To identify novel non-invasive biomarkers for improved detection, risk assessment and prognostic evaluation of cancer, expression profiles of circulating microRNAs are currently under evaluation. Circulating microRNAs are highly promising candidates in this context, as they present some key characteristics for cancer biomarkers: they are tissue-specific with reproducible expression and consistency among individuals from the same species, they are potentially derived directly from the tumour and therefore might correlate with tumour progression and recurrence, and they are bound to proteins or contained in subcellular particles, such as microvesicles or exosomes, making them highly stable and resistant to degradation. The present review highlights the origin of circulating microRNAs, their stability in blood samples, and techniques to isolate exosomal microRNAs, and then addresses the current evidence supporting potential clinical applications of circulating miRNAs for diagnostic and prognostic purposes.

INTRODUCTION

The identification and validation of biomarkers that can be measured routinely in easily accessible samples and that are able to diagnose cancer, predict treatment efficacy and the risk of progression or relapse is a major challenge in cancer research. In recent years, much work has focussed on a novel class of molecules that are promising candidates in this context: miRNAs (miRs). miRNAs are a well-known class of regulatory molecules, which control translation and stability of mRNA post-transcriptionally. It is postulated that one single miRNA can regulate the expression of mRNA from hundreds of genes [1]. Since the first release of miRBase (a database that overviews all known miRNAs) in 2002, the number of miRNA loci has increased constantly, with 24521 miRNAs identified in 206 species in 2013 [2]. Interestingly, a considerable number of these miRNAs map to regions of the human genome that are known to be altered in cancer [3]. Esquela-Kerschner and Slack [4] established the term ‘oncomirs’ for miRNAs with oncogenic function, implying that abnormalities in miRNA expression might either directly result in de-differentiation of cells allowing tumour formation to occur or that, on the other hand, alterations of miRNA expression could be used for tumour classification, diagnosis and prognosis in cancer.

Most early studies investigated miRNA expression patterns in fresh frozen samples, such as tumour biopsies or resection specimens, or in paraffin-embedded samples. However, miRNAs can also present as ‘circulating miRNAs’ and be found in a variety of human body fluids, such as urine, saliva, amniotic fluid and pleural fluid [59]. Since the first detection of circulating placenta-specific miRNAs in plasma reported by Chim et al. [10] in 2008, numerous studies have investigated miRNAs in blood samples. In 2012, Russo et al. [11] presented a database called miRandola. This collated information from 89 studies and contained 2132 entries with 581 circulating miRNAs identified in 21 sample types.

It has been shown that circulating miRNAs present tissue-specific, reproducible and consistent expression among individuals in the same species [5], derive potentially directly from tumour tissue, thereby possibly correlating with tumour progression and recurrence [12,13], are highly stable in the circulation due to RNA-binding proteins [1417] or subcellular particles such as microvesicles or exosomes [1720], and are resistant to unfavourable physiological conditions, such as repeated freeze–thawing cycles or low/high pH [12,21]. Moreover, Weber et al. [6] examined the expression of miRNAs in plasma, saliva, tears, urine, amniotic fluid, colostrum, breast milk, bronchial lavage, cerebrospinal fluid, peritoneal fluid, pleural fluid and seminal fluid from normal individuals and found miR-335, miR-509-5p, miR-515-3p, miR-873 and miR-616 to be among the most abundant miRNAs present in all or most of the fluid types. Several of the highly abundant miRNAs in these fluids were common among multiple fluid types; however, some of the miRNAs were enriched in specific fluids. The authors concluded that fluid-type-specific miRNAs may have functional roles associated with the surrounding tissues [6]. Pigati et al. [22] reported that the release of miRNAs into blood, milk and ductal fluids is selective and differs between normal and malignant mammary epithelial cells, implying a potential correlation of miRNA release and malignancy. It therefore seems very likely that circulating miRNAs might be valuable biomarkers for cancer diagnostics and prog-nostic evaluation. The present review paper summarizes the origin of circulating miRNAs, their stability in blood samples and techniques to isolate exosomal miRNAs. It also highlights potential clinical applications for circulating miRNAs as biomarkers for diagnostic and prognostic evaluation. In particular, the present review focuses on applications relevant to gastrointestinal malignancies. Of the 15 most frequent malignancies worldwide, five are cancers of the gastrointestinal tract (colorectum, stomach, liver, oesophagus and pancreas) and circulating miRNAs show potential as clinically relevant biomarkers in these cancers.

CIRCULATING miRNAs IN BLOOD SAMPLES

Origin of circulating miRNAs

There is increasing evidence that circulating miRNAs can be detected in blood samples from healthy individuals and patients with cancer, and several different mechanisms of origin for these circulating miRNAs have been established recently. Circulating miRNAs can originate from:(i) active secretion via microvesicles, including exosomes [2325], (ii) active secretion in a protein–miRNA complex [15,24,25], (iii) circulating miRNAs originating from blood cells such as immunocytes [25], and (iv) passive secretion by tumour apoptosis and necrosis [2426]. Consequently, when actively secreted, circulating miRNAs can show certain characteristics of the underlying tumour, such as metastatic potential [2729] or chemosensitivity [3033]. Furthermore, when passively secreted, they might be used as biomarkers of cell destruction, for example in the context of myocardial infarction [26,34] or liver cirrhosis [35]. Figure 1 shows an overview of the different origins of circulating miRNAs.

Different sources of circulating miRNAs

Figure 1
Different sources of circulating miRNAs

In the cell nucleus, primary miRNAs (pri-miRNAs) are transcribed and cleaved by the microprocessor complex Drosha/DGCR8 into shorter miRNA precursors (pre-miRNA), and finally transported in the cytoplasm. Mature miRNAs are then released via different ways into the circulation. (i) Released in microvesicles after for example fusion of multivesicular bodies (MVBs) with the plasma membrane (exosomes) or the release by the endocytic membrane transport pathway (endosomes). (ii) Released in an miRNA–protein complex, as for example bound to Ago2, HDL and LDL. (iii) Encapsulated into apoptotic bodies by (tumour) apoptosis and tumour necrosis. (iv) Derived from blood cells

Figure 1
Different sources of circulating miRNAs

In the cell nucleus, primary miRNAs (pri-miRNAs) are transcribed and cleaved by the microprocessor complex Drosha/DGCR8 into shorter miRNA precursors (pre-miRNA), and finally transported in the cytoplasm. Mature miRNAs are then released via different ways into the circulation. (i) Released in microvesicles after for example fusion of multivesicular bodies (MVBs) with the plasma membrane (exosomes) or the release by the endocytic membrane transport pathway (endosomes). (ii) Released in an miRNA–protein complex, as for example bound to Ago2, HDL and LDL. (iii) Encapsulated into apoptotic bodies by (tumour) apoptosis and tumour necrosis. (iv) Derived from blood cells

Circulating exosomal miRNAs

Cells can release three types of vesicles: (i) apoptotic bodies (500 nm to 3 μm in diameter; released by cells undergoing apoptosis), (ii) microvesicles (100 nm to 1 μm in diameter; released directly via plasma budding) and (iii) nanovesicles, including exosomes (30–100 nm and 40–100 nm respectively in diameter; released by exocytosis from endosome multivesicular bodies) [36,37]. Exosomes were first discovered in the 1970s [38]. After initial ‘mis-interpretation’ as cell debris [39], exosomes have been found in recent years to play a key role in cell transport, immune function and cancer. Exosomes are released into the extracellular environment via fusion of multivesicular bodies with the plasma membrane [40]. Many cells, including reticulocytes, dendritic cells, B- and T-cells, mast cells, epithelial cells and tumour cells, have the capacity to release exosomes [40]. Exosomes are present in nearly all body fluids [41], such as blood [42], urine [43], ascites [44], amniotic fluid [41] and saliva [45]. Exosomes share a lot of common (structural, adhesion and transportation) proteins, but they also contain cell-specific proteins such as MHC class II or intestinal A33 when derived from B-cells and colorectal cancer cells respectively [4648]. Valadi et al. [40] demonstrated that exosomes carry mRNA and miRNA when released from mast cells. Interestingly, many of these RNA species seem to be specifically packaged exclusively into exosomes. Additionally, mRNAs in the exosomes are translatable and encode proteins that are often exosome-specific [40,49,50]. However, exosomal RNA profiles also differ between exosomes from different parental cells [51,52], and tumour-derived exosomes may contain some ‘tumour-specific’ or ‘tumour-enriched’ miRNAs [42]. The release of exosomes might be regulated by tumour suppressor genes or oncogenes, as exosomes are found more abundantly in the bloodstream of patients with cancer compared with healthy individuals [42,53]. In addition, release of exosomes appears to be regulated by a ceramide-dependent secretory machinery [23].

Circulating miRNA–protein complex

Turchinovich et al. [16] investigated miRNA expression in plasma samples from healthy controls, and showed that the expression of miR-16, miR-21 and miR-24 did not differ significantly in the samples, compared with the results obtained by simply centrifuging and filtering (0.22 μm filters) samples (i.e. with exosomes) or with ultracentrifuged samples (i.e. without exosomes). They concluded that >97% of the miRNA amount was ‘exosome-free’. These extracellular miRNAs were demonstrated to bind to Argonaute proteins (Ago2 in particular, which is part of the RNA-inducing silencing complex), and to be extremely stable. The authors suggested that some miRNA–Ago2 complexes might be directly released by cells either as a by-product of cell death, or in a paracrine manner [15]. Another study investigating all four Argonaute proteins showed that only 14 out of 43 miRNAs presented similar binding patterns to either Ago1 or Ago2 in blood plasma, and that Argonaute-specific miRNA profiles in blood cells were markedly different from those in plasma. From these findings, the authors concluded that many cell types contribute to circulating miRNA and the majority of these are likely to be non-blood cells [16]. Other researchers demonstrated that HDLs (high-density lipoproteins) and LDLs (low-density lipoproteins), both involved in the transport of triacylglycerols (triglycerides) and cholesterol, have the capability to carry miRNAs in their core [54,55].

