Clinical applications of liquid biopsies in gastrointestinal oncology

Introduction

In recent years, interest in blood-based “liquid biopsies” to diagnose, monitor, and characterize solid tumors has surged. With the arrival of several commercially available blood-based assays, the use of these assays in clinical practice is increasingly routine. Although many of these assays are analytically sensitive, specific, and accurate, prospective evidence linking results with clinical utility remains limited. In the absence of this prospective evidence, the increased use of blood-based biopsies reflects a desire to minimize procedural risk to the patient, while applying therapies tailored to a patient’s specific tumor characteristics.

Liquid biopsies can be used to identify a range of circulating tumor products, including circulating tumor cells (CTCs), circulating tumor DNA (ctDNA), circulating messenger RNA (mRNA), circulating microRNA (miRNA), circulating exosomes, and tumor educated platelets (TEPs). Of these circulating tumor products, CTC and ctDNA assays are orderable through commercial vendors, and are most broadly incorporated into clinical practice. The detection of circulating tumor products has been well described for decades (1,2), but the assays used to detect these products have only recently been technically validated for routine clinical practice.

For patients with gastrointestinal (GI) malignancies, the use of blood-based biopsies to screen, diagnose, and genotype tumors may be particularly beneficial. First, tumor tissue samples—which are often limited to cytology or fine needle aspiration (FNA)—may be of insufficient quantity or tumor content. As a result, genomic profiling for clinical trial participation and personalized treatment strategies may not be possible. When tumor tissue is unavailable, clinicians must choose between a repeat tumor biopsy or treatment in the absence of genomic data. In some cases, metastatic tumor lesions are only available through diagnostic laparoscopy or other invasive procedures. Second, blood-based assays offer a dynamic “snapshot” of the entire tumor burden at a specific point in time. In addition to showing dominant mutations, this anatomically unbiased profiling often reveals clinically relevant subclones. These subclones may not be detected in primary tumor tissue, but may increase under the selective pressure of targeted therapy. Third, blood-based biopsies capture both inter and intratumoral heterogeneity. In some cases this heterogeneity may be substantial (3-5). Finally, given the relative ease and speed with which blood can be obtained, blood-based assays can be utilized to identify genomic changes associated with sensitivity and resistance to novel therapeutics. This can accelerate efforts to discover relevant resistance-conferring alterations, and may support efforts to further personalize treatment.

This review will examine current liquid biopsy technologies for patients with GI malignancies, and will assess the ability of circulating tumor products to detect new cancers, diagnose recurrent disease, identify genomic biomarkers of sensitivity and resistance, and predict prognosis.

Circulating tumor products: overview of current technologies

CTCs

CTCs originate from the primary tumor and metastatic sites. Most CTC assays rely on selection of tumor cells based on their biological (protein expression) and physical properties (size, density, deformability, electric charge). Similar to ctDNA, CTCs have been used to help determine tumor mutation status (6) and prognosis (7). There currently exists a FDA-approved CTC assay, CellSearch (Veridex, New Jersey, USA), for the enumeration of CTCs. Although commercially available, the use of CTCs in GI cancers is limited due to low rates of detection, and its prognostic, rather than predictive value (8). While CTCs can be detected in over 60% of patients with metastatic breast cancer and prostate cancer, CTCs can only be detected in 30–40% of patients with metastatic CRC (mCRC) (8-11). When CTCs are detected in mCRC, they are typically at a lower absolute number (median 1–2 CTC/7.5 mL) than when they are detected in metastatic breast and prostate cancer (median 6–7 CTC/7.5 mL) (12-14). The reason for this discrepancy is multifactorial. First, CRC CTCs often travel in adherent clusters due to increased surface adhesion molecules (15). Second, the circulation anatomy may affect cell collection (16). Specifically, hepatic drainage of portal blood may act as a filter, resulting in fewer peripheral blood CTCs. Tumor location also plays a role. For example, lower rectal tumors have increased CTCs in the central venous blood, compared with tumors in the upper rectum and colon. Lastly, as CTCs undergo epithelial-mesenchymal transition (EMT), cells lose their epithelial markers and become more difficult to detect. CellSearch relies on expression of EpCAM, a marker of epithelial phenotype (17). Even when the cells do express EpCAM, it has been shown that the specific EpCAM antibody used can greatly change the detection rate (18). There are currently efforts underway to detect CTCs lacking EpCAM expression (17,19-21). One example is capturing cells which express the tyrosine kinase receptor c-MET (21).

ctDNA

ctDNA is comprised of fragments of cell-free, tumor-derived DNA from necrotic or apoptotic tumor cells (22). ctDNA can originate from CTCs, metastases, or the primary tumor. There are two challenges facing detection of ctDNA. First, a tumor-encoded somatic aberration (e.g., mutation, translocation, or methylation event) must be present in the tumor genome if it is ever to be detected in the blood. While genes like APC, TP53 and KRAS are frequently mutated in GI malignancies, none is mutated in 100% of them, and a tumor suppressor gene has several mechanisms by which it can be inactivated, making for an unacceptably large potential search space for some technologies. Second, while poorly understood, quantity of ctDNA is likely related to tumor burden in a non-linear manner, meaning ctDNA is present in small amounts in early stage disease, sometimes occupying only 0.01% of total cell free DNA (cfDNA) (23). For example, in patients with localized GI tumors [colorectal cancer (CRC), gastric or gastroesophageal (GE) cancer, and pancreatic cancer] ctDNA can be detected in 48–73% of patients (24). However, as tumor burden increases, deep surveillance of sufficient genomic space increases ctDNA detection to nearly 100% (25). For this reason—and because tumor genotyping is currently less likely to influence early stage treatment decisions—ctDNA is now most useful in patients with advanced metastatic disease. The standard DNA sequencing approaches such as Sanger sequencing or pyrosequencing are only able to detect ctDNA in patients with significant tumor burden. To improve the ability to detect and analyze ctDNA, a variety of technologies have been developed, including digital PCR (26), digital NGS (27), beads, emulsion, amplification, and magnetics (BEAMing) (28), pyrophosphorolysis-activated polymerization (PAP) (29), cancer personalized profiling by deep sequencing (CAPPSeq) (30), and tagged-amplicon deep sequencing (TAm-Seq) (31).

The optimal liquid biopsy assay remains an area of active investigation. Nonetheless, compared with traditional CTC assays, detection of ctDNA is arguably more sensitive (32,33). For example, in a recent study of patients with advanced solid tumors, ctDNA was always detected when CTCs were present (25). On the other hand, ctDNA was often detected when CTCs were absent. However, it should be noted that this study did not utilize enrichment methods prior to detection of CTCs, something which is now standard amongst modern CTC tests.

In addition to the ongoing research regarding early detection of malignancy, ctDNA is also being developed to detect clinically actionable somatic point mutations or deletions (34), cancer surveillance after definitive surgery (24), and monitoring for the development of molecular resistance to targeted therapies (35).

