Sage Journals: Discover world-class research

Abstract

BACKGROUND:

Circulating tumor DNA (ctDNA) has been recognized as a promising biomarker for colorectal cancer (CRC) early diagnosis and postoperative monitoring. However, we hypothesize that the clinical value of ctDNA sequencing may differ for colon cancer (CC) and rectal cancer (RC).

METHODS:

Forty-three patients with primary CRC were prospectively enrolled. Tumor tissue samples, paired preoperative plasma samples and a series of postoperative plasma samples were obtained. Mutations in each sample were identified and compared.

RESULTS:

For 73.0% patients, at least one concordant mutation was detected in both tumor tissue DNA and paired preoperative ctDNA. The mutation concordance rate were higher in CC patients compared to RC patients (92.3% vs 45.5%; $p=$ 0.004). For early stage patients, the mutation concordance rate was 72.7%. The recurrence rate was 33.3% for patients with postoperative ctDNA positive mutations, and 3.4% for patients with negative ctDNA (HR 10.767; 95% CI 1.1–103.8; $p=$ 0.040).

CONCLUSION:

Liquid biopsy via ctDNA sequencing has great potential for the early detection and postoperative monitoring of CRC. The DNA of CC tissues is more likely to be released into blood than the DNA of RC tissues. This should be considered when diagnosing CRC patients with ctDNA sequencing technology.

Keywords

Circulating tumor DNA colorectal cancer early diagnosis predicting recurrence next generation sequencing

1. Introduction

The early diagnosis of colorectal cancer (CRC) offers great potential for reducing mortality. CRC is the third-most commonly diagnosed malignancy and the fourth leading cause of cancer-related deaths worldwide, accounting for about 1.4 million diagnosed cases and almost 700,000 deaths in 2012 [1]. The five-year survival rate for patients with stage I or II CRC reaches up to 90%, whereas it is only 12% for patients with stage IV CRC with multiple organ metastases. Unfortunately, approximately three fifths of patients are diagnosed at the late stage. CRC relapse occurs in 30–50% of patients, even in cases where treatment is effective [2, 3]. The survival rate can be rose, if the evaluation of each patient’s risk of recurrence can be improved. Therefore, there is a need for modified diagnostic tools in CRC. Colonoscopy has been widely used for the screening of CRC and contributed strongly to a drop in the mortality rate of CRC patients [4]. However, up to 25% of all colonoscopies fail due to extraordinary pain and inadequate bowel preparation [5]. Nowadays, carcinoembryonic antigen (CEA) is widely used as a potential biomarker for cancer diagnosis; however, it does not provide sufficient sensitivity and reliability since high CEA levels may be caused by infection and other diseases [6, 7]. Furthermore, CEA is not a sufficiently good indicator of the cancer severity since it may not increase in patients with advanced CRC [6]. Positron emission tomography (PET) and computed tomography (CT) are also effective diagnostic tools for cancers; however, they are associated with radiation exposure and also suffer from a high rate of false positivity [8]. Traditional recurrence risk evaluation methods that examine CEA and/or PET/CT also have not shown promising results [9, 10]. Almost 40% of patients with recurrence are not detected by CEA and/or PET/CT [11].

Fragments of circulating cell-free DNA (cfDNA) are released by different tissues in patients as well as healthy individuals. Noninvasive liquid biopsy based on cfDNA can be a great supplement of traditional diagnostic approaches [12, 13, 14]. Tumor-specific DNA mutations can be detected in circulating tumor DNA (ctDNA) [15], released from primary tumors or metastases, which is a significant component of the cfDNA [16]. Extensive studies have demonstrated that ctDNA could be used for the early diagnosis of a variety of cancers [17, 18, 19, 20], since it features higher sensitivity and specificity than traditional biomarkers [21, 22]. In addition, ctDNA can reflect the tumor burden in real time and offers both the potential for clinical implications in monitoring patients with cancer and prediction of recurrence risk [12, 21, 23, 24, 25]. The results of this study indicate the association between ctDNA and the occurrence and recurrence of colorectal cancer.

The obtained results also show that the ctDNA sequencing can provide different clinical outcome for colon cancer (CC) and rectal cancer (RC). Although CC and RC are collectively labeled CRC, their embryonic origin, anatomy, and function are different [26]. For example, the capillary networks of the colon are multi-layered and more abundant than those of the rectum [26]. Accordingly, we hypothesized that the ctDNA pattern of CC and RC patients may also differ. To further verify this hypothesis, the somatic mutations detected in the sequencing data of tumor tissues and paired plasma from patients with CC or RC were analyzed. The results show that the ctDNA can enable different clinical outcome for CC and RC.