Circulating miRNAs originating from blood cells

Previous studies have revealed that, at least in part, cell-free circulating miRNAs could originate from blood cells [16,56,57]. Pritchard et al. [56] compared the expression of ten miRNAs in plasma with blood cell counts. miRNA expression presented cell-type- and cell-count-specific patterns for red blood cells (miR-451, miR-92a, miR-16 and miR-486-5p) and myeloid blood cells (miR-223, let-7a, miR-197 and miR-574-3p). However, this has to be considered in the context of blood-cell-derived miRNAs, as expression of miRNAs, such as miR-16 and miR-451, can be affected by haemolysis [16,5658]. The contribution of blood cells to cell-free circulating miRNAs has been raised as a consideration in related cancer biomarker studies, and it has been suggested that some reported circulating cancer-associated miRNAs might be due to a secondary effect on blood cells and not originate from tumour cells.

Stability of circulating miRNAs

Circulating miRNAs are protected from endogenous RNase activity by either being bound to proteins or by being packed into microparticles, such as exosomes or microvesicles [17]. Li et al. [14] analysed the expression of selected miRNAs (miR-16, miR-223, miR-30a, miR-320b, let-7b, miR-92a, miR-423-5p and miR-21) and demonstrated that the vesicular structure of microvesicles provides general protection for stored miRNAs in healthy individuals. In addition, Taylor and Gercel-Taylor [42] investigated the stability of exosomal miRNAs under various storage conditions (short-time storage up to 96 h at 4°C compared with long-time storage up to 28 days at −70°C), and they found no significant differences in the stability of miRNAs between these groups. Finally, Köberle et al. [59] compared the stability of miRNAs packed in microvesicles with those bound to proteins after incubation with RNase A or RNase inhibitor. They showed that vesicle-associated miRNAs appeared to be more stable and more resistant towards RNase A treatment compared with protein-bound miRNAs [59].

Techniques for isolation of exosomal miRNAs

Today, various approaches have been established for the isolation of exosomal miRNAs. Of these, ultracentrifugation is by far the most widely accepted approach with potential for clinical application [60,93]. An alternative technique for miRNA isolation from exosomes is the immuno-isolation method. This uses magnetic beads coated with monoclonal antibodies (anti-HLA DP, DQ, DR) to couple exosomes through the proteins expressed on their surfaces [61,62]. An additional way to isolate exosomal miRNAs is the use of the commercially available ExoQuick™ kit. ExoQuick™ is a precipitation reagent that pellets exosomes simply by adding the reagent to pre-processed samples [63]. In addition a density gradient technique can also be used. For example, this is applied in the OptiPrep™ technique, which separates vesicles based on the iodixanol gradient [65]. Finally, a fairly easy technique and potentially useful step for reducing contamination with non-exosomal miRNA in the context of exosome isolation protocols might be treatment of serum samples with RNase, as this might allow vesicle-associated (i.e. better protected from RNase treatment) and non-vesicle-associated (i.e. less protected from RNase treatment) miRNAs to be distinguished [59].

All of the these methods for exosome isolation have distinct advantages and disadvantages. Momen-Heravi et al. [93] recently compared several of the isolation techniques for EVs (extracellular vesicles), including exosomes. Ultracentrifugation, for example, which is considered to be the current gold standard, is lengthy (4–5 h), requires an ultracentrifuge and recovers a relatively small proportion of the EVs. Furthermore, the sedimentation efficiency of EVs and sedimentation stability depend on the rotors being used, and viscosity also plays a role in the recovery of EVs [93]. On the other hand, immuno-isolation is a promising method for specific isolation and characterization of EVs and has been shown to be the most effective strategy for isolation of EVs when compared with differential centrifugation and density gradient separation, and this technique offers the advantage of flow cytometric, immunoblot, and electron microscopy analysis of bead–EVs complexes. However, this approach is not suited for large sample volumes, and captured EVs may not retain functionality [93].

It has been shown previously that ExoQuick™ produces the largest amount and highest quality of exosomal RNA and proteins when compared with ultracentrifugation, immuno-isolation or size-exclusion chromatography techniques. However, this technique isolates exosomes in general, and does not exhibit specificity for the originating cell [64,93]. Kalra et al. [65] investigated the density gradient technique compared with two other EV isolation techniques (differential centrifugation coupled with ultracentrifugation and epithelial cell adhesion molecule immunoaffinity pull-down) and showed that OptiPrep™ density gradient separation was superior to the two other approaches for isolating pure populations of exosomes. Finally, regarding the additional purification step of using RNase treatment, this technique is the least expensive approach, easiest to perform and can be performed using even smaller sample volumes [59]. However, it should be noted that surface proteins on vesicles can protect RNA from degradation, so in order for RNAse treatment to be effective it should be coupled with protease treatment [66].

General considerations about technical aspects of miRNA analysis in blood samples

As outlined in a recent review by Moldovan et al. [92], developing circulating miRNA profiles as biomarkers is still in the early stages. There are many aspects that contribute to variability or potential inconsistency of research results and these must be overcome before clinical translation can occur.

With regards to sample acquisition and miRNA extraction, several issues have to be considered when planning experiments and interpreting data. It has been shown that factors such selection of the appropriate sample (e.g. plasma or serum), collection tubes (EDTA, citrate or heparin), extraction methods (phenol/chloroform or silica: distinct differences in required fluid volume, yield, procedural contaminants etc.), quality and quantity control, fasting or blood draw timing all potentially have an impact on the results of miRNA analyses in blood samples.

With regards to miRNA profiling methods, there are a number of different technical approaches available, all of which present distinct advantages and disadvantages. These techniques include qRT–PCR (quantitative reverse transcription–PCR), microarrays, sequence-specific hybridization in solution followed by miRNA molecule counting based on reporter probes, and direct sequencing. Although qRT–PCR approaches are fairly easy to perform and provide the best sensitivity and specificity in combination with very high absolute quantification/accuracy and flexibility, this technique is somewhat limited by its moderate-to-low throughput. Microarray techniques provide very high throughput, but sensitivity/specificity/absolute quantification/accuracy and flexibility are lower compared with qRT–PCR approaches. In-solution hybridization yields moderate quality in all of the aspects mentioned above.

The most promising technology for miRNA profiling might be the next-generation sequencing, as it is the only technique that allows identification of novel miRNAs and is not limited to known miRNAs. In addition, this technique provides better specificity and flexibility (both very high) compared with microarray approaches and yields moderate sensitivity, but absolute quantification/accuracy. However, costs of this approach are very high, and significant computational infrastructure and bioinformatics support is required [92].

Circulating miRNAs: a surrogate for underlying malignant diseases?

Apart from the need to optimize and standardize methodology for circulating miRNA research, before clinical translation can be considered it is essential to confirm whether sensitive and specific circulating miRNA profiles can be clearly linked to specific tumour types and tumour stages. As a starting point towards this, it was first demonstrated that the amount of circulating EVs containing miRNA cargo is elevated in certain disease states, including various types of cancer [93,96]. However, since then there is strong debate as to whether profiles of circulating miRNAs reflect the expression pattern in the underlying tumours. Some authors have reported that circulating miRNAs in breast cancer or testicular germ cell tumour patients might not exhibit the same expression pattern as the underlying tumours [94,95]. However, others have shown that the expression pattern of miRNAs in circulating exosomes in lung cancer and ovarian cancer patients, for example, were similar to the expression profile in the underlying tumour [96,97,99], or that selected miRNAs were generally expressed lower in microvesicles in glioma patients, but correlated well with the tumour [98,99]. Chan et al. [95] studied miRNA profiles in paired breast cancer tissue and serum samples, and suggested that their data might indicate selective release of miRNAs from breast cancer tissue into the blood stream. Other studies have clearly demonstrated that specific miRNAs are selectively packaged into exosomes for export from tumour cells, and this results in a significant difference between the originating cancer cell miRNA profile compared with the cancer exosome miRNA profile [100,101]. Therefore it may not be reasonable to expect good concordance between tumour tissue miRNA profiles and tumour exosome miRNA profiles. Moreover, the issue of concordance between tissue and exosome miRNA profiles is probably less important than the question of whether miRNAs in the peripheral circulation correlate with tumour diagnosis and prognosis in cancer patients with sufficient sensitivity and specificity. This question will be examined in the following sections.

CLINICAL APPLICATIONS OF CIRCULATING miRNAs AS DIAGNOSTIC/PROGNOSTIC BIOMARKERS IN CANCER

Given the fact that circulating miRNAs are on the one hand very stable in the peripheral bloodstream, and on the other hand, even more importantly, mostly actively secreted into the blood, they might yield excellent biomarkers for cancer detection and prognosis. Hence analysis of expression profiles of circulating miRNAs might provide a new tool for cancer diagnostics and screening, or for prognostic evaluation in individual patients.

Circulating miRNAs as diagnostic markers

Molecular biomarkers that detect cancer should exhibit significant and reproducible differences in expression between patients with cancer compared with healthy or other non-cancer controls. If these criteria are met, then these biomarkers might be useful for cancer screening, diagnosis or surveillance. A number of studies have analysed the expression profiles of circulating miRNAs in a variety of cancers and investigated whether miRNA profiles differ between patients with compared with without cancer. There is increasing evidence demonstrating significant associations between the expression of certain miRNAs and different cancer types, highlighting their potential use as diagnostic biomarkers.