Circulating exosomes

Exosomes are extracellular vesicles—secreted by all cells—which contain proteins and nucleic acids. Cancer specific exosomes may have unique cell surface proteins which are distinguishable from normal exosomes (36). Glypican-1 (GPC1), a membrane anchored protein overexpressed in breast (37) and pancreatic cancer (38), was recently discovered to be detected exclusively in cancer exosomes (39). In a validation cohort of 56 patients with pancreatic ductal adenocarcinoma (PDA) (with patients from all four stages of cancer including carcinoma in situ), 6 patients with benign pancreatic disease such as chronic pancreatitis, and 20 healthy patients, GPC1⁺ circulating exosomes (crExos) were 100% sensitive and specific for distinguishing patients with PDAC (39). Additional validation studies are needed for this novel approach.

Circulating mRNA

Several circulating mRNA candidates have been studied for CRC screening and prognosis after surgical resection (40-42). Blood mRNA biomarkers such as CEA, CK20, CK19, human telomerase reverse transcriptase (hTERT), and guanylyl cyclase C (GCC) have all been examined in the perioperative setting (42). In most cases, persistence of tumor associated mRNA within 24 hours of tumor resection has been predictive of relapse (43). In a meta-analysis of nine studies with patients undergoing curative surgery for CRC, CTC detection based on CEA, CK19 and CK20 mRNA correlated with the development of hepatic metastases and decreased disease-free survival (44). Recently, Rodia et al. utilized a novel Transcriptome Mapper (TRAM) to identify TSPAN8, LGALS4, COL1A2, and CEACAM6 as candidate mRNAs for detection of CRC (40).

However, circulating mRNA also has limitations. Extracellular mRNA is susceptible to degradation by RNase in plasma (45). Attempts to use RNase inhibitors have been unsuccessful at protecting mRNA from degradation (45). Thus, the use of mRNA in cancer screening and prognosis remains investigational.

Circulating miRNA

miRNAs are small noncoding RNA molecules that regulate cellular processes (46). miRNAs are stable in plasma and serum, properties which make them an excellent candidate as a blood-based tumor diagnostic (47). A recent meta-analysis of 16 studies examined the diagnostic accuracy of circulating miRNA in CRC. This study found that miRNA-21, as opposed to miRNA panels, was promising as a diagnostic biomarker (47). Nonetheless, there are several technical limitations with circulating miRNA. First, there is no standard approach for miRNA isolation. Studies report using serum, plasma, peripheral blood mononuclear cells, and whole blood (48). Second, it has been shown that hemolysis via centrifugation can impact miRNA levels, highlighting the need for consistent enrichment techniques (49). Further standardization is needed before miRNA is appropriate for routine clinical use.

TEPs

Platelets are involved in hemostasis and wound healing. While platelets are anucleate, they contain cytoplasmic pre-mRNA which may be translated into protein in response to external stimuli (50). Tumor cells may influence (“educate”) neighboring platelets to undergo specific splicing of pre-mRNA, thereby altering the RNA profile of blood platelets (51). Using the mRNA of TEPs, Best et al. were able to distinguish between patients with malignancies versus healthy controls with 96% accuracy, and were able to determine the location of the primary tumor with 71% accuracy (52). This technology remains investigational.

Screening

Current cancer screening modalities include direct visualization, imaging, histologic evaluation, and blood protein tests. While current screening methods help identify patients with potentially curable early stage malignancies, there still exists several barriers to cancer screening (53,54). Many tests are uncomfortable (colonoscopies, pap smears, mammograms), expose patients to radiation (CT scans, mammography), are unpleasant to accomplish (stool studies), and almost all require scheduling additional healthcare visits. Thus, the optimal cancer screening assay remains an area of active research.

Two new non-invasive techniques were recently approved for CRC screening. In 2014, the FDA approved fecal DNA testing with Cologuard (Exact Sciences Corporation, Madison, WI, USA). In 2016, the FDA approved Epi proColon (Epigenomics, Berlin, Germany), a blood-based test that detects methylated Septin9 (SEPT9) DNA to identify CRC in patients who choose not to undergo colonoscopy and stool-based fecal immunochemical tests (55). Other assays to detect early stage GI malignancies are currently under development.

Already, liquid biopsies have been utilized to detect clinically silent malignancies. In a study of 125,426 women undergoing non-invasive prenatal testing (NIPT) for fetal aneuploidy via verifi Prenatal Test (Illumina, San Diego, California, USA), 3,757 women tested positive for one or more aneuploidies involving chromosome 13, 18, 21, X, or Y. Of these women, 39 had multiple aneuploidies, and seven (18%) were diagnosed with an occult malignancy (56,57). Three other women with single aneuploidy were also diagnosed with malignancies. In three of the ten cases, the discordant NIPT test prompted the evaluation for malignancy. One limitation of this approach to cancer screening is that it requires a comparison between two different types of tissue—in this case mother and fetus. One approach to overcome this barrier is the use of a “nucleosome footprint”—ctDNA nucleosome spacing is unique with respect to the nuclear architecture and gene expression and thus may inform the cell type (58).

Another method to overcome the use of a comparator is digital karyotyping (DK). Abnormal chromosomal content in the form of losses and gains of entire chromosomes, modification of chromosome arms, amplifications and deletions, and chromosomal rearrangements can all be analyzed via next generation sequencing (NGS) (59). Using the DK approach and NGS data from 81 cancer patients and 10,000 simulated controls, Leary et al. hypothesized that they could reach a sensitivity of >90% and 99% specificity with detection of breast cancer and CRC when ctDNA concentrations reached more than 0.75%. While this test may be sensitive and specific, the required ctDNA concentration is still greater than 0.10%—the sensitivity needed to detect tumors based on BEAMing assays in early stage CRC (23,24). Another challenge will be the detection of false positives, due to constitutional germline or mosaic structural alterations, which can occur in 1.9% of patients older than 75 (60). Given the current challenges identifying ctDNA in patients with low tumor burden (23,25), colonoscopy should remain the standard of care for CRC screening.

Surveillance for recurrent disease

ctDNA may be effective for detecting tumor recurrence. In breast cancer, for example, ctDNA has been shown to be reliable in detecting tumor progression or metastatic disease, often with clinical lead time of up to almost 1 year (61). Similar studies have found that liquid biopsies can also be used for surveillance of GI malignancies (62,63). In addition to detection of recurrent disease, the presence of circulating tumor products may eventually guide adjuvant chemotherapy decisions.

CRC

In patients with CRC, the BEAMing and patient specific somatic structural variants (SSVs) techniques have been investigated as tools to identify cancer recurrence (23,62). In both methods, ctDNA was found to be more reliable and sensitive at detecting tumor recurrence than CEA. Diehl et al. measured ctDNA in patients after both complete and incomplete resections. Those patients with undetectable ctDNA had 100% recurrence free survival (RFS) at 2 years, compared to less than 15% RFS for patients with detectable ctDNA (23). Frattini et al. reported similar findings in a cohort of 70 CRC patients who underwent surgery and then had plasma DNA and CEA levels followed for 16 months (64). Prior to surgery, circulating plasma DNA was elevated in all patients, whereas only 30% of patients had an elevated CEA. At follow up, plasma DNA levels increased in all patients who ultimately developed metastases or loco-regional relapse—however, CEA levels did not increase in all patients with recurrent disease, demonstrating the increased sensitivity of plasma DNA over CEA.

