Table 3.

Colorectal cancer.

Stage	Finding	Known Mut.	Method	Patient Number	Reference
/	Cell-free DNA (cfDNA) level cannot be used as a biomarker to distinguish subjects with premalignant from those without endoscopic lesions. cfDNA concentration can predict adenocarcinomas in FOBT positive patients	Yes	Mutant-enriched PCR qRT-PCR	179 healthy, FOBT positive	Perrone et al. (110)
/	Mutations in K-ras gene detected in plasma cfDNA are associated with risk of colorectal cancer (CRC)	Yes	PCR	240	Kopreski et al. (111)
n.s	No correlation between CEA and plasma cfDNA level. No association between plasma cfDNA level and age and gender of patient, location and size of tumor, histologic grading and Dukes’ stage. Changes in cfDNA level can be used for monitoring recurrence and prospectively to identify high-risk patients	No	DNA Dipstick Kit	70 20 healthy	Frattini et al. (112)
B,C,D	CRC patients have higher cfDNA level than healthy subjects at the time of surgery. cfDNA level can be used to confirm the presence of recurrence or metastasis. Correlation of CEA and cfDNA level was observed	Yes	Mutant-enriched PCR Fluorescent-methylation specific PCR	70 20 healthy	Frattini et al. (113)
I–IV	Circulating tumor DNA (ctDNA) analysis reveals disease recurrence earlier than conventional follow-up (lead time 10 months). ctDNA superior over CEA in monitoring CRC	Yes	NGS ddPCR	11	Reinert et al. (114)
A, B, C, D	Combined analysis of cfDNA and CEA has higher diagnostic capacity in CRC than each of markers alone	No	qRT-PCR	75 75 healthy	Flamini et al. (115)
I–IV	Mutations in plasma cfDNA with AF >0.1% showed clinical utility in monitoring tumor burden in CRC. Median plasma cfDNA level of healthy individuals, endoscopically resectable tumors and advanced CRCs are 4.2, 6.8, and 9.2 ng/ml, respectively	Yes	NGS ddPCR	44 9 healthy	Sato et al. (116)
II–IV	Changes in ctDNA level used to follow tumor dynamics	Yes	BEAMing qRT-PCR	18	Diehl et al. (48)
I–IV	No correlation between ctDNA and CEA. High preoperative ctDNA level correlate with poor prognosis, shorter progression-free survival (PFS) and overall survival (OS)	Yes	TEC-Seq	42 44 healthy	Phallen et al. (54)
IV	Plasma cfDNA level correlate with plasma mutant KRAS level. No difference in cfDNA level between KRAS and wt-positive disease. Concordance rate of 78% for KRAS mutation between primary and cfDNA	Yes	ARMS qPCR	108	Spindler et al. (117)
IV	High concordance rate between plasma cfDNA and tumor for BRAF, KRAS, and PIK3CA	Yes	BEAMing qRT-PCR	503	Tabernero et al. (118)
IV	Presence of KRAS mutation in plasma, but not in tumor is strong prognostic factor for PFS and OS. Positive correlation between cfDNA and LDH, but not with CEA	Yes	TheraScreen KRAS mutation kit	140	Spindler et al. (119)
II	Detection of ctDNA after resection of colon cancer indicates residual disease and identifies patients at high-risk for recurrence. Serial ctDNA level is more sensitive in predicting radiologic recurrence than CEA levels	Yes	Safe-Seq	230	Tie et al. (120)
IV	KRAS mutation detectable in plasma cfDNA 10 months before radiographic progression	Yes	iPLEX assay Exome sequencing BEAMing direct sequencing	18	Misale et al. (121)
IV	Re-challenge with EGFR-specific antibodies causes increase and decrease in percentage of mutant KRAS clones	Yes	HMRA Sanger sequencing Pyrosequencing BEAMing ddPCR NGS qRT-PCR	100	Siravegna et al. (122)
IV	Longer PFS in patients who experienced more than 10-fold change in ctDNA level after cycle 1 of chemotherapy. No correlation between OS and fold change in cDNA was observed	Yes	MPS	53	Tie et al. (123)
IV	Metastatic colorectal cancer (mCRC) patients with liver metastases and poorer performance had higher CTC count. 26% of mCRC patients at baseline had unfavorable (>3 CTCs/7.5 ml blood) CTC count. CTC number as prognostic and predictive factor at baseline and during follow-up	No	CellSearch	430	Cohen et al. (124)
I–IV	CTCs have been detected in all stages of CRC. Increased number of biphenotypic and mesenchymal CTCs in later stages of CRC. Presence of CTC with mesenchymal phenotype correlates with disease severity	No	CanPatrol	1,203	Zhao et al. (125)
I–IV	No clinicopathologic variables are associated with CTC detection in non-metastatic patients. Detection of the CTCs in non-metastatic patients associated with shorter OS and PFS. CTC count associated with the stage of disease. Preoperative detection of CTCs is strong prediction factor for disease progression and survival	No	CellSearch	287	Bork et al. (126)
n.s	Intra- and interpatient heterogeneity of CTCs with regard to EGFR amplification/expression and mutation profile in BRAF, KRAS, and PIK3CA gene	No	CellSearch qRT-PCR	49 mCRC 32 non-mCRC	Gasch et al. (127)
I–IV	90% of patients had at least 1 CTC/3 ml blood and 7% of those had CTC clusters. CTC number correlates to the disease stage, but not to CEA and CA19.9 markers. droplet digital PCR (ddPCR) is preferable technique to analyze KRAS mutation status in isolated CTCs	Yes	ScreenCell MB ddPCR TaqMeltPCR High-resolution melting Sanger sequencing MPS	35	Denis et al. (128)