Biology and significance of circulating and disseminated tumour cells in colorectal cancer

Transcription

1 DOI /s REVIEW ARTICLE Biology and significance of circulating and disseminated tumour cells in colorectal cancer Gunnar Steinert & Sebastian Schölch & Moritz Koch & Jürgen Weitz Received: 25 January 2012 / Accepted: 27 January 2012 # Springer-Verlag 2012 Abstract Purpose More than 130 years ago, circulating tumour cells (CTCs) and disseminated tumour cells (DTCs) have been linked to metastasis. Since then, a myriad of studies attempted to characterise and elucidate the clinical impact of CTCs/ DTCs, amongst others in colorectal cancer (CRC). Due to a flood of heterogeneous findings regarding CTCs/DTCs in CRC, this review aims to describe the known facts about CTC/DTC biology and clinical impact. Methods To identify the basic scientific literature regarding the biology and clinical impact of CTCs/DTCs in CRC, we reviewed the literature in the PubMed database. We focused on publications written in English and published until January As search terms, we used colorectal cancer (CRC), colon cancer (CC), CTC, DTC, bone marrow (BM), lymph node (LN), peripheral blood (PB), significance and prognosis. Results CTC detection and quantification under standardised conditions is feasible. Several studies in large patient settings have revealed prognostic impact of CTCs in CRC. CRCderived DTC detection and analysis in BM exhibits a more heterogeneous picture but also shows clinical value. Furthermore, the presence of DTCs in LN has a strong prognostic impact in CRC. Conclusions Clinical relevance and prognostic significance of CTCs/DTCs in CRC have been clearly demonstrated in many experimental studies. The major challenge in CTC/DTC research is now to harmonise the various identification and detection approaches and consequently to conduct large prospective G. Steinert : S. Schölch : M. Koch : J. Weitz (*) Department of General, Visceral and Transplant Surgery, University of Heidelberg, Im Neuenheimer Feld 110, Heidelberg, Germany juergen.weitz@med.uni-heidelberg.de multi-institutional trials to verify the use of CTCs/DTCs as a valid prognostic and predictive biomarker for clinical routine. Keywords Colorectal cancer. Circulating tumour cells. Disseminated tumour cells. Epithelial mesenchymal transition. Prognostic value. Clinical significance Introduction Cancer-related death is most commonly caused by metastases derived from epithelial tumours [1]. Among these, colorectal cancer (CRC) is ranked as the second most common cause of cancer-related death in both genders [2]. In 2008, 436,000 new cases of CRC and 212,000 CRC-related deaths were recorded [3, 4]. Successful metastasis requires the potential and ability of tumour cells to enter the circulation, attach to the endothelium, invade the target organ and subsequently form metastases. Since the development of modern molecular detection assays, the concept and existence of circulating tumour cells (CTCs) and the dissemination and settlement of these cells in secondary organs such as liver, bone marrow (BM), lymph nodes (LN) and lungs have been widely accepted [5]. The tumour cells are shed into the mesenteric circulation or lymph vessels and may lead to clinically overt metastases via those routes [6]. These cells are also the potential source for later tumour relapse despite a potentially curative resection. The most common soils for these circulating and disseminated tumour cells (DTCs) are liver, LN and BM. In this context, we call single tumour cells in LN or BM DTCs and tumour cells spreading freely through the blood CTCs. Already in 1869, T. Ashworth assumed that circulating cells are thought to be responsible for metastasis [7]. However, still

2 very little is known about the exact amount of tumour cells released into the blood from a tumour in human patients. The present review will highlight recent findings regarding CTCs and DTCs in the context of CRC. First, we will point out typical biological traits, promising biomarker candidates of CTCs/DTCs and the need and eligibility of mouse models in CTC/DTC research. Second, the clinical significance and prognostic impact of CTCs/DTCs with respect to their identification in LN, BM and peripheral blood (PB) will intensively be depicted and discussed. Metastatic cascade, CTC/DTC biology and biomarker Although CTCs were first described by Ashworth in 1869, it took more than a century until these cells were accessible to further analyses with regard to their biological traits [7]. By summarising the findings of CTC/DTC-associated biology, one intriguing issue is obviously the anoikis resistance of those cells. Normally, epithelial cells die upon detachment and loosing of cellular assembly, which is known as anoikis. Anoikis resistance could be an explanation for the survival in circulation [8]. In line with that, IGRF1 seems to support the survival of CTCs by inhibition of the anoikis cascade [9]. In summary, tumour cell dissemination and the metastatic cascade take place as follows: epithelial mesenchymal transition (EMT) of single tumour cells, invasion of the surrounding tissue, entry into the microvasculature of the lymph and blood vessel system (intravasation), survival in the circulation and translocation to microvessels of distant organs, extravasation from the bloodstream, mesenchymal epithelial transition (MET), survival in the new tissue-specific environment, quiescence/proliferation and finally formation of metastases [10]. At the present time, two time-dependent models of metastatic spread are still under debate, namely whether metastatic spread represents an early event with integrated dormant phase in the LN (lymphogenous spread) or is rather a continuous, delayed event via the blood stream without prior dormant interstation (haematogenous spread) [11]. Currently, we