In patients with stage II CRC, ctDNA is being evaluated in the post-operative setting to detect disease recurrence. Preliminary data on 190 patients demonstrate cancer recurrence in 5 of 6 patients with detectable ctDNA and 5 of 72 patients with undetectable ctDNA (63). Patients with detectable ctDNA had a shorter recurrence-free survival (median 234 days vs. undefined, HR 23.09, P<0.0001). For patients with stage II colon cancer—where overall prognosis is good and the benefit of adjuvant chemotherapy is limited, ctDNA might eventually be used to identify patients with the greatest risk of recurrence, and guide adjuvant chemotherapy decision-making (65-67).

Gastric cancer

ctDNA is also being studied as a tool to monitor disease recurrence following gastrectomy. Hamakawa et al. collected gastric cancer tissue and blood from 42 patients undergoing gastrectomy (68). Ten patients had a detectable P53 mutation. Of those 10 patients, pre-operative cfDNA was available in 6 cases and ctDNA was measurable in only 3 cases. ctDNA correlated with disease progression in all three patients, whereas cfDNA did not. To enhance the feasibility of this approach, future efforts will need to include a larger panel of detectable oncogenes.

Pancreatic cancer

Only a minority of patients with pancreatic cancer are candidates for surgical resection. Even when patients undergo resection, recurrence rates range from 70–90% (69). Current post-operative surveillance for pancreatic cancer includes CA19-9 and imaging, both of which lack sensitivity and specificity. Thus, the use of circulating tumor products may be helpful in this setting. A study showed that 14 of 46 patients who had undetectable ctDNA following resection had a longer time to disease recurrence (median time 545 vs. 471 days), but this difference was not statistically significant (HR =0.58; P=0.3) (70). Similar findings were reported by Sausen et al.—patients with detectable ctDNA after surgical resection were more likely to relapse compared with those with undetectable ctDNA (71). Additional larger studies are needed to validate the use of ctDNA as a biomarker for recurrence in resected pancreatic cancer.

Predictive biomarkers—molecular profiling of blood to identify markers of sensitivity and resistance

Precision medicine is based on the notion that molecular profiles (tumor, stroma, immune) can be used to personalize cancer therapy (72). Until recently, most precision medicine clinical trials utilized tissue biopsies—either from primary tumors or at progression—to identify potential predictors of sensitivity and resistance (73-75). While these precision medicine initiatives represent advancements in the treatment of cancer, the technical and operational challenges of relying exclusively on tumor tissue are increasingly recognized. For example, the International Working Group on Multidisciplinary Lung Adenocarcinoma Classification estimated that only 57% of such biopsies had sufficient tissue for genomic analysis after initial pathology diagnosis and staining (76). Since 2009, the NCCN has recommended that all patients with mCRC be tested for KRAS mutations upon diagnosis (65). Yet, in 2010, only 23% of patients with mCRC received guideline based genomic analysis (77,78). While many factors contribute to lack of testing (77), the use of liquid biopsies may ease the practical burden of genomic testing, and facilitate the use of predictive biomarkers in clinical practice.

CRC

One application of ctDNA in CRC is to identify predictive biomarkers of primary treatment resistance. Both KRAS and NRAS (RAS) mutations are well established biomarkers that predict resistance to anti-epidermal growth factor receptor (anti-EGFR) monoclonal antibody therapy (either cetuximab or panitumumab) (79-81). Current NCCN practice guidelines recommend expanded RAS tissue testing in all patients with mCRC (65). Bettegowda et al. evaluated KRAS mutation status in the plasma and tumors of 206 patients with mCRC (25). Of the 206 patients, ctDNA correctly identified 127 of the 128 patients with KRAS wild-type tumors and 69 of 79 patients with KRAS mutations, yielding a sensitivity of 87% and specificity of 99%. Theirry et al. tested 106 patient plasma samples and companion tumor samples for seven KRAS point mutations and yielded 98% specificity and 92% sensitivity with a concordance value of 96% (82). It is unlikely that liquid biopsies will soon supplant standard of care tissue testing for RAS and BRAF mutations. However, in cases where tissue is not safely accessible or is not available, blood-based genotyping may provide a surrogate for tissue.

A second application of ctDNA in CRC is to interrogate the genomics of acquired treatment resistance. Anti-EGFR therapies improve survival in patients with RAS WT mCRC (83-87), but nearly all patients develop resistance. Blood based assays—in particular ctDNA—could help identify acquired mutations that confer treatment resistance. Already, ctDNA has identified KRAS and NRAS mutations as the dominant drivers of acquired resistance to anti-EGFR therapies (88). These mutations may initially exist as rare subclones within a metastatic lesion, but increase under selective pressure (35). Diaz et al. examined the blood of 24 patients with KRAS WT tumors who received anti-EGFR therapy (35). Of the 24 patients, 9 patients were found to develop KRAS mutations in codons 12 and 13 during the course of therapy. The detection of KRAS mutations in blood preceded radiographic evidence of progression in 3 of 9 cases, with an average lead-time of 21 weeks. Additionally, Morelli et al. examined 62 patients who developed resistance to EGFR therapy. Of these 62 patients, 27 patients developed acquired KRAS mutations. In this same study, EGFR-ectodomain (ECD) mutations were detected in five patients treated with cetuximab but none of those treated with panitumumab (88). Since some EGFR ECD mutations do not confer cross-resistance to panitumumab, they are potentially targetable with panitumumab or other novel therapeutics (89-91).

Recently, ctDNA profiling has been utilized to identify other rare or overlapping genomic alterations associated with EGFR resistance. Bettegowda et al. performed ctDNA profiling on 24 patients who initially responded to EGFR therapy. The panel included not only KRAS codons 12 and 13, but also NRAS, BRAF, EGFR, and PIK3CA (25). Of these 24 cases, they discovered at least one mitogen-activated protein kinase pathway gene mutation in 23 patients (96%). The average number of mutations per each patient was 2.9 (range, 0–12). In total, they observed 70 somatic mutations, which had not been detected in the tumor or plasma prior to EGFR blockade.

The combination of ctDNA and tissue-based NGS may provide further insights into mechanisms of EGFR resistance. Siravegna et al. evaluated matched tissue and blood samples from 100 patients with KRAS WT mCRC (92). Ten out of 100 patients received an EGFR antibody but did not have clinical benefit. Of these ten patients, ctDNA revealed NRAS mutations in two patients, which had not been tested prior to treatment. In the remaining eight patients, NGS revealed ERBB2 amplification in four patients, an alteration associated with primary EGFR resistance (93). Another patient had a somatic variant of the MAP2K1 gene encoding the MEK protein, a rare mutation present in 1.5% of CRC tumors, and not yet associated with EGFR resistance.