2. Materials and methods

2.1 Ethics statement and patient cohorts

This study protocol was approved by the ethics committee of Shenzhen People’s Hospital on 4/6/2016 and the methods were carried out in accordance with the approved guidelines. All patients were counseled and consented according to the Declaration of Helsinki. Written informed consent was obtained from all patients for the use of blood and tumor tissue under the approval of the Ethics Committee. All samples and medical data used in this study have been irreversibly anonymized.

From September 2016 to May 2019, 43 patients with stage I, II, III, and IV CRC were prospectively enrolled at the Shenzhen People’s Hospital in Shenzhen, China. All patients were diagnosed with primary CRC and underwent radical resection (R0 resection), including 29 patients with CC and 14 patients with RC. For each patient, a fresh frozen tissues sample was collected during surgery and immediately stored at $-$ 80 ${}^{\circ}$ C until DNA extraction. Its paired plasma sample was obtained prior to surgery and centrifuged twice at 4 ${}^{\circ}$ C within 3 h. Postoperative plasma samples were collected successively one month (1 ${}^{\text{st}}$ ), two months (2 ${}^{\text{nd}}$ ), and half a year (3 ${}^{\text{rd}}$ ) after surgery. All plasma samples were frozen at $-$ 80 ${}^{\circ}$ C until cfDNA extraction. Part of the data from 11 patients has been reported previously [27].

2.2 DNA extraction and sequencing

Tumor DNA was extracted with the EasyPure ${}^{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}$ Genomic DNA Kit (Beijing TransGen Biotech, Beijing, China). CfDNA was extracted from at least 2 mL plasma using the Serum/Plasma Free DNA Extraction Kit (Ningbo ZD Biotechnology, Ningbo, China). DNA was quantified with the Qubit 2.0 Fluorometer and the Qubit dsDNA HS Assay kit (Life Technologies, Carlsbad, USA) according to the recommended protocol. CfDNA samples were also analyzed on the Agilent 4200 TapeStation system (Agilent Technologies, USA) and qualitatively measured by locating the mononucleosomal peak.

Table 1
Clinic pathological characteristics and their correlation with preoperative ctDNA

Characteristics	All patients ( $n=$ 43)	Preoperative ctDNA ( $n=$ 37)
		Positive ( $n=$ 29)	Negative ( $n=$ 8)	$p$
Age at surgery
Median age	58	61	49.5	0.169
Range	30–77	30–73	37–74
Gender, no. (%)
Male	23 (53.5%)	15 (51.7%)	5 (62.5%)	0.600
Female	20 (46.5%)	14 (48.3%)	3 (37.5%)
Tumor volume (cm ${}^{3}$ )
Median	12.45	16.88	7	0.070
Range	1.12–170.63	2.25–170.63	1.12–108
Tumor stage, no. (%)
I	9 (20.9%)	6 (20.7%)	3 (37.5%)	0.208
II	14 (32.6% )	10 (34.5%)	3 (37.5%)
III	15 (34.9%)	9 (31.0%)	2 (25.0%)
IV	5 (11.6%)	4 (13.8%)	0 (0.0%)
Localisation, no. (%)
Colon	29 (67.4%)	24 (82.8%)	2 (25.0%)	0.001
Rectum	14 (32.6%)	5 (17.2%)	6 (75.0%)
Vascular invasion
Yes	12 (27.9%)	9 (31.0%)	1 (12.5%)	0.287
No	29 (67.4%)	19 (65.5%)	7 (87.5%)
NA	2 (4.7%)	1 (3.4%)	0 (0.0%)
Clinical risk, no. (%)
High	27 (62.8%)	20 (69.0%)	4 (50.0%)	0.333
Low	16 (37.2%)	9 (31.0%)	4 (50.0%)
Family history, no. (%)
Yes	11 (25.6%)	8 (27.6%)	2 (25.0%)	1.000
No	30 (69.8%)	20 (69.0%)	5 (62.5%)
NA	2 (4.7%)	1 (3.4%)	1 (12.5%)
Recurrence, no. (%)
Yes	5 (11.6%)	5 (17.2%)	0 (0.0%)	0.199
No	36 (83.7%)	22 (75.9%)	8 (100%)
NA	2 (4.7%)	2 (6.9%)	0 (0.0%)
Total	43	29	8

About 200 ng tumor DNA fragments and 10–50 ng cfDNA were input to construct libraries by the KAPA Library Preparation Kit (KAPA Biosystems, USA) according to the manufacturer’s protocol. Target enrichment was performed with the custom SeqCap EZ Library Kit (Roche NimbleGen, Madison, WI, USA). The targeted capture probes were designed based on genomic regions of 85 genes that covered most CRC related mutations (Supplemental Table 1). Capture hybridization was conducted according to the manufacturer’s protocol. After that, the libraries were amplified for 7–10 PCR cycles using 1 $\times$ KAPA HiFi Hot Start Ready Mix and Pre-LM-PCR Oligos (KAPA Biosystems, USA) in 50 $\mu$ l reactions depending on the amount of DNA input. Sequencing was carried out to $>$ 10,000 $\times$ coverage on the Illumina NextSeq 500/550 instrument according to the manufacturer’s protocol using NextSeq 500/550 v2 Kits (Illumina, San Diego, CA, USA).