For ESCC (oesophageal squamous cell carcinoma), for example Hirajima et al. [30] have reported a significantly higher concentration of miR-18a in plasma from patients with cancer, compared with healthy volunteers [sensitivity/specificity: 86.8%/100%; AUC (area under the curve)=0.9449]. Another group [67] examined selected miRNAs and showed that relative expression levels of circulating miR-155 and miR-183 were significantly reduced in patients with cancer, compared with controls. For miR-155, the authors determined an AUC of 0.66 [67]. A further study examining selected miRNAs revealed significantly higher plasma levels of miR-21 in patients with cancer compared with healthy volunteers, whereas the levels of miR-375 and miR-184 were significantly lower. In addition, the miR-21/miR-375 ratio was significant higher in patients with cancer, and the authors calculated an AUC value of 0.816 for the miR-21/miR-375 ratio. Interestingly, the authors found a reduction in miR-21 levels after tumour removal. However, these results should be interpreted with caution as miR-184 could not be detected in 50% of the patients with cancer due to its low expression [68]. Finally, Tanaka et al. [69] showed that serum levels of miR-21, miR-145, miR-200c and let-7c were significantly higher in patients with cancer compared with healthy volunteers. Sheinerman et al. [70] compared the miRNA expression pattern in plasma samples of patients with gastrointestinal diseases (including oesophageal cancer, gastric cancer, colon cancer and Crohn's disease) and in patients with pulmonary diseases, such as asthma, pneumonia and non-small-cell lung cancer. By using pairs of specific miRNAs, the authors could distinguish between patients with gastrointestinal disease and healthy controls with an accuracy of 95% (miRNA pairs: miR-215/miR-30e-3p, miR-215/miR-145, miR-203/miR-30e-3p and miR-203/miR-145) or patients with pulmonary disease with an accuracy of 94% (miRNA pairs: miR-145/miR-155, miR-486-5p/miR-155, miR-145/miR-30e-3p and miR-192/miR-31). Unfortunately, the authors failed to provide details about the subtype of oesophageal cancer included in this study [70].

In gastric cancer, several authors have reported similar deregulation of circulating miRNAs in patients with cancer compared with healthy controls. Li et al. [71], for example, demonstrated an up-regulation of miR-21 and miR-223, and down-regulation of miR-218 in patients with cancercompared with controls (miR-21: sensitivity/specificity, 74.29%/75.71%, AUC=0.7432; miR-218: sensitivity/specificity, 94.29%/44.29%, AUC=0.7432; and miR-223: sensitivity/specificity, 84.29%/88.57%, AUC=0.9089) [71]. Another group confirmed significant miR-21 up-regulation in patients with gastric cancer compared with other malignancies (sensitivity/specificity, 56.7%/94.9%, AUC=0.81) [72]. Tsujiura et al. [73] showed plasma concentrations of miR-17-5p, miR-106a, miR-106b and again of miR-21 to be significantly higher in patients with cancer compared with controls, whereas the levels of let-7a were significantly reduced. By using expression ratios between up- compared with down-regulated miRNAs, the authors demonstrated further improvement in sensitivity and specificity respectively in AUC (up to 0.879). Interestingly, the authors investigated further whether the miRNA deregulation observed in the plasma samples could also be seen in the miRNAs from the corresponding cancer tissue specimens. They showed that miR-106b had a higher expression in primary gastric cancer tissue samples compared with adjacent normal mucosa from seven out of the eight patients (87.5%), whereas let-7a showed lower expression in seven of these patients. The authors concluded that the miRNA levels from the primary cancer tissue and the plasma samples showed similar tendencies in most patients, and suggested that the levels of plasma miRNAs might reflect the expression level of the underlying tumour [73,74]. Zhu et al. [74] recently published a detailed review and meta-analysis, which addressed deregulation of circulating miRNAs in patients with gastric cancer and included 18 articles from 2010 to 2013. The review highlighted alterations in expression levels of miR-21, miR-27a and miR-106b, and emphasized that five studies reported the up-regulation of miR-21 in patients with cancer [74].

Regarding colorectal cancer, there have been only a few publications that address the potential for circulating miRNAs to be used for diagnosis in this cancer type. Giráldez et al. [75], for example, performed a genome-wide profiling study of circulating miRNAs in a large number of plasma samples from patients with colorectal cancer, pre-malignant neoplastic lesions, such as advanced adenomas, and healthy controls. They found six miRNAs (miR-15b, miR-18a, miR-19a, miR-19b, miR-29a and miR-335) were significantly up-regulated in patients with cancer compared with controls (AUC between 0.80 and 0.70). Interestingly, in patients with colonic adenomas, only miR-18a was significantly up-regulated (AUC=0.64) [75]. Another study screened using microfluidic array technology the expression of 380 miRNAs in plasma samples from healthy controls compared with patients with adenomas or colorectal cancer [76]. The authors showed that a panel of eight miRNAs (namely miR-15b, miR-17, miR-142-3p, miR-195, miR-331, miR-532, miR-532-3p and miR-652) accurately identified patients with adenomas (AUC=0.868). Furthermore, a five-miRNA panel (miR-15b, miR-21, miR-142-3p, miR-331 and miR-339-3p) distinguished patients with advanced adenomas from patients with cancer (AUC=0.856) [76].

For pancreatic cancer, again a fairly large body of evidence has confirmed deregulation of circulating miRNAs in blood samples from patients with cancer compared with controls. Morimura et al. [77] demonstrated a significant up-regulation of miR-18a levels in plasma from patients with pancreatic cancer compared with healthy controls (AUC=0.9369). Furthermore, Wang et al. [78] reported the up-regulation of miR-21, miR-210, miR-155 and miR-196a in patients with ductal adenocarcinoma compared with healthy controls (sensitivity/specificity for this panel: 64%/89%, AUC=0.82). Similarly, Ho et al. [79] showed a 4-fold up-regulation of miR-210 (normalized to miR-54) in plasma samples from patients with pancreatic cancer compared with controls. Another study investigated miRNA expression in serum from patients with ductal adenocarcinoma, chronic pancreatitis and healthy controls [80]. They investigated the expression of seven selected miRNAs (miR-21, miR-155, miR-181a, miR-181b, miR-196a, miR-221 and miR-222), which were validated in previous studies and shown to be highly expressed in pancreatic ductal adenocarcinoma tissue. Interestingly, up-regulation of miR-21 distinguished patients with cancer from healthy controls or patients with pancreatitis, whereas miR-155 and miR-196a were up-regulated in patients with either cancer or pancreatitis compared with healthy controls [80]. Liu et al. [81] demonstrated the up-regulation of miR-26 and miR-196a in patients with cancer compared with controls. Furthermore, these authors showed that a combination of miR-26, miR-196a and the standard tumour marker CA 19-9 (carbohydrate antigen 19-9) achieved an extraordinarily high sensitivity and specificity of 92.0% and 95.6% respectively (AUC=0.979) for cancer compared with controls [81]. Another research group reported that the expression of a panel of seven miRNAs (miR-20a, miR-21, miR-24, miR-25, miR-99a, miR-185 and miR-191) was significantly up-regulated in patients with ductal adenocarcinoma compared with controls (accuracy 86.8%, AUC=0.993). Most interestingly, the accuracy of this panel was far better than established tumour markers such as CA 19-9 and CEA (carcinoembryonic antigen) [82]. Furthermore, Li et al. [83] detected significantly elevated levels of miR-1290 in patients with pancreatic cancer in comparison with healthy controls or patients with chronic pancreatitis (AUC=0.72), patients with pancreatic neuroendocrine tumours and patients with intraductal papillary mucinous neoplasm (AUC=0.76).

For HCC (hepatocellular carcinoma), a number of studies have reported that deregulation of circulating miRNAs correlated with tumours. Shen et al. [84], for example, found in a first set of plasma samples from patients with HCC, compared with healthy controls, seven miRNAs (miR-19b-1, miR-24, miR-29c, miR-92a, miR-376a, miR-378 and miR-520c-3p) to be significantly differentially expressed in patients with cancer. A validation cohort confirmed only miR-483-5p to be associated with an increased risk of HCC (sensitivity/specificity, 55.1%/85.7%) [84]. Luo et al. [85] found miR-122a levels were significantly lower in serum samples from patients with HCC compared with controls (sensitivity/specificity, 70.6%/67.1%, AUC=0.707). In combination with AFP (α-feto-protein), the detection rate of HCC could be improved further (sensitivity/specificity, 87.1%/98.8%) [85]. However, these data have to be interpreted with caution as miR-122a might also be involved in viral replication in HCV (hepatitis C virus)- and HBV (hepatitis B virus)-related liver cancer [86,87]. In addition, another study reported that five plasma miRNAs (miR-23a, miR-23b, miR-342-3p, miR-375 and miR-423) were up-regulated in HBV-related HCC patients, thereby providing a diagnostic miRNA signature to discriminate HBV-related HCC from controls (sensitivity/specificity, 99.9%/99.4%, AUC=0.999). Interestingly, the authors reported further a signature of four up-regulated miRNAs (miR-10a, miR-223, miR-375 and miR-423), which differentiate patients with hepatitis B from healthy controls (sensitivity/specificity, 99.3%/98.8%, AUC=0.999) [88]. Finally, Li et al. [89] found three miRNAs (miR-21, miR-222 and miR-224) were significantly overexpressed in serum samples of patients with HCC compared healthy volunteers.

To summarize, there is increasing evidence that levels of certain circulating miRNAs in patients with gastrointestinal malignancies (upper gastrointestinal, lower gastrointestinal and hepato-pancreato-biliary) differ from those in healthy or non-malignant controls. Table 1 summarizes the data described above. This demonstrates the emerging potential for circulating miRNAs to be potent novel biomarkers for screening, diagnosis or surveillance of cancer.

Table 1
Circulating miRNAs as diagnostic markers

An overview of the deregulation of specific miRNAs in gastrointestinal malignant tumours from the upper to the lower gastrointestinal tract, including hepato-pancreato-biliary tumours, compared with healthy or non-malignant controls. CA, cancer; GA, gastric cancer; CP, chronic pancreatitis; IPMN, intraductal papillary mucinous neoplasms; BPD, benign pancreatic disease; HBV-rel., HBV related.

graphic
 
graphic
 

Circulating miRNAs as prognostic markers

Another important aspect of clinical decision making for patients with cancer is the consideration of clinically relevant tumour characteristics, such as prognosis, metastatic potential or resistance towards conventional chemotherapies, and the ability to reliably predict these opens the possibility of tailored or individualized cancer treatment. Molecular biomarkers that predict cancer prognosis need to present significant and reproducible expression differences between different prognostic subgroups. For example, patients with locally advanced malignancy, lymph node involvement or metastases have a poorer prognosis, and if biomarkers could be used as surrogate markers that predict these factors and prognosis, or likely success of treatment, then treatments might be tailored to the individual patient and tumour. Indeed, some recent early studies have addressed the question whether or not circulating miRNAs might be potential prognostic biomarkers in different cancer types. Although only a few studies have addressed this, initial evidence highlights the potential for these markers.