CTCs have also been utilized to identify actionable targets in patients with CRC. Heitzer et al. isolated CTCs from patients with mCRC and compared 68 CRC-associated genes between primary tumors, metastases, and the corresponding CTCs (94). They found many mutations exclusively within the CTCs but not initially identified in the original tumor tissue.

Pancreatic and biliary cancers

For patients with pancreatic and biliary cancers, the use of tumor tissue to identify markers of sensitivity and resistance has proven challenging. First, these malignancies are often diagnosed by cytology or FNA, leaving limited tissue for molecular studies (95). Second, even when tissue is available, tumor cellularity is often less than 30%, making genomic analyses difficult (96). As a result of these challenges, genotype-directed clinical trials that rely on tissue testing have failed to demonstrate clinical benefit (97).

For patients with pancreatic cancer, ctDNA may offer advantages for genomic profiling. Over 75% of patients with metastatic PDA have detectable ctDNA (25). Zill et al. studied the sensitivity and specificity of ctDNA by analyzing ctDNA and primary tumor tissue of 26 patients with PDAs. They found that ctDNA had 92% sensitivity and 100% specificity in detecting genetic mutations when tumor tissue was used as the gold standard (98). Seven of nine patients (78%) in this study had clinically actionable mutations discovered in ctDNA but not in tumor tissue. One patient had an activating EGFR mutation discovered in blood, which was not seen in tumor until repeat biopsy 7 months later. That patient was treated with capecitabine and erlotinib with an exceptional response. Another patient found to have a FGFR2 mutation was enrolled in a FGFR inhibitor clinical trial. This is in stark contrast to the recent IMPaCT trial, which relied upon tumor tissue profiling, and enrolled no patients in genotype directed therapies (97). Future applications for ctDNA in pancreatic cancer may include identification of BRCA pathway mutations (99) and other potentially actionable genomic markers of DNA mismatch repair deficiency (100).

In addition to ctDNA, investigators have also used exosomal DNA (exoDNA) to assess the genomic profiles of pancreatic cancer (36). One advantage of exoDNA is that it is shielded from degradation in circulation (101). In a small proof-of-concept study, San Lucas et al. used whole exome sequencing to analyze three patients with pancreaticobiliary cancer and were able to find actionable mutations, including ERBB2 and NOTCH1 (102). In one of the three patients, the patient’s pleural fluid was used for analysis. His pleural effusion had fewer than 1% malignant cells on cytospin and a NGS assay was unable to detect tumor DNA. Nonetheless, there were numerous tumor-derived exosomes present for analysis, demonstrating the utility of this approach.

Gastric cancer

Up to one sixth of patients with gastric cancer have HER2 overexpression or amplification (103). Detection of HER2 gene amplification in ctDNA can be serially monitored in patients undergoing treatment (104). In addition to ctDNA, CTCs have also been utilized to predict tumor progression in patients treated with palliative chemotherapy (105).

Prognostic biomarkers

Cancer prognosis has traditionally relied on pathologic stage at the time of diagnosis. Because the presence or absence of circulating tumor products correlates with absolute tumor burden, it may represent another independent prognostic metric. CTCs and ctDNA have been found to correlate with survival in many GI cancers, including colon, pancreatic, and gastric cancer. A recent systematic review of 39 studies and 4,052 patients found that detection of ctDNA is associated with significantly worse overall survival (106).

CRC

In patients with mCRC, there is an inverse relationship between level of cfDNA and survival. In a study by Spindler et al., patients with cfDNA below the median had an overall survival of 12.2 months, versus only 4.5 months in patients with levels above the median (107). In addition to total levels of cfDNA, they were able to correlate plasma mutant KRAS (mutKRAS) levels with prognosis. Patients with high plasma mutKRAS levels had a 0% rate of disease control (stable disease or response), compared with 42% in patients with low plasma mutKRAS levels (108).

In addition to cfDNA and ctDNA, CTCs have also been found to be prognostic in patients with CRC. In a study using CellSearch to detect CTCs, 430 patients with mCRC were split into favorable and unfavorable groups based on the number of CTCs—those patients with more than 3 CTCs per 7.5 mL of blood were classified as unfavorable (109). Those patients with a low CTC count had nearly twice as long progression free survival and overall survival compared to those with a high CTC count. CTC count was prognostic in all subgroups analyzed, including age, ECOG status, liver involvement, receipt of first line chemotherapy, and receipt of various chemotherapy regimens. As technology emerges to detect CTCs in a greater percentage of patients, the utility of CTCs in clinical practice may increase.

Pancreatic cancer

The detection of CTCs has been found to be prognostic in pancreatic cancer. In a randomized controlled trial in patients with locally advanced pancreatic cancer, CTC detection with CellSearch was associated with inferior survival (110). These results have been replicated by several other studies, and a meta analysis of nine independent cohort studies concluded that detection of CTCs is associated with worse progression free survival and overall survival (111).

In addition to CTCs, ctDNA may also predict prognosis in patients with pancreatic cancer (71,112,113). In a pilot study of 45 patients with exocrine pancreatic cancer, patients with KRAS mutated ctDNA had significantly decreased 60 day overall survival compared with those with KRAS WT ctDNA (60 vs. 772 days) (114). In this same study, the detection of CTCs also strongly correlated with survival—however, one major difference was that CTCs were only detectable in patients with metastatic disease, whereas KRAS mutated ctDNA was detected in patients with both resectable and locally advanced disease. The lack of detection of CTCs in early stage disease may be due to the detection technique as well as tumor physiology. A mouse model by Rhim et al. found that only 27% of circulating epithelial cells express EpCAM (115). The use of CTCs and ctDNA to predict prognosis in patients with pancreatic cancer remains investigational.

Gastric cancer

A recent meta-analysis of 26 studies involving 1,950 patients examined the prognostic significance of CTCs in gastric cancer (116). Zhang et al. found that CTCs are associated with worse RFS and overall survival. The authors found that the time at which CTCs were collected was an important factor in predicting overall survival. CTCs collected prior to chemotherapy or surgery were predictive of metastases. CTCs collected immediately after surgery did not impart such prognostic information, as there is a transient increase in CTC detection after surgery, which has been demonstrated both in human and mouse models (117). CTCs collected in the surveillance period after surgery may be useful in indicating tumor recurrence. Thus, when using CTCs as a prognostic biomarker, it is important to specify when CTCs are collected, as blood collected prior to surgery and chemotherapy is likely more meaningful for overall survival, but CTCs collected after surgery are important for monitoring tumor recurrence.