think that EMT should be intensively investigated mainly by assuming a decisive role in making tumour cells moveable. Thus, we will focus on this topic more in detail to provide a better understanding of the sophisticated biology of CTCs/DTCs. Primarily, EMT was identified in early embryonic morphogenesis [12]. EMT-inducing transcription factors (EMTiTFs) like Snail, Slug, Twist1, def1, SIP1 and FOXC2 enable epithelial cells to be transformed in mesenchymal derivatives, whereas they are likewise able to revert back to the epithelial phenotype by undergoing MET [11, 13, 14]. EMT has been shown to endow cancer cells with motility, invasiveness, resistance to apoptosis, quiescence, self-renewal and multidrug resistance [13]. Notably, EMTmediated tumour cell migration and invasion are based on three major strategies: weakening cell cell cohesion, matrix degradation and modification of the cell cytoskeleton [12]. Recently, it has been reported that non-cancer stem cells (non-cscs) were transformed into a CSC-like state by EMT [15]. Along with involved signalling pathways like TGFβ, Notch or Wnt, EMTiTFs orchestrate to boost the EMTassociated traits and the expression of EMT-specific markers like E-cadherin, claudin and occludin [13, 14]. Changes in the cytoskeleton of epithelial cells such as the repression of cytokeratin (CK) and upregulation of the mesenchymal skeleton-specific vimentin can be observed as well [13, 16]. Interestingly, Vincan et al. could demonstrate a kind of reversible change between EMT and MET of LIM1863- Mph cells [17]. This interchangeable phenomenon probably represents a core event in CRC pathogenesis. The impact of EMT in CRC could also be demonstrated by Loboda et al. who identified a tightly correlated (Pearson R00.92, P< ) EMT gene signature in 326 colon cancer samples [18]. In contrast to the assumption that EMT is usually fully completed and thus completely abolishes epithelial traits on CTCs, Lecharpentier et al. described CTCs expressing a dual epithelial and mesenchymal phenotype in the PB of patients with metastatic carcinoma [19]. From the therapeutic point of view, EMT poses a dramatic challenge caused by the EMT-mediated resistance to conventional anticancer therapies. It has been shown that colon and ovarian carcinoma cell lines presenting an EMT phenotype were resistant to oxaliplatin and paclitaxel. Moreover, Snail and Twist are able to mediate radioresistance and chemoresistance in several human cancer cells [13]. A remarkable and intriguing approach arose from several mouse models which focused on metastasis. In line with these investigations, it was suggested that primary tumours exhibit a paracrine potential to direct metastasis. More precisely, factors secreted by the primary tumour (e.g. VEGF- A, PIGF or PSAP) are thought to mobilise BM-derived cells subsequently attracted to premetastatic sites. In turn, cells of the premetastatic niche release factors (e.g. SDF-1, S1000A8 or S100A9) which attract DTCs again [20 22]. However, the formation of metastasis is an absolutely sophisticated programme comprised of several successfully conducted single steps. Therefore, metastasis seems to be a fairly inefficient process [23, 24]. Nevertheless, a small number of tumour cells (~0.01%) succeed in completing each single step to form metastases [25]. Beyond the crucial role of EMT, what do we know exactly about the physiological and cellular traits of CTCs/DTCs enabling them to safely circulate and successfully seed the soil? For instance, tissue factor proteins exposed on the surface of individual cancer cells decoy dozens of aggregating platelets [26, 27]. From this, it follows that the detection of CTCs by surface markers could be hampered by a platelet envelope [11]. In addition, it has been reported that tumour

3 cells occasionally enter the circulation as multicellular aggregates or clusters. These clusters are also known as circulating tumour microemboli (CTM) which are thought to have high metastatic potential through aggressively proliferating into the specific organ site [28]. Unlike the concept of the platelet envelope and CTM, another theory claims that CTCs fuse with macrophages in the blood to increase their survival chances and acquire mesenchymal traits [29]. The formation of micrometastases, which seems to be the basis for further clinically overt metastases, requires an organ-specific and prosperous homing of the tumour cells. One reason for this can be the expression of specific proteins like integrins, which seem to play a key role in the organspecific homing process [30 32]. Hence, prostate cancer preferentially metastasizes to the bone and CRC mainly to the liver [6]. Kim et al. postulated that CTCs might colonise their tumours of origin, a process that they call tumour selfseeding, which takes place in CRC, too. In a breast cancer model, they observed that self-seeding accelerated tumour growth, angiogenesis and stromal recruitment through seedderived factors such as the chemokine CXCL [33]. Approximately 10 6 tumour cells per gram of primary tumour are daily released into the bloodstream [34]. However, shear forces in the physiological range can induce lethal damage in a high percentage of CTCs. It has been reported that up to 70% of B16 melanoma cells can be killed within 1 h when exposed to shear stress [35]. A possible explanation why CTCs are still detectable in the blood months to years after complete removal of the primary tumour may therefore be the circulation and exchange of tumour cells between different metastatic sites and compartments [36]. The first report with regard to the identification of CTCspecific biomarker and genes in CRC was published by Smirnov et al. in 2005 [37]. Smirnov and co-workers screened blood containing 100 EpCAM + -CTCs from one metastatic CRC (mcrc) patient and