For example, in ESCC, expression levels of miR-18a in plasma have been shown to be significantly higher in patients with early stage tumours, i.e. pTis or stage I tumours [30]. Another research group has demonstrated a significantly increased risk for ESCC to be associated with reduced expression of circulating miR-155 and miR-183 (even after adjusting their data for potential confounders, such as smoking status and alcohol consumption) [67]. Komatsu et al. [68] found that ESCC patients with high plasma miR-21 levels showed a tendency to a shorter long-term survival, whereas patients with elevated miR-375 levels survived longer. Interestingly, patients with both high miR-21 and low miR-375 levels had a significantly worse prognosis than patients with the opposite expression pattern for these miRNAs (3-year survival rate: 48.4 compared with 83.1%). Multivariate analysis confirmed the combination of high miR-21 and low miR-375 levels in plasma to be an independent prognostic factor [hazard ratio, 3.8 (1.14–12.5)] [68]. Currently, only one study has investigated the potential for circulating serum miRNAs to predict treatment response in ESCC. Tanaka et al. [69] found that low levels of miR-200c correlated with a good clinical response to 5-FU (5-fluorouracil) and cisplatin plus adriamycin chemotherapy regimens (evaluated via computer tomography, endoscopy and PET/CT), but not with a 5-FU and cisplatin plus docetaxel regimen (P=0.7167). High expression of miR-200c was also associated with shortened progression-free survival. The authors found no significant relationship between expression levels of miR-200c in pre-treatment biopsies of the primary tumour and response to chemotherapy [69].

In gastric cancer, Zheng et al. [90] demonstrated a significant association between elevated miR-21 levels and increasing TNM (tumour node metastasis) stage and tumour size (AUC value up to 0.853). Zhu et al. [74] evaluated in their meta-analysis the prognostic value of circulating miRNAs, and found that elevated levels of circulating miR-17-5p, miR-20a, miR-21 and miR-200c were significantly correlated with overall survival and predicted poor prognosis.

For patients with colorectal cancer, similar deregulations of certain miRNAs in different outcome-relevant subgroups have been described. Toiyama et al. [27] identified miR-200c as the serum miRNA being best associated with metastasis. These authors found significantly higher serum miR-200c levels with stage IV cancer compared with stages I–III, and high expression miR-200c was also significantly associated with lymph node metastasis, liver metastasis and the development of distant metastases. Furthermore, high miR-200c levels represented an independent predictor of lymph node metastasis in these patients [hazard ratio, 4.81 (95% confidence interval, 1.98–11.7)] [27]. Another two studies investigated the potential for circulating miRNAs to predict treatment response. Zhang et al. [31] used TaqMan low-density arrays to screen miRNA expression in serum from a large patient cohort that underwent chemotherapy and compared the expression pattern with treatment response. They found 17 serum miRNAs correlated with chemosensitivity, and finally selected a panel of five miRNAs (miR-20a, miR-130, miR-145, miR-216 and miR-372), which were significantly down-regulated in the responder group. This panel reached an AUC value of up to 0.918 for its ability to distinguish between responders and non-responders (AUC value for CEA=0.689; and AUC value for CA19-9=0.746) [31]. Finally, another research group compared the expression of 742 miRNAs in plasma samples from cancer patients before commencement and after four cycles of 5-FU/oxaliplatin chemotherapy. They identified three miRNAs (miR-106a, miR-130b and miR-484) which were significantly up-regulated before treatment in non-responders. Moreover, high expression of miR-27b, miR-148a and miR-326 was associated with decreased progression-free survival, and high miR-326 expression correlated with decreased overall survival [hazard ratio, 1.5 (95% confidence interval, 1.1–2.0)] [32].

In patients with pancreatic cancer, high levels of miR-196a in serum have been found to be significantly associated with unresectable disease, TNM staging and median survival (high compared with low expression of miR-196a: median survival of 6.1 months compared with 12 months) [80]. Similarly, another research group reported a panel of seven up-regulated miRNAs (miR-20a, miR-21, miR-24, miR-25, miR-99a, miR-185 and miR-191), which improved the detection of early stage pancreatic cancer, compared with CA 19-9 or CEA (detection rate of 46.2/62.5% and 30.8/62.5% respectively in stage I/stage II cancer). Furthermore, patients with high levels of miR-21 had poorer survival [82]. In addition, Wang et al. [91] demonstrated that elevated serum miR-21 levels were independent prognostic factors for both lower time-to-progression and overall survival. As miR-21 expression in cancer tissue has been reported to be associated with clinical outcome and the success of gemcitabine treatment in patients with pancreatic cancer, the authors suggested that miR-21 serum levels might serve as potential predictors of chemosensitivity [91]. Finally, serum miR-1290 levels have been shown to distinguish between patients with early-stage pancreatic cancer compared with controls better than CA19-9 levels, and miR-1290 has been identified to be independent predictor of pancreatic cancer prognosis [83].

For HCC, comparable results have been published with correlations between the expression of circulating miRNAs and factors also affecting outcome. Köberle et al. [59], for example, compared selected serum miRNA levels in patients with HCC and those with liver cirrhosis. Although serum miR-1 and miR-122 levels did not differ significantly between patients with cancer compared with liver cirrhosis without HCC, the authors found that patients with HCC with higher miR-1 and miR-122 serum levels survived longer than individuals with low expression. Furthermore, the authors demonstrated miR-1 serum levels to be independently associated with overall survival. Serum miR-122 levelscorrelated negatively with MELD (model for end-stage liver disease) score, supporting the hypothesis that patients with high miR-122 serum levels do have a better liver function and survive longer [59]. In addition, Luo et al. [85] found miR-122a expression to be significantly lower in males, and there was a trend towards lower expression in patients who smoked cigarettes, consumed alcohol, had a family history of HCC or were positive for HBV infection. Li et al. [89] showed that high levels of miR-221 correlated significantly with the existence of cirrhosis, tumour size and tumour stage in patients with HCC. In addition, these authors reported a trend towards increased miR-221 expression with progression of TNM stage (stage I<stage II<stages III–IV).

These initial results suggest that circulating miRNAs in patients with cancer might be useful new biomarkers which might aid the assessment of prognosis, response to treatment and outcome in gastrointestinal malignant tumours in the upper and lower gastrointestinal tract, and also in hepato-pancreato-biliary tumours. Table 2 summarizes the data described above.

Table 2
Circulating miRNAs as prognostic markers

An overview of the deregulation of specific miRNAs in prognostic relevant subgroups of patients with gastrointestinal malignant tumours from the upper to the lower gastrointestinal tract, including hepato-pancreato-biliary tumours. CTX, chemotherapy.

graphic
 
graphic
 

CONCLUSIONS

Since the first reported discovery of miRNAs several years ago, it has become very clear that these components of the epigenetic machinery play a key role in cancer initiation, development and progression. In 2008, came the first report of miRNAs detection in peripheral blood samples. Since then, a number of research groups have examined these so-called ‘circulating miRNAs’, and found that there are different types of origin for these molecules in the bloodstream, and that these circulating miRNAs are quite stable in the circulation. These characteristics suggest that circulating miRNAs might be useful blood-based biomarkers in patients with cancer. In this context, two aspects are of particular interest for the clinician: on the one hand, ideal biomarkers should present significant and reproducible differences in expression between patients with compared with without cancer in order to allow their use for screening purposes, diagnostics and surveillance of cancer. On the other hand, molecular biomarkers that have an impact on prognosis should present significant and reproducible differences in expression between different patient subgroups, thereby contributing to the prognostic assessment of patients with cancer. The early data available for gastrointestinal malignant tumours demonstrate a promising picture. A large number of miRNAs have been reported to show different expression patterns between patients with cancer compared with controls, and between subgroups of patients with better compared with worse prognosis. A number of these miRNA candidates have been detected in more than just one study, and their expression also correlates with the presence of tumours or cancer prognosis in different tumour types, highlighting their multifaceted potential clinical utility in cancer. Figure 2 presents in this context an overview of the most promising circulating miRNA candidates for diagnostic or prognostic evaluation.

Most promising circulating miRNA candidates for diagnostic or prognostic evaluation

Figure 2
Most promising circulating miRNA candidates for diagnostic or prognostic evaluation

A selection of circulating miRNAs are shown that were found in more than one study to be deregulated between cancer patients and controls (diagnostic) or between different prognostic relevant subgroups of cancer patients (prognostic). In total, 46 miRNAs were described to be deregulated in more than one study. However, we selected miRNAs that were named more than once. Red cells, up-regulation of miRNA; green cells, down-regulation of miRNA. miR-19 family includes miR-19a and miR-19b; miR-29 family includes miR-29a and miR-29c; miR-106 family includes miR-106a and miR-106b; and let-7 family includes let-7a and let-7c. CA, cancer.

Figure 2
Most promising circulating miRNA candidates for diagnostic or prognostic evaluation

A selection of circulating miRNAs are shown that were found in more than one study to be deregulated between cancer patients and controls (diagnostic) or between different prognostic relevant subgroups of cancer patients (prognostic). In total, 46 miRNAs were described to be deregulated in more than one study. However, we selected miRNAs that were named more than once. Red cells, up-regulation of miRNA; green cells, down-regulation of miRNA. miR-19 family includes miR-19a and miR-19b; miR-29 family includes miR-29a and miR-29c; miR-106 family includes miR-106a and miR-106b; and let-7 family includes let-7a and let-7c. CA, cancer.

