold standard for cancer diagnosis, but its limita- sents leading problem in diagnosis of cancer with- tions and potential risks of complications lead to in asymptomatic population, considered „healthy“. comprehensive search for less invasive techniques. Great number of tumor biomarkers are being ex- The discovery of cell free DNA (cfDNA) by Man- amined nowadays, but not many have reached del and Metais in 1948 (2) and circulating tumor clinical practice. Sensitivity and speciﬁ city of these cells (CTCs) in bloodstream revealed the new era biomarkers remain suboptimal for early-stage, or of molecular diagnostics of tumor diseases. Liquid

23 Lib Oncol. 2017;45(1):23–30 biopsy refers to less invasive technique which pro- early-stage cancer disease and can detect residual vides CTCs, cfDNA, cell free tumor DNA (ctD- lesions following therapy. High sensitivity and NA), circulating microRNA (cmiRNA) and exo- specifi city of ctDNA analytical methods can help somes from patient’s plasma. Development of mo- diagnose and monitoring of invasive cancer in lecular technologies such as PCR (polymerase early stage, together with recurrence of the cancer. chain reaction) or NGS (next generation sequencing) made analysis of ctDNA possible, in despite of low concentration in plasma. Characteristics of liquid biopsy compounds CTCs-circulating tumor cells Tissue biopsy and liquid biopsy- compared Origin of CTCs can be primary tumor or sec- ondary metastatic sites. Detached cancer cells Tissue biopsy provides enough material for have a role in metastasis through blood circula- pathohistological and molecular analysis giving tion. CTCs found in blood have epithelial /mesen- information on pathohistological and molecular chymal tumor origin and express epithelial mark- changes within tumor cells. Although it allows ers on their surface (6) for example EpCAM malicious and benign tumor diff erentiation and (Epithelial Cell Adhesion Molecules) or CDH2 classifi cation, it also has shortcomings. Fine or (Cadherin-2, N-cadherin) – a molecule included in core needle biopsy has risks of infl ammation, tumor metastasis induced by biopsy (3), (4), bleed- cell-to-cell adhesion and transendothelial metas- ing, or other complications, also it is painful for tasis of cancer cells (7). Epithelial tumor cells go patient. Further biopsies during treatment are of- through epithelial to mesenchymal transition ten impossible to be done due to the risks listed. (EMT) - changing surface molecules and weaken- Another disadvantage of tissue biopsy is aff ord- ing intercellular adhesion. This process aff ects the ability as it is performed by experienced profes- success of the isolation of CTCs using antibodies sional and is usually guided by ultrasound, com- targeting epithelial surface molecules. Transition- puted tomography scan (CT scan) or mammogra- al cells gain the ability to survive inside blood- phy. CT guided fi ne needle aspiration biopsy cost stream and invade distance tissue. Adhesion of for lung cancer patients was presented in a study. CTCs and progression of metastatic cancer de- The cost (including procedure, adverse events and pend on optimal fl ow of blood - “mechanic hy- pathology tests) was $4130,09 on average, while pothesis” and local microenvironment of the ad- blood-based genome test (including blood draw hesion site – the “seed and soil” hypothesis (8). and pathology tests) summed up to $836,45. An- Great number of researches had shown correla- other tissue biopsy disadvantage of concern is tion evidence between blood fl ow patt ern and me- turnaround time (TAT); while DNA analysis re- tastasis of primary tumor cells. CTCs (mean diam- sults can be obtained within 71 hours from blood eter of 12 to 20 μm) are usually trapped inside collection, biopsy results are usually given within capillaries of lungs and liver (diameter 3-8 μm), or 1 or 2 weeks from sampling. Molecular analysis other capillary perfused organs (7). CTCs need to obtained from formalin-fi xed paraffi n-embedded escape anoikis (programmed cell death that oc- (FFPE) tissue sample has weaknesses; non-stan- curs when cells