suggested genes such as AGR2, S100A14, S100A16 and FABP1 to be specific for CTCs. In contrast, Solmi et al. carried out a microarraybased screen and concluded that a suitable and consistent colon epithelium mrna marker might be difficult to identify [38]. Ziegelschmid et al. presented data from a multicenter study (196 PB samples were collected from 76 CRC patients) indicating a correlation between epidermal growth factor receptor (EGFR) and carcinoembryonic antigen (CEA) gene expression, disease progression and tumour stage [39]. Based on these results, Lankiewicz et al. analysed the PB of CRC patients and demonstrated 55% of patients to be positive for CEA and 18% for EGFR [40]. Interestingly, the same authors examined the expression profile of CTCs from CRC patients and observed increased CEA expression before and decreased CEA expression after chemotherapy; the same effect was demonstrated for EGFR vice versa [41]. In summary, many studies tried to identify biomarkers for CRC-derived CTCs but further investigations are required as the results of the currently available studies differ widely [42 45]. Limited by the fairly low concentration of DTC (10 5 to 10 6 ) in BM, the DTC identification also requires a preceding enrichment step comparable with those for CTC [46]. Interestingly, Klein et al. postulated that DTCs might be genomically heterogeneous [47]. Stöcklein and Klein acted on that suggestion and reviewed whether typical CRC mutations (KRAS and TP53) should be present in both primary tumour and metastasis or just in one of them [48]. The authors pointed out that KRAS mutations were observed in 0 60% and TP53 mutations in 20 60% of overall analysed cases, whereas in case of disparity, the mutations were mainly found in the metastasis. Based on the indicated genomic heterogeneity of DTCs, two fundamental concepts of systemic cancer progression are discussed by Klein: the linear progression and the parallel progression model. According to the linear progression model, the metastasisassembling cells ought to be genotypically highly similar to the primary tumour, whereas in the parallel progression model, the cells are rather heterogenous [49]. Weber et al. analysed 33 microsatellite markers from 17 different chromosomes and additionally sequenced KRAS in 38 colorectal tumours and corresponding synchronous liver metastases. They could demonstrate uniformity of the mutational status of the KRAS gene, and in contrast, a marked disparity of the allelic imbalances, whereas two thirds of the hepatic metastases displayed different allelotypes in comparison to the primary tumours [50]. Intriguingly, several reports showed that the KRAS mutation status of CRC-derived DTCs is not necessarily identical to those found in the corresponding primary tumour [51 53]. One final important biological property of DTC is dormancy, a non-proliferating state. This assumption is caused by the negativity for Ki-67, the gold standard biomarker for proliferating cells, and may also be responsible for partial resistance to adjuvant chemotherapy and the putative stem cell phenotype of DTCs [54, 55]. Moreover, one report suggested a cross-talk between ERK1/2 and p38α/β-signalling in DTC dormancy [56]. Implementation of mouse models in CTC/DTC research The rarity of CTCs/DTCs in CRC patients renders research with these cells cumbersome. In addition, as the biological properties of CTCs/DTCs must always be seen in context with the biology of the primary tumour, concurrent samples of primary tumour tissue, blood and BM from the same patient are required. This is obviously very difficult to achieve; therefore, mouse models of tumour cell dissemination in CRC may prove beneficial.

4 Although successfully metastasizing orthotopic tumour models as well as the feasibility of CTC isolation, quantification and characterization in mouse models of other tumour entities have been described [57 66]; there is currently no literature about CTCs/DTCs in mouse models of CRC. This is mostly due to technical issues; however, as several groups are currently working on these problems and major obstacles have already been overcome, mouse models of tumour cell dissemination in CRC are likely to be introduced in the near future. Clinical significance and prognostic value of CTCs/DTCs To date, numerous reports have been published on CTC detection and identification. In addition, the potential prognostic significance of CTCs in CRC has been intensively and extensively reviewed. Peach et al. showed in their systematic review that the occurrence of CTCs 24 h after resection of CRC represents an independent prognostic marker for future recurrence [67]. A recent meta-analysis by our group summarised all available studies on the prognostic impact of CTCs and DTCs in CRC [68]. Rahbari et al. concluded that the detection of CTCs in the PB is significantly associated with poor prognosis in CRC. In the following section, we will focus in detail on the current knowledge about the presence and clinical significance of CTCs and DTCs in CRC. Presence and clinical significance of CTCs in PB In theory, the detection and the count of CTCs reflect ongoing metastasis and can be used as an individual prognostic marker and may be linked to various clinical outcome parameters such as progression-free survival (PFS), disease-free survival (DFS), relapse-free survival (RFS) and overall survival (OS). Furthermore, decrease or at best disappearance of CTCs might serve as a predictive marker for response to cancer treatment [13]. In an early period of CTC research, several groups attempted to identify and detect CTCs in PB by reverse transcriptionpolymerase chain reaction (RT-PCR). Weitz et al. developed a sensitive and specific CK20 RT-PCR for the detection of CTCs of CRC [69]. In addition, Bessa et al. examined the presence of CTCs in PB from 