However, it is still too early to determine whether circulating miRNAs will influence daily clinical practice and decision making in the future. Data so far have been derived from early studies that have several limitations. Most of these studies have included only small groups of patients, and there have been no prospective evaluations of circulating miRNAs as potential biomarkers reported so far. Furthermore, there are a wide range of experimental techniques currently in use for the assessment of circulating miRNAs, and there is no standardization of approach to this area. As all of these approaches (including sample acquisition, preparation, isolation of miRNAs or EVs, miRNA profiling and interpretation of data among others) are varied, a definitive comparison of the results of different studies is not possible. Moreover, the question about the origin of circulating miRNAs has yet to be fully answered, although this will have an impact on our understanding of biological aspects of tumour initiation and progression, and the meaning of circulating miRNAs in the context of cell–cell interactions. Finally, as the present review has focussed on cancers of the gastrointestinal tract, caution should be applied when extrapolating to other cancer types. Each tumour type (or even subtype) will need to be evaluated separately if circulating miRNAs are to be defined as a new source of molecular biomarkers in the respective cancers. These considerations clearly show that further research efforts are required to definitively establish circulating miRNAs as clinically useful biomarkers that will have an impact on the assessment and treatment of patients in the clinical setting.

However, despite the early stage of research on circulating miRNAs to date, the initial data in this field appear positive, and support the contention that circulating miRNAs have great potential to improve diagnostic and prognostic evaluation in patients with cancer in the near future.

Abbreviations

     
  • Ago

    Argonaute

  •  
  • AUC

    area under the curve

  •  
  • CA 19-9

    carbohydrate antigen 19-9

  •  
  • ESCC

    oesophageal squamous cell carcinoma

  •  
  • EV

    extracellular vesicle

  •  
  • 5-FU

    5-fluorouracil

  •  
  • HBV

    hepatitis B virus

  •  
  • HCC

    hepatocellular carcinoma

  •  
  • HDL

    high-density lipoprotein

  •  
  • LDL

    low-density lipoprotein

  •  
  • qRT–PCR

    quantitative reverse transcription–PCR

  •  
  • TNM

    tumour node metastasis

References

References
1
Li
 
H.
Yang
 
B. B.
 
Friend or foe: the role of microRNA in chemotherapy resistance
Acta Pharmacol. Sin.
2013
, vol. 
34
 (pg. 
870
-
879
)
[PubMed]
2
Kozomara
 
A.
Griffiths-Jones
 
S.
 
miRBase: annotating high confidence microRNAs using deep sequencing data
Nucleic Acids Res.
2014
, vol. 
42
 (pg. 
68
-
73
)
3
Calin
 
G. A.
Sevignani
 
C.
Dumitru
 
C. D.
Hyslop
 
T.
Noch
 
E.
Yendamuri
 
S.
Shimizu
 
M.
Rattan
 
S.
Bullrich
 
F.
Negrini
 
M.
Croce
 
C. M.
 
Human microRNA genes are frequently located at fragile sites and genomic regions involved in cancers
Proc. Natl. Acad. Sci. U.S.A.
2004
, vol. 
101
 (pg. 
2999
-
3004
)
[PubMed]
4
Esquela-Kerscher
 
A.
Slack
 
F. J.
 
Oncomirs: microRNAs with a role in cancer
Nat. Rev. Cancer
2006
, vol. 
6
 (pg. 
259
-
69
)
[PubMed]
5
Fang
 
Y.
Fang
 
D.
Hu
 
J.
 
MicroRNA and its roles in esophageal cancer
Med. Sci. Monit.
2012
, vol. 
18
 (pg. 
22
-
30
)
6
Weber
 
J. A.
Baxter
 
D. H.
Zhang
 
S.
 
The microRNA spectrum in 12 body fluids
Clin. Chem.
2010
, vol. 
56
 (pg. 
1733
-
1741
)
[PubMed]
7
Xiao
 
Y. F.
Yong
 
X.
Fan
 
Y. H.
 
M. H.
Yang
 
S. M.
Hu
 
C. J.
 
microRNA detection in feces, sputum, pleural effusion and urine: Novel tools for cancer screening
Oncol. Rep.
2013
, vol. 
30
 (pg. 
535
-
544
)
[PubMed]
8
Allegra
 
A.
Alonci
 
A.
Campo
 
S.
Penna
 
G.
Petrungaro
 
A.
Gerace
 
D.
Musolino
 
C.
 
Circulating microRNAs: new biomarkers in diagnosis, prognosis and treatment of cancer
Int. J. Oncol.
2012
, vol. 
41
 (pg. 
1897
-
1912
)
[PubMed]
9
Iorio
 
M. V.
Croce
 
C. M.
 
MicroRNA dysregulation in cancer: diagnostics, monitoring and therapeutics. A comprehensive review
EMBO Mol. Med.
2012
, vol. 
4
 (pg. 
143
-
159
)
[PubMed]
10
Chim
 
S. S.
Shing
 
T. K.
Hung
 
E. C.
Leung
 
T. Y.
Lau
 
T. K.
Chiu
 
R. W.
Lo
 
Y. M.
 
Detection and characterization of placental microRNAs in maternal plasma
Clin. Chem.
2008
, vol. 
54
 (pg. 
482
-
490
)
[PubMed]
11
Russo
 
F.
Di Bella
 
S.
Nigita
 
G.
Macca
 
V.
Laganà
 
A.
Giugno
 
R.
Pulvirenti
 
A.
Ferro
 
A.
 
miRandola: extracellular circulating microRNAs database
PLoS ONE
2012
, vol. 
7
 pg. 
e47786
 
[PubMed]
12
Mo
 
M. H.
Chen
 
L.
Fu
 
Y.
Wang
 
W.
Fu
 
S. W.
 
Cell-free circulating miRNA biomarkers in cancer
J. Cancer
2012
, vol. 
3
 (pg. 
432
-
448
)
[PubMed]
13
Duttagupta
 
R.
Jiang
 
R.
Gollub
 
J.
Getts
 
R. C.
Jones
 
K. W.
 
Impact of cellular miRNAs on circulating miRNA biomarker signatures
PLoS ONE
2011
, vol. 
6
 pg. 
e20769
 
[PubMed]
14
Li
 
L.
Zhu
 
D.
Huang
 
L.
Zhang
 
J.
Bian
 
Z.
Chen
 
X.
Liu
 
Y.
Zhang
 
C. Y.
Zen
 
K.
 
Argonaute 2 complexes selectively protect the circulating microRNAs in cell-secreted microvesicles
PLoS ONE
2012
, vol. 
7
 pg. 
e46957
 
[PubMed]
15
Turchinovich
 
A.
Weiz
 
L.
Langheinz
 
A.
Burwinkel
 
B.
 
Characterization of extracellular circulating microRNA
Nucleic Acids Res.
2011
, vol. 
39
 (pg. 
7223
-
33
)
[PubMed]
16
Turchinovich
 
A.
Burwinkel
 
B.
 
Distinct AGO1 and AGO2 associated miRNA profiles in human cells and blood plasma
RNA Biol.
2012
, vol. 
9
 (pg. 
1066
-
1075
)
[PubMed]
17
Kim
 
T.
Reitmair
 
A.
 
Non-coding RNAs: functional aspects and diagnostic utility in oncology
Int. J. Mol. Sci.
2013
, vol. 
14
 (pg. 
4934
-
4968
)
[PubMed]
18
Valadi
 
H.
Ekström
 
K.
Bossios
 
A.
Sjöstrand
 
M.
Lee
 
J. J.
Lötvall
 
J. O.
 
Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells
Nat. Cell Biol.
2007
, vol. 
9
 (pg. 
654
-
659
)
[PubMed]
19
Montecalvo
 
A.
Larregina
 
A. T.
Shufesky
 
W. J.
Stolz
 
D. B.
Sullivan
 
M. L.
Karlsson
 
J. M.
Baty
 
C. J.
Gibson
 
G. A.
Erdos
 
G.
Wang
 
Z.
, et al 
Mechanism of transfer of functional microRNAs between mouse dendritic cells via exosomes
Blood
2012
, vol. 
119
 (pg. 
756
-
766
)
[PubMed]
20
Quackenbush
 
J. F.
Cassidy
 
P. B.
Pfeffer
 
L. M.
Boucher
 
K. M.
Hawkes
 
J. E.
Pfeffer
 
S. R.
Kopelovich
 
L.
Leachman
 
S. A.
 
Isolation of circulating microRNAs from microvesicles found in human plasma
Methods Mol. Biol.
2011
, vol. 
1102
 (pg. 
641
-
653
)
21
Cortez
 
M. A.
Bueso-Ramos
 
C.
Ferdin
 
J.
Lopez-Berestein
 
G.
Sood
 
A. K.
Calin
 
G. A.
 
MicroRNAs in body fluids: the mix of hormones and biomarkers
Nat. Rev. Clin. Oncol.
2011
, vol. 
8
 (pg. 
467
-
477
)
[PubMed]
22
Pigati
 
L.
Yaddanapudi
 
S. C.
Iyengar
 
R.
Kim
 
D. J.
Hearn
 
S. A.
Danforth
 
D.
Hastings
 
M. L.
Duelli
 
D. M.
 
Selective release of microRNA species from normal and malignant mammary epithelial cells
PLoS ONE
2010
, vol. 
5
 pg. 
e13515
 
[PubMed]
23
Kosaka
 
N.
Iguchi
 
H.
Yoshioka
 
Y.
Takeshita
 
F.
Matsuki
 
Y.
Ochiya
 
T.
 
Secretory mechanisms and intercellular transfer of microRNAs in living cells
J. Biol. Chem.
2010
, vol. 
285
 (pg. 
17442
-
1752
)
[PubMed]
24
Rayner
 
K. J.
Hennessy
 
E. J.
 
Extracellular communication via microRNA: lipid particles have a new message
J. Lipid Res.
2013
, vol. 
54
 (pg. 
1174
-
1181
)
[PubMed]
25
Ma
 
R.
Jiang
 
T.
Kang
 
X.
 
Circulating microRNAs in cancer: origin, function and application
J. Exp. Clin. Cancer Res.
2012
, vol. 
30
 (pg. 
31
-
38
)
26
Li
 
C.
Pei
 
F.
Zhu
 
X.
Duan
 
D. D.
Zeng
 
C.
 