Conclusions

Liquid biopsies have received widespread attention because of the implications for personalized medicine. From cancer screening and surveillance to molecular analysis of metastatic lesions, a liquid biopsy may obviate the need for invasive biopsies and help guide therapeutic decision making. Numerous studies have shown that circulating tumor products can predict prognosis and recurrence risk after surgery. However, there are still barriers that must be overcome before liquid biopsy technology replaces the current tissue-based gold standard. First, the most sensitive and specific methodology of detecting circulating tumor products remains an area of active investigation. Second, there is a paucity of prospective studies linking results from liquid biopsy assays to clinical benefit. Finally, while liquid biopsies may give us the ability to detect mutations, we still lack effective drugs for many genomic alterations. Despite these limitations, liquid biopsy technologies represent a promising step forward in the detection and monitoring of GI cancers.

Acknowledgements

We would like to thank Eric A. Collisson, MD (Helen Diller Family Comprehensive Cancer Center, University of California San Francisco) who provided insight and expertise.

Footnote

Conflicts of Interest: The authors have no conflicts of interest to declare.

References

1. Mandel P, Metais P. Les acides nucléiques du plasma sanguin chez l'homme. C R Seances Soc Biol Fil 1948;142:241-3. [PubMed]
2. Bendich A, Wilczok T, Borenfreund E. Circulating DNA as a possible factor in oncogenesis. Science 1965;148:374-6. [Crossref] [PubMed]
3. Swanton C. Intratumor heterogeneityevolution through space and time. Intratumor heterogeneityevolution through space and time. Cancer Res 2012;72:4875-82. [Crossref] [PubMed]
4. Carreira S, Romanel A, Goodall J, et al. Tumor clone dynamics in lethal prostate cancer. Sci Transl Med 2014;6:254ra125. [Crossref] [PubMed]
5. Gerlinger M, Rowan AJ, Horswell S, et al. Intratumor heterogeneity and branched evolution revealed by multiregion sequencing. N Engl J Med 2012;366:883-92. [Crossref] [PubMed]
6. Lyberopoulou A, Aravantinos G, Efstathopoulos EP, et al. Mutational analysis of circulating tumor cells from colorectal cancer patients and correlation with primary tumor tissue. PLoS One 2015;10e0123902.
7. Cohen SJ, Punt CJ, Iannotti N, et al. Relationship of circulating tumor cells to tumor response, progression-free survival, and overall survival in patients with metastatic colorectal cancer. J Clin Oncol 2008;26:3213-21. [Crossref] [PubMed]
8. Allard WJ, Matera J, Miller MC, et al. Tumor cells circulate in the peripheral blood of all major carcinomas but not in healthy subjects or patients with nonmalignant diseases. Clin Cancer Res 2004;10:6897-904. [Crossref] [PubMed]
9. Stott SL, Lee RJ, Nagrath S, et al. Isolation and characterization of circulating tumor cells from patients with localized and metastatic prostate cancer. Sci Transl Med 2010;2:25ra23. [Crossref] [PubMed]
10. Riethdorf S, Fritsche H, Müller V, et al. Detection of circulating tumor cells in peripheral blood of patients with metastatic breast cancera validation study of the CellSearch system. Clin Cancer Res 2007;13:920-8. [Crossref] [PubMed]
11. Hiraiwa K, Takeuchi H, Hasegawa H, et al. Clinical significance of circulating tumor cells in blood from patients with gastrointestinal cancers. Ann Surg Oncol 2008;15:3092-100. [PubMed]
12. Cohen SJ, Alpaugh RK, Gross S, et al. Isolation and characterization of circulating tumor cells in patients with metastatic colorectal cancer. Clin Colorectal Cancer 2006;6:125-32. [Crossref] [PubMed]
13. Cristofanilli M, Budd GT, Ellis MJ, et al. Circulating tumor cells, disease progression, and survival in metastatic breast cancer. N Engl J Med 2004;351:781-91. [Crossref] [PubMed]
14. Moreno JG, O'Hara SM, Gross S, et al. Changes in circulating carcinoma cells in patients with metastatic prostate cancer correlate with disease status. Urology 2001;58:386-92. [Crossref] [PubMed]
15. Molnar B, Ladanyi A, Tanko L, et al. Circulating tumor cell clusters in the peripheral blood of colorectal cancer patients. Clin Cancer Res 2001;7:4080-5. [PubMed]
16. Rahbari NN, Bork U, Kircher A, et al. Compartmental differences of circulating tumor cells in colorectal cancer. Ann Surg Oncol 2012;19:2195-202. [Crossref] [PubMed]
17. Satelli A, Mitra A, Brownlee Z, et al. Epithelial-mesenchymal transitioned circulating tumor cells capture for detecting tumor progression. Clin Cancer Res 2015;21:899-906. [Crossref] [PubMed]
18. Antolovic D, Galindo L, Carstens A, et al. Heterogeneous detection of circulating tumor cells in patients with colorectal cancer by immunomagnetic enrichment using different EpCAM-specific antibodies. BMC Biotechnol 2010;10:35. [Crossref] [PubMed]
19. Fehm T, Schneck H, Gierke B, et al. Abstract P2-02-13: EpCAM-independent enrichment approach for isolation of circulating tumor cells (CTCs) in breast cancer - What can be found in the EpCAM-depleted fraction? Cancer Res 2016;76:P2-02-13-P2-02-13.
20. Mikolajczyk SD, Millar LS, Tsinberg P, et al. Detection of EpCAM-Negative and Cytokeratin-Negative Circulating Tumor Cells in Peripheral Blood. J Oncol 2011;2011:252361.
21. Zhang T, Boominathan R, Foulk B, et al. Development of a Novel c-MET-Based CTC Detection Platform. Mol Cancer Res 2016.539-47. [Crossref] [PubMed]
22. Diaz LA Jr, Bardelli A. Liquid biopsiesgenotyping circulating tumor DNA. J Clin Oncol 2014;32:579-86. [Crossref] [PubMed]
23. Diehl F, Schmidt K, Choti MA, et al. Circulating mutant DNA to assess tumor dynamics. Nat Med 2008;14:985-90. [Crossref] [PubMed]
24. Diehl F, Li M, Dressman D, et al. Detection and quantification of mutations in the plasma of patients with colorectal tumors. Proc Natl Acad Sci U S A 2005;102:16368-73. [Crossref] [PubMed]
25. Bettegowda C, Sausen M, Leary RJ, et al. Detection of circulating tumor DNA in early- and late-stage human malignancies. Sci Transl Med 2014;6:224ra24. [Crossref] [PubMed]