95 consecutive patients with histologically confirmed CRC by assessing CEA mrna level. The study revealed no prognostic significance regarding the presence of CTCs in CRC [70]. Similarly, the same group could not ascertain the prognostic significance of postoperative CTC detection [71]. In contrast to this, Bosch et al. showed an independent prognostic significance for CTCs in patients undergoing surgery for CRC by using cytology and immunocytochemistry (ICC) [72]. Some years later, Koch et al. suggested that detection of CTCs in intraoperatively drawn PB samples may serve as an independent prognostic factor for tumour relapse after potentially curative resection of CRC liver metastases [73]. Beyond that, other studies evaluated CTC presence 24 h after curative resection of primary CRC by CEA or CK20 RT-PCR and demonstrated that the detection of CTCs is associated with an increased rate of recurrence [74]. Several consecutively performed studies tried to evaluate clinical prognostic values by introducing new biomarkers. For instance, Koyanagi et al. examined c-met, MAGEA3, GalNAc-T and CK20 and demonstrated a significant correlation of these markers with shorter DFS in CRC patients [75]. Surprisingly, the analysis of Lankiewicz et al. revealed a chemotherapy-induced biomarker expression change (CIBEC) in CRC-derived CTCs. The observed CIBEC comprised decreasing CEA and increasing EGFR expression levels [41]. Interestingly, several groups used multimarker detection technologies like Uen et al. who investigated the CTC status in 438 CRC patients by applying a membrane array targeting htert, CK19/20 and CEA. They demonstrated that patients with persistent presence of CTCs after curative resection had a worse RFS [76]. With the introduction of immunomagnetic detection devices such as the CellSearch System (Veridex, Warren, NJ), not only detection but also quantification of the CTCs became possible [77]. Currently, it is known that 30 40% of patients with mcrc show 3 CTC/7.5 ml blood [78]. In a multiinstitutional trial by Cohen et al., 430 mcrc patients were analysed for CTC count at baseline and during treatment using the CellSearch System. Measurement of CTCs was used to subdivide the entire patient cohort into favourable (<3 CTC/ 7.5 ml) and unfavourable ( 3 CTC/7.5 ml) prognostic groups. Elevated CTC counts were associated with a worse PFS and OS. Interestingly, patients who exhibited a therapyrelated decrease of CTCs revealed a significantly better outcome compared to those patients with constantly elevated CTC counts despite treatment [79]. In addition, Cohen et al. demonstrated that the baseline CTC level represents an important prognostic factor for specific subgroups defined by treatment or patient characteristics [80]. Another large study (Cairo-2) regarding CTC detection and numbers in 755 mcrc patients confirmed the findings of Cohen et al. [81]. In addition, Wong et al. confirmed the association of CTC count with higher stage and prognosis in varying stages of disease [82]. Interestingly, metastatic patients do not exclusively harbour CTCs in PB. Allard et al. reported that merely 30% of mcrc patients showed 2 CTC/7.5 ml blood, whereas the remaining patients were free of CTCs [77]. In conclusion, the abovementioned studies and the recently performed meta-analysis by our group have clearly shown that CTC detection and quantification may develop into a clinically relevant prognostic and predictive surrogate marker for the individual CRC patient. Presence and clinical significance of DTCs in BM Interestingly, DTCs are often detectable in the absence of LN (N0) or distant metastases (M0) [54, 83, 84]. In particular,

5 BM-associated DTCs in M0 patients strongly correlate with a decreased RFS and OS for CRC [85]. In addition, various reports revealed a significant correlation among metastatic relapse and the presence of DTCs in BM in diverse tumour types [46]. Several groups attempted to catch DTCs by utilising a pan-ck antibody and a cocktail of several anti-epithelial antibodies [46, 86 89]. The first study with a relevant number of patients was conducted by Flatmark et al. and included 275 CRC and 206 non-cancer patients in which DTCs were targeted by EpCAM. Unfortunately, the study failed to demonstrate a significant correlation of DTCs with disease stage or other clinical parameters [90]. Further studies have shown various results: no correlation among prognostic factors and the presence of BM-associated DTCs was shown in three studies [87, 91, 92] and two reports revealed a positive correlation of DTCs in BM with an increased recurrence rate [83, 93] or reduced OS, respectively [88, 94]. Ample studies were carried out and focussed on the detection of CK20 as a DTC marker. For instance, Vlems et al. failed to show a correlation of CK20 + cells in BM with survival by investigating BM of metastatic patients [95]. Other studies demonstrated an independent prognostic role for CK20 on mrna level [73, 96, 97]. Beyond utilising CK18 or CK20 as DTC detection markers, one small study examined the presence of CEA + -DTCs in ten patients [98]. The authors succeeded in detecting CEA + -DTCs in 66% of the investigated cases, whereas no cells were detected in the healthy control samples. Alike, the concurrent detection of CK20 and guanylyl cyclase C (GCC) showed 11% and 6% detection rates of DTCs in BM. Just one of the 109 investigated cases exhibited concurrent CK20 and GCC expression. However, no clinical correlation could be demonstrated for the single or collective expression [99]. Ausch et al. were interested in the correlation of soluble CK18 fragments (M30 and M65) which are released from human cancer cells during cell death and may indicate the presence of DTCs in BM. The serum level of M65 (necrosis and apoptosis) and M30 (apoptosis) was quantified preoperatively