detach from the surrounding ex- dardized preparation methods and toxic fi xatives tracellular matrix) and immune system defence induce DNA degradation, thus DNA integrity prior to metastasis (9). Some of these mechanisms verifi cation is necessary prior to molecular analy- include clustering with other cells, such as plate- sis (5). As it grows, tumor mutates and tumor tis- lets -“cloaking eff ect” while other mechanisms are sue becomes molecularly heterogeneous – one yet to be described. Platelets and monocytes needle aspiration can hardly collect all types of (found in CTC clusters) and macrophages mediate tumor cells, and sometimes re-aspiration is need- adhesion and thrive the metastatic progress ed -this is where liquid biopsy is welcomed. False (10,11). Value of CTCs number clinically correlates negative pathohistological or cytological results of with invasive progression of cancer, tumor mass patients with early-stage cancer can be conse- and decreased overall survival in patients with quence of imprecise needle aspiration. On the con- diff erent cancer varieties. Using diff erent plat- trary, ctDNA can be found in some patients with forms circulating tumor cells can be suffi ciently

24 Lib Oncol. 2017;45(1):23–30 separated from other blood cells, despite the low to tumor cell-DNA biopsy, if the samples were count in which they are found (1 CTC per 106 – 108 taken properly. Garcia-Olmoand coworkers es- white blood cells or 0-10 CTC/mL of blood (12)). tablished the hypothesis of genometastasis in 1999 Single tumor cells can be isolated using EpCAM with remarkable discovery of cancer transforma- expressed on their surface, or can be separated by tion of cultured cells induced by addition of plas- size using microporous fi lter membranes, since ma taken from tumor bearing rats (17). The fol- CTCs are bigger in size in relation to blood cells. lowing studies demonstrated oncogenic transfor- Isolated cells can be characterized by immunos- mation of susceptible cultured cells caused by taining of cytokeratin or other specifi c molecules addition of plasma containing ctDNA obtained expressed in tumor cells. On the other hand, clus- from colorectal carcinoma (CRC) patients (18). ters of CTCs, which have higher metastatic poten- Trejo-Becerril et al. brought evidence to the hy- tial, can be separated physically using microfl u- pothesis of genometastasis with research showing idic bifurcating traps (13). Morphological charac- the ability of ctDNA to prompt horizontal tumor teristics defi ne best isolation methods to be used progression in an immunocompetent colon-carci- in particular research purposes. nogenesis rat model (19). Oncogenic transformation of cells did not occur when serum and super- ctDNA – cell free tumor DNA natants were depleted of DNA, thus confi rming tumorigenic characteristics of ctDNA. The mecha- ctDNA can be isolated from plasma or serum nism of ctDNA uptake by the host cells remains of cancer patients, together with other cell-free unclear. Few studies point out circulating apop- DNA. Short half-life of ctDNA in blood (~ 2 hours) totic bodies as primary source of ctDNA, while refl ects real-time tumor stage(14), and can be clin- others suggest DNA/RNA complexes (“virto- ically used to monitor tumor progression or treat- somes”) as one of the forms in which ctDNA is ment effi cacy, while some protein biomarkers cir- taken by host cell (15). culate in bloodstream for several weeks. Con- centration range of cfDNA in healthy individuals Exosomes is 0-35 ng/mL while numbers can exceed 1000 ng/ mL in cancer patients group. Depending on can- Exosomes are microvesicles (3-200 nm in size) cer stage and metastasis ctDNA varies from 0,01% containing RNA, proteins and lipids inside lipid to >90% of cfDNA concentration (15). ctDNA can membrane. Physiological function of exosomes be either single- or double-stranded DNA, with circulating in blood is communication and trans- apoptotic or necrotic tumor cell origin, or can be fer of molecules between distant cells. Tumor cells spontaneously released into the bloodstream by use exosomes for stimulation of cell