26. Vogelstein B, Kinzler KW. Digital PCR. Proc Natl Acad Sci U S A 1999;96:9236-41. [Crossref] [PubMed]
27. Lanman RB, Mortimer SA, Zill OA, et al. Analytical and Clinical Validation of a Digital Sequencing Panel for Quantitative, Highly Accurate Evaluation of Cell-Free Circulating Tumor DNA. PLoS One 2015;10e0140712.
28. Dressman D, Yan H, Traverso G, et al. Transforming single DNA molecules into fluorescent magnetic particles for detection and enumeration of genetic variations. Proc Natl Acad Sci U S A 2003;100:8817-22. [Crossref] [PubMed]
29. Liu Q, Sommer SS. Pyrophosphorolysis-activated polymerization (PAP): application to allele-specific amplification. Biotechniques 2000;29:1072-6, 1078, 1080 passim. [PubMed]
30. Newman AM, Bratman SV, To J, et al. An ultrasensitive method for quantitating circulating tumor DNA with broad patient coverage. Nat Med 2014;20:548-54. [Crossref] [PubMed]
31. Forshew T, Murtaza M, Parkinson C, et al. Noninvasive identification and monitoring of cancer mutations by targeted deep sequencing of plasma DNA. Sci Transl Med 2012;4:136ra68. [Crossref] [PubMed]
32. Bidard FC, Madic J, Mariani P, et al. Detection rate and prognostic value of circulating tumor cells and circulating tumor DNA in metastatic uveal melanoma. Int J Cancer 2014;134:1207-13. [Crossref] [PubMed]
33. Freidin MB, Freydina DV, Leung M, et al. Circulating tumor DNA outperforms circulating tumor cells for KRAS mutation detection in thoracic malignancies. Clin Chem 2015;61:1299-304. [Crossref] [PubMed]
34. Sacher AG, Paweletz C, Dahlberg SE, et al. Prospective Validation of Rapid Plasma Genotyping for the Detection of EGFR and KRAS Mutations in Advanced Lung Cancer. JAMA Oncol 2016;2:1014-22. [Crossref] [PubMed]
35. Diaz LA Jr, Williams RT, Wu J, Kinde I, et al. The molecular evolution of acquired resistance to targeted EGFR blockade in colorectal cancers. Nature 2012;486:537-40. [PubMed]
36. Skog J, Würdinger T, van Rijn S, et al. Glioblastoma microvesicles transport RNA and proteins that promote tumour growth and provide diagnostic biomarkers. Nat Cell Biol 2008;10:1470-6. [Crossref] [PubMed]
37. Matsuda K, Maruyama H, Guo F, et al. Glypican-1 is overexpressed in human breast cancer and modulates the mitogenic effects of multiple heparin-binding growth factors in breast cancer cells. Cancer Res 2001;61:5562-9. [PubMed]
38. Kleeff J, Ishiwata T, Kumbasar A, et al. The cell-surface heparan sulfate proteoglycan glypican-1 regulates growth factor action in pancreatic carcinoma cells and is overexpressed in human pancreatic cancer. J Clin Invest 1998;102:1662-73. [Crossref] [PubMed]
39. Melo SA, Luecke LB, Kahlert C, et al. Glypican-1 identifies cancer exosomes and detects early pancreatic cancer. Nature 2015;523:177-82. [Crossref] [PubMed]
40. Rodia MT, Ugolini G, Mattei G, et al. Systematic large-scale meta-analysis identifies a panel of two mRNAs as blood biomarkers for colorectal cancer detection. Oncotarget 2016;7:30295-306. [PubMed]
41. Bujanda L, Sarasqueta C, Cosme A, et al. Evaluation of alpha 1-antitrypsin and the levels of mRNA expression of matrix metalloproteinase 7, urokinase type plasminogen activator receptor and COX-2 for the diagnosis of colorectal cancer. PLoS One 2013;8e51810.
42. Lim SH, Spring KJ, de Souza P, et al. Circulating tumour cells and circulating nucleic acids as a measure of tumour dissemination in non-metastatic colorectal cancer surgery. Eur J Surg Oncol 2015;41:309-14. [Crossref] [PubMed]
43. Allen-Mersh TG, McCullough TK, Patel H, et al. Role of circulating tumour cells in predicting recurrence after excision of primary colorectal carcinoma. Br J Surg 2007;94:96-105. [Crossref] [PubMed]
44. Katsuno H, Zacharakis E, Aziz O, et al. Does the presence of circulating tumor cells in the venous drainage of curative colorectal cancer resections determine prognosis? A meta-analysis. Ann Surg Oncol 2008;15:3083-91. [Crossref] [PubMed]
45. El-Hefnawy T, Raja S, Kelly L, et al. Characterization of amplifiable, circulating RNA in plasma and its potential as a tool for cancer diagnostics. Clin Chem 2004;50:564-73. [Crossref] [PubMed]
46. He L, Thomson JM, Hemann MT, et al. A microRNA polycistron as a potential human oncogene. Nature 2005;435:828-33. [Crossref] [PubMed]
47. Rokkas T, Kothonas F, Rokka A, et al. The role of circulating microRNAs as novel biomarkers in diagnosing colorectal cancera meta-analysis. Eur J Gastroenterol Hepatol 2015;27:819-25. [Crossref] [PubMed]
48. De Guire V, Robitaille R, Tétreault N, et al. Circulating miRNAs as sensitive and specific biomarkers for the diagnosis and monitoring of human diseasespromises and challenges. Clin Biochem 2013;46:846-60. [Crossref] [PubMed]
49. Pritchard CC, Kroh E, Wood B, et al. Blood cell origin of circulating microRNAsa cautionary note for cancer biomarker studies. Cancer Prev Res (Phila) 2012;5:492-7. [Crossref] [PubMed]
50. Macaulay IC, Carr P, Gusnanto A, et al. Platelet genomics and proteomics in human health and disease. J Clin Invest 2005;115:3370-7. [Crossref] [PubMed]
51. Calverley DC, Phang TL, Choudhury QG, et al. Significant downregulation of platelet gene expression in metastatic lung cancer. Clin Transl Sci 2010;3:227-32. [Crossref] [PubMed]
52. Best MG, Sol N, Kooi I, et al. RNA-Seq of Tumor-Educated Platelets Enables Blood-Based Pan-Cancer, Multiclass, and Molecular Pathway Cancer Diagnostics. Cancer Cell 2015;28:666-76. [Crossref] [PubMed]
53. Finney Rutten LJ, Nelson DE, et al. Meissner HI. Examination of population-wide trends in barriers to cancer screening from a diffusion of innovation perspective (1987-2000). Prev Med 2004;38:258-68. [Crossref] [PubMed]
54. Lo SH, Waller J, Wardle J, et al. Comparing barriers to colorectal cancer screening with barriers to breast and cervical screeninga population-based survey of screening-age women in Great Britain. J Med Screen 2013;20:73-9. [Crossref] [PubMed]
55. Potter NT, Hurban P, White MN, et al. Validation of a real-time PCR-based qualitative assay for the detection of methylated SEPT9 DNA in human plasma. Clin Chem 2014;60:1183-91. [Crossref] [PubMed]