Circulating microRNAs as novel and sensitive biomarkers of acute myocardial Infarction
Clin. Biochem.
2012
, vol. 
45
 (pg. 
727
-
732
)
[PubMed]
27
Toiyama
 
Y.
Hur
 
K.
Tanaka
 
K.
Inoue
 
Y.
Kusunoki
 
M.
Boland
 
C. R.
Goel
 
A.
 
Serum miR-200c is a novel prognostic and metastasis-predictive biomarker in patients with colorectal cancer
Ann. Surg.
2014
, vol. 
259
 (pg. 
735
-
743
)
[PubMed]
28
Li
 
X.
Zhang
 
Y.
Zhang
 
H.
Liu
 
X.
Gong
 
T.
Li
 
M.
Sun
 
L.
Ji
 
G.
Shi
 
Y.
Han
 
Z.
, et al 
miRNA-223 promotes gastric cancer invasion and metastasis by targeting tumor suppressor EPB41L3
Mol. Cancer Res.
2011
, vol. 
9
 (pg. 
824
-
833
)
[PubMed]
29
Eichelser
 
C.
Flesch-Janys
 
D.
Chang-Claude
 
J.
Pantel
 
K.
Schwarzenbach
 
H.
 
Deregulated serum concentrations of circulating cell-free microRNAs miR-17, miR-34a, miR-155, and miR-373 in human breast cancer development and progression
Clin. Chem.
2013
, vol. 
59
 (pg. 
1489
-
1496
)
[PubMed]
30
Hirajima
 
S.
Komatsu
 
S.
Ichikawa
 
D.
Takeshita
 
H.
Konishi
 
H.
Shiozaki
 
A.
Morimura
 
R.
Tsujiura
 
M.
Nagata
 
H.
Kawaguchi
 
T.
, et al 
Clinical impact of circulating miR-18a in plasma of patients with oesophageal squamous cell carcinoma
Br. J. Cancer
2013
, vol. 
108
 (pg. 
1822
-
1829
)
[PubMed]
31
Zhang
 
J.
Zhang
 
K.
Bi
 
M.
Jiao
 
X.
Zhang
 
D.
Dong
 
Q.
 
Circulating microRNA expressions in colorectal cancer as predictors of response to chemotherapy
Anticancer Drugs
2014
, vol. 
25
 (pg. 
346
-
352
)
[PubMed]
32
Kjersem
 
J. B.
Ikdahl
 
T.
Lingjaerde
 
O. C.
Guren
 
T.
Tveit
 
K. M.
Kure
 
E. H.
 
Plasma microRNAs predicting clinical outcome in metastatic colorectal cancer patients receiving first-line oxaliplatin-based treatment
Mol. Oncol.
2014
, vol. 
8
 (pg. 
59
-
67
)
[PubMed]
33
Tanaka
 
K.
Miyata
 
H.
Yamasaki
 
M.
Sugimura
 
K.
Takahashi
 
T.
Kurokawa
 
Y.
Nakajima
 
K.
Takiguchi
 
S.
Mori
 
M.
Doki
 
Y.
 
Circulating miR-200c levels significantly predict response to chemotherapy and prognosis of patients undergoing neoadjuvant chemotherapy for esophageal cancer
Ann. Surg. Oncol.
2012
, vol. 
20
 (pg. 
607
-
615
)
[PubMed]
34
van Empel
 
V. P.
De Windt
 
L. J.
da Costa Martins
 
P. A.
 
Circulating miRNAs: reflecting or affecting cardiovascular disease?
Curr. Hypertens. Rep.
2012
, vol. 
14
 (pg. 
498
-
509
)
[PubMed]
35
Takahashi
 
K.
Yan
 
I.
Wen
 
H. J.
Patel
 
T.
 
microRNAs in liver disease: from diagnostics to therapeutics
Clin. Biochem.
2013
, vol. 
46
 (pg. 
946
-
952
)
[PubMed]
36
Dragovic
 
R. A.
Gardiner
 
C.
Brooks
 
A. S.
Tannetta
 
D. S.
Ferguson
 
D. J.
Hole
 
P.
Carr
 
B.
Redman
 
C. W.
Harris
 
A. L.
Dobson
 
P. J.
, et al 
Sizing and phenotyping of cellular vesicles using nanoparticle tracking analysis
Nanomedicine
2011
, vol. 
7
 (pg. 
780
-
788
)
[PubMed]
37
Rani
 
S.
O’Brien
 
K.
Kelleher
 
F. C.
Corcoran
 
C.
Germano
 
S.
Radomski
 
M. W.
Crown
 
J.
O’Driscoll
 
L.
 
Isolation of exosomes for subsequent mRNA, microRNA, and protein profiling
Methods Mol. Biol.
2011
, vol. 
784
 (pg. 
181
-
195
)
[PubMed]
38
Johnstone
 
R. M.
 
Revisiting the road to the discovery of exosomes
Blood Cells Mol. Dis.
2005
, vol. 
34
 (pg. 
214
-
219
)
[PubMed]
39
Camussi
 
G.
Deregibus
 
M. C.
Bruno
 
S.
Cantaluppi
 
V.
Biancone
 
L.
 
Exosomes/microvesicles as a mechanism of cell-to-cell communication
Kidney Int.
2010
, vol. 
78
 (pg. 
838
-
848
)
[PubMed]
40
Valadi
 
H.
Ekström
 
K.
Bossios
 
A.
Sjöstrand
 
M.
Lee
 
J. J.
Lötvall
 
J. O.
 
Exosome-mediated transfer of mRNAs and microRNAs is a novel mechanism of genetic exchange between cells
Nat. Cell Biol.
2007
, vol. 
9
 (pg. 
654
-
959
)
[PubMed]
41
Keller
 
S.
Ridinger
 
J.
Rupp
 
A. K.
Janssen
 
J. W.
Altevogt
 
P.
 
Body fluid derived exosomes as a novel template for clinical diagnostics
J. Transl. Med.
2011
, vol. 
9
 pg. 
86
 
[PubMed]
42
Taylor
 
D. D.
Gercel-Taylor
 
C.
 
MicroRNA signatures of tumor-derived exosomes as diagnostic biomarkers of ovarian cancer
Gynecol. Oncol.
2008
, vol. 
110
 (pg. 
13
-
21
)
[PubMed]
43
Principe
 
S.
Jones
 
E. E.
Kim
 
Y.
Sinha
 
A.
Nyalwidhe
 
J. O.
Brooks
 
J.
Semmes
 
O. J.
Troyer
 
D. A.
Lance
 
R. S.
Kislinger
 
T.
Drake
 
R. R.
 
In-depth proteomic analyses of exosomes isolated from expressed prostatic secretions in urine
Proteomics
2013
, vol. 
13
 (pg. 
1667
-
1671
)
[PubMed]
44
Dai
 
S.
Wei
 
D.
Wu
 
Z.
Zhou
 
X.
Wei
 
X.
Huang
 
H.
Li
 
G.
 
Phase I clinical trial of autologous ascites-derived exosomes combined with GM-CSF for colorectal cancer
Mol. Ther.
2008
, vol. 
16
 (pg. 
782
-
790
)
[PubMed]
45
Michael
 
A.
Bajracharya
 
S. D.
Yuen
 
P. S.
Zhou
 
H.
Star
 
R. A.
Illei
 
G. G.
Alevizos
 
I.
 
Exosomes from humansaliva as a source of microRNAbiomarkers
Oral Dis.
2010
, vol. 
16
 (pg. 
34
-
38
)
[PubMed]
46
Théry
 
C.
Zitvogel
 
L.
Amigorena
 
S.
 
Exosomes: composition, biogenesis and function
Nat. Rev. Immunol.
2002
, vol. 
2
 (pg. 
569
-
579
)
[PubMed]
47
Théry
 
C.
Boussac
 
M.
Véron
 
P.
Ricciardi-Castagnoli
 
P.
Raposo
 
G.
Garin
 
J.
Amigorena
 
S.
 
Proteomic analysis of dendritic cell-derived exosomes: a secreted subcellular compartment distinct from apoptotic vesicles
J. Immunol.
2001
, vol. 
166
 (pg. 
7309
-
7318
)
[PubMed]
48
Mathivanan
 
S.
Ji
 
H.
Simpson
 
R. J.
 
Exosomes: extracellular organelles important in intercellular communication
J. Proteomics.
2010
, vol. 
73
 (pg. 
1907
-
1920
)
[PubMed]
49
Kosaka
 
N.
Iguchi
 
H.
Yoshioka
 
Y.
Takeshita
 
F.
Matsuki
 
Y.
Ochiya
 
T.
 
Secretory mechanisms and intercellular transfer of microRNAs in living cells
J. Biol. Chem.
2010
, vol. 
285
 (pg. 
17442
-
17452
)
[PubMed]
50
Ekström
 
K.
Valadi
 
H.
Sjöstrand
 
M.
Malmhäll
 
C.
Bossios
 
A.
Eldh
 
M.
Lötvall
 
J.
 
Characterization of mRNA and microRNA in human mast cell-derived exosomes and their transfer to other mast cells and blood CD34 progenitor cells
J. Extracell. Vesicles
2012
, vol. 
1
 pg. 
18389
 
[PubMed]
51
Eldh
 
M.
Ekström
 
K.
Valadi
 
H.
Sjöstrand
 
M.
Olsson
 
B.
Jernås
 
M.
Lötvall
 
J.
 
Exosomes communicate protective messages during oxidative stress; possible role of exosomal shuttle RNA
PLoS ONE
2010
, vol. 
5
 pg. 
e15353
 
[PubMed]
52
Montecalvo
 
A.
Larregina
 
A. T.
Shufesky
 
W. J.
Stolz
 
D. B.
Sullivan
 
M. L.
Karlsson
 
J. M.
Baty
 
C. J.
Gibson
 
G. A.
Erdos
 
G.
Wang
 
Z.
, et al 
Mechanism of transfer of functional microRNAs between mouse dendritic cells via exosomes
Blood
2012
, vol. 
119
 (pg. 
756
-
766
)
[PubMed]
53
Rosell
 
R.
Wei
 
J.
Taron
 
M.
 