proliferation, living tumor cells in the form of nucleosomes. Dif- angiogenesis and metastasis. On the molecular ferent studies fi nd ctDNA shorter than non-cancer point of view, exosomes released by tumor cells cell free DNA, while some fi nd it opposite – diff er- carry highly specifi c molecular patt erns and con- ent conclusions can be linked to the variations of tribute to the detailed study of cancer metabolism. detection methods. The approximal length of ctD- A study presented specifi c molecule – glypican-1 NA fragment (around 166 bp) indicates its apop- (GPC1(+)) isolated from circulating exosomes totic origin, during apoptosis DNA is cleaved in found in serum of patients with pancreatic cancer. between nucleosomes at ~ 180 bp intervals (the This molecule had absolute specifi city and sen- endonuclease cleavage leaves “ladder-like” pat- sitivity for distinguishing healthy subjects and tern on gel electrophoresis). Other evidences sup- patients with a benign pancreatic disease from port hypothesis of CTCs as one of the ctDNA patients with early- and late-stage pancreatic can- sources in blood, since mutations found in ctDNA cer (20). are identical to the ones found in CTC from the same patient sample(16). Tumor tissue biopsy re- cmiRNA – circulating micro RNA sults can prove ctDNA origin; point mutations, copy number variations or chromosome aberra- cmiRNA can be found inside exosomes or in tions as well as epigenetic molecular changes a free-form released into bloodstream by tumor or (DNA methylation, sulfonation) or cancer derived other cells (21). RNA is fragile molecule, and re- viral sequences found in ctDNA will be identical quires RNase inhibitors and RNase free equip-

25 Lib Oncol. 2017;45(1):23–30 ment in order to avoid destruction during isola- eration sequencing (NGS), pyrosequencing and tion and sample manipulation. However, miRNAs other similar methods bring improvement and ad- are relatively stable in blood containing RNases vances in cancer research and post-analytical (necessary for viral RNA degradation), but the phase, making it easier for clinicians to interpret. mechanism of RNase circumvention remains in- distinct (21). miRNAs are single-stranded, non- Pre-analytical phase coding RNA molecules that are 20-23 nucleotides The Biomarkers Consortium Workshop held long and have a role in silencing mRNA and post- in 2010 brought together the experts in the fi eld of transcriptional regulation of gene expression, while CTC research to share experience and compare specifi c cmiRNA patt erns can be found in cancer the performances of most promising assay tech- patients (22) such as circulating miR-21 which can nologies with aim to take initial step in clinical be used as potential tumor biomarker in CRC (23). validation of these methods (25). In order to enter clinical practice, pre-analytical phase variables TEP-tumor educated platelets should be standardized for each assay – sample One of the cells that also participates in tu- acquisition, processing, storage and transport mor growth and dissemination and aff ects their conditions should be defi ned. gene expression are platelets. RNA profi le of blood platelets is infl uenced by tumor type, pro- CTCs pre-analytical requirements: gression and metastasis as tumor alerts premRNA Liquid biopsy requires 5 to 10 mL of blood splicing. TEPs provide specifi c information on the taken on anticoagulant EDTA tubes used for blood location and characteristics of primary tumor. In collection need to be processed within 24 hours of 283 patients, Best et al. (24) have done TEPs mRNA blood collection to minimize cell degradation. sequencing and diff ered patients with primary tu- Procedures can diff er regarding stability, size and mor and metastasis from healthy ones with 96 % concentration in plasma in dependence of exam- accuracy. Location of primary tumor was correct- ined CTCs and method used. ly identifi ed in 71% of patients and diff erent mutations were described. TEPs isolation should be done within 48 hours after blood collection, but cfDNA/ ctDNA pre-analytical requirements the risk of change in