2.3 Sequencing data analysis and variant calling

Sequencing data were preprocessed by Fastp for profiling quality status, trimming sequencing adapters and filtering low quality reads [28]. The filtered rdata were than mapped and aligned to the reference human genome, hg19 genome (GRch37) using Burrows Wheeler Aligner after removal of terminal adaptor sequences and low-quality reads. Pileup files for properly paired reads with mapping quality $\geqslant$ 60 were generated using SAMtools. Somatic variants were called using VarScan2 (– min-coverage $=$ 100; – min-varfreq $=$ 0.05; – somatic- $p$ -value $=$ 0.01; – strand-filter $=$ 1; otherwise, default parameters). Variant allele frequencies (VAFs) were calculated for all Q30 bases. Previously identified tumor DNA mutations were intersected with a SAMtools mpileup file generated for each plasma DNA sample, using a custom Python script. The number and frequency were then calculated for each mutation. To detect mutations from plasma DNA without tumor DNA present, a mutation was called if $\geqslant$ 4 mutant reads were found in plasma with $\geqslant$ 1 read on each strand. No mutant reads were observed in matched genomic DNA from white blood cells.

Figure 1.

Patient enrolment and study overview.

Figure 2.

The yield of pre and post-operative cfDNA.

2.4 Statistical analysis

Differences in the sensitivity between ctDNA and CEA and between colon cancer and rectal cancer were assessed using the Fisher’s exact test. Correlation analyses between ctDNA status and clinical characteristics were evaluated using the Spearman’s coefficient of correlations. Hazard rations (HRs) were estimated using univariate Cox proportional hazard models. The two-tailed $p$ value $<$ 0.05 was considered to indicate significant differences for all statistical analyses.

3. Results

3.1 Patient characteristics

From September 2016 to May 2019, 43 CRC patients were prospectively enrolled at the Department of Gastrointestinal Surgery of the Shenzhen People’s Hospital in China. Patient characteristics and tumor features are summarized in Table 1 and Supplemental Table 2. The average age of patients was 58 and 53.5% patients were male. All eligible patients had histologically confirmed adenocarcinoma of the colon (67.4%) or rectum (32.6%) with stage I, II, III, and IV, and 39 (90.7%) of patients were moderately differentiated. All 43 patients underwent radical resection (R0 resection) and 29 (67.4%) patients received adjuvant chemotherapy at their clinician’s discretion. According to clinical and pathological features, 27 (62.8%) patients were identified with a high recurrence risk. By the end of 30 June 2019, the median follow-up time was 120 weeks (range, 7–146 weeks). At the end of follow-up, five patients (12.2%) experienced recurrence and of two (4.9%) died of disease progression. Patient enrolment and the study overview are presented in Fig. 1.

Figure 3.

Mutation statistics of tumor samples and preoperative plasma from 43 patients with CRC. The colored boxes indicate where the mutations were detected in tumor tissues, and the triangles where the mutations were detected in preoperative plasma ctDNA.

3.2 cfDNA yield in preoperative and postoperative plasma

The median cfDNA yield in properative and postoperative plasma was 7.50 ng and 7.64 ng per mL plasma, range from 1.8 ng to 73.8 ng and from 1.8 ng to 117.2 ng. There were neither correlation ( $p=$ 0.988) nor significant difference ( $p=$ 0.717) between them. The yield of properative cfDNA showed an increasing trend alone with tumor stage I, II and III (Fig. 2A). However, there did not show the same trend in stage IV, which possibly due to the small sample size. The yield of postperative cfDNA didn’t show any trend (Fig. 2B).

3.3 Preoperative plasma status and CRC diagnosis

To identify tumor somatic single nucleotide variants (SNV) and short insertion/deletion (InDel) mutations, they were detected in tumor tissues via targeted capture sequencing. At least one somatic mutation was identified in tumor tissue of 40 of the 43 (93.0%) eligible patients. The median mutation number was 3 (range, 0–15). The most frequent mutations were gene TP53 (51.2%), APC (48.8%), KRAS (37.2%), and PIK3CA (13.95%) in the enrolled patients, which were “hotspot” mutations consistent with previous studies. No mutation of these four genes was detected in only six patients (14.0%) (Fig. 3).