Circulating microRNA signatures of tumor-derived exosomes for early diagnosis of non-small-cell lung cancer
Clin. Lung Cancer
2009
, vol. 
10
 (pg. 
8
-
9
)
[PubMed]
54
Vickers
 
K. C.
Remaley
 
A. T.
 
Lipid-based carriers of microRNAs and intercellular communication
Curr. Opin. Lipidol.
2012
, vol. 
23
 (pg. 
91
-
97
)
[PubMed]
55
Wagner
 
J.
Riwanto
 
M.
Besler
 
C.
Knau
 
A.
Fichtlscherer
 
S.
Röxe
 
T.
Zeiher
 
A. M.
Landmesser
 
U.
Dimmeler
 
S.
 
Characterization of levels and cellular transfer of circulating lipoprotein-bound microRNAs
Arterioscler. Thromb. Vasc. Biol.
2013
, vol. 
33
 (pg. 
1392
-
1400
)
[PubMed]
56
Pritchard
 
C. C.
Kroh
 
E.
Wood
 
B.
Arroyo
 
J. D.
Dougherty
 
K. J.
Miyaji
 
M. M.
Tait
 
J. F.
Tewari
 
M.
 
Blood cell origin of circulating microRNAs: a cautionary note for cancer biomarker studies
Cancer Prev. Res. (Phila)
2012
, vol. 
5
 (pg. 
492
-
497
)
[PubMed]
57
Kirschner
 
M. B.
Kao
 
S. C.
Edelman
 
J. J.
Armstrong
 
N. J.
Vallely
 
M. P.
van Zandwijk
 
N.
Reid
 
G.
 
Haemolysis during sample preparation alters microRNA content of plasma
PLoS ONE
2011
, vol. 
6
 pg. 
e24145
 
[PubMed]
58
Kirschner
 
M. B.
Edelman
 
J. J.
Kao
 
S. C.
Vallely
 
M. P.
van Zandwijk
 
N.
Reid
 
G.
 
The impact of hemolysis on cell-free microRNA biomarkers
Front. Genet.
2013
, vol. 
4
 pg. 
94
 
[PubMed]
59
Köberle
 
V.
Pleli
 
T.
Schmithals
 
C.
Augusto Alonso
 
E.
Haupenthal
 
J.
Bönig
 
H.
Peveling-Oberhag
 
J.
Biondi
 
R. M.
Zeuzem
 
S.
Kronenberger
 
B.
, et al 
Differential stability of cell-free circulating microRNAs: implications for their utilization as biomarkers
PLoS ONE
2013
, vol. 
8
 pg. 
e75184
 
[PubMed]
60
Théry
 
C.
Amigorena
 
S.
Raposo
 
G.
Clayton
 
A.
 
Isolation and characterization of exosomes from cell culture supernatants and biological fluids
Curr. Protoc. Cell Biol.
2006
, vol. 
3
  
Chapter, Unit 3.22
[PubMed]
61
Clayton
 
A.
Court
 
J.
Navabi
 
H.
Adams
 
M.
Mason
 
M. D.
Hobot
 
J. A.
Newman
 
G. R.
Jasani
 
B.
 
Analysis of antigen presenting cell derived exosomes, based on immuno-magnetic isolation and flow cytometry
J. Immunol. Methods
2001
, vol. 
247
 (pg. 
163
-
174
)
[PubMed]
62
Koga
 
K.
Matsumoto
 
K.
Akiyoshi
 
T.
Kubo
 
M.
Yamanaka
 
N.
Tasaki
 
A.
Nakashima
 
H.
Nakamura
 
M.
Kuroki
 
S.
Tanaka
 
M.
Katano
 
M.
 
Purification, characterization and biological significance of tumor-derived exosomes
Anticancer Res.
2005
, vol. 
25
 (pg. 
3703
-
3707
)
[PubMed]
63
Yamada
 
T.
Inoshima
 
Y.
Matsuda
 
T.
Ishiguro
 
N.
 
Comparison of methods for isolating exosomes from bovine milk
J. Vet. Med. Sci.
2012
, vol. 
74
 (pg. 
1523
-
1525
)
[PubMed]
64
Taylor
 
D. D.
Zacharias
 
W.
Gercel-Taylor
 
C.
 
Exosome isolation for proteomic analyses and RNA profiling
Methods Mol. Biol.
2011
, vol. 
728
 (pg. 
235
-
246
)
[PubMed]
65
Kalra
 
H.
Adda
 
C. G.
Liem
 
M.
Ang
 
C. S.
Mechler
 
A.
Simpson
 
R. J.
Hulett
 
M. D.
Mathivanan
 
S.
 
Comparative proteomics evaluation of plasma exosome isolation techniques and assessment of the stability of exosomes in normal human blood plasma
Proteomics
2013
, vol. 
13
 (pg. 
3354
-
3364
)
[PubMed]
66
Hill
 
A. F.
Pegtel
 
D. M.
Lambertz
 
U.
Leonardi
 
T.
O’Driscoll
 
L.
Pluchino
 
S.
Ter-Ovanesyan
 
D.
Nolte-'t Hoen
 
E. N.
 
ISEV position paper: extracellular vesicle RNA analysis and bioinformatics
J. Extracell. Vesicles
2013
, vol. 
2
 pg. 
22859
 
[PubMed]
67
Liu
 
R.
Liao
 
J.
Yang
 
M.
Shi
 
Y.
Peng
 
Y.
Wang
 
Y.
Pan
 
E.
Guo
 
W.
Pu
 
Y.
Yin
 
L.
 
Circulating miR-155 expression in plasma: a potential biomarker for early diagnosis of esophageal cancer in humans
J. Toxicol. Environ Health A
2012
, vol. 
75
 (pg. 
1154
-
1162
)
[PubMed]
68
Komatsu
 
S.
Ichikawa
 
D.
Takeshita
 
H.
Tsujiura
 
M.
Morimura
 
R.
Nagata
 
H.
Kosuga
 
T.
Iitaka
 
D.
Konishi
 
H.
Shiozaki
 
A.
, et al 
Circulating microRNAs in plasma of patients with oesophageal squamous cell carcinoma
Br. J. Cancer
2011
, vol. 
105
 (pg. 
104
-
111
)
[PubMed]
69
Tanaka
 
K.
Miyata
 
H.
Yamasaki
 
M.
Sugimura
 
K.
Takahashi
 
T.
Kurokawa
 
Y.
Nakajima
 
K.
Takiguchi
 
S.
Mori
 
M.
Doki
 
Y.
 
Circulating miR-200c levels significantly predict response to chemotherapy and prognosis of patients undergoing neoadjuvant chemotherapy for esophageal cancer
Ann. Surg. Oncol.
2013
, vol. 
3
 (pg. 
607
-
615
)
70
Sheinerman
 
K. S.
Tsivinsky
 
V. G.
Umansky
 
S. R.
 
Analysis of organ-enriched microRNAs in plasma as an approach to development of universal screening test: feasibility study
J. Transl. Med.
2013
, vol. 
11
 pg. 
304
 
[PubMed]
71
Li
 
B. S.
Zhao
 
Y. L.
Guo
 
G.
Li
 
W.
Zhu
 
E. D.
Luo
 
X.
Mao
 
X. H.
Zou
 
Q. M.
Yu
 
P. W.
Zuo
 
Q. F.
, et al 
Plasma microRNAs, miR-223, miR-21 and miR-218, as novel potential biomarkers for gastric cancer detection
PLoS ONE
2012
, vol. 
7
 pg. 
e41629
 
[PubMed]
72
Wang
 
B.
Zhang
 
Q.
 
The expression and clinical significance of circulating microRNA-21 in serum of five solid tumors
J. Cancer Res. Clin. Oncol.
2012
, vol. 
138
 (pg. 
1659
-
1666
)
[PubMed]
73
Tsujiura
 
M.
Ichikawa
 
D.
Komatsu
 
S.
Shiozaki
 
A.
Takeshita
 
H.
Kosuga
 
T.
Konishi
 
H.
Morimura
 
R.
Deguchi
 
K.
Fujiwara
 
H.
, et al 
Circulating microRNAs in plasma of patients with gastric cancers
Br. J. Cancer
2010
, vol. 
102
 (pg. 
1174
-
1179
)
[PubMed]
74
Zhu
 
X.
Lv
 
M.
Wang
 
H.
Guan
 
W.
 
Identification of circulating microRNAs as novel potential biomarkers for gastric cancer detection: a systematic review and meta-analysis
Dig. Dis. Sci.
2014
, vol. 
59
 (pg. 
911
-
919
)
[PubMed]
75
Giráldez
 
M. D.
Lozano
 
J. J.
Ramírez
 
G.
Hijona
 
E.
Bujanda
 
L.
Castells
 
A.
Gironella
 
M.
 
Circulating microRNAs as biomarkers of colorectal cancer, results from a genome-wide profiling and validation study
Clin. Gastroenterol. Hepatol.
2013
, vol. 
11
 (pg. 
681
-
688
)
[PubMed]
76
Kanaan
 
Z.
Roberts
 
H.
Eichenberger
 
M. R.
Billeter
 
A.
Ocheretner
 
G.
Pan
 
J.
Rai
 
S. N.
Jorden
 
J.
Williford
 
A.
Galandiuk
 
S.
 
A plasma microRNA panel for detection of colorectal adenomas: a step toward more precise screening for colorectal cancer
Ann. Surg.
2013
, vol. 
258
 (pg. 
400
-
408
)
[PubMed]
77
Morimura
 
R.
Komatsu
 
S.
Ichikawa
 
D.
Takeshita
 
H.
Tsujiura
 
M.
Nagata
 
H.
Konishi
 
H.
Shiozaki
 
A.
Ikoma
 
H.
Okamoto
 
K.
, et al 
Novel diagnostic value of circulating miR-18a in plasma of patients with pancreatic cancer
Br. J. Cancer
2011
, vol. 
105
 (pg. 
1733
-
1740
)
[PubMed]
78
Wang
 
J.
Chen
 
J.
Chang
 
P.
LeBlanc
 
A.
Li
 
D.
Abbruzzesse
 
J. L.
Frazier
 
M. L.
Killary
 
A. M.
Sen
 
S.
 