mRNA profi le increases dur- Since ctDNA builds up a fraction in circulating that period. Diseases other than cancer can ing cell free DNA, cfDNA should be isolated in also alert platelet mRNA profi le, however, larger order to analyse ctDNA. Isolation from plasma or studies are required on this subject. serum sample is performed using isolation kits; shipping, storage or quality of sample, however, can aff ect isolation’s effi ciency. Blood cell degra- Methodologies in liquid biopsy analysis dation in pre-analytical phase can contaminate the sample with genomic DNA. Manufacturers give The selection of method should be made re- instructions on sample acquisition, processing garding the purpose of the analysis; it is hard to and storage - tested and optimized in studies that fi nd optimal method to be used in clinical routine preceded putt ing the isolation kit on market. Some due to complex nature of liquid biopsy analytes. manufacturers even examined the infl uence of Diff erent methods may be required for early-stage shaking and storage conditions in a way of con- recognition and monitoring therapy effi cacy. For tamination with genome DNA in plasma collected example: CTCs can be isolated intact and cultivat- in preservative tubes which they made for this ed for further therapy analysis but the method purpose especially, and compared it to other com- could be time consuming, while ctDNA concen- mercially available K EDTA tubes. tration analysis, which can help monitoring tumor 2 progression, could be done within hours. In the cmiRNA pre-analytical requirements fi eld of liquid biopsy, numerous researches are struggling with method optimization and pre-an- Studies on miRNA confi rmed advantages of alytical phase requests, which have to be met for plasma over serum samples (26). During clot pro- successful analytical phase. Nowadays, next gen- cess required for serum, non-specifi c miRNA is

26 Lib Oncol. 2017;45(1):23–30 released from platelets, lysed red blood cells, leu- (28). If viable, intact cells are required for cultiva- kocytes, and circulating tumor cells, contaminat- tion or other analysis, size based vacuum fi ltration ing the cmiRNA. Plasma is collected in tubes con- would be a method of choice, e.g. ISET (Isolation taining standard blood anticoagulants: EDTA or by Size of Epithelial Tumor cells) and Nucleopore sodium citrate followed by centrifugation. Hepa- Assay methods. Flow cytometry is also one of rin anticoagulant might interfere with analytic technologies with wide application in CTC analy- methods (inhibits polymerase and reverse tran- sis. Some technologies took a step forward and scriptase enzymes used in PCR), therefore, hepa- exploit diff erences in density and dielectric prop- rin tubes are not recommended (27). Standard erties for diff erentiation of cells (Dielectroforetic protocol for sample handling and storage of miR- Field Flow Fractionation – DFFF). Extraordinary NA samples have yet to be made, however, cryo- assays use nanodetectors binding EpCAM+ cells preservation on -80°C reduces degradation during in vivo, or even noninvasively count CTCs by in- long-term storage. All limitations listed can aff ect jection of tumor-specifi c fl uorescent ligand fol- the interpretation of miRNA results; therefore, lowed by multiphoton fl uorescence imaging of standardized protocols are indispensable. superfi cial blood vessels to quantitate the fl owing CTCs. CTC undergoing EMT can be separated us- Analytical phase ing negative enrichment – red cells are lysed and CD45+ cells (leukocytes) removed by immuno- Clinical practice has set the bar high when magnetic beads, but still have to be distinguished discussing analytical phase requirements; high from remaining epithelial cells. Manifold technol- sensitivity (with low limit detection) and specifi c- ogies are used in but still have not reached clinical ity are of main importance, especially for tumor practice. biomarkers since diagnostic criteria and therapeutic protocols depend on biomarker results. Report- able ranges have to be set prior to clinical use of cfDNA tumor marker, unfortunately, cancer