Thirty-seven (86.0%) paired preoperative plasma samples were collected and sequenced using targeted capture panel. In 73.0% (27/37) of all patients, at least one mutation could be detected in preoperative plasma, which was consistent with the paired tumor tissue. No mutations were detected in tumor tissue of patient H19, H34 and H41. However, a missense mutation in exon 2 of NRAS (G37A) was found in the preoperative plasma sample of H19, and the VAFs was 0.3%. KRAS G12D, TP53 R273H and SMAD4 G419V were found in the preoperative plasma sample of H34. And there wasn’t found any mutation in the preoperative plasma sample of H41. The other seven patients did not show any mutation in the preoperative ctDNA. The most frequently identified alterations in preoperative ctDNA were also APC (32.4%), TP53 (27.0%), KRAS (27.0%), and PIK3CA (8.1%), which was consistent with the tumor tissues. Furthermore, none of these four genes were mutated in twelve 32.4%) patients (Fig. 3), among 8 patients were negtive.

The sensitivity of ctDNA for diagnosis CRC reached up to 78.4% (29/37). The sensitivity was 72.7% in 22 patients with stage I and II CRC, and 86.7% (13/15) in patients with stage III and IV CRC. Preoperative CEA levels of 35 patients were collected among all eligible patients. However, only in 11 patients, the CEA were detected positive ( $>$ 5 $\mu$ g/L). The sensitivity of CEA for diagnosis CRC was only 31.4%, which was much lower than ctDNA. The sensitivity of CEA was even lower than 21.7% (5/23) in patients with stage I and II CRC. The patient characteristics, including tumor location, clinical risk factors, and the cancer stage, are shown at the bottom of Fig. 3.

Figure 4.

Dynamic ctDNA monitoring of patients H03 and H21. A. Dynamic ctDNA monitoring of patient H03. B. Dynamic ctDNA monitoring for patient H21. The orange shaded box shows the period of adjuvant chemotherapy. The blue lines shows the time point of surgery and died or reccured. VAF, variant allele fractions; ctDNA, circulating tumor DNA.

3.4 Differences of ctDNA between CC and RC preoperative blood samples

The yield of preoperative plasma cfDNA in CC was little higher than in RC (media yield of preoperative ctDNA per mL plasma was 8.25 ng vs 7.51 ng). Six of the eight preoperative ctDNA-negative patients had RC and only two had CC, which differed significantly ( $p=$ 0.015). In total, at least one concordant mutation was detected in both tumor tissue DNA and paired preoperative ctDNA of 84.6% (22/26) of CC patients and 45.5% (5/11) of RC patients. Preoperative ctDNA-positive was significantly associated with tumor location ( $p=$ 0.001); however, not with primary tumor stage ( $p=$ 0.208) and other clinical characteristics (Table 1). The sensitivity of ctDNA for diagnosis of CC and RC were 92.3% (24/26) and 45.5% (5/11) ( $p=$ 0.004), respectively. In addition, KRAS and PIK3CA mutations were not detected in preoperative ctDNA of RC patients, which may be caused by our limited sample size for differences in populations.

3.5 Postoperative blood ctDNA for CRC recurrence prediction

A total of 98 postoperative plasma samples were collected from 39 enrolled patients. The number of mutations detected in the 1 ${}^{\text{st}}$ postoperative ctDNA were significantly decreased compared to preoperative ctDNA ( $p<$ 0.001). Mutations were not detected in 24 (64.5%) patients. The VAFs of the mutations were also significantly decreased ( $p<$ 0.001). However, the VAFs of the mutations detected in the 1 ${}^{\text{st}}$ postoperative ctDNA of two patients, H03 and H21, continuously increased in 2 ${}^{\text{nd}}$ and 3 ${}^{\text{rd}}$ postoperative ctDNA (Supplemental Table 3). Patient H03 had stage IV rectal adenocarcinoma cancer. The mutation TP53 p.E286G was detected in this patient’s tumor tissues and paired preoperative ctDNA. The TP53 p.E286G VAF was 8.95% in preoperative ctDNA, 11.32% in the 1 ${}^{\text{st}}$ postoperative ctDNA, and 71.4% at the last. A new mutation PIK3CA p.E542K was also detected 85 days after surgery, and its VAF increased from 12.66% to 19.96% at the last postoperative ctDNA. Patient H03 died 177 days after surgery (Fig. 4). This case had been reported previously [24]. Patient H21 had stage III rectal adenocarcinoma cancer. The VAF of ARID1A p.Q515X was detected in tumor tissues and paired preoperative ctDNA of patient H21 increased from 0.65% to 1.3%, and up to 3.8% at last. The VAF of TP53 p.V272M decreased from 1.1% in preoperative ctDNA to 0.4% in the 1 ${}^{\text{st}}$ postoperative ctDNA. However, it increased up to 5.81% in the 2 ${}^{\text{nd}}$ postoperative ctDNA and eventually reached 5.24% (Fig. 4). The tumor was found to recur by CT scan 317 days after surgery.