and postoperatively using specific enzyme-linked immunosorbent assays and DTCs were detected by A45-B/B3 staining of aspirates. Within that study, M65 seems to be an additional surrogate marker to identify patient subgroups with a high incidence of systemic disease [100]. Another study examined the prognostic impact of EMPIRIN and evaluated its reliable application as a biomarker to detect DTCs in BM. Even though only 16% of DTCs in BM of CRC patients were stained positive for EMPIRIN, the authors postulated an association with clinical markers of tumour progression in patients with colorectal/gastric cancer [101]. Recently, Vogelaar et al. chose a different way to detect DTCs in BM by applying an ICC staining for CK (CK-ICC) in combination with automated microscopy or indirectly using RT-PCR. The authors were able to demonstrate that patients free of DTCs in BM as assessed by RT-PCR had a significantly better DFS after resection of their liver metastases (p00.02) and a significantly better OS (p00.002). Contradictory, however, CK-ICC did not reveal a worse clinical outcome [102]. By examining the follow-up of DTCs in the BM of CRC patients, Buxhofer-Ausch et al. reported that the presence of DTCs in the preoperative BM has no impact on the prognostic outcome. Remarkably, after 12 months, the BM status had changed in a quarter of patients [103]. Another recent study identified an association of DTC presence with adverse prognosis of the curatively resected patients [104]. In conclusion, the available reports regarding the presence and clinical impact of DTCs in CRC-associated BM depict a fairly heterogeneous picture. The heterogeneity is due to the non-standardised study designs and variable detection methods which makes a valid comparison difficult. Our conclusion is supported by Rahbari et al. who deduced the lack of a clear prognostic impact of DTC in CRC to the heterogeneity of study findings as well. Moreover, Rahbari et al. suggested that this could be due to the currently still low number of available studies [68]. Presence and clinical significance of DTCs in LN Besides BM, DTCs are most notably found in LN. Detection and identification of DTCs in LN are carried out by methods of conventional pathology, immunohistochemistry (IHC) and molecular biology. Detection rates of DTCs in histopathologically tumour-free LN (pn0) vary between 19% and 86%. In most instances, the antibodies employed in IHC stainings are directed against different CKs (8, 18 and 19) [2]. In addition, most of the reports regarding the clinical significance of DTCs in LN came to no clear conclusion. This facet will be discussed subsequently. Greenson et al. showed that the presence of CK + cells within LN correlated with a significantly poorer prognosis [105]. In contrast, Öberg et al. and Nakanishi et al. failed to show an influence of lymphatic DTC detection on survival [106, 107]. The next study investigating the extent of CRC cell dissemination in LN was conducted by Weitz et al., who utilised a CK20 RT-PCR assay to examine 279 LN samples. The data of Weitz et al. strongly suggested prognostic relevance of DTC detection in the apical LN and postulated that DTC detection could improve tumour-staging protocols [108]. By using CK20 and GCC to identify DTCs in LN, Conzelmann et al. detected 79% CK20 + and 68% GCC + cells in the LN of CRC patients [99]. Due to the uncertain prognostic value of DTC in histopathologically negative LN of CRC patients, Rosenberg et al. investigated 170 LN of 85 patients with curatively resected CRC at UICC stage I or II for the presence of CEA + -, CK and EpCAM + -DTCs. DTCs were identified in 27%, 28% and 27% of the

6 examined patients, respectively. Furthermore, triple positive cells could be identified in 21% of cases. In addition, the presence of CEA + - and CK20 + -DTCs correlated significantly with a worse OS (p<0.01). Interestingly, Rosenberg et al. identified the tumour site (colon versus rectal cancer, p<0.006) and the presence of CEA + -DTCs (p<0.03) as independent prognostic factors in a multivariate analysis [109]. Moreover, one study attempted to analyse the presence of DTCs in LN as well as the KRAS mutation status of the corresponding primary tumour synchronously. DTC presence was significantly higher in patients with KRAS codon 13 mutations in the primary tumour [110]. In the recent systematic review and meta-analysis, our group demonstrated that the molecular detection of DTCs in regional LN is strongly associated with an increased risk of disease recurrence and poor survival in patients with pn0 CRC [111]. Conclusion and future directions Since the idea that CTCs/DTCs may be responsible for metastasis has arisen, enormous effort was made to analyse the biological and clinical impact of those cells [7]. Moreover, scientists all over the world have developed different strategies and applicable technologies or devices to reliably identify, isolate and further analyse assumed CTCs/DTCs. Based on these technologies and strategies, many studies have been conducted to elucidate the clinical relevance and prognostic significance of CTCs/DTCs. Without any doubt, CTCs and DTCs have the potential for a suitable biomarker in the era of personalised cancer care. However, due to heterogenous detection methods and lack of prospective multicenter trials, CTCs/DTCs have not entered into clinical routine as a valid biomarker until now. Furthermore, the CTC/DTC research field is much more complex than previously assumed. Particularly, the heterogeneous biology of CTC makes it hard to believe that CTCs/DTCs express a single or a set of biomarkers uniformly. We are still in the early stages of entering into the mystery of CTC/DTC biology, and further research in this field should focus on a more precise molecular characterisation of these cells. This task can only be