diagnosis is cfDNA is isolated from plasma using special- sometimes hard to set due to complexity of cancer, ized DNA isolation kits prior to ctDNA analysis– and tumor biomarkers cannot be easily put in be- numerous articles compared diff erent commer- tween range boundaries. Assay reproducibility cially available kits (29) (e.g. QIAamp® circulating and robustness required for routine work repre- nucleic acid kit (QIA - Qiagen), PME free-circulat- sent one of the main diffi culties in implementing ing DNA Extraction Kit (PME - Analytik Jenna), liquid biopsy in everyday clinical practice. MagTMcfDNA Isolation Kit (NpM - BiooScientif- ic)). Steps following isolation include measure- CTCs ment with quantitative PCR – qPCR (based on CTCs assay technologies used in process are fl uorescence of probes att ached to amplifi ed DNA determined by target cells and assay outcome- segments), digital droplet PCR (ddPCR), quan- characterization, enumeration or enrichment. Im- titative kits for measuring cfDNA concentration munoassays are widely used in detection and en- (based on fl uorescence and standard curve) or se- richment of rare CTCs. Specifi c antigen-antibody quencing of cfDNA fragments (using NGS, pyro- reaction allows isolation of EpCAM+, CD45-, sequencing or other methods). Sequencing usual- HER2/neu+ or other tumor cells, following purifi - ly requires pre-amplifi cation of cfDNA and library cation can be performed with immunomagnetic construction using multiple-primer reaction. Ion beads or columns. Downstream processes usually Torrent™ technology by ThermoFisher Scientifi c include fl uorescent antibodies staining and label- is platform that allows generating gigabases of se- ling of intracellular components and confocal mi- quence per day, and enables new experiments on croscopy. Other technologies are based upon CTC complementary DNA (cDNA) sequencing. This size, for example, CTC Membrane Microfi lter is technology directly translates chemically encoded designed to exploit cell size diff erences between information (A, C, G, T) into digital information normal blood cell and tumor cells and following (0, 1) on a semiconductor chip, resulting in sim- characterization of the tumor cell is done by quan- pler, faster, more cost-eff ective and scalable next- tum dot-based immunofl uorescence detection generation sequencing. Technology evolution and

27 Lib Oncol. 2017;45(1):23–30 widening of gene databases contributes to screen- tionable stability of stock sample used in generating and early recognition of cancer diseases; cfD- ing curves. Microarray assay methodologies allow NA/ctDNA and NGS have potential to dramati- wide gene profi ling, but require software systems cally change diagnostic, monitoring and treatment for imaging and data analysis (31) and cannot be of cancer once used in clinical practice, as con- used for miRNA quantifi cation due to low speci- cluded on Circulating Tumor Cells (CTC) and fi city and sensitivity. 3D-Gene® microarray pro- Cell-Free DNA (cfDNA) Workshop 2016: Scientif- duced by Toray (Tokyo, Japan) is improved, high- ic Opportunities and Logistics for Cancer Clinical ly sensitive and reproducible assay expanding the Trial Incorporation conference (30). horizon for miRNA analysis. Massive parallel sequencing was a complex method of miRNA anal- cmiRNA ysis including RNA extraction and size fractionation of the small RNA, conversion to cDNA and Previous isolation of RNA included multiple PCR multiplication together with library construc- extraction steps with analyte loss and long dura- tion. It is replaced by the HTG EdgeSeq system tion of the process. Phenol-chloroform extraction (HTG molecular, Tuscon, AZ, USA) which relies (Trizol) required a large sample volume and in- on the specifi city of the pre-designed probes. cluded working with toxic reagents, but recently Currently it is ongoing validation to determine several isolation kits become commercially avail- specifi city and sensitivity in detecting cmiRNA. able for research purposes. MiRCURY™ RNA Iso- The advantages of HTG EdgeSeq: highly auto- lation Kit (Exiqon, Denmark), mirVana™ PARIS™ mated protocol (with reduced user variations), (Life Technologies, Grand Island, NY, USA) and adjusted sample preparation and input require- miRNeasy® (Qiagen, Venlo, Limburg, Belgium) ments, large number of targeted miRNAs made isolation of miRNA aff ordable, easy and (>2000), needless RNA extraction and absence of practical. The amount and quality of isolated ma- molecular library made it revolutionary in miR- terial could not be compared between these kits NA analysis. since the reports on this subjects haven’t been done yet. Quantifi cation of cmiRNA extracted can be done using spectrophotometer, but due to low The future of liquid biopsy in clinical practice amounts of cmiRNA, more sensitive technologies are required. Agilent Technologies 2100 Bioana- Liquid biopsy has been introduced in clini- lyzer (Agilent, Santa Clara, CA, USA) uses capil- cal practice through prenatal screening tests, lary electrophoresis to assess miRNA concentra- and consequently showed great potential to be tion (31). The integrity of isolated RNA represents implanted in other fi elds, especially for diagno- another problem in miRNA analysis – degraded sis and monitoring of cancer disease. Detection RNA is not a suitable analyte for NGS or microar- of specifi c mutations, overall survival predic- ray analysis. Reverse transcriptase PCR (RT-qP- tion, recurrence of disease, cultivation of CTCs CR) is a method of choice for miRNA analysis for further pharmacotherapy and therapeutic measuring changes of gene expression to a suit- targeting, as well as monitoring of therapy re- able internal control. This is the “weak point” of sponse can be provided from ctDNA and CTCs RT-qPCR method, since few endogenous microR- – the most frequently used liquid biopsy com- NAs can be used as an internal control to evaluate pounds. cmiRNA multiplication. This is due to variations CTCs are detectable in 60 % of metastatic of microRNA expression in several diseases and breast and prostate cancers, while lower abso- among individuals (32). Exogenous control refers lute number is found in metastatic CRC cancer to artifi cial microRNAs added to the samples pri- (~ 30 – 40% of cases) (33). The explanation for or to RNA extraction, but can contaminate prima- this event can be found (could be) in circulatory ry samples, and it is diffi cult to control the amount anatomy of intestine and hepatic drainage, of external RNA added to each sample. These dis- which prevents/disables CTCs to reach periph- advantages can be circumvented using standard eral blood. However, in patients with advanced curves for each microRNA. Unfortunately, this is solid tumors, ctDNA was present even when an expensive, time-consuming method with ques- CTCs were not detectable. The detection and

28 Lib Oncol. 2017;45(1):23–30 enumeration of CTC cells can be used in progno- dance between expressed phenotype and detect- sis of disease and prediction of disease-free and ed genotype (not all of the mutations found take overall survival. FDA approved CellSearch® part in cancer pathogenesis). From analytical (Veridex, New Jersey, USA) Circulating Tumor point of view, ctDNA is reliable, stable and eas- Cell Kit for enumeration based on EpCAM+ an- ily isolated analyte, but requires a priori knowl- tibodies. Specifi city of antibodies, however, can edge and target of interest for clinical practice. vary and the detection rate can be greatly changed (34). This has to be considered when in- REFERENCES terpreting the result. Consequently, serial CTC testing using CellSearch® should be used to- 1. Uk CR. Worldwide cancer statistics. Cancer Res UK. gether with other clinical methods for monitor- 2014;2012:2012–5. ing metastatic breast, colorectal, and prostate 2. Mandel P, Metais P. Les acides nucléiques du plasma sanguin chez l’homme. C R Seances Soc Biol Ses Fil. cancer. CTC go through dynamic changes in tu- 1948;142(3–4):241–3. mor progression, therefore characterization of 3. Loughran CF, Keeling CR. Seeding of tumour cells fol- EMT features can detect tumor cells resistance to lowing breast biopsy: a literature review. Br J Radiol. given therapy (35). CTCs obtained from patient 2011;84(1006):869–74. before surgery can help predict the overall sur- 4. A L Young JPAL. Needle track seeding following bi- vival, but more importantly CTCs obtained after opsy of liver lesions in the diagnosis of hepatocellular surgical therapy can help guide therapeutic de- cancer: a systematic review and meta-analysis. Gut. cision (36), which represents an endless chal- 2008;57(11):1592–6. lenge for oncologists worldwide. 