In this study, ctDNA was also used to predict recurrence in CRC. The 1 ${}^{\text{st}}$ postoperative ctDNA of nine patients (23.1%) was positive, three of which had CRC recurrence (33.3%). However, only one of the patients (1/30, 3.3%) with postoperative ctDNA-negative recurred (HR 10.767; 95% CI 1.1–103.8; $p=$ 0.040). There was no difference in recurrence between patients with clinical high-risk of recurrence and clinical low-risk of recurrence (HR 2.311; 95% CI 0.258–20.722; $p=$ 0.454). The sensitivity and specificity of the 1 ${}^{\text{st}}$ postoperative ctDNA status for predicting the risk of recurrence in CRC were 75% and 81.8%, respectively. The sensitivity and specificity of clinical risk factors were 75% and 45.5%, respectively. At least twice postoperative CEA levels were provided in 29 patients, 5 of which were detected positive ( $>$ 5 $\mu$ g/L). Two patient with positive CEA (2/5, 40.0%) and 2 of 24 (8.3%) patients with negative CEA recurred. The sensitivity and specificity of postoperative CEA for monitoring CRC recurrence were 40% and 88%, respectively. The sensitivity would increase to 100% if ctDNA and CEA biomarkers were combined, but the specificity would decrease to 76.0%.

4. Discussion

In this study, the clinical implications for CRC diagnosis was investigated with preoperative plasma ctDNA. The obtained data is consistent with previous studies [29]. The sensitivity of preoperative ctDNA (78.4%) is much higher than that of CEA (31.4%). In particular, in this study, ctDNA was also found to be effective for diagnosing early stage CRC. The sensitivity can reach up to 72.7%, while other biomarkers such as CEA and ECT2 are less sensitive [30, 31].

The recent technological improvement for detecting low-frequency mutations has improved ctDNA-based cancer diagnosis. The quantity of ctDNA likely depends on tumor load. For a tumor load of 30 g, only 0.15% of the cell-free circulating DNA would be released from the tumor [32]. Therefore, the sequencing depth was increased to about 10,000 $\times$ to improve the detection rate of ctDNA, accounting for 0.1% of cfDNA. Conseqently, the sensitivity of diagnosing stage I/II CRC using ctDNA could reach up to 72.7%, which is comparable to a previous study using the TEC-Seq method (targeted error correction sequencing) developed for the noninvasive detection of early-stage tumors [19].

As ctDNA is mostly released from the apoptosis and necrosis of tumor cells, it can provide more comprehensive information for the comparison of tumor tissues, and can be used to overcome problems introduced by tumor heterogeneity [33, 34]. For example, a missense mutation in the exon2 of NRAS (G37A) was detected in preoperative plasma ctDNA of patient H19, but it was not detected in paired tumor tissue.

The clinical value of preoperative ctDNA in CC and RC was found to differ. The concentration of preoperative plasma cfDNA in CC exceeded that of RC. Furthermore, the mutation-positive rate of preoperative ctDNA in CC was also higher than in RC (92.3% vs 45.5%, $p=$ 0.004), which was consistent with previous studies (93.6% vs 77%, $p=$ 0.01) [35, 36]. This indicates that DNA from CC tissues was more likely to be released into blood than DNA from RC tissues, which is caused by histological and biological differences between CC and RC. Vascular invasion was one possible impact factor, which was observed in twelve patients. However, this factor ( $p=$ 0.287) was not associated with preoperative ctDNA-positive or tumor location ( $p=$ 0.564). Compared to a tumor in the rectum, a tumor located in colon might be more likely to damage the barrier between tumor and blood vessels. Furthermore, the mechanisms of oncogenesis might also be different between CC and RC. A tumor located in the rectum might contain more nuclear $\beta$ -catenin, which exhibits additional functions independent of APC in the APC/ $\beta$ -catenin pathway [37]. Understanding the differences between CC and RC is conducive to obtaining more accurate information, which guides the clinical practice. Since the number of patients enrolled in this study is relatively small, it is expected that these results can be validated in future studies. Particularly, since the number of recurrent cases are too few, we can only make a very limited speculation that postoperative ctDNA can be a potential biomarker for predicting CRC recurrence, and more effort should be paid to evaluate its effectiveness.