accomplished by a combined effort of different research groups, as exemplified by the KFO 227 (Clinical Research Group 227: Colorectal cancer: from primary tumour progression towards metastases) which is supported by the DFG. Conflicts of interest References None. 1. Pantel K, Brakenhoff RH (2004) Dissecting the metastatic cascade. Nat Rev Cancer 4(6): Wolfrum F et al (2005) Detection and clinical implications of minimal residual disease in gastro-intestinal cancer. Langenbecks Arch Surg 390(5): Ferlay J, Parkin DM, Steliarova-Foucher E (2010) Estimates of cancer incidence and mortality in Europe in Eur J Cancer 46 (4): Thorsteinsson M, Jess P (2011) The clinical significance of circulating tumor cells in non-metastatic colorectal cancer a review. Eur J Surg Oncol 37(6): Weitz J et al (2005) Colorectal cancer. Lancet 365(9454): Hess KR et al (2006) Metastatic patterns in adenocarcinoma. Cancer 106(7): Ashworth, TR (1869) A case of cancer in which cells similar to those in the tumours were seen in the blood after death. Aust Med J 14: Zhan M, Zhao H, Han ZC (2004) Signalling mechanisms of anoikis. Histol Histopathol 19(3): Sachdev D et al (2010) The type I insulin-like growth factor receptor regulates cancer metastasis independently of primary tumor growth by promoting invasion and survival. Oncogene 29(2): Fidler IJ (2003) The pathogenesis of cancer metastasis: the seed and soil hypothesis revisited. Nat Rev Cancer 3(6): Chaffer CL, Weinberg RA (2011) A perspective on cancer cell metastasis. Science 331(6024): Thiery JP et al (2009) Epithelial-mesenchymal transitions in development and disease. Cell 139(5): Sun YF et al (2011) Circulating tumor cells: advances in detection methods, biological issues, and clinical relevance. J Cancer Res Clin Oncol 137(8): Yang J, Weinberg RA (2008) Epithelial-mesenchymal transition: at the crossroads of development and tumor metastasis. Dev Cell 14 (6): Morel AP et al (2008) Generation of breast cancer stem cells through epithelial-mesenchymal transition. PLoS One 3(8):e Sabbah M et al (2008) Molecular signature and therapeutic perspective of the epithelial-to-mesenchymal transitions in epithelial cancers. Drug Resist Updat 11(4 5): Vincan E et al (2007) A human three-dimensional cell line model allows the study of dynamic and reversible epithelial-mesenchymal and mesenchymal-epithelial transition that underpins colorectal carcinogenesis. Cells Tissues Organs 185(1 3): Loboda A et al (2011) EMT is the dominant program in human colon cancer. BMC Med Genomics 4:9 19. Lecharpentier A et al (2011) Detection of circulating tumour cells with a hybrid (epithelial/mesenchymal) phenotype in patients with metastatic non-small cell lung cancer. Br J Cancer 105(9): Hiratsuka S et al (2006) Tumour-mediated upregulation of chemoattractants and recruitment of myeloid cells predetermines lung metastasis. Nat Cell Biol 8(12): Kang SY et al (2009) Prosaposin inhibits tumor metastasis via paracrine and endocrine stimulation of stromal p53 and Tsp-1. Proc Natl Acad Sci U S A 106(29): Kaplan RN et al (2005) VEGFR1-positive haematopoietic bone marrow progenitors initiate the pre-metastatic niche. Nature 438 (7069): Mehlen P, Puisieux A (2006) Metastasis: a question of life or death. Nat Rev Cancer 6(6): Weiss L (1990) Metastatic inefficiency. Adv Cancer Res 54: Gout S, Tremblay PL, Huot J (2008) Selectins and selectin ligands in extravasation of cancer cells and organ selectivity of metastasis. Clin Exp Metastasis 25(4): Camerer E et al (2004) Platelets, protease-activated receptors, and fibrinogen in hematogenous metastasis. Blood 104(2): Nieswandt B et al (1999) Lysis of tumor cells by natural killer cells in mice is impeded by platelets. Cancer Res 59 (6):

7 28. Hou JM et al (2011) Circulating tumor cells as a window on metastasis biology in lung cancer. Am J Pathol 178(3): Pawelek JM, Chakraborty AK (2008) The cancer cell leukocyte fusion theory of metastasis. Adv Cancer Res 101: Abdel-Ghany M et al (2001) The breast cancer beta 4 integrin and endothelial human CLCA2 mediate lung metastasis. J Biol Chem 276(27): Brown DM, Ruoslahti E (2004) Metadherin, a cell surface protein in breast tumors that mediates lung metastasis. Cancer Cell 5(4): Wang H et al (2004) Tumor cell alpha3beta1 integrin and vascular laminin-5 mediate pulmonary arrest and metastasis. J Cell Biol 164 (6): Kim MY et al (2009) Tumor self-seeding by circulating cancer cells. Cell 139(7): Chang YS et al (2000) Mosaic blood vessels in tumors: frequency of cancer cells in contact with flowing blood. Proc Natl Acad Sci U S A 97(26): Brooks DE (1984) The biorheology of tumor cells. Biorheology 21(1 2): Pantel K, Brakenhoff RH, Brandt B (2008) Detection, clinical relevance and specific biological properties of disseminating tumour cells. Nat Rev Cancer 8(5): Smirnov DA et al (2005) Global gene expression profiling of circulating tumor cells. Cancer Res 65(12): Solmi R et al (2006) Microarray-based identification and RT-PCR test screening for epithelial-specific mrnas in peripheral blood of patients with colon cancer. BMC Cancer 6: Zieglschmid V et al (2007) Tumor-associated gene expression in disseminated tumor cells correlates with disease progression and tumor stage in colorectal cancer. Anticancer Res 27(4A): Lankiewicz S et al (2008) Tumour-associated transcripts and EGFR deletion variants in colorectal cancer in primary tumour, metastases and circulating tumour cells. Cell Oncol 30(6): Lankiewicz S et al (2008) Circulating tumour cells as a predictive factor for response to systemic chemotherapy in patients with advanced colorectal cancer. Mol Oncol 2(4): Findeisen P et al (2008) Systematic identification and validation of candidate genes for detection of circulating tumor cells in peripheral blood specimens of colorectal cancer patients. Int J Oncol 33 (5): Gazzaniga P et al (2010) Molecular markers in circulating tumour cells from metastatic colorectal cancer patients. J Cell Mol Med 14(8): Wong SC et al (2011) Clinical significance of CDX2-positive circulating tumour cells in colorectal cancer patients. Br J Cancer 104(6): Xi L et al (2007) Optimal markers for real-time quantitative reverse transcription PCR detection of circulating tumor cells from melanoma, breast, colon, esophageal, head and neck, and lung cancers. Clin Chem 53(7): Riethdorf S, Wikman H, Pantel K (2008) Review: biological relevance of disseminated tumor cells in cancer patients. Int J Cancer 123(9): Klein CA et al (2002) Genetic heterogeneity of single disseminated tumour cells in minimal residual cancer. Lancet 360(9334): Stoecklein NH, Klein CA (2010) Genetic disparity between primary tumours, disseminated tumour cells, and manifest metastasis. Int J Cancer 126(3): Klein CA (2009) Parallel progression of primary tumours and metastases. Nat Rev Cancer 9(4): Weber JC et al (2007) Allelotyping analyses of synchronous primary and metastasis CIN colon cancers identified different subtypes. Int J Cancer 120(3): Conzelmann M, Linnemann U, Berger MR (2005) Molecular detection of clinical colorectal cancer metastasis: how should multiple markers be put to use? Int J Colorectal Dis 20(2): Kraus MC, Linnemann U, Berger MR (2011) Disseminated colorectal tumor cells in organs prone to metastasis detected by new double enriched nested-pcr in comparison with recognized assays. Oncol Rep 25(5): Tortola S et al (2001) Discordance between K-ras mutations in bone marrow micrometastases and the primary tumor in colorectal cancer. J Clin Oncol 19(11): Braun S et al (2000) Lack of effect of adjuvant chemotherapy on the elimination of single dormant tumor cells in bone marrow of high-risk breast cancer patients. J Clin Oncol 18(1): Wikman H, Vessella R, Pantel K (2008) Cancer micrometastasis and tumour dormancy. APMIS 116(7 8): Sosa MS et al (2011) ERK1/2 and p38alpha/beta signaling in tumor cell quiescence: opportunities to control dormant residual disease. Clin Cancer Res 17(18): Alencar H et al (2005) A novel mouse model for segmental orthotopic colon cancer. Int J Cancer 117(3): Bhullar JS et al (2011) A novel nonoperative orthotopic colorectal cancer murine model using electrocoagulation. J Am Coll Surg 213(1):54 60, discussion Bresalier RS et al (1987) An animal model for colon cancer metastasis: establishment and characterization of murine cell lines with enhanced liver-metastasizing ability. Cancer Res 47(5): Donigan M et al (2010) A metastatic colon cancer model using nonoperative transanal rectal injection. Surg Endosc 24(3): Eliane JP et al (2008) Monitoring serial changes in circulating human breast cancer cells in murine xenograft models. Cancer Res 68(14): Fu XY et al (1991) Models of human metastatic colon cancer in nude mice orthotopically constructed by using histologically intact patient specimens. Proc Natl Acad Sci U S A 88(20): Galanzha EI et al (2009) In vivo, noninvasive, label-free detection and eradication of circulating metastatic melanoma cells using two-color photoacoustic flow cytometry with a diode laser. Cancer Res 69(20): Kolostova K et al (2011) Circulating human prostate cancer cells from an orthotopic mouse model rapidly captured by immunomagnetic beads and imaged by GFP expression. Anticancer Res 31(5): Lupu CM et al (2006) An orthotopic colon cancer model for studying the B7-H3 antitumor effect in vivo. J Gastrointest Surg 10 (5): Zigmond E et al (2011) Utilization of murine colonoscopy for orthotopic implantation of colorectal cancer. PLoS One 6(12):e Peach G et al (2010) Prognostic significance of circulating tumour cells following surgical resection of colorectal cancers: a systematic review. Br J Cancer 102(9): Rahbari NN et al (2010) Meta-analysis shows that detection of circulating tumor cells indicates poor prognosis in patients with colorectal cancer. Gastroenterology 138(5): Weitz J et al (1998) Dissemination of tumor cells in patients undergoing surgery for colorectal cancer. Clin Cancer Res 4 (2): Bessa X et al (2001) Lack of prognostic influence of circulating tumor cells in peripheral blood of patients with colorectal cancer. Gastroenterology 120(5): Bessa X et al (2003) Prognostic value of postoperative detection of blood circulating tumor cells in patients with colorectal cancer operated on for cure. Ann Surg 237(3): Bosch B et al (2003) Perioperative detection of disseminated tumour cells is an independent prognostic factor in patients with colorectal cancer. Br J Surg 90(7): Koch M et al (2005) Detection of hematogenous tumor cell dissemination predicts tumor relapse in patients undergoing surgical resection of colorectal liver metastases. Ann Surg 241 (2):

8 74. Allen-Mersh TG et al (2007) Role of circulating tumour cells in predicting recurrence after excision of primary colorectal carcinoma. Br J Surg 94(1): Koyanagi K et al (2008) Prognostic relevance of occult nodal micrometastases and circulating tumor cells in colorectal cancer in a prospective multicenter trial. Clin Cancer Res 14(22): Uen YH et al (2008) Persistent presence of postoperative circulating tumor cells is a poor prognostic factor for patients with stage I III colorectal cancer after curative resection. Ann Surg Oncol 15 (8): Allard WJ et al (2004) Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res 10(20): Negin BP, Cohen SJ (2010) Circulating tumor cells in colorectal cancer: past, present, and future challenges. Curr Treat Options Oncol 11(1 2): Cohen SJ et al (2008) Relationship of circulating tumor cells to tumor response, progression-free survival, and overall survival in patients with metastatic colorectal cancer. J Clin Oncol 26(19): Cohen SJ et al (2009) Prognostic significance of circulating tumor cells in patients with metastatic colorectal cancer. Ann Oncol 