MicroRNAs in plasma of pancreatic ductal adenocarcinoma patients as novel blood-based biomarkers of disease
Cancer Prev. Res. (Phila)
2009
, vol. 
2
 (pg. 
807
-
813
)
[PubMed]
79
Ho
 
A. S.
Huang
 
X.
Cao
 
H.
Christman-Skieller
 
C.
Bennewith
 
K.
Le
 
Q. T.
Koong
 
A. C.
 
Circulating miR-210 as a novel hypoxia marker in pancreatic cancer
Transl. Oncol.
2010
, vol. 
3
 (pg. 
109
-
113
)
[PubMed]
80
Kong
 
X.
Du
 
Y.
Wang
 
G.
Gao
 
J.
Gong
 
Y.
Li
 
L.
Zhang
 
Z.
Zhu
 
J.
Jing
 
Q.
Qin
 
Y.
Li
 
Z.
 
Detection of differentially expressed microRNAs in serum of pancreatic ductal adenocarcinoma patients: miR-196a could be a potential marker for poor prognosis
Dig. Dis. Sci.
2011
, vol. 
56
 (pg. 
602
-
609
)
[PubMed]
81
Liu
 
J.
Gao
 
J.
Du
 
Y.
Li
 
Z.
Ren
 
Y.
Gu
 
J.
Wang
 
X.
Gong
 
Y.
Wang
 
W.
Kong
 
X.
 
Combination of plasma microRNAs with serum CA19–9 for early detection of pancreatic cancer
Int. J. Cancer
2012
, vol. 
131
 (pg. 
683
-
691
)
[PubMed]
82
Liu
 
R.
Chen
 
X.
Du
 
Y.
Yao
 
W.
Shen
 
L.
Wang
 
C.
Hu
 
Z.
Zhuang
 
R.
Ning
 
G.
Zhang
 
C.
, et al 
Serum microRNA expression profile as a biomarker in the diagnosis and prognosis of pancreatic cancer
Clin. Chem.
2012
, vol. 
58
 (pg. 
610
-
618
)
[PubMed]
83
Li
 
A.
Yu
 
J.
Kim
 
H.
Wolfgang
 
C. L.
Canto
 
M. I.
Hruban
 
R. H.
Goggins
 
M.
 
MicroRNA array analysis finds elevated serum miR-1290 accurately distinguishes patients with low-stage pancreatic cancer from healthy and disease controls
Clin. Cancer Res.
2013
, vol. 
19
 (pg. 
3600
-
3610
)
[PubMed]
84
Shen
 
J.
Wang
 
A.
Wang
 
Q.
Gurvich
 
I.
Siegel
 
A. B.
Remotti
 
H.
Santella
 
R. M.
 
Exploration of genome-wide circulating microRNA in hepatocellular carcinoma:miR-483–5p as a potential biomarker
Cancer Epidemiol. Biomarkers Prev.
2013
, vol. 
22
 (pg. 
2364
-
2373
)
[PubMed]
85
Luo
 
J.
Chen
 
M.
Huang
 
H.
Yuan
 
T.
Zhang
 
M.
Zhang
 
K.
Deng
 
S.
 
Circulating microRNA-122a as a diagnostic marker for hepatocellular carcinoma
Onco Targets Ther.
2013
, vol. 
6
 (pg. 
577
-
583
)
[PubMed]
86
Lanford
 
R. E.
Hildebrandt-Eriksen
 
E. S.
Petri
 
A.
Persson
 
R.
Lindow
 
M.
Munk
 
M. E.
Kauppinen
 
S.
Ørum
 
H.
 
Therapeutic silencing of microRNA-122 in primates with chronic hepatitis C virus infection
Science
2010
, vol. 
327
 (pg. 
198
-
201
)
[PubMed]
87
Qiu
 
L.
Fan
 
H.
Jin
 
W.
Zhao
 
B.
Wang
 
Y.
Ju
 
Y.
Chen
 
L.
Chen
 
Y.
Duan
 
Z.
Meng
 
S.
 
miR-122-induced down-regulation of HO-1 negatively affects miR-122-mediated suppression of HBV
Biochem. Biophys. Res. Commun.
2010
, vol. 
398
 (pg. 
771
-
777
)
[PubMed]
88
Li
 
L. M.
Hu
 
Z. B.
Zhou
 
Z. X.
Chen
 
X.
Liu
 
F. Y.
Zhang
 
J. F.
Shen
 
H. B.
Zhang
 
C. Y.
Zen
 
K.
 
Serum microRNA profiles serve as novel biomarkers for HBV infection and diagnosis of HBV-positive hepatocarcinoma
Cancer Res.
2010
, vol. 
70
 (pg. 
9798
-
9807
)
[PubMed]
89
Li
 
J.
Wang
 
Y.
Yu
 
W.
Chen
 
J.
Luo
 
J.
 
Expression of serum miR-221 in human hepatocellular carcinoma and its prognostic significance
Biochem. Biophys. Res. Commun.
2011
, vol. 
406
 (pg. 
70
-
73
)
[PubMed]
90
Zheng
 
Y.
Cui
 
L.
Sun
 
W.
Zhou
 
H.
Yuan
 
X.
Huo
 
M.
Chen
 
J.
Lou
 
Y.
Guo
 
J.
 
MicroRNA-21 is a new marker of circulating tumor cells in gastric cancer patients
Cancer Biomark.
2011
, vol. 
10
 (pg. 
71
-
77
)
[PubMed]
91
Wang
 
P.
Zhuang
 
L.
Zhang
 
J.
Fan
 
J.
Luo
 
J.
Chen
 
H.
Wang
 
K.
Liu
 
L.
Chen
 
Z.
Meng
 
Z.
 
The serum miR-21 level serves as a predictor for the chemosensitivity of advanced pancreatic cancer, and miR-21 expression confers chemoresistance by targeting FasL
Mol. Oncol.
2013
, vol. 
7
 (pg. 
334
-
345
)
[PubMed]
92
Moldovan
 
L.
Batte
 
K. E.
Trgovcich
 
J.
Wisler
 
J.
Marsh
 
C. B.
Piper
 
M.
 
Methodological challenges in utilizing miRNAs as circulating biomarkers
J. Cell. Mol. Med.
2014
, vol. 
18
 (pg. 
371
-
390
)
[PubMed]
93
Momen-Heravi
 
F.
Balaj
 
L.
Alian
 
S.
Mantel
 
P. Y.
Halleck
 
A. E.
Trachtenberg
 
A. J.
Soria
 
C. E.
Oquin
 
S.
Bonebreak
 
C. M.
Saracoglu
 
E.
, et al 
Current methods for the isolation of extracellularvesicles
Biol Chem.
2013
, vol. 
394
 (pg. 
1253
-
1262
)
[PubMed]
94
Dieckmann
 
K. P.
Spiekermann
 
M.
Balks
 
T.
Flor
 
I.
Löning
 
T.
Bullerdiek
 
J.
Belge
 
G.
 
MicroRNAs miR-371–3 in serum as diagnostic tools in the management of testicular germ cell tumours
Br. J. Cancer
2012
, vol. 
107
 (pg. 
1754
-
1760
)
[PubMed]
95
Chan
 
M.
Liaw
 
C. S.
Ji
 
S. M.
Tan
 
H. H.
Wong
 
C. Y.
Thike
 
A. A.
Tan
 
P. H.
Ho
 
G. H.
Lee
 
A. S.
 
Identification of circulating microRNA signatures for breast cancer detection
Clin Cancer Res.
2013
, vol. 
19
 (pg. 
4477
-
4487
)
[PubMed]
96
Rabinowits
 
G.
Gerçel-Taylor
 
C.
Day
 
J. M.
Taylor
 
D. D.
Kloecker
 
G. H.
 
Exosomal microRNA: a diagnostic marker for lung cancer
Clin. Lung Cancer
2009
, vol. 
10
 (pg. 
42
-
46
)
[PubMed]
97
Taylor
 
D. D.
Gercel-Taylor
 
C.
 
MicroRNA signatures of tumor-derived exosomes as diagnostic biomarkers of ovarian cancer
Gynecol. Oncol.
2008
, vol. 
110
 (pg. 
13
-
21
)
[PubMed]
98
Skog
 
J.
Würdinger
 
T.
van Rijn
 
S.
Meijer
 
D. H.
Gainche
 
L.
Sena-Esteves
 
M.
Curry
 
W. T.
Carter
 
B. S.
Krichevsky
 
A. M.
Breakefield
 
X. O.
 
Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers
Nat. Cell Biol.
2008
, vol. 
10
 (pg. 
1470
-
1476
)
[PubMed]
99
Hummel
 
R.
Hussey
 
D. J.
Haier
 
J.
 
MicroRNAs: predictors and modifiers of chemo- and radiotherapy in different tumour types
Eur. J. Cancer
2010
, vol. 
46
 (pg. 
298
-
311
)
[PubMed]
100
Pigati
 
L.
Yaddanapudi
 
S. C.
Iyengar
 
R.
Kim
 
D. J.
Hearn
 
S. A.
Danforth
 
D.
Hastings
 
M. L.
Duelli
 
D. M.
 
Selective release of microRNA species from normal and malignant mammary epithelial cells
PLoS ONE
2010
, vol. 
5
 pg. 
e13515
 
[PubMed]
101
Jaiswal
 
R.
Luk
 
F.
Gong
 
J.
Mathys
 
J. M.
Grau
 
G. E.
Bebawy
 
M.
 
Microparticle conferred microRNA profiles: implications in the transfer and dominance of cancer traits
Mol. Cancer
2012
, vol. 
11
 pg. 
37
 
[PubMed]