5. Srinivasan M, Sedmak D, Jewell S. Eff ect of fi xatives and tissue processing on the content and integrity of ctDNA gives real-time information on mo- nucleic acids. AmJPathol. 2002;161(0002–9440 (Print)): lecular changes inside growing tumor including 1961–71. copy number aberrations and other somatic mu- 6. Wu S, Liu S, Liu Z, Huang J, Pu X, Li J, et al. Classifi ca- tations and deletions. Clinical correlation of ctD- tion of circulating tumor cells by epithelial-mesenchy- NA and tumor burden depends on tumor stage, mal transition markers. PLoS One. 2015;10(4). localization and type. Another study detected 7. Ramis-Conde I, Chaplain MAJ, Anderson ARA, Dras- ctDNA in 48-73% patients with gastrointestinal do D. Multi-scale modelling of cancer cell intravasa- tion: the role of cadherins in metastasis. Phys Biol. cancer, while all patients with metastatic GI can- 2009;6(1):16008. cer had higher levels of ctDNA (36). Circulating 8. Azevedo AS, Follain G, Patt habhiraman S, Harlepp S, tumor DNA can be used in screening for CRC; Goetz JG. Metastasis of circulating tumor cells: Favor- FDA approved EpiProColon blood-based test able soil or suitable biomechanics, or both? Vol. 9, Cell that identifi es methylated Septin 9 DNA. High Adhesion and Migration. 2015. p. 345–56. preoperative ctDNA concentration is associated 9. Joosse SA, Gorges TM, Pantel K. Biology, detection, with poorer prognosis and shorter overall sur- and clinical implications of circulating tumor cells. vival. Postoperative ctDNA is more reliable and EMBO Mol Med. 2015;7(1):1–11. sensitive for detection of tumor recurrence than 10. Labelle M, Begum S, Hynes RO. Direct Signaling be- other tumor biomarkers, such as CEA (carcino- tween Platelets and Cancer Cells Induces an Epitheli- al-Mesenchymal-Like Transition and Promotes Me- embryonic antigen) in CRC or CA 15-3 (cancer tastasis. Cancer Cell. 2011;20(5):576–90. antigen 15-3) in breast cancer. Patients were fol- 11. Qian B, Deng Y, Im JH, Muschel RJ, Zou Y, Li J, et al. A lowed for 2 years after surgery: the disease reoc- distinct macrophage population mediates metastatic curred in 85% patients with detectable postsur- breast cancer cell extravasation, establishment and gical ctDNA levels while 100% patients with un- growth. PLoS One. 2009;4(8). detectable ctDNA were recurrence-free. ctDNA 12. Miller MC, Doyle G V, Terstappen LWMM. Signifi - can identify patients with higher risk of recur- cance of Circulating Tumor Cells Detected by the Cell- rence and help guide adjuvant therapy. Chemo- Search System in Patients with Metastatic Breast Colorectal and Prostate Cancer. Oncology. 2010;2010: therapy or radiotherapy cause extensive ctDNA 617421. release, but it is unclear if ctDNA is released 13. Sarioglu a F, Aceto N, Kojic N, Donaldson MC, Zeinali from dying cells or higher levels of ctDNA are M, Hamza B, et al. A microfl uidic device for label-free, consequence of cancer resistance to therapy. An- physical capture of circulating tumor cell clusters. Nat other disadvantage of ctDNA is discorrespon- Methods. 2015;12(April):1–10.

29 Lib Oncol. 2017;45(1):23–30