56. Romero R, Mahoney MJ. Noninvasive Prenatal Testing and Detection of Maternal Cancer. JAMA 2015;314:131-3. [Crossref] [PubMed]
57. Bianchi DW, Chudova D, Sehnert AJ, et al. Noninvasive Prenatal Testing and Incidental Detection of Occult Maternal Malignancies. JAMA 2015;314:162-9. [Crossref] [PubMed]
58. Snyder MW, Kircher M, Hill AJ, et al. Cell-free DNA Comprises an In Vivo Nucleosome Footprint that Informs Its Tissues-Of-Origin. Cell 2016;164:57-68. [Crossref] [PubMed]
59. Leary RJ, Sausen M, Kinde I, et al. Detection of chromosomal alterations in the circulation of cancer patients with whole-genome sequencing. Sci Transl Med 2012;4:162ra154. [Crossref] [PubMed]
60. Jacobs KB, Yeager M, Zhou W. Detectable clonal mosaicism and its relationship to aging and cancer. Nat Genet 2012;44:651-8. [Crossref] [PubMed]
61. Olsson E, Winter C, George A, et al. Serial monitoring of circulating tumor DNA in patients with primary breast cancer for detection of occult metastatic disease. EMBO Mol Med 2015;7:1034-47. [Crossref] [PubMed]
62. Reinert T, Schøler LV, Thomsen R, et al. Analysis of circulating tumour DNA to monitor disease burden following colorectal cancer surgery. Gut 2016;65:625-34. [Crossref] [PubMed]
63. Tie J, Kinde I, Wang Y, et al. Circulating tumor DNA (ctDNA) as a marker of recurrence risk in stage II colon cancer (CC). 2014 ASCO Annual Meeting Proceedings. Chicago, 2014.
64. Frattini M, Gallino G, Signoroni S, et al. Quantitative and qualitative characterization of plasma DNA identifies primary and recurrent colorectal cancer. Cancer Lett 2008;263:170-81. [Crossref] [PubMed]
65. NCCN clinical practice guidelines in oncology: Colon Cancer. 2016. Available online: http://www.nccn.org/professionals/physician_gls/pdf/colon.pdf
66. Efficacy of adjuvant fluorouracil and folinic acid in colon cancer. International Multicentre Pooled Analysis of Colon Cancer Trials (IMPACT) investigators. Lancet 1995;345:939-44. [Crossref] [PubMed]
67. Quasar Collaborative Group, Gray R, Barnwell J, et al. Adjuvant chemotherapy versus observation in patients with colorectal cancer a randomised study. Lancet 2007;370:2020-9. [Crossref] [PubMed]
68. Hamakawa T, Kukita Y, Kurokawa Y, et al. Monitoring gastric cancer progression with circulating tumour DNA. Br J Cancer 2015;112:352-6. [Crossref] [PubMed]
69. Oettle H, Post S, Neuhaus P, et al. Adjuvant chemotherapy with gemcitabine vs observation in patients undergoing curative-intent resection of pancreatic cancer a randomized controlled trial. JAMA 2007;297:267-77. [Crossref] [PubMed]
70. Wang JS-Z, Sausen M, Parpart-Li S, et al. Circulating tumor DNA (ctDNA) as a prognostic marker for recurrence in resected pancreas cancer. 2015 ASCO Annual Meeting Proceedings. Chicago, 2015.
71. Sausen M, Phallen J, Adleff V, et al. Clinical implications of genomic alterations in the tumour and circulation of pancreatic cancer patients. Nat Commun 2015;6:7686. [Crossref] [PubMed]
72. Schork NJ. Personalized medicine: Time for one-person trials. Nature 2015;520:609-11. [Crossref] [PubMed]
73. McNeil C. NCI-MATCH launch highlights new trial design in precision-medicine era. J Natl Cancer Inst 2015.107. [PubMed]
74. Herbst RS, Gandara DR, Hirsch FR, et al. Lung Master Protocol (Lung-MAP)-A Biomarker-Driven Protocol for Accelerating Development of Therapies for Squamous Cell Lung CancerSWOG S1400. Clin Cancer Res 2015;21:1514-24. [Crossref] [PubMed]
75. NCI launches ALCHEMIST. Cancer Discov 2014;4OF9.
76. Travis WD, Brambilla E, Noguchi M, et al. International association for the study of lung cancer/american thoracic society/european respiratory society international multidisciplinary classification of lung adenocarcinoma. J Thorac Oncol 2011;6:244-85. [Crossref] [PubMed]
77. Charlton ME, Karlitz JJ, Schlichting JA, et al. Factors Associated With Guideline-recommended KRAS Testing in Colorectal Cancer Patients: A Population-based Study. Am J Clin Oncol 2015. [Epub ahead of print]. [Crossref] [PubMed]
78. Webster J, Kauffman TL, Feigelson HS, et al. KRAS testing and epidermal growth factor receptor inhibitor treatment for colorectal cancer in community settings. Cancer Epidemiol Biomarkers Prev 2013;22:91-101. [Crossref] [PubMed]
79. Karapetis CS, Khambata-Ford S, Jonker DJ, et al. K-ras mutations and benefit from cetuximab in advanced colorectal cancer. N Engl J Med 2008;359:1757-65. [Crossref] [PubMed]
80. Qiu LX, Mao C, Zhang J, et al. Predictive and prognostic value of KRAS mutations in metastatic colorectal cancer patients treated with cetuximaba meta-analysis of 22 studies. Eur J Cancer 2010;46:2781-7. [Crossref] [PubMed]
81. Douillard JY, Oliner KS, Siena S, et al. Panitumumab-FOLFOX4 treatment and RAS mutations in colorectal cancer. N Engl J Med 2013;369:1023-34. [Crossref] [PubMed]
82. Thierry AR, Mouliere F, El Messaoudi S, et al. Clinical validation of the detection of KRAS and BRAF mutations from circulating tumor DNA. Nat Med 2014;20:430-5. [Crossref] [PubMed]
83. Douillard JY, Siena S, Cassidy J, et al. Final results from PRIME: randomized phase III study of panitumumab with FOLFOX4 for first-line treatment of metastatic colorectal cancer. Ann Oncol 2014;25:1346-55. [Crossref] [PubMed]
84. Van Cutsem E, Köhne CH, Hitre E, et al. Cetuximab and chemotherapy as initial treatment for metastatic colorectal cancer. N Engl J Med 2009;360:1408-17. [Crossref] [PubMed]
85. Van Cutsem E, Peeters M, Siena S, et al. Open-label phase III trial of panitumumab plus best supportive care compared with best supportive care alone in patients with chemotherapy-refractory metastatic colorectal cancer. J Clin Oncol 2007;25:1658-64. [Crossref] [PubMed]
86. Cunningham D, Humblet Y, Siena S, et al. Cetuximab monotherapy and cetuximab plus irinotecan in irinotecan-refractory metastatic colorectal cancer. N Engl J Med 2004;351:337-45. [Crossref] [PubMed]
87. Jonker DJ, O'Callaghan CJ, Karapetis CS, et al. Cetuximab for the treatment of colorectal cancer. N Engl J Med 2007;357:2040-8. [Crossref] [PubMed]