Patients with increased VAFs of mutations detected in the 1 ${}^{\text{st}}$ postoperative ctDNA would tend to recur. The continued increase in VAF will be more predictable. This finding is consistent with previous studies. For example, Tie et al. reported that the postoperative ctDNA mutation status is associated with prognosis in patients with CRC [35, 36].

In this study, 85 genes mutated most frequently in CRC were analyzed via targeted capture sequencing. The sensitivity and specificity were 75% and 81.8%, respectively, which are superior to conventional clinical risk factors. Furthermore, the sensitivity and specificity can be further improved by using larger targeted capture panels.

5. Conclusion

This study suggests three potential values of ctDNA sequencing for diagnosing and monitoring of CRC. Firstly, mutations can be detected from preoperative ctDNA samples for even stage I/II CRC patients, suggesting that ctDNA can be a potentially promising biomarker for the diagnosis of CRC. Secondly, CC tumors may be more likely to release tumor DNA into blood than RC tumors, suggesting that different thresholds for CC and RC patients may be considered when evaluating sequencing results. Finally, the mutation statuses of postoperative ctDNA are significantly correlated with recurrence for CRC patients, suggesting that it may be beneficial to monitor CRC patients by sequencing their post-surgery ctDNA samples.

Footnotes

Acknowledgments

We thank the clinicians took part in the study design, sample collection and clinical information collection, meanwhile, the researchers and engineers from HaploX Biotechnology took part in the experiments and sequencing data analysis. The present study was funded by grants from Health and Family Planning Commission of Shenzhen Municipality (grant No. 201601016) for patient enrolment and sample collection, by Shenzhen Science and Technology Innovation Committee Technical Research Project (grant No. JSGG20180703164202084) for DNA sequencing, and by Shenzhen Strategic Emerging Industry Development Special Fund (grant No. 20170922151538732) for data analysis.

Conflict of interest

None.

Supplementary data

The supplementary files are available to download from http://dx.doi.org/10.3233/CBM-190257.

References

Arnold

Sierra

M.S.

Laversanne

Soerjomataram

Jemal

and Bray

, Global patterns and trends in colorectal cancer incidence and mortality, Gut 66 (2017), 683–691.

Young

P.E.

Womeldorph

C.M.

Johnson

E.K.

Maykel

J.A.

Brucher

Stojadinovic

Avital

Nissan

and Steele

S.R.

, Early detection of colorectal cancer recurrence in patients undergoing surgery with curative intent: current status and challenges, J Cancer 5 (2014), 262–271.

Van Cutsem

Cervantes

Adam

Sobrero

Van Krieken

J.H.

Aderka

Aguilar

E.A.

Bardelli

Benson

and Bodoky

, ESMO consensus guidelines for the management of patients with metastatic colorectal cancer, Ann. Oncol 27 (2016), 1386–1422.

Siegel

R.L.

Miller

K.D.

and Jemal

, Cancer statistics, 2018, Ca A Cancer J. Clin 68 (2018), 7–30.

Johnson

D.A.

Barkun

A.N.

Cohen

L.B.

Dominitz

J.A.

Kaltenbach

Martel

Robertson

D.J.

Boland

C.R.

Giardello

F.M.

and Lieberman

D.A.

, Optimizing adequacy of bowel cleansing for colonoscopy: recommendations from the U.S. Multi-Society Task Force on Colorectal Cancer, Gastroenterology 147 (2014), 903–924.

Tanaka

and Ishigamori

, Biomarkers for colorectal cancer, Int. J. Mol. Sci 11 (2010), 3209–3225.

Konishi

Shimada

Hsu

Tufts

Jimenez-Rodriguez

Cercek

Yaeger

Saltz

Smith

J.J.

and Nash

G.M.

, Association of preoperative and postoperative serum carcinoembryonic antigen and colon cancer outcome, Jama Oncol 4 (2017), E1–E6.

Chao

and Gibbs

, Caution is required before recommending routine carcinoembryonic antigen and imaging follow-up for patients with early-stage colon cancer, J. Clin. Oncol 27 (2009), e279–e280.

Quah

H.M.

Chou

J.F.

Gonen

Shia

Schrag

Landmann

R.G.

Guillem

J.G.

Paty

P.B.

Temple

L.K.

and Wong

W.D.

, Identification of patients with high-risk stage II colon cancer for adjuvant therapy, Dis. Colon Rectum 51 (2008), 503–507.

10.

Donna

Bertagnolli

M.M.

Warren

R.S.

Compton

C.C.

Kemeny

N.E.

Al Bowen

Gail

E.S.