20(7): Tol J et al (2008) A randomised phase III study on capecitabine, oxaliplatin and bevacizumab with or without cetuximab in firstline advanced colorectal cancer, the CAIRO2 study of the Dutch Colorectal Cancer Group (DCCG). An interim analysis of toxicity. Ann Oncol 19(4): Wong SC et al (2009) Clinical significance of cytokeratin 20- positive circulating tumor cells detected by a refined immunomagnetic enrichment assay in colorectal cancer patients. Clin Cancer Res 15(3): Lindemann F et al (1992) Prognostic significance of micrometastatic tumour cells in bone marrow of colorectal cancer patients. Lancet 340(8821): Pantel K et al (1996) Frequency and prognostic significance of isolated tumour cells in bone marrow of patients with non-smallcell lung cancer without overt metastases. Lancet 347(9002): Leinung S et al (2000) Cytokeratin-positive cells in bone marrow in comparison with other prognostic factors in colon carcinoma. Langenbecks Arch Surg 385(5): Broll R et al (1996) Tumor cell dissemination in bone marrow and peritoneal cavity. An immunocytochemical study of patients with stomach or colorectal carcinoma. Langenbecks Arch Chir 381(1): Cohen AM et al (1998) In vitro detection of occult bone marrow metastases in patients with colorectal cancer hepatic metastases. Dis Colon Rectum 41(9): Juhl H et al (1994) Immunocytological detection of micrometastatic cells: comparative evaluation of findings in the peritoneal cavity and the bone marrow of gastric, colorectal and pancreatic cancer patients. Int J Cancer 57(3): Schott A et al (1998) Isolated tumor cells are frequently detectable in the peritoneal cavity of gastric and colorectal cancer patients and serve as a new prognostic marker. Ann Surg 227(3): Flatmark K et al (2002) Immunomagnetic detection of micrometastatic cells in bone marrow of colorectal cancer patients. Clin Cancer Res 8(2): Pantel K et al (1994) Methodological analysis of immunocytochemical screening for disseminated epithelial tumor cells in bone marrow. J Hematother 3(3): Schoppmeyer K et al (2006) Tumor cell dissemination in colon cancer does not predict extrahepatic recurrence in patients undergoing surgery for hepatic metastases. Oncol Rep 15(2): O Sullivan GC et al (1997) Micrometastases: marker of metastatic potential or evidence of residual disease? Gut 40(4): Leinung S et al (2000) Detection of cytokeratin-positive cells in bone marrow in breast cancer and colorectal carcinoma in comparison with other factors of prognosis. J Hematother Stem Cell Res 9 (6): Vlems FA et al (2003) Detection of disseminated tumour cells in blood and bone marrow samples of patients undergoing hepatic resection for metastasis of colorectal cancer. Br J Surg 90(8): Soeth E et al (1996) The detection of disseminated tumor cells in bone marrow from colorectal-cancer patients by a cytokeratin-20- specific nested reverse-transcriptase-polymerase-chain reaction is related to the stage of disease. Int J Cancer 69(4): Soeth E et al (1997) Comparative analysis of bone marrow and venous blood isolates from gastrointestinal cancer patients for the detection of disseminated tumor cells using reverse transcription PCR. Cancer Res 57(15): Gerhard M et al (1994) Specific detection of carcinoembryonic antigen-expressing tumor cells in bone marrow aspirates by polymerase chain reaction. J Clin Oncol 12(4): Conzelmann M et al (2003) Cytokeratin 20 and guanylyl cyclase C mrna is largely present in lymph node and liver specimens of colorectal cancer patients. Int J Cancer 107(4): Ausch C et al (2009) Circulating cytokeratin 18 fragment m65-a potential marker of malignancy in colorectal cancer patients. J Gastrointest Surg 13(11): Buergy D et al (2009) Prognostic impact of extracellular matrix metalloprotease inducer: immunohistochemical analyses of colorectal tumors and immunocytochemical screening of disseminated tumor cells in bone marrow from patients with gastrointestinal cancer. Cancer 115(20): Vogelaar FJ et al (2010) Clinical impact of different detection methods for disseminated tumor cells in bone marrow of patients undergoing surgical resection of colorectal liver metastases: a prospective follow-up study. BMC Cancer 10: Buxhofer-Ausch V et al (2010) Spontaneous changes in tumour cell dissemination to bone marrow in colorectal cancer. Colorectal Dis 12(8): Valladares-Ayerbes M et al (2011) Prognostic impact of disseminated tumor cells and microrna cluster deregulation in gastrointestinal cancer. Int J Oncol 39(5): Greenson JK et al (1994) Identification of occult micrometastases in pericolic lymph nodes of Duke s B colorectal cancer patients using monoclonal antibodies against cytokeratin and CC49. Correlation with long-term survival. Cancer 73(3): Nakanishi Y et al (1999) Clinical implications of lymph node micrometastases in patients with colorectal cancers. A case control study. Oncology 57(4): Oberg A et al (1998) Are lymph node micrometastases of any clinical significance in Dukes stages A and B colorectal cancer? Dis Colon Rectum 41(10): Weitz J et al (1999) Detection of disseminated colorectal cancer cells in lymph nodes, blood and bone marrow. Clin Cancer Res 5 (7): Rosenberg R et al (2004) Prognostic evaluation and review of immunohistochemically detected disseminatedtumorcellsin peritumoral lymph nodes of patients with pn0 colorectal cancer. Int J Colorectal Dis 19(5): Conzelmann M, Linnemann U, Berger MR (2004) K-ras codon 12 and 13 mutations are correlated with differential patterns of tumor cell dissemination in colorectal cancer patients. Int J Oncol 24 (6): Rahbari NN, Bork U, Motschall E, Thorlund K, Büchler MW, Koch M, Weitz J (2012) Molecular detection of tumor cells in regional lymph nodes is associated with disease recurrence and poor survival in node-negative colorectal cancer: a systematic review and metaanalysis. J Clin Oncol 30(1): doi: /jco