88. Morelli MP, Overman MJ, Dasari A, et al. Characterizing the patterns of clonal selection in circulating tumor DNA from patients with colorectal cancer refractory to anti-EGFR treatment. Ann Oncol 2015;26:731-6. [Crossref] [PubMed]
89. Sánchez-Martín FJ, Bellosillo B, Gelabert-Baldrich M, et al. The First-in-class Anti-EGFR Antibody Mixture Sym004 Overcomes Cetuximab Resistance Mediated by EGFR Extracellular Domain Mutations in Colorectal Cancer. Clin Cancer Res 2016;22:3260-7. [Crossref] [PubMed]
90. Arena S, Siravegna G, Mussolin B, et al. MM-151 overcomes acquired resistance to cetuximab and panitumumab in colorectal cancers harboring EGFR extracellular domain mutations. Sci Transl Med 2016 3;8:324ra14.
91. Montagut C, Dalmases A, Bellosillo B, et al. Identification of a mutation in the extracellular domain of the Epidermal Growth Factor Receptor conferring cetuximab resistance in colorectal cancer. Nat Med 2012;18:221-3. [Crossref] [PubMed]
92. Siravegna G, Mussolin B, Buscarino M, et al. Clonal evolution and resistance to EGFR blockade in the blood of colorectal cancer patients. Nat Med 2015;21:795-801. [Crossref] [PubMed]
93. Yonesaka K, Zejnullahu K, Okamoto I, et al. Activation of ERBB2 signaling causes resistance to the EGFR-directed therapeutic antibody cetuximab. Sci Transl Med 2011;3:99ra86. [Crossref] [PubMed]
94. Heitzer E, Auer M, Gasch C, et al. Complex tumor genomes inferred from single circulating tumor cells by array-CGH and next-generation sequencing. Cancer Res 2013;73:2965-75. [Crossref] [PubMed]
95. Butturini G, Stocken DD, Wente MN, et al. Influence of resection margins and treatment on survival in patients with pancreatic cancermeta-analysis of randomized controlled trials. Arch Surg 2008;143:75-83. [Crossref] [PubMed]
96. Song S, Nones K, Miller D, et al. qpure: A tool to estimate tumor cellularity from genome-wide single-nucleotide polymorphism profiles. PLoS One 2012;7:e45835. [Crossref] [PubMed]
97. Chantrill LA, Nagrial AM, Watson C, et al. Precision Medicine for Advanced Pancreas CancerThe Individualized Molecular Pancreatic Cancer Therapy (IMPaCT) Trial. Clin Cancer Res 2015;21:2029-37. [Crossref] [PubMed]
98. Zill OA, Greene C, Sebisanovic D, et al. Cell-Free DNA Next-Generation Sequencing in Pancreatobiliary Carcinomas. Cancer Discov 2015;5:1040-8. [Crossref] [PubMed]
99. Waddell N, Pajic M, Patch AM, et al. Whole genomes redefine the mutational landscape of pancreatic cancer. Nature 2015;518:495-501. [Crossref] [PubMed]
100. Le DT, Uram JN, Wang H, et al. PD-1 Blockade in Tumors with Mismatch-Repair Deficiency. N Engl J Med 2015;372:2509-20. [Crossref] [PubMed]
101. Kahlert C, Kalluri R. Exosomes in tumor microenvironment influence cancer progression and metastasis. J Mol Med (Berl) 2013;91:431-7. [Crossref] [PubMed]
102. San Lucas FA, Allenson K, Bernard V, et al. Minimally invasive genomic and transcriptomic profiling of visceral cancers by next-generation sequencing of circulating exosomes. Ann Oncol 2016;27:635-41. [Crossref] [PubMed]
103. Gravalos C, Jimeno A. HER2 in gastric cancera new prognostic factor and a novel therapeutic target. Ann Oncol 2008;19:1523-9. [Crossref] [PubMed]
104. Shoda K, Masuda K, Ichikawa D, et al. HER2 amplification detected in the circulating DNA of patients with gastric cancera retrospective pilot study. Gastric Cancer 2015;18:698-710. [Crossref] [PubMed]
105. Uenosono Y, Arigami T, Kozono T, et al. Clinical significance of circulating tumor cells in peripheral blood from patients with gastric cancer. Cancer 2013;119:3984-91. [Crossref] [PubMed]
106. Ocaña A, Díez-González L, García-Olmo DC, et al. Circulating DNA and Survival in Solid Tumors. Cancer Epidemiol Biomarkers Prev 2016;25:399-406. [Crossref] [PubMed]
107. Spindler KL, Pallisgaard N, Andersen RF, et al. Circulating free DNA as biomarker and source for mutation detection in metastatic colorectal cancer. PLoS One 2015;10:e0108247. [Crossref] [PubMed]
108. Spindler KL, Pallisgaard N, Vogelius I, et al. Quantitative cell-free DNA, KRAS, and BRAF mutations in plasma from patients with metastatic colorectal cancer during treatment with cetuximab and irinotecan. Clin Cancer Res 2012;18:1177-85. [Crossref] [PubMed]
109. Cohen SJ, Punt CJ, Iannotti N, et al. Prognostic significance of circulating tumor cells in patients with metastatic colorectal cancer. Ann Oncol 2009;20:1223-9. [Crossref] [PubMed]
110. Bidard FC, Huguet F, Louvet C, et al. Circulating tumor cells in locally advanced pancreatic adenocarcinoma: the ancillary CirCe 07 study to the LAP 07 trial. Ann Oncol 2013;24:2057-61. [Crossref] [PubMed]
111. Han L, Chen W, Zhao Q. Prognostic value of circulating tumor cells in patients with pancreatic cancera meta-analysis. Tumour Biol 2014;35:2473-80. [Crossref] [PubMed]
112. Kinugasa H, Nouso K, Miyahara K, et al. Detection of K-ras gene mutation by liquid biopsy in patients with pancreatic cancer. Cancer 2015;121:2271-80. [Crossref] [PubMed]
113. Chen H, Tu H, Meng ZQ, et al. K-ras mutational status predicts poor prognosis in unresectable pancreatic cancer. Eur J Surg Oncol 2010;36:657-62. [Crossref] [PubMed]
114. Earl J, Garcia-Nieto S, Martinez-Avila JC, et al. Circulating tumor cells (Ctc) and kras mutant circulating free Dna (cfdna) detection in peripheral blood as biomarkers in patients diagnosed with exocrine pancreatic cancer. BMC Cancer 2015;15:797. [Crossref] [PubMed]
115. Rhim AD, Mirek ET, Aiello NM, et al. EMT and dissemination precede pancreatic tumor formation. Cell 2012;148:349-61. [Crossref] [PubMed]
116. Zhang ZY, Dai ZL, Yin XW, et al. Meta-analysis shows that circulating tumor cells including circulating microRNAs are useful to predict the survival of patients with gastric cancer. BMC Cancer 2014 21;14:773.
117. Miyazono F, Natsugoe S, Takao S, et al. Surgical maneuvers enhance molecular detection of circulating tumor cells during gastric cancer surgery. Ann Surg 2001;233:189-94. [Crossref] [PubMed]