Steven

Porjosh

G.N.

and Kerr

D.J.

, Documenting the natural history of patients with resected stage II adenocarcinoma of the colon after random assignment to adjuvant treatment with edrecolomab or observation: results from CALGB 9581, J. Clin. Oncol. Off. J. Am. Soc. Clin. Oncol 29 (2011), 3146–3152.

11.

Anya

Andrea

Neil

Diane

R.L.

Stadler

Z.K.

Yaeger

R.D.

Kemeny

N.E.

Weiser

M.R.

Pessin

M.S.

and Leonard

, False-positive elevations of carcinoembryonic antigen in patients with a history of resected colorectal cancer, J Natl Compr Canc Netw 12 (2014), 907–913.

12.

Schøler

L.V.

Reinert

Ørntoft

et al., Clinical implications of monitoring circulating tumor DNA in patients with colorectal cancer, Clin. Cancer Res 23 (2017), 5437.

13.

Evelyn

Kyra

Matthew

Valentina

Visser

B.C.

Andre

Bertram

Norton

J.A.

Mark

and Jeffrey

S.S.

, Mutation profiling of tumor DNA from plasma and tumor tissue of colorectal cancer patients with a novel, high-sensitivity multiplexed mutation detection platform, Oncotarget 6 (2015), 2549–2561.

14.

Beije

Helmijr

J.C.

Weerts

M.J.A.

Beaufort

C.M.

Wiggin

Marziali

Verhoef

Sleijfer

Jansen

M.P.H.M.

and Martens

J.W.M.

, Somatic mutation detection using various targeted detection assays in paired samples of circulating tumor DNA, primary tumor and metastases from patients undergoing resection of colorectal liver metastases, Mol. Oncol 10 (2016), 1575–1584.

15.

Siravegna

Marsoni

Siena

and Bardelli

, Integrating liquid biopsies into the management of cancer, Nat. Rev. Clin. Oncol 14 (2017), 531.

16.

Wan

J.C.M.

Massie

Garciacorbacho

Mouliere

Brenton

J.D.

Caldas

Pacey

Baird

and Rosenfeld

, Liquid biopsies come of age: towards implementation of circulating tumour DNA, Nat. Rev. Cancer 17 (2017), 223.

17.

Aravanis

A.M.

Lee

and Klausner

R.D.

, Next-generation sequencing of circulating tumor DNA for early cancer detection, Cell 168 (2017), 571–574.

18.

Chan

Jks

King

Bcy

Wkj

Chan

S.L.

Chu

Mak

Iol

and Sym

, Analysis of plasma epstein-barr virus DNA to screen for nasopharyngeal cancer, N. Engl. J. Med 377 (2017), 513.

19.

Phallen

Sausen

Adleff

Leal

Hruban

White

Anagnostou

Fiksel

Cristiano

and Papp

, Direct detection of early-stage cancers using circulating tumor DNA, Sci. Transl. Med 9 (2017), eaan2415.

20.

R.H.

Wei

Krawczyk

Wang

Luo

Flagg

Shi

Quan

and Li

, Circulating tumour DNA methylation markers for diagnosis and prognosis of hepatocellular carcinoma, Nat. Mater 16 (2017), 1155.

21.

Tie

Wang

Tomasetti

Springer

Kinde

Silliman

Tacey

Wong

H.L.

Christie

Kosmider

Skinner

Wong

Steel

Tran

Desai

Jones

Haydon

Hayes

Price

T.J.

Strausberg

R.L.

Diaz

L.A.

Jr Papadopoulos

Kinzler

K.W.

Vogelstein

and Gibbs

, Circulating tumor DNA analysis detects minimal residual disease and predicts recurrence in patients with stage II colon cancer, Sci. Transl. Med 11 (2017), 117–118.

22.

Zhou

Chang

Guan

Yang

Xia

Cui

and Lin

, Application of circulating tumor DNA as a non-invasive tool for monitoring the progression of colorectal cancer, PLoS One 11 (2016), e0159708.

23.

Reinert

Schøler

L.V.

Thomsen

Tobiasen

Vang

Nordentoft

Lamy

Kannerup

A.S.

Mortensen

F.V.

and Stribolt

, Analysis of circulating tumour DNA to monitor disease burden following colorectal cancer surgery, Gut 65 (2016), 625–634.

24.

Sato

K.A.

Tsuyoshi

Takeshi

Kohei

Teppei

Keisuke

Yukito

Risaburo

Takayuki

and Koki

, Individualized mutation detection in circulating tumor DNA for monitoring colorectal tumor burden using a cancer-associated gene sequencing panel, PLoS One 11 (2016), e0146275.

25.