14. Diehl F, Schmidt K, Choti M a, Romans K, Goodman Infl uence Circulating microRNA Biomarker Levels. S, Li M, et al. Circulating mutant DNA to assess tumor PLoS One. 2013;8(6). dynamics. Nat Med. 2008;14(9):985–90. 27. Al-Soud WA, Rådström P. Purifi cation and character- 15. Cheng F, Su L, Qian C, Cheng F, Su L, Qian C. Circu- ization of PCR-inhibitory components in blood cells. J lating tumor DNA: a promising biomarker in the liq- Clin Microbiol. 2001;39(2):485–93. uid biopsy of cancer. Oncotarget. 2016;7(30):48832–41. 28. Zhou L, Yan J, Tong L, Han X, Wu X, Guo P. Quantum 16. Bett egowda C, Sausen M, Leary RJ, Kinde I, Wang Y, Dot-based Immunohistochemistry for Pathological Agrawal N, et al. Detection of Circulating Tumor Applications. Cancer Transl Med. 2016;2(1):21–8. DNA in Early- and Late-Stage Human Malignancies. 29. Sorber L, Zwaenepoel K, Deschoolmeester V, Roeyen Sci Transl Med. 2014;6(224):224ra24-224ra24. G, Lardon F, Rolfo C, et al. A Comparison of Cell-Free 17. García-Olmo D. Tumor DNA circulating in the plasma DNA Isolation Kits. J Mol Diagnostics. 2016;(Novem- might play a role in metastasis. The hypothesis of the ber):1–7. genometastasis. Histol Histopathol. 1999;14(4):1159–64. 30. Lowes LE, Bratman S V., Ditt amore R, Done S, Kelley 18. García-Olmo DC, Domínguez C, García-Arranz M, SO, Mai S, et al. Circulating tumor cells (CTC) and Anker P, Stroun M, García-Verdugo JM, et al. Cell-free cell-free DNA (cfDNA)workshop 2016: Scientifi c op- nucleic acids circulating in the plasma of colorectal portunities and logistics for cancer clinical trial incor- cancer patients induce the oncogenic transformation poration. Int J Mol Sci. 2016;17(9). of susceptible cultured cells. Cancer Res. 2010;70(2): 31. Schroeder A, Mueller O, Stocker S, Salowsky R, Leiber 560–7. M, Gassmann M, et al. The RIN: an RNA integrity 19. Trejo-Becerril C, Pérez-Cárdenas E, Taja-Chayeb L, number for assigning integrity values to RNA mea- Anker P, Herrera-Goepfert R, Medina-Velázquez LA, surements. BMC Mol Biol. 2006;7(1):3. et al. Cancer Progression Mediated by Horizontal 32. Ono S, Lam S, Nagahara M, Hoon D. Circulating mi- Gene Transfer in an In Vivo Model. PLoS One. 2012; croRNA Biomarkers as Liquid Biopsy for Cancer Pa- 7(12). tients: Pros and Cons of Current Assays. J Clin Med. 20. Melo S a., Luecke LB, Kahlert C, Fernandez AF, Gam- 2015;4(10):1890–907. mon ST, Kaye J, et al. Glypican-1 identifi es cancer exo- 33. Katsiampoura A, Kopetz S. Clinical applications of liq- somes and detects early pancreatic cancer. Nature. uid biopsies in gastrointestinal oncology. Gastrointest 2015;523(7559):177–82. Cancer Res. 2014;7(4):S8–12. 21. Schrauder MG, Strick R, Schulz-Wendtland R, Strissel 34. Antolovic D, Galindo L, Carstens A, Rahbari N, PL, Kahmann L, Loehberg CR, et al. Circulating mi- Büchler MW, Weitz J, et al. Heterogeneous detection of cro-RNAs as potential blood-based markers for early circulating tumor cells in patients with colorectal can- stage breast cancer detection. PLoS One. 2012;7(1). cer by immunomagnetic enrichment using diff erent 22. Izzott i A, Carozzo S, Pulliero A, Zhabayeva D, Ravett i EpCAM-specifi c antibodies. BMC Biotechnol. 2010; JL, Bersimbaev R. Extracellular MicroRNA in liquid 10:35. biopsy: applicability in cancer diagnosis and preven- 35. Yu M, Bardia A, Witt ner BS, Stott SL, Smas ME, Ting tion. Am J Cancer Res. 2016;6(7):1461–93. DT, et al. Circulating breast tumor cells exhibit dy- 23. Toiyama Y, Takahashi M, Hur K, Nagasaka T, Tanaka namic changes in epithelial and mesenchymal compo- K, Inoue Y, et al. Serum miR-21 as a diagnostic and sition. Science. 2013;339(6119):580–4. prognostic biomarker in colorectal cancer. J Natl Can- 36. Diehl F, Li M, Dressman D, He Y, Shen D, Szabo S, et cer Inst. 2013;105(12):849–59. al. Detection and quantifi cation of mutations in the 24. Best MG, Sol N, Kooi I, Tannous J, Westerman BA, plasma of patients with colorectal tumors. Proc Natl Rustenburg F, et al. RNA-Seq of Tumor-Educated Acad Sci U S A. 2005;102(45):16368–73. Platelets Enables Blood-Based Pan-Cancer, Multiclass, and Molecular Pathway Cancer Diagnostics. Cancer Cell. 2015;28(5):666–76. 25. Allan AL, Keeney M. Circulating tumor cell analysis: Technical and statistical considerations for application Corresponding author: Josipa Periša, Department of Medi- to the clinic. Vol. 2010, Journal of Oncology. 2010. cal Biochemistry in Oncology, University Hosipital for Tu- 26. Cheng HH, Yi HS, Kim Y, Kroh EM, Chien JW, Eaton mors, Sestre milosrdnice University Hospital Center, Ilica KD, et al. Plasma Processing Conditions Substantially 197, 10000 Zagreb, Croatia; e-mail: [email protected]