Vidal

Muinelo

Dalmases

Jones

Edelstein

Iglesias

Orrillo

Abalo

Rodríguez

and Brozos

, Plasma ctDNA RAS mutation analysis for the diagnosis and treatment monitoring of metastatic colorectal cancer patients, Ann. Oncol. Off. J. Eur. Soc. Med. Oncol 28 (2017), 1325.

26.

J.Y.

Zhao

A.L.

and Gu

, Colorectal cancer or colon and rectal cancer? Clinicopathological comparison between colonic and rectal carcinomas, Oncology 73 (2007), 52–57.

27.

Sun

Huang

T.X.

Huang

Cheng

F.S.

Huang

K.B.

Liu

M.W.

Zhang

X.N.

M.Y.

Chen

S.F.

and Xia

L.G.

, Monitoring colorectal cancer following surgery using plasma circulating tumor DNA, Oncol. Lett 15 (2018), 4365–4375.

28.

Chen

S.F.

Zhou

Y.Q.

Chen

Y.R.

and Jia

, Fastp: an ultra-fast all-in-one FASTQ preprocessor, Bioinformatics 34 (2018), i884–i890.

29.

Tie

Wang

Kinde

Steel

and Gibbs

, Circulating tumor DNA (ctDNA) in nonmetastatic colorectal cancer (CRC): potential role as a screening tool, J. Clin. Oncol 33 (2010), 518.

30.

Chen

C.J.

Sung

W.W.

Chen

H.C.

Chern

Y.J.

Hsu

H.T.

Lin

Y.M.

Lin

S.H.

Peck

and Yeh

K.T.

, Early assessment of colorectal cancer by quantifying circulating tumor cells in peripheral blood: ECT2 in diagnosis of colorectal cancer, Int. J. Mol. Sci 18 (2017), 743.

31.

Chen

H.D.

Werner

Butt

Zörnig

Knebel

Michel

Eichmüller

S.B.

Jäger

Waterboer

and Pawlita

, Prospective evaluation of 64 serum autoantibodies as biomarkers for early detection of colorectal cancer in a true screening setting, Oncotarget 7 (2016), 16420–16432.

32.

Frank

Meng

Devin

Yiping

Dong

Steve

Diaz

L.A.

Goodman

S.N.

David

K.A.

and Hartmut

, Detection and quantification of mutations in the plasma of patients with colorectal tumors, Proc Natl Acad Sci U S A 102 (2005), 16368–16373.

33.

Baldus

S.E.

Karl-L

Rainer

Dinah

Stoecklein

N.H.

and Gabbert

H.E.

, Prevalence and heterogeneity of KRAS, BRAF, and PIK3CA mutations in primary colorectal adenocarcinomas and their corresponding metastases, Clin. Cancer Res 16 (2010), 790–799.

34.

Russo

Siravegna

Blaszkowsky

L.S.

Corti

Crisafulli

Ahronian

L.G.

Mussolin

Kwak

E.L.

Buscarino

and Lazzari

, Tumor heterogeneity and lesion-specific response to targeted therapy in colorectal cancer, Cancer Discov 6 (2016), 147–153.

35.

Tie

Kinde

Wang

Wong

H.L.

Roebert

Christie

Tacey

Wong

Singh

and Karapetis

C.S.

, Circulating tumor DNA as an early marker of therapeutic response in patients with metastatic colorectal cancer, Ann. Oncol 26 (2015), 1715–1722.

36.

Tie

Cohen

J.D.

Wang

Christie

Simons

Elsaleh

Kosmider

Wong

and Yip

, Serial circulating tumour DNA analysis during multimodality treatment of locally advanced rectal cancer: a prospective biomarker study, Gut (2018), gutjnl-2017-315852.

37.

Barry

D.S.

, Colorectal Cancer & You: A guide for people living with colorectal cancer, in: Colorectal Cancer Association of Canada, 2008.

Circulating tumor DNA sequencing for colorectal cancers: A comparative analysis of colon cancer and rectal cancer data

Abstract

BACKGROUND:

METHODS:

RESULTS:

CONCLUSION:

Keywords

1. Introduction

2. Materials and methods

2.1 Ethics statement and patient cohorts

2.2 DNA extraction and sequencing

Table 1 Clinic pathological characteristics and their correlation with preoperative ctDNA

3. Results

3.1 Patient characteristics

3.3 Preoperative plasma status and CRC diagnosis

3.5 Postoperative blood ctDNA for CRC recurrence prediction

4. Discussion

5. Conclusion

Footnotes

Acknowledgments

Conflict of interest

Supplementary data

References

Table 1
Clinic pathological characteristics and their correlation with preoperative ctDNA